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": 8.97461281833028,
+      "learning_rate": 3.2258064516129035e-07,
+      "loss": 1.1706,
+      "step": 1
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 9.755779813660224,
+      "learning_rate": 6.451612903225807e-07,
+      "loss": 1.2432,
+      "step": 2
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 9.35961225876907,
+      "learning_rate": 9.67741935483871e-07,
+      "loss": 1.2064,
+      "step": 3
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 9.541530822682375,
+      "learning_rate": 1.2903225806451614e-06,
+      "loss": 1.2315,
+      "step": 4
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 8.597425770350325,
+      "learning_rate": 1.6129032258064516e-06,
+      "loss": 1.2042,
+      "step": 5
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 8.257445749607397,
+      "learning_rate": 1.935483870967742e-06,
+      "loss": 1.1953,
+      "step": 6
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 5.674696683897444,
+      "learning_rate": 2.2580645161290324e-06,
+      "loss": 1.1315,
+      "step": 7
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 5.111622098903344,
+      "learning_rate": 2.580645161290323e-06,
+      "loss": 1.1211,
+      "step": 8
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 3.3716646839459865,
+      "learning_rate": 2.903225806451613e-06,
+      "loss": 1.1006,
+      "step": 9
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 3.200175573102256,
+      "learning_rate": 3.225806451612903e-06,
+      "loss": 1.0807,
+      "step": 10
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 2.988477967943152,
+      "learning_rate": 3.548387096774194e-06,
+      "loss": 1.0514,
+      "step": 11
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 3.8728195924732116,
+      "learning_rate": 3.870967741935484e-06,
+      "loss": 1.0337,
+      "step": 12
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 4.009313234209644,
+      "learning_rate": 4.193548387096774e-06,
+      "loss": 1.04,
+      "step": 13
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 3.7510890226627116,
+      "learning_rate": 4.516129032258065e-06,
+      "loss": 1.0209,
+      "step": 14
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 2.9901510667948945,
+      "learning_rate": 4.838709677419355e-06,
+      "loss": 1.0063,
+      "step": 15
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 2.510114404648719,
+      "learning_rate": 5.161290322580646e-06,
+      "loss": 0.9627,
+      "step": 16
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 2.7495137699415353,
+      "learning_rate": 5.483870967741935e-06,
+      "loss": 0.9285,
+      "step": 17
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 2.4404013320722338,
+      "learning_rate": 5.806451612903226e-06,
+      "loss": 0.9089,
+      "step": 18
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.8090807483456413,
+      "learning_rate": 6.129032258064517e-06,
+      "loss": 0.8975,
+      "step": 19
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.4196916011915979,
+      "learning_rate": 6.451612903225806e-06,
+      "loss": 0.8753,
+      "step": 20
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.297121977361323,
+      "learning_rate": 6.774193548387097e-06,
+      "loss": 0.8458,
+      "step": 21
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.4005691104902367,
+      "learning_rate": 7.096774193548388e-06,
+      "loss": 0.8696,
+      "step": 22
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.279055770268444,
+      "learning_rate": 7.4193548387096784e-06,
+      "loss": 0.8686,
+      "step": 23
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.1286272029710493,
+      "learning_rate": 7.741935483870968e-06,
+      "loss": 0.8357,
+      "step": 24
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 0.9647487202974877,
+      "learning_rate": 8.064516129032258e-06,
+      "loss": 0.8432,
+      "step": 25
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 0.9158938570938183,
+      "learning_rate": 8.387096774193549e-06,
+      "loss": 0.8314,
+      "step": 26
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 0.9741777128961578,
+      "learning_rate": 8.70967741935484e-06,
+      "loss": 0.8192,
+      "step": 27
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 0.9516738101183386,
+      "learning_rate": 9.03225806451613e-06,
+      "loss": 0.8272,
+      "step": 28
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 0.9053450324049076,
+      "learning_rate": 9.35483870967742e-06,
+      "loss": 0.8079,
+      "step": 29
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 0.7452116634263862,
+      "learning_rate": 9.67741935483871e-06,
+      "loss": 0.7968,
+      "step": 30
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 0.8714240728320455,
+      "learning_rate": 1e-05,
+      "loss": 0.8067,
+      "step": 31
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 0.9969347011448508,
+      "learning_rate": 1.0322580645161291e-05,
+      "loss": 0.8044,
+      "step": 32
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 0.8438117784031082,
+      "learning_rate": 1.0645161290322582e-05,
+      "loss": 0.7832,
+      "step": 33
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 0.8117310933171908,
+      "learning_rate": 1.096774193548387e-05,
+      "loss": 0.7781,
+      "step": 34
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 1.0992586514974958,
+      "learning_rate": 1.1290322580645164e-05,
+      "loss": 0.7725,
+      "step": 35
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 0.957133672543808,
+      "learning_rate": 1.1612903225806453e-05,
+      "loss": 0.7769,
+      "step": 36
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 1.0115453245499644,
+      "learning_rate": 1.1935483870967743e-05,
+      "loss": 0.7806,
+      "step": 37
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.807507183227898,
+      "learning_rate": 1.2258064516129034e-05,
+      "loss": 0.7631,
+      "step": 38
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 0.7571845084477026,
+      "learning_rate": 1.2580645161290324e-05,
+      "loss": 0.761,
+      "step": 39
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 0.8124725339751027,
+      "learning_rate": 1.2903225806451613e-05,
+      "loss": 0.7667,
+      "step": 40
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 0.7240369916077818,
+      "learning_rate": 1.3225806451612903e-05,
+      "loss": 0.7604,
+      "step": 41
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 0.9623740382538682,
+      "learning_rate": 1.3548387096774194e-05,
+      "loss": 0.7626,
+      "step": 42
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 0.7442385310069625,
+      "learning_rate": 1.3870967741935486e-05,
+      "loss": 0.7554,
+      "step": 43
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 0.9007045097754852,
+      "learning_rate": 1.4193548387096776e-05,
+      "loss": 0.7416,
+      "step": 44
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 0.8321089648405688,
+      "learning_rate": 1.4516129032258066e-05,
+      "loss": 0.7485,
+      "step": 45
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 0.9184543857452137,
+      "learning_rate": 1.4838709677419357e-05,
+      "loss": 0.74,
+      "step": 46
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 0.707782024529352,
+      "learning_rate": 1.5161290322580646e-05,
+      "loss": 0.7539,
+      "step": 47
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 0.8049529238198098,
+      "learning_rate": 1.5483870967741936e-05,
+      "loss": 0.7353,
+      "step": 48
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 0.7085241525830759,
+      "learning_rate": 1.5806451612903226e-05,
+      "loss": 0.7545,
+      "step": 49
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 0.8450481574184178,
+      "learning_rate": 1.6129032258064517e-05,
+      "loss": 0.7443,
+      "step": 50
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 0.8381941501655076,
+      "learning_rate": 1.6451612903225807e-05,
+      "loss": 0.737,
+      "step": 51
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 0.8229048452411964,
+      "learning_rate": 1.6774193548387098e-05,
+      "loss": 0.744,
+      "step": 52
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 0.8146165431213764,
+      "learning_rate": 1.7096774193548388e-05,
+      "loss": 0.7422,
+      "step": 53
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 1.0732298831956086,
+      "learning_rate": 1.741935483870968e-05,
+      "loss": 0.7167,
+      "step": 54
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 0.69285965205896,
+      "learning_rate": 1.774193548387097e-05,
+      "loss": 0.7255,
+      "step": 55
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 1.0160743477301444,
+      "learning_rate": 1.806451612903226e-05,
+      "loss": 0.7436,
+      "step": 56
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.9512104944037822,
+      "learning_rate": 1.838709677419355e-05,
+      "loss": 0.7276,
+      "step": 57
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 0.7835167660691672,
+      "learning_rate": 1.870967741935484e-05,
+      "loss": 0.7335,
+      "step": 58
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 0.9578749516364524,
+      "learning_rate": 1.903225806451613e-05,
+      "loss": 0.7192,
+      "step": 59
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 0.8171847606925552,
+      "learning_rate": 1.935483870967742e-05,
+      "loss": 0.7263,
+      "step": 60
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 0.8576762404193319,
+      "learning_rate": 1.967741935483871e-05,
+      "loss": 0.7253,
+      "step": 61
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 0.7127835127423651,
+      "learning_rate": 2e-05,
+      "loss": 0.7133,
+      "step": 62
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 0.9803409293439198,
+      "learning_rate": 2.0322580645161292e-05,
+      "loss": 0.727,
+      "step": 63
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 0.8304627627024237,
+      "learning_rate": 2.0645161290322582e-05,
+      "loss": 0.73,
+      "step": 64
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.0698477803481856,
+      "learning_rate": 2.0967741935483873e-05,
+      "loss": 0.7157,
+      "step": 65
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 1.0344761359957897,
+      "learning_rate": 2.1290322580645163e-05,
+      "loss": 0.7229,
+      "step": 66
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 0.8407311301709058,
+      "learning_rate": 2.161290322580645e-05,
+      "loss": 0.7341,
+      "step": 67
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 0.9603925415283671,
+      "learning_rate": 2.193548387096774e-05,
+      "loss": 0.7222,
+      "step": 68
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.0498729615175724,
+      "learning_rate": 2.225806451612903e-05,
+      "loss": 0.7164,
+      "step": 69
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 0.8391733181836435,
+      "learning_rate": 2.2580645161290328e-05,
+      "loss": 0.7354,
+      "step": 70
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.005649694209137,
+      "learning_rate": 2.2903225806451615e-05,
+      "loss": 0.7226,
+      "step": 71
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.5470360218427381,
+      "learning_rate": 2.3225806451612906e-05,
+      "loss": 0.7084,
+      "step": 72
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 0.9114490100662501,
+      "learning_rate": 2.3548387096774196e-05,
+      "loss": 0.713,
+      "step": 73
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.8244138394073541,
+      "learning_rate": 2.3870967741935486e-05,
+      "loss": 0.7152,
+      "step": 74
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.2199210801437352,
+      "learning_rate": 2.4193548387096777e-05,
+      "loss": 0.6991,
+      "step": 75
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.8837854886462218,
+      "learning_rate": 2.4516129032258067e-05,
+      "loss": 0.7278,
+      "step": 76
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.443935126686266,
+      "learning_rate": 2.4838709677419358e-05,
+      "loss": 0.7145,
+      "step": 77
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.5269840720466061,
+      "learning_rate": 2.5161290322580648e-05,
+      "loss": 0.7066,
+      "step": 78
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.4741084962177176,
+      "learning_rate": 2.548387096774194e-05,
+      "loss": 0.7182,
+      "step": 79
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.2905569618713717,
+      "learning_rate": 2.5806451612903226e-05,
+      "loss": 0.713,
+      "step": 80
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.3107747835066688,
+      "learning_rate": 2.6129032258064516e-05,
+      "loss": 0.7214,
+      "step": 81
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.0954597336935217,
+      "learning_rate": 2.6451612903225806e-05,
+      "loss": 0.712,
+      "step": 82
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.1288092867345743,
+      "learning_rate": 2.6774193548387097e-05,
+      "loss": 0.712,
+      "step": 83
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.0804836164080145,
+      "learning_rate": 2.7096774193548387e-05,
+      "loss": 0.7084,
+      "step": 84
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.1259335556050527,
+      "learning_rate": 2.7419354838709678e-05,
+      "loss": 0.7148,
+      "step": 85
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.0227476081622406,
+      "learning_rate": 2.774193548387097e-05,
+      "loss": 0.7134,
+      "step": 86
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 0.9320989616893572,
+      "learning_rate": 2.8064516129032262e-05,
+      "loss": 0.7054,
+      "step": 87
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.071571934061012,
+      "learning_rate": 2.8387096774193552e-05,
+      "loss": 0.695,
+      "step": 88
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.0295599533058397,
+      "learning_rate": 2.8709677419354843e-05,
+      "loss": 0.7093,
+      "step": 89
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.3173460273520448,
+      "learning_rate": 2.9032258064516133e-05,
+      "loss": 0.7074,
+      "step": 90
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.2585954400536044,
+      "learning_rate": 2.9354838709677423e-05,
+      "loss": 0.7088,
+      "step": 91
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 0.8818604538507505,
+      "learning_rate": 2.9677419354838714e-05,
+      "loss": 0.6976,
+      "step": 92
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.1977853343326466,
+      "learning_rate": 3.0000000000000004e-05,
+      "loss": 0.6977,
+      "step": 93
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.1431500226494722,
+      "learning_rate": 3.032258064516129e-05,
+      "loss": 0.7062,
+      "step": 94
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.1307668243250135,
+      "learning_rate": 3.0645161290322585e-05,
+      "loss": 0.7111,
+      "step": 95
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.0615423223239975,
+      "learning_rate": 3.096774193548387e-05,
+      "loss": 0.6978,
+      "step": 96
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.412008352013034,
+      "learning_rate": 3.1290322580645166e-05,
+      "loss": 0.698,
+      "step": 97
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.1983415978717316,
+      "learning_rate": 3.161290322580645e-05,
+      "loss": 0.6986,
+      "step": 98
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.038388678907693,
+      "learning_rate": 3.193548387096774e-05,
+      "loss": 0.7026,
+      "step": 99
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 2.133733102954517,
+      "learning_rate": 3.2258064516129034e-05,
+      "loss": 0.7001,
+      "step": 100
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.4647411173603935,
+      "learning_rate": 3.258064516129033e-05,
+      "loss": 0.6979,
+      "step": 101
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 2.183173830496467,
+      "learning_rate": 3.2903225806451614e-05,
+      "loss": 0.7145,
+      "step": 102
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 2.101154049987277,
+      "learning_rate": 3.322580645161291e-05,
+      "loss": 0.7214,
+      "step": 103
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.588783647797691,
+      "learning_rate": 3.3548387096774195e-05,
+      "loss": 0.703,
+      "step": 104
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.688715300224427,
+      "learning_rate": 3.387096774193549e-05,
+      "loss": 0.6984,
+      "step": 105
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.260085281171982,
+      "learning_rate": 3.4193548387096776e-05,
+      "loss": 0.7026,
+      "step": 106
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.230785932524029,
+      "learning_rate": 3.451612903225807e-05,
+      "loss": 0.7029,
+      "step": 107
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.289987440051931,
+      "learning_rate": 3.483870967741936e-05,
+      "loss": 0.6941,
+      "step": 108
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.0515210639779737,
+      "learning_rate": 3.516129032258065e-05,
+      "loss": 0.708,
+      "step": 109
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.3023576731183046,
+      "learning_rate": 3.548387096774194e-05,
+      "loss": 0.7062,
+      "step": 110
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 0.9985629681400807,
+      "learning_rate": 3.5806451612903225e-05,
+      "loss": 0.6983,
+      "step": 111
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.1672598526176874,
+      "learning_rate": 3.612903225806452e-05,
+      "loss": 0.7072,
+      "step": 112
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.152896588408172,
+      "learning_rate": 3.6451612903225805e-05,
+      "loss": 0.7064,
+      "step": 113
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.9259236058735657,
+      "learning_rate": 3.67741935483871e-05,
+      "loss": 0.6973,
+      "step": 114
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.1066117680735188,
+      "learning_rate": 3.7096774193548386e-05,
+      "loss": 0.688,
+      "step": 115
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 0.9413326013879276,
+      "learning_rate": 3.741935483870968e-05,
+      "loss": 0.6928,
+      "step": 116
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.0169844152269307,
+      "learning_rate": 3.7741935483870974e-05,
+      "loss": 0.6835,
+      "step": 117
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 0.9574581464963522,
+      "learning_rate": 3.806451612903226e-05,
+      "loss": 0.7019,
+      "step": 118
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 0.9838915296646057,
+      "learning_rate": 3.8387096774193555e-05,
+      "loss": 0.6946,
+      "step": 119
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 0.9863005840229495,
+      "learning_rate": 3.870967741935484e-05,
+      "loss": 0.6845,
+      "step": 120
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 0.9856495020777085,
+      "learning_rate": 3.9032258064516135e-05,
+      "loss": 0.7065,
+      "step": 121
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.3642059842170808,
+      "learning_rate": 3.935483870967742e-05,
+      "loss": 0.7012,
+      "step": 122
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.0406983230783544,
+      "learning_rate": 3.9677419354838716e-05,
+      "loss": 0.7036,
+      "step": 123
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 1.3786914773516044,
+      "learning_rate": 4e-05,
+      "loss": 0.695,
+      "step": 124
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 0.8679368168488655,
+      "learning_rate": 3.9999920040266474e-05,
+      "loss": 0.6912,
+      "step": 125
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.3188205696142914,
+      "learning_rate": 3.999968016170525e-05,
+      "loss": 0.7059,
+      "step": 126
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 0.8974718531403549,
+      "learning_rate": 3.9999280366234394e-05,
+      "loss": 0.7181,
+      "step": 127
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.052462176878189,
+      "learning_rate": 3.999872065705065e-05,
+      "loss": 0.7,
+      "step": 128
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.2419830814111212,
+      "learning_rate": 3.999800103862945e-05,
+      "loss": 0.6993,
+      "step": 129
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.0229647242066775,
+      "learning_rate": 3.9997121516724834e-05,
+      "loss": 0.6907,
+      "step": 130
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.0567147482927461,
+      "learning_rate": 3.999608209836944e-05,
+      "loss": 0.7109,
+      "step": 131
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.16651207730761,
+      "learning_rate": 3.999488279187443e-05,
+      "loss": 0.706,
+      "step": 132
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.106194964952565,
+      "learning_rate": 3.9993523606829425e-05,
+      "loss": 0.706,
+      "step": 133
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 0.9519550104224455,
+      "learning_rate": 3.9992004554102435e-05,
+      "loss": 0.7123,
+      "step": 134
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.4320834822742146,
+      "learning_rate": 3.999032564583976e-05,
+      "loss": 0.7027,
+      "step": 135
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 0.7897764685974833,
+      "learning_rate": 3.998848689546591e-05,
+      "loss": 0.7028,
+      "step": 136
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.1252580480911993,
+      "learning_rate": 3.9986488317683496e-05,
+      "loss": 0.6899,
+      "step": 137
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 0.9517428905046706,
+      "learning_rate": 3.9984329928473067e-05,
+      "loss": 0.6949,
+      "step": 138
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.12202342024344,
+      "learning_rate": 3.998201174509306e-05,
+      "loss": 0.697,
+      "step": 139
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.005720817363771,
+      "learning_rate": 3.997953378607961e-05,
+      "loss": 0.6959,
+      "step": 140
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.0438192845119851,
+      "learning_rate": 3.9976896071246403e-05,
+      "loss": 0.6963,
+      "step": 141
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 0.9708420641102506,
+      "learning_rate": 3.997409862168454e-05,
+      "loss": 0.7037,
+      "step": 142
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 0.9578376838066693,
+      "learning_rate": 3.997114145976237e-05,
+      "loss": 0.6826,
+      "step": 143
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.244100816107012,
+      "learning_rate": 3.9968024609125245e-05,
+      "loss": 0.7007,
+      "step": 144
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 0.8603685701122628,
+      "learning_rate": 3.996474809469545e-05,
+      "loss": 0.6834,
+      "step": 145
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.0568814710755285,
+      "learning_rate": 3.996131194267189e-05,
+      "loss": 0.6877,
+      "step": 146
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 0.7685882212109898,
+      "learning_rate": 3.995771618052995e-05,
+      "loss": 0.6812,
+      "step": 147
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 0.9011342884481083,
+      "learning_rate": 3.9953960837021256e-05,
+      "loss": 0.6904,
+      "step": 148
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 0.8565291327440087,
+      "learning_rate": 3.995004594217343e-05,
+      "loss": 0.6861,
+      "step": 149
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 0.8927841138567424,
+      "learning_rate": 3.994597152728987e-05,
+      "loss": 0.6923,
+      "step": 150
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.0846365997278467,
+      "learning_rate": 3.994173762494947e-05,
+      "loss": 0.6888,
+      "step": 151
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.7235924961335207,
+      "learning_rate": 3.9937344269006424e-05,
+      "loss": 0.6961,
+      "step": 152
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.0447148182664239,
+      "learning_rate": 3.993279149458988e-05,
+      "loss": 0.6741,
+      "step": 153
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 0.7643080440664354,
+      "learning_rate": 3.9928079338103695e-05,
+      "loss": 0.6923,
+      "step": 154
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 0.8443853419019671,
+      "learning_rate": 3.992320783722616e-05,
+      "loss": 0.6914,
+      "step": 155
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 0.9007637974352969,
+      "learning_rate": 3.991817703090965e-05,
+      "loss": 0.6853,
+      "step": 156
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 0.9877305983251269,
+      "learning_rate": 3.991298695938038e-05,
+      "loss": 0.6833,
+      "step": 157
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 0.7032264975314625,
+      "learning_rate": 3.9907637664138e-05,
+      "loss": 0.6764,
+      "step": 158
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 0.9447612040177417,
+      "learning_rate": 3.990212918795534e-05,
+      "loss": 0.6824,
+      "step": 159
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 1.0401155675978846,
+      "learning_rate": 3.9896461574878036e-05,
+      "loss": 0.6804,
+      "step": 160
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.0159398838662401,
+      "learning_rate": 3.989063487022416e-05,
+      "loss": 0.6884,
+      "step": 161
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 0.7121417949530757,
+      "learning_rate": 3.9884649120583903e-05,
+      "loss": 0.675,
+      "step": 162
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 0.7724339986503345,
+      "learning_rate": 3.987850437381915e-05,
+      "loss": 0.6745,
+      "step": 163
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 0.6767034798093763,
+      "learning_rate": 3.987220067906313e-05,
+      "loss": 0.6876,
+      "step": 164
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 0.7806540717198109,
+      "learning_rate": 3.986573808672002e-05,
+      "loss": 0.687,
+      "step": 165
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 0.6639006341638573,
+      "learning_rate": 3.9859116648464534e-05,
+      "loss": 0.6762,
+      "step": 166
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 0.732256702360303,
+      "learning_rate": 3.985233641724152e-05,
+      "loss": 0.6836,
+      "step": 167
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 0.7576680958729621,
+      "learning_rate": 3.984539744726554e-05,
+      "loss": 0.6773,
+      "step": 168
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 0.7605697779644685,
+      "learning_rate": 3.983829979402038e-05,
+      "loss": 0.684,
+      "step": 169
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 0.7104136417042677,
+      "learning_rate": 3.983104351425871e-05,
+      "loss": 0.6782,
+      "step": 170
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.805404193663185,
+      "learning_rate": 3.982362866600154e-05,
+      "loss": 0.6807,
+      "step": 171
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 0.7984877706504858,
+      "learning_rate": 3.9816055308537806e-05,
+      "loss": 0.6804,
+      "step": 172
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 0.8389515953675403,
+      "learning_rate": 3.980832350242388e-05,
+      "loss": 0.6646,
+      "step": 173
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 0.9152913957214306,
+      "learning_rate": 3.980043330948306e-05,
+      "loss": 0.6735,
+      "step": 174
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 0.7991563605829256,
+      "learning_rate": 3.9792384792805124e-05,
+      "loss": 0.6794,
+      "step": 175
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 0.8084461427241927,
+      "learning_rate": 3.978417801674579e-05,
+      "loss": 0.6907,
+      "step": 176
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 0.6351366346898705,
+      "learning_rate": 3.9775813046926244e-05,
+      "loss": 0.6758,
+      "step": 177
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 0.8749044379535599,
+      "learning_rate": 3.976728995023254e-05,
+      "loss": 0.6819,
+      "step": 178
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 0.9137942254769006,
+      "learning_rate": 3.975860879481514e-05,
+      "loss": 0.6789,
+      "step": 179
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 0.72331767796892,
+      "learning_rate": 3.974976965008833e-05,
+      "loss": 0.6846,
+      "step": 180
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 0.691017540114063,
+      "learning_rate": 3.9740772586729674e-05,
+      "loss": 0.6714,
+      "step": 181
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 0.8669303780934151,
+      "learning_rate": 3.973161767667946e-05,
+      "loss": 0.6833,
+      "step": 182
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 0.8646330432172081,
+      "learning_rate": 3.972230499314009e-05,
+      "loss": 0.6719,
+      "step": 183
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 0.6523740107691184,
+      "learning_rate": 3.9712834610575545e-05,
+      "loss": 0.6739,
+      "step": 184
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 0.6430736651702299,
+      "learning_rate": 3.9703206604710746e-05,
+      "loss": 0.6758,
+      "step": 185
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 0.8628809846749322,
+      "learning_rate": 3.969342105253098e-05,
+      "loss": 0.6715,
+      "step": 186
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 0.7814922404045013,
+      "learning_rate": 3.9683478032281246e-05,
+      "loss": 0.6874,
+      "step": 187
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 0.8188345721797253,
+      "learning_rate": 3.967337762346568e-05,
+      "loss": 0.6794,
+      "step": 188
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 0.8123288789342519,
+      "learning_rate": 3.9663119906846885e-05,
+      "loss": 0.6909,
+      "step": 189
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.7986416469561298,
+      "learning_rate": 3.965270496444528e-05,
+      "loss": 0.6833,
+      "step": 190
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 0.9150891636761573,
+      "learning_rate": 3.9642132879538465e-05,
+      "loss": 0.6712,
+      "step": 191
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 0.7677420605592837,
+      "learning_rate": 3.9631403736660566e-05,
+      "loss": 0.688,
+      "step": 192
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 0.7294852002923622,
+      "learning_rate": 3.962051762160151e-05,
+      "loss": 0.6599,
+      "step": 193
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 0.8476014489128755,
+      "learning_rate": 3.960947462140638e-05,
+      "loss": 0.6675,
+      "step": 194
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 0.8275922714720053,
+      "learning_rate": 3.959827482437472e-05,
+      "loss": 0.6707,
+      "step": 195
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 0.7616602904989631,
+      "learning_rate": 3.958691832005981e-05,
+      "loss": 0.6647,
+      "step": 196
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 0.7113175677159013,
+      "learning_rate": 3.9575405199267944e-05,
+      "loss": 0.6772,
+      "step": 197
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 0.8337717237681868,
+      "learning_rate": 3.956373555405774e-05,
+      "loss": 0.6666,
+      "step": 198
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 0.6577857364571946,
+      "learning_rate": 3.955190947773937e-05,
+      "loss": 0.6616,
+      "step": 199
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 0.6301854623033886,
+      "learning_rate": 3.9539927064873824e-05,
+      "loss": 0.6647,
+      "step": 200
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 0.5766203497830974,
+      "learning_rate": 3.952778841127214e-05,
+      "loss": 0.6684,
+      "step": 201
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 0.6640001585072334,
+      "learning_rate": 3.95154936139947e-05,
+      "loss": 0.6766,
+      "step": 202
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 0.7365990692597033,
+      "learning_rate": 3.950304277135034e-05,
+      "loss": 0.6725,
+      "step": 203
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 0.6644702849056594,
+      "learning_rate": 3.94904359828957e-05,
+      "loss": 0.6696,
+      "step": 204
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 0.6500927249444098,
+      "learning_rate": 3.94776733494343e-05,
+      "loss": 0.6772,
+      "step": 205
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 0.5856163493314395,
+      "learning_rate": 3.9464754973015826e-05,
+      "loss": 0.6689,
+      "step": 206
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 0.5686685036570529,
+      "learning_rate": 3.9451680956935275e-05,
+      "loss": 0.6693,
+      "step": 207
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 0.6659310141974335,
+      "learning_rate": 3.943845140573213e-05,
+      "loss": 0.6774,
+      "step": 208
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.562719922102353,
+      "learning_rate": 3.942506642518952e-05,
+      "loss": 0.6648,
+      "step": 209
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 0.5897298952299841,
+      "learning_rate": 3.9411526122333396e-05,
+      "loss": 0.6663,
+      "step": 210
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 0.6091776530192319,
+      "learning_rate": 3.939783060543167e-05,
+      "loss": 0.6667,
+      "step": 211
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 0.5927037161469282,
+      "learning_rate": 3.938397998399333e-05,
+      "loss": 0.6724,
+      "step": 212
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 0.6303746794675272,
+      "learning_rate": 3.936997436876756e-05,
+      "loss": 0.6696,
+      "step": 213
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 0.7117321875852997,
+      "learning_rate": 3.935581387174289e-05,
+      "loss": 0.6778,
+      "step": 214
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 0.562789774408806,
+      "learning_rate": 3.9341498606146295e-05,
+      "loss": 0.6613,
+      "step": 215
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 0.8013850793665591,
+      "learning_rate": 3.932702868644224e-05,
+      "loss": 0.6745,
+      "step": 216
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 0.6089129063203483,
+      "learning_rate": 3.931240422833181e-05,
+      "loss": 0.664,
+      "step": 217
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 0.7183286982933746,
+      "learning_rate": 3.9297625348751806e-05,
+      "loss": 0.6688,
+      "step": 218
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 0.6190571197753483,
+      "learning_rate": 3.9282692165873744e-05,
+      "loss": 0.6721,
+      "step": 219
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.7005912345068807,
+      "learning_rate": 3.9267604799102945e-05,
+      "loss": 0.6834,
+      "step": 220
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 0.6159456830693482,
+      "learning_rate": 3.9252363369077615e-05,
+      "loss": 0.6641,
+      "step": 221
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 0.6557539194436985,
+      "learning_rate": 3.9236967997667796e-05,
+      "loss": 0.6805,
+      "step": 222
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 0.6781366711945995,
+      "learning_rate": 3.9221418807974496e-05,
+      "loss": 0.6665,
+      "step": 223
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 1.0198625087066957,
+      "learning_rate": 3.92057159243286e-05,
+      "loss": 0.6496,
+      "step": 224
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 0.7302225220455655,
+      "learning_rate": 3.9189859472289956e-05,
+      "loss": 0.6673,
+      "step": 225
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 0.9077782657376843,
+      "learning_rate": 3.9173849578646324e-05,
+      "loss": 0.6769,
+      "step": 226
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 1.0789099729627538,
+      "learning_rate": 3.9157686371412395e-05,
+      "loss": 0.682,
+      "step": 227
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.8200396105520883,
+      "learning_rate": 3.914136997982874e-05,
+      "loss": 0.6745,
+      "step": 228
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 0.7197055779122763,
+      "learning_rate": 3.912490053436079e-05,
+      "loss": 0.67,
+      "step": 229
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 0.8302470432649802,
+      "learning_rate": 3.9108278166697805e-05,
+      "loss": 0.6727,
+      "step": 230
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 0.7702887604063612,
+      "learning_rate": 3.909150300975178e-05,
+      "loss": 0.6671,
+      "step": 231
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 0.6847236786586546,
+      "learning_rate": 3.907457519765642e-05,
+      "loss": 0.6584,
+      "step": 232
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 0.8486623354389392,
+      "learning_rate": 3.905749486576607e-05,
+      "loss": 0.6719,
+      "step": 233
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 0.8175738933939216,
+      "learning_rate": 3.90402621506546e-05,
+      "loss": 0.6717,
+      "step": 234
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 0.621608174621054,
+      "learning_rate": 3.902287719011435e-05,
+      "loss": 0.6673,
+      "step": 235
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 0.6042307047619724,
+      "learning_rate": 3.900534012315499e-05,
+      "loss": 0.6686,
+      "step": 236
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 0.7271895084082575,
+      "learning_rate": 3.8987651090002454e-05,
+      "loss": 0.6712,
+      "step": 237
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 0.5342166342211299,
+      "learning_rate": 3.896981023209777e-05,
+      "loss": 0.671,
+      "step": 238
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 0.6712848507302626,
+      "learning_rate": 3.8951817692095966e-05,
+      "loss": 0.6582,
+      "step": 239
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 0.597510464598331,
+      "learning_rate": 3.893367361386491e-05,
+      "loss": 0.6506,
+      "step": 240
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 0.5487423197334942,
+      "learning_rate": 3.891537814248417e-05,
+      "loss": 0.6723,
+      "step": 241
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 0.634013180605427,
+      "learning_rate": 3.889693142424384e-05,
+      "loss": 0.6653,
+      "step": 242
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 0.712009817128921,
+      "learning_rate": 3.8878333606643405e-05,
+      "loss": 0.6606,
+      "step": 243
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 0.7337195409097608,
+      "learning_rate": 3.885958483839049e-05,
+      "loss": 0.6711,
+      "step": 244
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 0.7078751142621397,
+      "learning_rate": 3.8840685269399786e-05,
+      "loss": 0.6572,
+      "step": 245
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 0.672404199589782,
+      "learning_rate": 3.882163505079171e-05,
+      "loss": 0.6616,
+      "step": 246
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7833018759447947,
+      "learning_rate": 3.880243433489132e-05,
+      "loss": 0.6748,
+      "step": 247
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 0.6533976283663586,
+      "learning_rate": 3.878308327522702e-05,
+      "loss": 0.5987,
+      "step": 248
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 0.7522328850745098,
+      "learning_rate": 3.876358202652936e-05,
+      "loss": 0.583,
+      "step": 249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 0.8190032851894639,
+      "learning_rate": 3.8743930744729835e-05,
+      "loss": 0.5824,
+      "step": 250
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 0.8252357129991155,
+      "learning_rate": 3.8724129586959535e-05,
+      "loss": 0.5913,
+      "step": 251
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 0.7627644549805407,
+      "learning_rate": 3.870417871154801e-05,
+      "loss": 0.5827,
+      "step": 252
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 0.7187175596104006,
+      "learning_rate": 3.868407827802192e-05,
+      "loss": 0.576,
+      "step": 253
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 0.7465461209009953,
+      "learning_rate": 3.8663828447103804e-05,
+      "loss": 0.5953,
+      "step": 254
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 0.8647231946821289,
+      "learning_rate": 3.8643429380710765e-05,
+      "loss": 0.5729,
+      "step": 255
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 0.8943049522847307,
+      "learning_rate": 3.862288124195319e-05,
+      "loss": 0.5784,
+      "step": 256
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 0.7771953417189341,
+      "learning_rate": 3.8602184195133464e-05,
+      "loss": 0.5795,
+      "step": 257
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 0.8188468932530452,
+      "learning_rate": 3.8581338405744604e-05,
+      "loss": 0.5672,
+      "step": 258
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 0.7504146388501137,
+      "learning_rate": 3.8560344040468996e-05,
+      "loss": 0.5847,
+      "step": 259
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 0.6503710363199894,
+      "learning_rate": 3.853920126717703e-05,
+      "loss": 0.5738,
+      "step": 260
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 0.7690414884115693,
+      "learning_rate": 3.851791025492574e-05,
+      "loss": 0.5902,
+      "step": 261
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 0.946516126229426,
+      "learning_rate": 3.8496471173957515e-05,
+      "loss": 0.5727,
+      "step": 262
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 0.6792246347429378,
+      "learning_rate": 3.847488419569866e-05,
+      "loss": 0.5756,
+      "step": 263
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 0.704019072643891,
+      "learning_rate": 3.8453149492758077e-05,
+      "loss": 0.5801,
+      "step": 264
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 0.6267019157809601,
+      "learning_rate": 3.8431267238925886e-05,
+      "loss": 0.5686,
+      "step": 265
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.6952982383008113,
+      "learning_rate": 3.840923760917199e-05,
+      "loss": 0.5762,
+      "step": 266
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 0.6605910666447081,
+      "learning_rate": 3.838706077964473e-05,
+      "loss": 0.5653,
+      "step": 267
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 0.701814264681331,
+      "learning_rate": 3.836473692766944e-05,
+      "loss": 0.5739,
+      "step": 268
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 0.6393552999819505,
+      "learning_rate": 3.834226623174705e-05,
+      "loss": 0.568,
+      "step": 269
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 0.7104954759651017,
+      "learning_rate": 3.831964887155264e-05,
+      "loss": 0.5757,
+      "step": 270
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 0.7066056887934199,
+      "learning_rate": 3.829688502793403e-05,
+      "loss": 0.5685,
+      "step": 271
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 0.7266855835555316,
+      "learning_rate": 3.827397488291029e-05,
+      "loss": 0.5811,
+      "step": 272
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 0.6740861482548771,
+      "learning_rate": 3.8250918619670343e-05,
+      "loss": 0.5743,
+      "step": 273
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 0.6373854464407009,
+      "learning_rate": 3.822771642257145e-05,
+      "loss": 0.5874,
+      "step": 274
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 0.6811572675469807,
+      "learning_rate": 3.820436847713776e-05,
+      "loss": 0.5729,
+      "step": 275
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 0.5781780175229592,
+      "learning_rate": 3.8180874970058826e-05,
+      "loss": 0.5581,
+      "step": 276
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 0.6413522935795214,
+      "learning_rate": 3.8157236089188103e-05,
+      "loss": 0.5756,
+      "step": 277
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 0.6538410320045799,
+      "learning_rate": 3.813345202354145e-05,
+      "loss": 0.5722,
+      "step": 278
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 0.5289942500442774,
+      "learning_rate": 3.810952296329563e-05,
+      "loss": 0.584,
+      "step": 279
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 0.7762788553008584,
+      "learning_rate": 3.808544909978676e-05,
+      "loss": 0.5757,
+      "step": 280
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 0.5173429524888455,
+      "learning_rate": 3.806123062550882e-05,
+      "loss": 0.583,
+      "step": 281
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 0.7235180411680129,
+      "learning_rate": 3.803686773411208e-05,
+      "loss": 0.566,
+      "step": 282
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 0.6238523562068127,
+      "learning_rate": 3.801236062040158e-05,
+      "loss": 0.5789,
+      "step": 283
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 0.7445441970321809,
+      "learning_rate": 3.798770948033553e-05,
+      "loss": 0.5799,
+      "step": 284
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.617319908117489,
+      "learning_rate": 3.796291451102381e-05,
+      "loss": 0.5663,
+      "step": 285
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 0.6007622321406768,
+      "learning_rate": 3.793797591072631e-05,
+      "loss": 0.564,
+      "step": 286
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 0.7349189376503978,
+      "learning_rate": 3.791289387885144e-05,
+      "loss": 0.568,
+      "step": 287
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 0.5960213462485423,
+      "learning_rate": 3.788766861595443e-05,
+      "loss": 0.5794,
+      "step": 288
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 0.7048487025552601,
+      "learning_rate": 3.7862300323735835e-05,
+      "loss": 0.5867,
+      "step": 289
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 1.2908470580652365,
+      "learning_rate": 3.783678920503983e-05,
+      "loss": 0.5691,
+      "step": 290
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 0.6669678711945466,
+      "learning_rate": 3.781113546385265e-05,
+      "loss": 0.5839,
+      "step": 291
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 0.6169886909575514,
+      "learning_rate": 3.7785339305300916e-05,
+      "loss": 0.5789,
+      "step": 292
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 0.6267170147384689,
+      "learning_rate": 3.7759400935650025e-05,
+      "loss": 0.5749,
+      "step": 293
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 0.5735683760474032,
+      "learning_rate": 3.77333205623025e-05,
+      "loss": 0.5649,
+      "step": 294
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 0.6333323697198419,
+      "learning_rate": 3.7707098393796295e-05,
+      "loss": 0.5704,
+      "step": 295
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 0.6037336680763106,
+      "learning_rate": 3.7680734639803185e-05,
+      "loss": 0.5755,
+      "step": 296
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 0.698153984506721,
+      "learning_rate": 3.765422951112704e-05,
+      "loss": 0.5802,
+      "step": 297
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 0.6283362095529275,
+      "learning_rate": 3.762758321970216e-05,
+      "loss": 0.5808,
+      "step": 298
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 0.6809328770327113,
+      "learning_rate": 3.7600795978591584e-05,
+      "loss": 0.5636,
+      "step": 299
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 0.6075065618153778,
+      "learning_rate": 3.757386800198538e-05,
+      "loss": 0.5936,
+      "step": 300
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 0.7183550958967533,
+      "learning_rate": 3.7546799505198925e-05,
+      "loss": 0.5854,
+      "step": 301
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 0.6167164572873178,
+      "learning_rate": 3.7519590704671197e-05,
+      "loss": 0.5849,
+      "step": 302
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 0.6678456600089896,
+      "learning_rate": 3.749224181796305e-05,
+      "loss": 0.5796,
+      "step": 303
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.6447685773534853,
+      "learning_rate": 3.746475306375545e-05,
+      "loss": 0.5609,
+      "step": 304
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 0.6262469315325815,
+      "learning_rate": 3.743712466184774e-05,
+      "loss": 0.5814,
+      "step": 305
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 0.6051020801765525,
+      "learning_rate": 3.7409356833155885e-05,
+      "loss": 0.5834,
+      "step": 306
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 0.5633131230854308,
+      "learning_rate": 3.73814497997107e-05,
+      "loss": 0.5749,
+      "step": 307
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 0.5773340491945368,
+      "learning_rate": 3.7353403784656095e-05,
+      "loss": 0.5796,
+      "step": 308
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 0.7035207544279037,
+      "learning_rate": 3.732521901224724e-05,
+      "loss": 0.5728,
+      "step": 309
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 0.6745372482199073,
+      "learning_rate": 3.7296895707848844e-05,
+      "loss": 0.5743,
+      "step": 310
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 0.6857357704254586,
+      "learning_rate": 3.7268434097933275e-05,
+      "loss": 0.5667,
+      "step": 311
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 0.4716603186081334,
+      "learning_rate": 3.723983441007882e-05,
+      "loss": 0.5708,
+      "step": 312
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 0.6145176056154371,
+      "learning_rate": 3.721109687296781e-05,
+      "loss": 0.5847,
+      "step": 313
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 0.5620830364491717,
+      "learning_rate": 3.718222171638484e-05,
+      "loss": 0.5858,
+      "step": 314
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 0.5566497763131408,
+      "learning_rate": 3.715320917121488e-05,
+      "loss": 0.5795,
+      "step": 315
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 0.5519212095801592,
+      "learning_rate": 3.712405946944149e-05,
+      "loss": 0.5773,
+      "step": 316
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 0.5195551870343489,
+      "learning_rate": 3.709477284414488e-05,
+      "loss": 0.5714,
+      "step": 317
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 0.5453367353736083,
+      "learning_rate": 3.7065349529500136e-05,
+      "loss": 0.5786,
+      "step": 318
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 0.597311300832169,
+      "learning_rate": 3.7035789760775305e-05,
+      "loss": 0.5882,
+      "step": 319
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 0.5732330472994642,
+      "learning_rate": 3.70060937743295e-05,
+      "loss": 0.5703,
+      "step": 320
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 0.5266125626137899,
+      "learning_rate": 3.6976261807611045e-05,
+      "loss": 0.5706,
+      "step": 321
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 0.6370521293776403,
+      "learning_rate": 3.694629409915555e-05,
+      "loss": 0.5758,
+      "step": 322
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.6458772205921188,
+      "learning_rate": 3.691619088858401e-05,
+      "loss": 0.5814,
+      "step": 323
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 0.49872078424568583,
+      "learning_rate": 3.68859524166009e-05,
+      "loss": 0.5768,
+      "step": 324
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 0.7298426618554872,
+      "learning_rate": 3.6855578924992226e-05,
+      "loss": 0.5809,
+      "step": 325
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 0.5195832253340972,
+      "learning_rate": 3.6825070656623626e-05,
+      "loss": 0.5762,
+      "step": 326
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 0.8354993356377957,
+      "learning_rate": 3.67944278554384e-05,
+      "loss": 0.5766,
+      "step": 327
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 0.6259139407255918,
+      "learning_rate": 3.676365076645557e-05,
+      "loss": 0.5742,
+      "step": 328
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 0.6531936152826266,
+      "learning_rate": 3.673273963576791e-05,
+      "loss": 0.5653,
+      "step": 329
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 0.5432286079400924,
+      "learning_rate": 3.670169471054001e-05,
+      "loss": 0.5828,
+      "step": 330
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 0.6041983173520113,
+      "learning_rate": 3.6670516239006254e-05,
+      "loss": 0.5694,
+      "step": 331
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 0.600601442684894,
+      "learning_rate": 3.663920447046888e-05,
+      "loss": 0.586,
+      "step": 332
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 0.6244842830611561,
+      "learning_rate": 3.6607759655295954e-05,
+      "loss": 0.5918,
+      "step": 333
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 0.6423401615437673,
+      "learning_rate": 3.657618204491937e-05,
+      "loss": 0.5845,
+      "step": 334
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 0.6306626360115847,
+      "learning_rate": 3.6544471891832875e-05,
+      "loss": 0.5829,
+      "step": 335
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 0.5451278054540003,
+      "learning_rate": 3.6512629449589983e-05,
+      "loss": 0.5708,
+      "step": 336
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 0.6101817583885952,
+      "learning_rate": 3.6480654972802045e-05,
+      "loss": 0.5739,
+      "step": 337
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 0.6881870754619792,
+      "learning_rate": 3.644854871713611e-05,
+      "loss": 0.5856,
+      "step": 338
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 0.4994913082675211,
+      "learning_rate": 3.6416310939312924e-05,
+      "loss": 0.5744,
+      "step": 339
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 0.770651247418208,
+      "learning_rate": 3.638394189710493e-05,
+      "loss": 0.5672,
+      "step": 340
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 0.5849387484355894,
+      "learning_rate": 3.635144184933412e-05,
+      "loss": 0.5734,
+      "step": 341
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.788346098586097,
+      "learning_rate": 3.6318811055869994e-05,
+      "loss": 0.5817,
+      "step": 342
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 0.8394349771423787,
+      "learning_rate": 3.628604977762752e-05,
+      "loss": 0.5753,
+      "step": 343
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 0.519752083684247,
+      "learning_rate": 3.6253158276565004e-05,
+      "loss": 0.5631,
+      "step": 344
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 0.5659530913541807,
+      "learning_rate": 3.6220136815682016e-05,
+      "loss": 0.5826,
+      "step": 345
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 0.5989854284998389,
+      "learning_rate": 3.618698565901727e-05,
+      "loss": 0.5915,
+      "step": 346
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 0.5154387793654174,
+      "learning_rate": 3.615370507164654e-05,
+      "loss": 0.5867,
+      "step": 347
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 0.5363982628357308,
+      "learning_rate": 3.61202953196805e-05,
+      "loss": 0.5554,
+      "step": 348
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 0.6942197720374867,
+      "learning_rate": 3.6086756670262654e-05,
+      "loss": 0.5731,
+      "step": 349
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 0.5098988961716602,
+      "learning_rate": 3.6053089391567135e-05,
+      "loss": 0.5713,
+      "step": 350
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 0.6070847179298107,
+      "learning_rate": 3.601929375279662e-05,
+      "loss": 0.5841,
+      "step": 351
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 0.5467662582640342,
+      "learning_rate": 3.598537002418012e-05,
+      "loss": 0.5702,
+      "step": 352
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 0.5982082967221476,
+      "learning_rate": 3.5951318476970885e-05,
+      "loss": 0.5698,
+      "step": 353
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 0.535485934080746,
+      "learning_rate": 3.591713938344416e-05,
+      "loss": 0.5807,
+      "step": 354
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 0.5839172301681578,
+      "learning_rate": 3.588283301689507e-05,
+      "loss": 0.5738,
+      "step": 355
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 0.5666786525887844,
+      "learning_rate": 3.5848399651636424e-05,
+      "loss": 0.5806,
+      "step": 356
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 0.5052583545847535,
+      "learning_rate": 3.5813839562996473e-05,
+      "loss": 0.5722,
+      "step": 357
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 0.6136400010638171,
+      "learning_rate": 3.5779153027316775e-05,
+      "loss": 0.5831,
+      "step": 358
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 0.5812688271846456,
+      "learning_rate": 3.574434032194994e-05,
+      "loss": 0.5679,
+      "step": 359
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 0.48976518859690316,
+      "learning_rate": 3.5709401725257436e-05,
+      "loss": 0.5803,
+      "step": 360
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.5753548350251354,
+      "learning_rate": 3.5674337516607346e-05,
+      "loss": 0.5734,
+      "step": 361
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 0.546697327206454,
+      "learning_rate": 3.563914797637216e-05,
+      "loss": 0.5649,
+      "step": 362
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 0.6087123160524148,
+      "learning_rate": 3.560383338592649e-05,
+      "loss": 0.5662,
+      "step": 363
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 0.6456548881252534,
+      "learning_rate": 3.556839402764487e-05,
+      "loss": 0.5688,
+      "step": 364
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 0.5973381527120813,
+      "learning_rate": 3.553283018489946e-05,
+      "loss": 0.5786,
+      "step": 365
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 0.5662727642321289,
+      "learning_rate": 3.54971421420578e-05,
+      "loss": 0.5772,
+      "step": 366
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 0.6961357382639561,
+      "learning_rate": 3.5461330184480524e-05,
+      "loss": 0.5792,
+      "step": 367
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 0.6956856438118255,
+      "learning_rate": 3.54253945985191e-05,
+      "loss": 0.5762,
+      "step": 368
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 0.5292784376199834,
+      "learning_rate": 3.5389335671513524e-05,
+      "loss": 0.5633,
+      "step": 369
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 0.6488229304846795,
+      "learning_rate": 3.5353153691789986e-05,
+      "loss": 0.5718,
+      "step": 370
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 0.604501358429834,
+      "learning_rate": 3.531684894865866e-05,
+      "loss": 0.5824,
+      "step": 371
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 0.5816042149340864,
+      "learning_rate": 3.528042173241128e-05,
+      "loss": 0.5755,
+      "step": 372
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 0.6050102600289597,
+      "learning_rate": 3.524387233431892e-05,
+      "loss": 0.5775,
+      "step": 373
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 0.5794505683530962,
+      "learning_rate": 3.520720104662958e-05,
+      "loss": 0.5848,
+      "step": 374
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 0.614248214590476,
+      "learning_rate": 3.5170408162565904e-05,
+      "loss": 0.577,
+      "step": 375
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 0.6089878871882058,
+      "learning_rate": 3.513349397632281e-05,
+      "loss": 0.5745,
+      "step": 376
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 0.7344159844655079,
+      "learning_rate": 3.5096458783065145e-05,
+      "loss": 0.5889,
+      "step": 377
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 0.544082563596434,
+      "learning_rate": 3.505930287892533e-05,
+      "loss": 0.5853,
+      "step": 378
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 0.6582880868500904,
+      "learning_rate": 3.502202656100099e-05,
+      "loss": 0.5813,
+      "step": 379
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.6324561533493505,
+      "learning_rate": 3.498463012735256e-05,
+      "loss": 0.5796,
+      "step": 380
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 0.6716786601151193,
+      "learning_rate": 3.494711387700094e-05,
+      "loss": 0.5624,
+      "step": 381
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 0.6149291808476118,
+      "learning_rate": 3.490947810992505e-05,
+      "loss": 0.5756,
+      "step": 382
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 0.5785374099711249,
+      "learning_rate": 3.48717231270595e-05,
+      "loss": 0.5645,
+      "step": 383
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 0.5564631554861159,
+      "learning_rate": 3.483384923029211e-05,
+      "loss": 0.5839,
+      "step": 384
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 0.6255734729651925,
+      "learning_rate": 3.479585672246156e-05,
+      "loss": 0.5795,
+      "step": 385
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 0.5846116356927361,
+      "learning_rate": 3.4757745907354924e-05,
+      "loss": 0.5821,
+      "step": 386
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 0.5352886861600319,
+      "learning_rate": 3.471951708970527e-05,
+      "loss": 0.5721,
+      "step": 387
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 0.6274708363611826,
+      "learning_rate": 3.468117057518921e-05,
+      "loss": 0.5838,
+      "step": 388
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 0.4896376111225778,
+      "learning_rate": 3.4642706670424426e-05,
+      "loss": 0.5737,
+      "step": 389
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 0.5050026986595124,
+      "learning_rate": 3.460412568296731e-05,
+      "loss": 0.5845,
+      "step": 390
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 0.552971635496396,
+      "learning_rate": 3.456542792131039e-05,
+      "loss": 0.5655,
+      "step": 391
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 0.5873176761587965,
+      "learning_rate": 3.4526613694879936e-05,
+      "loss": 0.5734,
+      "step": 392
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 0.4918355507076112,
+      "learning_rate": 3.448768331403348e-05,
+      "loss": 0.5758,
+      "step": 393
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 0.5506736387250489,
+      "learning_rate": 3.44486370900573e-05,
+      "loss": 0.5724,
+      "step": 394
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 0.4951000953907766,
+      "learning_rate": 3.440947533516396e-05,
+      "loss": 0.5673,
+      "step": 395
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 0.6107456041567669,
+      "learning_rate": 3.4370198362489816e-05,
+      "loss": 0.5917,
+      "step": 396
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 0.6070053163581968,
+      "learning_rate": 3.43308064860925e-05,
+      "loss": 0.5762,
+      "step": 397
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 0.5629467634994745,
+      "learning_rate": 3.4291300020948393e-05,
+      "loss": 0.5808,
+      "step": 398
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.5861014444814027,
+      "learning_rate": 3.425167928295015e-05,
+      "loss": 0.5777,
+      "step": 399
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 0.5338684560866042,
+      "learning_rate": 3.421194458890411e-05,
+      "loss": 0.5774,
+      "step": 400
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 0.5263017307978562,
+      "learning_rate": 3.4172096256527865e-05,
+      "loss": 0.5739,
+      "step": 401
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 0.5669028949920282,
+      "learning_rate": 3.4132134604447595e-05,
+      "loss": 0.5643,
+      "step": 402
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 0.5501817126979249,
+      "learning_rate": 3.409205995219561e-05,
+      "loss": 0.5797,
+      "step": 403
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 0.5239763664480283,
+      "learning_rate": 3.4051872620207765e-05,
+      "loss": 0.5818,
+      "step": 404
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 0.6001702325230553,
+      "learning_rate": 3.40115729298209e-05,
+      "loss": 0.5798,
+      "step": 405
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 0.6357204043829063,
+      "learning_rate": 3.397116120327025e-05,
+      "loss": 0.5845,
+      "step": 406
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 0.5784402007941786,
+      "learning_rate": 3.3930637763686916e-05,
+      "loss": 0.5767,
+      "step": 407
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 0.5872778949520026,
+      "learning_rate": 3.389000293509524e-05,
+      "loss": 0.5682,
+      "step": 408
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 0.5086359268893824,
+      "learning_rate": 3.3849257042410225e-05,
+      "loss": 0.5615,
+      "step": 409
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 0.4585913498228958,
+      "learning_rate": 3.380840041143494e-05,
+      "loss": 0.5724,
+      "step": 410
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 0.5024323899597434,
+      "learning_rate": 3.376743336885792e-05,
+      "loss": 0.5713,
+      "step": 411
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 0.4551879635974102,
+      "learning_rate": 3.372635624225055e-05,
+      "loss": 0.5723,
+      "step": 412
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 0.4705477904886718,
+      "learning_rate": 3.3685169360064445e-05,
+      "loss": 0.5703,
+      "step": 413
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 0.491996278920646,
+      "learning_rate": 3.3643873051628804e-05,
+      "loss": 0.5651,
+      "step": 414
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 0.49756141818246935,
+      "learning_rate": 3.3602467647147814e-05,
+      "loss": 0.5855,
+      "step": 415
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 0.5115503626102201,
+      "learning_rate": 3.356095347769799e-05,
+      "loss": 0.5757,
+      "step": 416
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 0.5041602395737734,
+      "learning_rate": 3.351933087522552e-05,
+      "loss": 0.5715,
+      "step": 417
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.46803486032849767,
+      "learning_rate": 3.3477600172543626e-05,
+      "loss": 0.5652,
+      "step": 418
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 0.5138837457909423,
+      "learning_rate": 3.3435761703329894e-05,
+      "loss": 0.5754,
+      "step": 419
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 0.4957665056462923,
+      "learning_rate": 3.3393815802123606e-05,
+      "loss": 0.5559,
+      "step": 420
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 0.5027195379010544,
+      "learning_rate": 3.335176280432308e-05,
+      "loss": 0.5765,
+      "step": 421
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 0.5236483936749398,
+      "learning_rate": 3.330960304618295e-05,
+      "loss": 0.5801,
+      "step": 422
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 0.49538209913012177,
+      "learning_rate": 3.326733686481153e-05,
+      "loss": 0.5796,
+      "step": 423
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 0.5213070345785242,
+      "learning_rate": 3.3224964598168084e-05,
+      "loss": 0.5731,
+      "step": 424
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 0.4609095842328061,
+      "learning_rate": 3.318248658506012e-05,
+      "loss": 0.5786,
+      "step": 425
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 0.5172273983589015,
+      "learning_rate": 3.313990316514069e-05,
+      "loss": 0.5803,
+      "step": 426
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 0.48232428295341495,
+      "learning_rate": 3.309721467890571e-05,
+      "loss": 0.5809,
+      "step": 427
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 0.4549190254042141,
+      "learning_rate": 3.305442146769115e-05,
+      "loss": 0.5638,
+      "step": 428
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 0.5271979713459393,
+      "learning_rate": 3.30115238736704e-05,
+      "loss": 0.5834,
+      "step": 429
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 0.455576877922238,
+      "learning_rate": 3.296852223985148e-05,
+      "loss": 0.5734,
+      "step": 430
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 0.5659675116385305,
+      "learning_rate": 3.2925416910074315e-05,
+      "loss": 0.5673,
+      "step": 431
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 0.4811656828608784,
+      "learning_rate": 3.288220822900796e-05,
+      "loss": 0.577,
+      "step": 432
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 0.49439012303715196,
+      "learning_rate": 3.283889654214788e-05,
+      "loss": 0.5662,
+      "step": 433
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 0.548207613114792,
+      "learning_rate": 3.279548219581317e-05,
+      "loss": 0.5757,
+      "step": 434
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 0.5140922147055464,
+      "learning_rate": 3.275196553714379e-05,
+      "loss": 0.5647,
+      "step": 435
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 0.48971000155351596,
+      "learning_rate": 3.270834691409778e-05,
+      "loss": 0.564,
+      "step": 436
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.46373813614560966,
+      "learning_rate": 3.266462667544849e-05,
+      "loss": 0.5725,
+      "step": 437
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 0.4765081797584862,
+      "learning_rate": 3.262080517078178e-05,
+      "loss": 0.5734,
+      "step": 438
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 0.48437726146704724,
+      "learning_rate": 3.257688275049323e-05,
+      "loss": 0.573,
+      "step": 439
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 0.5266053842160988,
+      "learning_rate": 3.253285976578535e-05,
+      "loss": 0.5572,
+      "step": 440
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 0.47535690583884144,
+      "learning_rate": 3.2488736568664756e-05,
+      "loss": 0.571,
+      "step": 441
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 0.5278660963713033,
+      "learning_rate": 3.244451351193935e-05,
+      "loss": 0.5723,
+      "step": 442
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 0.46068883780152264,
+      "learning_rate": 3.240019094921551e-05,
+      "loss": 0.568,
+      "step": 443
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 0.5004369562157935,
+      "learning_rate": 3.235576923489527e-05,
+      "loss": 0.564,
+      "step": 444
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 0.6608088888744782,
+      "learning_rate": 3.2311248724173476e-05,
+      "loss": 0.5791,
+      "step": 445
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 0.46157315495379736,
+      "learning_rate": 3.226662977303494e-05,
+      "loss": 0.5682,
+      "step": 446
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 0.5886214001809562,
+      "learning_rate": 3.2221912738251626e-05,
+      "loss": 0.5837,
+      "step": 447
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 0.5673061366882541,
+      "learning_rate": 3.2177097977379736e-05,
+      "loss": 0.5671,
+      "step": 448
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 0.6294855822529412,
+      "learning_rate": 3.21321858487569e-05,
+      "loss": 0.5659,
+      "step": 449
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 0.5383225442240591,
+      "learning_rate": 3.208717671149931e-05,
+      "loss": 0.5717,
+      "step": 450
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 0.6192770334081542,
+      "learning_rate": 3.204207092549882e-05,
+      "loss": 0.5697,
+      "step": 451
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 0.5405993306443383,
+      "learning_rate": 3.19968688514201e-05,
+      "loss": 0.5729,
+      "step": 452
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 0.5973631161303004,
+      "learning_rate": 3.195157085069774e-05,
+      "loss": 0.5852,
+      "step": 453
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 0.5700061415869571,
+      "learning_rate": 3.190617728553332e-05,
+      "loss": 0.5806,
+      "step": 454
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 0.5152599449384585,
+      "learning_rate": 3.1860688518892606e-05,
+      "loss": 0.5797,
+      "step": 455
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.5916480939721651,
+      "learning_rate": 3.181510491450254e-05,
+      "loss": 0.5819,
+      "step": 456
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 0.4708682156293534,
+      "learning_rate": 3.176942683684842e-05,
+      "loss": 0.582,
+      "step": 457
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 0.495568670203384,
+      "learning_rate": 3.1723654651170934e-05,
+      "loss": 0.5661,
+      "step": 458
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 0.5384166169569073,
+      "learning_rate": 3.167778872346327e-05,
+      "loss": 0.5768,
+      "step": 459
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 0.47271774500833125,
+      "learning_rate": 3.163182942046815e-05,
+      "loss": 0.5737,
+      "step": 460
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 0.579079343199453,
+      "learning_rate": 3.158577710967494e-05,
+      "loss": 0.5663,
+      "step": 461
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 0.5498260293854788,
+      "learning_rate": 3.153963215931669e-05,
+      "loss": 0.5645,
+      "step": 462
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 0.6236058315191131,
+      "learning_rate": 3.14933949383672e-05,
+      "loss": 0.5849,
+      "step": 463
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 0.7454393308545267,
+      "learning_rate": 3.144706581653805e-05,
+      "loss": 0.5629,
+      "step": 464
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 0.5288343875804269,
+      "learning_rate": 3.140064516427566e-05,
+      "loss": 0.5829,
+      "step": 465
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 0.8255189944738969,
+      "learning_rate": 3.135413335275833e-05,
+      "loss": 0.5733,
+      "step": 466
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 0.5947311332285182,
+      "learning_rate": 3.130753075389327e-05,
+      "loss": 0.5678,
+      "step": 467
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 0.6741702868520348,
+      "learning_rate": 3.1260837740313625e-05,
+      "loss": 0.5748,
+      "step": 468
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 0.5662507505519576,
+      "learning_rate": 3.121405468537547e-05,
+      "loss": 0.5721,
+      "step": 469
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 0.6318405247143001,
+      "learning_rate": 3.116718196315488e-05,
+      "loss": 0.5633,
+      "step": 470
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 0.5342653145246871,
+      "learning_rate": 3.1120219948444894e-05,
+      "loss": 0.5649,
+      "step": 471
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 0.5562388606479534,
+      "learning_rate": 3.107316901675252e-05,
+      "loss": 0.5777,
+      "step": 472
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 0.5193509648285766,
+      "learning_rate": 3.1026029544295745e-05,
+      "loss": 0.5764,
+      "step": 473
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 0.5331607288799004,
+      "learning_rate": 3.097880190800057e-05,
+      "loss": 0.5767,
+      "step": 474
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.5307103551999701,
+      "learning_rate": 3.093148648549788e-05,
+      "loss": 0.5726,
+      "step": 475
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 0.5195639410491685,
+      "learning_rate": 3.088408365512055e-05,
+      "loss": 0.5661,
+      "step": 476
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 0.5469503146837309,
+      "learning_rate": 3.083659379590034e-05,
+      "loss": 0.5764,
+      "step": 477
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 0.5226732320729406,
+      "learning_rate": 3.0789017287564913e-05,
+      "loss": 0.5736,
+      "step": 478
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 0.5556267272823231,
+      "learning_rate": 3.074135451053475e-05,
+      "loss": 0.5618,
+      "step": 479
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 0.5350751032479102,
+      "learning_rate": 3.069360584592015e-05,
+      "loss": 0.5642,
+      "step": 480
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 0.5172535597743249,
+      "learning_rate": 3.064577167551816e-05,
+      "loss": 0.5658,
+      "step": 481
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 0.558030608963143,
+      "learning_rate": 3.059785238180953e-05,
+      "loss": 0.5692,
+      "step": 482
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 0.5179070163953278,
+      "learning_rate": 3.054984834795566e-05,
+      "loss": 0.5581,
+      "step": 483
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 0.5903848563192007,
+      "learning_rate": 3.0501759957795537e-05,
+      "loss": 0.5719,
+      "step": 484
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 0.5311402293082551,
+      "learning_rate": 3.0453587595842627e-05,
+      "loss": 0.5781,
+      "step": 485
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 0.5274583466072534,
+      "learning_rate": 3.0405331647281856e-05,
+      "loss": 0.5811,
+      "step": 486
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 0.5857739638066775,
+      "learning_rate": 3.0356992497966508e-05,
+      "loss": 0.553,
+      "step": 487
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 0.4847393023046402,
+      "learning_rate": 3.0308570534415137e-05,
+      "loss": 0.562,
+      "step": 488
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 0.620992529766653,
+      "learning_rate": 3.0260066143808467e-05,
+      "loss": 0.5816,
+      "step": 489
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 0.5521122842032309,
+      "learning_rate": 3.021147971398632e-05,
+      "loss": 0.5817,
+      "step": 490
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 0.5259849708856554,
+      "learning_rate": 3.0162811633444478e-05,
+      "loss": 0.5701,
+      "step": 491
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 0.5658232289564129,
+      "learning_rate": 3.0114062291331635e-05,
+      "loss": 0.5768,
+      "step": 492
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 0.5163215485961835,
+      "learning_rate": 3.0065232077446217e-05,
+      "loss": 0.5729,
+      "step": 493
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.5428414300261472,
+      "learning_rate": 3.001632138223332e-05,
+      "loss": 0.5631,
+      "step": 494
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 0.7277299060883373,
+      "learning_rate": 2.9967330596781565e-05,
+      "loss": 0.4535,
+      "step": 495
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 0.8189564152107975,
+      "learning_rate": 2.9918260112819952e-05,
+      "loss": 0.4527,
+      "step": 496
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 0.7356482419832299,
+      "learning_rate": 2.9869110322714778e-05,
+      "loss": 0.4456,
+      "step": 497
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 0.6512607303863459,
+      "learning_rate": 2.9819881619466447e-05,
+      "loss": 0.4464,
+      "step": 498
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 0.5675975642937546,
+      "learning_rate": 2.9770574396706364e-05,
+      "loss": 0.4477,
+      "step": 499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 0.6306955683042804,
+      "learning_rate": 2.972118904869376e-05,
+      "loss": 0.43,
+      "step": 500
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 0.6420989392328846,
+      "learning_rate": 2.9671725970312563e-05,
+      "loss": 0.4523,
+      "step": 501
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 0.5536745632431284,
+      "learning_rate": 2.9622185557068238e-05,
+      "loss": 0.4335,
+      "step": 502
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 0.5706706769162428,
+      "learning_rate": 2.95725682050846e-05,
+      "loss": 0.4476,
+      "step": 503
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 0.5257510520101587,
+      "learning_rate": 2.9522874311100684e-05,
+      "loss": 0.4473,
+      "step": 504
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 0.5269203437016025,
+      "learning_rate": 2.9473104272467526e-05,
+      "loss": 0.4421,
+      "step": 505
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 0.6020166924917083,
+      "learning_rate": 2.942325848714505e-05,
+      "loss": 0.4321,
+      "step": 506
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 0.5634655968117813,
+      "learning_rate": 2.9373337353698804e-05,
+      "loss": 0.4526,
+      "step": 507
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 0.6686099799547104,
+      "learning_rate": 2.932334127129686e-05,
+      "loss": 0.4444,
+      "step": 508
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 0.5262295183761894,
+      "learning_rate": 2.9273270639706547e-05,
+      "loss": 0.4292,
+      "step": 509
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 0.5036770163747739,
+      "learning_rate": 2.922312585929131e-05,
+      "loss": 0.4453,
+      "step": 510
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 0.5886875163305816,
+      "learning_rate": 2.9172907331007478e-05,
+      "loss": 0.437,
+      "step": 511
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 0.5211487878831738,
+      "learning_rate": 2.9122615456401053e-05,
+      "loss": 0.4544,
+      "step": 512
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.5211449075973712,
+      "learning_rate": 2.9072250637604535e-05,
+      "loss": 0.4414,
+      "step": 513
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 0.635555375200411,
+      "learning_rate": 2.9021813277333667e-05,
+      "loss": 0.4467,
+      "step": 514
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 0.49125312209194855,
+      "learning_rate": 2.8971303778884244e-05,
+      "loss": 0.4398,
+      "step": 515
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 0.679978257317568,
+      "learning_rate": 2.8920722546128863e-05,
+      "loss": 0.4391,
+      "step": 516
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 0.4869238888764651,
+      "learning_rate": 2.887006998351372e-05,
+      "loss": 0.4567,
+      "step": 517
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 0.6229087604081991,
+      "learning_rate": 2.881934649605535e-05,
+      "loss": 0.4491,
+      "step": 518
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 0.613124376108937,
+      "learning_rate": 2.8768552489337408e-05,
+      "loss": 0.4345,
+      "step": 519
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 0.514549225874971,
+      "learning_rate": 2.8717688369507425e-05,
+      "loss": 0.443,
+      "step": 520
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 0.5379170341363162,
+      "learning_rate": 2.8666754543273535e-05,
+      "loss": 0.434,
+      "step": 521
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 0.5167699946657967,
+      "learning_rate": 2.8615751417901273e-05,
+      "loss": 0.4277,
+      "step": 522
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 0.4858233350037585,
+      "learning_rate": 2.856467940121025e-05,
+      "loss": 0.456,
+      "step": 523
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 0.5222570692242859,
+      "learning_rate": 2.8513538901570964e-05,
+      "loss": 0.4241,
+      "step": 524
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 0.46182723893285893,
+      "learning_rate": 2.846233032790149e-05,
+      "loss": 0.4437,
+      "step": 525
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 0.4829836658174888,
+      "learning_rate": 2.84110540896642e-05,
+      "loss": 0.4405,
+      "step": 526
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 0.4673039728532001,
+      "learning_rate": 2.8359710596862555e-05,
+      "loss": 0.4414,
+      "step": 527
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 0.5062814234221203,
+      "learning_rate": 2.8308300260037734e-05,
+      "loss": 0.4399,
+      "step": 528
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 0.4801559275138607,
+      "learning_rate": 2.8256823490265424e-05,
+      "loss": 0.4331,
+      "step": 529
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 0.482697718460908,
+      "learning_rate": 2.8205280699152515e-05,
+      "loss": 0.4488,
+      "step": 530
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 0.6106898415600357,
+      "learning_rate": 2.8153672298833777e-05,
+      "loss": 0.4484,
+      "step": 531
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.4994360874423692,
+      "learning_rate": 2.8101998701968615e-05,
+      "loss": 0.4419,
+      "step": 532
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 0.5061547214570566,
+      "learning_rate": 2.805026032173773e-05,
+      "loss": 0.4502,
+      "step": 533
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 0.5249093147870095,
+      "learning_rate": 2.7998457571839832e-05,
+      "loss": 0.4408,
+      "step": 534
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 0.5305125455991501,
+      "learning_rate": 2.7946590866488322e-05,
+      "loss": 0.4406,
+      "step": 535
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 0.48307952946898614,
+      "learning_rate": 2.7894660620408e-05,
+      "loss": 0.4492,
+      "step": 536
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 0.5220756644894622,
+      "learning_rate": 2.784266724883173e-05,
+      "loss": 0.4352,
+      "step": 537
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 0.5537772853128828,
+      "learning_rate": 2.7790611167497128e-05,
+      "loss": 0.4427,
+      "step": 538
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 0.5128400211067883,
+      "learning_rate": 2.7738492792643224e-05,
+      "loss": 0.4362,
+      "step": 539
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 0.518978011341626,
+      "learning_rate": 2.768631254100716e-05,
+      "loss": 0.4439,
+      "step": 540
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 0.5105663115497837,
+      "learning_rate": 2.7634070829820836e-05,
+      "loss": 0.4486,
+      "step": 541
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 0.4471663696184521,
+      "learning_rate": 2.7581768076807592e-05,
+      "loss": 0.4468,
+      "step": 542
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 0.5233900585593475,
+      "learning_rate": 2.7529404700178832e-05,
+      "loss": 0.4377,
+      "step": 543
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 0.4403871823554908,
+      "learning_rate": 2.7476981118630735e-05,
+      "loss": 0.4351,
+      "step": 544
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 0.4612300790121619,
+      "learning_rate": 2.742449775134085e-05,
+      "loss": 0.436,
+      "step": 545
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 0.46577918885819,
+      "learning_rate": 2.7371955017964788e-05,
+      "loss": 0.4593,
+      "step": 546
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 0.477851525831599,
+      "learning_rate": 2.7319353338632845e-05,
+      "loss": 0.4459,
+      "step": 547
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 0.4473397072249732,
+      "learning_rate": 2.7266693133946645e-05,
+      "loss": 0.4421,
+      "step": 548
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 0.521899197925737,
+      "learning_rate": 2.7213974824975785e-05,
+      "loss": 0.4457,
+      "step": 549
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 0.49310610778159214,
+      "learning_rate": 2.7161198833254464e-05,
+      "loss": 0.4437,
+      "step": 550
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.5159032790314152,
+      "learning_rate": 2.7108365580778095e-05,
+      "loss": 0.4474,
+      "step": 551
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 0.4975208996324524,
+      "learning_rate": 2.705547548999996e-05,
+      "loss": 0.4515,
+      "step": 552
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 0.5359574899373757,
+      "learning_rate": 2.7002528983827817e-05,
+      "loss": 0.4403,
+      "step": 553
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 0.4748303815792672,
+      "learning_rate": 2.6949526485620517e-05,
+      "loss": 0.4454,
+      "step": 554
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 0.5717262423979245,
+      "learning_rate": 2.6896468419184623e-05,
+      "loss": 0.4525,
+      "step": 555
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 0.4800264714687931,
+      "learning_rate": 2.6843355208771024e-05,
+      "loss": 0.4388,
+      "step": 556
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 0.5247942217650406,
+      "learning_rate": 2.6790187279071536e-05,
+      "loss": 0.4544,
+      "step": 557
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 0.536027458957431,
+      "learning_rate": 2.67369650552155e-05,
+      "loss": 0.4366,
+      "step": 558
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 0.45500912033698715,
+      "learning_rate": 2.6683688962766407e-05,
+      "loss": 0.4362,
+      "step": 559
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 0.5426686995716679,
+      "learning_rate": 2.6630359427718468e-05,
+      "loss": 0.4436,
+      "step": 560
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 0.4448005930547741,
+      "learning_rate": 2.6576976876493225e-05,
+      "loss": 0.4347,
+      "step": 561
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 0.48795217869306323,
+      "learning_rate": 2.652354173593614e-05,
+      "loss": 0.4387,
+      "step": 562
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 0.4783673787558888,
+      "learning_rate": 2.647005443331317e-05,
+      "loss": 0.4653,
+      "step": 563
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 0.4850590294383947,
+      "learning_rate": 2.6416515396307354e-05,
+      "loss": 0.4554,
+      "step": 564
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 0.4836007639209731,
+      "learning_rate": 2.636292505301542e-05,
+      "loss": 0.4407,
+      "step": 565
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 0.49267233668728755,
+      "learning_rate": 2.630928383194431e-05,
+      "loss": 0.4511,
+      "step": 566
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 0.40908954288203586,
+      "learning_rate": 2.625559216200781e-05,
+      "loss": 0.4469,
+      "step": 567
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 0.5103823026145429,
+      "learning_rate": 2.620185047252308e-05,
+      "loss": 0.4443,
+      "step": 568
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 0.4763968203605892,
+      "learning_rate": 2.614805919320723e-05,
+      "loss": 0.4511,
+      "step": 569
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.4851936130830955,
+      "learning_rate": 2.6094218754173895e-05,
+      "loss": 0.4538,
+      "step": 570
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 0.4332435537034786,
+      "learning_rate": 2.60403295859298e-05,
+      "loss": 0.4363,
+      "step": 571
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 0.5262609515175539,
+      "learning_rate": 2.5986392119371297e-05,
+      "loss": 0.4523,
+      "step": 572
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 0.4674592866296592,
+      "learning_rate": 2.5932406785780928e-05,
+      "loss": 0.4484,
+      "step": 573
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 0.4717261238065866,
+      "learning_rate": 2.5878374016823987e-05,
+      "loss": 0.4373,
+      "step": 574
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 0.46493963430218416,
+      "learning_rate": 2.5824294244545045e-05,
+      "loss": 0.4583,
+      "step": 575
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 0.4752393100794034,
+      "learning_rate": 2.577016790136453e-05,
+      "loss": 0.4535,
+      "step": 576
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 0.4931109838303442,
+      "learning_rate": 2.5715995420075225e-05,
+      "loss": 0.4339,
+      "step": 577
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 0.48332047372799974,
+      "learning_rate": 2.5661777233838864e-05,
+      "loss": 0.4463,
+      "step": 578
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 0.4914185545988274,
+      "learning_rate": 2.5607513776182616e-05,
+      "loss": 0.4453,
+      "step": 579
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 0.44131359902483985,
+      "learning_rate": 2.5553205480995633e-05,
+      "loss": 0.4417,
+      "step": 580
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 0.4914836977522714,
+      "learning_rate": 2.5498852782525604e-05,
+      "loss": 0.4484,
+      "step": 581
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 0.4716688977727445,
+      "learning_rate": 2.5444456115375252e-05,
+      "loss": 0.4515,
+      "step": 582
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 0.5150581532543587,
+      "learning_rate": 2.539001591449889e-05,
+      "loss": 0.4331,
+      "step": 583
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 0.45038879809106674,
+      "learning_rate": 2.53355326151989e-05,
+      "loss": 0.4466,
+      "step": 584
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 0.5008613907767624,
+      "learning_rate": 2.5281006653122288e-05,
+      "loss": 0.4351,
+      "step": 585
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 0.4644195065600521,
+      "learning_rate": 2.5226438464257217e-05,
+      "loss": 0.4675,
+      "step": 586
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 0.5013139439747594,
+      "learning_rate": 2.5171828484929447e-05,
+      "loss": 0.4493,
+      "step": 587
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 0.4346687718008206,
+      "learning_rate": 2.5117177151798934e-05,
+      "loss": 0.4375,
+      "step": 588
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.45621295449986815,
+      "learning_rate": 2.506248490185627e-05,
+      "loss": 0.4489,
+      "step": 589
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 0.46041238033892146,
+      "learning_rate": 2.500775217241923e-05,
+      "loss": 0.4439,
+      "step": 590
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 0.42630558366461785,
+      "learning_rate": 2.4952979401129266e-05,
+      "loss": 0.4513,
+      "step": 591
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 0.4604435793930857,
+      "learning_rate": 2.4898167025948e-05,
+      "loss": 0.4474,
+      "step": 592
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 0.47623292976412746,
+      "learning_rate": 2.4843315485153703e-05,
+      "loss": 0.4447,
+      "step": 593
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 0.49008163851591807,
+      "learning_rate": 2.4788425217337856e-05,
+      "loss": 0.4476,
+      "step": 594
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 0.49767271504043303,
+      "learning_rate": 2.4733496661401572e-05,
+      "loss": 0.4494,
+      "step": 595
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 0.46873970443938234,
+      "learning_rate": 2.4678530256552114e-05,
+      "loss": 0.4448,
+      "step": 596
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 0.549552775911276,
+      "learning_rate": 2.46235264422994e-05,
+      "loss": 0.4416,
+      "step": 597
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 0.45516396379497287,
+      "learning_rate": 2.4568485658452452e-05,
+      "loss": 0.4373,
+      "step": 598
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 0.6341154707843578,
+      "learning_rate": 2.451340834511592e-05,
+      "loss": 0.443,
+      "step": 599
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 0.4555942519715106,
+      "learning_rate": 2.4458294942686528e-05,
+      "loss": 0.4306,
+      "step": 600
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 0.5452058928789751,
+      "learning_rate": 2.440314589184958e-05,
+      "loss": 0.4508,
+      "step": 601
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 0.48468940001091654,
+      "learning_rate": 2.434796163357541e-05,
+      "loss": 0.4456,
+      "step": 602
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 0.4862534409422184,
+      "learning_rate": 2.429274260911588e-05,
+      "loss": 0.4351,
+      "step": 603
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 0.514698132848713,
+      "learning_rate": 2.423748926000084e-05,
+      "loss": 0.4511,
+      "step": 604
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 0.46355729603008156,
+      "learning_rate": 2.418220202803459e-05,
+      "loss": 0.4494,
+      "step": 605
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 0.5305200649767189,
+      "learning_rate": 2.412688135529237e-05,
+      "loss": 0.4353,
+      "step": 606
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 0.43386288503535403,
+      "learning_rate": 2.4071527684116807e-05,
+      "loss": 0.4506,
+      "step": 607
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.49583646632493406,
+      "learning_rate": 2.4016141457114372e-05,
+      "loss": 0.448,
+      "step": 608
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 0.46009820240845806,
+      "learning_rate": 2.396072311715187e-05,
+      "loss": 0.446,
+      "step": 609
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 0.5130043620628927,
+      "learning_rate": 2.3905273107352863e-05,
+      "loss": 0.4524,
+      "step": 610
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 0.46435628675596785,
+      "learning_rate": 2.3849791871094157e-05,
+      "loss": 0.4461,
+      "step": 611
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 0.5063229797023708,
+      "learning_rate": 2.379427985200223e-05,
+      "loss": 0.4499,
+      "step": 612
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 0.4481758528431784,
+      "learning_rate": 2.373873749394972e-05,
+      "loss": 0.4603,
+      "step": 613
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 0.48780373634578944,
+      "learning_rate": 2.3683165241051836e-05,
+      "loss": 0.4507,
+      "step": 614
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 0.4994437705714549,
+      "learning_rate": 2.3627563537662823e-05,
+      "loss": 0.4454,
+      "step": 615
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 0.477883016455413,
+      "learning_rate": 2.3571932828372433e-05,
+      "loss": 0.4387,
+      "step": 616
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 0.49109976798673205,
+      "learning_rate": 2.351627355800233e-05,
+      "loss": 0.4532,
+      "step": 617
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 0.4841506510153999,
+      "learning_rate": 2.346058617160256e-05,
+      "loss": 0.4514,
+      "step": 618
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 0.510774908201177,
+      "learning_rate": 2.3404871114447976e-05,
+      "loss": 0.4475,
+      "step": 619
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 0.5016353316760236,
+      "learning_rate": 2.334912883203469e-05,
+      "loss": 0.4532,
+      "step": 620
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 0.5270998889963494,
+      "learning_rate": 2.3293359770076505e-05,
+      "loss": 0.4599,
+      "step": 621
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 0.46746243966219475,
+      "learning_rate": 2.323756437450136e-05,
+      "loss": 0.4427,
+      "step": 622
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 0.5821893088227773,
+      "learning_rate": 2.3181743091447754e-05,
+      "loss": 0.4366,
+      "step": 623
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 0.4423657447098324,
+      "learning_rate": 2.3125896367261164e-05,
+      "loss": 0.4552,
+      "step": 624
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 0.528815304908352,
+      "learning_rate": 2.307002464849052e-05,
+      "loss": 0.4408,
+      "step": 625
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 0.48507936386395434,
+      "learning_rate": 2.30141283818846e-05,
+      "loss": 0.4475,
+      "step": 626
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.46676713370903833,
+      "learning_rate": 2.2958208014388462e-05,
+      "loss": 0.4398,
+      "step": 627
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 0.47379279426000354,
+      "learning_rate": 2.2902263993139857e-05,
+      "loss": 0.449,
+      "step": 628
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 0.4297732471896876,
+      "learning_rate": 2.2846296765465708e-05,
+      "loss": 0.4475,
+      "step": 629
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 0.46511784467009243,
+      "learning_rate": 2.2790306778878465e-05,
+      "loss": 0.438,
+      "step": 630
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 0.47135791396070176,
+      "learning_rate": 2.273429448107257e-05,
+      "loss": 0.4393,
+      "step": 631
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 0.4297904434999469,
+      "learning_rate": 2.2678260319920868e-05,
+      "loss": 0.4353,
+      "step": 632
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 0.4988448263897187,
+      "learning_rate": 2.262220474347102e-05,
+      "loss": 0.4401,
+      "step": 633
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 0.45352687242882067,
+      "learning_rate": 2.2566128199941918e-05,
+      "loss": 0.4474,
+      "step": 634
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 0.4716383161351407,
+      "learning_rate": 2.2510031137720104e-05,
+      "loss": 0.4394,
+      "step": 635
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 0.43496820524475277,
+      "learning_rate": 2.2453914005356206e-05,
+      "loss": 0.4397,
+      "step": 636
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 0.4411238246781254,
+      "learning_rate": 2.2397777251561302e-05,
+      "loss": 0.4525,
+      "step": 637
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 0.487210435040213,
+      "learning_rate": 2.2341621325203393e-05,
+      "loss": 0.4484,
+      "step": 638
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 0.448759324379628,
+      "learning_rate": 2.2285446675303764e-05,
+      "loss": 0.4376,
+      "step": 639
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 0.5002441114932388,
+      "learning_rate": 2.2229253751033423e-05,
+      "loss": 0.4347,
+      "step": 640
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 0.4784720487646929,
+      "learning_rate": 2.2173043001709492e-05,
+      "loss": 0.458,
+      "step": 641
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 0.46820009456571043,
+      "learning_rate": 2.2116814876791627e-05,
+      "loss": 0.4571,
+      "step": 642
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 0.5083326109245436,
+      "learning_rate": 2.2060569825878413e-05,
+      "loss": 0.4522,
+      "step": 643
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 0.46547553026681815,
+      "learning_rate": 2.2004308298703777e-05,
+      "loss": 0.4377,
+      "step": 644
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 0.45358842288298545,
+      "learning_rate": 2.1948030745133396e-05,
+      "loss": 0.4511,
+      "step": 645
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.4345941679901805,
+      "learning_rate": 2.1891737615161083e-05,
+      "loss": 0.4479,
+      "step": 646
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 0.45453786816951613,
+      "learning_rate": 2.1835429358905218e-05,
+      "loss": 0.4375,
+      "step": 647
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 0.45688058149906047,
+      "learning_rate": 2.17791064266051e-05,
+      "loss": 0.4385,
+      "step": 648
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 0.4333334702754955,
+      "learning_rate": 2.172276926861741e-05,
+      "loss": 0.437,
+      "step": 649
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 0.45666521964354234,
+      "learning_rate": 2.166641833541255e-05,
+      "loss": 0.4426,
+      "step": 650
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 0.45266581030753006,
+      "learning_rate": 2.161005407757109e-05,
+      "loss": 0.4329,
+      "step": 651
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 0.46238520877831263,
+      "learning_rate": 2.155367694578013e-05,
+      "loss": 0.445,
+      "step": 652
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 0.4705804619810313,
+      "learning_rate": 2.1497287390829712e-05,
+      "loss": 0.4632,
+      "step": 653
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 0.45846596230638514,
+      "learning_rate": 2.1440885863609208e-05,
+      "loss": 0.4247,
+      "step": 654
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 0.4978120653130781,
+      "learning_rate": 2.1384472815103737e-05,
+      "loss": 0.4636,
+      "step": 655
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 0.4449371159510662,
+      "learning_rate": 2.1328048696390534e-05,
+      "loss": 0.4408,
+      "step": 656
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 0.515342942780243,
+      "learning_rate": 2.1271613958635342e-05,
+      "loss": 0.4454,
+      "step": 657
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 0.4575747817138289,
+      "learning_rate": 2.1215169053088825e-05,
+      "loss": 0.4506,
+      "step": 658
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 0.4713000463673081,
+      "learning_rate": 2.1158714431082937e-05,
+      "loss": 0.4524,
+      "step": 659
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 0.4488692541574474,
+      "learning_rate": 2.110225054402733e-05,
+      "loss": 0.4567,
+      "step": 660
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 0.5014227906182664,
+      "learning_rate": 2.1045777843405747e-05,
+      "loss": 0.4598,
+      "step": 661
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 0.45512195659307453,
+      "learning_rate": 2.09892967807724e-05,
+      "loss": 0.4453,
+      "step": 662
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 0.5031461037421029,
+      "learning_rate": 2.093280780774835e-05,
+      "loss": 0.4584,
+      "step": 663
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 0.45354996121334445,
+      "learning_rate": 2.087631137601793e-05,
+      "loss": 0.437,
+      "step": 664
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.4951329448709223,
+      "learning_rate": 2.0819807937325104e-05,
+      "loss": 0.4545,
+      "step": 665
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 0.44573353472227406,
+      "learning_rate": 2.0763297943469847e-05,
+      "loss": 0.4309,
+      "step": 666
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 0.5011566984094569,
+      "learning_rate": 2.070678184630458e-05,
+      "loss": 0.4493,
+      "step": 667
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 0.4697168098734015,
+      "learning_rate": 2.06502600977305e-05,
+      "loss": 0.453,
+      "step": 668
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 0.4437862417476549,
+      "learning_rate": 2.0593733149694008e-05,
+      "loss": 0.4409,
+      "step": 669
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 0.46263382022536503,
+      "learning_rate": 2.0537201454183074e-05,
+      "loss": 0.4505,
+      "step": 670
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 0.47979681222890735,
+      "learning_rate": 2.0480665463223626e-05,
+      "loss": 0.4453,
+      "step": 671
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 0.49368017901028094,
+      "learning_rate": 2.042412562887594e-05,
+      "loss": 0.4505,
+      "step": 672
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 0.9316120358540896,
+      "learning_rate": 2.0367582403231033e-05,
+      "loss": 0.4609,
+      "step": 673
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 0.4289389044095614,
+      "learning_rate": 2.0311036238407023e-05,
+      "loss": 0.4465,
+      "step": 674
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 0.5096552229415786,
+      "learning_rate": 2.0254487586545534e-05,
+      "loss": 0.4382,
+      "step": 675
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 0.4803783685262513,
+      "learning_rate": 2.0197936899808088e-05,
+      "loss": 0.4481,
+      "step": 676
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 0.5112320995274277,
+      "learning_rate": 2.0141384630372466e-05,
+      "loss": 0.4383,
+      "step": 677
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 0.48387710734083,
+      "learning_rate": 2.0084831230429106e-05,
+      "loss": 0.4626,
+      "step": 678
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 0.4459248206622319,
+      "learning_rate": 2.0028277152177492e-05,
+      "loss": 0.4468,
+      "step": 679
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 0.4543564861084283,
+      "learning_rate": 1.9971722847822518e-05,
+      "loss": 0.4533,
+      "step": 680
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 0.4297968532721812,
+      "learning_rate": 1.99151687695709e-05,
+      "loss": 0.4278,
+      "step": 681
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 0.47722672955481643,
+      "learning_rate": 1.9858615369627537e-05,
+      "loss": 0.437,
+      "step": 682
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 0.45425747513232356,
+      "learning_rate": 1.9802063100191916e-05,
+      "loss": 0.4476,
+      "step": 683
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.4449751545395058,
+      "learning_rate": 1.974551241345447e-05,
+      "loss": 0.444,
+      "step": 684
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 0.41817986727461,
+      "learning_rate": 1.9688963761592984e-05,
+      "loss": 0.4401,
+      "step": 685
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 0.4795934551092186,
+      "learning_rate": 1.963241759676897e-05,
+      "loss": 0.4483,
+      "step": 686
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 0.489489932556473,
+      "learning_rate": 1.9575874371124066e-05,
+      "loss": 0.4509,
+      "step": 687
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 0.4512565580307455,
+      "learning_rate": 1.951933453677638e-05,
+      "loss": 0.4434,
+      "step": 688
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 0.4449512806006609,
+      "learning_rate": 1.9462798545816932e-05,
+      "loss": 0.4365,
+      "step": 689
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 0.47957987697176724,
+      "learning_rate": 1.9406266850305996e-05,
+      "loss": 0.4334,
+      "step": 690
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 0.42406424956224525,
+      "learning_rate": 1.9349739902269503e-05,
+      "loss": 0.44,
+      "step": 691
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 0.4643349539854223,
+      "learning_rate": 1.9293218153695425e-05,
+      "loss": 0.4374,
+      "step": 692
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 0.47272563286366076,
+      "learning_rate": 1.9236702056530153e-05,
+      "loss": 0.4568,
+      "step": 693
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 0.4286444675331695,
+      "learning_rate": 1.9180192062674907e-05,
+      "loss": 0.4387,
+      "step": 694
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 0.45714908128076237,
+      "learning_rate": 1.9123688623982076e-05,
+      "loss": 0.4527,
+      "step": 695
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 0.4337394199918063,
+      "learning_rate": 1.906719219225165e-05,
+      "loss": 0.4399,
+      "step": 696
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 0.45463230697714563,
+      "learning_rate": 1.9010703219227603e-05,
+      "loss": 0.4497,
+      "step": 697
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 0.3975969394733958,
+      "learning_rate": 1.8954222156594256e-05,
+      "loss": 0.4595,
+      "step": 698
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 0.4911802378848793,
+      "learning_rate": 1.8897749455972673e-05,
+      "loss": 0.4474,
+      "step": 699
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 0.40714516994918937,
+      "learning_rate": 1.884128556891707e-05,
+      "loss": 0.4496,
+      "step": 700
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 0.4549467086936069,
+      "learning_rate": 1.878483094691119e-05,
+      "loss": 0.4462,
+      "step": 701
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 0.4034551403932841,
+      "learning_rate": 1.8728386041364664e-05,
+      "loss": 0.4439,
+      "step": 702
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.4107310671763944,
+      "learning_rate": 1.867195130360947e-05,
+      "loss": 0.4569,
+      "step": 703
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 0.42189010441162617,
+      "learning_rate": 1.8615527184896263e-05,
+      "loss": 0.4389,
+      "step": 704
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 0.3910122769812341,
+      "learning_rate": 1.8559114136390795e-05,
+      "loss": 0.4362,
+      "step": 705
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 0.46022341597159677,
+      "learning_rate": 1.8502712609170298e-05,
+      "loss": 0.4426,
+      "step": 706
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 0.431500123066856,
+      "learning_rate": 1.8446323054219876e-05,
+      "loss": 0.4473,
+      "step": 707
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 0.4838860771700523,
+      "learning_rate": 1.838994592242891e-05,
+      "loss": 0.4534,
+      "step": 708
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 0.4387576010915568,
+      "learning_rate": 1.8333581664587453e-05,
+      "loss": 0.4482,
+      "step": 709
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 0.48912766985415823,
+      "learning_rate": 1.8277230731382593e-05,
+      "loss": 0.443,
+      "step": 710
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 0.5075499423343519,
+      "learning_rate": 1.82208935733949e-05,
+      "loss": 0.4344,
+      "step": 711
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 0.4796373411805534,
+      "learning_rate": 1.8164570641094793e-05,
+      "loss": 0.4459,
+      "step": 712
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 0.4879127690607013,
+      "learning_rate": 1.810826238483892e-05,
+      "loss": 0.4393,
+      "step": 713
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 0.4389758652096589,
+      "learning_rate": 1.8051969254866608e-05,
+      "loss": 0.4496,
+      "step": 714
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 0.4726804151677781,
+      "learning_rate": 1.7995691701296226e-05,
+      "loss": 0.4423,
+      "step": 715
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 0.41155911664347733,
+      "learning_rate": 1.7939430174121593e-05,
+      "loss": 0.4422,
+      "step": 716
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 0.4949062156114871,
+      "learning_rate": 1.788318512320838e-05,
+      "loss": 0.4465,
+      "step": 717
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 0.42298285164911925,
+      "learning_rate": 1.782695699829051e-05,
+      "loss": 0.4493,
+      "step": 718
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 0.5184521817661292,
+      "learning_rate": 1.7770746248966587e-05,
+      "loss": 0.4358,
+      "step": 719
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 0.46867427176561255,
+      "learning_rate": 1.7714553324696243e-05,
+      "loss": 0.4433,
+      "step": 720
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 0.47519556258554396,
+      "learning_rate": 1.7658378674796614e-05,
+      "loss": 0.4597,
+      "step": 721
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.5617449765875315,
+      "learning_rate": 1.76022227484387e-05,
+      "loss": 0.4504,
+      "step": 722
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 0.41170880112800484,
+      "learning_rate": 1.7546085994643807e-05,
+      "loss": 0.4497,
+      "step": 723
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 0.4633721889708841,
+      "learning_rate": 1.7489968862279902e-05,
+      "loss": 0.4423,
+      "step": 724
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 0.46198932719965385,
+      "learning_rate": 1.743387180005809e-05,
+      "loss": 0.4478,
+      "step": 725
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 0.42847063645000055,
+      "learning_rate": 1.737779525652899e-05,
+      "loss": 0.4467,
+      "step": 726
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 0.44742676107104995,
+      "learning_rate": 1.732173968007914e-05,
+      "loss": 0.4606,
+      "step": 727
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 0.464163929777104,
+      "learning_rate": 1.7265705518927437e-05,
+      "loss": 0.4444,
+      "step": 728
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 0.3964385434174572,
+      "learning_rate": 1.7209693221121542e-05,
+      "loss": 0.4404,
+      "step": 729
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 0.47931311477145533,
+      "learning_rate": 1.7153703234534302e-05,
+      "loss": 0.4372,
+      "step": 730
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 0.411205778750304,
+      "learning_rate": 1.709773600686015e-05,
+      "loss": 0.4496,
+      "step": 731
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 0.42521553555024094,
+      "learning_rate": 1.7041791985611548e-05,
+      "loss": 0.4463,
+      "step": 732
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 0.40840176525442845,
+      "learning_rate": 1.69858716181154e-05,
+      "loss": 0.443,
+      "step": 733
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 0.43552905185271573,
+      "learning_rate": 1.692997535150948e-05,
+      "loss": 0.4589,
+      "step": 734
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 0.40932706669559815,
+      "learning_rate": 1.687410363273884e-05,
+      "loss": 0.4421,
+      "step": 735
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 0.4628467450061272,
+      "learning_rate": 1.6818256908552257e-05,
+      "loss": 0.4549,
+      "step": 736
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 0.4218821388007794,
+      "learning_rate": 1.6762435625498646e-05,
+      "loss": 0.4464,
+      "step": 737
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 0.4244260629263333,
+      "learning_rate": 1.67066402299235e-05,
+      "loss": 0.4526,
+      "step": 738
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 0.5248554986889205,
+      "learning_rate": 1.6650871167965313e-05,
+      "loss": 0.4475,
+      "step": 739
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 0.4031283893995207,
+      "learning_rate": 1.6595128885552028e-05,
+      "loss": 0.4402,
+      "step": 740
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.44075705628008954,
+      "learning_rate": 1.653941382839745e-05,
+      "loss": 0.4205,
+      "step": 741
+    },
+    {
+      "epoch": 3.0040485829959516,
+      "grad_norm": 0.8447947892012599,
+      "learning_rate": 1.6483726441997673e-05,
+      "loss": 0.3241,
+      "step": 742
+    },
+    {
+      "epoch": 3.0080971659919027,
+      "grad_norm": 0.5372009204809357,
+      "learning_rate": 1.642806717162757e-05,
+      "loss": 0.3287,
+      "step": 743
+    },
+    {
+      "epoch": 3.0121457489878543,
+      "grad_norm": 1.0026075931670777,
+      "learning_rate": 1.637243646233718e-05,
+      "loss": 0.3184,
+      "step": 744
+    },
+    {
+      "epoch": 3.016194331983806,
+      "grad_norm": 0.6348835680054623,
+      "learning_rate": 1.6316834758948174e-05,
+      "loss": 0.3276,
+      "step": 745
+    },
+    {
+      "epoch": 3.020242914979757,
+      "grad_norm": 0.6543327998094357,
+      "learning_rate": 1.6261262506050282e-05,
+      "loss": 0.3223,
+      "step": 746
+    },
+    {
+      "epoch": 3.0242914979757085,
+      "grad_norm": 0.6507007839128218,
+      "learning_rate": 1.620572014799777e-05,
+      "loss": 0.3266,
+      "step": 747
+    },
+    {
+      "epoch": 3.02834008097166,
+      "grad_norm": 0.5902129083064382,
+      "learning_rate": 1.6150208128905857e-05,
+      "loss": 0.3172,
+      "step": 748
+    },
+    {
+      "epoch": 3.032388663967611,
+      "grad_norm": 0.529819877230645,
+      "learning_rate": 1.6094726892647147e-05,
+      "loss": 0.3138,
+      "step": 749
+    },
+    {
+      "epoch": 3.0364372469635628,
+      "grad_norm": 0.6130183153724829,
+      "learning_rate": 1.6039276882848135e-05,
+      "loss": 0.324,
+      "step": 750
+    },
+    {
+      "epoch": 3.0404858299595143,
+      "grad_norm": 0.5491884015461056,
+      "learning_rate": 1.5983858542885635e-05,
+      "loss": 0.3229,
+      "step": 751
+    },
+    {
+      "epoch": 3.0445344129554655,
+      "grad_norm": 0.5563866597727088,
+      "learning_rate": 1.5928472315883203e-05,
+      "loss": 0.3196,
+      "step": 752
+    },
+    {
+      "epoch": 3.048582995951417,
+      "grad_norm": 0.49783454188565557,
+      "learning_rate": 1.5873118644707633e-05,
+      "loss": 0.3171,
+      "step": 753
+    },
+    {
+      "epoch": 3.0526315789473686,
+      "grad_norm": 0.5359222080122537,
+      "learning_rate": 1.5817797971965413e-05,
+      "loss": 0.3193,
+      "step": 754
+    },
+    {
+      "epoch": 3.0566801619433197,
+      "grad_norm": 0.5075140553905693,
+      "learning_rate": 1.576251073999917e-05,
+      "loss": 0.3209,
+      "step": 755
+    },
+    {
+      "epoch": 3.0607287449392713,
+      "grad_norm": 0.4688089047021019,
+      "learning_rate": 1.5707257390884126e-05,
+      "loss": 0.3232,
+      "step": 756
+    },
+    {
+      "epoch": 3.064777327935223,
+      "grad_norm": 0.49205219983147586,
+      "learning_rate": 1.5652038366424595e-05,
+      "loss": 0.292,
+      "step": 757
+    },
+    {
+      "epoch": 3.068825910931174,
+      "grad_norm": 0.5223677610652214,
+      "learning_rate": 1.5596854108150424e-05,
+      "loss": 0.3219,
+      "step": 758
+    },
+    {
+      "epoch": 3.0728744939271255,
+      "grad_norm": 0.47635408529070444,
+      "learning_rate": 1.5541705057313476e-05,
+      "loss": 0.3083,
+      "step": 759
+    },
+    {
+      "epoch": 3.076923076923077,
+      "grad_norm": 0.46992693392801005,
+      "learning_rate": 1.5486591654884086e-05,
+      "loss": 0.3199,
+      "step": 760
+    },
+    {
+      "epoch": 3.080971659919028,
+      "grad_norm": 0.46771871682840727,
+      "learning_rate": 1.543151434154755e-05,
+      "loss": 0.3117,
+      "step": 761
+    },
+    {
+      "epoch": 3.08502024291498,
+      "grad_norm": 0.48428079278287645,
+      "learning_rate": 1.5376473557700615e-05,
+      "loss": 0.312,
+      "step": 762
+    },
+    {
+      "epoch": 3.0890688259109313,
+      "grad_norm": 0.48507986786870233,
+      "learning_rate": 1.5321469743447893e-05,
+      "loss": 0.3233,
+      "step": 763
+    },
+    {
+      "epoch": 3.0931174089068825,
+      "grad_norm": 0.4739172670798252,
+      "learning_rate": 1.5266503338598434e-05,
+      "loss": 0.3177,
+      "step": 764
+    },
+    {
+      "epoch": 3.097165991902834,
+      "grad_norm": 0.5193723721490996,
+      "learning_rate": 1.5211574782662149e-05,
+      "loss": 0.3213,
+      "step": 765
+    },
+    {
+      "epoch": 3.1012145748987856,
+      "grad_norm": 0.4608866400402393,
+      "learning_rate": 1.5156684514846307e-05,
+      "loss": 0.3193,
+      "step": 766
+    },
+    {
+      "epoch": 3.1052631578947367,
+      "grad_norm": 0.4546014949142791,
+      "learning_rate": 1.5101832974052012e-05,
+      "loss": 0.3123,
+      "step": 767
+    },
+    {
+      "epoch": 3.1093117408906883,
+      "grad_norm": 0.482079428432936,
+      "learning_rate": 1.5047020598870737e-05,
+      "loss": 0.3119,
+      "step": 768
+    },
+    {
+      "epoch": 3.11336032388664,
+      "grad_norm": 0.4534679960479946,
+      "learning_rate": 1.4992247827580778e-05,
+      "loss": 0.3219,
+      "step": 769
+    },
+    {
+      "epoch": 3.117408906882591,
+      "grad_norm": 0.47529270419527425,
+      "learning_rate": 1.493751509814374e-05,
+      "loss": 0.3207,
+      "step": 770
+    },
+    {
+      "epoch": 3.1214574898785425,
+      "grad_norm": 0.4228781111411663,
+      "learning_rate": 1.4882822848201073e-05,
+      "loss": 0.3208,
+      "step": 771
+    },
+    {
+      "epoch": 3.125506072874494,
+      "grad_norm": 0.44857101734286187,
+      "learning_rate": 1.4828171515070553e-05,
+      "loss": 0.3167,
+      "step": 772
+    },
+    {
+      "epoch": 3.1295546558704452,
+      "grad_norm": 0.41979179817527923,
+      "learning_rate": 1.4773561535742793e-05,
+      "loss": 0.3239,
+      "step": 773
+    },
+    {
+      "epoch": 3.133603238866397,
+      "grad_norm": 0.4158682998294857,
+      "learning_rate": 1.4718993346877715e-05,
+      "loss": 0.3074,
+      "step": 774
+    },
+    {
+      "epoch": 3.1376518218623484,
+      "grad_norm": 0.4376798293906905,
+      "learning_rate": 1.466446738480111e-05,
+      "loss": 0.3062,
+      "step": 775
+    },
+    {
+      "epoch": 3.1417004048582995,
+      "grad_norm": 0.41132459681107714,
+      "learning_rate": 1.460998408550112e-05,
+      "loss": 0.3151,
+      "step": 776
+    },
+    {
+      "epoch": 3.145748987854251,
+      "grad_norm": 0.4189277556428406,
+      "learning_rate": 1.4555543884624751e-05,
+      "loss": 0.313,
+      "step": 777
+    },
+    {
+      "epoch": 3.1497975708502026,
+      "grad_norm": 0.4279889813921973,
+      "learning_rate": 1.4501147217474402e-05,
+      "loss": 0.3175,
+      "step": 778
+    },
+    {
+      "epoch": 3.1538461538461537,
+      "grad_norm": 0.4255430172472349,
+      "learning_rate": 1.444679451900437e-05,
+      "loss": 0.3233,
+      "step": 779
+    },
+    {
+      "epoch": 3.1578947368421053,
+      "grad_norm": 0.4307403833715584,
+      "learning_rate": 1.4392486223817397e-05,
+      "loss": 0.3202,
+      "step": 780
+    },
+    {
+      "epoch": 3.161943319838057,
+      "grad_norm": 0.4120924021236646,
+      "learning_rate": 1.4338222766161143e-05,
+      "loss": 0.3106,
+      "step": 781
+    },
+    {
+      "epoch": 3.165991902834008,
+      "grad_norm": 0.42934095727925825,
+      "learning_rate": 1.4284004579924777e-05,
+      "loss": 0.3368,
+      "step": 782
+    },
+    {
+      "epoch": 3.1700404858299596,
+      "grad_norm": 0.40436276243451563,
+      "learning_rate": 1.4229832098635479e-05,
+      "loss": 0.3084,
+      "step": 783
+    },
+    {
+      "epoch": 3.174089068825911,
+      "grad_norm": 0.43567845773186126,
+      "learning_rate": 1.4175705755454963e-05,
+      "loss": 0.316,
+      "step": 784
+    },
+    {
+      "epoch": 3.1781376518218623,
+      "grad_norm": 0.44953305592569115,
+      "learning_rate": 1.412162598317602e-05,
+      "loss": 0.322,
+      "step": 785
+    },
+    {
+      "epoch": 3.182186234817814,
+      "grad_norm": 0.438050060188759,
+      "learning_rate": 1.406759321421907e-05,
+      "loss": 0.3298,
+      "step": 786
+    },
+    {
+      "epoch": 3.1862348178137654,
+      "grad_norm": 0.45107599367469275,
+      "learning_rate": 1.401360788062871e-05,
+      "loss": 0.3147,
+      "step": 787
+    },
+    {
+      "epoch": 3.1902834008097165,
+      "grad_norm": 0.44019277506275495,
+      "learning_rate": 1.3959670414070204e-05,
+      "loss": 0.3226,
+      "step": 788
+    },
+    {
+      "epoch": 3.194331983805668,
+      "grad_norm": 0.4695874889650123,
+      "learning_rate": 1.3905781245826108e-05,
+      "loss": 0.3155,
+      "step": 789
+    },
+    {
+      "epoch": 3.1983805668016196,
+      "grad_norm": 0.4221513739603117,
+      "learning_rate": 1.3851940806792778e-05,
+      "loss": 0.3272,
+      "step": 790
+    },
+    {
+      "epoch": 3.2024291497975708,
+      "grad_norm": 0.42724640389578744,
+      "learning_rate": 1.379814952747693e-05,
+      "loss": 0.3201,
+      "step": 791
+    },
+    {
+      "epoch": 3.2064777327935223,
+      "grad_norm": 0.4484247414694738,
+      "learning_rate": 1.3744407837992193e-05,
+      "loss": 0.3123,
+      "step": 792
+    },
+    {
+      "epoch": 3.2105263157894735,
+      "grad_norm": 0.40731343163017864,
+      "learning_rate": 1.3690716168055692e-05,
+      "loss": 0.3219,
+      "step": 793
+    },
+    {
+      "epoch": 3.214574898785425,
+      "grad_norm": 0.4533165673709925,
+      "learning_rate": 1.363707494698459e-05,
+      "loss": 0.331,
+      "step": 794
+    },
+    {
+      "epoch": 3.2186234817813766,
+      "grad_norm": 0.41909380741811675,
+      "learning_rate": 1.358348460369265e-05,
+      "loss": 0.3221,
+      "step": 795
+    },
+    {
+      "epoch": 3.2226720647773277,
+      "grad_norm": 0.4381157100819422,
+      "learning_rate": 1.352994556668684e-05,
+      "loss": 0.3102,
+      "step": 796
+    },
+    {
+      "epoch": 3.2267206477732793,
+      "grad_norm": 0.4044632644628445,
+      "learning_rate": 1.3476458264063865e-05,
+      "loss": 0.2999,
+      "step": 797
+    },
+    {
+      "epoch": 3.230769230769231,
+      "grad_norm": 0.4609963402955929,
+      "learning_rate": 1.342302312350678e-05,
+      "loss": 0.3161,
+      "step": 798
+    },
+    {
+      "epoch": 3.234817813765182,
+      "grad_norm": 0.43348547565917134,
+      "learning_rate": 1.3369640572281537e-05,
+      "loss": 0.3284,
+      "step": 799
+    },
+    {
+      "epoch": 3.2388663967611335,
+      "grad_norm": 0.44666294274195595,
+      "learning_rate": 1.3316311037233596e-05,
+      "loss": 0.3141,
+      "step": 800
+    },
+    {
+      "epoch": 3.242914979757085,
+      "grad_norm": 0.41240559210024047,
+      "learning_rate": 1.326303494478451e-05,
+      "loss": 0.325,
+      "step": 801
+    },
+    {
+      "epoch": 3.246963562753036,
+      "grad_norm": 0.44038280416867553,
+      "learning_rate": 1.3209812720928472e-05,
+      "loss": 0.3136,
+      "step": 802
+    },
+    {
+      "epoch": 3.251012145748988,
+      "grad_norm": 0.4176350807654786,
+      "learning_rate": 1.315664479122898e-05,
+      "loss": 0.3216,
+      "step": 803
+    },
+    {
+      "epoch": 3.2550607287449393,
+      "grad_norm": 0.42274540709901165,
+      "learning_rate": 1.3103531580815378e-05,
+      "loss": 0.3375,
+      "step": 804
+    },
+    {
+      "epoch": 3.2591093117408905,
+      "grad_norm": 0.4338505176099702,
+      "learning_rate": 1.3050473514379488e-05,
+      "loss": 0.3192,
+      "step": 805
+    },
+    {
+      "epoch": 3.263157894736842,
+      "grad_norm": 0.4304768693997427,
+      "learning_rate": 1.2997471016172188e-05,
+      "loss": 0.3114,
+      "step": 806
+    },
+    {
+      "epoch": 3.2672064777327936,
+      "grad_norm": 0.4655963770681824,
+      "learning_rate": 1.2944524510000042e-05,
+      "loss": 0.3162,
+      "step": 807
+    },
+    {
+      "epoch": 3.2712550607287447,
+      "grad_norm": 0.42969268820287215,
+      "learning_rate": 1.2891634419221908e-05,
+      "loss": 0.3038,
+      "step": 808
+    },
+    {
+      "epoch": 3.2753036437246963,
+      "grad_norm": 0.4327761607397563,
+      "learning_rate": 1.2838801166745545e-05,
+      "loss": 0.3183,
+      "step": 809
+    },
+    {
+      "epoch": 3.279352226720648,
+      "grad_norm": 0.41573433607268284,
+      "learning_rate": 1.2786025175024216e-05,
+      "loss": 0.314,
+      "step": 810
+    },
+    {
+      "epoch": 3.283400809716599,
+      "grad_norm": 0.42968102777000833,
+      "learning_rate": 1.2733306866053357e-05,
+      "loss": 0.33,
+      "step": 811
+    },
+    {
+      "epoch": 3.2874493927125505,
+      "grad_norm": 0.39167945307602303,
+      "learning_rate": 1.2680646661367163e-05,
+      "loss": 0.3245,
+      "step": 812
+    },
+    {
+      "epoch": 3.291497975708502,
+      "grad_norm": 0.422048106648319,
+      "learning_rate": 1.2628044982035219e-05,
+      "loss": 0.33,
+      "step": 813
+    },
+    {
+      "epoch": 3.2955465587044532,
+      "grad_norm": 0.4232074825597713,
+      "learning_rate": 1.2575502248659155e-05,
+      "loss": 0.314,
+      "step": 814
+    },
+    {
+      "epoch": 3.299595141700405,
+      "grad_norm": 0.4315408479139215,
+      "learning_rate": 1.2523018881369269e-05,
+      "loss": 0.3174,
+      "step": 815
+    },
+    {
+      "epoch": 3.3036437246963564,
+      "grad_norm": 0.3957575663259354,
+      "learning_rate": 1.2470595299821176e-05,
+      "loss": 0.3292,
+      "step": 816
+    },
+    {
+      "epoch": 3.3076923076923075,
+      "grad_norm": 0.40491582318532154,
+      "learning_rate": 1.2418231923192416e-05,
+      "loss": 0.3204,
+      "step": 817
+    },
+    {
+      "epoch": 3.311740890688259,
+      "grad_norm": 0.40566088999717553,
+      "learning_rate": 1.2365929170179165e-05,
+      "loss": 0.322,
+      "step": 818
+    },
+    {
+      "epoch": 3.3157894736842106,
+      "grad_norm": 0.4679022651741593,
+      "learning_rate": 1.2313687458992848e-05,
+      "loss": 0.3125,
+      "step": 819
+    },
+    {
+      "epoch": 3.3198380566801617,
+      "grad_norm": 0.4083162347627205,
+      "learning_rate": 1.2261507207356785e-05,
+      "loss": 0.3197,
+      "step": 820
+    },
+    {
+      "epoch": 3.3238866396761133,
+      "grad_norm": 0.42946174169848716,
+      "learning_rate": 1.220938883250288e-05,
+      "loss": 0.3171,
+      "step": 821
+    },
+    {
+      "epoch": 3.327935222672065,
+      "grad_norm": 0.40921218749116095,
+      "learning_rate": 1.215733275116827e-05,
+      "loss": 0.3087,
+      "step": 822
+    },
+    {
+      "epoch": 3.331983805668016,
+      "grad_norm": 0.4001462259645613,
+      "learning_rate": 1.2105339379592005e-05,
+      "loss": 0.3168,
+      "step": 823
+    },
+    {
+      "epoch": 3.3360323886639676,
+      "grad_norm": 0.426623170247207,
+      "learning_rate": 1.2053409133511681e-05,
+      "loss": 0.3174,
+      "step": 824
+    },
+    {
+      "epoch": 3.340080971659919,
+      "grad_norm": 0.42442286179670463,
+      "learning_rate": 1.2001542428160173e-05,
+      "loss": 0.308,
+      "step": 825
+    },
+    {
+      "epoch": 3.3441295546558703,
+      "grad_norm": 0.41437459624533396,
+      "learning_rate": 1.1949739678262278e-05,
+      "loss": 0.3117,
+      "step": 826
+    },
+    {
+      "epoch": 3.348178137651822,
+      "grad_norm": 0.4190021082699289,
+      "learning_rate": 1.189800129803139e-05,
+      "loss": 0.3123,
+      "step": 827
+    },
+    {
+      "epoch": 3.3522267206477734,
+      "grad_norm": 0.4096876313003511,
+      "learning_rate": 1.1846327701166228e-05,
+      "loss": 0.3025,
+      "step": 828
+    },
+    {
+      "epoch": 3.3562753036437245,
+      "grad_norm": 0.45148894768343756,
+      "learning_rate": 1.1794719300847493e-05,
+      "loss": 0.3112,
+      "step": 829
+    },
+    {
+      "epoch": 3.360323886639676,
+      "grad_norm": 0.4408411806610938,
+      "learning_rate": 1.1743176509734582e-05,
+      "loss": 0.3163,
+      "step": 830
+    },
+    {
+      "epoch": 3.3643724696356276,
+      "grad_norm": 0.406056775065985,
+      "learning_rate": 1.1691699739962275e-05,
+      "loss": 0.3106,
+      "step": 831
+    },
+    {
+      "epoch": 3.3684210526315788,
+      "grad_norm": 0.3952974155550243,
+      "learning_rate": 1.1640289403137456e-05,
+      "loss": 0.3176,
+      "step": 832
+    },
+    {
+      "epoch": 3.3724696356275303,
+      "grad_norm": 0.4076384524172983,
+      "learning_rate": 1.1588945910335802e-05,
+      "loss": 0.3195,
+      "step": 833
+    },
+    {
+      "epoch": 3.376518218623482,
+      "grad_norm": 0.406126752878918,
+      "learning_rate": 1.1537669672098517e-05,
+      "loss": 0.3183,
+      "step": 834
+    },
+    {
+      "epoch": 3.380566801619433,
+      "grad_norm": 0.41980329580855036,
+      "learning_rate": 1.1486461098429036e-05,
+      "loss": 0.3189,
+      "step": 835
+    },
+    {
+      "epoch": 3.3846153846153846,
+      "grad_norm": 0.388240022733057,
+      "learning_rate": 1.1435320598789747e-05,
+      "loss": 0.3196,
+      "step": 836
+    },
+    {
+      "epoch": 3.388663967611336,
+      "grad_norm": 0.39241224594750124,
+      "learning_rate": 1.1384248582098736e-05,
+      "loss": 0.3161,
+      "step": 837
+    },
+    {
+      "epoch": 3.3927125506072873,
+      "grad_norm": 0.40470176090425797,
+      "learning_rate": 1.1333245456726467e-05,
+      "loss": 0.3151,
+      "step": 838
+    },
+    {
+      "epoch": 3.396761133603239,
+      "grad_norm": 0.39971815709576985,
+      "learning_rate": 1.128231163049258e-05,
+      "loss": 0.3326,
+      "step": 839
+    },
+    {
+      "epoch": 3.4008097165991904,
+      "grad_norm": 0.4242566032865865,
+      "learning_rate": 1.123144751066259e-05,
+      "loss": 0.3105,
+      "step": 840
+    },
+    {
+      "epoch": 3.4048582995951415,
+      "grad_norm": 0.40183320445413007,
+      "learning_rate": 1.1180653503944657e-05,
+      "loss": 0.301,
+      "step": 841
+    },
+    {
+      "epoch": 3.408906882591093,
+      "grad_norm": 0.37152197633310236,
+      "learning_rate": 1.1129930016486287e-05,
+      "loss": 0.3304,
+      "step": 842
+    },
+    {
+      "epoch": 3.4129554655870447,
+      "grad_norm": 0.39761807642930935,
+      "learning_rate": 1.107927745387114e-05,
+      "loss": 0.3233,
+      "step": 843
+    },
+    {
+      "epoch": 3.417004048582996,
+      "grad_norm": 0.42076017710504077,
+      "learning_rate": 1.1028696221115766e-05,
+      "loss": 0.2981,
+      "step": 844
+    },
+    {
+      "epoch": 3.4210526315789473,
+      "grad_norm": 0.399626055206577,
+      "learning_rate": 1.0978186722666341e-05,
+      "loss": 0.3348,
+      "step": 845
+    },
+    {
+      "epoch": 3.425101214574899,
+      "grad_norm": 0.41387225043422265,
+      "learning_rate": 1.0927749362395475e-05,
+      "loss": 0.3106,
+      "step": 846
+    },
+    {
+      "epoch": 3.42914979757085,
+      "grad_norm": 0.4184568914486274,
+      "learning_rate": 1.0877384543598954e-05,
+      "loss": 0.3115,
+      "step": 847
+    },
+    {
+      "epoch": 3.4331983805668016,
+      "grad_norm": 0.3996512590396983,
+      "learning_rate": 1.0827092668992529e-05,
+      "loss": 0.3148,
+      "step": 848
+    },
+    {
+      "epoch": 3.437246963562753,
+      "grad_norm": 0.4362122628766555,
+      "learning_rate": 1.077687414070869e-05,
+      "loss": 0.3196,
+      "step": 849
+    },
+    {
+      "epoch": 3.4412955465587043,
+      "grad_norm": 0.3986208847714447,
+      "learning_rate": 1.0726729360293451e-05,
+      "loss": 0.3085,
+      "step": 850
+    },
+    {
+      "epoch": 3.445344129554656,
+      "grad_norm": 0.3980480541021148,
+      "learning_rate": 1.0676658728703147e-05,
+      "loss": 0.3377,
+      "step": 851
+    },
+    {
+      "epoch": 3.4493927125506074,
+      "grad_norm": 0.426923244222778,
+      "learning_rate": 1.0626662646301198e-05,
+      "loss": 0.3171,
+      "step": 852
+    },
+    {
+      "epoch": 3.4534412955465585,
+      "grad_norm": 0.3989438758060905,
+      "learning_rate": 1.0576741512854958e-05,
+      "loss": 0.3278,
+      "step": 853
+    },
+    {
+      "epoch": 3.45748987854251,
+      "grad_norm": 0.3960252411335677,
+      "learning_rate": 1.052689572753247e-05,
+      "loss": 0.3149,
+      "step": 854
+    },
+    {
+      "epoch": 3.4615384615384617,
+      "grad_norm": 0.4027693361208235,
+      "learning_rate": 1.0477125688899328e-05,
+      "loss": 0.3133,
+      "step": 855
+    },
+    {
+      "epoch": 3.465587044534413,
+      "grad_norm": 0.4007701884545717,
+      "learning_rate": 1.0427431794915406e-05,
+      "loss": 0.3145,
+      "step": 856
+    },
+    {
+      "epoch": 3.4696356275303644,
+      "grad_norm": 0.40687412608038387,
+      "learning_rate": 1.0377814442931769e-05,
+      "loss": 0.3229,
+      "step": 857
+    },
+    {
+      "epoch": 3.473684210526316,
+      "grad_norm": 0.40049308290510216,
+      "learning_rate": 1.0328274029687439e-05,
+      "loss": 0.3144,
+      "step": 858
+    },
+    {
+      "epoch": 3.477732793522267,
+      "grad_norm": 0.41663337832337183,
+      "learning_rate": 1.0278810951306253e-05,
+      "loss": 0.3246,
+      "step": 859
+    },
+    {
+      "epoch": 3.4817813765182186,
+      "grad_norm": 0.4225990822087912,
+      "learning_rate": 1.0229425603293647e-05,
+      "loss": 0.2974,
+      "step": 860
+    },
+    {
+      "epoch": 3.48582995951417,
+      "grad_norm": 0.424137773018942,
+      "learning_rate": 1.0180118380533558e-05,
+      "loss": 0.3159,
+      "step": 861
+    },
+    {
+      "epoch": 3.4898785425101213,
+      "grad_norm": 0.388029318944334,
+      "learning_rate": 1.0130889677285225e-05,
+      "loss": 0.3199,
+      "step": 862
+    },
+    {
+      "epoch": 3.493927125506073,
+      "grad_norm": 0.4146010509042746,
+      "learning_rate": 1.008173988718005e-05,
+      "loss": 0.3097,
+      "step": 863
+    },
+    {
+      "epoch": 3.4979757085020244,
+      "grad_norm": 0.410214548247326,
+      "learning_rate": 1.0032669403218439e-05,
+      "loss": 0.2941,
+      "step": 864
+    },
+    {
+      "epoch": 3.5020242914979756,
+      "grad_norm": 0.3999175402990461,
+      "learning_rate": 9.983678617766677e-06,
+      "loss": 0.3173,
+      "step": 865
+    },
+    {
+      "epoch": 3.506072874493927,
+      "grad_norm": 0.40768412428837963,
+      "learning_rate": 9.934767922553788e-06,
+      "loss": 0.3308,
+      "step": 866
+    },
+    {
+      "epoch": 3.5101214574898787,
+      "grad_norm": 0.38690066851084665,
+      "learning_rate": 9.885937708668371e-06,
+      "loss": 0.3249,
+      "step": 867
+    },
+    {
+      "epoch": 3.51417004048583,
+      "grad_norm": 0.3830985080151568,
+      "learning_rate": 9.837188366555524e-06,
+      "loss": 0.3099,
+      "step": 868
+    },
+    {
+      "epoch": 3.5182186234817814,
+      "grad_norm": 0.4100218641632111,
+      "learning_rate": 9.788520286013691e-06,
+      "loss": 0.3331,
+      "step": 869
+    },
+    {
+      "epoch": 3.522267206477733,
+      "grad_norm": 0.4088163511154281,
+      "learning_rate": 9.739933856191537e-06,
+      "loss": 0.3027,
+      "step": 870
+    },
+    {
+      "epoch": 3.526315789473684,
+      "grad_norm": 0.39184947295676614,
+      "learning_rate": 9.691429465584866e-06,
+      "loss": 0.3103,
+      "step": 871
+    },
+    {
+      "epoch": 3.5303643724696356,
+      "grad_norm": 0.42986253243669387,
+      "learning_rate": 9.643007502033492e-06,
+      "loss": 0.3022,
+      "step": 872
+    },
+    {
+      "epoch": 3.534412955465587,
+      "grad_norm": 0.4041788474994215,
+      "learning_rate": 9.594668352718152e-06,
+      "loss": 0.31,
+      "step": 873
+    },
+    {
+      "epoch": 3.5384615384615383,
+      "grad_norm": 0.41155684940217174,
+      "learning_rate": 9.546412404157383e-06,
+      "loss": 0.3307,
+      "step": 874
+    },
+    {
+      "epoch": 3.54251012145749,
+      "grad_norm": 0.41511608686547835,
+      "learning_rate": 9.498240042204471e-06,
+      "loss": 0.3158,
+      "step": 875
+    },
+    {
+      "epoch": 3.5465587044534415,
+      "grad_norm": 0.4312491597771405,
+      "learning_rate": 9.450151652044339e-06,
+      "loss": 0.3066,
+      "step": 876
+    },
+    {
+      "epoch": 3.5506072874493926,
+      "grad_norm": 0.4305210556676289,
+      "learning_rate": 9.402147618190472e-06,
+      "loss": 0.3177,
+      "step": 877
+    },
+    {
+      "epoch": 3.554655870445344,
+      "grad_norm": 0.40142189894378993,
+      "learning_rate": 9.354228324481841e-06,
+      "loss": 0.3335,
+      "step": 878
+    },
+    {
+      "epoch": 3.5587044534412957,
+      "grad_norm": 0.41237898523309086,
+      "learning_rate": 9.306394154079852e-06,
+      "loss": 0.3109,
+      "step": 879
+    },
+    {
+      "epoch": 3.562753036437247,
+      "grad_norm": 0.41426137566830445,
+      "learning_rate": 9.258645489465254e-06,
+      "loss": 0.3013,
+      "step": 880
+    },
+    {
+      "epoch": 3.5668016194331984,
+      "grad_norm": 0.4044605956920581,
+      "learning_rate": 9.21098271243509e-06,
+      "loss": 0.3045,
+      "step": 881
+    },
+    {
+      "epoch": 3.57085020242915,
+      "grad_norm": 0.44146965014857853,
+      "learning_rate": 9.163406204099659e-06,
+      "loss": 0.3211,
+      "step": 882
+    },
+    {
+      "epoch": 3.574898785425101,
+      "grad_norm": 0.4078262884020033,
+      "learning_rate": 9.115916344879452e-06,
+      "loss": 0.3215,
+      "step": 883
+    },
+    {
+      "epoch": 3.5789473684210527,
+      "grad_norm": 0.480379766015715,
+      "learning_rate": 9.068513514502126e-06,
+      "loss": 0.3136,
+      "step": 884
+    },
+    {
+      "epoch": 3.582995951417004,
+      "grad_norm": 0.3960114941235987,
+      "learning_rate": 9.021198091999439e-06,
+      "loss": 0.3044,
+      "step": 885
+    },
+    {
+      "epoch": 3.5870445344129553,
+      "grad_norm": 0.4107451140908343,
+      "learning_rate": 8.973970455704251e-06,
+      "loss": 0.2953,
+      "step": 886
+    },
+    {
+      "epoch": 3.591093117408907,
+      "grad_norm": 0.4128237735223163,
+      "learning_rate": 8.926830983247495e-06,
+      "loss": 0.3041,
+      "step": 887
+    },
+    {
+      "epoch": 3.5951417004048585,
+      "grad_norm": 0.4054155063294461,
+      "learning_rate": 8.879780051555116e-06,
+      "loss": 0.3125,
+      "step": 888
+    },
+    {
+      "epoch": 3.5991902834008096,
+      "grad_norm": 0.39312185635573593,
+      "learning_rate": 8.83281803684512e-06,
+      "loss": 0.3138,
+      "step": 889
+    },
+    {
+      "epoch": 3.603238866396761,
+      "grad_norm": 0.42248585629666513,
+      "learning_rate": 8.785945314624532e-06,
+      "loss": 0.2977,
+      "step": 890
+    },
+    {
+      "epoch": 3.6072874493927127,
+      "grad_norm": 0.42226909852974537,
+      "learning_rate": 8.739162259686385e-06,
+      "loss": 0.303,
+      "step": 891
+    },
+    {
+      "epoch": 3.611336032388664,
+      "grad_norm": 0.38980823691977895,
+      "learning_rate": 8.692469246106734e-06,
+      "loss": 0.3153,
+      "step": 892
+    },
+    {
+      "epoch": 3.6153846153846154,
+      "grad_norm": 0.3954863134130458,
+      "learning_rate": 8.645866647241677e-06,
+      "loss": 0.3043,
+      "step": 893
+    },
+    {
+      "epoch": 3.619433198380567,
+      "grad_norm": 0.43118301989971436,
+      "learning_rate": 8.599354835724348e-06,
+      "loss": 0.3235,
+      "step": 894
+    },
+    {
+      "epoch": 3.623481781376518,
+      "grad_norm": 0.4107398770154618,
+      "learning_rate": 8.552934183461958e-06,
+      "loss": 0.317,
+      "step": 895
+    },
+    {
+      "epoch": 3.6275303643724697,
+      "grad_norm": 0.4279116079822966,
+      "learning_rate": 8.506605061632805e-06,
+      "loss": 0.311,
+      "step": 896
+    },
+    {
+      "epoch": 3.6315789473684212,
+      "grad_norm": 0.40761415001854806,
+      "learning_rate": 8.460367840683308e-06,
+      "loss": 0.3038,
+      "step": 897
+    },
+    {
+      "epoch": 3.6356275303643724,
+      "grad_norm": 0.40611020971600315,
+      "learning_rate": 8.414222890325065e-06,
+      "loss": 0.3313,
+      "step": 898
+    },
+    {
+      "epoch": 3.639676113360324,
+      "grad_norm": 0.40471946183793683,
+      "learning_rate": 8.368170579531855e-06,
+      "loss": 0.3108,
+      "step": 899
+    },
+    {
+      "epoch": 3.6437246963562755,
+      "grad_norm": 0.41496089547223947,
+      "learning_rate": 8.322211276536733e-06,
+      "loss": 0.3068,
+      "step": 900
+    },
+    {
+      "epoch": 3.6477732793522266,
+      "grad_norm": 0.42158123266844644,
+      "learning_rate": 8.27634534882907e-06,
+      "loss": 0.3176,
+      "step": 901
+    },
+    {
+      "epoch": 3.651821862348178,
+      "grad_norm": 0.4101617462602249,
+      "learning_rate": 8.230573163151585e-06,
+      "loss": 0.3133,
+      "step": 902
+    },
+    {
+      "epoch": 3.6558704453441297,
+      "grad_norm": 0.42169953531083165,
+      "learning_rate": 8.184895085497466e-06,
+      "loss": 0.3257,
+      "step": 903
+    },
+    {
+      "epoch": 3.659919028340081,
+      "grad_norm": 0.4101721821550569,
+      "learning_rate": 8.139311481107401e-06,
+      "loss": 0.3143,
+      "step": 904
+    },
+    {
+      "epoch": 3.6639676113360324,
+      "grad_norm": 0.39429189289313066,
+      "learning_rate": 8.093822714466686e-06,
+      "loss": 0.3048,
+      "step": 905
+    },
+    {
+      "epoch": 3.668016194331984,
+      "grad_norm": 0.40039909211322405,
+      "learning_rate": 8.048429149302271e-06,
+      "loss": 0.3241,
+      "step": 906
+    },
+    {
+      "epoch": 3.672064777327935,
+      "grad_norm": 0.40039752847221377,
+      "learning_rate": 8.003131148579902e-06,
+      "loss": 0.3056,
+      "step": 907
+    },
+    {
+      "epoch": 3.6761133603238867,
+      "grad_norm": 0.42325139589895827,
+      "learning_rate": 7.957929074501182e-06,
+      "loss": 0.3222,
+      "step": 908
+    },
+    {
+      "epoch": 3.6801619433198383,
+      "grad_norm": 0.3937807444897557,
+      "learning_rate": 7.912823288500695e-06,
+      "loss": 0.3074,
+      "step": 909
+    },
+    {
+      "epoch": 3.6842105263157894,
+      "grad_norm": 0.40879514797261124,
+      "learning_rate": 7.867814151243102e-06,
+      "loss": 0.2993,
+      "step": 910
+    },
+    {
+      "epoch": 3.688259109311741,
+      "grad_norm": 0.4315854203673398,
+      "learning_rate": 7.822902022620267e-06,
+      "loss": 0.3143,
+      "step": 911
+    },
+    {
+      "epoch": 3.6923076923076925,
+      "grad_norm": 0.4357258456383694,
+      "learning_rate": 7.778087261748377e-06,
+      "loss": 0.307,
+      "step": 912
+    },
+    {
+      "epoch": 3.6963562753036436,
+      "grad_norm": 0.41053272764111715,
+      "learning_rate": 7.733370226965056e-06,
+      "loss": 0.315,
+      "step": 913
+    },
+    {
+      "epoch": 3.700404858299595,
+      "grad_norm": 0.41966042904286244,
+      "learning_rate": 7.688751275826527e-06,
+      "loss": 0.3107,
+      "step": 914
+    },
+    {
+      "epoch": 3.7044534412955468,
+      "grad_norm": 0.4006218200021428,
+      "learning_rate": 7.644230765104735e-06,
+      "loss": 0.3194,
+      "step": 915
+    },
+    {
+      "epoch": 3.708502024291498,
+      "grad_norm": 0.4300305093205582,
+      "learning_rate": 7.599809050784502e-06,
+      "loss": 0.3163,
+      "step": 916
+    },
+    {
+      "epoch": 3.7125506072874495,
+      "grad_norm": 0.4089890290216391,
+      "learning_rate": 7.555486488060661e-06,
+      "loss": 0.3174,
+      "step": 917
+    },
+    {
+      "epoch": 3.716599190283401,
+      "grad_norm": 0.47084530938202035,
+      "learning_rate": 7.511263431335249e-06,
+      "loss": 0.3139,
+      "step": 918
+    },
+    {
+      "epoch": 3.720647773279352,
+      "grad_norm": 0.41756084486147005,
+      "learning_rate": 7.467140234214656e-06,
+      "loss": 0.3187,
+      "step": 919
+    },
+    {
+      "epoch": 3.7246963562753037,
+      "grad_norm": 0.427706610933873,
+      "learning_rate": 7.423117249506777e-06,
+      "loss": 0.3158,
+      "step": 920
+    },
+    {
+      "epoch": 3.7287449392712553,
+      "grad_norm": 0.4074059596014088,
+      "learning_rate": 7.379194829218231e-06,
+      "loss": 0.3191,
+      "step": 921
+    },
+    {
+      "epoch": 3.7327935222672064,
+      "grad_norm": 0.39795032030080923,
+      "learning_rate": 7.335373324551518e-06,
+      "loss": 0.3105,
+      "step": 922
+    },
+    {
+      "epoch": 3.736842105263158,
+      "grad_norm": 0.40110129054228233,
+      "learning_rate": 7.291653085902224e-06,
+      "loss": 0.3134,
+      "step": 923
+    },
+    {
+      "epoch": 3.7408906882591095,
+      "grad_norm": 0.43960375408919766,
+      "learning_rate": 7.248034462856213e-06,
+      "loss": 0.3236,
+      "step": 924
+    },
+    {
+      "epoch": 3.7449392712550607,
+      "grad_norm": 0.3909107439286301,
+      "learning_rate": 7.20451780418683e-06,
+      "loss": 0.3168,
+      "step": 925
+    },
+    {
+      "epoch": 3.748987854251012,
+      "grad_norm": 0.40192357409261603,
+      "learning_rate": 7.161103457852128e-06,
+      "loss": 0.3172,
+      "step": 926
+    },
+    {
+      "epoch": 3.753036437246964,
+      "grad_norm": 0.4365387442848415,
+      "learning_rate": 7.117791770992046e-06,
+      "loss": 0.3159,
+      "step": 927
+    },
+    {
+      "epoch": 3.757085020242915,
+      "grad_norm": 0.3850089221399938,
+      "learning_rate": 7.074583089925688e-06,
+      "loss": 0.3074,
+      "step": 928
+    },
+    {
+      "epoch": 3.7611336032388665,
+      "grad_norm": 0.4039137972341113,
+      "learning_rate": 7.0314777601485155e-06,
+      "loss": 0.3104,
+      "step": 929
+    },
+    {
+      "epoch": 3.765182186234818,
+      "grad_norm": 0.40614054566186053,
+      "learning_rate": 6.988476126329604e-06,
+      "loss": 0.3184,
+      "step": 930
+    },
+    {
+      "epoch": 3.769230769230769,
+      "grad_norm": 0.4055132875246551,
+      "learning_rate": 6.945578532308857e-06,
+      "loss": 0.3134,
+      "step": 931
+    },
+    {
+      "epoch": 3.7732793522267207,
+      "grad_norm": 0.4002523670560014,
+      "learning_rate": 6.902785321094301e-06,
+      "loss": 0.3211,
+      "step": 932
+    },
+    {
+      "epoch": 3.7773279352226723,
+      "grad_norm": 0.38407847959509206,
+      "learning_rate": 6.8600968348593115e-06,
+      "loss": 0.3279,
+      "step": 933
+    },
+    {
+      "epoch": 3.7813765182186234,
+      "grad_norm": 0.4087311976323667,
+      "learning_rate": 6.817513414939892e-06,
+      "loss": 0.3238,
+      "step": 934
+    },
+    {
+      "epoch": 3.785425101214575,
+      "grad_norm": 0.4210160146182284,
+      "learning_rate": 6.775035401831924e-06,
+      "loss": 0.319,
+      "step": 935
+    },
+    {
+      "epoch": 3.7894736842105265,
+      "grad_norm": 0.39448256022311123,
+      "learning_rate": 6.732663135188473e-06,
+      "loss": 0.311,
+      "step": 936
+    },
+    {
+      "epoch": 3.7935222672064777,
+      "grad_norm": 0.41213696237836056,
+      "learning_rate": 6.690396953817055e-06,
+      "loss": 0.3083,
+      "step": 937
+    },
+    {
+      "epoch": 3.7975708502024292,
+      "grad_norm": 0.3977790591928748,
+      "learning_rate": 6.648237195676927e-06,
+      "loss": 0.3175,
+      "step": 938
+    },
+    {
+      "epoch": 3.801619433198381,
+      "grad_norm": 0.41324281143560293,
+      "learning_rate": 6.606184197876395e-06,
+      "loss": 0.3148,
+      "step": 939
+    },
+    {
+      "epoch": 3.805668016194332,
+      "grad_norm": 3.1939886838268623,
+      "learning_rate": 6.564238296670109e-06,
+      "loss": 0.4812,
+      "step": 940
+    },
+    {
+      "epoch": 3.8097165991902835,
+      "grad_norm": 0.3934159299568829,
+      "learning_rate": 6.52239982745638e-06,
+      "loss": 0.3226,
+      "step": 941
+    },
+    {
+      "epoch": 3.813765182186235,
+      "grad_norm": 0.37742021186445124,
+      "learning_rate": 6.480669124774484e-06,
+      "loss": 0.322,
+      "step": 942
+    },
+    {
+      "epoch": 3.817813765182186,
+      "grad_norm": 0.3982824640948855,
+      "learning_rate": 6.439046522302013e-06,
+      "loss": 0.3112,
+      "step": 943
+    },
+    {
+      "epoch": 3.8218623481781377,
+      "grad_norm": 0.39162102485632505,
+      "learning_rate": 6.397532352852191e-06,
+      "loss": 0.3117,
+      "step": 944
+    },
+    {
+      "epoch": 3.8259109311740893,
+      "grad_norm": 0.38255990293951087,
+      "learning_rate": 6.356126948371202e-06,
+      "loss": 0.3227,
+      "step": 945
+    },
+    {
+      "epoch": 3.8299595141700404,
+      "grad_norm": 0.37490738944080265,
+      "learning_rate": 6.314830639935559e-06,
+      "loss": 0.3118,
+      "step": 946
+    },
+    {
+      "epoch": 3.834008097165992,
+      "grad_norm": 0.367368328275876,
+      "learning_rate": 6.273643757749448e-06,
+      "loss": 0.3293,
+      "step": 947
+    },
+    {
+      "epoch": 3.8380566801619436,
+      "grad_norm": 0.41461370745997683,
+      "learning_rate": 6.2325666311420856e-06,
+      "loss": 0.2961,
+      "step": 948
+    },
+    {
+      "epoch": 3.8421052631578947,
+      "grad_norm": 0.3857449955551371,
+      "learning_rate": 6.1915995885650675e-06,
+      "loss": 0.3055,
+      "step": 949
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 0.39530457410309977,
+      "learning_rate": 6.150742957589784e-06,
+      "loss": 0.3115,
+      "step": 950
+    },
+    {
+      "epoch": 3.850202429149798,
+      "grad_norm": 0.39352317331189257,
+      "learning_rate": 6.109997064904774e-06,
+      "loss": 0.3206,
+      "step": 951
+    },
+    {
+      "epoch": 3.854251012145749,
+      "grad_norm": 0.4002028708067021,
+      "learning_rate": 6.069362236313092e-06,
+      "loss": 0.3112,
+      "step": 952
+    },
+    {
+      "epoch": 3.8582995951417005,
+      "grad_norm": 0.39256406493507834,
+      "learning_rate": 6.0288387967297565e-06,
+      "loss": 0.3138,
+      "step": 953
+    },
+    {
+      "epoch": 3.862348178137652,
+      "grad_norm": 0.3859313241606953,
+      "learning_rate": 5.988427070179108e-06,
+      "loss": 0.3178,
+      "step": 954
+    },
+    {
+      "epoch": 3.866396761133603,
+      "grad_norm": 0.3917169772237801,
+      "learning_rate": 5.948127379792234e-06,
+      "loss": 0.3169,
+      "step": 955
+    },
+    {
+      "epoch": 3.8704453441295548,
+      "grad_norm": 0.3864276053509235,
+      "learning_rate": 5.90794004780439e-06,
+      "loss": 0.3104,
+      "step": 956
+    },
+    {
+      "epoch": 3.8744939271255063,
+      "grad_norm": 0.4033244648788553,
+      "learning_rate": 5.867865395552405e-06,
+      "loss": 0.3163,
+      "step": 957
+    },
+    {
+      "epoch": 3.8785425101214575,
+      "grad_norm": 0.373341126223814,
+      "learning_rate": 5.8279037434721364e-06,
+      "loss": 0.2938,
+      "step": 958
+    },
+    {
+      "epoch": 3.882591093117409,
+      "grad_norm": 0.38958590570464663,
+      "learning_rate": 5.788055411095892e-06,
+      "loss": 0.31,
+      "step": 959
+    },
+    {
+      "epoch": 3.8866396761133606,
+      "grad_norm": 0.4113629517875376,
+      "learning_rate": 5.7483207170498625e-06,
+      "loss": 0.3131,
+      "step": 960
+    },
+    {
+      "epoch": 3.8906882591093117,
+      "grad_norm": 0.4176690410123093,
+      "learning_rate": 5.708699979051613e-06,
+      "loss": 0.3031,
+      "step": 961
+    },
+    {
+      "epoch": 3.8947368421052633,
+      "grad_norm": 0.396136754239487,
+      "learning_rate": 5.66919351390751e-06,
+      "loss": 0.312,
+      "step": 962
+    },
+    {
+      "epoch": 3.898785425101215,
+      "grad_norm": 0.37982852571812503,
+      "learning_rate": 5.629801637510188e-06,
+      "loss": 0.3127,
+      "step": 963
+    },
+    {
+      "epoch": 3.902834008097166,
+      "grad_norm": 0.3860199652806424,
+      "learning_rate": 5.590524664836043e-06,
+      "loss": 0.3156,
+      "step": 964
+    },
+    {
+      "epoch": 3.9068825910931175,
+      "grad_norm": 0.4125143825138135,
+      "learning_rate": 5.551362909942708e-06,
+      "loss": 0.3034,
+      "step": 965
+    },
+    {
+      "epoch": 3.910931174089069,
+      "grad_norm": 0.3840390814510789,
+      "learning_rate": 5.51231668596653e-06,
+      "loss": 0.3079,
+      "step": 966
+    },
+    {
+      "epoch": 3.91497975708502,
+      "grad_norm": 0.3869791139962296,
+      "learning_rate": 5.47338630512007e-06,
+      "loss": 0.3132,
+      "step": 967
+    },
+    {
+      "epoch": 3.919028340080972,
+      "grad_norm": 0.40404928677549473,
+      "learning_rate": 5.434572078689617e-06,
+      "loss": 0.3151,
+      "step": 968
+    },
+    {
+      "epoch": 3.9230769230769234,
+      "grad_norm": 0.3596918893880109,
+      "learning_rate": 5.395874317032694e-06,
+      "loss": 0.3137,
+      "step": 969
+    },
+    {
+      "epoch": 3.9271255060728745,
+      "grad_norm": 0.3686349848532875,
+      "learning_rate": 5.357293329575573e-06,
+      "loss": 0.3001,
+      "step": 970
+    },
+    {
+      "epoch": 3.931174089068826,
+      "grad_norm": 0.3680632012876226,
+      "learning_rate": 5.318829424810797e-06,
+      "loss": 0.3375,
+      "step": 971
+    },
+    {
+      "epoch": 3.9352226720647776,
+      "grad_norm": 0.3827581536283051,
+      "learning_rate": 5.280482910294728e-06,
+      "loss": 0.3091,
+      "step": 972
+    },
+    {
+      "epoch": 3.9392712550607287,
+      "grad_norm": 0.388884027025834,
+      "learning_rate": 5.242254092645078e-06,
+      "loss": 0.316,
+      "step": 973
+    },
+    {
+      "epoch": 3.9433198380566803,
+      "grad_norm": 0.37968754376659886,
+      "learning_rate": 5.2041432775384445e-06,
+      "loss": 0.3061,
+      "step": 974
+    },
+    {
+      "epoch": 3.9473684210526314,
+      "grad_norm": 0.37183077734836745,
+      "learning_rate": 5.166150769707892e-06,
+      "loss": 0.3165,
+      "step": 975
+    },
+    {
+      "epoch": 3.951417004048583,
+      "grad_norm": 0.37362996419868477,
+      "learning_rate": 5.128276872940503e-06,
+      "loss": 0.3164,
+      "step": 976
+    },
+    {
+      "epoch": 3.9554655870445345,
+      "grad_norm": 0.40207789128799165,
+      "learning_rate": 5.090521890074953e-06,
+      "loss": 0.2961,
+      "step": 977
+    },
+    {
+      "epoch": 3.9595141700404857,
+      "grad_norm": 0.3673992556859058,
+      "learning_rate": 5.0528861229990675e-06,
+      "loss": 0.3234,
+      "step": 978
+    },
+    {
+      "epoch": 3.9635627530364372,
+      "grad_norm": 0.37007567004654895,
+      "learning_rate": 5.015369872647442e-06,
+      "loss": 0.3108,
+      "step": 979
+    },
+    {
+      "epoch": 3.967611336032389,
+      "grad_norm": 0.39900224751868707,
+      "learning_rate": 4.9779734389990195e-06,
+      "loss": 0.299,
+      "step": 980
+    },
+    {
+      "epoch": 3.97165991902834,
+      "grad_norm": 0.3874735449357752,
+      "learning_rate": 4.940697121074675e-06,
+      "loss": 0.3184,
+      "step": 981
+    },
+    {
+      "epoch": 3.9757085020242915,
+      "grad_norm": 0.36303882558641887,
+      "learning_rate": 4.903541216934864e-06,
+      "loss": 0.3022,
+      "step": 982
+    },
+    {
+      "epoch": 3.979757085020243,
+      "grad_norm": 0.37855138221402246,
+      "learning_rate": 4.8665060236772e-06,
+      "loss": 0.3091,
+      "step": 983
+    },
+    {
+      "epoch": 3.983805668016194,
+      "grad_norm": 0.3853009494414105,
+      "learning_rate": 4.829591837434102e-06,
+      "loss": 0.322,
+      "step": 984
+    },
+    {
+      "epoch": 3.9878542510121457,
+      "grad_norm": 0.3930801865594734,
+      "learning_rate": 4.792798953370424e-06,
+      "loss": 0.3093,
+      "step": 985
+    },
+    {
+      "epoch": 3.9919028340080973,
+      "grad_norm": 0.3911278827929508,
+      "learning_rate": 4.756127665681083e-06,
+      "loss": 0.3096,
+      "step": 986
+    },
+    {
+      "epoch": 3.9959514170040484,
+      "grad_norm": 0.3838964020448454,
+      "learning_rate": 4.719578267588725e-06,
+      "loss": 0.3254,
+      "step": 987
+    },
+    {
+      "epoch": 4.0,
+      "grad_norm": 0.37551733142260557,
+      "learning_rate": 4.6831510513413525e-06,
+      "loss": 0.3032,
+      "step": 988
+    },
+    {
+      "epoch": 4.004048582995951,
+      "grad_norm": 0.6708538591030712,
+      "learning_rate": 4.6468463082100176e-06,
+      "loss": 0.2351,
+      "step": 989
+    },
+    {
+      "epoch": 4.008097165991903,
+      "grad_norm": 0.5551200461653889,
+      "learning_rate": 4.610664328486485e-06,
+      "loss": 0.2202,
+      "step": 990
+    },
+    {
+      "epoch": 4.012145748987854,
+      "grad_norm": 0.3969463435006643,
+      "learning_rate": 4.5746054014809e-06,
+      "loss": 0.2156,
+      "step": 991
+    },
+    {
+      "epoch": 4.016194331983805,
+      "grad_norm": 0.5725329162568793,
+      "learning_rate": 4.5386698155194766e-06,
+      "loss": 0.2379,
+      "step": 992
+    },
+    {
+      "epoch": 4.020242914979757,
+      "grad_norm": 0.7179093482462969,
+      "learning_rate": 4.502857857942204e-06,
+      "loss": 0.2238,
+      "step": 993
+    },
+    {
+      "epoch": 4.0242914979757085,
+      "grad_norm": 0.6666261887465794,
+      "learning_rate": 4.467169815100549e-06,
+      "loss": 0.221,
+      "step": 994
+    },
+    {
+      "epoch": 4.02834008097166,
+      "grad_norm": 0.48441023851241044,
+      "learning_rate": 4.431605972355137e-06,
+      "loss": 0.2364,
+      "step": 995
+    },
+    {
+      "epoch": 4.032388663967612,
+      "grad_norm": 0.5247548613928057,
+      "learning_rate": 4.396166614073516e-06,
+      "loss": 0.237,
+      "step": 996
+    },
+    {
+      "epoch": 4.036437246963563,
+      "grad_norm": 0.6285915040062444,
+      "learning_rate": 4.360852023627846e-06,
+      "loss": 0.25,
+      "step": 997
+    },
+    {
+      "epoch": 4.040485829959514,
+      "grad_norm": 0.5717386196536701,
+      "learning_rate": 4.325662483392656e-06,
+      "loss": 0.218,
+      "step": 998
+    },
+    {
+      "epoch": 4.044534412955466,
+      "grad_norm": 0.4730558763226804,
+      "learning_rate": 4.29059827474257e-06,
+      "loss": 0.2185,
+      "step": 999
+    },
+    {
+      "epoch": 4.048582995951417,
+      "grad_norm": 0.4106639922677574,
+      "learning_rate": 4.255659678050063e-06,
+      "loss": 0.2349,
+      "step": 1000
+    },
+    {
+      "epoch": 4.052631578947368,
+      "grad_norm": 0.43951469307867636,
+      "learning_rate": 4.22084697268323e-06,
+      "loss": 0.2186,
+      "step": 1001
+    },
+    {
+      "epoch": 4.05668016194332,
+      "grad_norm": 0.4440563287698763,
+      "learning_rate": 4.186160437003535e-06,
+      "loss": 0.2373,
+      "step": 1002
+    },
+    {
+      "epoch": 4.060728744939271,
+      "grad_norm": 0.48465711330716577,
+      "learning_rate": 4.151600348363582e-06,
+      "loss": 0.2334,
+      "step": 1003
+    },
+    {
+      "epoch": 4.064777327935222,
+      "grad_norm": 0.44717928529761974,
+      "learning_rate": 4.11716698310493e-06,
+      "loss": 0.229,
+      "step": 1004
+    },
+    {
+      "epoch": 4.068825910931174,
+      "grad_norm": 0.41563707398638566,
+      "learning_rate": 4.082860616555848e-06,
+      "loss": 0.2208,
+      "step": 1005
+    },
+    {
+      "epoch": 4.0728744939271255,
+      "grad_norm": 0.38620714606394774,
+      "learning_rate": 4.048681523029125e-06,
+      "loss": 0.2219,
+      "step": 1006
+    },
+    {
+      "epoch": 4.076923076923077,
+      "grad_norm": 0.4002347168162354,
+      "learning_rate": 4.014629975819882e-06,
+      "loss": 0.2167,
+      "step": 1007
+    },
+    {
+      "epoch": 4.080971659919029,
+      "grad_norm": 0.40978198423243484,
+      "learning_rate": 3.9807062472033855e-06,
+      "loss": 0.2349,
+      "step": 1008
+    },
+    {
+      "epoch": 4.08502024291498,
+      "grad_norm": 0.3978919779911618,
+      "learning_rate": 3.94691060843287e-06,
+      "loss": 0.2333,
+      "step": 1009
+    },
+    {
+      "epoch": 4.089068825910931,
+      "grad_norm": 0.39511815983866755,
+      "learning_rate": 3.913243329737353e-06,
+      "loss": 0.2435,
+      "step": 1010
+    },
+    {
+      "epoch": 4.093117408906883,
+      "grad_norm": 0.4005248935013769,
+      "learning_rate": 3.879704680319503e-06,
+      "loss": 0.2197,
+      "step": 1011
+    },
+    {
+      "epoch": 4.097165991902834,
+      "grad_norm": 0.3927694463076842,
+      "learning_rate": 3.846294928353471e-06,
+      "loss": 0.228,
+      "step": 1012
+    },
+    {
+      "epoch": 4.101214574898785,
+      "grad_norm": 0.39124392982454465,
+      "learning_rate": 3.8130143409827345e-06,
+      "loss": 0.2277,
+      "step": 1013
+    },
+    {
+      "epoch": 4.105263157894737,
+      "grad_norm": 0.4000761232809199,
+      "learning_rate": 3.7798631843179887e-06,
+      "loss": 0.2349,
+      "step": 1014
+    },
+    {
+      "epoch": 4.109311740890688,
+      "grad_norm": 0.4098083022169801,
+      "learning_rate": 3.7468417234349973e-06,
+      "loss": 0.2273,
+      "step": 1015
+    },
+    {
+      "epoch": 4.113360323886639,
+      "grad_norm": 0.3978503062520131,
+      "learning_rate": 3.7139502223724823e-06,
+      "loss": 0.2383,
+      "step": 1016
+    },
+    {
+      "epoch": 4.117408906882591,
+      "grad_norm": 0.3710357146327907,
+      "learning_rate": 3.6811889441300095e-06,
+      "loss": 0.2293,
+      "step": 1017
+    },
+    {
+      "epoch": 4.1214574898785425,
+      "grad_norm": 0.36872792776131325,
+      "learning_rate": 3.648558150665886e-06,
+      "loss": 0.2225,
+      "step": 1018
+    },
+    {
+      "epoch": 4.125506072874494,
+      "grad_norm": 0.376518478569443,
+      "learning_rate": 3.616058102895075e-06,
+      "loss": 0.2332,
+      "step": 1019
+    },
+    {
+      "epoch": 4.129554655870446,
+      "grad_norm": 0.3706868019132022,
+      "learning_rate": 3.583689060687079e-06,
+      "loss": 0.2285,
+      "step": 1020
+    },
+    {
+      "epoch": 4.133603238866397,
+      "grad_norm": 0.385613811352824,
+      "learning_rate": 3.5514512828639023e-06,
+      "loss": 0.2248,
+      "step": 1021
+    },
+    {
+      "epoch": 4.137651821862348,
+      "grad_norm": 0.36463778211091247,
+      "learning_rate": 3.51934502719796e-06,
+      "loss": 0.2217,
+      "step": 1022
+    },
+    {
+      "epoch": 4.1417004048583,
+      "grad_norm": 0.38103512922476573,
+      "learning_rate": 3.4873705504100185e-06,
+      "loss": 0.2189,
+      "step": 1023
+    },
+    {
+      "epoch": 4.145748987854251,
+      "grad_norm": 0.35153105257571815,
+      "learning_rate": 3.4555281081671345e-06,
+      "loss": 0.2295,
+      "step": 1024
+    },
+    {
+      "epoch": 4.149797570850202,
+      "grad_norm": 0.4010271386026946,
+      "learning_rate": 3.4238179550806326e-06,
+      "loss": 0.2342,
+      "step": 1025
+    },
+    {
+      "epoch": 4.153846153846154,
+      "grad_norm": 0.38980686823427807,
+      "learning_rate": 3.3922403447040497e-06,
+      "loss": 0.2184,
+      "step": 1026
+    },
+    {
+      "epoch": 4.157894736842105,
+      "grad_norm": 0.38966806992597725,
+      "learning_rate": 3.3607955295311247e-06,
+      "loss": 0.2394,
+      "step": 1027
+    },
+    {
+      "epoch": 4.161943319838056,
+      "grad_norm": 0.3540547314733533,
+      "learning_rate": 3.32948376099375e-06,
+      "loss": 0.2258,
+      "step": 1028
+    },
+    {
+      "epoch": 4.165991902834008,
+      "grad_norm": 0.37231232937830017,
+      "learning_rate": 3.2983052894599976e-06,
+      "loss": 0.2042,
+      "step": 1029
+    },
+    {
+      "epoch": 4.17004048582996,
+      "grad_norm": 0.37252458985419257,
+      "learning_rate": 3.2672603642320945e-06,
+      "loss": 0.2328,
+      "step": 1030
+    },
+    {
+      "epoch": 4.174089068825911,
+      "grad_norm": 0.3844026651198638,
+      "learning_rate": 3.236349233544438e-06,
+      "loss": 0.2112,
+      "step": 1031
+    },
+    {
+      "epoch": 4.178137651821863,
+      "grad_norm": 0.3558431211602861,
+      "learning_rate": 3.205572144561604e-06,
+      "loss": 0.2329,
+      "step": 1032
+    },
+    {
+      "epoch": 4.182186234817814,
+      "grad_norm": 0.37780444164250254,
+      "learning_rate": 3.174929343376374e-06,
+      "loss": 0.2272,
+      "step": 1033
+    },
+    {
+      "epoch": 4.186234817813765,
+      "grad_norm": 0.35168917160243024,
+      "learning_rate": 3.1444210750077777e-06,
+      "loss": 0.2424,
+      "step": 1034
+    },
+    {
+      "epoch": 4.190283400809717,
+      "grad_norm": 0.37115003696162585,
+      "learning_rate": 3.1140475833991046e-06,
+      "loss": 0.2305,
+      "step": 1035
+    },
+    {
+      "epoch": 4.194331983805668,
+      "grad_norm": 0.3759549966906988,
+      "learning_rate": 3.0838091114159895e-06,
+      "loss": 0.2189,
+      "step": 1036
+    },
+    {
+      "epoch": 4.198380566801619,
+      "grad_norm": 0.3785366724776867,
+      "learning_rate": 3.0537059008444524e-06,
+      "loss": 0.2157,
+      "step": 1037
+    },
+    {
+      "epoch": 4.202429149797571,
+      "grad_norm": 0.36410690321676187,
+      "learning_rate": 3.0237381923889563e-06,
+      "loss": 0.2237,
+      "step": 1038
+    },
+    {
+      "epoch": 4.206477732793522,
+      "grad_norm": 0.36131221532504765,
+      "learning_rate": 2.9939062256704997e-06,
+      "loss": 0.2125,
+      "step": 1039
+    },
+    {
+      "epoch": 4.2105263157894735,
+      "grad_norm": 0.366735219421801,
+      "learning_rate": 2.964210239224696e-06,
+      "loss": 0.2324,
+      "step": 1040
+    },
+    {
+      "epoch": 4.2145748987854255,
+      "grad_norm": 0.38601564956371587,
+      "learning_rate": 2.934650470499869e-06,
+      "loss": 0.2282,
+      "step": 1041
+    },
+    {
+      "epoch": 4.218623481781377,
+      "grad_norm": 0.38038037666249674,
+      "learning_rate": 2.905227155855128e-06,
+      "loss": 0.2207,
+      "step": 1042
+    },
+    {
+      "epoch": 4.222672064777328,
+      "grad_norm": 0.36015602653837625,
+      "learning_rate": 2.875940530558519e-06,
+      "loss": 0.2197,
+      "step": 1043
+    },
+    {
+      "epoch": 4.22672064777328,
+      "grad_norm": 0.380525131812513,
+      "learning_rate": 2.8467908287851154e-06,
+      "loss": 0.2298,
+      "step": 1044
+    },
+    {
+      "epoch": 4.230769230769231,
+      "grad_norm": 0.3850127382112041,
+      "learning_rate": 2.8177782836151603e-06,
+      "loss": 0.2123,
+      "step": 1045
+    },
+    {
+      "epoch": 4.234817813765182,
+      "grad_norm": 0.37323070777697,
+      "learning_rate": 2.7889031270321874e-06,
+      "loss": 0.2164,
+      "step": 1046
+    },
+    {
+      "epoch": 4.238866396761134,
+      "grad_norm": 0.35103015692739065,
+      "learning_rate": 2.760165589921184e-06,
+      "loss": 0.2312,
+      "step": 1047
+    },
+    {
+      "epoch": 4.242914979757085,
+      "grad_norm": 0.37519676038424005,
+      "learning_rate": 2.7315659020667305e-06,
+      "loss": 0.2083,
+      "step": 1048
+    },
+    {
+      "epoch": 4.246963562753036,
+      "grad_norm": 0.37106929943579686,
+      "learning_rate": 2.703104292151164e-06,
+      "loss": 0.2347,
+      "step": 1049
+    },
+    {
+      "epoch": 4.251012145748988,
+      "grad_norm": 0.36954145295077506,
+      "learning_rate": 2.6747809877527607e-06,
+      "loss": 0.2254,
+      "step": 1050
+    },
+    {
+      "epoch": 4.255060728744939,
+      "grad_norm": 0.379075272495497,
+      "learning_rate": 2.6465962153439082e-06,
+      "loss": 0.225,
+      "step": 1051
+    },
+    {
+      "epoch": 4.2591093117408905,
+      "grad_norm": 0.35285319202633997,
+      "learning_rate": 2.618550200289303e-06,
+      "loss": 0.2317,
+      "step": 1052
+    },
+    {
+      "epoch": 4.2631578947368425,
+      "grad_norm": 0.3608464898944512,
+      "learning_rate": 2.59064316684412e-06,
+      "loss": 0.2359,
+      "step": 1053
+    },
+    {
+      "epoch": 4.267206477732794,
+      "grad_norm": 0.3623641891246903,
+      "learning_rate": 2.562875338152264e-06,
+      "loss": 0.2174,
+      "step": 1054
+    },
+    {
+      "epoch": 4.271255060728745,
+      "grad_norm": 0.36697182714516674,
+      "learning_rate": 2.5352469362445552e-06,
+      "loss": 0.2308,
+      "step": 1055
+    },
+    {
+      "epoch": 4.275303643724697,
+      "grad_norm": 0.3745803150836529,
+      "learning_rate": 2.5077581820369523e-06,
+      "loss": 0.2106,
+      "step": 1056
+    },
+    {
+      "epoch": 4.279352226720648,
+      "grad_norm": 0.3607851160438666,
+      "learning_rate": 2.480409295328803e-06,
+      "loss": 0.217,
+      "step": 1057
+    },
+    {
+      "epoch": 4.283400809716599,
+      "grad_norm": 0.3784675561538219,
+      "learning_rate": 2.45320049480108e-06,
+      "loss": 0.2222,
+      "step": 1058
+    },
+    {
+      "epoch": 4.287449392712551,
+      "grad_norm": 0.34923339189204233,
+      "learning_rate": 2.4261319980146293e-06,
+      "loss": 0.2334,
+      "step": 1059
+    },
+    {
+      "epoch": 4.291497975708502,
+      "grad_norm": 0.37848533931042055,
+      "learning_rate": 2.399204021408421e-06,
+      "loss": 0.2424,
+      "step": 1060
+    },
+    {
+      "epoch": 4.295546558704453,
+      "grad_norm": 0.36266946951083545,
+      "learning_rate": 2.3724167802978395e-06,
+      "loss": 0.2291,
+      "step": 1061
+    },
+    {
+      "epoch": 4.299595141700405,
+      "grad_norm": 0.3686182588347215,
+      "learning_rate": 2.3457704888729625e-06,
+      "loss": 0.2226,
+      "step": 1062
+    },
+    {
+      "epoch": 4.303643724696356,
+      "grad_norm": 0.3681448379506579,
+      "learning_rate": 2.3192653601968162e-06,
+      "loss": 0.2274,
+      "step": 1063
+    },
+    {
+      "epoch": 4.3076923076923075,
+      "grad_norm": 0.371664756695541,
+      "learning_rate": 2.2929016062037057e-06,
+      "loss": 0.2334,
+      "step": 1064
+    },
+    {
+      "epoch": 4.3117408906882595,
+      "grad_norm": 0.3648736949654214,
+      "learning_rate": 2.2666794376975055e-06,
+      "loss": 0.2161,
+      "step": 1065
+    },
+    {
+      "epoch": 4.315789473684211,
+      "grad_norm": 0.37208747988053975,
+      "learning_rate": 2.24059906434998e-06,
+      "loss": 0.2379,
+      "step": 1066
+    },
+    {
+      "epoch": 4.319838056680162,
+      "grad_norm": 0.35758903481656246,
+      "learning_rate": 2.214660694699091e-06,
+      "loss": 0.2331,
+      "step": 1067
+    },
+    {
+      "epoch": 4.323886639676114,
+      "grad_norm": 0.3601808787436978,
+      "learning_rate": 2.188864536147355e-06,
+      "loss": 0.2189,
+      "step": 1068
+    },
+    {
+      "epoch": 4.327935222672065,
+      "grad_norm": 0.3471523167323254,
+      "learning_rate": 2.1632107949601757e-06,
+      "loss": 0.2384,
+      "step": 1069
+    },
+    {
+      "epoch": 4.331983805668016,
+      "grad_norm": 0.3800588348265646,
+      "learning_rate": 2.137699676264171e-06,
+      "loss": 0.2272,
+      "step": 1070
+    },
+    {
+      "epoch": 4.336032388663968,
+      "grad_norm": 0.36746234139656975,
+      "learning_rate": 2.1123313840455738e-06,
+      "loss": 0.2214,
+      "step": 1071
+    },
+    {
+      "epoch": 4.340080971659919,
+      "grad_norm": 0.3690070877700327,
+      "learning_rate": 2.08710612114857e-06,
+      "loss": 0.2385,
+      "step": 1072
+    },
+    {
+      "epoch": 4.34412955465587,
+      "grad_norm": 0.3373355762273791,
+      "learning_rate": 2.0620240892736953e-06,
+      "loss": 0.2348,
+      "step": 1073
+    },
+    {
+      "epoch": 4.348178137651822,
+      "grad_norm": 0.363631925724982,
+      "learning_rate": 2.0370854889762e-06,
+      "loss": 0.2281,
+      "step": 1074
+    },
+    {
+      "epoch": 4.352226720647773,
+      "grad_norm": 0.3708054321485275,
+      "learning_rate": 2.0122905196644726e-06,
+      "loss": 0.2205,
+      "step": 1075
+    },
+    {
+      "epoch": 4.3562753036437245,
+      "grad_norm": 0.37252318280586383,
+      "learning_rate": 1.987639379598427e-06,
+      "loss": 0.2263,
+      "step": 1076
+    },
+    {
+      "epoch": 4.3603238866396765,
+      "grad_norm": 0.37775136503096124,
+      "learning_rate": 1.963132265887919e-06,
+      "loss": 0.2266,
+      "step": 1077
+    },
+    {
+      "epoch": 4.364372469635628,
+      "grad_norm": 0.3678793343304089,
+      "learning_rate": 1.9387693744911784e-06,
+      "loss": 0.2184,
+      "step": 1078
+    },
+    {
+      "epoch": 4.368421052631579,
+      "grad_norm": 0.3710560171377015,
+      "learning_rate": 1.9145509002132368e-06,
+      "loss": 0.2199,
+      "step": 1079
+    },
+    {
+      "epoch": 4.372469635627531,
+      "grad_norm": 0.35332133172428803,
+      "learning_rate": 1.8904770367043723e-06,
+      "loss": 0.2264,
+      "step": 1080
+    },
+    {
+      "epoch": 4.376518218623482,
+      "grad_norm": 0.36713719799262984,
+      "learning_rate": 1.8665479764585504e-06,
+      "loss": 0.2214,
+      "step": 1081
+    },
+    {
+      "epoch": 4.380566801619433,
+      "grad_norm": 0.36869232996903717,
+      "learning_rate": 1.8427639108118977e-06,
+      "loss": 0.2207,
+      "step": 1082
+    },
+    {
+      "epoch": 4.384615384615385,
+      "grad_norm": 0.36605210083882833,
+      "learning_rate": 1.8191250299411755e-06,
+      "loss": 0.2178,
+      "step": 1083
+    },
+    {
+      "epoch": 4.388663967611336,
+      "grad_norm": 0.37749682971367365,
+      "learning_rate": 1.7956315228622446e-06,
+      "loss": 0.2265,
+      "step": 1084
+    },
+    {
+      "epoch": 4.392712550607287,
+      "grad_norm": 0.3925102453071479,
+      "learning_rate": 1.7722835774285553e-06,
+      "loss": 0.2186,
+      "step": 1085
+    },
+    {
+      "epoch": 4.396761133603239,
+      "grad_norm": 0.35933324278055623,
+      "learning_rate": 1.7490813803296624e-06,
+      "loss": 0.2228,
+      "step": 1086
+    },
+    {
+      "epoch": 4.40080971659919,
+      "grad_norm": 0.3479949103965139,
+      "learning_rate": 1.726025117089718e-06,
+      "loss": 0.2393,
+      "step": 1087
+    },
+    {
+      "epoch": 4.4048582995951415,
+      "grad_norm": 0.3608761169327039,
+      "learning_rate": 1.70311497206598e-06,
+      "loss": 0.2171,
+      "step": 1088
+    },
+    {
+      "epoch": 4.4089068825910935,
+      "grad_norm": 0.3570285142707633,
+      "learning_rate": 1.6803511284473638e-06,
+      "loss": 0.228,
+      "step": 1089
+    },
+    {
+      "epoch": 4.412955465587045,
+      "grad_norm": 0.367846185825811,
+      "learning_rate": 1.6577337682529538e-06,
+      "loss": 0.2297,
+      "step": 1090
+    },
+    {
+      "epoch": 4.417004048582996,
+      "grad_norm": 0.35335766264530827,
+      "learning_rate": 1.6352630723305617e-06,
+      "loss": 0.2363,
+      "step": 1091
+    },
+    {
+      "epoch": 4.421052631578947,
+      "grad_norm": 0.3550303386454913,
+      "learning_rate": 1.612939220355272e-06,
+      "loss": 0.2196,
+      "step": 1092
+    },
+    {
+      "epoch": 4.425101214574899,
+      "grad_norm": 0.3708844005843545,
+      "learning_rate": 1.5907623908280112e-06,
+      "loss": 0.2369,
+      "step": 1093
+    },
+    {
+      "epoch": 4.42914979757085,
+      "grad_norm": 0.3518184050146964,
+      "learning_rate": 1.568732761074121e-06,
+      "loss": 0.2226,
+      "step": 1094
+    },
+    {
+      "epoch": 4.433198380566802,
+      "grad_norm": 0.3671105974235395,
+      "learning_rate": 1.5468505072419237e-06,
+      "loss": 0.2453,
+      "step": 1095
+    },
+    {
+      "epoch": 4.437246963562753,
+      "grad_norm": 0.3488867645976264,
+      "learning_rate": 1.5251158043013448e-06,
+      "loss": 0.2252,
+      "step": 1096
+    },
+    {
+      "epoch": 4.441295546558704,
+      "grad_norm": 0.362583000950518,
+      "learning_rate": 1.5035288260424885e-06,
+      "loss": 0.2435,
+      "step": 1097
+    },
+    {
+      "epoch": 4.445344129554655,
+      "grad_norm": 0.3727194869441794,
+      "learning_rate": 1.4820897450742623e-06,
+      "loss": 0.1979,
+      "step": 1098
+    },
+    {
+      "epoch": 4.449392712550607,
+      "grad_norm": 0.3575270304522218,
+      "learning_rate": 1.4607987328229767e-06,
+      "loss": 0.2245,
+      "step": 1099
+    },
+    {
+      "epoch": 4.4534412955465585,
+      "grad_norm": 0.3675719930801693,
+      "learning_rate": 1.4396559595310056e-06,
+      "loss": 0.2133,
+      "step": 1100
+    },
+    {
+      "epoch": 4.4574898785425106,
+      "grad_norm": 0.36832061790187703,
+      "learning_rate": 1.4186615942553973e-06,
+      "loss": 0.2093,
+      "step": 1101
+    },
+    {
+      "epoch": 4.461538461538462,
+      "grad_norm": 0.37097833652974993,
+      "learning_rate": 1.3978158048665424e-06,
+      "loss": 0.2175,
+      "step": 1102
+    },
+    {
+      "epoch": 4.465587044534413,
+      "grad_norm": 0.34796873299160264,
+      "learning_rate": 1.3771187580468115e-06,
+      "loss": 0.2167,
+      "step": 1103
+    },
+    {
+      "epoch": 4.469635627530364,
+      "grad_norm": 0.3674733756650656,
+      "learning_rate": 1.3565706192892392e-06,
+      "loss": 0.2273,
+      "step": 1104
+    },
+    {
+      "epoch": 4.473684210526316,
+      "grad_norm": 0.35453837599570415,
+      "learning_rate": 1.3361715528961993e-06,
+      "loss": 0.2249,
+      "step": 1105
+    },
+    {
+      "epoch": 4.477732793522267,
+      "grad_norm": 0.3724054759493854,
+      "learning_rate": 1.3159217219780817e-06,
+      "loss": 0.2257,
+      "step": 1106
+    },
+    {
+      "epoch": 4.481781376518219,
+      "grad_norm": 0.35296442057991745,
+      "learning_rate": 1.2958212884519949e-06,
+      "loss": 0.2396,
+      "step": 1107
+    },
+    {
+      "epoch": 4.48582995951417,
+      "grad_norm": 5.509625210573818,
+      "learning_rate": 1.2758704130404675e-06,
+      "loss": 0.3214,
+      "step": 1108
+    },
+    {
+      "epoch": 4.489878542510121,
+      "grad_norm": 0.36169358570908944,
+      "learning_rate": 1.2560692552701738e-06,
+      "loss": 0.2381,
+      "step": 1109
+    },
+    {
+      "epoch": 4.493927125506072,
+      "grad_norm": 0.370163089540907,
+      "learning_rate": 1.2364179734706382e-06,
+      "loss": 0.2189,
+      "step": 1110
+    },
+    {
+      "epoch": 4.497975708502024,
+      "grad_norm": 0.3550194476021693,
+      "learning_rate": 1.2169167247729852e-06,
+      "loss": 0.2262,
+      "step": 1111
+    },
+    {
+      "epoch": 4.502024291497976,
+      "grad_norm": 0.3672284630626532,
+      "learning_rate": 1.1975656651086864e-06,
+      "loss": 0.2121,
+      "step": 1112
+    },
+    {
+      "epoch": 4.506072874493928,
+      "grad_norm": 0.3555206545820634,
+      "learning_rate": 1.1783649492082926e-06,
+      "loss": 0.2299,
+      "step": 1113
+    },
+    {
+      "epoch": 4.510121457489879,
+      "grad_norm": 0.3640251941007677,
+      "learning_rate": 1.1593147306002183e-06,
+      "loss": 0.2201,
+      "step": 1114
+    },
+    {
+      "epoch": 4.51417004048583,
+      "grad_norm": 0.35368724745439517,
+      "learning_rate": 1.140415161609505e-06,
+      "loss": 0.2417,
+      "step": 1115
+    },
+    {
+      "epoch": 4.518218623481781,
+      "grad_norm": 0.3531971737925385,
+      "learning_rate": 1.121666393356604e-06,
+      "loss": 0.2364,
+      "step": 1116
+    },
+    {
+      "epoch": 4.522267206477733,
+      "grad_norm": 0.3604366827059777,
+      "learning_rate": 1.1030685757561632e-06,
+      "loss": 0.2226,
+      "step": 1117
+    },
+    {
+      "epoch": 4.526315789473684,
+      "grad_norm": 0.3651025946861312,
+      "learning_rate": 1.0846218575158373e-06,
+      "loss": 0.2328,
+      "step": 1118
+    },
+    {
+      "epoch": 4.530364372469636,
+      "grad_norm": 0.36004561346673986,
+      "learning_rate": 1.066326386135097e-06,
+      "loss": 0.2249,
+      "step": 1119
+    },
+    {
+      "epoch": 4.534412955465587,
+      "grad_norm": 0.36736602767468185,
+      "learning_rate": 1.0481823079040398e-06,
+      "loss": 0.2186,
+      "step": 1120
+    },
+    {
+      "epoch": 4.538461538461538,
+      "grad_norm": 0.3661413559076187,
+      "learning_rate": 1.0301897679022344e-06,
+      "loss": 0.2146,
+      "step": 1121
+    },
+    {
+      "epoch": 4.5425101214574894,
+      "grad_norm": 0.35259652624498017,
+      "learning_rate": 1.0123489099975491e-06,
+      "loss": 0.2284,
+      "step": 1122
+    },
+    {
+      "epoch": 4.5465587044534415,
+      "grad_norm": 0.3590260492153483,
+      "learning_rate": 9.9465987684501e-07,
+      "loss": 0.2257,
+      "step": 1123
+    },
+    {
+      "epoch": 4.550607287449393,
+      "grad_norm": 0.37167200183890814,
+      "learning_rate": 9.771228098856534e-07,
+      "loss": 0.2287,
+      "step": 1124
+    },
+    {
+      "epoch": 4.554655870445345,
+      "grad_norm": 0.3718932086299448,
+      "learning_rate": 9.597378493454013e-07,
+      "loss": 0.2201,
+      "step": 1125
+    },
+    {
+      "epoch": 4.558704453441296,
+      "grad_norm": 0.3544484472439895,
+      "learning_rate": 9.425051342339353e-07,
+      "loss": 0.2381,
+      "step": 1126
+    },
+    {
+      "epoch": 4.562753036437247,
+      "grad_norm": 0.347403886288815,
+      "learning_rate": 9.254248023435864e-07,
+      "loss": 0.2374,
+      "step": 1127
+    },
+    {
+      "epoch": 4.566801619433198,
+      "grad_norm": 0.35214173178441505,
+      "learning_rate": 9.084969902482288e-07,
+      "loss": 0.2192,
+      "step": 1128
+    },
+    {
+      "epoch": 4.57085020242915,
+      "grad_norm": 0.36781436542600615,
+      "learning_rate": 8.917218333021993e-07,
+      "loss": 0.2252,
+      "step": 1129
+    },
+    {
+      "epoch": 4.574898785425101,
+      "grad_norm": 0.36106829566334636,
+      "learning_rate": 8.750994656392109e-07,
+      "loss": 0.2135,
+      "step": 1130
+    },
+    {
+      "epoch": 4.578947368421053,
+      "grad_norm": 0.356353279470934,
+      "learning_rate": 8.58630020171265e-07,
+      "loss": 0.2198,
+      "step": 1131
+    },
+    {
+      "epoch": 4.582995951417004,
+      "grad_norm": 0.3752780397008564,
+      "learning_rate": 8.423136285876099e-07,
+      "loss": 0.2313,
+      "step": 1132
+    },
+    {
+      "epoch": 4.587044534412955,
+      "grad_norm": 0.3759209109496245,
+      "learning_rate": 8.26150421353682e-07,
+      "loss": 0.2235,
+      "step": 1133
+    },
+    {
+      "epoch": 4.5910931174089065,
+      "grad_norm": 0.35003140246346603,
+      "learning_rate": 8.101405277100549e-07,
+      "loss": 0.218,
+      "step": 1134
+    },
+    {
+      "epoch": 4.5951417004048585,
+      "grad_norm": 0.3670547117295844,
+      "learning_rate": 7.942840756714077e-07,
+      "loss": 0.2216,
+      "step": 1135
+    },
+    {
+      "epoch": 4.59919028340081,
+      "grad_norm": 0.36373976824580107,
+      "learning_rate": 7.785811920255093e-07,
+      "loss": 0.2151,
+      "step": 1136
+    },
+    {
+      "epoch": 4.603238866396762,
+      "grad_norm": 0.3650533995321982,
+      "learning_rate": 7.630320023322024e-07,
+      "loss": 0.2253,
+      "step": 1137
+    },
+    {
+      "epoch": 4.607287449392713,
+      "grad_norm": 0.37082364449881094,
+      "learning_rate": 7.476366309223903e-07,
+      "loss": 0.2256,
+      "step": 1138
+    },
+    {
+      "epoch": 4.611336032388664,
+      "grad_norm": 0.3710780573435566,
+      "learning_rate": 7.323952008970537e-07,
+      "loss": 0.2149,
+      "step": 1139
+    },
+    {
+      "epoch": 4.615384615384615,
+      "grad_norm": 0.36847133434156526,
+      "learning_rate": 7.173078341262618e-07,
+      "loss": 0.2099,
+      "step": 1140
+    },
+    {
+      "epoch": 4.619433198380567,
+      "grad_norm": 0.37192237617649176,
+      "learning_rate": 7.023746512481989e-07,
+      "loss": 0.2349,
+      "step": 1141
+    },
+    {
+      "epoch": 4.623481781376518,
+      "grad_norm": 0.37168392671086475,
+      "learning_rate": 6.875957716681902e-07,
+      "loss": 0.2206,
+      "step": 1142
+    },
+    {
+      "epoch": 4.62753036437247,
+      "grad_norm": 0.3604970921541651,
+      "learning_rate": 6.729713135577687e-07,
+      "loss": 0.2194,
+      "step": 1143
+    },
+    {
+      "epoch": 4.631578947368421,
+      "grad_norm": 0.34586958233498477,
+      "learning_rate": 6.585013938537143e-07,
+      "loss": 0.2248,
+      "step": 1144
+    },
+    {
+      "epoch": 4.635627530364372,
+      "grad_norm": 0.3628846529287771,
+      "learning_rate": 6.441861282571115e-07,
+      "loss": 0.2314,
+      "step": 1145
+    },
+    {
+      "epoch": 4.6396761133603235,
+      "grad_norm": 0.35584275777779584,
+      "learning_rate": 6.300256312324471e-07,
+      "loss": 0.2283,
+      "step": 1146
+    },
+    {
+      "epoch": 4.6437246963562755,
+      "grad_norm": 0.3422664404845478,
+      "learning_rate": 6.160200160066776e-07,
+      "loss": 0.2317,
+      "step": 1147
+    },
+    {
+      "epoch": 4.647773279352227,
+      "grad_norm": 0.3591233696762835,
+      "learning_rate": 6.021693945683327e-07,
+      "loss": 0.228,
+      "step": 1148
+    },
+    {
+      "epoch": 4.651821862348179,
+      "grad_norm": 0.3738399355896656,
+      "learning_rate": 5.884738776666044e-07,
+      "loss": 0.2167,
+      "step": 1149
+    },
+    {
+      "epoch": 4.65587044534413,
+      "grad_norm": 0.35642320382512144,
+      "learning_rate": 5.749335748104878e-07,
+      "loss": 0.2282,
+      "step": 1150
+    },
+    {
+      "epoch": 4.659919028340081,
+      "grad_norm": 0.37408520135281575,
+      "learning_rate": 5.615485942678778e-07,
+      "loss": 0.2181,
+      "step": 1151
+    },
+    {
+      "epoch": 4.663967611336032,
+      "grad_norm": 0.3603865491590705,
+      "learning_rate": 5.483190430647289e-07,
+      "loss": 0.2236,
+      "step": 1152
+    },
+    {
+      "epoch": 4.668016194331984,
+      "grad_norm": 0.35954774274659335,
+      "learning_rate": 5.352450269841747e-07,
+      "loss": 0.2157,
+      "step": 1153
+    },
+    {
+      "epoch": 4.672064777327935,
+      "grad_norm": 0.359712381360597,
+      "learning_rate": 5.22326650565701e-07,
+      "loss": 0.2416,
+      "step": 1154
+    },
+    {
+      "epoch": 4.676113360323887,
+      "grad_norm": 0.359893638011839,
+      "learning_rate": 5.095640171043048e-07,
+      "loss": 0.2345,
+      "step": 1155
+    },
+    {
+      "epoch": 4.680161943319838,
+      "grad_norm": 0.3533565745481335,
+      "learning_rate": 4.969572286496594e-07,
+      "loss": 0.2262,
+      "step": 1156
+    },
+    {
+      "epoch": 4.684210526315789,
+      "grad_norm": 0.36803543688361995,
+      "learning_rate": 4.845063860053079e-07,
+      "loss": 0.2022,
+      "step": 1157
+    },
+    {
+      "epoch": 4.6882591093117405,
+      "grad_norm": 0.3676048053259113,
+      "learning_rate": 4.7221158872785997e-07,
+      "loss": 0.2029,
+      "step": 1158
+    },
+    {
+      "epoch": 4.6923076923076925,
+      "grad_norm": 0.356832294668356,
+      "learning_rate": 4.600729351261857e-07,
+      "loss": 0.2237,
+      "step": 1159
+    },
+    {
+      "epoch": 4.696356275303644,
+      "grad_norm": 0.37508334432412177,
+      "learning_rate": 4.4809052226063355e-07,
+      "loss": 0.23,
+      "step": 1160
+    },
+    {
+      "epoch": 4.700404858299595,
+      "grad_norm": 0.35321466933263035,
+      "learning_rate": 4.3626444594226046e-07,
+      "loss": 0.2117,
+      "step": 1161
+    },
+    {
+      "epoch": 4.704453441295547,
+      "grad_norm": 0.3764167270910701,
+      "learning_rate": 4.245948007320588e-07,
+      "loss": 0.1979,
+      "step": 1162
+    },
+    {
+      "epoch": 4.708502024291498,
+      "grad_norm": 0.36771092517194165,
+      "learning_rate": 4.130816799401971e-07,
+      "loss": 0.2384,
+      "step": 1163
+    },
+    {
+      "epoch": 4.712550607287449,
+      "grad_norm": 0.34278477282608333,
+      "learning_rate": 4.017251756252827e-07,
+      "loss": 0.2142,
+      "step": 1164
+    },
+    {
+      "epoch": 4.716599190283401,
+      "grad_norm": 0.36114125047049483,
+      "learning_rate": 3.9052537859362473e-07,
+      "loss": 0.227,
+      "step": 1165
+    },
+    {
+      "epoch": 4.720647773279352,
+      "grad_norm": 0.38848744419962783,
+      "learning_rate": 3.7948237839849687e-07,
+      "loss": 0.2223,
+      "step": 1166
+    },
+    {
+      "epoch": 4.724696356275303,
+      "grad_norm": 0.3532998519492891,
+      "learning_rate": 3.6859626333943797e-07,
+      "loss": 0.2218,
+      "step": 1167
+    },
+    {
+      "epoch": 4.728744939271255,
+      "grad_norm": 0.3467783704116183,
+      "learning_rate": 3.578671204615347e-07,
+      "loss": 0.2073,
+      "step": 1168
+    },
+    {
+      "epoch": 4.732793522267206,
+      "grad_norm": 0.3683546939228451,
+      "learning_rate": 3.472950355547267e-07,
+      "loss": 0.2282,
+      "step": 1169
+    },
+    {
+      "epoch": 4.7368421052631575,
+      "grad_norm": 0.34704700093071433,
+      "learning_rate": 3.3688009315312244e-07,
+      "loss": 0.2138,
+      "step": 1170
+    },
+    {
+      "epoch": 4.7408906882591095,
+      "grad_norm": 0.34203130111386437,
+      "learning_rate": 3.266223765343224e-07,
+      "loss": 0.2216,
+      "step": 1171
+    },
+    {
+      "epoch": 4.744939271255061,
+      "grad_norm": 0.3521505941384088,
+      "learning_rate": 3.16521967718757e-07,
+      "loss": 0.2106,
+      "step": 1172
+    },
+    {
+      "epoch": 4.748987854251012,
+      "grad_norm": 0.35589395252876704,
+      "learning_rate": 3.0657894746902726e-07,
+      "loss": 0.237,
+      "step": 1173
+    },
+    {
+      "epoch": 4.753036437246964,
+      "grad_norm": 0.35517150969335354,
+      "learning_rate": 2.9679339528925654e-07,
+      "loss": 0.2157,
+      "step": 1174
+    },
+    {
+      "epoch": 4.757085020242915,
+      "grad_norm": 0.35514304913103834,
+      "learning_rate": 2.871653894244597e-07,
+      "loss": 0.2297,
+      "step": 1175
+    },
+    {
+      "epoch": 4.761133603238866,
+      "grad_norm": 0.36468686062169453,
+      "learning_rate": 2.7769500685991266e-07,
+      "loss": 0.2172,
+      "step": 1176
+    },
+    {
+      "epoch": 4.765182186234818,
+      "grad_norm": 0.35084991908411833,
+      "learning_rate": 2.683823233205485e-07,
+      "loss": 0.2307,
+      "step": 1177
+    },
+    {
+      "epoch": 4.769230769230769,
+      "grad_norm": 0.34209752535676297,
+      "learning_rate": 2.592274132703287e-07,
+      "loss": 0.231,
+      "step": 1178
+    },
+    {
+      "epoch": 4.77327935222672,
+      "grad_norm": 0.3740377762624594,
+      "learning_rate": 2.502303499116754e-07,
+      "loss": 0.22,
+      "step": 1179
+    },
+    {
+      "epoch": 4.777327935222672,
+      "grad_norm": 0.3659780648954515,
+      "learning_rate": 2.413912051848666e-07,
+      "loss": 0.2218,
+      "step": 1180
+    },
+    {
+      "epoch": 4.781376518218623,
+      "grad_norm": 0.3614868565292996,
+      "learning_rate": 2.3271004976746613e-07,
+      "loss": 0.2386,
+      "step": 1181
+    },
+    {
+      "epoch": 4.7854251012145745,
+      "grad_norm": 0.3641258155264177,
+      "learning_rate": 2.2418695307376393e-07,
+      "loss": 0.2195,
+      "step": 1182
+    },
+    {
+      "epoch": 4.7894736842105265,
+      "grad_norm": 0.36221426459161316,
+      "learning_rate": 2.1582198325420968e-07,
+      "loss": 0.2168,
+      "step": 1183
+    },
+    {
+      "epoch": 4.793522267206478,
+      "grad_norm": 0.3570102429792326,
+      "learning_rate": 2.0761520719488447e-07,
+      "loss": 0.2166,
+      "step": 1184
+    },
+    {
+      "epoch": 4.797570850202429,
+      "grad_norm": 0.3515190552134478,
+      "learning_rate": 1.9956669051694798e-07,
+      "loss": 0.2253,
+      "step": 1185
+    },
+    {
+      "epoch": 4.801619433198381,
+      "grad_norm": 0.34397773045954955,
+      "learning_rate": 1.916764975761254e-07,
+      "loss": 0.2176,
+      "step": 1186
+    },
+    {
+      "epoch": 4.805668016194332,
+      "grad_norm": 0.3687516159160382,
+      "learning_rate": 1.839446914621923e-07,
+      "loss": 0.2281,
+      "step": 1187
+    },
+    {
+      "epoch": 4.809716599190283,
+      "grad_norm": 0.3588471293540617,
+      "learning_rate": 1.7637133399845962e-07,
+      "loss": 0.2337,
+      "step": 1188
+    },
+    {
+      "epoch": 4.813765182186235,
+      "grad_norm": 0.361216230525289,
+      "learning_rate": 1.6895648574129397e-07,
+      "loss": 0.2106,
+      "step": 1189
+    },
+    {
+      "epoch": 4.817813765182186,
+      "grad_norm": 0.3685307470413959,
+      "learning_rate": 1.6170020597962465e-07,
+      "loss": 0.2017,
+      "step": 1190
+    },
+    {
+      "epoch": 4.821862348178137,
+      "grad_norm": 0.35912611454450033,
+      "learning_rate": 1.5460255273447078e-07,
+      "loss": 0.2183,
+      "step": 1191
+    },
+    {
+      "epoch": 4.825910931174089,
+      "grad_norm": 0.3584653771891412,
+      "learning_rate": 1.4766358275847715e-07,
+      "loss": 0.2291,
+      "step": 1192
+    },
+    {
+      "epoch": 4.82995951417004,
+      "grad_norm": 0.343172443551189,
+      "learning_rate": 1.4088335153546573e-07,
+      "loss": 0.2383,
+      "step": 1193
+    },
+    {
+      "epoch": 4.834008097165992,
+      "grad_norm": 0.35355381997254376,
+      "learning_rate": 1.3426191327998496e-07,
+      "loss": 0.2277,
+      "step": 1194
+    },
+    {
+      "epoch": 4.838056680161944,
+      "grad_norm": 0.3551796372322166,
+      "learning_rate": 1.277993209368744e-07,
+      "loss": 0.2288,
+      "step": 1195
+    },
+    {
+      "epoch": 4.842105263157895,
+      "grad_norm": 0.3533415376018305,
+      "learning_rate": 1.214956261808542e-07,
+      "loss": 0.2307,
+      "step": 1196
+    },
+    {
+      "epoch": 4.846153846153846,
+      "grad_norm": 0.36983899744797877,
+      "learning_rate": 1.1535087941609846e-07,
+      "loss": 0.2187,
+      "step": 1197
+    },
+    {
+      "epoch": 4.850202429149798,
+      "grad_norm": 0.355596342703396,
+      "learning_rate": 1.0936512977583802e-07,
+      "loss": 0.2324,
+      "step": 1198
+    },
+    {
+      "epoch": 4.854251012145749,
+      "grad_norm": 0.3553784185156337,
+      "learning_rate": 1.0353842512196732e-07,
+      "loss": 0.2296,
+      "step": 1199
+    },
+    {
+      "epoch": 4.8582995951417,
+      "grad_norm": 0.34379615520065393,
+      "learning_rate": 9.787081204466253e-08,
+      "loss": 0.2222,
+      "step": 1200
+    },
+    {
+      "epoch": 4.862348178137652,
+      "grad_norm": 0.35950951194510433,
+      "learning_rate": 9.236233586200627e-08,
+      "loss": 0.2225,
+      "step": 1201
+    },
+    {
+      "epoch": 4.866396761133603,
+      "grad_norm": 0.37144662802015416,
+      "learning_rate": 8.70130406196279e-08,
+      "loss": 0.226,
+      "step": 1202
+    },
+    {
+      "epoch": 4.870445344129554,
+      "grad_norm": 0.35915447267801726,
+      "learning_rate": 8.182296909035047e-08,
+      "loss": 0.2176,
+      "step": 1203
+    },
+    {
+      "epoch": 4.874493927125506,
+      "grad_norm": 0.3557333488322219,
+      "learning_rate": 7.679216277384439e-08,
+      "loss": 0.2315,
+      "step": 1204
+    },
+    {
+      "epoch": 4.8785425101214575,
+      "grad_norm": 0.36615724464269084,
+      "learning_rate": 7.19206618963053e-08,
+      "loss": 0.2196,
+      "step": 1205
+    },
+    {
+      "epoch": 4.882591093117409,
+      "grad_norm": 0.35297198858376494,
+      "learning_rate": 6.720850541012347e-08,
+      "loss": 0.2204,
+      "step": 1206
+    },
+    {
+      "epoch": 4.886639676113361,
+      "grad_norm": 0.3583838701659506,
+      "learning_rate": 6.265573099357714e-08,
+      "loss": 0.2419,
+      "step": 1207
+    },
+    {
+      "epoch": 4.890688259109312,
+      "grad_norm": 0.36563431369170435,
+      "learning_rate": 5.826237505053067e-08,
+      "loss": 0.2182,
+      "step": 1208
+    },
+    {
+      "epoch": 4.894736842105263,
+      "grad_norm": 0.372588316681097,
+      "learning_rate": 5.4028472710137e-08,
+      "loss": 0.2235,
+      "step": 1209
+    },
+    {
+      "epoch": 4.898785425101215,
+      "grad_norm": 0.35070375946543236,
+      "learning_rate": 4.9954057826571146e-08,
+      "loss": 0.23,
+      "step": 1210
+    },
+    {
+      "epoch": 4.902834008097166,
+      "grad_norm": 0.349297829357981,
+      "learning_rate": 4.6039162978743776e-08,
+      "loss": 0.2228,
+      "step": 1211
+    },
+    {
+      "epoch": 4.906882591093117,
+      "grad_norm": 0.3676635061238659,
+      "learning_rate": 4.228381947004812e-08,
+      "loss": 0.2178,
+      "step": 1212
+    },
+    {
+      "epoch": 4.910931174089069,
+      "grad_norm": 0.34253550935625715,
+      "learning_rate": 3.868805732811343e-08,
+      "loss": 0.2394,
+      "step": 1213
+    },
+    {
+      "epoch": 4.91497975708502,
+      "grad_norm": 0.352638133148164,
+      "learning_rate": 3.525190530455636e-08,
+      "loss": 0.2291,
+      "step": 1214
+    },
+    {
+      "epoch": 4.919028340080971,
+      "grad_norm": 0.36648341487512326,
+      "learning_rate": 3.197539087475887e-08,
+      "loss": 0.2362,
+      "step": 1215
+    },
+    {
+      "epoch": 4.923076923076923,
+      "grad_norm": 0.3745702299881134,
+      "learning_rate": 2.8858540237641786e-08,
+      "loss": 0.2292,
+      "step": 1216
+    },
+    {
+      "epoch": 4.9271255060728745,
+      "grad_norm": 0.3543647788541752,
+      "learning_rate": 2.5901378315458247e-08,
+      "loss": 0.2286,
+      "step": 1217
+    },
+    {
+      "epoch": 4.931174089068826,
+      "grad_norm": 0.35902814954055645,
+      "learning_rate": 2.3103928753598347e-08,
+      "loss": 0.2226,
+      "step": 1218
+    },
+    {
+      "epoch": 4.935222672064778,
+      "grad_norm": 0.3697294923726762,
+      "learning_rate": 2.0466213920393718e-08,
+      "loss": 0.2325,
+      "step": 1219
+    },
+    {
+      "epoch": 4.939271255060729,
+      "grad_norm": 0.3763834153963268,
+      "learning_rate": 1.7988254906942115e-08,
+      "loss": 0.2248,
+      "step": 1220
+    },
+    {
+      "epoch": 4.94331983805668,
+      "grad_norm": 0.3634967760923329,
+      "learning_rate": 1.5670071526936447e-08,
+      "loss": 0.2257,
+      "step": 1221
+    },
+    {
+      "epoch": 4.947368421052632,
+      "grad_norm": 0.37196699845937403,
+      "learning_rate": 1.3511682316509344e-08,
+      "loss": 0.2135,
+      "step": 1222
+    },
+    {
+      "epoch": 4.951417004048583,
+      "grad_norm": 0.3548226597363989,
+      "learning_rate": 1.1513104534086605e-08,
+      "loss": 0.2142,
+      "step": 1223
+    },
+    {
+      "epoch": 4.955465587044534,
+      "grad_norm": 0.34873427535751,
+      "learning_rate": 9.67435416023843e-09,
+      "loss": 0.217,
+      "step": 1224
+    },
+    {
+      "epoch": 4.959514170040486,
+      "grad_norm": 0.35365504559267813,
+      "learning_rate": 7.995445897566178e-09,
+      "loss": 0.2141,
+      "step": 1225
+    },
+    {
+      "epoch": 4.963562753036437,
+      "grad_norm": 0.34795641741942585,
+      "learning_rate": 6.476393170578021e-09,
+      "loss": 0.2427,
+      "step": 1226
+    },
+    {
+      "epoch": 4.967611336032388,
+      "grad_norm": 0.3628896454052226,
+      "learning_rate": 5.11720812557348e-09,
+      "loss": 0.2247,
+      "step": 1227
+    },
+    {
+      "epoch": 4.97165991902834,
+      "grad_norm": 0.3529782060829724,
+      "learning_rate": 3.917901630563492e-09,
+      "loss": 0.2207,
+      "step": 1228
+    },
+    {
+      "epoch": 4.9757085020242915,
+      "grad_norm": 0.3579944070227318,
+      "learning_rate": 2.878483275170485e-09,
+      "loss": 0.2136,
+      "step": 1229
+    },
+    {
+      "epoch": 4.979757085020243,
+      "grad_norm": 0.3487499963085914,
+      "learning_rate": 1.998961370555108e-09,
+      "loss": 0.2255,
+      "step": 1230
+    },
+    {
+      "epoch": 4.983805668016195,
+      "grad_norm": 0.3661605943694416,
+      "learning_rate": 1.2793429493518361e-09,
+      "loss": 0.2293,
+      "step": 1231
+    },
+    {
+      "epoch": 4.987854251012146,
+      "grad_norm": 0.36091534988488383,
+      "learning_rate": 7.196337656112384e-10,
+      "loss": 0.2257,
+      "step": 1232
+    },
+    {
+      "epoch": 4.991902834008097,
+      "grad_norm": 0.3470947745618737,
+      "learning_rate": 3.198382947511292e-10,
+      "loss": 0.2392,
+      "step": 1233
+    },
+    {
+      "epoch": 4.995951417004049,
+      "grad_norm": 0.3580462274878544,
+      "learning_rate": 7.995973352770137e-11,
+      "loss": 0.2313,
+      "step": 1234
+    },
+    {
+      "epoch": 5.0,
+      "grad_norm": 0.36561355474848917,
+      "learning_rate": 0.0,
+      "loss": 0.2105,
+      "step": 1235
+    },
+    {
+      "epoch": 5.0,
+      "step": 1235,
+      "total_flos": 1203201120075776.0,
+      "train_loss": 0.45908891277515934,
+      "train_runtime": 13629.1302,
+      "train_samples_per_second": 11.593,
+      "train_steps_per_second": 0.091
+    }
+  ],
+  "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": 1203201120075776.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}