diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20238 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2885,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003466204506065858,
+      "grad_norm": 4618.282804934648,
+      "learning_rate": 0.0,
+      "loss": 16.0147,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006932409012131716,
+      "grad_norm": 5213.5842520780825,
+      "learning_rate": 1.1494252873563219e-07,
+      "loss": 17.8753,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010398613518197574,
+      "grad_norm": 5409.710276374634,
+      "learning_rate": 2.2988505747126437e-07,
+      "loss": 18.1862,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013864818024263432,
+      "grad_norm": 4136.051829421708,
+      "learning_rate": 3.4482758620689656e-07,
+      "loss": 16.5366,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017331022530329288,
+      "grad_norm": 3334.2961288135266,
+      "learning_rate": 4.5977011494252875e-07,
+      "loss": 16.1778,
+      "step": 5
+    },
+    {
+      "epoch": 0.0020797227036395147,
+      "grad_norm": 2125.034308601018,
+      "learning_rate": 5.747126436781609e-07,
+      "loss": 16.6772,
+      "step": 6
+    },
+    {
+      "epoch": 0.0024263431542461003,
+      "grad_norm": 1998.850515461059,
+      "learning_rate": 6.896551724137931e-07,
+      "loss": 15.0461,
+      "step": 7
+    },
+    {
+      "epoch": 0.0027729636048526864,
+      "grad_norm": 2201.588748275882,
+      "learning_rate": 8.045977011494253e-07,
+      "loss": 15.787,
+      "step": 8
+    },
+    {
+      "epoch": 0.003119584055459272,
+      "grad_norm": 2684.608244320096,
+      "learning_rate": 9.195402298850575e-07,
+      "loss": 15.9028,
+      "step": 9
+    },
+    {
+      "epoch": 0.0034662045060658577,
+      "grad_norm": 1335.784839763534,
+      "learning_rate": 1.0344827586206898e-06,
+      "loss": 13.8919,
+      "step": 10
+    },
+    {
+      "epoch": 0.0038128249566724438,
+      "grad_norm": 1166.3967879549855,
+      "learning_rate": 1.1494252873563219e-06,
+      "loss": 12.5174,
+      "step": 11
+    },
+    {
+      "epoch": 0.004159445407279029,
+      "grad_norm": 1266.4769253934846,
+      "learning_rate": 1.2643678160919542e-06,
+      "loss": 12.9142,
+      "step": 12
+    },
+    {
+      "epoch": 0.004506065857885615,
+      "grad_norm": 1106.558629680182,
+      "learning_rate": 1.3793103448275862e-06,
+      "loss": 11.3865,
+      "step": 13
+    },
+    {
+      "epoch": 0.004852686308492201,
+      "grad_norm": 864.7507360825546,
+      "learning_rate": 1.4942528735632185e-06,
+      "loss": 9.4808,
+      "step": 14
+    },
+    {
+      "epoch": 0.005199306759098787,
+      "grad_norm": 1066.7853605185871,
+      "learning_rate": 1.6091954022988506e-06,
+      "loss": 10.2843,
+      "step": 15
+    },
+    {
+      "epoch": 0.005545927209705373,
+      "grad_norm": 690.8549588092386,
+      "learning_rate": 1.724137931034483e-06,
+      "loss": 8.0828,
+      "step": 16
+    },
+    {
+      "epoch": 0.0058925476603119585,
+      "grad_norm": 528.0957808889266,
+      "learning_rate": 1.839080459770115e-06,
+      "loss": 7.5071,
+      "step": 17
+    },
+    {
+      "epoch": 0.006239168110918544,
+      "grad_norm": 293.7810445697396,
+      "learning_rate": 1.9540229885057475e-06,
+      "loss": 5.936,
+      "step": 18
+    },
+    {
+      "epoch": 0.00658578856152513,
+      "grad_norm": 144.22721031890694,
+      "learning_rate": 2.0689655172413796e-06,
+      "loss": 5.2956,
+      "step": 19
+    },
+    {
+      "epoch": 0.006932409012131715,
+      "grad_norm": 171.90176865437178,
+      "learning_rate": 2.1839080459770117e-06,
+      "loss": 5.1679,
+      "step": 20
+    },
+    {
+      "epoch": 0.007279029462738302,
+      "grad_norm": 204.46252527677265,
+      "learning_rate": 2.2988505747126437e-06,
+      "loss": 5.5122,
+      "step": 21
+    },
+    {
+      "epoch": 0.0076256499133448875,
+      "grad_norm": 169.05792515283744,
+      "learning_rate": 2.4137931034482762e-06,
+      "loss": 4.4633,
+      "step": 22
+    },
+    {
+      "epoch": 0.007972270363951473,
+      "grad_norm": 228.50563921692242,
+      "learning_rate": 2.5287356321839083e-06,
+      "loss": 4.9628,
+      "step": 23
+    },
+    {
+      "epoch": 0.008318890814558059,
+      "grad_norm": 110.87530744693845,
+      "learning_rate": 2.6436781609195404e-06,
+      "loss": 4.2628,
+      "step": 24
+    },
+    {
+      "epoch": 0.008665511265164644,
+      "grad_norm": 136.7812086806913,
+      "learning_rate": 2.7586206896551725e-06,
+      "loss": 3.9873,
+      "step": 25
+    },
+    {
+      "epoch": 0.00901213171577123,
+      "grad_norm": 130.54742877150514,
+      "learning_rate": 2.8735632183908046e-06,
+      "loss": 3.6381,
+      "step": 26
+    },
+    {
+      "epoch": 0.009358752166377816,
+      "grad_norm": 118.48273003224699,
+      "learning_rate": 2.988505747126437e-06,
+      "loss": 3.7173,
+      "step": 27
+    },
+    {
+      "epoch": 0.009705372616984401,
+      "grad_norm": 122.72173665454332,
+      "learning_rate": 3.103448275862069e-06,
+      "loss": 3.6341,
+      "step": 28
+    },
+    {
+      "epoch": 0.010051993067590987,
+      "grad_norm": 250.72882647757098,
+      "learning_rate": 3.2183908045977012e-06,
+      "loss": 3.3925,
+      "step": 29
+    },
+    {
+      "epoch": 0.010398613518197574,
+      "grad_norm": 328.9700692788651,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 4.0645,
+      "step": 30
+    },
+    {
+      "epoch": 0.01074523396880416,
+      "grad_norm": 54.113767902695244,
+      "learning_rate": 3.448275862068966e-06,
+      "loss": 3.1032,
+      "step": 31
+    },
+    {
+      "epoch": 0.011091854419410746,
+      "grad_norm": 75.58281310986887,
+      "learning_rate": 3.563218390804598e-06,
+      "loss": 3.0049,
+      "step": 32
+    },
+    {
+      "epoch": 0.011438474870017331,
+      "grad_norm": 63.9685641561165,
+      "learning_rate": 3.67816091954023e-06,
+      "loss": 2.8288,
+      "step": 33
+    },
+    {
+      "epoch": 0.011785095320623917,
+      "grad_norm": 4716.267761872261,
+      "learning_rate": 3.793103448275862e-06,
+      "loss": 20.9533,
+      "step": 34
+    },
+    {
+      "epoch": 0.012131715771230503,
+      "grad_norm": 70.08804565058698,
+      "learning_rate": 3.908045977011495e-06,
+      "loss": 2.6855,
+      "step": 35
+    },
+    {
+      "epoch": 0.012478336221837088,
+      "grad_norm": 53.66522473592411,
+      "learning_rate": 4.022988505747127e-06,
+      "loss": 2.5901,
+      "step": 36
+    },
+    {
+      "epoch": 0.012824956672443674,
+      "grad_norm": 76.50821364486562,
+      "learning_rate": 4.137931034482759e-06,
+      "loss": 2.4186,
+      "step": 37
+    },
+    {
+      "epoch": 0.01317157712305026,
+      "grad_norm": 30.501907225904844,
+      "learning_rate": 4.252873563218391e-06,
+      "loss": 2.2111,
+      "step": 38
+    },
+    {
+      "epoch": 0.013518197573656845,
+      "grad_norm": 60.87101409968745,
+      "learning_rate": 4.367816091954023e-06,
+      "loss": 2.0748,
+      "step": 39
+    },
+    {
+      "epoch": 0.01386481802426343,
+      "grad_norm": 45.92327473924702,
+      "learning_rate": 4.482758620689656e-06,
+      "loss": 1.9156,
+      "step": 40
+    },
+    {
+      "epoch": 0.014211438474870018,
+      "grad_norm": 152.27216820128604,
+      "learning_rate": 4.5977011494252875e-06,
+      "loss": 2.0127,
+      "step": 41
+    },
+    {
+      "epoch": 0.014558058925476604,
+      "grad_norm": 68.88232980028772,
+      "learning_rate": 4.71264367816092e-06,
+      "loss": 1.8441,
+      "step": 42
+    },
+    {
+      "epoch": 0.01490467937608319,
+      "grad_norm": 68.9411443950094,
+      "learning_rate": 4.8275862068965525e-06,
+      "loss": 1.9401,
+      "step": 43
+    },
+    {
+      "epoch": 0.015251299826689775,
+      "grad_norm": 89.41449331118261,
+      "learning_rate": 4.942528735632184e-06,
+      "loss": 1.7181,
+      "step": 44
+    },
+    {
+      "epoch": 0.01559792027729636,
+      "grad_norm": 46.76695542782983,
+      "learning_rate": 5.057471264367817e-06,
+      "loss": 1.7787,
+      "step": 45
+    },
+    {
+      "epoch": 0.015944540727902946,
+      "grad_norm": 33.522202099542,
+      "learning_rate": 5.172413793103449e-06,
+      "loss": 1.7505,
+      "step": 46
+    },
+    {
+      "epoch": 0.016291161178509532,
+      "grad_norm": 30.121462865785542,
+      "learning_rate": 5.287356321839081e-06,
+      "loss": 1.6952,
+      "step": 47
+    },
+    {
+      "epoch": 0.016637781629116118,
+      "grad_norm": 64.84583447300922,
+      "learning_rate": 5.402298850574713e-06,
+      "loss": 1.7638,
+      "step": 48
+    },
+    {
+      "epoch": 0.016984402079722703,
+      "grad_norm": 40.720694375534876,
+      "learning_rate": 5.517241379310345e-06,
+      "loss": 1.5699,
+      "step": 49
+    },
+    {
+      "epoch": 0.01733102253032929,
+      "grad_norm": 26.83292220306791,
+      "learning_rate": 5.6321839080459775e-06,
+      "loss": 1.6617,
+      "step": 50
+    },
+    {
+      "epoch": 0.017677642980935875,
+      "grad_norm": 176.30878620102465,
+      "learning_rate": 5.747126436781609e-06,
+      "loss": 1.7026,
+      "step": 51
+    },
+    {
+      "epoch": 0.01802426343154246,
+      "grad_norm": 45.91874619543963,
+      "learning_rate": 5.862068965517242e-06,
+      "loss": 1.5127,
+      "step": 52
+    },
+    {
+      "epoch": 0.018370883882149046,
+      "grad_norm": 58.64407565008856,
+      "learning_rate": 5.977011494252874e-06,
+      "loss": 1.6313,
+      "step": 53
+    },
+    {
+      "epoch": 0.01871750433275563,
+      "grad_norm": 18.40269627142795,
+      "learning_rate": 6.091954022988507e-06,
+      "loss": 1.4517,
+      "step": 54
+    },
+    {
+      "epoch": 0.019064124783362217,
+      "grad_norm": 21.316706590065447,
+      "learning_rate": 6.206896551724138e-06,
+      "loss": 1.3059,
+      "step": 55
+    },
+    {
+      "epoch": 0.019410745233968803,
+      "grad_norm": 21.771733234858406,
+      "learning_rate": 6.321839080459771e-06,
+      "loss": 1.5169,
+      "step": 56
+    },
+    {
+      "epoch": 0.01975736568457539,
+      "grad_norm": 14.740175557476102,
+      "learning_rate": 6.4367816091954025e-06,
+      "loss": 1.4375,
+      "step": 57
+    },
+    {
+      "epoch": 0.020103986135181974,
+      "grad_norm": 18.690154937614523,
+      "learning_rate": 6.551724137931035e-06,
+      "loss": 1.2524,
+      "step": 58
+    },
+    {
+      "epoch": 0.020450606585788563,
+      "grad_norm": 17.18921531208569,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.462,
+      "step": 59
+    },
+    {
+      "epoch": 0.02079722703639515,
+      "grad_norm": 10.014412121147148,
+      "learning_rate": 6.781609195402299e-06,
+      "loss": 1.5189,
+      "step": 60
+    },
+    {
+      "epoch": 0.021143847487001734,
+      "grad_norm": 31.145774628242833,
+      "learning_rate": 6.896551724137932e-06,
+      "loss": 1.3397,
+      "step": 61
+    },
+    {
+      "epoch": 0.02149046793760832,
+      "grad_norm": 12.861553515552583,
+      "learning_rate": 7.011494252873564e-06,
+      "loss": 1.3147,
+      "step": 62
+    },
+    {
+      "epoch": 0.021837088388214906,
+      "grad_norm": 34.569436404710416,
+      "learning_rate": 7.126436781609196e-06,
+      "loss": 1.2923,
+      "step": 63
+    },
+    {
+      "epoch": 0.02218370883882149,
+      "grad_norm": 12.158689317607632,
+      "learning_rate": 7.241379310344828e-06,
+      "loss": 1.3198,
+      "step": 64
+    },
+    {
+      "epoch": 0.022530329289428077,
+      "grad_norm": 28.456131885408436,
+      "learning_rate": 7.35632183908046e-06,
+      "loss": 1.3265,
+      "step": 65
+    },
+    {
+      "epoch": 0.022876949740034663,
+      "grad_norm": 17.773673987493257,
+      "learning_rate": 7.4712643678160925e-06,
+      "loss": 1.3047,
+      "step": 66
+    },
+    {
+      "epoch": 0.023223570190641248,
+      "grad_norm": 13.560055071977681,
+      "learning_rate": 7.586206896551724e-06,
+      "loss": 1.3564,
+      "step": 67
+    },
+    {
+      "epoch": 0.023570190641247834,
+      "grad_norm": 19.115383865559867,
+      "learning_rate": 7.701149425287356e-06,
+      "loss": 1.2955,
+      "step": 68
+    },
+    {
+      "epoch": 0.02391681109185442,
+      "grad_norm": 18.24881928921815,
+      "learning_rate": 7.81609195402299e-06,
+      "loss": 1.2496,
+      "step": 69
+    },
+    {
+      "epoch": 0.024263431542461005,
+      "grad_norm": 10.871460931159795,
+      "learning_rate": 7.93103448275862e-06,
+      "loss": 1.3906,
+      "step": 70
+    },
+    {
+      "epoch": 0.02461005199306759,
+      "grad_norm": 10.737724969783773,
+      "learning_rate": 8.045977011494253e-06,
+      "loss": 1.3147,
+      "step": 71
+    },
+    {
+      "epoch": 0.024956672443674176,
+      "grad_norm": 18.408899639448737,
+      "learning_rate": 8.160919540229886e-06,
+      "loss": 1.2045,
+      "step": 72
+    },
+    {
+      "epoch": 0.025303292894280762,
+      "grad_norm": 13.445831976654624,
+      "learning_rate": 8.275862068965518e-06,
+      "loss": 1.2127,
+      "step": 73
+    },
+    {
+      "epoch": 0.025649913344887348,
+      "grad_norm": 16.706869117491483,
+      "learning_rate": 8.390804597701149e-06,
+      "loss": 1.203,
+      "step": 74
+    },
+    {
+      "epoch": 0.025996533795493933,
+      "grad_norm": 9.122593255784889,
+      "learning_rate": 8.505747126436782e-06,
+      "loss": 1.2925,
+      "step": 75
+    },
+    {
+      "epoch": 0.02634315424610052,
+      "grad_norm": 13.570792876352762,
+      "learning_rate": 8.620689655172414e-06,
+      "loss": 1.1558,
+      "step": 76
+    },
+    {
+      "epoch": 0.026689774696707105,
+      "grad_norm": 11.937609286247422,
+      "learning_rate": 8.735632183908047e-06,
+      "loss": 1.1361,
+      "step": 77
+    },
+    {
+      "epoch": 0.02703639514731369,
+      "grad_norm": 13.65435941991185,
+      "learning_rate": 8.85057471264368e-06,
+      "loss": 1.0866,
+      "step": 78
+    },
+    {
+      "epoch": 0.027383015597920276,
+      "grad_norm": 7.74366103090689,
+      "learning_rate": 8.965517241379312e-06,
+      "loss": 1.1831,
+      "step": 79
+    },
+    {
+      "epoch": 0.02772963604852686,
+      "grad_norm": 11.150433448859165,
+      "learning_rate": 9.080459770114942e-06,
+      "loss": 1.1897,
+      "step": 80
+    },
+    {
+      "epoch": 0.028076256499133447,
+      "grad_norm": 57.612587288945676,
+      "learning_rate": 9.195402298850575e-06,
+      "loss": 1.1184,
+      "step": 81
+    },
+    {
+      "epoch": 0.028422876949740036,
+      "grad_norm": 35.55721441411927,
+      "learning_rate": 9.310344827586207e-06,
+      "loss": 1.0859,
+      "step": 82
+    },
+    {
+      "epoch": 0.028769497400346622,
+      "grad_norm": 5.994600732482119,
+      "learning_rate": 9.42528735632184e-06,
+      "loss": 1.1023,
+      "step": 83
+    },
+    {
+      "epoch": 0.029116117850953208,
+      "grad_norm": 17.20725820230211,
+      "learning_rate": 9.54022988505747e-06,
+      "loss": 1.2,
+      "step": 84
+    },
+    {
+      "epoch": 0.029462738301559793,
+      "grad_norm": 9.113666839020665,
+      "learning_rate": 9.655172413793105e-06,
+      "loss": 1.1211,
+      "step": 85
+    },
+    {
+      "epoch": 0.02980935875216638,
+      "grad_norm": 14.750757960251278,
+      "learning_rate": 9.770114942528738e-06,
+      "loss": 1.0665,
+      "step": 86
+    },
+    {
+      "epoch": 0.030155979202772965,
+      "grad_norm": 14.784606769941155,
+      "learning_rate": 9.885057471264368e-06,
+      "loss": 1.1422,
+      "step": 87
+    },
+    {
+      "epoch": 0.03050259965337955,
+      "grad_norm": 14.972141203807944,
+      "learning_rate": 1e-05,
+      "loss": 1.0264,
+      "step": 88
+    },
+    {
+      "epoch": 0.030849220103986136,
+      "grad_norm": 13.531586912546912,
+      "learning_rate": 9.999996848304237e-06,
+      "loss": 1.121,
+      "step": 89
+    },
+    {
+      "epoch": 0.03119584055459272,
+      "grad_norm": 10.36302857840998,
+      "learning_rate": 9.999987393220914e-06,
+      "loss": 0.9816,
+      "step": 90
+    },
+    {
+      "epoch": 0.03154246100519931,
+      "grad_norm": 20.445734727626593,
+      "learning_rate": 9.999971634761957e-06,
+      "loss": 1.1702,
+      "step": 91
+    },
+    {
+      "epoch": 0.03188908145580589,
+      "grad_norm": 7.185933264792868,
+      "learning_rate": 9.999949572947229e-06,
+      "loss": 1.0624,
+      "step": 92
+    },
+    {
+      "epoch": 0.03223570190641248,
+      "grad_norm": 5.554528883771991,
+      "learning_rate": 9.999921207804544e-06,
+      "loss": 1.1265,
+      "step": 93
+    },
+    {
+      "epoch": 0.032582322357019064,
+      "grad_norm": 5.49977923031017,
+      "learning_rate": 9.99988653936966e-06,
+      "loss": 1.0979,
+      "step": 94
+    },
+    {
+      "epoch": 0.03292894280762565,
+      "grad_norm": 8.91695296315903,
+      "learning_rate": 9.999845567686284e-06,
+      "loss": 0.9849,
+      "step": 95
+    },
+    {
+      "epoch": 0.033275563258232235,
+      "grad_norm": 8.743913886651002,
+      "learning_rate": 9.999798292806068e-06,
+      "loss": 1.1372,
+      "step": 96
+    },
+    {
+      "epoch": 0.03362218370883882,
+      "grad_norm": 13.372099659870027,
+      "learning_rate": 9.999744714788611e-06,
+      "loss": 0.922,
+      "step": 97
+    },
+    {
+      "epoch": 0.03396880415944541,
+      "grad_norm": 6.172677688454842,
+      "learning_rate": 9.999684833701455e-06,
+      "loss": 1.091,
+      "step": 98
+    },
+    {
+      "epoch": 0.03431542461005199,
+      "grad_norm": 5.434346865122065,
+      "learning_rate": 9.999618649620093e-06,
+      "loss": 1.013,
+      "step": 99
+    },
+    {
+      "epoch": 0.03466204506065858,
+      "grad_norm": 15.70379135059192,
+      "learning_rate": 9.999546162627962e-06,
+      "loss": 1.0481,
+      "step": 100
+    },
+    {
+      "epoch": 0.03500866551126516,
+      "grad_norm": 6.830049925276472,
+      "learning_rate": 9.999467372816445e-06,
+      "loss": 0.9946,
+      "step": 101
+    },
+    {
+      "epoch": 0.03535528596187175,
+      "grad_norm": 9.638228223897892,
+      "learning_rate": 9.99938228028487e-06,
+      "loss": 0.9778,
+      "step": 102
+    },
+    {
+      "epoch": 0.035701906412478335,
+      "grad_norm": 8.268958304083215,
+      "learning_rate": 9.99929088514051e-06,
+      "loss": 1.1513,
+      "step": 103
+    },
+    {
+      "epoch": 0.03604852686308492,
+      "grad_norm": 8.293163221055886,
+      "learning_rate": 9.999193187498586e-06,
+      "loss": 0.8943,
+      "step": 104
+    },
+    {
+      "epoch": 0.036395147313691506,
+      "grad_norm": 10.29285794142903,
+      "learning_rate": 9.999089187482265e-06,
+      "loss": 0.951,
+      "step": 105
+    },
+    {
+      "epoch": 0.03674176776429809,
+      "grad_norm": 10.44168531886649,
+      "learning_rate": 9.998978885222654e-06,
+      "loss": 1.0452,
+      "step": 106
+    },
+    {
+      "epoch": 0.03708838821490468,
+      "grad_norm": 19.143895877179677,
+      "learning_rate": 9.998862280858811e-06,
+      "loss": 1.0641,
+      "step": 107
+    },
+    {
+      "epoch": 0.03743500866551126,
+      "grad_norm": 8.335046549371656,
+      "learning_rate": 9.998739374537737e-06,
+      "loss": 1.0137,
+      "step": 108
+    },
+    {
+      "epoch": 0.03778162911611785,
+      "grad_norm": 3.9759151076014656,
+      "learning_rate": 9.998610166414375e-06,
+      "loss": 1.0126,
+      "step": 109
+    },
+    {
+      "epoch": 0.038128249566724434,
+      "grad_norm": 4.223280201045921,
+      "learning_rate": 9.998474656651618e-06,
+      "loss": 1.0842,
+      "step": 110
+    },
+    {
+      "epoch": 0.03847487001733102,
+      "grad_norm": 6.455802564591906,
+      "learning_rate": 9.998332845420298e-06,
+      "loss": 1.0139,
+      "step": 111
+    },
+    {
+      "epoch": 0.038821490467937605,
+      "grad_norm": 7.896816749832506,
+      "learning_rate": 9.998184732899193e-06,
+      "loss": 0.9858,
+      "step": 112
+    },
+    {
+      "epoch": 0.03916811091854419,
+      "grad_norm": 10.511366626644786,
+      "learning_rate": 9.998030319275028e-06,
+      "loss": 0.892,
+      "step": 113
+    },
+    {
+      "epoch": 0.03951473136915078,
+      "grad_norm": 10.313762759809007,
+      "learning_rate": 9.997869604742466e-06,
+      "loss": 1.0112,
+      "step": 114
+    },
+    {
+      "epoch": 0.03986135181975736,
+      "grad_norm": 5.983695994334371,
+      "learning_rate": 9.997702589504118e-06,
+      "loss": 0.9765,
+      "step": 115
+    },
+    {
+      "epoch": 0.04020797227036395,
+      "grad_norm": 7.363269469938932,
+      "learning_rate": 9.997529273770535e-06,
+      "loss": 1.0649,
+      "step": 116
+    },
+    {
+      "epoch": 0.04055459272097054,
+      "grad_norm": 7.989769825461357,
+      "learning_rate": 9.997349657760212e-06,
+      "loss": 1.1034,
+      "step": 117
+    },
+    {
+      "epoch": 0.040901213171577126,
+      "grad_norm": 6.2940601431984255,
+      "learning_rate": 9.997163741699591e-06,
+      "loss": 0.9551,
+      "step": 118
+    },
+    {
+      "epoch": 0.04124783362218371,
+      "grad_norm": 8.175615115832349,
+      "learning_rate": 9.996971525823048e-06,
+      "loss": 1.0398,
+      "step": 119
+    },
+    {
+      "epoch": 0.0415944540727903,
+      "grad_norm": 4.577047921745155,
+      "learning_rate": 9.996773010372906e-06,
+      "loss": 1.1334,
+      "step": 120
+    },
+    {
+      "epoch": 0.04194107452339688,
+      "grad_norm": 5.879849647241637,
+      "learning_rate": 9.996568195599433e-06,
+      "loss": 0.9006,
+      "step": 121
+    },
+    {
+      "epoch": 0.04228769497400347,
+      "grad_norm": 4.452720615229334,
+      "learning_rate": 9.99635708176083e-06,
+      "loss": 0.9304,
+      "step": 122
+    },
+    {
+      "epoch": 0.042634315424610054,
+      "grad_norm": 23.54986263184362,
+      "learning_rate": 9.996139669123246e-06,
+      "loss": 0.9703,
+      "step": 123
+    },
+    {
+      "epoch": 0.04298093587521664,
+      "grad_norm": 7.071398108733023,
+      "learning_rate": 9.995915957960766e-06,
+      "loss": 1.1328,
+      "step": 124
+    },
+    {
+      "epoch": 0.043327556325823226,
+      "grad_norm": 6.977639007971496,
+      "learning_rate": 9.995685948555423e-06,
+      "loss": 1.0668,
+      "step": 125
+    },
+    {
+      "epoch": 0.04367417677642981,
+      "grad_norm": 8.769970583890121,
+      "learning_rate": 9.995449641197178e-06,
+      "loss": 0.9827,
+      "step": 126
+    },
+    {
+      "epoch": 0.0440207972270364,
+      "grad_norm": 5.090974107547331,
+      "learning_rate": 9.995207036183944e-06,
+      "loss": 1.0529,
+      "step": 127
+    },
+    {
+      "epoch": 0.04436741767764298,
+      "grad_norm": 63.232020804006645,
+      "learning_rate": 9.994958133821565e-06,
+      "loss": 0.8144,
+      "step": 128
+    },
+    {
+      "epoch": 0.04471403812824957,
+      "grad_norm": 11.003988880347633,
+      "learning_rate": 9.994702934423828e-06,
+      "loss": 0.9168,
+      "step": 129
+    },
+    {
+      "epoch": 0.045060658578856154,
+      "grad_norm": 5.735828324223102,
+      "learning_rate": 9.994441438312456e-06,
+      "loss": 1.009,
+      "step": 130
+    },
+    {
+      "epoch": 0.04540727902946274,
+      "grad_norm": 5.062951375713483,
+      "learning_rate": 9.994173645817114e-06,
+      "loss": 0.9499,
+      "step": 131
+    },
+    {
+      "epoch": 0.045753899480069325,
+      "grad_norm": 8.007008250064091,
+      "learning_rate": 9.9938995572754e-06,
+      "loss": 0.9,
+      "step": 132
+    },
+    {
+      "epoch": 0.04610051993067591,
+      "grad_norm": 6.50536185037283,
+      "learning_rate": 9.993619173032852e-06,
+      "loss": 0.9743,
+      "step": 133
+    },
+    {
+      "epoch": 0.046447140381282497,
+      "grad_norm": 7.328791140335307,
+      "learning_rate": 9.993332493442944e-06,
+      "loss": 1.0833,
+      "step": 134
+    },
+    {
+      "epoch": 0.04679376083188908,
+      "grad_norm": 6.1951458087859566,
+      "learning_rate": 9.993039518867087e-06,
+      "loss": 1.0842,
+      "step": 135
+    },
+    {
+      "epoch": 0.04714038128249567,
+      "grad_norm": 5.973291284040921,
+      "learning_rate": 9.992740249674628e-06,
+      "loss": 0.8745,
+      "step": 136
+    },
+    {
+      "epoch": 0.04748700173310225,
+      "grad_norm": 6.282580816514619,
+      "learning_rate": 9.99243468624285e-06,
+      "loss": 1.0852,
+      "step": 137
+    },
+    {
+      "epoch": 0.04783362218370884,
+      "grad_norm": 4.435856165619402,
+      "learning_rate": 9.99212282895697e-06,
+      "loss": 0.8327,
+      "step": 138
+    },
+    {
+      "epoch": 0.048180242634315425,
+      "grad_norm": 4.703058614711892,
+      "learning_rate": 9.991804678210137e-06,
+      "loss": 0.8821,
+      "step": 139
+    },
+    {
+      "epoch": 0.04852686308492201,
+      "grad_norm": 5.500479236315137,
+      "learning_rate": 9.991480234403438e-06,
+      "loss": 0.9016,
+      "step": 140
+    },
+    {
+      "epoch": 0.048873483535528596,
+      "grad_norm": 5.198873486373792,
+      "learning_rate": 9.991149497945896e-06,
+      "loss": 0.9411,
+      "step": 141
+    },
+    {
+      "epoch": 0.04922010398613518,
+      "grad_norm": 5.70941204451936,
+      "learning_rate": 9.990812469254458e-06,
+      "loss": 0.9456,
+      "step": 142
+    },
+    {
+      "epoch": 0.04956672443674177,
+      "grad_norm": 5.659073602511497,
+      "learning_rate": 9.990469148754012e-06,
+      "loss": 0.8876,
+      "step": 143
+    },
+    {
+      "epoch": 0.04991334488734835,
+      "grad_norm": 7.898182759316434,
+      "learning_rate": 9.990119536877373e-06,
+      "loss": 0.9635,
+      "step": 144
+    },
+    {
+      "epoch": 0.05025996533795494,
+      "grad_norm": 9.257262288140792,
+      "learning_rate": 9.98976363406529e-06,
+      "loss": 0.9374,
+      "step": 145
+    },
+    {
+      "epoch": 0.050606585788561524,
+      "grad_norm": 15.488269597693538,
+      "learning_rate": 9.989401440766443e-06,
+      "loss": 0.969,
+      "step": 146
+    },
+    {
+      "epoch": 0.05095320623916811,
+      "grad_norm": 11.952044437369421,
+      "learning_rate": 9.989032957437439e-06,
+      "loss": 0.8264,
+      "step": 147
+    },
+    {
+      "epoch": 0.051299826689774695,
+      "grad_norm": 6.693612811903062,
+      "learning_rate": 9.988658184542817e-06,
+      "loss": 0.9157,
+      "step": 148
+    },
+    {
+      "epoch": 0.05164644714038128,
+      "grad_norm": 7.674287521230887,
+      "learning_rate": 9.988277122555048e-06,
+      "loss": 0.9873,
+      "step": 149
+    },
+    {
+      "epoch": 0.05199306759098787,
+      "grad_norm": 7.606580262695174,
+      "learning_rate": 9.987889771954525e-06,
+      "loss": 1.0545,
+      "step": 150
+    },
+    {
+      "epoch": 0.05233968804159445,
+      "grad_norm": 5.344315375749218,
+      "learning_rate": 9.987496133229575e-06,
+      "loss": 1.0092,
+      "step": 151
+    },
+    {
+      "epoch": 0.05268630849220104,
+      "grad_norm": 6.757354290844761,
+      "learning_rate": 9.987096206876447e-06,
+      "loss": 1.0383,
+      "step": 152
+    },
+    {
+      "epoch": 0.053032928942807624,
+      "grad_norm": 5.955273052833046,
+      "learning_rate": 9.986689993399324e-06,
+      "loss": 0.9329,
+      "step": 153
+    },
+    {
+      "epoch": 0.05337954939341421,
+      "grad_norm": 3.8798724039083248,
+      "learning_rate": 9.986277493310308e-06,
+      "loss": 0.9835,
+      "step": 154
+    },
+    {
+      "epoch": 0.053726169844020795,
+      "grad_norm": 7.333450734961028,
+      "learning_rate": 9.985858707129427e-06,
+      "loss": 1.0155,
+      "step": 155
+    },
+    {
+      "epoch": 0.05407279029462738,
+      "grad_norm": 14.789622436951639,
+      "learning_rate": 9.985433635384638e-06,
+      "loss": 0.9509,
+      "step": 156
+    },
+    {
+      "epoch": 0.054419410745233966,
+      "grad_norm": 21.20529070261329,
+      "learning_rate": 9.985002278611819e-06,
+      "loss": 0.9137,
+      "step": 157
+    },
+    {
+      "epoch": 0.05476603119584055,
+      "grad_norm": 8.929439285178203,
+      "learning_rate": 9.984564637354773e-06,
+      "loss": 0.8297,
+      "step": 158
+    },
+    {
+      "epoch": 0.05511265164644714,
+      "grad_norm": 5.30638589909566,
+      "learning_rate": 9.984120712165224e-06,
+      "loss": 0.8873,
+      "step": 159
+    },
+    {
+      "epoch": 0.05545927209705372,
+      "grad_norm": 6.634003345198622,
+      "learning_rate": 9.983670503602817e-06,
+      "loss": 0.8723,
+      "step": 160
+    },
+    {
+      "epoch": 0.05580589254766031,
+      "grad_norm": 5.150545793521682,
+      "learning_rate": 9.983214012235124e-06,
+      "loss": 0.9554,
+      "step": 161
+    },
+    {
+      "epoch": 0.056152512998266894,
+      "grad_norm": 21.83433402973597,
+      "learning_rate": 9.982751238637633e-06,
+      "loss": 0.8614,
+      "step": 162
+    },
+    {
+      "epoch": 0.05649913344887349,
+      "grad_norm": 4.564145975710755,
+      "learning_rate": 9.982282183393751e-06,
+      "loss": 1.0083,
+      "step": 163
+    },
+    {
+      "epoch": 0.05684575389948007,
+      "grad_norm": 3.9993687681932295,
+      "learning_rate": 9.981806847094806e-06,
+      "loss": 0.7877,
+      "step": 164
+    },
+    {
+      "epoch": 0.05719237435008666,
+      "grad_norm": 4.84941241265614,
+      "learning_rate": 9.981325230340045e-06,
+      "loss": 1.0033,
+      "step": 165
+    },
+    {
+      "epoch": 0.057538994800693244,
+      "grad_norm": 34.205862661016205,
+      "learning_rate": 9.98083733373663e-06,
+      "loss": 0.8925,
+      "step": 166
+    },
+    {
+      "epoch": 0.05788561525129983,
+      "grad_norm": 6.487323055781911,
+      "learning_rate": 9.980343157899643e-06,
+      "loss": 0.9161,
+      "step": 167
+    },
+    {
+      "epoch": 0.058232235701906415,
+      "grad_norm": 5.481669364924128,
+      "learning_rate": 9.979842703452083e-06,
+      "loss": 0.9097,
+      "step": 168
+    },
+    {
+      "epoch": 0.058578856152513,
+      "grad_norm": 6.718904643936427,
+      "learning_rate": 9.979335971024857e-06,
+      "loss": 0.7979,
+      "step": 169
+    },
+    {
+      "epoch": 0.058925476603119586,
+      "grad_norm": 5.174536127441246,
+      "learning_rate": 9.978822961256796e-06,
+      "loss": 0.9229,
+      "step": 170
+    },
+    {
+      "epoch": 0.05927209705372617,
+      "grad_norm": 5.033762100073308,
+      "learning_rate": 9.978303674794637e-06,
+      "loss": 0.943,
+      "step": 171
+    },
+    {
+      "epoch": 0.05961871750433276,
+      "grad_norm": 6.496924206935665,
+      "learning_rate": 9.977778112293038e-06,
+      "loss": 0.8844,
+      "step": 172
+    },
+    {
+      "epoch": 0.05996533795493934,
+      "grad_norm": 5.800831568169819,
+      "learning_rate": 9.977246274414559e-06,
+      "loss": 0.8944,
+      "step": 173
+    },
+    {
+      "epoch": 0.06031195840554593,
+      "grad_norm": 3.9658597665622137,
+      "learning_rate": 9.976708161829677e-06,
+      "loss": 0.8569,
+      "step": 174
+    },
+    {
+      "epoch": 0.060658578856152515,
+      "grad_norm": 5.603267908980764,
+      "learning_rate": 9.976163775216783e-06,
+      "loss": 0.8298,
+      "step": 175
+    },
+    {
+      "epoch": 0.0610051993067591,
+      "grad_norm": 5.049121392538148,
+      "learning_rate": 9.975613115262169e-06,
+      "loss": 0.8586,
+      "step": 176
+    },
+    {
+      "epoch": 0.061351819757365686,
+      "grad_norm": 5.476999771372688,
+      "learning_rate": 9.975056182660043e-06,
+      "loss": 0.9005,
+      "step": 177
+    },
+    {
+      "epoch": 0.06169844020797227,
+      "grad_norm": 12.7676394099054,
+      "learning_rate": 9.974492978112516e-06,
+      "loss": 0.8544,
+      "step": 178
+    },
+    {
+      "epoch": 0.06204506065857886,
+      "grad_norm": 5.801638300513387,
+      "learning_rate": 9.97392350232961e-06,
+      "loss": 1.0963,
+      "step": 179
+    },
+    {
+      "epoch": 0.06239168110918544,
+      "grad_norm": 18.38148780874575,
+      "learning_rate": 9.973347756029249e-06,
+      "loss": 0.9184,
+      "step": 180
+    },
+    {
+      "epoch": 0.06273830155979203,
+      "grad_norm": 5.595633229671469,
+      "learning_rate": 9.97276573993726e-06,
+      "loss": 1.0555,
+      "step": 181
+    },
+    {
+      "epoch": 0.06308492201039861,
+      "grad_norm": 14.202794372855356,
+      "learning_rate": 9.972177454787386e-06,
+      "loss": 1.0282,
+      "step": 182
+    },
+    {
+      "epoch": 0.0634315424610052,
+      "grad_norm": 5.302233643265699,
+      "learning_rate": 9.971582901321261e-06,
+      "loss": 0.9061,
+      "step": 183
+    },
+    {
+      "epoch": 0.06377816291161179,
+      "grad_norm": 5.372888709867176,
+      "learning_rate": 9.970982080288428e-06,
+      "loss": 0.9373,
+      "step": 184
+    },
+    {
+      "epoch": 0.06412478336221837,
+      "grad_norm": 5.472100818041015,
+      "learning_rate": 9.970374992446323e-06,
+      "loss": 0.8255,
+      "step": 185
+    },
+    {
+      "epoch": 0.06447140381282496,
+      "grad_norm": 5.046735230029842,
+      "learning_rate": 9.969761638560296e-06,
+      "loss": 0.9286,
+      "step": 186
+    },
+    {
+      "epoch": 0.06481802426343154,
+      "grad_norm": 4.571287377957819,
+      "learning_rate": 9.969142019403584e-06,
+      "loss": 0.9356,
+      "step": 187
+    },
+    {
+      "epoch": 0.06516464471403813,
+      "grad_norm": 10.72865473269801,
+      "learning_rate": 9.96851613575733e-06,
+      "loss": 0.896,
+      "step": 188
+    },
+    {
+      "epoch": 0.06551126516464471,
+      "grad_norm": 6.693356599257177,
+      "learning_rate": 9.96788398841057e-06,
+      "loss": 0.8806,
+      "step": 189
+    },
+    {
+      "epoch": 0.0658578856152513,
+      "grad_norm": 9.158255841570336,
+      "learning_rate": 9.967245578160239e-06,
+      "loss": 0.8702,
+      "step": 190
+    },
+    {
+      "epoch": 0.06620450606585788,
+      "grad_norm": 13.999244300603626,
+      "learning_rate": 9.966600905811168e-06,
+      "loss": 0.8411,
+      "step": 191
+    },
+    {
+      "epoch": 0.06655112651646447,
+      "grad_norm": 9.17175206313389,
+      "learning_rate": 9.965949972176081e-06,
+      "loss": 1.0145,
+      "step": 192
+    },
+    {
+      "epoch": 0.06689774696707106,
+      "grad_norm": 16.6998144790754,
+      "learning_rate": 9.965292778075596e-06,
+      "loss": 0.8929,
+      "step": 193
+    },
+    {
+      "epoch": 0.06724436741767764,
+      "grad_norm": 6.428903620608322,
+      "learning_rate": 9.964629324338221e-06,
+      "loss": 0.863,
+      "step": 194
+    },
+    {
+      "epoch": 0.06759098786828423,
+      "grad_norm": 7.452428900092671,
+      "learning_rate": 9.963959611800364e-06,
+      "loss": 0.9771,
+      "step": 195
+    },
+    {
+      "epoch": 0.06793760831889081,
+      "grad_norm": 5.478331507314076,
+      "learning_rate": 9.96328364130631e-06,
+      "loss": 0.9101,
+      "step": 196
+    },
+    {
+      "epoch": 0.0682842287694974,
+      "grad_norm": 7.890731498823678,
+      "learning_rate": 9.962601413708243e-06,
+      "loss": 0.8246,
+      "step": 197
+    },
+    {
+      "epoch": 0.06863084922010398,
+      "grad_norm": 6.131758495238151,
+      "learning_rate": 9.961912929866234e-06,
+      "loss": 0.9943,
+      "step": 198
+    },
+    {
+      "epoch": 0.06897746967071057,
+      "grad_norm": 4.504922658136465,
+      "learning_rate": 9.961218190648236e-06,
+      "loss": 0.9438,
+      "step": 199
+    },
+    {
+      "epoch": 0.06932409012131716,
+      "grad_norm": 4.632797294893285,
+      "learning_rate": 9.960517196930097e-06,
+      "loss": 0.9857,
+      "step": 200
+    },
+    {
+      "epoch": 0.06967071057192374,
+      "grad_norm": 7.548654450277067,
+      "learning_rate": 9.95980994959554e-06,
+      "loss": 0.9298,
+      "step": 201
+    },
+    {
+      "epoch": 0.07001733102253033,
+      "grad_norm": 4.673694201204682,
+      "learning_rate": 9.959096449536179e-06,
+      "loss": 0.9197,
+      "step": 202
+    },
+    {
+      "epoch": 0.07036395147313691,
+      "grad_norm": 13.227259014078731,
+      "learning_rate": 9.958376697651506e-06,
+      "loss": 0.8358,
+      "step": 203
+    },
+    {
+      "epoch": 0.0707105719237435,
+      "grad_norm": 4.607516984413122,
+      "learning_rate": 9.957650694848897e-06,
+      "loss": 0.7839,
+      "step": 204
+    },
+    {
+      "epoch": 0.07105719237435008,
+      "grad_norm": 8.693292202117366,
+      "learning_rate": 9.95691844204361e-06,
+      "loss": 0.8864,
+      "step": 205
+    },
+    {
+      "epoch": 0.07140381282495667,
+      "grad_norm": 14.248196298898993,
+      "learning_rate": 9.956179940158779e-06,
+      "loss": 0.9146,
+      "step": 206
+    },
+    {
+      "epoch": 0.07175043327556326,
+      "grad_norm": 7.584473470361438,
+      "learning_rate": 9.955435190125415e-06,
+      "loss": 0.9129,
+      "step": 207
+    },
+    {
+      "epoch": 0.07209705372616984,
+      "grad_norm": 5.3268633891501525,
+      "learning_rate": 9.954684192882414e-06,
+      "loss": 0.9517,
+      "step": 208
+    },
+    {
+      "epoch": 0.07244367417677643,
+      "grad_norm": 3.5836203272792626,
+      "learning_rate": 9.953926949376536e-06,
+      "loss": 0.8868,
+      "step": 209
+    },
+    {
+      "epoch": 0.07279029462738301,
+      "grad_norm": 5.835091393215384,
+      "learning_rate": 9.953163460562422e-06,
+      "loss": 0.9242,
+      "step": 210
+    },
+    {
+      "epoch": 0.0731369150779896,
+      "grad_norm": 4.462079458845755,
+      "learning_rate": 9.95239372740259e-06,
+      "loss": 0.8895,
+      "step": 211
+    },
+    {
+      "epoch": 0.07348353552859618,
+      "grad_norm": 7.971131789415096,
+      "learning_rate": 9.951617750867423e-06,
+      "loss": 0.8535,
+      "step": 212
+    },
+    {
+      "epoch": 0.07383015597920277,
+      "grad_norm": 5.225104503545673,
+      "learning_rate": 9.950835531935178e-06,
+      "loss": 1.0235,
+      "step": 213
+    },
+    {
+      "epoch": 0.07417677642980935,
+      "grad_norm": 6.0972169144399615,
+      "learning_rate": 9.950047071591982e-06,
+      "loss": 0.92,
+      "step": 214
+    },
+    {
+      "epoch": 0.07452339688041594,
+      "grad_norm": 12.284047168770785,
+      "learning_rate": 9.949252370831827e-06,
+      "loss": 1.0013,
+      "step": 215
+    },
+    {
+      "epoch": 0.07487001733102253,
+      "grad_norm": 8.600473638759373,
+      "learning_rate": 9.948451430656581e-06,
+      "loss": 0.9705,
+      "step": 216
+    },
+    {
+      "epoch": 0.07521663778162911,
+      "grad_norm": 5.854863943239285,
+      "learning_rate": 9.947644252075968e-06,
+      "loss": 0.9232,
+      "step": 217
+    },
+    {
+      "epoch": 0.0755632582322357,
+      "grad_norm": 7.903942975290449,
+      "learning_rate": 9.946830836107579e-06,
+      "loss": 0.9343,
+      "step": 218
+    },
+    {
+      "epoch": 0.07590987868284228,
+      "grad_norm": 4.598241308555538,
+      "learning_rate": 9.946011183776872e-06,
+      "loss": 0.8641,
+      "step": 219
+    },
+    {
+      "epoch": 0.07625649913344887,
+      "grad_norm": 6.555536832229999,
+      "learning_rate": 9.945185296117165e-06,
+      "loss": 1.0905,
+      "step": 220
+    },
+    {
+      "epoch": 0.07660311958405545,
+      "grad_norm": 5.913219565342373,
+      "learning_rate": 9.944353174169638e-06,
+      "loss": 0.84,
+      "step": 221
+    },
+    {
+      "epoch": 0.07694974003466204,
+      "grad_norm": 7.196717572404395,
+      "learning_rate": 9.943514818983326e-06,
+      "loss": 0.8858,
+      "step": 222
+    },
+    {
+      "epoch": 0.07729636048526863,
+      "grad_norm": 4.427284551642813,
+      "learning_rate": 9.942670231615127e-06,
+      "loss": 0.9744,
+      "step": 223
+    },
+    {
+      "epoch": 0.07764298093587521,
+      "grad_norm": 7.494576372807999,
+      "learning_rate": 9.941819413129794e-06,
+      "loss": 0.8841,
+      "step": 224
+    },
+    {
+      "epoch": 0.0779896013864818,
+      "grad_norm": 7.591818888103855,
+      "learning_rate": 9.940962364599933e-06,
+      "loss": 0.909,
+      "step": 225
+    },
+    {
+      "epoch": 0.07833622183708838,
+      "grad_norm": 5.80820759349654,
+      "learning_rate": 9.94009908710601e-06,
+      "loss": 0.8553,
+      "step": 226
+    },
+    {
+      "epoch": 0.07868284228769497,
+      "grad_norm": 6.985676538257756,
+      "learning_rate": 9.939229581736339e-06,
+      "loss": 0.9114,
+      "step": 227
+    },
+    {
+      "epoch": 0.07902946273830155,
+      "grad_norm": 7.270393927996093,
+      "learning_rate": 9.938353849587086e-06,
+      "loss": 0.7604,
+      "step": 228
+    },
+    {
+      "epoch": 0.07937608318890814,
+      "grad_norm": 10.684428855333744,
+      "learning_rate": 9.937471891762267e-06,
+      "loss": 0.9413,
+      "step": 229
+    },
+    {
+      "epoch": 0.07972270363951472,
+      "grad_norm": 5.290228381961196,
+      "learning_rate": 9.93658370937375e-06,
+      "loss": 0.9619,
+      "step": 230
+    },
+    {
+      "epoch": 0.08006932409012131,
+      "grad_norm": 6.140164815954596,
+      "learning_rate": 9.935689303541243e-06,
+      "loss": 0.9887,
+      "step": 231
+    },
+    {
+      "epoch": 0.0804159445407279,
+      "grad_norm": 22.665376271780527,
+      "learning_rate": 9.934788675392306e-06,
+      "loss": 0.9879,
+      "step": 232
+    },
+    {
+      "epoch": 0.0807625649913345,
+      "grad_norm": 6.842189914584052,
+      "learning_rate": 9.933881826062344e-06,
+      "loss": 0.9964,
+      "step": 233
+    },
+    {
+      "epoch": 0.08110918544194108,
+      "grad_norm": 3.6846896736625867,
+      "learning_rate": 9.932968756694597e-06,
+      "loss": 0.9829,
+      "step": 234
+    },
+    {
+      "epoch": 0.08145580589254767,
+      "grad_norm": 7.01855742428931,
+      "learning_rate": 9.932049468440157e-06,
+      "loss": 0.9747,
+      "step": 235
+    },
+    {
+      "epoch": 0.08180242634315425,
+      "grad_norm": 4.210728623592585,
+      "learning_rate": 9.931123962457947e-06,
+      "loss": 0.8327,
+      "step": 236
+    },
+    {
+      "epoch": 0.08214904679376084,
+      "grad_norm": 7.49241016550849,
+      "learning_rate": 9.930192239914733e-06,
+      "loss": 1.0304,
+      "step": 237
+    },
+    {
+      "epoch": 0.08249566724436742,
+      "grad_norm": 30.27089468614627,
+      "learning_rate": 9.929254301985119e-06,
+      "loss": 0.9654,
+      "step": 238
+    },
+    {
+      "epoch": 0.08284228769497401,
+      "grad_norm": 5.750266958378385,
+      "learning_rate": 9.928310149851541e-06,
+      "loss": 0.925,
+      "step": 239
+    },
+    {
+      "epoch": 0.0831889081455806,
+      "grad_norm": 5.753854129931636,
+      "learning_rate": 9.927359784704274e-06,
+      "loss": 0.987,
+      "step": 240
+    },
+    {
+      "epoch": 0.08353552859618718,
+      "grad_norm": 7.028260255730405,
+      "learning_rate": 9.92640320774142e-06,
+      "loss": 1.0824,
+      "step": 241
+    },
+    {
+      "epoch": 0.08388214904679377,
+      "grad_norm": 6.332764151242251,
+      "learning_rate": 9.925440420168917e-06,
+      "loss": 0.9068,
+      "step": 242
+    },
+    {
+      "epoch": 0.08422876949740035,
+      "grad_norm": 8.474324770591991,
+      "learning_rate": 9.924471423200527e-06,
+      "loss": 0.8261,
+      "step": 243
+    },
+    {
+      "epoch": 0.08457538994800694,
+      "grad_norm": 4.692499042482626,
+      "learning_rate": 9.923496218057846e-06,
+      "loss": 0.9399,
+      "step": 244
+    },
+    {
+      "epoch": 0.08492201039861352,
+      "grad_norm": 3.850816132646767,
+      "learning_rate": 9.922514805970293e-06,
+      "loss": 0.9499,
+      "step": 245
+    },
+    {
+      "epoch": 0.08526863084922011,
+      "grad_norm": 4.584545211398445,
+      "learning_rate": 9.921527188175115e-06,
+      "loss": 0.8225,
+      "step": 246
+    },
+    {
+      "epoch": 0.0856152512998267,
+      "grad_norm": 9.425533996302024,
+      "learning_rate": 9.920533365917379e-06,
+      "loss": 0.9203,
+      "step": 247
+    },
+    {
+      "epoch": 0.08596187175043328,
+      "grad_norm": 4.974570356345345,
+      "learning_rate": 9.919533340449974e-06,
+      "loss": 0.9994,
+      "step": 248
+    },
+    {
+      "epoch": 0.08630849220103987,
+      "grad_norm": 12.221918978405679,
+      "learning_rate": 9.918527113033612e-06,
+      "loss": 1.0773,
+      "step": 249
+    },
+    {
+      "epoch": 0.08665511265164645,
+      "grad_norm": 6.6525642970219545,
+      "learning_rate": 9.917514684936819e-06,
+      "loss": 0.9275,
+      "step": 250
+    },
+    {
+      "epoch": 0.08700173310225304,
+      "grad_norm": 6.023132012201478,
+      "learning_rate": 9.916496057435947e-06,
+      "loss": 0.8546,
+      "step": 251
+    },
+    {
+      "epoch": 0.08734835355285962,
+      "grad_norm": 13.607659736196476,
+      "learning_rate": 9.915471231815153e-06,
+      "loss": 0.82,
+      "step": 252
+    },
+    {
+      "epoch": 0.08769497400346621,
+      "grad_norm": 4.503005848183179,
+      "learning_rate": 9.914440209366415e-06,
+      "loss": 0.8482,
+      "step": 253
+    },
+    {
+      "epoch": 0.0880415944540728,
+      "grad_norm": 4.576941309677782,
+      "learning_rate": 9.913402991389516e-06,
+      "loss": 0.9583,
+      "step": 254
+    },
+    {
+      "epoch": 0.08838821490467938,
+      "grad_norm": 11.693399661867932,
+      "learning_rate": 9.912359579192059e-06,
+      "loss": 0.8112,
+      "step": 255
+    },
+    {
+      "epoch": 0.08873483535528597,
+      "grad_norm": 5.159200557667021,
+      "learning_rate": 9.911309974089449e-06,
+      "loss": 0.8505,
+      "step": 256
+    },
+    {
+      "epoch": 0.08908145580589255,
+      "grad_norm": 5.055966911981114,
+      "learning_rate": 9.910254177404901e-06,
+      "loss": 0.8884,
+      "step": 257
+    },
+    {
+      "epoch": 0.08942807625649914,
+      "grad_norm": 4.242248760098905,
+      "learning_rate": 9.909192190469437e-06,
+      "loss": 0.8132,
+      "step": 258
+    },
+    {
+      "epoch": 0.08977469670710572,
+      "grad_norm": 4.764359066894779,
+      "learning_rate": 9.908124014621876e-06,
+      "loss": 0.8885,
+      "step": 259
+    },
+    {
+      "epoch": 0.09012131715771231,
+      "grad_norm": 5.636548371945377,
+      "learning_rate": 9.90704965120885e-06,
+      "loss": 0.9501,
+      "step": 260
+    },
+    {
+      "epoch": 0.0904679376083189,
+      "grad_norm": 6.2833719462988675,
+      "learning_rate": 9.90596910158478e-06,
+      "loss": 0.8201,
+      "step": 261
+    },
+    {
+      "epoch": 0.09081455805892548,
+      "grad_norm": 5.409416941614016,
+      "learning_rate": 9.904882367111894e-06,
+      "loss": 0.791,
+      "step": 262
+    },
+    {
+      "epoch": 0.09116117850953206,
+      "grad_norm": 20.328683900610987,
+      "learning_rate": 9.903789449160215e-06,
+      "loss": 0.8829,
+      "step": 263
+    },
+    {
+      "epoch": 0.09150779896013865,
+      "grad_norm": 22.2148385328222,
+      "learning_rate": 9.90269034910756e-06,
+      "loss": 0.8731,
+      "step": 264
+    },
+    {
+      "epoch": 0.09185441941074524,
+      "grad_norm": 5.899483335842463,
+      "learning_rate": 9.901585068339542e-06,
+      "loss": 0.9079,
+      "step": 265
+    },
+    {
+      "epoch": 0.09220103986135182,
+      "grad_norm": 5.166252289947941,
+      "learning_rate": 9.900473608249566e-06,
+      "loss": 0.8739,
+      "step": 266
+    },
+    {
+      "epoch": 0.09254766031195841,
+      "grad_norm": 3.8984257783868177,
+      "learning_rate": 9.89935597023882e-06,
+      "loss": 0.9745,
+      "step": 267
+    },
+    {
+      "epoch": 0.09289428076256499,
+      "grad_norm": 4.783678958566108,
+      "learning_rate": 9.89823215571629e-06,
+      "loss": 0.736,
+      "step": 268
+    },
+    {
+      "epoch": 0.09324090121317158,
+      "grad_norm": 4.164755878761951,
+      "learning_rate": 9.897102166098744e-06,
+      "loss": 0.868,
+      "step": 269
+    },
+    {
+      "epoch": 0.09358752166377816,
+      "grad_norm": 3.781013894258776,
+      "learning_rate": 9.895966002810735e-06,
+      "loss": 0.8753,
+      "step": 270
+    },
+    {
+      "epoch": 0.09393414211438475,
+      "grad_norm": 12.921084780522774,
+      "learning_rate": 9.8948236672846e-06,
+      "loss": 0.7907,
+      "step": 271
+    },
+    {
+      "epoch": 0.09428076256499134,
+      "grad_norm": 11.593691780552392,
+      "learning_rate": 9.893675160960456e-06,
+      "loss": 0.84,
+      "step": 272
+    },
+    {
+      "epoch": 0.09462738301559792,
+      "grad_norm": 10.847542406998437,
+      "learning_rate": 9.8925204852862e-06,
+      "loss": 0.9405,
+      "step": 273
+    },
+    {
+      "epoch": 0.0949740034662045,
+      "grad_norm": 3.741299245772857,
+      "learning_rate": 9.891359641717508e-06,
+      "loss": 0.9546,
+      "step": 274
+    },
+    {
+      "epoch": 0.09532062391681109,
+      "grad_norm": 4.3873543040702465,
+      "learning_rate": 9.89019263171783e-06,
+      "loss": 0.8604,
+      "step": 275
+    },
+    {
+      "epoch": 0.09566724436741768,
+      "grad_norm": 5.47197700979158,
+      "learning_rate": 9.889019456758387e-06,
+      "loss": 0.9183,
+      "step": 276
+    },
+    {
+      "epoch": 0.09601386481802426,
+      "grad_norm": 4.318511163251727,
+      "learning_rate": 9.88784011831818e-06,
+      "loss": 0.8678,
+      "step": 277
+    },
+    {
+      "epoch": 0.09636048526863085,
+      "grad_norm": 5.585014482368944,
+      "learning_rate": 9.886654617883971e-06,
+      "loss": 0.7906,
+      "step": 278
+    },
+    {
+      "epoch": 0.09670710571923744,
+      "grad_norm": 5.66359289047795,
+      "learning_rate": 9.885462956950295e-06,
+      "loss": 0.7688,
+      "step": 279
+    },
+    {
+      "epoch": 0.09705372616984402,
+      "grad_norm": 4.969285702171553,
+      "learning_rate": 9.884265137019457e-06,
+      "loss": 0.9467,
+      "step": 280
+    },
+    {
+      "epoch": 0.0974003466204506,
+      "grad_norm": 6.408822174214393,
+      "learning_rate": 9.883061159601521e-06,
+      "loss": 0.8849,
+      "step": 281
+    },
+    {
+      "epoch": 0.09774696707105719,
+      "grad_norm": 3.5674918208836655,
+      "learning_rate": 9.881851026214315e-06,
+      "loss": 0.9078,
+      "step": 282
+    },
+    {
+      "epoch": 0.09809358752166378,
+      "grad_norm": 128.23071918047899,
+      "learning_rate": 9.880634738383426e-06,
+      "loss": 0.971,
+      "step": 283
+    },
+    {
+      "epoch": 0.09844020797227036,
+      "grad_norm": 6.049960512121416,
+      "learning_rate": 9.879412297642202e-06,
+      "loss": 0.7585,
+      "step": 284
+    },
+    {
+      "epoch": 0.09878682842287695,
+      "grad_norm": 3.859086835545687,
+      "learning_rate": 9.87818370553175e-06,
+      "loss": 0.8498,
+      "step": 285
+    },
+    {
+      "epoch": 0.09913344887348353,
+      "grad_norm": 11.003833490416248,
+      "learning_rate": 9.876948963600927e-06,
+      "loss": 0.9073,
+      "step": 286
+    },
+    {
+      "epoch": 0.09948006932409012,
+      "grad_norm": 4.51278780241575,
+      "learning_rate": 9.875708073406348e-06,
+      "loss": 0.8559,
+      "step": 287
+    },
+    {
+      "epoch": 0.0998266897746967,
+      "grad_norm": 160.12962146059328,
+      "learning_rate": 9.874461036512375e-06,
+      "loss": 1.5114,
+      "step": 288
+    },
+    {
+      "epoch": 0.10017331022530329,
+      "grad_norm": 6.603994681561948,
+      "learning_rate": 9.873207854491118e-06,
+      "loss": 0.9038,
+      "step": 289
+    },
+    {
+      "epoch": 0.10051993067590988,
+      "grad_norm": 6.596000847513638,
+      "learning_rate": 9.87194852892244e-06,
+      "loss": 0.8397,
+      "step": 290
+    },
+    {
+      "epoch": 0.10086655112651646,
+      "grad_norm": 4.986704433712958,
+      "learning_rate": 9.870683061393945e-06,
+      "loss": 0.9513,
+      "step": 291
+    },
+    {
+      "epoch": 0.10121317157712305,
+      "grad_norm": 14.151787471568616,
+      "learning_rate": 9.869411453500978e-06,
+      "loss": 0.951,
+      "step": 292
+    },
+    {
+      "epoch": 0.10155979202772963,
+      "grad_norm": 6.968871622602454,
+      "learning_rate": 9.86813370684663e-06,
+      "loss": 0.9818,
+      "step": 293
+    },
+    {
+      "epoch": 0.10190641247833622,
+      "grad_norm": 5.140207058583939,
+      "learning_rate": 9.866849823041724e-06,
+      "loss": 0.7592,
+      "step": 294
+    },
+    {
+      "epoch": 0.1022530329289428,
+      "grad_norm": 7.275304128655026,
+      "learning_rate": 9.865559803704832e-06,
+      "loss": 0.9728,
+      "step": 295
+    },
+    {
+      "epoch": 0.10259965337954939,
+      "grad_norm": 5.317022242793877,
+      "learning_rate": 9.864263650462247e-06,
+      "loss": 0.8217,
+      "step": 296
+    },
+    {
+      "epoch": 0.10294627383015598,
+      "grad_norm": 15.97954069585634,
+      "learning_rate": 9.862961364948005e-06,
+      "loss": 0.8292,
+      "step": 297
+    },
+    {
+      "epoch": 0.10329289428076256,
+      "grad_norm": 47.353200317685925,
+      "learning_rate": 9.861652948803866e-06,
+      "loss": 0.7884,
+      "step": 298
+    },
+    {
+      "epoch": 0.10363951473136915,
+      "grad_norm": 7.843341576614204,
+      "learning_rate": 9.860338403679323e-06,
+      "loss": 0.9103,
+      "step": 299
+    },
+    {
+      "epoch": 0.10398613518197573,
+      "grad_norm": 4.4522655237508415,
+      "learning_rate": 9.859017731231597e-06,
+      "loss": 0.7231,
+      "step": 300
+    },
+    {
+      "epoch": 0.10433275563258232,
+      "grad_norm": 17.245345096509755,
+      "learning_rate": 9.857690933125628e-06,
+      "loss": 0.8121,
+      "step": 301
+    },
+    {
+      "epoch": 0.1046793760831889,
+      "grad_norm": 4.939783343157512,
+      "learning_rate": 9.856358011034083e-06,
+      "loss": 0.8744,
+      "step": 302
+    },
+    {
+      "epoch": 0.10502599653379549,
+      "grad_norm": 7.22485789834502,
+      "learning_rate": 9.855018966637346e-06,
+      "loss": 0.9005,
+      "step": 303
+    },
+    {
+      "epoch": 0.10537261698440208,
+      "grad_norm": 4.099236064974072,
+      "learning_rate": 9.853673801623526e-06,
+      "loss": 0.9054,
+      "step": 304
+    },
+    {
+      "epoch": 0.10571923743500866,
+      "grad_norm": 5.657665024003485,
+      "learning_rate": 9.852322517688437e-06,
+      "loss": 0.8918,
+      "step": 305
+    },
+    {
+      "epoch": 0.10606585788561525,
+      "grad_norm": 4.579286735791722,
+      "learning_rate": 9.850965116535618e-06,
+      "loss": 0.8214,
+      "step": 306
+    },
+    {
+      "epoch": 0.10641247833622183,
+      "grad_norm": 3.479382933377304,
+      "learning_rate": 9.849601599876315e-06,
+      "loss": 0.9229,
+      "step": 307
+    },
+    {
+      "epoch": 0.10675909878682842,
+      "grad_norm": 3.833247122025002,
+      "learning_rate": 9.84823196942948e-06,
+      "loss": 1.0045,
+      "step": 308
+    },
+    {
+      "epoch": 0.107105719237435,
+      "grad_norm": 4.117961927667102,
+      "learning_rate": 9.846856226921782e-06,
+      "loss": 0.7634,
+      "step": 309
+    },
+    {
+      "epoch": 0.10745233968804159,
+      "grad_norm": 12.164068505092871,
+      "learning_rate": 9.845474374087584e-06,
+      "loss": 0.9649,
+      "step": 310
+    },
+    {
+      "epoch": 0.10779896013864818,
+      "grad_norm": 4.639152300523412,
+      "learning_rate": 9.844086412668963e-06,
+      "loss": 0.8452,
+      "step": 311
+    },
+    {
+      "epoch": 0.10814558058925476,
+      "grad_norm": 3.86851167800573,
+      "learning_rate": 9.842692344415688e-06,
+      "loss": 0.8762,
+      "step": 312
+    },
+    {
+      "epoch": 0.10849220103986135,
+      "grad_norm": 9.745676327281428,
+      "learning_rate": 9.841292171085232e-06,
+      "loss": 0.8646,
+      "step": 313
+    },
+    {
+      "epoch": 0.10883882149046793,
+      "grad_norm": 5.360994003405478,
+      "learning_rate": 9.839885894442763e-06,
+      "loss": 0.953,
+      "step": 314
+    },
+    {
+      "epoch": 0.10918544194107452,
+      "grad_norm": 9.896024052610795,
+      "learning_rate": 9.838473516261144e-06,
+      "loss": 0.8807,
+      "step": 315
+    },
+    {
+      "epoch": 0.1095320623916811,
+      "grad_norm": 4.222591592651269,
+      "learning_rate": 9.83705503832093e-06,
+      "loss": 0.7793,
+      "step": 316
+    },
+    {
+      "epoch": 0.10987868284228769,
+      "grad_norm": 4.876017561391574,
+      "learning_rate": 9.835630462410363e-06,
+      "loss": 0.915,
+      "step": 317
+    },
+    {
+      "epoch": 0.11022530329289427,
+      "grad_norm": 3.6308374950067366,
+      "learning_rate": 9.834199790325378e-06,
+      "loss": 0.8504,
+      "step": 318
+    },
+    {
+      "epoch": 0.11057192374350086,
+      "grad_norm": 10.247859144676358,
+      "learning_rate": 9.832763023869592e-06,
+      "loss": 0.7913,
+      "step": 319
+    },
+    {
+      "epoch": 0.11091854419410745,
+      "grad_norm": 6.031378269390875,
+      "learning_rate": 9.8313201648543e-06,
+      "loss": 0.9686,
+      "step": 320
+    },
+    {
+      "epoch": 0.11126516464471403,
+      "grad_norm": 5.702358261961062,
+      "learning_rate": 9.82987121509849e-06,
+      "loss": 0.9374,
+      "step": 321
+    },
+    {
+      "epoch": 0.11161178509532062,
+      "grad_norm": 2.8418095361988147,
+      "learning_rate": 9.828416176428819e-06,
+      "loss": 0.852,
+      "step": 322
+    },
+    {
+      "epoch": 0.1119584055459272,
+      "grad_norm": 4.444882686325085,
+      "learning_rate": 9.82695505067962e-06,
+      "loss": 0.9356,
+      "step": 323
+    },
+    {
+      "epoch": 0.11230502599653379,
+      "grad_norm": 4.376674962931048,
+      "learning_rate": 9.825487839692907e-06,
+      "loss": 0.9113,
+      "step": 324
+    },
+    {
+      "epoch": 0.11265164644714037,
+      "grad_norm": 3.8028743505729756,
+      "learning_rate": 9.824014545318359e-06,
+      "loss": 0.7406,
+      "step": 325
+    },
+    {
+      "epoch": 0.11299826689774697,
+      "grad_norm": 5.366151246998004,
+      "learning_rate": 9.822535169413325e-06,
+      "loss": 0.8962,
+      "step": 326
+    },
+    {
+      "epoch": 0.11334488734835356,
+      "grad_norm": 7.88357629131342,
+      "learning_rate": 9.821049713842824e-06,
+      "loss": 0.7851,
+      "step": 327
+    },
+    {
+      "epoch": 0.11369150779896015,
+      "grad_norm": 5.231419072131993,
+      "learning_rate": 9.819558180479538e-06,
+      "loss": 0.8638,
+      "step": 328
+    },
+    {
+      "epoch": 0.11403812824956673,
+      "grad_norm": 7.8273682169456,
+      "learning_rate": 9.818060571203807e-06,
+      "loss": 0.9453,
+      "step": 329
+    },
+    {
+      "epoch": 0.11438474870017332,
+      "grad_norm": 4.450910170011337,
+      "learning_rate": 9.816556887903639e-06,
+      "loss": 0.8797,
+      "step": 330
+    },
+    {
+      "epoch": 0.1147313691507799,
+      "grad_norm": 5.091319407302553,
+      "learning_rate": 9.815047132474693e-06,
+      "loss": 0.9707,
+      "step": 331
+    },
+    {
+      "epoch": 0.11507798960138649,
+      "grad_norm": 3.7031323039580917,
+      "learning_rate": 9.813531306820284e-06,
+      "loss": 0.8831,
+      "step": 332
+    },
+    {
+      "epoch": 0.11542461005199307,
+      "grad_norm": 6.490571768479035,
+      "learning_rate": 9.812009412851382e-06,
+      "loss": 0.7311,
+      "step": 333
+    },
+    {
+      "epoch": 0.11577123050259966,
+      "grad_norm": 6.567230689198332,
+      "learning_rate": 9.810481452486606e-06,
+      "loss": 0.9197,
+      "step": 334
+    },
+    {
+      "epoch": 0.11611785095320624,
+      "grad_norm": 5.497839181786425,
+      "learning_rate": 9.808947427652223e-06,
+      "loss": 0.9303,
+      "step": 335
+    },
+    {
+      "epoch": 0.11646447140381283,
+      "grad_norm": 4.232263826153935,
+      "learning_rate": 9.80740734028214e-06,
+      "loss": 1.0051,
+      "step": 336
+    },
+    {
+      "epoch": 0.11681109185441942,
+      "grad_norm": 4.272395860446608,
+      "learning_rate": 9.805861192317918e-06,
+      "loss": 0.8154,
+      "step": 337
+    },
+    {
+      "epoch": 0.117157712305026,
+      "grad_norm": 4.690511309927586,
+      "learning_rate": 9.804308985708748e-06,
+      "loss": 0.8985,
+      "step": 338
+    },
+    {
+      "epoch": 0.11750433275563259,
+      "grad_norm": 4.83573927323973,
+      "learning_rate": 9.802750722411465e-06,
+      "loss": 0.9745,
+      "step": 339
+    },
+    {
+      "epoch": 0.11785095320623917,
+      "grad_norm": 6.446592093759288,
+      "learning_rate": 9.80118640439054e-06,
+      "loss": 0.8724,
+      "step": 340
+    },
+    {
+      "epoch": 0.11819757365684576,
+      "grad_norm": 4.338927704547092,
+      "learning_rate": 9.799616033618069e-06,
+      "loss": 0.8613,
+      "step": 341
+    },
+    {
+      "epoch": 0.11854419410745234,
+      "grad_norm": 4.847085256611815,
+      "learning_rate": 9.798039612073789e-06,
+      "loss": 0.8381,
+      "step": 342
+    },
+    {
+      "epoch": 0.11889081455805893,
+      "grad_norm": 3.9017969244964257,
+      "learning_rate": 9.796457141745057e-06,
+      "loss": 0.8398,
+      "step": 343
+    },
+    {
+      "epoch": 0.11923743500866552,
+      "grad_norm": 9.271431982723342,
+      "learning_rate": 9.794868624626863e-06,
+      "loss": 0.8719,
+      "step": 344
+    },
+    {
+      "epoch": 0.1195840554592721,
+      "grad_norm": 4.900623300892105,
+      "learning_rate": 9.793274062721814e-06,
+      "loss": 0.8036,
+      "step": 345
+    },
+    {
+      "epoch": 0.11993067590987869,
+      "grad_norm": 4.33756586960525,
+      "learning_rate": 9.79167345804014e-06,
+      "loss": 0.9719,
+      "step": 346
+    },
+    {
+      "epoch": 0.12027729636048527,
+      "grad_norm": 2.9284336999170164,
+      "learning_rate": 9.790066812599687e-06,
+      "loss": 0.9358,
+      "step": 347
+    },
+    {
+      "epoch": 0.12062391681109186,
+      "grad_norm": 6.689181810098891,
+      "learning_rate": 9.788454128425921e-06,
+      "loss": 0.827,
+      "step": 348
+    },
+    {
+      "epoch": 0.12097053726169844,
+      "grad_norm": 4.890161030023105,
+      "learning_rate": 9.786835407551915e-06,
+      "loss": 0.8867,
+      "step": 349
+    },
+    {
+      "epoch": 0.12131715771230503,
+      "grad_norm": 4.335726830921586,
+      "learning_rate": 9.785210652018358e-06,
+      "loss": 0.8037,
+      "step": 350
+    },
+    {
+      "epoch": 0.12166377816291161,
+      "grad_norm": 3.566002861195123,
+      "learning_rate": 9.783579863873543e-06,
+      "loss": 0.7532,
+      "step": 351
+    },
+    {
+      "epoch": 0.1220103986135182,
+      "grad_norm": 4.782892936454329,
+      "learning_rate": 9.781943045173368e-06,
+      "loss": 0.8477,
+      "step": 352
+    },
+    {
+      "epoch": 0.12235701906412479,
+      "grad_norm": 3.982660285310167,
+      "learning_rate": 9.780300197981334e-06,
+      "loss": 0.8027,
+      "step": 353
+    },
+    {
+      "epoch": 0.12270363951473137,
+      "grad_norm": 6.623353059750746,
+      "learning_rate": 9.778651324368546e-06,
+      "loss": 0.7175,
+      "step": 354
+    },
+    {
+      "epoch": 0.12305025996533796,
+      "grad_norm": 8.170454600725773,
+      "learning_rate": 9.776996426413703e-06,
+      "loss": 0.944,
+      "step": 355
+    },
+    {
+      "epoch": 0.12339688041594454,
+      "grad_norm": 23.87061198044083,
+      "learning_rate": 9.775335506203097e-06,
+      "loss": 0.8967,
+      "step": 356
+    },
+    {
+      "epoch": 0.12374350086655113,
+      "grad_norm": 3.3122117322605864,
+      "learning_rate": 9.773668565830613e-06,
+      "loss": 0.7265,
+      "step": 357
+    },
+    {
+      "epoch": 0.12409012131715771,
+      "grad_norm": 3.4658127206701925,
+      "learning_rate": 9.77199560739773e-06,
+      "loss": 0.8605,
+      "step": 358
+    },
+    {
+      "epoch": 0.1244367417677643,
+      "grad_norm": 5.840513660217553,
+      "learning_rate": 9.770316633013508e-06,
+      "loss": 0.7202,
+      "step": 359
+    },
+    {
+      "epoch": 0.12478336221837089,
+      "grad_norm": 19.95314370886294,
+      "learning_rate": 9.768631644794595e-06,
+      "loss": 0.7472,
+      "step": 360
+    },
+    {
+      "epoch": 0.12512998266897746,
+      "grad_norm": 13.159860568053347,
+      "learning_rate": 9.766940644865216e-06,
+      "loss": 0.9189,
+      "step": 361
+    },
+    {
+      "epoch": 0.12547660311958406,
+      "grad_norm": 4.827198790126389,
+      "learning_rate": 9.765243635357185e-06,
+      "loss": 0.8089,
+      "step": 362
+    },
+    {
+      "epoch": 0.12582322357019063,
+      "grad_norm": 5.414123688991217,
+      "learning_rate": 9.763540618409876e-06,
+      "loss": 0.7938,
+      "step": 363
+    },
+    {
+      "epoch": 0.12616984402079723,
+      "grad_norm": 4.965162748198596,
+      "learning_rate": 9.761831596170252e-06,
+      "loss": 0.8101,
+      "step": 364
+    },
+    {
+      "epoch": 0.1265164644714038,
+      "grad_norm": 6.312609280171994,
+      "learning_rate": 9.760116570792839e-06,
+      "loss": 0.9829,
+      "step": 365
+    },
+    {
+      "epoch": 0.1268630849220104,
+      "grad_norm": 4.454995766787344,
+      "learning_rate": 9.75839554443973e-06,
+      "loss": 0.9319,
+      "step": 366
+    },
+    {
+      "epoch": 0.12720970537261697,
+      "grad_norm": 12.486574741113571,
+      "learning_rate": 9.756668519280587e-06,
+      "loss": 0.8174,
+      "step": 367
+    },
+    {
+      "epoch": 0.12755632582322357,
+      "grad_norm": 5.016418105011129,
+      "learning_rate": 9.754935497492634e-06,
+      "loss": 0.761,
+      "step": 368
+    },
+    {
+      "epoch": 0.12790294627383014,
+      "grad_norm": 4.620912339603629,
+      "learning_rate": 9.753196481260654e-06,
+      "loss": 0.894,
+      "step": 369
+    },
+    {
+      "epoch": 0.12824956672443674,
+      "grad_norm": 4.390737768437102,
+      "learning_rate": 9.751451472776983e-06,
+      "loss": 0.8048,
+      "step": 370
+    },
+    {
+      "epoch": 0.1285961871750433,
+      "grad_norm": 3.6999715615822364,
+      "learning_rate": 9.749700474241522e-06,
+      "loss": 0.8326,
+      "step": 371
+    },
+    {
+      "epoch": 0.1289428076256499,
+      "grad_norm": 11.006109075954225,
+      "learning_rate": 9.74794348786171e-06,
+      "loss": 0.908,
+      "step": 372
+    },
+    {
+      "epoch": 0.1292894280762565,
+      "grad_norm": 6.417674917275881,
+      "learning_rate": 9.746180515852549e-06,
+      "loss": 0.7921,
+      "step": 373
+    },
+    {
+      "epoch": 0.12963604852686308,
+      "grad_norm": 3.4800276179524405,
+      "learning_rate": 9.744411560436572e-06,
+      "loss": 0.7403,
+      "step": 374
+    },
+    {
+      "epoch": 0.12998266897746968,
+      "grad_norm": 4.856171851368818,
+      "learning_rate": 9.74263662384387e-06,
+      "loss": 0.9516,
+      "step": 375
+    },
+    {
+      "epoch": 0.13032928942807626,
+      "grad_norm": 7.033642975005388,
+      "learning_rate": 9.74085570831206e-06,
+      "loss": 0.8553,
+      "step": 376
+    },
+    {
+      "epoch": 0.13067590987868286,
+      "grad_norm": 4.203841749038082,
+      "learning_rate": 9.73906881608631e-06,
+      "loss": 0.8366,
+      "step": 377
+    },
+    {
+      "epoch": 0.13102253032928943,
+      "grad_norm": 13.73295532603178,
+      "learning_rate": 9.73727594941931e-06,
+      "loss": 0.7827,
+      "step": 378
+    },
+    {
+      "epoch": 0.13136915077989603,
+      "grad_norm": 67.35488443253173,
+      "learning_rate": 9.735477110571294e-06,
+      "loss": 0.8826,
+      "step": 379
+    },
+    {
+      "epoch": 0.1317157712305026,
+      "grad_norm": 8.367752829409739,
+      "learning_rate": 9.733672301810016e-06,
+      "loss": 0.9922,
+      "step": 380
+    },
+    {
+      "epoch": 0.1320623916811092,
+      "grad_norm": 10.291907525166378,
+      "learning_rate": 9.731861525410758e-06,
+      "loss": 0.825,
+      "step": 381
+    },
+    {
+      "epoch": 0.13240901213171577,
+      "grad_norm": 4.537051980012283,
+      "learning_rate": 9.73004478365633e-06,
+      "loss": 0.8405,
+      "step": 382
+    },
+    {
+      "epoch": 0.13275563258232237,
+      "grad_norm": 4.166247144567635,
+      "learning_rate": 9.728222078837056e-06,
+      "loss": 0.7571,
+      "step": 383
+    },
+    {
+      "epoch": 0.13310225303292894,
+      "grad_norm": 4.532139879256794,
+      "learning_rate": 9.726393413250781e-06,
+      "loss": 0.8611,
+      "step": 384
+    },
+    {
+      "epoch": 0.13344887348353554,
+      "grad_norm": 4.483738051611331,
+      "learning_rate": 9.724558789202867e-06,
+      "loss": 0.8523,
+      "step": 385
+    },
+    {
+      "epoch": 0.1337954939341421,
+      "grad_norm": 5.283025545423599,
+      "learning_rate": 9.722718209006181e-06,
+      "loss": 0.8257,
+      "step": 386
+    },
+    {
+      "epoch": 0.1341421143847487,
+      "grad_norm": 5.099156168809862,
+      "learning_rate": 9.720871674981102e-06,
+      "loss": 0.7261,
+      "step": 387
+    },
+    {
+      "epoch": 0.13448873483535528,
+      "grad_norm": 5.694141635753971,
+      "learning_rate": 9.71901918945552e-06,
+      "loss": 0.8455,
+      "step": 388
+    },
+    {
+      "epoch": 0.13483535528596188,
+      "grad_norm": 10.84031270205953,
+      "learning_rate": 9.717160754764821e-06,
+      "loss": 0.7924,
+      "step": 389
+    },
+    {
+      "epoch": 0.13518197573656845,
+      "grad_norm": 3.4923983557817784,
+      "learning_rate": 9.715296373251891e-06,
+      "loss": 0.7612,
+      "step": 390
+    },
+    {
+      "epoch": 0.13552859618717505,
+      "grad_norm": 3.047166247254343,
+      "learning_rate": 9.713426047267117e-06,
+      "loss": 0.8382,
+      "step": 391
+    },
+    {
+      "epoch": 0.13587521663778163,
+      "grad_norm": 7.8354926754777106,
+      "learning_rate": 9.711549779168379e-06,
+      "loss": 0.8414,
+      "step": 392
+    },
+    {
+      "epoch": 0.13622183708838823,
+      "grad_norm": 9.551631461723723,
+      "learning_rate": 9.709667571321049e-06,
+      "loss": 0.8524,
+      "step": 393
+    },
+    {
+      "epoch": 0.1365684575389948,
+      "grad_norm": 4.865692549216927,
+      "learning_rate": 9.70777942609798e-06,
+      "loss": 0.8219,
+      "step": 394
+    },
+    {
+      "epoch": 0.1369150779896014,
+      "grad_norm": 3.260514698101462,
+      "learning_rate": 9.705885345879523e-06,
+      "loss": 0.7548,
+      "step": 395
+    },
+    {
+      "epoch": 0.13726169844020797,
+      "grad_norm": 3.4579320547103767,
+      "learning_rate": 9.703985333053498e-06,
+      "loss": 0.8335,
+      "step": 396
+    },
+    {
+      "epoch": 0.13760831889081457,
+      "grad_norm": 5.17540000091191,
+      "learning_rate": 9.702079390015213e-06,
+      "loss": 0.944,
+      "step": 397
+    },
+    {
+      "epoch": 0.13795493934142114,
+      "grad_norm": 5.619588941390446,
+      "learning_rate": 9.700167519167448e-06,
+      "loss": 0.8861,
+      "step": 398
+    },
+    {
+      "epoch": 0.13830155979202774,
+      "grad_norm": 4.275758391911119,
+      "learning_rate": 9.698249722920456e-06,
+      "loss": 0.974,
+      "step": 399
+    },
+    {
+      "epoch": 0.1386481802426343,
+      "grad_norm": 4.1514463936874,
+      "learning_rate": 9.696326003691964e-06,
+      "loss": 0.8397,
+      "step": 400
+    },
+    {
+      "epoch": 0.1389948006932409,
+      "grad_norm": 7.727731230661998,
+      "learning_rate": 9.69439636390716e-06,
+      "loss": 0.8563,
+      "step": 401
+    },
+    {
+      "epoch": 0.13934142114384748,
+      "grad_norm": 6.24855441944965,
+      "learning_rate": 9.692460805998703e-06,
+      "loss": 0.8237,
+      "step": 402
+    },
+    {
+      "epoch": 0.13968804159445408,
+      "grad_norm": 5.344414619677134,
+      "learning_rate": 9.690519332406706e-06,
+      "loss": 0.8639,
+      "step": 403
+    },
+    {
+      "epoch": 0.14003466204506065,
+      "grad_norm": 6.585907717937583,
+      "learning_rate": 9.688571945578742e-06,
+      "loss": 0.864,
+      "step": 404
+    },
+    {
+      "epoch": 0.14038128249566725,
+      "grad_norm": 4.827205394418977,
+      "learning_rate": 9.686618647969841e-06,
+      "loss": 0.9067,
+      "step": 405
+    },
+    {
+      "epoch": 0.14072790294627383,
+      "grad_norm": 2.9269348211525323,
+      "learning_rate": 9.684659442042482e-06,
+      "loss": 0.854,
+      "step": 406
+    },
+    {
+      "epoch": 0.14107452339688042,
+      "grad_norm": 14.792834493349774,
+      "learning_rate": 9.682694330266593e-06,
+      "loss": 0.8688,
+      "step": 407
+    },
+    {
+      "epoch": 0.141421143847487,
+      "grad_norm": 4.836061420572961,
+      "learning_rate": 9.68072331511955e-06,
+      "loss": 0.8685,
+      "step": 408
+    },
+    {
+      "epoch": 0.1417677642980936,
+      "grad_norm": 4.515342018415354,
+      "learning_rate": 9.678746399086167e-06,
+      "loss": 0.8929,
+      "step": 409
+    },
+    {
+      "epoch": 0.14211438474870017,
+      "grad_norm": 3.7360001863946,
+      "learning_rate": 9.676763584658699e-06,
+      "loss": 0.6826,
+      "step": 410
+    },
+    {
+      "epoch": 0.14246100519930677,
+      "grad_norm": 6.075263811330759,
+      "learning_rate": 9.674774874336839e-06,
+      "loss": 0.7316,
+      "step": 411
+    },
+    {
+      "epoch": 0.14280762564991334,
+      "grad_norm": 2.71257591077479,
+      "learning_rate": 9.672780270627708e-06,
+      "loss": 0.8107,
+      "step": 412
+    },
+    {
+      "epoch": 0.14315424610051994,
+      "grad_norm": 3.499639644376886,
+      "learning_rate": 9.670779776045864e-06,
+      "loss": 0.713,
+      "step": 413
+    },
+    {
+      "epoch": 0.1435008665511265,
+      "grad_norm": 4.2006708605814955,
+      "learning_rate": 9.668773393113282e-06,
+      "loss": 0.7734,
+      "step": 414
+    },
+    {
+      "epoch": 0.1438474870017331,
+      "grad_norm": 3.751510075671731,
+      "learning_rate": 9.66676112435937e-06,
+      "loss": 0.8847,
+      "step": 415
+    },
+    {
+      "epoch": 0.14419410745233968,
+      "grad_norm": 5.006111869927531,
+      "learning_rate": 9.664742972320949e-06,
+      "loss": 0.846,
+      "step": 416
+    },
+    {
+      "epoch": 0.14454072790294628,
+      "grad_norm": 3.6154671147927613,
+      "learning_rate": 9.66271893954226e-06,
+      "loss": 0.7825,
+      "step": 417
+    },
+    {
+      "epoch": 0.14488734835355285,
+      "grad_norm": 3.8141676440838177,
+      "learning_rate": 9.660689028574955e-06,
+      "loss": 0.6949,
+      "step": 418
+    },
+    {
+      "epoch": 0.14523396880415945,
+      "grad_norm": 4.418508598794215,
+      "learning_rate": 9.658653241978103e-06,
+      "loss": 0.7528,
+      "step": 419
+    },
+    {
+      "epoch": 0.14558058925476602,
+      "grad_norm": 6.039938328253148,
+      "learning_rate": 9.656611582318173e-06,
+      "loss": 0.9021,
+      "step": 420
+    },
+    {
+      "epoch": 0.14592720970537262,
+      "grad_norm": 5.240180600916672,
+      "learning_rate": 9.654564052169044e-06,
+      "loss": 0.8341,
+      "step": 421
+    },
+    {
+      "epoch": 0.1462738301559792,
+      "grad_norm": 4.230053996199884,
+      "learning_rate": 9.65251065411199e-06,
+      "loss": 0.778,
+      "step": 422
+    },
+    {
+      "epoch": 0.1466204506065858,
+      "grad_norm": 4.0501199990905254,
+      "learning_rate": 9.650451390735685e-06,
+      "loss": 0.8184,
+      "step": 423
+    },
+    {
+      "epoch": 0.14696707105719237,
+      "grad_norm": 3.7098275646057846,
+      "learning_rate": 9.6483862646362e-06,
+      "loss": 0.8549,
+      "step": 424
+    },
+    {
+      "epoch": 0.14731369150779897,
+      "grad_norm": 5.735741367984041,
+      "learning_rate": 9.646315278416993e-06,
+      "loss": 0.7534,
+      "step": 425
+    },
+    {
+      "epoch": 0.14766031195840554,
+      "grad_norm": 2.791634290920191,
+      "learning_rate": 9.644238434688913e-06,
+      "loss": 0.9053,
+      "step": 426
+    },
+    {
+      "epoch": 0.14800693240901214,
+      "grad_norm": 9.528025251807803,
+      "learning_rate": 9.64215573607019e-06,
+      "loss": 0.7455,
+      "step": 427
+    },
+    {
+      "epoch": 0.1483535528596187,
+      "grad_norm": 7.058995215433996,
+      "learning_rate": 9.640067185186436e-06,
+      "loss": 0.8538,
+      "step": 428
+    },
+    {
+      "epoch": 0.1487001733102253,
+      "grad_norm": 2.8839240501303314,
+      "learning_rate": 9.637972784670648e-06,
+      "loss": 0.7574,
+      "step": 429
+    },
+    {
+      "epoch": 0.14904679376083188,
+      "grad_norm": 10.73722678726303,
+      "learning_rate": 9.635872537163183e-06,
+      "loss": 0.8607,
+      "step": 430
+    },
+    {
+      "epoch": 0.14939341421143848,
+      "grad_norm": 4.953586189255954,
+      "learning_rate": 9.633766445311783e-06,
+      "loss": 0.8434,
+      "step": 431
+    },
+    {
+      "epoch": 0.14974003466204505,
+      "grad_norm": 9.11146454893727,
+      "learning_rate": 9.631654511771549e-06,
+      "loss": 0.8888,
+      "step": 432
+    },
+    {
+      "epoch": 0.15008665511265165,
+      "grad_norm": 4.140584926116786,
+      "learning_rate": 9.629536739204955e-06,
+      "loss": 0.8787,
+      "step": 433
+    },
+    {
+      "epoch": 0.15043327556325822,
+      "grad_norm": 4.751520174928703,
+      "learning_rate": 9.627413130281823e-06,
+      "loss": 0.8299,
+      "step": 434
+    },
+    {
+      "epoch": 0.15077989601386482,
+      "grad_norm": 4.129752741586027,
+      "learning_rate": 9.625283687679347e-06,
+      "loss": 0.8147,
+      "step": 435
+    },
+    {
+      "epoch": 0.1511265164644714,
+      "grad_norm": 5.301103534216404,
+      "learning_rate": 9.623148414082067e-06,
+      "loss": 0.8011,
+      "step": 436
+    },
+    {
+      "epoch": 0.151473136915078,
+      "grad_norm": 3.4598532407392386,
+      "learning_rate": 9.621007312181878e-06,
+      "loss": 1.0039,
+      "step": 437
+    },
+    {
+      "epoch": 0.15181975736568457,
+      "grad_norm": 9.95001292132511,
+      "learning_rate": 9.618860384678017e-06,
+      "loss": 0.7014,
+      "step": 438
+    },
+    {
+      "epoch": 0.15216637781629117,
+      "grad_norm": 3.1529611714364236,
+      "learning_rate": 9.61670763427707e-06,
+      "loss": 0.8759,
+      "step": 439
+    },
+    {
+      "epoch": 0.15251299826689774,
+      "grad_norm": 3.2285067819679174,
+      "learning_rate": 9.614549063692966e-06,
+      "loss": 0.8444,
+      "step": 440
+    },
+    {
+      "epoch": 0.15285961871750434,
+      "grad_norm": 3.097786698139052,
+      "learning_rate": 9.612384675646965e-06,
+      "loss": 0.8862,
+      "step": 441
+    },
+    {
+      "epoch": 0.1532062391681109,
+      "grad_norm": 3.8514885734799518,
+      "learning_rate": 9.610214472867662e-06,
+      "loss": 0.9347,
+      "step": 442
+    },
+    {
+      "epoch": 0.1535528596187175,
+      "grad_norm": 49.17593594334729,
+      "learning_rate": 9.60803845809099e-06,
+      "loss": 0.869,
+      "step": 443
+    },
+    {
+      "epoch": 0.15389948006932408,
+      "grad_norm": 4.664026698322477,
+      "learning_rate": 9.605856634060203e-06,
+      "loss": 0.7389,
+      "step": 444
+    },
+    {
+      "epoch": 0.15424610051993068,
+      "grad_norm": 8.696776877562456,
+      "learning_rate": 9.603669003525872e-06,
+      "loss": 0.9661,
+      "step": 445
+    },
+    {
+      "epoch": 0.15459272097053725,
+      "grad_norm": 3.5698121906512736,
+      "learning_rate": 9.601475569245905e-06,
+      "loss": 0.7722,
+      "step": 446
+    },
+    {
+      "epoch": 0.15493934142114385,
+      "grad_norm": 3.6544436873810193,
+      "learning_rate": 9.59927633398551e-06,
+      "loss": 0.8701,
+      "step": 447
+    },
+    {
+      "epoch": 0.15528596187175042,
+      "grad_norm": 4.658991752661569,
+      "learning_rate": 9.59707130051722e-06,
+      "loss": 0.7968,
+      "step": 448
+    },
+    {
+      "epoch": 0.15563258232235702,
+      "grad_norm": 9.376931655217444,
+      "learning_rate": 9.594860471620868e-06,
+      "loss": 0.731,
+      "step": 449
+    },
+    {
+      "epoch": 0.1559792027729636,
+      "grad_norm": 3.9472435226158464,
+      "learning_rate": 9.592643850083602e-06,
+      "loss": 0.8201,
+      "step": 450
+    },
+    {
+      "epoch": 0.1563258232235702,
+      "grad_norm": 8.06378351186663,
+      "learning_rate": 9.59042143869987e-06,
+      "loss": 0.7573,
+      "step": 451
+    },
+    {
+      "epoch": 0.15667244367417676,
+      "grad_norm": 4.922433097754393,
+      "learning_rate": 9.58819324027141e-06,
+      "loss": 0.8618,
+      "step": 452
+    },
+    {
+      "epoch": 0.15701906412478336,
+      "grad_norm": 6.085569592612043,
+      "learning_rate": 9.585959257607272e-06,
+      "loss": 0.9485,
+      "step": 453
+    },
+    {
+      "epoch": 0.15736568457538994,
+      "grad_norm": 5.267602986297629,
+      "learning_rate": 9.583719493523784e-06,
+      "loss": 0.8984,
+      "step": 454
+    },
+    {
+      "epoch": 0.15771230502599654,
+      "grad_norm": 6.004697462757485,
+      "learning_rate": 9.581473950844571e-06,
+      "loss": 0.8915,
+      "step": 455
+    },
+    {
+      "epoch": 0.1580589254766031,
+      "grad_norm": 5.298501408984141,
+      "learning_rate": 9.57922263240054e-06,
+      "loss": 0.8601,
+      "step": 456
+    },
+    {
+      "epoch": 0.1584055459272097,
+      "grad_norm": 3.797602443884916,
+      "learning_rate": 9.576965541029877e-06,
+      "loss": 0.8227,
+      "step": 457
+    },
+    {
+      "epoch": 0.15875216637781628,
+      "grad_norm": 2.5921219767853034,
+      "learning_rate": 9.57470267957805e-06,
+      "loss": 0.7573,
+      "step": 458
+    },
+    {
+      "epoch": 0.15909878682842288,
+      "grad_norm": 7.165847021487318,
+      "learning_rate": 9.572434050897798e-06,
+      "loss": 0.9293,
+      "step": 459
+    },
+    {
+      "epoch": 0.15944540727902945,
+      "grad_norm": 3.503317324742654,
+      "learning_rate": 9.570159657849131e-06,
+      "loss": 0.9606,
+      "step": 460
+    },
+    {
+      "epoch": 0.15979202772963605,
+      "grad_norm": 6.342729136013537,
+      "learning_rate": 9.567879503299331e-06,
+      "loss": 1.0241,
+      "step": 461
+    },
+    {
+      "epoch": 0.16013864818024262,
+      "grad_norm": 16.130698065770936,
+      "learning_rate": 9.565593590122938e-06,
+      "loss": 0.7665,
+      "step": 462
+    },
+    {
+      "epoch": 0.16048526863084922,
+      "grad_norm": 3.2029226100964996,
+      "learning_rate": 9.56330192120175e-06,
+      "loss": 0.8648,
+      "step": 463
+    },
+    {
+      "epoch": 0.1608318890814558,
+      "grad_norm": 4.228630378916828,
+      "learning_rate": 9.561004499424828e-06,
+      "loss": 0.825,
+      "step": 464
+    },
+    {
+      "epoch": 0.1611785095320624,
+      "grad_norm": 13.603181790442237,
+      "learning_rate": 9.55870132768848e-06,
+      "loss": 0.8251,
+      "step": 465
+    },
+    {
+      "epoch": 0.161525129982669,
+      "grad_norm": 3.678108852364823,
+      "learning_rate": 9.556392408896265e-06,
+      "loss": 0.7768,
+      "step": 466
+    },
+    {
+      "epoch": 0.16187175043327556,
+      "grad_norm": 3.843895269518157,
+      "learning_rate": 9.554077745958987e-06,
+      "loss": 0.8498,
+      "step": 467
+    },
+    {
+      "epoch": 0.16221837088388216,
+      "grad_norm": 6.266938925873624,
+      "learning_rate": 9.551757341794689e-06,
+      "loss": 0.8461,
+      "step": 468
+    },
+    {
+      "epoch": 0.16256499133448873,
+      "grad_norm": 3.6909732829162616,
+      "learning_rate": 9.549431199328659e-06,
+      "loss": 0.8584,
+      "step": 469
+    },
+    {
+      "epoch": 0.16291161178509533,
+      "grad_norm": 3.35426607055521,
+      "learning_rate": 9.547099321493411e-06,
+      "loss": 0.7688,
+      "step": 470
+    },
+    {
+      "epoch": 0.1632582322357019,
+      "grad_norm": 6.948714962503842,
+      "learning_rate": 9.544761711228692e-06,
+      "loss": 0.7621,
+      "step": 471
+    },
+    {
+      "epoch": 0.1636048526863085,
+      "grad_norm": 3.1201139631467174,
+      "learning_rate": 9.54241837148148e-06,
+      "loss": 0.7838,
+      "step": 472
+    },
+    {
+      "epoch": 0.16395147313691508,
+      "grad_norm": 4.293739566235541,
+      "learning_rate": 9.54006930520597e-06,
+      "loss": 0.8173,
+      "step": 473
+    },
+    {
+      "epoch": 0.16429809358752168,
+      "grad_norm": 3.7407636725887348,
+      "learning_rate": 9.537714515363578e-06,
+      "loss": 0.7303,
+      "step": 474
+    },
+    {
+      "epoch": 0.16464471403812825,
+      "grad_norm": 7.878758075533462,
+      "learning_rate": 9.535354004922941e-06,
+      "loss": 0.7379,
+      "step": 475
+    },
+    {
+      "epoch": 0.16499133448873485,
+      "grad_norm": 4.805631731826863,
+      "learning_rate": 9.532987776859897e-06,
+      "loss": 0.7611,
+      "step": 476
+    },
+    {
+      "epoch": 0.16533795493934142,
+      "grad_norm": 7.3276520153863585,
+      "learning_rate": 9.530615834157504e-06,
+      "loss": 0.828,
+      "step": 477
+    },
+    {
+      "epoch": 0.16568457538994802,
+      "grad_norm": 8.317087935373227,
+      "learning_rate": 9.528238179806015e-06,
+      "loss": 0.9127,
+      "step": 478
+    },
+    {
+      "epoch": 0.1660311958405546,
+      "grad_norm": 3.671575174074881,
+      "learning_rate": 9.525854816802889e-06,
+      "loss": 0.8774,
+      "step": 479
+    },
+    {
+      "epoch": 0.1663778162911612,
+      "grad_norm": 4.996256660835925,
+      "learning_rate": 9.52346574815278e-06,
+      "loss": 0.7828,
+      "step": 480
+    },
+    {
+      "epoch": 0.16672443674176776,
+      "grad_norm": 6.946978861452984,
+      "learning_rate": 9.521070976867535e-06,
+      "loss": 0.754,
+      "step": 481
+    },
+    {
+      "epoch": 0.16707105719237436,
+      "grad_norm": 4.711905012722177,
+      "learning_rate": 9.518670505966189e-06,
+      "loss": 0.8588,
+      "step": 482
+    },
+    {
+      "epoch": 0.16741767764298093,
+      "grad_norm": 4.224048162558031,
+      "learning_rate": 9.516264338474963e-06,
+      "loss": 0.8992,
+      "step": 483
+    },
+    {
+      "epoch": 0.16776429809358753,
+      "grad_norm": 4.064864617529167,
+      "learning_rate": 9.513852477427264e-06,
+      "loss": 0.6614,
+      "step": 484
+    },
+    {
+      "epoch": 0.1681109185441941,
+      "grad_norm": 4.27405265326448,
+      "learning_rate": 9.51143492586367e-06,
+      "loss": 0.8102,
+      "step": 485
+    },
+    {
+      "epoch": 0.1684575389948007,
+      "grad_norm": 4.8179447556102355,
+      "learning_rate": 9.509011686831936e-06,
+      "loss": 0.8608,
+      "step": 486
+    },
+    {
+      "epoch": 0.16880415944540728,
+      "grad_norm": 5.622310910211739,
+      "learning_rate": 9.506582763386987e-06,
+      "loss": 0.756,
+      "step": 487
+    },
+    {
+      "epoch": 0.16915077989601388,
+      "grad_norm": 3.593671094002545,
+      "learning_rate": 9.504148158590915e-06,
+      "loss": 0.7679,
+      "step": 488
+    },
+    {
+      "epoch": 0.16949740034662045,
+      "grad_norm": 3.0923512420975765,
+      "learning_rate": 9.501707875512972e-06,
+      "loss": 0.8022,
+      "step": 489
+    },
+    {
+      "epoch": 0.16984402079722705,
+      "grad_norm": 6.615792605672805,
+      "learning_rate": 9.49926191722957e-06,
+      "loss": 0.8826,
+      "step": 490
+    },
+    {
+      "epoch": 0.17019064124783362,
+      "grad_norm": 3.6105449980110067,
+      "learning_rate": 9.496810286824278e-06,
+      "loss": 0.7954,
+      "step": 491
+    },
+    {
+      "epoch": 0.17053726169844022,
+      "grad_norm": 8.851430896795382,
+      "learning_rate": 9.494352987387811e-06,
+      "loss": 0.9134,
+      "step": 492
+    },
+    {
+      "epoch": 0.1708838821490468,
+      "grad_norm": 3.839935260201139,
+      "learning_rate": 9.491890022018033e-06,
+      "loss": 0.8136,
+      "step": 493
+    },
+    {
+      "epoch": 0.1712305025996534,
+      "grad_norm": 2.836073513175657,
+      "learning_rate": 9.489421393819953e-06,
+      "loss": 0.7811,
+      "step": 494
+    },
+    {
+      "epoch": 0.17157712305025996,
+      "grad_norm": 5.48201943395182,
+      "learning_rate": 9.486947105905713e-06,
+      "loss": 0.7797,
+      "step": 495
+    },
+    {
+      "epoch": 0.17192374350086656,
+      "grad_norm": 6.332598631795086,
+      "learning_rate": 9.4844671613946e-06,
+      "loss": 0.8088,
+      "step": 496
+    },
+    {
+      "epoch": 0.17227036395147313,
+      "grad_norm": 3.597283948998224,
+      "learning_rate": 9.48198156341302e-06,
+      "loss": 0.767,
+      "step": 497
+    },
+    {
+      "epoch": 0.17261698440207973,
+      "grad_norm": 9.27068009734496,
+      "learning_rate": 9.479490315094515e-06,
+      "loss": 0.8142,
+      "step": 498
+    },
+    {
+      "epoch": 0.1729636048526863,
+      "grad_norm": 3.18032975664549,
+      "learning_rate": 9.476993419579751e-06,
+      "loss": 0.8388,
+      "step": 499
+    },
+    {
+      "epoch": 0.1733102253032929,
+      "grad_norm": 4.818761047182729,
+      "learning_rate": 9.474490880016507e-06,
+      "loss": 0.8584,
+      "step": 500
+    },
+    {
+      "epoch": 0.17365684575389947,
+      "grad_norm": 14.432099983645323,
+      "learning_rate": 9.471982699559679e-06,
+      "loss": 0.8851,
+      "step": 501
+    },
+    {
+      "epoch": 0.17400346620450607,
+      "grad_norm": 2.697028289669232,
+      "learning_rate": 9.469468881371277e-06,
+      "loss": 0.888,
+      "step": 502
+    },
+    {
+      "epoch": 0.17435008665511265,
+      "grad_norm": 6.072303901489455,
+      "learning_rate": 9.466949428620418e-06,
+      "loss": 0.7758,
+      "step": 503
+    },
+    {
+      "epoch": 0.17469670710571925,
+      "grad_norm": 3.4157658163697247,
+      "learning_rate": 9.46442434448332e-06,
+      "loss": 0.8982,
+      "step": 504
+    },
+    {
+      "epoch": 0.17504332755632582,
+      "grad_norm": 4.751499987818415,
+      "learning_rate": 9.461893632143302e-06,
+      "loss": 0.9031,
+      "step": 505
+    },
+    {
+      "epoch": 0.17538994800693242,
+      "grad_norm": 2.717159164261551,
+      "learning_rate": 9.459357294790778e-06,
+      "loss": 0.771,
+      "step": 506
+    },
+    {
+      "epoch": 0.175736568457539,
+      "grad_norm": 10.567165697524384,
+      "learning_rate": 9.456815335623256e-06,
+      "loss": 0.8224,
+      "step": 507
+    },
+    {
+      "epoch": 0.1760831889081456,
+      "grad_norm": 8.401276864329072,
+      "learning_rate": 9.454267757845326e-06,
+      "loss": 0.8751,
+      "step": 508
+    },
+    {
+      "epoch": 0.17642980935875216,
+      "grad_norm": 4.517927071557469,
+      "learning_rate": 9.451714564668664e-06,
+      "loss": 0.7842,
+      "step": 509
+    },
+    {
+      "epoch": 0.17677642980935876,
+      "grad_norm": 3.1326601409493255,
+      "learning_rate": 9.449155759312028e-06,
+      "loss": 0.8041,
+      "step": 510
+    },
+    {
+      "epoch": 0.17712305025996533,
+      "grad_norm": 15.203224634113898,
+      "learning_rate": 9.446591345001247e-06,
+      "loss": 0.8512,
+      "step": 511
+    },
+    {
+      "epoch": 0.17746967071057193,
+      "grad_norm": 3.9275812363521703,
+      "learning_rate": 9.444021324969219e-06,
+      "loss": 0.8274,
+      "step": 512
+    },
+    {
+      "epoch": 0.1778162911611785,
+      "grad_norm": 4.307057087675799,
+      "learning_rate": 9.441445702455916e-06,
+      "loss": 0.8734,
+      "step": 513
+    },
+    {
+      "epoch": 0.1781629116117851,
+      "grad_norm": 3.0785724267947456,
+      "learning_rate": 9.43886448070837e-06,
+      "loss": 0.8505,
+      "step": 514
+    },
+    {
+      "epoch": 0.17850953206239167,
+      "grad_norm": 22.51955245415556,
+      "learning_rate": 9.43627766298067e-06,
+      "loss": 0.9522,
+      "step": 515
+    },
+    {
+      "epoch": 0.17885615251299827,
+      "grad_norm": 3.1523782625743864,
+      "learning_rate": 9.433685252533962e-06,
+      "loss": 0.9217,
+      "step": 516
+    },
+    {
+      "epoch": 0.17920277296360484,
+      "grad_norm": 9.412318642817155,
+      "learning_rate": 9.431087252636441e-06,
+      "loss": 0.8439,
+      "step": 517
+    },
+    {
+      "epoch": 0.17954939341421144,
+      "grad_norm": 8.30093057373676,
+      "learning_rate": 9.428483666563347e-06,
+      "loss": 0.9167,
+      "step": 518
+    },
+    {
+      "epoch": 0.17989601386481802,
+      "grad_norm": 3.238733520408378,
+      "learning_rate": 9.425874497596967e-06,
+      "loss": 0.8337,
+      "step": 519
+    },
+    {
+      "epoch": 0.18024263431542462,
+      "grad_norm": 4.493731897989322,
+      "learning_rate": 9.423259749026624e-06,
+      "loss": 0.805,
+      "step": 520
+    },
+    {
+      "epoch": 0.1805892547660312,
+      "grad_norm": 4.759472309950165,
+      "learning_rate": 9.420639424148675e-06,
+      "loss": 0.7751,
+      "step": 521
+    },
+    {
+      "epoch": 0.1809358752166378,
+      "grad_norm": 4.150528004291723,
+      "learning_rate": 9.418013526266502e-06,
+      "loss": 0.8678,
+      "step": 522
+    },
+    {
+      "epoch": 0.18128249566724436,
+      "grad_norm": 3.6019488502476587,
+      "learning_rate": 9.415382058690524e-06,
+      "loss": 0.766,
+      "step": 523
+    },
+    {
+      "epoch": 0.18162911611785096,
+      "grad_norm": 3.934052141929442,
+      "learning_rate": 9.41274502473817e-06,
+      "loss": 0.8482,
+      "step": 524
+    },
+    {
+      "epoch": 0.18197573656845753,
+      "grad_norm": 4.507257139955182,
+      "learning_rate": 9.410102427733895e-06,
+      "loss": 0.8062,
+      "step": 525
+    },
+    {
+      "epoch": 0.18232235701906413,
+      "grad_norm": 4.077584784197728,
+      "learning_rate": 9.407454271009162e-06,
+      "loss": 0.8982,
+      "step": 526
+    },
+    {
+      "epoch": 0.1826689774696707,
+      "grad_norm": 4.781173241898481,
+      "learning_rate": 9.404800557902445e-06,
+      "loss": 0.7808,
+      "step": 527
+    },
+    {
+      "epoch": 0.1830155979202773,
+      "grad_norm": 14.287802899385268,
+      "learning_rate": 9.402141291759223e-06,
+      "loss": 0.8286,
+      "step": 528
+    },
+    {
+      "epoch": 0.18336221837088387,
+      "grad_norm": 4.28752261526284,
+      "learning_rate": 9.399476475931977e-06,
+      "loss": 0.8147,
+      "step": 529
+    },
+    {
+      "epoch": 0.18370883882149047,
+      "grad_norm": 2.9698506018310473,
+      "learning_rate": 9.396806113780177e-06,
+      "loss": 0.8235,
+      "step": 530
+    },
+    {
+      "epoch": 0.18405545927209704,
+      "grad_norm": 4.895731017251158,
+      "learning_rate": 9.394130208670296e-06,
+      "loss": 0.7493,
+      "step": 531
+    },
+    {
+      "epoch": 0.18440207972270364,
+      "grad_norm": 3.6973778956948484,
+      "learning_rate": 9.391448763975786e-06,
+      "loss": 0.9047,
+      "step": 532
+    },
+    {
+      "epoch": 0.18474870017331022,
+      "grad_norm": 3.2601526331161015,
+      "learning_rate": 9.388761783077088e-06,
+      "loss": 0.8304,
+      "step": 533
+    },
+    {
+      "epoch": 0.18509532062391681,
+      "grad_norm": 3.479074912439342,
+      "learning_rate": 9.38606926936162e-06,
+      "loss": 1.0105,
+      "step": 534
+    },
+    {
+      "epoch": 0.1854419410745234,
+      "grad_norm": 4.277093201363125,
+      "learning_rate": 9.383371226223777e-06,
+      "loss": 0.769,
+      "step": 535
+    },
+    {
+      "epoch": 0.18578856152512999,
+      "grad_norm": 3.6103021659110675,
+      "learning_rate": 9.38066765706492e-06,
+      "loss": 0.7729,
+      "step": 536
+    },
+    {
+      "epoch": 0.18613518197573656,
+      "grad_norm": 3.641223585293136,
+      "learning_rate": 9.377958565293386e-06,
+      "loss": 0.9166,
+      "step": 537
+    },
+    {
+      "epoch": 0.18648180242634316,
+      "grad_norm": 3.5285419981922796,
+      "learning_rate": 9.375243954324462e-06,
+      "loss": 0.6577,
+      "step": 538
+    },
+    {
+      "epoch": 0.18682842287694973,
+      "grad_norm": 5.746057230637648,
+      "learning_rate": 9.372523827580402e-06,
+      "loss": 0.9474,
+      "step": 539
+    },
+    {
+      "epoch": 0.18717504332755633,
+      "grad_norm": 18.808244857052596,
+      "learning_rate": 9.369798188490409e-06,
+      "loss": 0.9083,
+      "step": 540
+    },
+    {
+      "epoch": 0.1875216637781629,
+      "grad_norm": 4.2545740609245595,
+      "learning_rate": 9.36706704049064e-06,
+      "loss": 0.7834,
+      "step": 541
+    },
+    {
+      "epoch": 0.1878682842287695,
+      "grad_norm": 5.678212015069833,
+      "learning_rate": 9.364330387024192e-06,
+      "loss": 0.7611,
+      "step": 542
+    },
+    {
+      "epoch": 0.18821490467937607,
+      "grad_norm": 3.3789498263595794,
+      "learning_rate": 9.361588231541105e-06,
+      "loss": 0.8336,
+      "step": 543
+    },
+    {
+      "epoch": 0.18856152512998267,
+      "grad_norm": 4.762432337870867,
+      "learning_rate": 9.358840577498356e-06,
+      "loss": 0.7448,
+      "step": 544
+    },
+    {
+      "epoch": 0.18890814558058924,
+      "grad_norm": 8.969855927783087,
+      "learning_rate": 9.35608742835985e-06,
+      "loss": 0.7938,
+      "step": 545
+    },
+    {
+      "epoch": 0.18925476603119584,
+      "grad_norm": 21.270178322823174,
+      "learning_rate": 9.353328787596426e-06,
+      "loss": 0.895,
+      "step": 546
+    },
+    {
+      "epoch": 0.18960138648180241,
+      "grad_norm": 3.826933563642163,
+      "learning_rate": 9.350564658685842e-06,
+      "loss": 0.842,
+      "step": 547
+    },
+    {
+      "epoch": 0.189948006932409,
+      "grad_norm": 4.411834194910977,
+      "learning_rate": 9.34779504511277e-06,
+      "loss": 0.8004,
+      "step": 548
+    },
+    {
+      "epoch": 0.19029462738301559,
+      "grad_norm": 3.820853636354766,
+      "learning_rate": 9.34501995036881e-06,
+      "loss": 0.7274,
+      "step": 549
+    },
+    {
+      "epoch": 0.19064124783362218,
+      "grad_norm": 5.7276543194817044,
+      "learning_rate": 9.342239377952457e-06,
+      "loss": 0.73,
+      "step": 550
+    },
+    {
+      "epoch": 0.19098786828422876,
+      "grad_norm": 3.5449831226271336,
+      "learning_rate": 9.339453331369124e-06,
+      "loss": 0.996,
+      "step": 551
+    },
+    {
+      "epoch": 0.19133448873483536,
+      "grad_norm": 2.6765211925786483,
+      "learning_rate": 9.336661814131114e-06,
+      "loss": 0.722,
+      "step": 552
+    },
+    {
+      "epoch": 0.19168110918544193,
+      "grad_norm": 5.159335729673817,
+      "learning_rate": 9.333864829757634e-06,
+      "loss": 0.829,
+      "step": 553
+    },
+    {
+      "epoch": 0.19202772963604853,
+      "grad_norm": 6.009371590143707,
+      "learning_rate": 9.331062381774785e-06,
+      "loss": 0.8431,
+      "step": 554
+    },
+    {
+      "epoch": 0.1923743500866551,
+      "grad_norm": 4.385258639325029,
+      "learning_rate": 9.328254473715548e-06,
+      "loss": 0.8122,
+      "step": 555
+    },
+    {
+      "epoch": 0.1927209705372617,
+      "grad_norm": 29.630001087828408,
+      "learning_rate": 9.325441109119795e-06,
+      "loss": 0.8766,
+      "step": 556
+    },
+    {
+      "epoch": 0.19306759098786827,
+      "grad_norm": 3.2318726937411677,
+      "learning_rate": 9.322622291534271e-06,
+      "loss": 0.8386,
+      "step": 557
+    },
+    {
+      "epoch": 0.19341421143847487,
+      "grad_norm": 3.939617240596644,
+      "learning_rate": 9.3197980245126e-06,
+      "loss": 0.8347,
+      "step": 558
+    },
+    {
+      "epoch": 0.19376083188908147,
+      "grad_norm": 13.949798288132383,
+      "learning_rate": 9.316968311615273e-06,
+      "loss": 0.7168,
+      "step": 559
+    },
+    {
+      "epoch": 0.19410745233968804,
+      "grad_norm": 5.982758479246912,
+      "learning_rate": 9.314133156409648e-06,
+      "loss": 0.6596,
+      "step": 560
+    },
+    {
+      "epoch": 0.19445407279029464,
+      "grad_norm": 6.319924574978242,
+      "learning_rate": 9.311292562469943e-06,
+      "loss": 0.7316,
+      "step": 561
+    },
+    {
+      "epoch": 0.1948006932409012,
+      "grad_norm": 4.539007403314948,
+      "learning_rate": 9.308446533377236e-06,
+      "loss": 0.8613,
+      "step": 562
+    },
+    {
+      "epoch": 0.1951473136915078,
+      "grad_norm": 3.3349823191657735,
+      "learning_rate": 9.30559507271945e-06,
+      "loss": 0.8528,
+      "step": 563
+    },
+    {
+      "epoch": 0.19549393414211438,
+      "grad_norm": 5.97275032465287,
+      "learning_rate": 9.302738184091363e-06,
+      "loss": 0.8216,
+      "step": 564
+    },
+    {
+      "epoch": 0.19584055459272098,
+      "grad_norm": 4.595575908217814,
+      "learning_rate": 9.29987587109459e-06,
+      "loss": 0.9377,
+      "step": 565
+    },
+    {
+      "epoch": 0.19618717504332756,
+      "grad_norm": 6.4657279193684385,
+      "learning_rate": 9.29700813733759e-06,
+      "loss": 0.774,
+      "step": 566
+    },
+    {
+      "epoch": 0.19653379549393415,
+      "grad_norm": 3.612939144751095,
+      "learning_rate": 9.294134986435647e-06,
+      "loss": 0.866,
+      "step": 567
+    },
+    {
+      "epoch": 0.19688041594454073,
+      "grad_norm": 17.46463040883144,
+      "learning_rate": 9.291256422010888e-06,
+      "loss": 0.7464,
+      "step": 568
+    },
+    {
+      "epoch": 0.19722703639514733,
+      "grad_norm": 6.462136886373985,
+      "learning_rate": 9.288372447692252e-06,
+      "loss": 0.8073,
+      "step": 569
+    },
+    {
+      "epoch": 0.1975736568457539,
+      "grad_norm": 3.0329456683185483,
+      "learning_rate": 9.2854830671155e-06,
+      "loss": 0.883,
+      "step": 570
+    },
+    {
+      "epoch": 0.1979202772963605,
+      "grad_norm": 4.95573244426877,
+      "learning_rate": 9.282588283923217e-06,
+      "loss": 0.8446,
+      "step": 571
+    },
+    {
+      "epoch": 0.19826689774696707,
+      "grad_norm": 6.704307312608394,
+      "learning_rate": 9.279688101764792e-06,
+      "loss": 0.9902,
+      "step": 572
+    },
+    {
+      "epoch": 0.19861351819757367,
+      "grad_norm": 4.135426850850625,
+      "learning_rate": 9.276782524296421e-06,
+      "loss": 0.8479,
+      "step": 573
+    },
+    {
+      "epoch": 0.19896013864818024,
+      "grad_norm": 5.48307131163695,
+      "learning_rate": 9.273871555181101e-06,
+      "loss": 0.8261,
+      "step": 574
+    },
+    {
+      "epoch": 0.19930675909878684,
+      "grad_norm": 6.913864152184309,
+      "learning_rate": 9.270955198088631e-06,
+      "loss": 0.8877,
+      "step": 575
+    },
+    {
+      "epoch": 0.1996533795493934,
+      "grad_norm": 5.342951080630691,
+      "learning_rate": 9.268033456695595e-06,
+      "loss": 0.8104,
+      "step": 576
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.5644536989402087,
+      "learning_rate": 9.265106334685372e-06,
+      "loss": 0.676,
+      "step": 577
+    },
+    {
+      "epoch": 0.20034662045060658,
+      "grad_norm": 3.2254915291008714,
+      "learning_rate": 9.26217383574812e-06,
+      "loss": 0.7308,
+      "step": 578
+    },
+    {
+      "epoch": 0.20069324090121318,
+      "grad_norm": 3.1860310686056748,
+      "learning_rate": 9.25923596358078e-06,
+      "loss": 0.8733,
+      "step": 579
+    },
+    {
+      "epoch": 0.20103986135181975,
+      "grad_norm": 3.299031295156969,
+      "learning_rate": 9.256292721887056e-06,
+      "loss": 0.8104,
+      "step": 580
+    },
+    {
+      "epoch": 0.20138648180242635,
+      "grad_norm": 7.7914378231536965,
+      "learning_rate": 9.253344114377438e-06,
+      "loss": 0.842,
+      "step": 581
+    },
+    {
+      "epoch": 0.20173310225303293,
+      "grad_norm": 5.8003861882117205,
+      "learning_rate": 9.250390144769165e-06,
+      "loss": 0.7609,
+      "step": 582
+    },
+    {
+      "epoch": 0.20207972270363952,
+      "grad_norm": 2.9361539890692114,
+      "learning_rate": 9.247430816786245e-06,
+      "loss": 0.8579,
+      "step": 583
+    },
+    {
+      "epoch": 0.2024263431542461,
+      "grad_norm": 4.433506779587395,
+      "learning_rate": 9.244466134159439e-06,
+      "loss": 0.8301,
+      "step": 584
+    },
+    {
+      "epoch": 0.2027729636048527,
+      "grad_norm": 3.9504761667756174,
+      "learning_rate": 9.241496100626257e-06,
+      "loss": 0.7632,
+      "step": 585
+    },
+    {
+      "epoch": 0.20311958405545927,
+      "grad_norm": 4.733936830786511,
+      "learning_rate": 9.238520719930957e-06,
+      "loss": 0.6923,
+      "step": 586
+    },
+    {
+      "epoch": 0.20346620450606587,
+      "grad_norm": 5.99248017638386,
+      "learning_rate": 9.235539995824533e-06,
+      "loss": 0.8403,
+      "step": 587
+    },
+    {
+      "epoch": 0.20381282495667244,
+      "grad_norm": 4.749256404855942,
+      "learning_rate": 9.232553932064727e-06,
+      "loss": 0.7713,
+      "step": 588
+    },
+    {
+      "epoch": 0.20415944540727904,
+      "grad_norm": 4.414112287650165,
+      "learning_rate": 9.229562532415996e-06,
+      "loss": 0.8079,
+      "step": 589
+    },
+    {
+      "epoch": 0.2045060658578856,
+      "grad_norm": 3.922837383986131,
+      "learning_rate": 9.22656580064954e-06,
+      "loss": 0.7669,
+      "step": 590
+    },
+    {
+      "epoch": 0.2048526863084922,
+      "grad_norm": 3.3523270661553437,
+      "learning_rate": 9.223563740543269e-06,
+      "loss": 0.8707,
+      "step": 591
+    },
+    {
+      "epoch": 0.20519930675909878,
+      "grad_norm": 5.272127444209614,
+      "learning_rate": 9.220556355881817e-06,
+      "loss": 0.818,
+      "step": 592
+    },
+    {
+      "epoch": 0.20554592720970538,
+      "grad_norm": 5.486915473374795,
+      "learning_rate": 9.217543650456528e-06,
+      "loss": 0.7514,
+      "step": 593
+    },
+    {
+      "epoch": 0.20589254766031195,
+      "grad_norm": 4.484343071630026,
+      "learning_rate": 9.214525628065453e-06,
+      "loss": 0.7991,
+      "step": 594
+    },
+    {
+      "epoch": 0.20623916811091855,
+      "grad_norm": 5.280896453993813,
+      "learning_rate": 9.21150229251335e-06,
+      "loss": 0.7079,
+      "step": 595
+    },
+    {
+      "epoch": 0.20658578856152512,
+      "grad_norm": 5.936611173417885,
+      "learning_rate": 9.208473647611674e-06,
+      "loss": 0.8552,
+      "step": 596
+    },
+    {
+      "epoch": 0.20693240901213172,
+      "grad_norm": 3.486440118961097,
+      "learning_rate": 9.205439697178567e-06,
+      "loss": 0.8433,
+      "step": 597
+    },
+    {
+      "epoch": 0.2072790294627383,
+      "grad_norm": 6.232686243819952,
+      "learning_rate": 9.202400445038866e-06,
+      "loss": 0.6964,
+      "step": 598
+    },
+    {
+      "epoch": 0.2076256499133449,
+      "grad_norm": 21.79792002994311,
+      "learning_rate": 9.199355895024091e-06,
+      "loss": 0.8397,
+      "step": 599
+    },
+    {
+      "epoch": 0.20797227036395147,
+      "grad_norm": 3.900493027118092,
+      "learning_rate": 9.19630605097244e-06,
+      "loss": 0.8483,
+      "step": 600
+    },
+    {
+      "epoch": 0.20831889081455807,
+      "grad_norm": 4.856617746715711,
+      "learning_rate": 9.193250916728789e-06,
+      "loss": 0.7529,
+      "step": 601
+    },
+    {
+      "epoch": 0.20866551126516464,
+      "grad_norm": 6.561427290965392,
+      "learning_rate": 9.190190496144673e-06,
+      "loss": 0.893,
+      "step": 602
+    },
+    {
+      "epoch": 0.20901213171577124,
+      "grad_norm": 6.087571183623038,
+      "learning_rate": 9.187124793078302e-06,
+      "loss": 0.7749,
+      "step": 603
+    },
+    {
+      "epoch": 0.2093587521663778,
+      "grad_norm": 3.791884863185793,
+      "learning_rate": 9.18405381139454e-06,
+      "loss": 0.8541,
+      "step": 604
+    },
+    {
+      "epoch": 0.2097053726169844,
+      "grad_norm": 9.135702869613137,
+      "learning_rate": 9.180977554964906e-06,
+      "loss": 0.7899,
+      "step": 605
+    },
+    {
+      "epoch": 0.21005199306759098,
+      "grad_norm": 13.052090195375834,
+      "learning_rate": 9.177896027667572e-06,
+      "loss": 0.8523,
+      "step": 606
+    },
+    {
+      "epoch": 0.21039861351819758,
+      "grad_norm": 8.661531623907187,
+      "learning_rate": 9.17480923338735e-06,
+      "loss": 0.72,
+      "step": 607
+    },
+    {
+      "epoch": 0.21074523396880415,
+      "grad_norm": 3.604844042315808,
+      "learning_rate": 9.171717176015698e-06,
+      "loss": 0.8384,
+      "step": 608
+    },
+    {
+      "epoch": 0.21109185441941075,
+      "grad_norm": 4.4087269944785445,
+      "learning_rate": 9.168619859450702e-06,
+      "loss": 0.6791,
+      "step": 609
+    },
+    {
+      "epoch": 0.21143847487001732,
+      "grad_norm": 8.654752483429995,
+      "learning_rate": 9.165517287597085e-06,
+      "loss": 0.7201,
+      "step": 610
+    },
+    {
+      "epoch": 0.21178509532062392,
+      "grad_norm": 12.744136448320251,
+      "learning_rate": 9.16240946436619e-06,
+      "loss": 0.8324,
+      "step": 611
+    },
+    {
+      "epoch": 0.2121317157712305,
+      "grad_norm": 5.031132651864972,
+      "learning_rate": 9.159296393675982e-06,
+      "loss": 0.7951,
+      "step": 612
+    },
+    {
+      "epoch": 0.2124783362218371,
+      "grad_norm": 6.095970228060085,
+      "learning_rate": 9.156178079451044e-06,
+      "loss": 0.7534,
+      "step": 613
+    },
+    {
+      "epoch": 0.21282495667244367,
+      "grad_norm": 4.832448165220876,
+      "learning_rate": 9.153054525622563e-06,
+      "loss": 0.7941,
+      "step": 614
+    },
+    {
+      "epoch": 0.21317157712305027,
+      "grad_norm": 9.116095961718605,
+      "learning_rate": 9.14992573612834e-06,
+      "loss": 0.6995,
+      "step": 615
+    },
+    {
+      "epoch": 0.21351819757365684,
+      "grad_norm": 4.8382746825912095,
+      "learning_rate": 9.14679171491277e-06,
+      "loss": 0.8109,
+      "step": 616
+    },
+    {
+      "epoch": 0.21386481802426344,
+      "grad_norm": 4.431576021484964,
+      "learning_rate": 9.143652465926846e-06,
+      "loss": 0.8172,
+      "step": 617
+    },
+    {
+      "epoch": 0.21421143847487,
+      "grad_norm": 5.354720488860779,
+      "learning_rate": 9.140507993128152e-06,
+      "loss": 0.7984,
+      "step": 618
+    },
+    {
+      "epoch": 0.2145580589254766,
+      "grad_norm": 4.083007277194521,
+      "learning_rate": 9.137358300480854e-06,
+      "loss": 0.8131,
+      "step": 619
+    },
+    {
+      "epoch": 0.21490467937608318,
+      "grad_norm": 5.593413651060834,
+      "learning_rate": 9.134203391955702e-06,
+      "loss": 0.8586,
+      "step": 620
+    },
+    {
+      "epoch": 0.21525129982668978,
+      "grad_norm": 4.162343665479354,
+      "learning_rate": 9.131043271530024e-06,
+      "loss": 0.8527,
+      "step": 621
+    },
+    {
+      "epoch": 0.21559792027729635,
+      "grad_norm": 4.755044984546613,
+      "learning_rate": 9.12787794318771e-06,
+      "loss": 0.8583,
+      "step": 622
+    },
+    {
+      "epoch": 0.21594454072790295,
+      "grad_norm": 3.6044677741072246,
+      "learning_rate": 9.124707410919225e-06,
+      "loss": 0.8074,
+      "step": 623
+    },
+    {
+      "epoch": 0.21629116117850952,
+      "grad_norm": 2.8771763655291895,
+      "learning_rate": 9.121531678721589e-06,
+      "loss": 0.8113,
+      "step": 624
+    },
+    {
+      "epoch": 0.21663778162911612,
+      "grad_norm": 4.001101179134556,
+      "learning_rate": 9.118350750598377e-06,
+      "loss": 0.8443,
+      "step": 625
+    },
+    {
+      "epoch": 0.2169844020797227,
+      "grad_norm": 3.137965203230599,
+      "learning_rate": 9.115164630559719e-06,
+      "loss": 0.7821,
+      "step": 626
+    },
+    {
+      "epoch": 0.2173310225303293,
+      "grad_norm": 3.227089174634959,
+      "learning_rate": 9.111973322622284e-06,
+      "loss": 0.8574,
+      "step": 627
+    },
+    {
+      "epoch": 0.21767764298093586,
+      "grad_norm": 5.785849174419331,
+      "learning_rate": 9.108776830809289e-06,
+      "loss": 0.6666,
+      "step": 628
+    },
+    {
+      "epoch": 0.21802426343154246,
+      "grad_norm": 3.981418725958829,
+      "learning_rate": 9.105575159150477e-06,
+      "loss": 0.8531,
+      "step": 629
+    },
+    {
+      "epoch": 0.21837088388214904,
+      "grad_norm": 7.581543119740569,
+      "learning_rate": 9.102368311682131e-06,
+      "loss": 0.6852,
+      "step": 630
+    },
+    {
+      "epoch": 0.21871750433275564,
+      "grad_norm": 4.8316554397976725,
+      "learning_rate": 9.099156292447048e-06,
+      "loss": 0.8448,
+      "step": 631
+    },
+    {
+      "epoch": 0.2190641247833622,
+      "grad_norm": 5.0649675338153,
+      "learning_rate": 9.095939105494557e-06,
+      "loss": 0.7124,
+      "step": 632
+    },
+    {
+      "epoch": 0.2194107452339688,
+      "grad_norm": 4.301635734807242,
+      "learning_rate": 9.092716754880493e-06,
+      "loss": 0.8121,
+      "step": 633
+    },
+    {
+      "epoch": 0.21975736568457538,
+      "grad_norm": 10.005117774558611,
+      "learning_rate": 9.089489244667205e-06,
+      "loss": 0.8114,
+      "step": 634
+    },
+    {
+      "epoch": 0.22010398613518198,
+      "grad_norm": 4.778616157652271,
+      "learning_rate": 9.086256578923542e-06,
+      "loss": 0.715,
+      "step": 635
+    },
+    {
+      "epoch": 0.22045060658578855,
+      "grad_norm": 3.466821674067941,
+      "learning_rate": 9.083018761724858e-06,
+      "loss": 0.7762,
+      "step": 636
+    },
+    {
+      "epoch": 0.22079722703639515,
+      "grad_norm": 5.371541645825729,
+      "learning_rate": 9.079775797152999e-06,
+      "loss": 0.748,
+      "step": 637
+    },
+    {
+      "epoch": 0.22114384748700172,
+      "grad_norm": 6.759298798645191,
+      "learning_rate": 9.076527689296299e-06,
+      "loss": 0.8492,
+      "step": 638
+    },
+    {
+      "epoch": 0.22149046793760832,
+      "grad_norm": 3.21818779982304,
+      "learning_rate": 9.073274442249577e-06,
+      "loss": 0.8399,
+      "step": 639
+    },
+    {
+      "epoch": 0.2218370883882149,
+      "grad_norm": 4.973363999698727,
+      "learning_rate": 9.070016060114132e-06,
+      "loss": 0.7945,
+      "step": 640
+    },
+    {
+      "epoch": 0.2221837088388215,
+      "grad_norm": 3.178284602118559,
+      "learning_rate": 9.066752546997734e-06,
+      "loss": 0.872,
+      "step": 641
+    },
+    {
+      "epoch": 0.22253032928942806,
+      "grad_norm": 7.553086864917782,
+      "learning_rate": 9.063483907014626e-06,
+      "loss": 0.8993,
+      "step": 642
+    },
+    {
+      "epoch": 0.22287694974003466,
+      "grad_norm": 5.081266341157601,
+      "learning_rate": 9.060210144285509e-06,
+      "loss": 0.8308,
+      "step": 643
+    },
+    {
+      "epoch": 0.22322357019064124,
+      "grad_norm": 2.5614804134689617,
+      "learning_rate": 9.056931262937547e-06,
+      "loss": 0.7841,
+      "step": 644
+    },
+    {
+      "epoch": 0.22357019064124783,
+      "grad_norm": 2.7853390785041383,
+      "learning_rate": 9.053647267104352e-06,
+      "loss": 0.8145,
+      "step": 645
+    },
+    {
+      "epoch": 0.2239168110918544,
+      "grad_norm": 5.039642623350357,
+      "learning_rate": 9.050358160925988e-06,
+      "loss": 0.7092,
+      "step": 646
+    },
+    {
+      "epoch": 0.224263431542461,
+      "grad_norm": 4.832623706048755,
+      "learning_rate": 9.047063948548959e-06,
+      "loss": 0.7888,
+      "step": 647
+    },
+    {
+      "epoch": 0.22461005199306758,
+      "grad_norm": 4.401507227761821,
+      "learning_rate": 9.043764634126206e-06,
+      "loss": 0.8582,
+      "step": 648
+    },
+    {
+      "epoch": 0.22495667244367418,
+      "grad_norm": 4.974755203397646,
+      "learning_rate": 9.040460221817107e-06,
+      "loss": 0.7558,
+      "step": 649
+    },
+    {
+      "epoch": 0.22530329289428075,
+      "grad_norm": 3.7225059709497237,
+      "learning_rate": 9.037150715787457e-06,
+      "loss": 0.8424,
+      "step": 650
+    },
+    {
+      "epoch": 0.22564991334488735,
+      "grad_norm": 3.9866144494496796,
+      "learning_rate": 9.033836120209484e-06,
+      "loss": 0.6812,
+      "step": 651
+    },
+    {
+      "epoch": 0.22599653379549395,
+      "grad_norm": 4.304105001079437,
+      "learning_rate": 9.030516439261825e-06,
+      "loss": 0.8586,
+      "step": 652
+    },
+    {
+      "epoch": 0.22634315424610052,
+      "grad_norm": 6.491209073785222,
+      "learning_rate": 9.027191677129528e-06,
+      "loss": 0.7335,
+      "step": 653
+    },
+    {
+      "epoch": 0.22668977469670712,
+      "grad_norm": 6.514351879083053,
+      "learning_rate": 9.02386183800405e-06,
+      "loss": 0.7963,
+      "step": 654
+    },
+    {
+      "epoch": 0.2270363951473137,
+      "grad_norm": 4.745702915878149,
+      "learning_rate": 9.020526926083248e-06,
+      "loss": 0.7949,
+      "step": 655
+    },
+    {
+      "epoch": 0.2273830155979203,
+      "grad_norm": 5.30077426868309,
+      "learning_rate": 9.01718694557137e-06,
+      "loss": 0.7786,
+      "step": 656
+    },
+    {
+      "epoch": 0.22772963604852686,
+      "grad_norm": 2.658229762526705,
+      "learning_rate": 9.013841900679061e-06,
+      "loss": 0.8784,
+      "step": 657
+    },
+    {
+      "epoch": 0.22807625649913346,
+      "grad_norm": 17.035110307724516,
+      "learning_rate": 9.010491795623345e-06,
+      "loss": 0.6765,
+      "step": 658
+    },
+    {
+      "epoch": 0.22842287694974003,
+      "grad_norm": 7.8287594882421665,
+      "learning_rate": 9.007136634627622e-06,
+      "loss": 0.8591,
+      "step": 659
+    },
+    {
+      "epoch": 0.22876949740034663,
+      "grad_norm": 6.250455993472959,
+      "learning_rate": 9.003776421921678e-06,
+      "loss": 0.751,
+      "step": 660
+    },
+    {
+      "epoch": 0.2291161178509532,
+      "grad_norm": 3.4682378075053855,
+      "learning_rate": 9.000411161741656e-06,
+      "loss": 0.7064,
+      "step": 661
+    },
+    {
+      "epoch": 0.2294627383015598,
+      "grad_norm": 4.618055756952564,
+      "learning_rate": 8.997040858330068e-06,
+      "loss": 0.9343,
+      "step": 662
+    },
+    {
+      "epoch": 0.22980935875216638,
+      "grad_norm": 5.443694016916337,
+      "learning_rate": 8.99366551593578e-06,
+      "loss": 0.9651,
+      "step": 663
+    },
+    {
+      "epoch": 0.23015597920277298,
+      "grad_norm": 4.781507150819373,
+      "learning_rate": 8.99028513881402e-06,
+      "loss": 0.8648,
+      "step": 664
+    },
+    {
+      "epoch": 0.23050259965337955,
+      "grad_norm": 5.926711538312498,
+      "learning_rate": 8.986899731226346e-06,
+      "loss": 0.7852,
+      "step": 665
+    },
+    {
+      "epoch": 0.23084922010398615,
+      "grad_norm": 6.403355243496209,
+      "learning_rate": 8.983509297440674e-06,
+      "loss": 0.6644,
+      "step": 666
+    },
+    {
+      "epoch": 0.23119584055459272,
+      "grad_norm": 4.816954933889582,
+      "learning_rate": 8.98011384173125e-06,
+      "loss": 0.6589,
+      "step": 667
+    },
+    {
+      "epoch": 0.23154246100519932,
+      "grad_norm": 3.920638614830443,
+      "learning_rate": 8.976713368378651e-06,
+      "loss": 0.8198,
+      "step": 668
+    },
+    {
+      "epoch": 0.2318890814558059,
+      "grad_norm": 4.717286903921587,
+      "learning_rate": 8.97330788166978e-06,
+      "loss": 0.791,
+      "step": 669
+    },
+    {
+      "epoch": 0.2322357019064125,
+      "grad_norm": 4.657561943478897,
+      "learning_rate": 8.969897385897857e-06,
+      "loss": 0.9009,
+      "step": 670
+    },
+    {
+      "epoch": 0.23258232235701906,
+      "grad_norm": 5.551476338322548,
+      "learning_rate": 8.966481885362426e-06,
+      "loss": 0.777,
+      "step": 671
+    },
+    {
+      "epoch": 0.23292894280762566,
+      "grad_norm": 6.392541432155706,
+      "learning_rate": 8.963061384369329e-06,
+      "loss": 0.8869,
+      "step": 672
+    },
+    {
+      "epoch": 0.23327556325823223,
+      "grad_norm": 4.122056911712834,
+      "learning_rate": 8.959635887230719e-06,
+      "loss": 0.7914,
+      "step": 673
+    },
+    {
+      "epoch": 0.23362218370883883,
+      "grad_norm": 10.043039817758022,
+      "learning_rate": 8.95620539826505e-06,
+      "loss": 0.8146,
+      "step": 674
+    },
+    {
+      "epoch": 0.2339688041594454,
+      "grad_norm": 5.377622312811555,
+      "learning_rate": 8.95276992179706e-06,
+      "loss": 0.8001,
+      "step": 675
+    },
+    {
+      "epoch": 0.234315424610052,
+      "grad_norm": 3.9550600697000386,
+      "learning_rate": 8.949329462157778e-06,
+      "loss": 0.8415,
+      "step": 676
+    },
+    {
+      "epoch": 0.23466204506065858,
+      "grad_norm": 3.198165512578629,
+      "learning_rate": 8.945884023684523e-06,
+      "loss": 0.7326,
+      "step": 677
+    },
+    {
+      "epoch": 0.23500866551126517,
+      "grad_norm": 2.5466864573302304,
+      "learning_rate": 8.942433610720882e-06,
+      "loss": 0.7455,
+      "step": 678
+    },
+    {
+      "epoch": 0.23535528596187175,
+      "grad_norm": 3.0115716643436885,
+      "learning_rate": 8.938978227616715e-06,
+      "loss": 0.715,
+      "step": 679
+    },
+    {
+      "epoch": 0.23570190641247835,
+      "grad_norm": 3.7183903133912644,
+      "learning_rate": 8.935517878728149e-06,
+      "loss": 0.9231,
+      "step": 680
+    },
+    {
+      "epoch": 0.23604852686308492,
+      "grad_norm": 4.956900750944687,
+      "learning_rate": 8.93205256841757e-06,
+      "loss": 0.8129,
+      "step": 681
+    },
+    {
+      "epoch": 0.23639514731369152,
+      "grad_norm": 3.3761834500526016,
+      "learning_rate": 8.928582301053621e-06,
+      "loss": 0.87,
+      "step": 682
+    },
+    {
+      "epoch": 0.2367417677642981,
+      "grad_norm": 3.9545811725742057,
+      "learning_rate": 8.925107081011191e-06,
+      "loss": 0.8475,
+      "step": 683
+    },
+    {
+      "epoch": 0.2370883882149047,
+      "grad_norm": 4.930725058726986,
+      "learning_rate": 8.921626912671416e-06,
+      "loss": 0.818,
+      "step": 684
+    },
+    {
+      "epoch": 0.23743500866551126,
+      "grad_norm": 7.680469731102449,
+      "learning_rate": 8.918141800421668e-06,
+      "loss": 0.8885,
+      "step": 685
+    },
+    {
+      "epoch": 0.23778162911611786,
+      "grad_norm": 4.011590470322957,
+      "learning_rate": 8.914651748655555e-06,
+      "loss": 0.851,
+      "step": 686
+    },
+    {
+      "epoch": 0.23812824956672443,
+      "grad_norm": 4.109160393943097,
+      "learning_rate": 8.911156761772903e-06,
+      "loss": 0.8244,
+      "step": 687
+    },
+    {
+      "epoch": 0.23847487001733103,
+      "grad_norm": 25.668940658651135,
+      "learning_rate": 8.907656844179773e-06,
+      "loss": 0.8616,
+      "step": 688
+    },
+    {
+      "epoch": 0.2388214904679376,
+      "grad_norm": 6.710358470496926,
+      "learning_rate": 8.904152000288432e-06,
+      "loss": 0.7289,
+      "step": 689
+    },
+    {
+      "epoch": 0.2391681109185442,
+      "grad_norm": 5.761858657464603,
+      "learning_rate": 8.90064223451736e-06,
+      "loss": 0.7308,
+      "step": 690
+    },
+    {
+      "epoch": 0.23951473136915077,
+      "grad_norm": 3.836781340478025,
+      "learning_rate": 8.897127551291245e-06,
+      "loss": 0.7135,
+      "step": 691
+    },
+    {
+      "epoch": 0.23986135181975737,
+      "grad_norm": 13.683839056116632,
+      "learning_rate": 8.893607955040971e-06,
+      "loss": 0.8189,
+      "step": 692
+    },
+    {
+      "epoch": 0.24020797227036395,
+      "grad_norm": 4.603513465905895,
+      "learning_rate": 8.890083450203615e-06,
+      "loss": 0.809,
+      "step": 693
+    },
+    {
+      "epoch": 0.24055459272097054,
+      "grad_norm": 7.298232308033326,
+      "learning_rate": 8.886554041222445e-06,
+      "loss": 0.6286,
+      "step": 694
+    },
+    {
+      "epoch": 0.24090121317157712,
+      "grad_norm": 10.637488735143245,
+      "learning_rate": 8.883019732546912e-06,
+      "loss": 0.901,
+      "step": 695
+    },
+    {
+      "epoch": 0.24124783362218372,
+      "grad_norm": 4.710789415085181,
+      "learning_rate": 8.87948052863264e-06,
+      "loss": 0.7469,
+      "step": 696
+    },
+    {
+      "epoch": 0.2415944540727903,
+      "grad_norm": 7.321473106776083,
+      "learning_rate": 8.875936433941427e-06,
+      "loss": 0.7592,
+      "step": 697
+    },
+    {
+      "epoch": 0.2419410745233969,
+      "grad_norm": 3.0146492877288855,
+      "learning_rate": 8.872387452941237e-06,
+      "loss": 0.8989,
+      "step": 698
+    },
+    {
+      "epoch": 0.24228769497400346,
+      "grad_norm": 3.4046534893516185,
+      "learning_rate": 8.868833590106191e-06,
+      "loss": 0.7958,
+      "step": 699
+    },
+    {
+      "epoch": 0.24263431542461006,
+      "grad_norm": 2.9927398939566046,
+      "learning_rate": 8.865274849916571e-06,
+      "loss": 0.7547,
+      "step": 700
+    },
+    {
+      "epoch": 0.24298093587521663,
+      "grad_norm": 39.01061622291388,
+      "learning_rate": 8.861711236858802e-06,
+      "loss": 0.8382,
+      "step": 701
+    },
+    {
+      "epoch": 0.24332755632582323,
+      "grad_norm": 4.138707793482425,
+      "learning_rate": 8.85814275542545e-06,
+      "loss": 0.7306,
+      "step": 702
+    },
+    {
+      "epoch": 0.2436741767764298,
+      "grad_norm": 7.354885560102154,
+      "learning_rate": 8.854569410115225e-06,
+      "loss": 0.7724,
+      "step": 703
+    },
+    {
+      "epoch": 0.2440207972270364,
+      "grad_norm": 5.403251388213113,
+      "learning_rate": 8.850991205432968e-06,
+      "loss": 0.8427,
+      "step": 704
+    },
+    {
+      "epoch": 0.24436741767764297,
+      "grad_norm": 3.1192702755815103,
+      "learning_rate": 8.847408145889643e-06,
+      "loss": 0.7301,
+      "step": 705
+    },
+    {
+      "epoch": 0.24471403812824957,
+      "grad_norm": 17.393681722143082,
+      "learning_rate": 8.843820236002332e-06,
+      "loss": 0.832,
+      "step": 706
+    },
+    {
+      "epoch": 0.24506065857885614,
+      "grad_norm": 2.693486458779027,
+      "learning_rate": 8.840227480294237e-06,
+      "loss": 0.746,
+      "step": 707
+    },
+    {
+      "epoch": 0.24540727902946274,
+      "grad_norm": 4.217328519973262,
+      "learning_rate": 8.83662988329467e-06,
+      "loss": 0.8439,
+      "step": 708
+    },
+    {
+      "epoch": 0.24575389948006932,
+      "grad_norm": 4.444547770124441,
+      "learning_rate": 8.833027449539039e-06,
+      "loss": 0.7152,
+      "step": 709
+    },
+    {
+      "epoch": 0.24610051993067592,
+      "grad_norm": 4.302143524634476,
+      "learning_rate": 8.829420183568857e-06,
+      "loss": 0.81,
+      "step": 710
+    },
+    {
+      "epoch": 0.2464471403812825,
+      "grad_norm": 34.995543876240745,
+      "learning_rate": 8.825808089931727e-06,
+      "loss": 0.7531,
+      "step": 711
+    },
+    {
+      "epoch": 0.2467937608318891,
+      "grad_norm": 4.979313028465498,
+      "learning_rate": 8.822191173181333e-06,
+      "loss": 0.76,
+      "step": 712
+    },
+    {
+      "epoch": 0.24714038128249566,
+      "grad_norm": 4.379744474390453,
+      "learning_rate": 8.818569437877447e-06,
+      "loss": 0.8456,
+      "step": 713
+    },
+    {
+      "epoch": 0.24748700173310226,
+      "grad_norm": 32.69851652020799,
+      "learning_rate": 8.814942888585911e-06,
+      "loss": 0.8282,
+      "step": 714
+    },
+    {
+      "epoch": 0.24783362218370883,
+      "grad_norm": 4.203625206747043,
+      "learning_rate": 8.811311529878636e-06,
+      "loss": 0.9249,
+      "step": 715
+    },
+    {
+      "epoch": 0.24818024263431543,
+      "grad_norm": 3.110094930871569,
+      "learning_rate": 8.807675366333598e-06,
+      "loss": 0.7533,
+      "step": 716
+    },
+    {
+      "epoch": 0.248526863084922,
+      "grad_norm": 10.756630664461735,
+      "learning_rate": 8.804034402534831e-06,
+      "loss": 0.8054,
+      "step": 717
+    },
+    {
+      "epoch": 0.2488734835355286,
+      "grad_norm": 4.834079458906423,
+      "learning_rate": 8.800388643072418e-06,
+      "loss": 0.8043,
+      "step": 718
+    },
+    {
+      "epoch": 0.24922010398613517,
+      "grad_norm": 3.5639720706641165,
+      "learning_rate": 8.796738092542487e-06,
+      "loss": 0.8009,
+      "step": 719
+    },
+    {
+      "epoch": 0.24956672443674177,
+      "grad_norm": 3.6715900746986225,
+      "learning_rate": 8.79308275554721e-06,
+      "loss": 0.859,
+      "step": 720
+    },
+    {
+      "epoch": 0.24991334488734834,
+      "grad_norm": 3.85941167500905,
+      "learning_rate": 8.789422636694791e-06,
+      "loss": 0.8079,
+      "step": 721
+    },
+    {
+      "epoch": 0.2502599653379549,
+      "grad_norm": 5.367474275073313,
+      "learning_rate": 8.785757740599462e-06,
+      "loss": 0.8189,
+      "step": 722
+    },
+    {
+      "epoch": 0.25060658578856154,
+      "grad_norm": 5.065366072974775,
+      "learning_rate": 8.782088071881478e-06,
+      "loss": 0.7898,
+      "step": 723
+    },
+    {
+      "epoch": 0.2509532062391681,
+      "grad_norm": 55.62171628559595,
+      "learning_rate": 8.77841363516711e-06,
+      "loss": 0.7296,
+      "step": 724
+    },
+    {
+      "epoch": 0.2512998266897747,
+      "grad_norm": 5.936297038817151,
+      "learning_rate": 8.774734435088642e-06,
+      "loss": 0.7678,
+      "step": 725
+    },
+    {
+      "epoch": 0.25164644714038126,
+      "grad_norm": 7.758167955518023,
+      "learning_rate": 8.77105047628436e-06,
+      "loss": 0.7204,
+      "step": 726
+    },
+    {
+      "epoch": 0.2519930675909879,
+      "grad_norm": 3.3562304888305725,
+      "learning_rate": 8.767361763398552e-06,
+      "loss": 0.8137,
+      "step": 727
+    },
+    {
+      "epoch": 0.25233968804159446,
+      "grad_norm": 3.496495840070481,
+      "learning_rate": 8.763668301081498e-06,
+      "loss": 0.8747,
+      "step": 728
+    },
+    {
+      "epoch": 0.25268630849220103,
+      "grad_norm": 4.587300414368256,
+      "learning_rate": 8.759970093989466e-06,
+      "loss": 0.8929,
+      "step": 729
+    },
+    {
+      "epoch": 0.2530329289428076,
+      "grad_norm": 7.279531546985616,
+      "learning_rate": 8.756267146784707e-06,
+      "loss": 0.7852,
+      "step": 730
+    },
+    {
+      "epoch": 0.2533795493934142,
+      "grad_norm": 4.703951272675372,
+      "learning_rate": 8.752559464135443e-06,
+      "loss": 0.9131,
+      "step": 731
+    },
+    {
+      "epoch": 0.2537261698440208,
+      "grad_norm": 4.32150667368508,
+      "learning_rate": 8.748847050715873e-06,
+      "loss": 0.7954,
+      "step": 732
+    },
+    {
+      "epoch": 0.25407279029462737,
+      "grad_norm": 3.6137610478703537,
+      "learning_rate": 8.745129911206154e-06,
+      "loss": 0.8272,
+      "step": 733
+    },
+    {
+      "epoch": 0.25441941074523394,
+      "grad_norm": 5.299058881053377,
+      "learning_rate": 8.741408050292401e-06,
+      "loss": 0.7869,
+      "step": 734
+    },
+    {
+      "epoch": 0.25476603119584057,
+      "grad_norm": 3.2080643913068627,
+      "learning_rate": 8.737681472666686e-06,
+      "loss": 0.7746,
+      "step": 735
+    },
+    {
+      "epoch": 0.25511265164644714,
+      "grad_norm": 14.940364968035711,
+      "learning_rate": 8.733950183027023e-06,
+      "loss": 0.7164,
+      "step": 736
+    },
+    {
+      "epoch": 0.2554592720970537,
+      "grad_norm": 4.769940976699451,
+      "learning_rate": 8.73021418607737e-06,
+      "loss": 0.7797,
+      "step": 737
+    },
+    {
+      "epoch": 0.2558058925476603,
+      "grad_norm": 9.020068883800164,
+      "learning_rate": 8.726473486527615e-06,
+      "loss": 0.8874,
+      "step": 738
+    },
+    {
+      "epoch": 0.2561525129982669,
+      "grad_norm": 3.7336833079879836,
+      "learning_rate": 8.72272808909358e-06,
+      "loss": 0.7241,
+      "step": 739
+    },
+    {
+      "epoch": 0.2564991334488735,
+      "grad_norm": 24.438263773586225,
+      "learning_rate": 8.718977998497003e-06,
+      "loss": 0.8472,
+      "step": 740
+    },
+    {
+      "epoch": 0.25684575389948006,
+      "grad_norm": 4.650188385604917,
+      "learning_rate": 8.715223219465542e-06,
+      "loss": 0.8404,
+      "step": 741
+    },
+    {
+      "epoch": 0.2571923743500866,
+      "grad_norm": 3.8861429484429673,
+      "learning_rate": 8.711463756732769e-06,
+      "loss": 0.7932,
+      "step": 742
+    },
+    {
+      "epoch": 0.25753899480069326,
+      "grad_norm": 4.539513305718361,
+      "learning_rate": 8.707699615038152e-06,
+      "loss": 0.9359,
+      "step": 743
+    },
+    {
+      "epoch": 0.2578856152512998,
+      "grad_norm": 8.017661151952668,
+      "learning_rate": 8.703930799127068e-06,
+      "loss": 0.8922,
+      "step": 744
+    },
+    {
+      "epoch": 0.2582322357019064,
+      "grad_norm": 3.6026029354566127,
+      "learning_rate": 8.700157313750776e-06,
+      "loss": 0.853,
+      "step": 745
+    },
+    {
+      "epoch": 0.258578856152513,
+      "grad_norm": 5.317326118007965,
+      "learning_rate": 8.696379163666431e-06,
+      "loss": 0.6375,
+      "step": 746
+    },
+    {
+      "epoch": 0.2589254766031196,
+      "grad_norm": 3.649995167653761,
+      "learning_rate": 8.692596353637066e-06,
+      "loss": 0.8097,
+      "step": 747
+    },
+    {
+      "epoch": 0.25927209705372617,
+      "grad_norm": 3.819367799754751,
+      "learning_rate": 8.688808888431583e-06,
+      "loss": 0.8382,
+      "step": 748
+    },
+    {
+      "epoch": 0.25961871750433274,
+      "grad_norm": 4.694128648744788,
+      "learning_rate": 8.68501677282476e-06,
+      "loss": 0.8502,
+      "step": 749
+    },
+    {
+      "epoch": 0.25996533795493937,
+      "grad_norm": 3.600321643270588,
+      "learning_rate": 8.681220011597235e-06,
+      "loss": 0.8796,
+      "step": 750
+    },
+    {
+      "epoch": 0.26031195840554594,
+      "grad_norm": 2.683416411408272,
+      "learning_rate": 8.677418609535502e-06,
+      "loss": 0.7519,
+      "step": 751
+    },
+    {
+      "epoch": 0.2606585788561525,
+      "grad_norm": 6.493323892200009,
+      "learning_rate": 8.673612571431907e-06,
+      "loss": 0.7373,
+      "step": 752
+    },
+    {
+      "epoch": 0.2610051993067591,
+      "grad_norm": 49.051233786119276,
+      "learning_rate": 8.669801902084636e-06,
+      "loss": 0.7728,
+      "step": 753
+    },
+    {
+      "epoch": 0.2613518197573657,
+      "grad_norm": 11.239886200034823,
+      "learning_rate": 8.665986606297723e-06,
+      "loss": 0.8615,
+      "step": 754
+    },
+    {
+      "epoch": 0.2616984402079723,
+      "grad_norm": 5.662937842749021,
+      "learning_rate": 8.662166688881024e-06,
+      "loss": 0.8479,
+      "step": 755
+    },
+    {
+      "epoch": 0.26204506065857885,
+      "grad_norm": 3.0547970532970203,
+      "learning_rate": 8.658342154650228e-06,
+      "loss": 0.6848,
+      "step": 756
+    },
+    {
+      "epoch": 0.2623916811091854,
+      "grad_norm": 3.312013484728972,
+      "learning_rate": 8.654513008426842e-06,
+      "loss": 0.7379,
+      "step": 757
+    },
+    {
+      "epoch": 0.26273830155979205,
+      "grad_norm": 4.884942153414987,
+      "learning_rate": 8.650679255038186e-06,
+      "loss": 0.7395,
+      "step": 758
+    },
+    {
+      "epoch": 0.2630849220103986,
+      "grad_norm": 3.845083199439453,
+      "learning_rate": 8.646840899317392e-06,
+      "loss": 0.8032,
+      "step": 759
+    },
+    {
+      "epoch": 0.2634315424610052,
+      "grad_norm": 3.33217439611148,
+      "learning_rate": 8.642997946103393e-06,
+      "loss": 0.8529,
+      "step": 760
+    },
+    {
+      "epoch": 0.26377816291161177,
+      "grad_norm": 8.89880586243316,
+      "learning_rate": 8.639150400240911e-06,
+      "loss": 0.768,
+      "step": 761
+    },
+    {
+      "epoch": 0.2641247833622184,
+      "grad_norm": 4.455524576791198,
+      "learning_rate": 8.635298266580472e-06,
+      "loss": 0.8126,
+      "step": 762
+    },
+    {
+      "epoch": 0.26447140381282497,
+      "grad_norm": 4.597003688965997,
+      "learning_rate": 8.63144154997837e-06,
+      "loss": 0.7213,
+      "step": 763
+    },
+    {
+      "epoch": 0.26481802426343154,
+      "grad_norm": 24.034895149235446,
+      "learning_rate": 8.627580255296686e-06,
+      "loss": 0.7183,
+      "step": 764
+    },
+    {
+      "epoch": 0.2651646447140381,
+      "grad_norm": 5.215598997967027,
+      "learning_rate": 8.623714387403273e-06,
+      "loss": 0.8117,
+      "step": 765
+    },
+    {
+      "epoch": 0.26551126516464474,
+      "grad_norm": 5.288864605651122,
+      "learning_rate": 8.619843951171744e-06,
+      "loss": 0.856,
+      "step": 766
+    },
+    {
+      "epoch": 0.2658578856152513,
+      "grad_norm": 3.5587582662662247,
+      "learning_rate": 8.615968951481475e-06,
+      "loss": 0.8607,
+      "step": 767
+    },
+    {
+      "epoch": 0.2662045060658579,
+      "grad_norm": 5.048609579338562,
+      "learning_rate": 8.612089393217594e-06,
+      "loss": 0.7554,
+      "step": 768
+    },
+    {
+      "epoch": 0.26655112651646445,
+      "grad_norm": 6.734800849200541,
+      "learning_rate": 8.608205281270975e-06,
+      "loss": 0.7516,
+      "step": 769
+    },
+    {
+      "epoch": 0.2668977469670711,
+      "grad_norm": 4.329074975944373,
+      "learning_rate": 8.604316620538235e-06,
+      "loss": 0.8677,
+      "step": 770
+    },
+    {
+      "epoch": 0.26724436741767765,
+      "grad_norm": 7.09211550871103,
+      "learning_rate": 8.600423415921727e-06,
+      "loss": 0.8116,
+      "step": 771
+    },
+    {
+      "epoch": 0.2675909878682842,
+      "grad_norm": 3.8856530258997335,
+      "learning_rate": 8.596525672329522e-06,
+      "loss": 0.7144,
+      "step": 772
+    },
+    {
+      "epoch": 0.2679376083188908,
+      "grad_norm": 6.479894896578386,
+      "learning_rate": 8.592623394675426e-06,
+      "loss": 0.7736,
+      "step": 773
+    },
+    {
+      "epoch": 0.2682842287694974,
+      "grad_norm": 4.3082370823537435,
+      "learning_rate": 8.588716587878957e-06,
+      "loss": 0.907,
+      "step": 774
+    },
+    {
+      "epoch": 0.268630849220104,
+      "grad_norm": 4.550617256682194,
+      "learning_rate": 8.584805256865336e-06,
+      "loss": 0.74,
+      "step": 775
+    },
+    {
+      "epoch": 0.26897746967071057,
+      "grad_norm": 9.090831734891506,
+      "learning_rate": 8.5808894065655e-06,
+      "loss": 0.7584,
+      "step": 776
+    },
+    {
+      "epoch": 0.26932409012131714,
+      "grad_norm": 3.5756813849928912,
+      "learning_rate": 8.576969041916072e-06,
+      "loss": 0.8191,
+      "step": 777
+    },
+    {
+      "epoch": 0.26967071057192377,
+      "grad_norm": 2.749761415655523,
+      "learning_rate": 8.573044167859373e-06,
+      "loss": 0.7401,
+      "step": 778
+    },
+    {
+      "epoch": 0.27001733102253034,
+      "grad_norm": 4.091507958362611,
+      "learning_rate": 8.569114789343402e-06,
+      "loss": 0.7335,
+      "step": 779
+    },
+    {
+      "epoch": 0.2703639514731369,
+      "grad_norm": 5.916956370352241,
+      "learning_rate": 8.565180911321846e-06,
+      "loss": 0.7909,
+      "step": 780
+    },
+    {
+      "epoch": 0.2707105719237435,
+      "grad_norm": 7.1356974179799755,
+      "learning_rate": 8.56124253875406e-06,
+      "loss": 0.7723,
+      "step": 781
+    },
+    {
+      "epoch": 0.2710571923743501,
+      "grad_norm": 4.1973574336878094,
+      "learning_rate": 8.557299676605064e-06,
+      "loss": 0.7308,
+      "step": 782
+    },
+    {
+      "epoch": 0.2714038128249567,
+      "grad_norm": 3.12255921259908,
+      "learning_rate": 8.553352329845535e-06,
+      "loss": 0.8248,
+      "step": 783
+    },
+    {
+      "epoch": 0.27175043327556325,
+      "grad_norm": 5.542786104343248,
+      "learning_rate": 8.549400503451813e-06,
+      "loss": 0.9728,
+      "step": 784
+    },
+    {
+      "epoch": 0.2720970537261698,
+      "grad_norm": 3.7077396717212063,
+      "learning_rate": 8.545444202405875e-06,
+      "loss": 0.8073,
+      "step": 785
+    },
+    {
+      "epoch": 0.27244367417677645,
+      "grad_norm": 17.22971287677723,
+      "learning_rate": 8.541483431695347e-06,
+      "loss": 0.8266,
+      "step": 786
+    },
+    {
+      "epoch": 0.272790294627383,
+      "grad_norm": 10.273712108515017,
+      "learning_rate": 8.537518196313487e-06,
+      "loss": 0.7806,
+      "step": 787
+    },
+    {
+      "epoch": 0.2731369150779896,
+      "grad_norm": 35.14325043578418,
+      "learning_rate": 8.533548501259179e-06,
+      "loss": 0.7886,
+      "step": 788
+    },
+    {
+      "epoch": 0.27348353552859617,
+      "grad_norm": 5.866794791116208,
+      "learning_rate": 8.529574351536933e-06,
+      "loss": 0.84,
+      "step": 789
+    },
+    {
+      "epoch": 0.2738301559792028,
+      "grad_norm": 10.109092700551574,
+      "learning_rate": 8.525595752156871e-06,
+      "loss": 0.7316,
+      "step": 790
+    },
+    {
+      "epoch": 0.27417677642980937,
+      "grad_norm": 2.8605598729466903,
+      "learning_rate": 8.52161270813473e-06,
+      "loss": 0.8302,
+      "step": 791
+    },
+    {
+      "epoch": 0.27452339688041594,
+      "grad_norm": 11.268069691691032,
+      "learning_rate": 8.517625224491847e-06,
+      "loss": 0.7202,
+      "step": 792
+    },
+    {
+      "epoch": 0.2748700173310225,
+      "grad_norm": 3.255484374979812,
+      "learning_rate": 8.513633306255153e-06,
+      "loss": 0.8454,
+      "step": 793
+    },
+    {
+      "epoch": 0.27521663778162914,
+      "grad_norm": 2.9360531938898036,
+      "learning_rate": 8.509636958457176e-06,
+      "loss": 0.7919,
+      "step": 794
+    },
+    {
+      "epoch": 0.2755632582322357,
+      "grad_norm": 3.6104695625735492,
+      "learning_rate": 8.505636186136024e-06,
+      "loss": 0.7956,
+      "step": 795
+    },
+    {
+      "epoch": 0.2759098786828423,
+      "grad_norm": 2.7316683465046534,
+      "learning_rate": 8.501630994335384e-06,
+      "loss": 0.7823,
+      "step": 796
+    },
+    {
+      "epoch": 0.27625649913344885,
+      "grad_norm": 2.9068073948698254,
+      "learning_rate": 8.497621388104514e-06,
+      "loss": 0.7909,
+      "step": 797
+    },
+    {
+      "epoch": 0.2766031195840555,
+      "grad_norm": 5.871005549459595,
+      "learning_rate": 8.493607372498236e-06,
+      "loss": 0.8253,
+      "step": 798
+    },
+    {
+      "epoch": 0.27694974003466205,
+      "grad_norm": 4.195723254996051,
+      "learning_rate": 8.489588952576935e-06,
+      "loss": 0.8501,
+      "step": 799
+    },
+    {
+      "epoch": 0.2772963604852686,
+      "grad_norm": 10.369782027350366,
+      "learning_rate": 8.485566133406547e-06,
+      "loss": 0.7942,
+      "step": 800
+    },
+    {
+      "epoch": 0.2776429809358752,
+      "grad_norm": 6.445482836544835,
+      "learning_rate": 8.481538920058547e-06,
+      "loss": 0.7728,
+      "step": 801
+    },
+    {
+      "epoch": 0.2779896013864818,
+      "grad_norm": 5.43820651060236,
+      "learning_rate": 8.477507317609961e-06,
+      "loss": 0.7012,
+      "step": 802
+    },
+    {
+      "epoch": 0.2783362218370884,
+      "grad_norm": 8.012920693282723,
+      "learning_rate": 8.473471331143341e-06,
+      "loss": 0.8412,
+      "step": 803
+    },
+    {
+      "epoch": 0.27868284228769497,
+      "grad_norm": 3.8804346639586944,
+      "learning_rate": 8.469430965746767e-06,
+      "loss": 0.746,
+      "step": 804
+    },
+    {
+      "epoch": 0.27902946273830154,
+      "grad_norm": 4.881936262850953,
+      "learning_rate": 8.46538622651384e-06,
+      "loss": 0.8072,
+      "step": 805
+    },
+    {
+      "epoch": 0.27937608318890816,
+      "grad_norm": 5.977625151280764,
+      "learning_rate": 8.461337118543678e-06,
+      "loss": 0.8321,
+      "step": 806
+    },
+    {
+      "epoch": 0.27972270363951474,
+      "grad_norm": 3.2525894499647854,
+      "learning_rate": 8.4572836469409e-06,
+      "loss": 0.9119,
+      "step": 807
+    },
+    {
+      "epoch": 0.2800693240901213,
+      "grad_norm": 4.443929902258599,
+      "learning_rate": 8.453225816815629e-06,
+      "loss": 0.6918,
+      "step": 808
+    },
+    {
+      "epoch": 0.2804159445407279,
+      "grad_norm": 3.8359431189958695,
+      "learning_rate": 8.449163633283488e-06,
+      "loss": 0.817,
+      "step": 809
+    },
+    {
+      "epoch": 0.2807625649913345,
+      "grad_norm": 3.812580841489284,
+      "learning_rate": 8.445097101465579e-06,
+      "loss": 0.6989,
+      "step": 810
+    },
+    {
+      "epoch": 0.2811091854419411,
+      "grad_norm": 3.907212333065873,
+      "learning_rate": 8.441026226488493e-06,
+      "loss": 0.7779,
+      "step": 811
+    },
+    {
+      "epoch": 0.28145580589254765,
+      "grad_norm": 5.635415142435284,
+      "learning_rate": 8.436951013484294e-06,
+      "loss": 0.7999,
+      "step": 812
+    },
+    {
+      "epoch": 0.2818024263431542,
+      "grad_norm": 53.69358115502805,
+      "learning_rate": 8.43287146759051e-06,
+      "loss": 0.7593,
+      "step": 813
+    },
+    {
+      "epoch": 0.28214904679376085,
+      "grad_norm": 3.5919922291728845,
+      "learning_rate": 8.428787593950144e-06,
+      "loss": 0.8189,
+      "step": 814
+    },
+    {
+      "epoch": 0.2824956672443674,
+      "grad_norm": 4.719179850300029,
+      "learning_rate": 8.424699397711642e-06,
+      "loss": 0.7804,
+      "step": 815
+    },
+    {
+      "epoch": 0.282842287694974,
+      "grad_norm": 6.632734957192689,
+      "learning_rate": 8.420606884028904e-06,
+      "loss": 0.777,
+      "step": 816
+    },
+    {
+      "epoch": 0.28318890814558056,
+      "grad_norm": 2.6823959697738773,
+      "learning_rate": 8.416510058061273e-06,
+      "loss": 0.717,
+      "step": 817
+    },
+    {
+      "epoch": 0.2835355285961872,
+      "grad_norm": 3.5300802144735166,
+      "learning_rate": 8.412408924973532e-06,
+      "loss": 0.7243,
+      "step": 818
+    },
+    {
+      "epoch": 0.28388214904679376,
+      "grad_norm": 2.7216746323554455,
+      "learning_rate": 8.408303489935887e-06,
+      "loss": 0.7562,
+      "step": 819
+    },
+    {
+      "epoch": 0.28422876949740034,
+      "grad_norm": 3.4244688058792017,
+      "learning_rate": 8.404193758123974e-06,
+      "loss": 0.8359,
+      "step": 820
+    },
+    {
+      "epoch": 0.2845753899480069,
+      "grad_norm": 4.242448596774674,
+      "learning_rate": 8.400079734718839e-06,
+      "loss": 0.7558,
+      "step": 821
+    },
+    {
+      "epoch": 0.28492201039861353,
+      "grad_norm": 4.423593384628689,
+      "learning_rate": 8.395961424906941e-06,
+      "loss": 0.752,
+      "step": 822
+    },
+    {
+      "epoch": 0.2852686308492201,
+      "grad_norm": 4.6441164671738715,
+      "learning_rate": 8.391838833880152e-06,
+      "loss": 0.8105,
+      "step": 823
+    },
+    {
+      "epoch": 0.2856152512998267,
+      "grad_norm": 4.228442098945691,
+      "learning_rate": 8.387711966835727e-06,
+      "loss": 0.7731,
+      "step": 824
+    },
+    {
+      "epoch": 0.28596187175043325,
+      "grad_norm": 3.5855936525951235,
+      "learning_rate": 8.383580828976318e-06,
+      "loss": 0.7636,
+      "step": 825
+    },
+    {
+      "epoch": 0.2863084922010399,
+      "grad_norm": 3.32534335958499,
+      "learning_rate": 8.37944542550996e-06,
+      "loss": 0.851,
+      "step": 826
+    },
+    {
+      "epoch": 0.28665511265164645,
+      "grad_norm": 3.223997270440258,
+      "learning_rate": 8.375305761650069e-06,
+      "loss": 0.9143,
+      "step": 827
+    },
+    {
+      "epoch": 0.287001733102253,
+      "grad_norm": 7.282218115255968,
+      "learning_rate": 8.37116184261543e-06,
+      "loss": 0.8901,
+      "step": 828
+    },
+    {
+      "epoch": 0.2873483535528596,
+      "grad_norm": 7.047303720108605,
+      "learning_rate": 8.367013673630189e-06,
+      "loss": 0.774,
+      "step": 829
+    },
+    {
+      "epoch": 0.2876949740034662,
+      "grad_norm": 6.181104580268031,
+      "learning_rate": 8.362861259923853e-06,
+      "loss": 0.8009,
+      "step": 830
+    },
+    {
+      "epoch": 0.2880415944540728,
+      "grad_norm": 3.143308786928076,
+      "learning_rate": 8.358704606731283e-06,
+      "loss": 0.7898,
+      "step": 831
+    },
+    {
+      "epoch": 0.28838821490467936,
+      "grad_norm": 4.0207242501552765,
+      "learning_rate": 8.354543719292676e-06,
+      "loss": 0.8437,
+      "step": 832
+    },
+    {
+      "epoch": 0.28873483535528593,
+      "grad_norm": 10.787342779247131,
+      "learning_rate": 8.350378602853578e-06,
+      "loss": 0.8715,
+      "step": 833
+    },
+    {
+      "epoch": 0.28908145580589256,
+      "grad_norm": 6.345751702330577,
+      "learning_rate": 8.346209262664859e-06,
+      "loss": 0.7599,
+      "step": 834
+    },
+    {
+      "epoch": 0.28942807625649913,
+      "grad_norm": 4.244261310460595,
+      "learning_rate": 8.342035703982714e-06,
+      "loss": 0.798,
+      "step": 835
+    },
+    {
+      "epoch": 0.2897746967071057,
+      "grad_norm": 3.817161060728082,
+      "learning_rate": 8.337857932068661e-06,
+      "loss": 0.8213,
+      "step": 836
+    },
+    {
+      "epoch": 0.2901213171577123,
+      "grad_norm": 4.186759375434355,
+      "learning_rate": 8.333675952189522e-06,
+      "loss": 0.9006,
+      "step": 837
+    },
+    {
+      "epoch": 0.2904679376083189,
+      "grad_norm": 3.3805606320587778,
+      "learning_rate": 8.329489769617433e-06,
+      "loss": 0.8151,
+      "step": 838
+    },
+    {
+      "epoch": 0.2908145580589255,
+      "grad_norm": 3.982475088415894,
+      "learning_rate": 8.325299389629821e-06,
+      "loss": 0.8603,
+      "step": 839
+    },
+    {
+      "epoch": 0.29116117850953205,
+      "grad_norm": 14.407602558222175,
+      "learning_rate": 8.321104817509409e-06,
+      "loss": 0.7626,
+      "step": 840
+    },
+    {
+      "epoch": 0.2915077989601387,
+      "grad_norm": 5.445546486890854,
+      "learning_rate": 8.316906058544201e-06,
+      "loss": 0.9216,
+      "step": 841
+    },
+    {
+      "epoch": 0.29185441941074525,
+      "grad_norm": 4.584534921827242,
+      "learning_rate": 8.312703118027483e-06,
+      "loss": 0.8022,
+      "step": 842
+    },
+    {
+      "epoch": 0.2922010398613518,
+      "grad_norm": 3.5483771354613345,
+      "learning_rate": 8.30849600125781e-06,
+      "loss": 0.7924,
+      "step": 843
+    },
+    {
+      "epoch": 0.2925476603119584,
+      "grad_norm": 3.0068699630616322,
+      "learning_rate": 8.304284713539002e-06,
+      "loss": 0.8513,
+      "step": 844
+    },
+    {
+      "epoch": 0.292894280762565,
+      "grad_norm": 4.132510162427907,
+      "learning_rate": 8.300069260180138e-06,
+      "loss": 0.7561,
+      "step": 845
+    },
+    {
+      "epoch": 0.2932409012131716,
+      "grad_norm": 3.548418938225706,
+      "learning_rate": 8.29584964649555e-06,
+      "loss": 0.7945,
+      "step": 846
+    },
+    {
+      "epoch": 0.29358752166377816,
+      "grad_norm": 5.6194316567305,
+      "learning_rate": 8.291625877804816e-06,
+      "loss": 0.8409,
+      "step": 847
+    },
+    {
+      "epoch": 0.29393414211438473,
+      "grad_norm": 3.2377929032421853,
+      "learning_rate": 8.287397959432744e-06,
+      "loss": 0.7132,
+      "step": 848
+    },
+    {
+      "epoch": 0.29428076256499136,
+      "grad_norm": 3.96440467277706,
+      "learning_rate": 8.283165896709381e-06,
+      "loss": 0.8332,
+      "step": 849
+    },
+    {
+      "epoch": 0.29462738301559793,
+      "grad_norm": 2.734688518789168,
+      "learning_rate": 8.27892969497e-06,
+      "loss": 0.8095,
+      "step": 850
+    },
+    {
+      "epoch": 0.2949740034662045,
+      "grad_norm": 40.60767971142399,
+      "learning_rate": 8.274689359555083e-06,
+      "loss": 0.8279,
+      "step": 851
+    },
+    {
+      "epoch": 0.2953206239168111,
+      "grad_norm": 5.769002495566961,
+      "learning_rate": 8.270444895810332e-06,
+      "loss": 0.7859,
+      "step": 852
+    },
+    {
+      "epoch": 0.2956672443674177,
+      "grad_norm": 4.116612520825564,
+      "learning_rate": 8.266196309086652e-06,
+      "loss": 0.7356,
+      "step": 853
+    },
+    {
+      "epoch": 0.2960138648180243,
+      "grad_norm": 3.477482498977536,
+      "learning_rate": 8.26194360474014e-06,
+      "loss": 0.7508,
+      "step": 854
+    },
+    {
+      "epoch": 0.29636048526863085,
+      "grad_norm": 8.718522071210328,
+      "learning_rate": 8.257686788132093e-06,
+      "loss": 0.9481,
+      "step": 855
+    },
+    {
+      "epoch": 0.2967071057192374,
+      "grad_norm": 3.9235899797275913,
+      "learning_rate": 8.253425864628984e-06,
+      "loss": 0.8052,
+      "step": 856
+    },
+    {
+      "epoch": 0.29705372616984405,
+      "grad_norm": 4.475847357109287,
+      "learning_rate": 8.249160839602465e-06,
+      "loss": 0.7799,
+      "step": 857
+    },
+    {
+      "epoch": 0.2974003466204506,
+      "grad_norm": 5.472226356227598,
+      "learning_rate": 8.244891718429365e-06,
+      "loss": 0.8466,
+      "step": 858
+    },
+    {
+      "epoch": 0.2977469670710572,
+      "grad_norm": 25.566651097648116,
+      "learning_rate": 8.24061850649167e-06,
+      "loss": 0.8636,
+      "step": 859
+    },
+    {
+      "epoch": 0.29809358752166376,
+      "grad_norm": 7.7686474057931045,
+      "learning_rate": 8.236341209176526e-06,
+      "loss": 0.8803,
+      "step": 860
+    },
+    {
+      "epoch": 0.2984402079722704,
+      "grad_norm": 4.6702252117260485,
+      "learning_rate": 8.23205983187623e-06,
+      "loss": 0.7934,
+      "step": 861
+    },
+    {
+      "epoch": 0.29878682842287696,
+      "grad_norm": 4.1669315267369225,
+      "learning_rate": 8.227774379988218e-06,
+      "loss": 0.8559,
+      "step": 862
+    },
+    {
+      "epoch": 0.29913344887348353,
+      "grad_norm": 4.374077212220792,
+      "learning_rate": 8.223484858915071e-06,
+      "loss": 0.889,
+      "step": 863
+    },
+    {
+      "epoch": 0.2994800693240901,
+      "grad_norm": 4.0719418298126335,
+      "learning_rate": 8.219191274064493e-06,
+      "loss": 0.7463,
+      "step": 864
+    },
+    {
+      "epoch": 0.29982668977469673,
+      "grad_norm": 3.1161529911148373,
+      "learning_rate": 8.214893630849311e-06,
+      "loss": 0.7738,
+      "step": 865
+    },
+    {
+      "epoch": 0.3001733102253033,
+      "grad_norm": 9.994064197260785,
+      "learning_rate": 8.210591934687475e-06,
+      "loss": 0.7484,
+      "step": 866
+    },
+    {
+      "epoch": 0.3005199306759099,
+      "grad_norm": 3.4991718182281097,
+      "learning_rate": 8.206286191002036e-06,
+      "loss": 0.7664,
+      "step": 867
+    },
+    {
+      "epoch": 0.30086655112651645,
+      "grad_norm": 4.515758368786123,
+      "learning_rate": 8.201976405221155e-06,
+      "loss": 0.697,
+      "step": 868
+    },
+    {
+      "epoch": 0.3012131715771231,
+      "grad_norm": 3.6439523915278045,
+      "learning_rate": 8.197662582778084e-06,
+      "loss": 0.7443,
+      "step": 869
+    },
+    {
+      "epoch": 0.30155979202772965,
+      "grad_norm": 4.093146614489871,
+      "learning_rate": 8.193344729111162e-06,
+      "loss": 0.67,
+      "step": 870
+    },
+    {
+      "epoch": 0.3019064124783362,
+      "grad_norm": 7.301032327626104,
+      "learning_rate": 8.189022849663819e-06,
+      "loss": 0.7447,
+      "step": 871
+    },
+    {
+      "epoch": 0.3022530329289428,
+      "grad_norm": 11.5709913537412,
+      "learning_rate": 8.184696949884552e-06,
+      "loss": 0.7688,
+      "step": 872
+    },
+    {
+      "epoch": 0.3025996533795494,
+      "grad_norm": 3.622442869436363,
+      "learning_rate": 8.18036703522693e-06,
+      "loss": 0.7936,
+      "step": 873
+    },
+    {
+      "epoch": 0.302946273830156,
+      "grad_norm": 4.915858767121732,
+      "learning_rate": 8.176033111149582e-06,
+      "loss": 0.7624,
+      "step": 874
+    },
+    {
+      "epoch": 0.30329289428076256,
+      "grad_norm": 6.009953822004191,
+      "learning_rate": 8.17169518311619e-06,
+      "loss": 0.6856,
+      "step": 875
+    },
+    {
+      "epoch": 0.30363951473136913,
+      "grad_norm": 3.7096197214116593,
+      "learning_rate": 8.167353256595488e-06,
+      "loss": 0.7266,
+      "step": 876
+    },
+    {
+      "epoch": 0.30398613518197576,
+      "grad_norm": 6.182622511277383,
+      "learning_rate": 8.163007337061248e-06,
+      "loss": 0.8324,
+      "step": 877
+    },
+    {
+      "epoch": 0.30433275563258233,
+      "grad_norm": 4.163342369550136,
+      "learning_rate": 8.158657429992279e-06,
+      "loss": 0.7619,
+      "step": 878
+    },
+    {
+      "epoch": 0.3046793760831889,
+      "grad_norm": 4.492885673872653,
+      "learning_rate": 8.15430354087241e-06,
+      "loss": 0.7438,
+      "step": 879
+    },
+    {
+      "epoch": 0.3050259965337955,
+      "grad_norm": 4.425283305570679,
+      "learning_rate": 8.149945675190497e-06,
+      "loss": 0.9176,
+      "step": 880
+    },
+    {
+      "epoch": 0.3053726169844021,
+      "grad_norm": 3.182903535547134,
+      "learning_rate": 8.145583838440407e-06,
+      "loss": 0.8124,
+      "step": 881
+    },
+    {
+      "epoch": 0.3057192374350087,
+      "grad_norm": 5.174912225211832,
+      "learning_rate": 8.141218036121012e-06,
+      "loss": 0.6125,
+      "step": 882
+    },
+    {
+      "epoch": 0.30606585788561524,
+      "grad_norm": 3.64885059564314,
+      "learning_rate": 8.136848273736184e-06,
+      "loss": 0.8505,
+      "step": 883
+    },
+    {
+      "epoch": 0.3064124783362218,
+      "grad_norm": 5.83483643082411,
+      "learning_rate": 8.13247455679479e-06,
+      "loss": 0.8491,
+      "step": 884
+    },
+    {
+      "epoch": 0.30675909878682844,
+      "grad_norm": 3.857200807664267,
+      "learning_rate": 8.128096890810678e-06,
+      "loss": 0.7609,
+      "step": 885
+    },
+    {
+      "epoch": 0.307105719237435,
+      "grad_norm": 4.458754695634749,
+      "learning_rate": 8.123715281302675e-06,
+      "loss": 0.78,
+      "step": 886
+    },
+    {
+      "epoch": 0.3074523396880416,
+      "grad_norm": 6.452102305121821,
+      "learning_rate": 8.119329733794583e-06,
+      "loss": 0.8492,
+      "step": 887
+    },
+    {
+      "epoch": 0.30779896013864816,
+      "grad_norm": 8.466769594336206,
+      "learning_rate": 8.11494025381517e-06,
+      "loss": 0.7579,
+      "step": 888
+    },
+    {
+      "epoch": 0.3081455805892548,
+      "grad_norm": 6.602920406983597,
+      "learning_rate": 8.110546846898151e-06,
+      "loss": 0.7689,
+      "step": 889
+    },
+    {
+      "epoch": 0.30849220103986136,
+      "grad_norm": 4.2445598598340535,
+      "learning_rate": 8.106149518582204e-06,
+      "loss": 0.772,
+      "step": 890
+    },
+    {
+      "epoch": 0.30883882149046793,
+      "grad_norm": 3.7716283514749573,
+      "learning_rate": 8.101748274410943e-06,
+      "loss": 0.778,
+      "step": 891
+    },
+    {
+      "epoch": 0.3091854419410745,
+      "grad_norm": 20.710073796657927,
+      "learning_rate": 8.097343119932924e-06,
+      "loss": 0.7879,
+      "step": 892
+    },
+    {
+      "epoch": 0.30953206239168113,
+      "grad_norm": 6.9636562757011715,
+      "learning_rate": 8.092934060701627e-06,
+      "loss": 0.7475,
+      "step": 893
+    },
+    {
+      "epoch": 0.3098786828422877,
+      "grad_norm": 5.898883114801366,
+      "learning_rate": 8.088521102275455e-06,
+      "loss": 0.8511,
+      "step": 894
+    },
+    {
+      "epoch": 0.31022530329289427,
+      "grad_norm": 3.2713518938280863,
+      "learning_rate": 8.084104250217736e-06,
+      "loss": 0.8842,
+      "step": 895
+    },
+    {
+      "epoch": 0.31057192374350084,
+      "grad_norm": 2.961796537058863,
+      "learning_rate": 8.079683510096693e-06,
+      "loss": 0.7299,
+      "step": 896
+    },
+    {
+      "epoch": 0.31091854419410747,
+      "grad_norm": 3.4329331522384634,
+      "learning_rate": 8.075258887485463e-06,
+      "loss": 0.8572,
+      "step": 897
+    },
+    {
+      "epoch": 0.31126516464471404,
+      "grad_norm": 6.542115423723931,
+      "learning_rate": 8.070830387962064e-06,
+      "loss": 0.7784,
+      "step": 898
+    },
+    {
+      "epoch": 0.3116117850953206,
+      "grad_norm": 7.28572953184547,
+      "learning_rate": 8.066398017109416e-06,
+      "loss": 0.8252,
+      "step": 899
+    },
+    {
+      "epoch": 0.3119584055459272,
+      "grad_norm": 3.1385572094437255,
+      "learning_rate": 8.061961780515311e-06,
+      "loss": 0.7452,
+      "step": 900
+    },
+    {
+      "epoch": 0.3123050259965338,
+      "grad_norm": 5.3112749267079655,
+      "learning_rate": 8.057521683772413e-06,
+      "loss": 0.7068,
+      "step": 901
+    },
+    {
+      "epoch": 0.3126516464471404,
+      "grad_norm": 3.5114169077062645,
+      "learning_rate": 8.053077732478261e-06,
+      "loss": 0.8817,
+      "step": 902
+    },
+    {
+      "epoch": 0.31299826689774696,
+      "grad_norm": 2.5963763229546397,
+      "learning_rate": 8.048629932235245e-06,
+      "loss": 0.8223,
+      "step": 903
+    },
+    {
+      "epoch": 0.31334488734835353,
+      "grad_norm": 5.044570650143747,
+      "learning_rate": 8.044178288650609e-06,
+      "loss": 0.7994,
+      "step": 904
+    },
+    {
+      "epoch": 0.31369150779896016,
+      "grad_norm": 4.390351831765179,
+      "learning_rate": 8.039722807336447e-06,
+      "loss": 0.8785,
+      "step": 905
+    },
+    {
+      "epoch": 0.31403812824956673,
+      "grad_norm": 5.48280599641561,
+      "learning_rate": 8.035263493909683e-06,
+      "loss": 0.8024,
+      "step": 906
+    },
+    {
+      "epoch": 0.3143847487001733,
+      "grad_norm": 3.144804220759166,
+      "learning_rate": 8.030800353992083e-06,
+      "loss": 0.7334,
+      "step": 907
+    },
+    {
+      "epoch": 0.31473136915077987,
+      "grad_norm": 7.101668473663387,
+      "learning_rate": 8.026333393210226e-06,
+      "loss": 0.675,
+      "step": 908
+    },
+    {
+      "epoch": 0.3150779896013865,
+      "grad_norm": 5.683286011097988,
+      "learning_rate": 8.021862617195513e-06,
+      "loss": 0.8611,
+      "step": 909
+    },
+    {
+      "epoch": 0.31542461005199307,
+      "grad_norm": 5.515272057561956,
+      "learning_rate": 8.017388031584155e-06,
+      "loss": 0.8489,
+      "step": 910
+    },
+    {
+      "epoch": 0.31577123050259964,
+      "grad_norm": 7.172685114257443,
+      "learning_rate": 8.012909642017164e-06,
+      "loss": 0.8043,
+      "step": 911
+    },
+    {
+      "epoch": 0.3161178509532062,
+      "grad_norm": 12.964613700419305,
+      "learning_rate": 8.00842745414035e-06,
+      "loss": 0.7497,
+      "step": 912
+    },
+    {
+      "epoch": 0.31646447140381284,
+      "grad_norm": 4.9089389154777,
+      "learning_rate": 8.003941473604312e-06,
+      "loss": 0.8236,
+      "step": 913
+    },
+    {
+      "epoch": 0.3168110918544194,
+      "grad_norm": 4.0799127930893855,
+      "learning_rate": 7.999451706064425e-06,
+      "loss": 0.8109,
+      "step": 914
+    },
+    {
+      "epoch": 0.317157712305026,
+      "grad_norm": 6.550089739101057,
+      "learning_rate": 7.994958157180842e-06,
+      "loss": 0.735,
+      "step": 915
+    },
+    {
+      "epoch": 0.31750433275563256,
+      "grad_norm": 3.80748438197191,
+      "learning_rate": 7.99046083261848e-06,
+      "loss": 0.7517,
+      "step": 916
+    },
+    {
+      "epoch": 0.3178509532062392,
+      "grad_norm": 6.099064033448688,
+      "learning_rate": 7.985959738047026e-06,
+      "loss": 0.823,
+      "step": 917
+    },
+    {
+      "epoch": 0.31819757365684576,
+      "grad_norm": 4.477473097017026,
+      "learning_rate": 7.981454879140906e-06,
+      "loss": 0.8235,
+      "step": 918
+    },
+    {
+      "epoch": 0.3185441941074523,
+      "grad_norm": 6.544269492078495,
+      "learning_rate": 7.9769462615793e-06,
+      "loss": 0.7461,
+      "step": 919
+    },
+    {
+      "epoch": 0.3188908145580589,
+      "grad_norm": 2.9039393509811173,
+      "learning_rate": 7.972433891046123e-06,
+      "loss": 0.8328,
+      "step": 920
+    },
+    {
+      "epoch": 0.3192374350086655,
+      "grad_norm": 5.570864010136834,
+      "learning_rate": 7.967917773230025e-06,
+      "loss": 0.8457,
+      "step": 921
+    },
+    {
+      "epoch": 0.3195840554592721,
+      "grad_norm": 4.7172718923045815,
+      "learning_rate": 7.963397913824376e-06,
+      "loss": 0.8125,
+      "step": 922
+    },
+    {
+      "epoch": 0.31993067590987867,
+      "grad_norm": 3.9188831034198466,
+      "learning_rate": 7.958874318527265e-06,
+      "loss": 0.7815,
+      "step": 923
+    },
+    {
+      "epoch": 0.32027729636048524,
+      "grad_norm": 3.6883484609120973,
+      "learning_rate": 7.95434699304149e-06,
+      "loss": 0.7585,
+      "step": 924
+    },
+    {
+      "epoch": 0.32062391681109187,
+      "grad_norm": 5.893401921902247,
+      "learning_rate": 7.949815943074555e-06,
+      "loss": 0.802,
+      "step": 925
+    },
+    {
+      "epoch": 0.32097053726169844,
+      "grad_norm": 17.111434758026032,
+      "learning_rate": 7.945281174338652e-06,
+      "loss": 0.8619,
+      "step": 926
+    },
+    {
+      "epoch": 0.321317157712305,
+      "grad_norm": 2.461131318546762,
+      "learning_rate": 7.94074269255067e-06,
+      "loss": 0.6881,
+      "step": 927
+    },
+    {
+      "epoch": 0.3216637781629116,
+      "grad_norm": 3.4849505509197307,
+      "learning_rate": 7.936200503432169e-06,
+      "loss": 0.858,
+      "step": 928
+    },
+    {
+      "epoch": 0.3220103986135182,
+      "grad_norm": 10.586547650132625,
+      "learning_rate": 7.931654612709393e-06,
+      "loss": 0.792,
+      "step": 929
+    },
+    {
+      "epoch": 0.3223570190641248,
+      "grad_norm": 3.9566096537164346,
+      "learning_rate": 7.927105026113246e-06,
+      "loss": 0.7545,
+      "step": 930
+    },
+    {
+      "epoch": 0.32270363951473136,
+      "grad_norm": 3.216064881305468,
+      "learning_rate": 7.922551749379295e-06,
+      "loss": 0.7981,
+      "step": 931
+    },
+    {
+      "epoch": 0.323050259965338,
+      "grad_norm": 4.620583338847734,
+      "learning_rate": 7.917994788247754e-06,
+      "loss": 0.6702,
+      "step": 932
+    },
+    {
+      "epoch": 0.32339688041594455,
+      "grad_norm": 4.447356995259258,
+      "learning_rate": 7.913434148463486e-06,
+      "loss": 0.8688,
+      "step": 933
+    },
+    {
+      "epoch": 0.3237435008665511,
+      "grad_norm": 3.5142756414824703,
+      "learning_rate": 7.908869835775991e-06,
+      "loss": 0.8106,
+      "step": 934
+    },
+    {
+      "epoch": 0.3240901213171577,
+      "grad_norm": 3.7191444691296685,
+      "learning_rate": 7.904301855939401e-06,
+      "loss": 0.6938,
+      "step": 935
+    },
+    {
+      "epoch": 0.3244367417677643,
+      "grad_norm": 2.8603950253426165,
+      "learning_rate": 7.899730214712466e-06,
+      "loss": 0.7351,
+      "step": 936
+    },
+    {
+      "epoch": 0.3247833622183709,
+      "grad_norm": 3.880561103920444,
+      "learning_rate": 7.895154917858559e-06,
+      "loss": 0.7948,
+      "step": 937
+    },
+    {
+      "epoch": 0.32512998266897747,
+      "grad_norm": 4.461763587659822,
+      "learning_rate": 7.890575971145651e-06,
+      "loss": 0.7563,
+      "step": 938
+    },
+    {
+      "epoch": 0.32547660311958404,
+      "grad_norm": 4.903452977935244,
+      "learning_rate": 7.885993380346328e-06,
+      "loss": 0.8241,
+      "step": 939
+    },
+    {
+      "epoch": 0.32582322357019067,
+      "grad_norm": 23.275683298383786,
+      "learning_rate": 7.881407151237756e-06,
+      "loss": 0.7719,
+      "step": 940
+    },
+    {
+      "epoch": 0.32616984402079724,
+      "grad_norm": 5.61730161619977,
+      "learning_rate": 7.8768172896017e-06,
+      "loss": 0.6821,
+      "step": 941
+    },
+    {
+      "epoch": 0.3265164644714038,
+      "grad_norm": 2.3698495399267534,
+      "learning_rate": 7.872223801224496e-06,
+      "loss": 0.7835,
+      "step": 942
+    },
+    {
+      "epoch": 0.3268630849220104,
+      "grad_norm": 3.4487069932641763,
+      "learning_rate": 7.867626691897056e-06,
+      "loss": 0.7746,
+      "step": 943
+    },
+    {
+      "epoch": 0.327209705372617,
+      "grad_norm": 6.026031271078959,
+      "learning_rate": 7.863025967414856e-06,
+      "loss": 0.7303,
+      "step": 944
+    },
+    {
+      "epoch": 0.3275563258232236,
+      "grad_norm": 6.508346042842791,
+      "learning_rate": 7.858421633577929e-06,
+      "loss": 0.8047,
+      "step": 945
+    },
+    {
+      "epoch": 0.32790294627383015,
+      "grad_norm": 11.444302070648359,
+      "learning_rate": 7.853813696190862e-06,
+      "loss": 0.837,
+      "step": 946
+    },
+    {
+      "epoch": 0.3282495667244367,
+      "grad_norm": 3.599568202094595,
+      "learning_rate": 7.849202161062776e-06,
+      "loss": 0.7622,
+      "step": 947
+    },
+    {
+      "epoch": 0.32859618717504335,
+      "grad_norm": 4.105029240472894,
+      "learning_rate": 7.844587034007337e-06,
+      "loss": 0.6992,
+      "step": 948
+    },
+    {
+      "epoch": 0.3289428076256499,
+      "grad_norm": 3.581680831729903,
+      "learning_rate": 7.839968320842737e-06,
+      "loss": 0.803,
+      "step": 949
+    },
+    {
+      "epoch": 0.3292894280762565,
+      "grad_norm": 2.456065037774029,
+      "learning_rate": 7.835346027391682e-06,
+      "loss": 0.768,
+      "step": 950
+    },
+    {
+      "epoch": 0.32963604852686307,
+      "grad_norm": 38.55449809667737,
+      "learning_rate": 7.830720159481404e-06,
+      "loss": 0.647,
+      "step": 951
+    },
+    {
+      "epoch": 0.3299826689774697,
+      "grad_norm": 5.805154903782192,
+      "learning_rate": 7.826090722943627e-06,
+      "loss": 0.7821,
+      "step": 952
+    },
+    {
+      "epoch": 0.33032928942807627,
+      "grad_norm": 6.897018562883929,
+      "learning_rate": 7.821457723614587e-06,
+      "loss": 0.7594,
+      "step": 953
+    },
+    {
+      "epoch": 0.33067590987868284,
+      "grad_norm": 3.60659809983751,
+      "learning_rate": 7.816821167335005e-06,
+      "loss": 0.7925,
+      "step": 954
+    },
+    {
+      "epoch": 0.3310225303292894,
+      "grad_norm": 3.5746239674949867,
+      "learning_rate": 7.812181059950084e-06,
+      "loss": 0.701,
+      "step": 955
+    },
+    {
+      "epoch": 0.33136915077989604,
+      "grad_norm": 4.596015729928653,
+      "learning_rate": 7.807537407309508e-06,
+      "loss": 0.7456,
+      "step": 956
+    },
+    {
+      "epoch": 0.3317157712305026,
+      "grad_norm": 3.0045413718419063,
+      "learning_rate": 7.802890215267429e-06,
+      "loss": 0.8666,
+      "step": 957
+    },
+    {
+      "epoch": 0.3320623916811092,
+      "grad_norm": 3.2382258603601266,
+      "learning_rate": 7.798239489682463e-06,
+      "loss": 0.7956,
+      "step": 958
+    },
+    {
+      "epoch": 0.33240901213171575,
+      "grad_norm": 5.668231319749799,
+      "learning_rate": 7.793585236417678e-06,
+      "loss": 0.6837,
+      "step": 959
+    },
+    {
+      "epoch": 0.3327556325823224,
+      "grad_norm": 3.7555147356343603,
+      "learning_rate": 7.788927461340589e-06,
+      "loss": 0.8594,
+      "step": 960
+    },
+    {
+      "epoch": 0.33310225303292895,
+      "grad_norm": 2.9726114158673878,
+      "learning_rate": 7.784266170323152e-06,
+      "loss": 0.8321,
+      "step": 961
+    },
+    {
+      "epoch": 0.3334488734835355,
+      "grad_norm": 4.150824778876615,
+      "learning_rate": 7.779601369241757e-06,
+      "loss": 0.7662,
+      "step": 962
+    },
+    {
+      "epoch": 0.3337954939341421,
+      "grad_norm": 4.325048797591292,
+      "learning_rate": 7.774933063977218e-06,
+      "loss": 0.7346,
+      "step": 963
+    },
+    {
+      "epoch": 0.3341421143847487,
+      "grad_norm": 2.8964015166123724,
+      "learning_rate": 7.770261260414763e-06,
+      "loss": 0.8789,
+      "step": 964
+    },
+    {
+      "epoch": 0.3344887348353553,
+      "grad_norm": 23.138870419053067,
+      "learning_rate": 7.765585964444034e-06,
+      "loss": 0.757,
+      "step": 965
+    },
+    {
+      "epoch": 0.33483535528596187,
+      "grad_norm": 2.914994070073769,
+      "learning_rate": 7.760907181959079e-06,
+      "loss": 0.7728,
+      "step": 966
+    },
+    {
+      "epoch": 0.33518197573656844,
+      "grad_norm": 4.380629283898593,
+      "learning_rate": 7.756224918858334e-06,
+      "loss": 0.8505,
+      "step": 967
+    },
+    {
+      "epoch": 0.33552859618717507,
+      "grad_norm": 7.004229453174149,
+      "learning_rate": 7.751539181044628e-06,
+      "loss": 0.808,
+      "step": 968
+    },
+    {
+      "epoch": 0.33587521663778164,
+      "grad_norm": 3.801783488927696,
+      "learning_rate": 7.746849974425167e-06,
+      "loss": 0.6931,
+      "step": 969
+    },
+    {
+      "epoch": 0.3362218370883882,
+      "grad_norm": 3.2594002690722585,
+      "learning_rate": 7.742157304911533e-06,
+      "loss": 0.8092,
+      "step": 970
+    },
+    {
+      "epoch": 0.3365684575389948,
+      "grad_norm": 3.220157212353576,
+      "learning_rate": 7.737461178419676e-06,
+      "loss": 0.893,
+      "step": 971
+    },
+    {
+      "epoch": 0.3369150779896014,
+      "grad_norm": 3.1597487712871906,
+      "learning_rate": 7.732761600869896e-06,
+      "loss": 0.76,
+      "step": 972
+    },
+    {
+      "epoch": 0.337261698440208,
+      "grad_norm": 3.730271055511726,
+      "learning_rate": 7.728058578186853e-06,
+      "loss": 0.9373,
+      "step": 973
+    },
+    {
+      "epoch": 0.33760831889081455,
+      "grad_norm": 3.592066901994553,
+      "learning_rate": 7.723352116299541e-06,
+      "loss": 0.7299,
+      "step": 974
+    },
+    {
+      "epoch": 0.3379549393414211,
+      "grad_norm": 4.408848151750255,
+      "learning_rate": 7.718642221141298e-06,
+      "loss": 0.7591,
+      "step": 975
+    },
+    {
+      "epoch": 0.33830155979202775,
+      "grad_norm": 8.106652337748837,
+      "learning_rate": 7.713928898649782e-06,
+      "loss": 0.8056,
+      "step": 976
+    },
+    {
+      "epoch": 0.3386481802426343,
+      "grad_norm": 3.8053698246904157,
+      "learning_rate": 7.709212154766984e-06,
+      "loss": 0.7995,
+      "step": 977
+    },
+    {
+      "epoch": 0.3389948006932409,
+      "grad_norm": 4.563112899221183,
+      "learning_rate": 7.704491995439192e-06,
+      "loss": 0.7827,
+      "step": 978
+    },
+    {
+      "epoch": 0.33934142114384747,
+      "grad_norm": 3.061520820926026,
+      "learning_rate": 7.699768426617015e-06,
+      "loss": 0.7711,
+      "step": 979
+    },
+    {
+      "epoch": 0.3396880415944541,
+      "grad_norm": 5.0860193699723215,
+      "learning_rate": 7.695041454255352e-06,
+      "loss": 0.8197,
+      "step": 980
+    },
+    {
+      "epoch": 0.34003466204506066,
+      "grad_norm": 3.725259539281139,
+      "learning_rate": 7.690311084313394e-06,
+      "loss": 0.6818,
+      "step": 981
+    },
+    {
+      "epoch": 0.34038128249566724,
+      "grad_norm": 5.089541094576722,
+      "learning_rate": 7.685577322754615e-06,
+      "loss": 0.7291,
+      "step": 982
+    },
+    {
+      "epoch": 0.3407279029462738,
+      "grad_norm": 3.664555366691614,
+      "learning_rate": 7.680840175546764e-06,
+      "loss": 0.6747,
+      "step": 983
+    },
+    {
+      "epoch": 0.34107452339688044,
+      "grad_norm": 3.2146795344758514,
+      "learning_rate": 7.676099648661866e-06,
+      "loss": 0.8358,
+      "step": 984
+    },
+    {
+      "epoch": 0.341421143847487,
+      "grad_norm": 2.281619813681752,
+      "learning_rate": 7.671355748076195e-06,
+      "loss": 0.7776,
+      "step": 985
+    },
+    {
+      "epoch": 0.3417677642980936,
+      "grad_norm": 3.6486072609962257,
+      "learning_rate": 7.666608479770285e-06,
+      "loss": 0.8119,
+      "step": 986
+    },
+    {
+      "epoch": 0.34211438474870015,
+      "grad_norm": 2.6941972371435874,
+      "learning_rate": 7.661857849728914e-06,
+      "loss": 0.6774,
+      "step": 987
+    },
+    {
+      "epoch": 0.3424610051993068,
+      "grad_norm": 10.206262339800437,
+      "learning_rate": 7.657103863941099e-06,
+      "loss": 0.7437,
+      "step": 988
+    },
+    {
+      "epoch": 0.34280762564991335,
+      "grad_norm": 4.881496281733532,
+      "learning_rate": 7.652346528400085e-06,
+      "loss": 0.8515,
+      "step": 989
+    },
+    {
+      "epoch": 0.3431542461005199,
+      "grad_norm": 6.013674578787092,
+      "learning_rate": 7.647585849103343e-06,
+      "loss": 0.7547,
+      "step": 990
+    },
+    {
+      "epoch": 0.3435008665511265,
+      "grad_norm": 13.3106585020306,
+      "learning_rate": 7.642821832052558e-06,
+      "loss": 0.7368,
+      "step": 991
+    },
+    {
+      "epoch": 0.3438474870017331,
+      "grad_norm": 4.090070751870045,
+      "learning_rate": 7.638054483253624e-06,
+      "loss": 0.8306,
+      "step": 992
+    },
+    {
+      "epoch": 0.3441941074523397,
+      "grad_norm": 40.7841711385337,
+      "learning_rate": 7.633283808716631e-06,
+      "loss": 0.7461,
+      "step": 993
+    },
+    {
+      "epoch": 0.34454072790294626,
+      "grad_norm": 5.021805601507357,
+      "learning_rate": 7.628509814455869e-06,
+      "loss": 0.6273,
+      "step": 994
+    },
+    {
+      "epoch": 0.34488734835355284,
+      "grad_norm": 5.477815101331064,
+      "learning_rate": 7.623732506489806e-06,
+      "loss": 0.786,
+      "step": 995
+    },
+    {
+      "epoch": 0.34523396880415946,
+      "grad_norm": 3.2669607101944704,
+      "learning_rate": 7.618951890841091e-06,
+      "loss": 0.8131,
+      "step": 996
+    },
+    {
+      "epoch": 0.34558058925476604,
+      "grad_norm": 2.8267494370545836,
+      "learning_rate": 7.614167973536543e-06,
+      "loss": 0.72,
+      "step": 997
+    },
+    {
+      "epoch": 0.3459272097053726,
+      "grad_norm": 5.074436280310562,
+      "learning_rate": 7.609380760607143e-06,
+      "loss": 0.7537,
+      "step": 998
+    },
+    {
+      "epoch": 0.3462738301559792,
+      "grad_norm": 7.360662361226735,
+      "learning_rate": 7.604590258088027e-06,
+      "loss": 0.8058,
+      "step": 999
+    },
+    {
+      "epoch": 0.3466204506065858,
+      "grad_norm": 2.6410323645754827,
+      "learning_rate": 7.599796472018477e-06,
+      "loss": 0.8098,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3469670710571924,
+      "grad_norm": 19.32845867783963,
+      "learning_rate": 7.5949994084419146e-06,
+      "loss": 0.7125,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34731369150779895,
+      "grad_norm": 5.822182352874723,
+      "learning_rate": 7.590199073405893e-06,
+      "loss": 0.7495,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3476603119584055,
+      "grad_norm": 4.529329487759585,
+      "learning_rate": 7.585395472962091e-06,
+      "loss": 0.7867,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34800693240901215,
+      "grad_norm": 4.582461674687221,
+      "learning_rate": 7.580588613166306e-06,
+      "loss": 0.788,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3483535528596187,
+      "grad_norm": 6.02811904849661,
+      "learning_rate": 7.575778500078439e-06,
+      "loss": 0.7997,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3487001733102253,
+      "grad_norm": 8.849143761067074,
+      "learning_rate": 7.570965139762496e-06,
+      "loss": 0.7049,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34904679376083186,
+      "grad_norm": 7.230170857220133,
+      "learning_rate": 7.566148538286575e-06,
+      "loss": 0.7644,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3493934142114385,
+      "grad_norm": 2.676986426403692,
+      "learning_rate": 7.561328701722863e-06,
+      "loss": 0.8794,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34974003466204506,
+      "grad_norm": 3.3076560757085534,
+      "learning_rate": 7.556505636147623e-06,
+      "loss": 0.8669,
+      "step": 1009
+    },
+    {
+      "epoch": 0.35008665511265163,
+      "grad_norm": 8.679309512263528,
+      "learning_rate": 7.551679347641188e-06,
+      "loss": 0.7728,
+      "step": 1010
+    },
+    {
+      "epoch": 0.3504332755632582,
+      "grad_norm": 4.763839905005605,
+      "learning_rate": 7.546849842287955e-06,
+      "loss": 0.8479,
+      "step": 1011
+    },
+    {
+      "epoch": 0.35077989601386483,
+      "grad_norm": 5.173002110144368,
+      "learning_rate": 7.542017126176377e-06,
+      "loss": 0.8308,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3511265164644714,
+      "grad_norm": 4.394664496464305,
+      "learning_rate": 7.537181205398957e-06,
+      "loss": 0.7905,
+      "step": 1013
+    },
+    {
+      "epoch": 0.351473136915078,
+      "grad_norm": 10.803052215664584,
+      "learning_rate": 7.532342086052231e-06,
+      "loss": 0.7522,
+      "step": 1014
+    },
+    {
+      "epoch": 0.35181975736568455,
+      "grad_norm": 3.8779840392320746,
+      "learning_rate": 7.527499774236774e-06,
+      "loss": 0.7892,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3521663778162912,
+      "grad_norm": 7.270242510908785,
+      "learning_rate": 7.522654276057184e-06,
+      "loss": 0.7783,
+      "step": 1016
+    },
+    {
+      "epoch": 0.35251299826689775,
+      "grad_norm": 3.859524178278526,
+      "learning_rate": 7.517805597622075e-06,
+      "loss": 0.8451,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3528596187175043,
+      "grad_norm": 2.735788592856702,
+      "learning_rate": 7.51295374504407e-06,
+      "loss": 0.8732,
+      "step": 1018
+    },
+    {
+      "epoch": 0.3532062391681109,
+      "grad_norm": 4.9503465478051645,
+      "learning_rate": 7.508098724439797e-06,
+      "loss": 0.7311,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3535528596187175,
+      "grad_norm": 19.186204610603546,
+      "learning_rate": 7.503240541929869e-06,
+      "loss": 0.8194,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3538994800693241,
+      "grad_norm": 2.945457835089021,
+      "learning_rate": 7.498379203638897e-06,
+      "loss": 0.7885,
+      "step": 1021
+    },
+    {
+      "epoch": 0.35424610051993066,
+      "grad_norm": 3.6753213873875468,
+      "learning_rate": 7.493514715695464e-06,
+      "loss": 0.8443,
+      "step": 1022
+    },
+    {
+      "epoch": 0.35459272097053723,
+      "grad_norm": 3.389365459287772,
+      "learning_rate": 7.488647084232123e-06,
+      "loss": 0.7708,
+      "step": 1023
+    },
+    {
+      "epoch": 0.35493934142114386,
+      "grad_norm": 17.12637059711508,
+      "learning_rate": 7.483776315385391e-06,
+      "loss": 0.7863,
+      "step": 1024
+    },
+    {
+      "epoch": 0.35528596187175043,
+      "grad_norm": 2.8570389556091786,
+      "learning_rate": 7.478902415295741e-06,
+      "loss": 0.7659,
+      "step": 1025
+    },
+    {
+      "epoch": 0.355632582322357,
+      "grad_norm": 4.170205949603099,
+      "learning_rate": 7.474025390107595e-06,
+      "loss": 0.7251,
+      "step": 1026
+    },
+    {
+      "epoch": 0.35597920277296363,
+      "grad_norm": 5.90936735043685,
+      "learning_rate": 7.4691452459693105e-06,
+      "loss": 0.8217,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3563258232235702,
+      "grad_norm": 6.753942646889834,
+      "learning_rate": 7.464261989033182e-06,
+      "loss": 0.7474,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3566724436741768,
+      "grad_norm": 3.731197016530116,
+      "learning_rate": 7.459375625455421e-06,
+      "loss": 0.7869,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35701906412478335,
+      "grad_norm": 5.45828525924612,
+      "learning_rate": 7.454486161396164e-06,
+      "loss": 0.6696,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35736568457539,
+      "grad_norm": 18.416365504021044,
+      "learning_rate": 7.449593603019451e-06,
+      "loss": 0.8589,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35771230502599655,
+      "grad_norm": 7.738932086863267,
+      "learning_rate": 7.444697956493225e-06,
+      "loss": 0.9086,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3580589254766031,
+      "grad_norm": 3.4220482978927533,
+      "learning_rate": 7.439799227989319e-06,
+      "loss": 0.7306,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3584055459272097,
+      "grad_norm": 2.50945919974729,
+      "learning_rate": 7.434897423683456e-06,
+      "loss": 0.7417,
+      "step": 1034
+    },
+    {
+      "epoch": 0.3587521663778163,
+      "grad_norm": 5.19697743963603,
+      "learning_rate": 7.429992549755234e-06,
+      "loss": 0.8007,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3590987868284229,
+      "grad_norm": 11.035246353863023,
+      "learning_rate": 7.425084612388122e-06,
+      "loss": 0.6742,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35944540727902946,
+      "grad_norm": 2.6283299237616427,
+      "learning_rate": 7.420173617769448e-06,
+      "loss": 0.8003,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35979202772963603,
+      "grad_norm": 4.948250559440026,
+      "learning_rate": 7.415259572090397e-06,
+      "loss": 0.8366,
+      "step": 1038
+    },
+    {
+      "epoch": 0.36013864818024266,
+      "grad_norm": 9.196834520237585,
+      "learning_rate": 7.410342481546002e-06,
+      "loss": 0.8265,
+      "step": 1039
+    },
+    {
+      "epoch": 0.36048526863084923,
+      "grad_norm": 3.552964349566498,
+      "learning_rate": 7.4054223523351295e-06,
+      "loss": 0.8689,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3608318890814558,
+      "grad_norm": 3.5419593290165343,
+      "learning_rate": 7.400499190660482e-06,
+      "loss": 0.714,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3611785095320624,
+      "grad_norm": 7.230234100025652,
+      "learning_rate": 7.39557300272858e-06,
+      "loss": 0.7552,
+      "step": 1042
+    },
+    {
+      "epoch": 0.361525129982669,
+      "grad_norm": 4.28107543424058,
+      "learning_rate": 7.390643794749763e-06,
+      "loss": 0.857,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3618717504332756,
+      "grad_norm": 4.43212720652161,
+      "learning_rate": 7.385711572938178e-06,
+      "loss": 0.7113,
+      "step": 1044
+    },
+    {
+      "epoch": 0.36221837088388215,
+      "grad_norm": 3.0182882408306204,
+      "learning_rate": 7.380776343511769e-06,
+      "loss": 0.7772,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3625649913344887,
+      "grad_norm": 4.442144119629415,
+      "learning_rate": 7.375838112692271e-06,
+      "loss": 0.8008,
+      "step": 1046
+    },
+    {
+      "epoch": 0.36291161178509534,
+      "grad_norm": 6.469874752189617,
+      "learning_rate": 7.3708968867052065e-06,
+      "loss": 0.8312,
+      "step": 1047
+    },
+    {
+      "epoch": 0.3632582322357019,
+      "grad_norm": 5.136907875369403,
+      "learning_rate": 7.36595267177987e-06,
+      "loss": 0.7899,
+      "step": 1048
+    },
+    {
+      "epoch": 0.3636048526863085,
+      "grad_norm": 5.9071196506413335,
+      "learning_rate": 7.36100547414933e-06,
+      "loss": 0.7635,
+      "step": 1049
+    },
+    {
+      "epoch": 0.36395147313691506,
+      "grad_norm": 4.363472731778042,
+      "learning_rate": 7.356055300050407e-06,
+      "loss": 0.7585,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3642980935875217,
+      "grad_norm": 4.1789784325878045,
+      "learning_rate": 7.35110215572368e-06,
+      "loss": 0.7276,
+      "step": 1051
+    },
+    {
+      "epoch": 0.36464471403812826,
+      "grad_norm": 3.5112159842537314,
+      "learning_rate": 7.3461460474134695e-06,
+      "loss": 0.7507,
+      "step": 1052
+    },
+    {
+      "epoch": 0.36499133448873483,
+      "grad_norm": 5.273860763209748,
+      "learning_rate": 7.341186981367835e-06,
+      "loss": 0.7091,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3653379549393414,
+      "grad_norm": 14.303654176051209,
+      "learning_rate": 7.336224963838563e-06,
+      "loss": 0.7241,
+      "step": 1054
+    },
+    {
+      "epoch": 0.36568457538994803,
+      "grad_norm": 4.495422931103704,
+      "learning_rate": 7.3312600010811615e-06,
+      "loss": 0.7698,
+      "step": 1055
+    },
+    {
+      "epoch": 0.3660311958405546,
+      "grad_norm": 3.902039688134791,
+      "learning_rate": 7.32629209935485e-06,
+      "loss": 0.7014,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3663778162911612,
+      "grad_norm": 4.149369957970063,
+      "learning_rate": 7.321321264922558e-06,
+      "loss": 0.7785,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36672443674176775,
+      "grad_norm": 4.274426103952399,
+      "learning_rate": 7.316347504050903e-06,
+      "loss": 0.7825,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3670710571923744,
+      "grad_norm": 3.215914706014133,
+      "learning_rate": 7.311370823010204e-06,
+      "loss": 0.6743,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36741767764298094,
+      "grad_norm": 3.047474674373949,
+      "learning_rate": 7.30639122807445e-06,
+      "loss": 0.693,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3677642980935875,
+      "grad_norm": 3.334211616395141,
+      "learning_rate": 7.301408725521308e-06,
+      "loss": 0.7689,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3681109185441941,
+      "grad_norm": 3.9667706809557695,
+      "learning_rate": 7.296423321632116e-06,
+      "loss": 0.7881,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3684575389948007,
+      "grad_norm": 7.8750219074552374,
+      "learning_rate": 7.291435022691859e-06,
+      "loss": 0.8339,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3688041594454073,
+      "grad_norm": 5.617312665200226,
+      "learning_rate": 7.286443834989178e-06,
+      "loss": 0.8266,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36915077989601386,
+      "grad_norm": 3.4366363891334024,
+      "learning_rate": 7.281449764816359e-06,
+      "loss": 0.9836,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36949740034662043,
+      "grad_norm": 8.037036663818999,
+      "learning_rate": 7.276452818469315e-06,
+      "loss": 0.8637,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36984402079722706,
+      "grad_norm": 5.1069238020573176,
+      "learning_rate": 7.271453002247588e-06,
+      "loss": 0.6843,
+      "step": 1067
+    },
+    {
+      "epoch": 0.37019064124783363,
+      "grad_norm": 10.577351755305097,
+      "learning_rate": 7.266450322454338e-06,
+      "loss": 0.6919,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3705372616984402,
+      "grad_norm": 4.110647150872241,
+      "learning_rate": 7.261444785396334e-06,
+      "loss": 0.8007,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3708838821490468,
+      "grad_norm": 14.066346706476839,
+      "learning_rate": 7.256436397383951e-06,
+      "loss": 0.7154,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3712305025996534,
+      "grad_norm": 4.447416458473875,
+      "learning_rate": 7.251425164731153e-06,
+      "loss": 0.681,
+      "step": 1071
+    },
+    {
+      "epoch": 0.37157712305025997,
+      "grad_norm": 2.89754474750456,
+      "learning_rate": 7.246411093755491e-06,
+      "loss": 0.763,
+      "step": 1072
+    },
+    {
+      "epoch": 0.37192374350086654,
+      "grad_norm": 4.157982344620452,
+      "learning_rate": 7.241394190778097e-06,
+      "loss": 0.8055,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3722703639514731,
+      "grad_norm": 3.251113383333912,
+      "learning_rate": 7.236374462123672e-06,
+      "loss": 0.7002,
+      "step": 1074
+    },
+    {
+      "epoch": 0.37261698440207974,
+      "grad_norm": 3.1740413167888075,
+      "learning_rate": 7.231351914120479e-06,
+      "loss": 0.8918,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3729636048526863,
+      "grad_norm": 7.161967116777736,
+      "learning_rate": 7.226326553100334e-06,
+      "loss": 0.8219,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3733102253032929,
+      "grad_norm": 15.380321853393228,
+      "learning_rate": 7.221298385398604e-06,
+      "loss": 0.7399,
+      "step": 1077
+    },
+    {
+      "epoch": 0.37365684575389946,
+      "grad_norm": 2.857145571324421,
+      "learning_rate": 7.216267417354186e-06,
+      "loss": 0.6845,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3740034662045061,
+      "grad_norm": 3.2587370794340385,
+      "learning_rate": 7.2112336553095155e-06,
+      "loss": 0.7536,
+      "step": 1079
+    },
+    {
+      "epoch": 0.37435008665511266,
+      "grad_norm": 2.7766712008487078,
+      "learning_rate": 7.206197105610547e-06,
+      "loss": 0.703,
+      "step": 1080
+    },
+    {
+      "epoch": 0.37469670710571923,
+      "grad_norm": 3.545469311289142,
+      "learning_rate": 7.2011577746067484e-06,
+      "loss": 0.7663,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3750433275563258,
+      "grad_norm": 2.419124004424902,
+      "learning_rate": 7.196115668651095e-06,
+      "loss": 0.7298,
+      "step": 1082
+    },
+    {
+      "epoch": 0.37538994800693243,
+      "grad_norm": 5.1152805537375485,
+      "learning_rate": 7.191070794100061e-06,
+      "loss": 0.7921,
+      "step": 1083
+    },
+    {
+      "epoch": 0.375736568457539,
+      "grad_norm": 2.862960681673759,
+      "learning_rate": 7.186023157313611e-06,
+      "loss": 0.6611,
+      "step": 1084
+    },
+    {
+      "epoch": 0.37608318890814557,
+      "grad_norm": 4.997524006001159,
+      "learning_rate": 7.180972764655189e-06,
+      "loss": 0.7531,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37642980935875214,
+      "grad_norm": 3.2991425396438685,
+      "learning_rate": 7.1759196224917175e-06,
+      "loss": 0.8896,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37677642980935877,
+      "grad_norm": 2.8828347038098388,
+      "learning_rate": 7.170863737193582e-06,
+      "loss": 0.7903,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37712305025996534,
+      "grad_norm": 2.9162987960118643,
+      "learning_rate": 7.1658051151346275e-06,
+      "loss": 0.8317,
+      "step": 1088
+    },
+    {
+      "epoch": 0.3774696707105719,
+      "grad_norm": 2.7088837844246503,
+      "learning_rate": 7.1607437626921505e-06,
+      "loss": 0.7796,
+      "step": 1089
+    },
+    {
+      "epoch": 0.3778162911611785,
+      "grad_norm": 8.42096952935219,
+      "learning_rate": 7.1556796862468865e-06,
+      "loss": 0.6642,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3781629116117851,
+      "grad_norm": 7.6963045911371095,
+      "learning_rate": 7.150612892183009e-06,
+      "loss": 0.787,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3785095320623917,
+      "grad_norm": 3.177629141611351,
+      "learning_rate": 7.145543386888111e-06,
+      "loss": 0.7231,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37885615251299826,
+      "grad_norm": 3.6549126523271482,
+      "learning_rate": 7.140471176753213e-06,
+      "loss": 0.8019,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37920277296360483,
+      "grad_norm": 4.886730198767973,
+      "learning_rate": 7.135396268172737e-06,
+      "loss": 0.7361,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37954939341421146,
+      "grad_norm": 3.23940025760448,
+      "learning_rate": 7.130318667544511e-06,
+      "loss": 0.7632,
+      "step": 1095
+    },
+    {
+      "epoch": 0.379896013864818,
+      "grad_norm": 18.854678677975713,
+      "learning_rate": 7.1252383812697565e-06,
+      "loss": 0.9475,
+      "step": 1096
+    },
+    {
+      "epoch": 0.3802426343154246,
+      "grad_norm": 4.502905031614254,
+      "learning_rate": 7.120155415753081e-06,
+      "loss": 0.7493,
+      "step": 1097
+    },
+    {
+      "epoch": 0.38058925476603117,
+      "grad_norm": 6.944416076012668,
+      "learning_rate": 7.115069777402467e-06,
+      "loss": 0.6947,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3809358752166378,
+      "grad_norm": 2.6175984710592655,
+      "learning_rate": 7.109981472629268e-06,
+      "loss": 0.7415,
+      "step": 1099
+    },
+    {
+      "epoch": 0.38128249566724437,
+      "grad_norm": 3.467334587143381,
+      "learning_rate": 7.104890507848201e-06,
+      "loss": 0.6888,
+      "step": 1100
+    },
+    {
+      "epoch": 0.38162911611785094,
+      "grad_norm": 5.889512362919233,
+      "learning_rate": 7.099796889477335e-06,
+      "loss": 0.701,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3819757365684575,
+      "grad_norm": 2.9732214935298455,
+      "learning_rate": 7.0947006239380825e-06,
+      "loss": 0.7701,
+      "step": 1102
+    },
+    {
+      "epoch": 0.38232235701906414,
+      "grad_norm": 3.7388257036807366,
+      "learning_rate": 7.089601717655197e-06,
+      "loss": 0.7604,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3826689774696707,
+      "grad_norm": 3.9233708580489535,
+      "learning_rate": 7.084500177056758e-06,
+      "loss": 0.7263,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3830155979202773,
+      "grad_norm": 3.235607816488042,
+      "learning_rate": 7.079396008574167e-06,
+      "loss": 0.8268,
+      "step": 1105
+    },
+    {
+      "epoch": 0.38336221837088386,
+      "grad_norm": 3.9849998276127896,
+      "learning_rate": 7.074289218642137e-06,
+      "loss": 0.812,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3837088388214905,
+      "grad_norm": 5.671737393042273,
+      "learning_rate": 7.06917981369869e-06,
+      "loss": 0.7665,
+      "step": 1107
+    },
+    {
+      "epoch": 0.38405545927209706,
+      "grad_norm": 3.7662894120238337,
+      "learning_rate": 7.064067800185142e-06,
+      "loss": 0.8531,
+      "step": 1108
+    },
+    {
+      "epoch": 0.3844020797227036,
+      "grad_norm": 3.9090767649592792,
+      "learning_rate": 7.058953184546093e-06,
+      "loss": 0.7808,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3847487001733102,
+      "grad_norm": 4.773535637791278,
+      "learning_rate": 7.053835973229433e-06,
+      "loss": 0.8659,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3850953206239168,
+      "grad_norm": 10.06397812127479,
+      "learning_rate": 7.048716172686321e-06,
+      "loss": 0.7253,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3854419410745234,
+      "grad_norm": 4.834202499469988,
+      "learning_rate": 7.043593789371171e-06,
+      "loss": 0.8763,
+      "step": 1112
+    },
+    {
+      "epoch": 0.38578856152512997,
+      "grad_norm": 3.780598901943384,
+      "learning_rate": 7.038468829741666e-06,
+      "loss": 0.7054,
+      "step": 1113
+    },
+    {
+      "epoch": 0.38613518197573654,
+      "grad_norm": 3.582200654171567,
+      "learning_rate": 7.033341300258729e-06,
+      "loss": 0.7444,
+      "step": 1114
+    },
+    {
+      "epoch": 0.38648180242634317,
+      "grad_norm": 3.442475781033951,
+      "learning_rate": 7.028211207386527e-06,
+      "loss": 0.6882,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38682842287694974,
+      "grad_norm": 9.829529547377257,
+      "learning_rate": 7.023078557592455e-06,
+      "loss": 0.7284,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3871750433275563,
+      "grad_norm": 6.978644938085211,
+      "learning_rate": 7.017943357347136e-06,
+      "loss": 0.8903,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38752166377816294,
+      "grad_norm": 5.483460552542588,
+      "learning_rate": 7.012805613124404e-06,
+      "loss": 0.8039,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3878682842287695,
+      "grad_norm": 3.709341569465285,
+      "learning_rate": 7.007665331401299e-06,
+      "loss": 0.7756,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3882149046793761,
+      "grad_norm": 3.065007183906629,
+      "learning_rate": 7.002522518658066e-06,
+      "loss": 0.7149,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38856152512998265,
+      "grad_norm": 4.224710699185205,
+      "learning_rate": 6.997377181378138e-06,
+      "loss": 0.7288,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3889081455805893,
+      "grad_norm": 7.853878888126959,
+      "learning_rate": 6.992229326048128e-06,
+      "loss": 0.7238,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38925476603119585,
+      "grad_norm": 3.168779590070741,
+      "learning_rate": 6.987078959157825e-06,
+      "loss": 0.8097,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3896013864818024,
+      "grad_norm": 3.4524379525350857,
+      "learning_rate": 6.981926087200188e-06,
+      "loss": 0.7245,
+      "step": 1124
+    },
+    {
+      "epoch": 0.389948006932409,
+      "grad_norm": 3.7385443425591087,
+      "learning_rate": 6.976770716671327e-06,
+      "loss": 0.8496,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3902946273830156,
+      "grad_norm": 3.594343867068319,
+      "learning_rate": 6.9716128540705084e-06,
+      "loss": 0.8877,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3906412478336222,
+      "grad_norm": 5.507867840198834,
+      "learning_rate": 6.966452505900138e-06,
+      "loss": 0.7679,
+      "step": 1127
+    },
+    {
+      "epoch": 0.39098786828422877,
+      "grad_norm": 3.391792488392336,
+      "learning_rate": 6.9612896786657524e-06,
+      "loss": 0.7824,
+      "step": 1128
+    },
+    {
+      "epoch": 0.39133448873483534,
+      "grad_norm": 2.562723368592942,
+      "learning_rate": 6.956124378876018e-06,
+      "loss": 0.8679,
+      "step": 1129
+    },
+    {
+      "epoch": 0.39168110918544197,
+      "grad_norm": 2.9604001254260455,
+      "learning_rate": 6.950956613042714e-06,
+      "loss": 0.764,
+      "step": 1130
+    },
+    {
+      "epoch": 0.39202772963604854,
+      "grad_norm": 3.263196114771287,
+      "learning_rate": 6.945786387680735e-06,
+      "loss": 0.7099,
+      "step": 1131
+    },
+    {
+      "epoch": 0.3923743500866551,
+      "grad_norm": 6.199134174819675,
+      "learning_rate": 6.940613709308067e-06,
+      "loss": 0.8294,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3927209705372617,
+      "grad_norm": 3.7077320687587414,
+      "learning_rate": 6.935438584445795e-06,
+      "loss": 0.7217,
+      "step": 1133
+    },
+    {
+      "epoch": 0.3930675909878683,
+      "grad_norm": 4.685155423951412,
+      "learning_rate": 6.930261019618089e-06,
+      "loss": 0.7216,
+      "step": 1134
+    },
+    {
+      "epoch": 0.3934142114384749,
+      "grad_norm": 3.636743389268508,
+      "learning_rate": 6.925081021352189e-06,
+      "loss": 0.7157,
+      "step": 1135
+    },
+    {
+      "epoch": 0.39376083188908145,
+      "grad_norm": 6.8885780665021565,
+      "learning_rate": 6.91989859617841e-06,
+      "loss": 0.7363,
+      "step": 1136
+    },
+    {
+      "epoch": 0.394107452339688,
+      "grad_norm": 4.786917032547147,
+      "learning_rate": 6.91471375063012e-06,
+      "loss": 0.7758,
+      "step": 1137
+    },
+    {
+      "epoch": 0.39445407279029465,
+      "grad_norm": 3.328171857213517,
+      "learning_rate": 6.909526491243742e-06,
+      "loss": 0.752,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3948006932409012,
+      "grad_norm": 4.258114241982606,
+      "learning_rate": 6.904336824558742e-06,
+      "loss": 0.8123,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3951473136915078,
+      "grad_norm": 3.6364778017838684,
+      "learning_rate": 6.899144757117621e-06,
+      "loss": 0.8575,
+      "step": 1140
+    },
+    {
+      "epoch": 0.39549393414211437,
+      "grad_norm": 2.5718609212411616,
+      "learning_rate": 6.893950295465905e-06,
+      "loss": 0.8091,
+      "step": 1141
+    },
+    {
+      "epoch": 0.395840554592721,
+      "grad_norm": 3.027074522817448,
+      "learning_rate": 6.888753446152139e-06,
+      "loss": 0.7099,
+      "step": 1142
+    },
+    {
+      "epoch": 0.39618717504332757,
+      "grad_norm": 5.861906621056083,
+      "learning_rate": 6.883554215727876e-06,
+      "loss": 0.8481,
+      "step": 1143
+    },
+    {
+      "epoch": 0.39653379549393414,
+      "grad_norm": 8.059211006233603,
+      "learning_rate": 6.8783526107476784e-06,
+      "loss": 0.7541,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3968804159445407,
+      "grad_norm": 3.08997409640164,
+      "learning_rate": 6.873148637769091e-06,
+      "loss": 0.712,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39722703639514734,
+      "grad_norm": 6.529302766338545,
+      "learning_rate": 6.867942303352653e-06,
+      "loss": 0.8556,
+      "step": 1146
+    },
+    {
+      "epoch": 0.3975736568457539,
+      "grad_norm": 10.067450061089987,
+      "learning_rate": 6.862733614061876e-06,
+      "loss": 0.7868,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3979202772963605,
+      "grad_norm": 17.23739475239773,
+      "learning_rate": 6.857522576463243e-06,
+      "loss": 0.8026,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39826689774696705,
+      "grad_norm": 6.803690648840442,
+      "learning_rate": 6.852309197126194e-06,
+      "loss": 0.7271,
+      "step": 1149
+    },
+    {
+      "epoch": 0.3986135181975737,
+      "grad_norm": 2.915362818596441,
+      "learning_rate": 6.847093482623125e-06,
+      "loss": 0.7757,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39896013864818025,
+      "grad_norm": 7.091939718509309,
+      "learning_rate": 6.841875439529373e-06,
+      "loss": 0.8251,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3993067590987868,
+      "grad_norm": 2.9520953040588616,
+      "learning_rate": 6.8366550744232145e-06,
+      "loss": 0.7772,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3996533795493934,
+      "grad_norm": 7.178716919816926,
+      "learning_rate": 6.831432393885844e-06,
+      "loss": 0.8874,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 6.100542109522188,
+      "learning_rate": 6.8262074045013894e-06,
+      "loss": 0.7261,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4003466204506066,
+      "grad_norm": 81.34881243965263,
+      "learning_rate": 6.820980112856877e-06,
+      "loss": 0.708,
+      "step": 1155
+    },
+    {
+      "epoch": 0.40069324090121317,
+      "grad_norm": 4.437567925250368,
+      "learning_rate": 6.815750525542239e-06,
+      "loss": 0.77,
+      "step": 1156
+    },
+    {
+      "epoch": 0.40103986135181974,
+      "grad_norm": 2.6698220679456246,
+      "learning_rate": 6.8105186491503065e-06,
+      "loss": 0.9222,
+      "step": 1157
+    },
+    {
+      "epoch": 0.40138648180242636,
+      "grad_norm": 4.212154799763013,
+      "learning_rate": 6.805284490276789e-06,
+      "loss": 0.8628,
+      "step": 1158
+    },
+    {
+      "epoch": 0.40173310225303294,
+      "grad_norm": 3.5338025421098584,
+      "learning_rate": 6.800048055520279e-06,
+      "loss": 0.6834,
+      "step": 1159
+    },
+    {
+      "epoch": 0.4020797227036395,
+      "grad_norm": 2.3856343434097202,
+      "learning_rate": 6.794809351482236e-06,
+      "loss": 0.705,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4024263431542461,
+      "grad_norm": 4.844457526553181,
+      "learning_rate": 6.7895683847669806e-06,
+      "loss": 0.7625,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4027729636048527,
+      "grad_norm": 5.030944951497747,
+      "learning_rate": 6.7843251619816855e-06,
+      "loss": 0.7597,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4031195840554593,
+      "grad_norm": 3.1229208878654604,
+      "learning_rate": 6.779079689736367e-06,
+      "loss": 0.8207,
+      "step": 1163
+    },
+    {
+      "epoch": 0.40346620450606585,
+      "grad_norm": 4.190846157690851,
+      "learning_rate": 6.7738319746438784e-06,
+      "loss": 0.7428,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4038128249566724,
+      "grad_norm": 8.126942328788516,
+      "learning_rate": 6.768582023319903e-06,
+      "loss": 0.8241,
+      "step": 1165
+    },
+    {
+      "epoch": 0.40415944540727905,
+      "grad_norm": 8.65941185783664,
+      "learning_rate": 6.763329842382938e-06,
+      "loss": 0.7265,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4045060658578856,
+      "grad_norm": 10.15131313323171,
+      "learning_rate": 6.758075438454295e-06,
+      "loss": 0.7401,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4048526863084922,
+      "grad_norm": 3.646156147140761,
+      "learning_rate": 6.752818818158085e-06,
+      "loss": 0.7168,
+      "step": 1168
+    },
+    {
+      "epoch": 0.40519930675909877,
+      "grad_norm": 3.8020908518610836,
+      "learning_rate": 6.747559988121217e-06,
+      "loss": 0.8088,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4055459272097054,
+      "grad_norm": 3.2851903175927135,
+      "learning_rate": 6.742298954973385e-06,
+      "loss": 0.7209,
+      "step": 1170
+    },
+    {
+      "epoch": 0.40589254766031196,
+      "grad_norm": 7.216908431317702,
+      "learning_rate": 6.7370357253470564e-06,
+      "loss": 0.8564,
+      "step": 1171
+    },
+    {
+      "epoch": 0.40623916811091854,
+      "grad_norm": 5.4400893516668765,
+      "learning_rate": 6.731770305877474e-06,
+      "loss": 0.67,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4065857885615251,
+      "grad_norm": 3.029567755020331,
+      "learning_rate": 6.726502703202635e-06,
+      "loss": 0.7203,
+      "step": 1173
+    },
+    {
+      "epoch": 0.40693240901213173,
+      "grad_norm": 5.105848428564404,
+      "learning_rate": 6.721232923963293e-06,
+      "loss": 0.7539,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4072790294627383,
+      "grad_norm": 3.345107108451253,
+      "learning_rate": 6.7159609748029456e-06,
+      "loss": 0.7534,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4076256499133449,
+      "grad_norm": 3.157127870731139,
+      "learning_rate": 6.7106868623678215e-06,
+      "loss": 0.7933,
+      "step": 1176
+    },
+    {
+      "epoch": 0.40797227036395145,
+      "grad_norm": 5.431933740396925,
+      "learning_rate": 6.705410593306882e-06,
+      "loss": 0.8088,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4083188908145581,
+      "grad_norm": 2.8999843424856135,
+      "learning_rate": 6.7001321742718066e-06,
+      "loss": 0.8485,
+      "step": 1178
+    },
+    {
+      "epoch": 0.40866551126516465,
+      "grad_norm": 6.032743414792963,
+      "learning_rate": 6.694851611916979e-06,
+      "loss": 0.7093,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4090121317157712,
+      "grad_norm": 8.498533425575786,
+      "learning_rate": 6.689568912899494e-06,
+      "loss": 0.8701,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4093587521663778,
+      "grad_norm": 3.3839469813414267,
+      "learning_rate": 6.684284083879134e-06,
+      "loss": 0.7863,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4097053726169844,
+      "grad_norm": 3.6956454544715167,
+      "learning_rate": 6.6789971315183695e-06,
+      "loss": 0.801,
+      "step": 1182
+    },
+    {
+      "epoch": 0.410051993067591,
+      "grad_norm": 7.496571707014741,
+      "learning_rate": 6.6737080624823446e-06,
+      "loss": 0.7578,
+      "step": 1183
+    },
+    {
+      "epoch": 0.41039861351819756,
+      "grad_norm": 4.007952250776024,
+      "learning_rate": 6.668416883438875e-06,
+      "loss": 0.7714,
+      "step": 1184
+    },
+    {
+      "epoch": 0.41074523396880414,
+      "grad_norm": 3.2385946811080712,
+      "learning_rate": 6.663123601058435e-06,
+      "loss": 0.7313,
+      "step": 1185
+    },
+    {
+      "epoch": 0.41109185441941076,
+      "grad_norm": 4.741613044737179,
+      "learning_rate": 6.657828222014152e-06,
+      "loss": 0.8693,
+      "step": 1186
+    },
+    {
+      "epoch": 0.41143847487001733,
+      "grad_norm": 5.128140611917805,
+      "learning_rate": 6.652530752981792e-06,
+      "loss": 0.8421,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4117850953206239,
+      "grad_norm": 6.159798712833919,
+      "learning_rate": 6.647231200639766e-06,
+      "loss": 0.6805,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4121317157712305,
+      "grad_norm": 3.3906935069860387,
+      "learning_rate": 6.641929571669096e-06,
+      "loss": 0.8184,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4124783362218371,
+      "grad_norm": 4.7407963730939064,
+      "learning_rate": 6.636625872753438e-06,
+      "loss": 0.7603,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4128249566724437,
+      "grad_norm": 4.090832313067579,
+      "learning_rate": 6.631320110579046e-06,
+      "loss": 0.8763,
+      "step": 1191
+    },
+    {
+      "epoch": 0.41317157712305025,
+      "grad_norm": 3.0671033872254165,
+      "learning_rate": 6.626012291834781e-06,
+      "loss": 0.7674,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4135181975736568,
+      "grad_norm": 4.443544780740881,
+      "learning_rate": 6.620702423212095e-06,
+      "loss": 0.6878,
+      "step": 1193
+    },
+    {
+      "epoch": 0.41386481802426345,
+      "grad_norm": 6.794405503517979,
+      "learning_rate": 6.615390511405022e-06,
+      "loss": 0.655,
+      "step": 1194
+    },
+    {
+      "epoch": 0.41421143847487,
+      "grad_norm": 7.9995897006729235,
+      "learning_rate": 6.610076563110177e-06,
+      "loss": 0.8138,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4145580589254766,
+      "grad_norm": 4.932543796040504,
+      "learning_rate": 6.6047605850267395e-06,
+      "loss": 0.7455,
+      "step": 1196
+    },
+    {
+      "epoch": 0.41490467937608316,
+      "grad_norm": 4.229103334696119,
+      "learning_rate": 6.599442583856443e-06,
+      "loss": 0.6983,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4152512998266898,
+      "grad_norm": 2.665694008020218,
+      "learning_rate": 6.5941225663035826e-06,
+      "loss": 0.6618,
+      "step": 1198
+    },
+    {
+      "epoch": 0.41559792027729636,
+      "grad_norm": 4.071436163436493,
+      "learning_rate": 6.588800539074984e-06,
+      "loss": 0.7107,
+      "step": 1199
+    },
+    {
+      "epoch": 0.41594454072790293,
+      "grad_norm": 6.954567574788078,
+      "learning_rate": 6.5834765088800135e-06,
+      "loss": 0.8521,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4162911611785095,
+      "grad_norm": 10.126537448435405,
+      "learning_rate": 6.57815048243056e-06,
+      "loss": 0.7223,
+      "step": 1201
+    },
+    {
+      "epoch": 0.41663778162911613,
+      "grad_norm": 3.21488566200229,
+      "learning_rate": 6.57282246644103e-06,
+      "loss": 0.746,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4169844020797227,
+      "grad_norm": 3.017279829670175,
+      "learning_rate": 6.567492467628338e-06,
+      "loss": 0.7029,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4173310225303293,
+      "grad_norm": 4.382950049258224,
+      "learning_rate": 6.562160492711896e-06,
+      "loss": 0.7339,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41767764298093585,
+      "grad_norm": 4.8602361792839375,
+      "learning_rate": 6.556826548413612e-06,
+      "loss": 0.699,
+      "step": 1205
+    },
+    {
+      "epoch": 0.4180242634315425,
+      "grad_norm": 6.469465988045293,
+      "learning_rate": 6.551490641457873e-06,
+      "loss": 0.7192,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41837088388214905,
+      "grad_norm": 3.1860435667809393,
+      "learning_rate": 6.546152778571538e-06,
+      "loss": 0.6882,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4187175043327556,
+      "grad_norm": 4.572668065665973,
+      "learning_rate": 6.5408129664839395e-06,
+      "loss": 0.7123,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4190641247833622,
+      "grad_norm": 4.011766979410526,
+      "learning_rate": 6.535471211926859e-06,
+      "loss": 0.8527,
+      "step": 1209
+    },
+    {
+      "epoch": 0.4194107452339688,
+      "grad_norm": 4.014787993142043,
+      "learning_rate": 6.5301275216345324e-06,
+      "loss": 0.7329,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4197573656845754,
+      "grad_norm": 7.003387219724084,
+      "learning_rate": 6.524781902343636e-06,
+      "loss": 0.8326,
+      "step": 1211
+    },
+    {
+      "epoch": 0.42010398613518196,
+      "grad_norm": 2.8622234725860665,
+      "learning_rate": 6.519434360793272e-06,
+      "loss": 0.7371,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4204506065857886,
+      "grad_norm": 3.595344481708059,
+      "learning_rate": 6.514084903724972e-06,
+      "loss": 0.8262,
+      "step": 1213
+    },
+    {
+      "epoch": 0.42079722703639516,
+      "grad_norm": 10.004236131939258,
+      "learning_rate": 6.508733537882681e-06,
+      "loss": 0.7967,
+      "step": 1214
+    },
+    {
+      "epoch": 0.42114384748700173,
+      "grad_norm": 6.782892928639603,
+      "learning_rate": 6.5033802700127505e-06,
+      "loss": 0.7344,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4214904679376083,
+      "grad_norm": 7.383216847531722,
+      "learning_rate": 6.4980251068639275e-06,
+      "loss": 0.6548,
+      "step": 1216
+    },
+    {
+      "epoch": 0.42183708838821493,
+      "grad_norm": 5.064959080669278,
+      "learning_rate": 6.49266805518735e-06,
+      "loss": 0.7504,
+      "step": 1217
+    },
+    {
+      "epoch": 0.4221837088388215,
+      "grad_norm": 8.417629780055375,
+      "learning_rate": 6.487309121736536e-06,
+      "loss": 0.6625,
+      "step": 1218
+    },
+    {
+      "epoch": 0.4225303292894281,
+      "grad_norm": 6.721657597609613,
+      "learning_rate": 6.481948313267381e-06,
+      "loss": 0.6929,
+      "step": 1219
+    },
+    {
+      "epoch": 0.42287694974003465,
+      "grad_norm": 4.397646828262417,
+      "learning_rate": 6.476585636538135e-06,
+      "loss": 0.8296,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4232235701906413,
+      "grad_norm": 2.370271804896577,
+      "learning_rate": 6.47122109830941e-06,
+      "loss": 0.8313,
+      "step": 1221
+    },
+    {
+      "epoch": 0.42357019064124785,
+      "grad_norm": 3.8507088661977744,
+      "learning_rate": 6.465854705344164e-06,
+      "loss": 0.7833,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4239168110918544,
+      "grad_norm": 3.7517787966225398,
+      "learning_rate": 6.460486464407692e-06,
+      "loss": 0.7483,
+      "step": 1223
+    },
+    {
+      "epoch": 0.424263431542461,
+      "grad_norm": 4.765177076687678,
+      "learning_rate": 6.455116382267616e-06,
+      "loss": 0.7521,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4246100519930676,
+      "grad_norm": 5.564907213510305,
+      "learning_rate": 6.449744465693886e-06,
+      "loss": 0.7412,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4249566724436742,
+      "grad_norm": 2.9446291090359002,
+      "learning_rate": 6.4443707214587595e-06,
+      "loss": 0.7476,
+      "step": 1226
+    },
+    {
+      "epoch": 0.42530329289428076,
+      "grad_norm": 4.368564501122817,
+      "learning_rate": 6.438995156336797e-06,
+      "loss": 0.7935,
+      "step": 1227
+    },
+    {
+      "epoch": 0.42564991334488733,
+      "grad_norm": 3.350812482556908,
+      "learning_rate": 6.433617777104858e-06,
+      "loss": 0.9283,
+      "step": 1228
+    },
+    {
+      "epoch": 0.42599653379549396,
+      "grad_norm": 4.455946827551343,
+      "learning_rate": 6.428238590542091e-06,
+      "loss": 0.7041,
+      "step": 1229
+    },
+    {
+      "epoch": 0.42634315424610053,
+      "grad_norm": 4.021637962842656,
+      "learning_rate": 6.422857603429915e-06,
+      "loss": 0.7314,
+      "step": 1230
+    },
+    {
+      "epoch": 0.4266897746967071,
+      "grad_norm": 4.545121641168341,
+      "learning_rate": 6.417474822552025e-06,
+      "loss": 0.8213,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4270363951473137,
+      "grad_norm": 3.450031468048601,
+      "learning_rate": 6.412090254694381e-06,
+      "loss": 0.8412,
+      "step": 1232
+    },
+    {
+      "epoch": 0.4273830155979203,
+      "grad_norm": 3.317723913641202,
+      "learning_rate": 6.4067039066451846e-06,
+      "loss": 0.8559,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4277296360485269,
+      "grad_norm": 6.473278483493087,
+      "learning_rate": 6.4013157851948896e-06,
+      "loss": 0.6523,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42807625649913345,
+      "grad_norm": 3.028589047830632,
+      "learning_rate": 6.3959258971361845e-06,
+      "loss": 0.7445,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42842287694974,
+      "grad_norm": 3.5496901723069603,
+      "learning_rate": 6.390534249263985e-06,
+      "loss": 0.7594,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42876949740034664,
+      "grad_norm": 3.7288694497415023,
+      "learning_rate": 6.385140848375424e-06,
+      "loss": 0.7117,
+      "step": 1237
+    },
+    {
+      "epoch": 0.4291161178509532,
+      "grad_norm": 3.5607543607389704,
+      "learning_rate": 6.379745701269844e-06,
+      "loss": 0.7171,
+      "step": 1238
+    },
+    {
+      "epoch": 0.4294627383015598,
+      "grad_norm": 4.99075392003206,
+      "learning_rate": 6.374348814748792e-06,
+      "loss": 0.6943,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42980935875216636,
+      "grad_norm": 4.337572640792187,
+      "learning_rate": 6.368950195616002e-06,
+      "loss": 0.7353,
+      "step": 1240
+    },
+    {
+      "epoch": 0.430155979202773,
+      "grad_norm": 3.6019790175955007,
+      "learning_rate": 6.3635498506774005e-06,
+      "loss": 0.8074,
+      "step": 1241
+    },
+    {
+      "epoch": 0.43050259965337956,
+      "grad_norm": 9.492438069654835,
+      "learning_rate": 6.358147786741082e-06,
+      "loss": 0.673,
+      "step": 1242
+    },
+    {
+      "epoch": 0.43084922010398613,
+      "grad_norm": 3.9773357314887114,
+      "learning_rate": 6.352744010617315e-06,
+      "loss": 0.727,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4311958405545927,
+      "grad_norm": 2.3925480280875964,
+      "learning_rate": 6.347338529118517e-06,
+      "loss": 0.7093,
+      "step": 1244
+    },
+    {
+      "epoch": 0.43154246100519933,
+      "grad_norm": 2.901964543847465,
+      "learning_rate": 6.341931349059266e-06,
+      "loss": 0.7005,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4318890814558059,
+      "grad_norm": 4.578376260837734,
+      "learning_rate": 6.336522477256275e-06,
+      "loss": 0.6859,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4322357019064125,
+      "grad_norm": 3.4084156222213067,
+      "learning_rate": 6.331111920528393e-06,
+      "loss": 0.7052,
+      "step": 1247
+    },
+    {
+      "epoch": 0.43258232235701904,
+      "grad_norm": 4.959941540197702,
+      "learning_rate": 6.325699685696588e-06,
+      "loss": 0.6103,
+      "step": 1248
+    },
+    {
+      "epoch": 0.43292894280762567,
+      "grad_norm": 6.34816717189547,
+      "learning_rate": 6.32028577958395e-06,
+      "loss": 0.7821,
+      "step": 1249
+    },
+    {
+      "epoch": 0.43327556325823224,
+      "grad_norm": 2.7290246564282064,
+      "learning_rate": 6.314870209015674e-06,
+      "loss": 0.6687,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4336221837088388,
+      "grad_norm": 2.75082312068807,
+      "learning_rate": 6.3094529808190466e-06,
+      "loss": 0.7227,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4339688041594454,
+      "grad_norm": 2.8080539576060404,
+      "learning_rate": 6.304034101823455e-06,
+      "loss": 0.708,
+      "step": 1252
+    },
+    {
+      "epoch": 0.434315424610052,
+      "grad_norm": 9.00582479880191,
+      "learning_rate": 6.298613578860362e-06,
+      "loss": 0.6876,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4346620450606586,
+      "grad_norm": 2.943954210778076,
+      "learning_rate": 6.293191418763301e-06,
+      "loss": 0.8127,
+      "step": 1254
+    },
+    {
+      "epoch": 0.43500866551126516,
+      "grad_norm": 2.9682129380065985,
+      "learning_rate": 6.287767628367873e-06,
+      "loss": 0.6852,
+      "step": 1255
+    },
+    {
+      "epoch": 0.43535528596187173,
+      "grad_norm": 3.6439601312638197,
+      "learning_rate": 6.282342214511734e-06,
+      "loss": 0.7026,
+      "step": 1256
+    },
+    {
+      "epoch": 0.43570190641247836,
+      "grad_norm": 8.259308421903109,
+      "learning_rate": 6.276915184034584e-06,
+      "loss": 0.731,
+      "step": 1257
+    },
+    {
+      "epoch": 0.43604852686308493,
+      "grad_norm": 3.8654747305622204,
+      "learning_rate": 6.271486543778162e-06,
+      "loss": 0.7244,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4363951473136915,
+      "grad_norm": 3.6523596729675645,
+      "learning_rate": 6.2660563005862395e-06,
+      "loss": 0.7578,
+      "step": 1259
+    },
+    {
+      "epoch": 0.43674176776429807,
+      "grad_norm": 2.9946709165339622,
+      "learning_rate": 6.260624461304605e-06,
+      "loss": 0.8379,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4370883882149047,
+      "grad_norm": 4.2700522848949864,
+      "learning_rate": 6.255191032781059e-06,
+      "loss": 0.791,
+      "step": 1261
+    },
+    {
+      "epoch": 0.43743500866551127,
+      "grad_norm": 14.69320484973424,
+      "learning_rate": 6.249756021865409e-06,
+      "loss": 0.8394,
+      "step": 1262
+    },
+    {
+      "epoch": 0.43778162911611784,
+      "grad_norm": 3.0957079311092133,
+      "learning_rate": 6.244319435409453e-06,
+      "loss": 0.9254,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4381282495667244,
+      "grad_norm": 4.603079312794203,
+      "learning_rate": 6.238881280266979e-06,
+      "loss": 0.7208,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43847487001733104,
+      "grad_norm": 2.599573673618702,
+      "learning_rate": 6.233441563293754e-06,
+      "loss": 0.7893,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4388214904679376,
+      "grad_norm": 12.696908116501413,
+      "learning_rate": 6.228000291347506e-06,
+      "loss": 0.7637,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4391681109185442,
+      "grad_norm": 3.4462894919316343,
+      "learning_rate": 6.222557471287932e-06,
+      "loss": 0.7627,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43951473136915076,
+      "grad_norm": 3.449543144037233,
+      "learning_rate": 6.217113109976676e-06,
+      "loss": 0.7752,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4398613518197574,
+      "grad_norm": 5.991191194455595,
+      "learning_rate": 6.211667214277326e-06,
+      "loss": 0.7893,
+      "step": 1269
+    },
+    {
+      "epoch": 0.44020797227036396,
+      "grad_norm": 4.240038590788041,
+      "learning_rate": 6.206219791055406e-06,
+      "loss": 0.8222,
+      "step": 1270
+    },
+    {
+      "epoch": 0.44055459272097053,
+      "grad_norm": 7.9313904964372615,
+      "learning_rate": 6.200770847178364e-06,
+      "loss": 0.8448,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4409012131715771,
+      "grad_norm": 15.319050519161792,
+      "learning_rate": 6.195320389515561e-06,
+      "loss": 0.8161,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4412478336221837,
+      "grad_norm": 6.8256834369047406,
+      "learning_rate": 6.1898684249382775e-06,
+      "loss": 0.6973,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4415944540727903,
+      "grad_norm": 3.65560976039272,
+      "learning_rate": 6.1844149603196835e-06,
+      "loss": 0.7745,
+      "step": 1274
+    },
+    {
+      "epoch": 0.44194107452339687,
+      "grad_norm": 3.9672471120659805,
+      "learning_rate": 6.178960002534843e-06,
+      "loss": 0.7322,
+      "step": 1275
+    },
+    {
+      "epoch": 0.44228769497400344,
+      "grad_norm": 3.0274354086339272,
+      "learning_rate": 6.173503558460703e-06,
+      "loss": 0.7814,
+      "step": 1276
+    },
+    {
+      "epoch": 0.44263431542461007,
+      "grad_norm": 5.110640549948285,
+      "learning_rate": 6.168045634976086e-06,
+      "loss": 0.7732,
+      "step": 1277
+    },
+    {
+      "epoch": 0.44298093587521664,
+      "grad_norm": 3.747970477183729,
+      "learning_rate": 6.162586238961677e-06,
+      "loss": 0.8026,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4433275563258232,
+      "grad_norm": 3.8759071277628308,
+      "learning_rate": 6.1571253773000165e-06,
+      "loss": 0.7741,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4436741767764298,
+      "grad_norm": 22.49088114295837,
+      "learning_rate": 6.151663056875496e-06,
+      "loss": 0.8135,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4440207972270364,
+      "grad_norm": 3.7079705262905325,
+      "learning_rate": 6.146199284574346e-06,
+      "loss": 0.8984,
+      "step": 1281
+    },
+    {
+      "epoch": 0.444367417677643,
+      "grad_norm": 3.6609304701403,
+      "learning_rate": 6.1407340672846195e-06,
+      "loss": 0.7561,
+      "step": 1282
+    },
+    {
+      "epoch": 0.44471403812824956,
+      "grad_norm": 5.718203319918465,
+      "learning_rate": 6.135267411896205e-06,
+      "loss": 0.6549,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4450606585788561,
+      "grad_norm": 3.117072201919077,
+      "learning_rate": 6.129799325300792e-06,
+      "loss": 0.5708,
+      "step": 1284
+    },
+    {
+      "epoch": 0.44540727902946275,
+      "grad_norm": 4.154569968388288,
+      "learning_rate": 6.124329814391878e-06,
+      "loss": 0.7492,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4457538994800693,
+      "grad_norm": 9.331673599893278,
+      "learning_rate": 6.11885888606476e-06,
+      "loss": 0.7752,
+      "step": 1286
+    },
+    {
+      "epoch": 0.4461005199306759,
+      "grad_norm": 2.9153620688919553,
+      "learning_rate": 6.113386547216516e-06,
+      "loss": 0.6952,
+      "step": 1287
+    },
+    {
+      "epoch": 0.44644714038128247,
+      "grad_norm": 3.5210828025142824,
+      "learning_rate": 6.107912804746008e-06,
+      "loss": 0.7652,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4467937608318891,
+      "grad_norm": 3.1733750990239455,
+      "learning_rate": 6.1024376655538595e-06,
+      "loss": 0.7866,
+      "step": 1289
+    },
+    {
+      "epoch": 0.44714038128249567,
+      "grad_norm": 2.5873852325638143,
+      "learning_rate": 6.096961136542463e-06,
+      "loss": 0.7394,
+      "step": 1290
+    },
+    {
+      "epoch": 0.44748700173310224,
+      "grad_norm": 17.79019159835679,
+      "learning_rate": 6.091483224615961e-06,
+      "loss": 0.775,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4478336221837088,
+      "grad_norm": 3.807140611012404,
+      "learning_rate": 6.086003936680234e-06,
+      "loss": 0.7764,
+      "step": 1292
+    },
+    {
+      "epoch": 0.44818024263431544,
+      "grad_norm": 5.504471386748529,
+      "learning_rate": 6.0805232796429055e-06,
+      "loss": 0.7854,
+      "step": 1293
+    },
+    {
+      "epoch": 0.448526863084922,
+      "grad_norm": 18.511684050889677,
+      "learning_rate": 6.0750412604133195e-06,
+      "loss": 0.6684,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4488734835355286,
+      "grad_norm": 3.1059983940653866,
+      "learning_rate": 6.069557885902537e-06,
+      "loss": 0.7548,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44922010398613516,
+      "grad_norm": 6.1837969698033035,
+      "learning_rate": 6.064073163023331e-06,
+      "loss": 0.7371,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4495667244367418,
+      "grad_norm": 6.678994088658321,
+      "learning_rate": 6.058587098690173e-06,
+      "loss": 0.7159,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44991334488734835,
+      "grad_norm": 3.5929284110714517,
+      "learning_rate": 6.053099699819226e-06,
+      "loss": 0.8331,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4502599653379549,
+      "grad_norm": 3.1768016397688092,
+      "learning_rate": 6.0476109733283305e-06,
+      "loss": 0.7298,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4506065857885615,
+      "grad_norm": 3.594845532737384,
+      "learning_rate": 6.042120926137009e-06,
+      "loss": 0.7732,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4509532062391681,
+      "grad_norm": 18.995549286902325,
+      "learning_rate": 6.036629565166445e-06,
+      "loss": 0.6681,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4512998266897747,
+      "grad_norm": 8.387599502869483,
+      "learning_rate": 6.031136897339476e-06,
+      "loss": 0.7514,
+      "step": 1302
+    },
+    {
+      "epoch": 0.45164644714038127,
+      "grad_norm": 3.8419299753637617,
+      "learning_rate": 6.02564292958059e-06,
+      "loss": 0.8182,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4519930675909879,
+      "grad_norm": 8.157449563915657,
+      "learning_rate": 6.0201476688159145e-06,
+      "loss": 0.763,
+      "step": 1304
+    },
+    {
+      "epoch": 0.45233968804159447,
+      "grad_norm": 4.763723691951883,
+      "learning_rate": 6.014651121973203e-06,
+      "loss": 0.8271,
+      "step": 1305
+    },
+    {
+      "epoch": 0.45268630849220104,
+      "grad_norm": 3.066013324983718,
+      "learning_rate": 6.009153295981834e-06,
+      "loss": 0.7305,
+      "step": 1306
+    },
+    {
+      "epoch": 0.4530329289428076,
+      "grad_norm": 2.412267567824713,
+      "learning_rate": 6.003654197772796e-06,
+      "loss": 0.7649,
+      "step": 1307
+    },
+    {
+      "epoch": 0.45337954939341424,
+      "grad_norm": 8.28921102882463,
+      "learning_rate": 5.998153834278685e-06,
+      "loss": 0.7198,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4537261698440208,
+      "grad_norm": 3.0646851982223144,
+      "learning_rate": 5.9926522124336894e-06,
+      "loss": 0.7556,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4540727902946274,
+      "grad_norm": 4.048698486743088,
+      "learning_rate": 5.987149339173584e-06,
+      "loss": 0.6988,
+      "step": 1310
+    },
+    {
+      "epoch": 0.45441941074523395,
+      "grad_norm": 8.290158154592186,
+      "learning_rate": 5.981645221435723e-06,
+      "loss": 0.7746,
+      "step": 1311
+    },
+    {
+      "epoch": 0.4547660311958406,
+      "grad_norm": 6.3046785478959295,
+      "learning_rate": 5.976139866159028e-06,
+      "loss": 0.8941,
+      "step": 1312
+    },
+    {
+      "epoch": 0.45511265164644715,
+      "grad_norm": 5.601069192014951,
+      "learning_rate": 5.970633280283978e-06,
+      "loss": 0.6622,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4554592720970537,
+      "grad_norm": 2.943141711580514,
+      "learning_rate": 5.965125470752609e-06,
+      "loss": 0.7426,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4558058925476603,
+      "grad_norm": 2.7020381256923907,
+      "learning_rate": 5.959616444508497e-06,
+      "loss": 0.6885,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4561525129982669,
+      "grad_norm": 6.359578669148671,
+      "learning_rate": 5.954106208496751e-06,
+      "loss": 0.7378,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4564991334488735,
+      "grad_norm": 3.1477461694147344,
+      "learning_rate": 5.948594769664007e-06,
+      "loss": 0.8248,
+      "step": 1317
+    },
+    {
+      "epoch": 0.45684575389948007,
+      "grad_norm": 2.6778905744511468,
+      "learning_rate": 5.943082134958414e-06,
+      "loss": 0.7729,
+      "step": 1318
+    },
+    {
+      "epoch": 0.45719237435008664,
+      "grad_norm": 4.427929932838776,
+      "learning_rate": 5.937568311329634e-06,
+      "loss": 0.8385,
+      "step": 1319
+    },
+    {
+      "epoch": 0.45753899480069327,
+      "grad_norm": 4.486144360240464,
+      "learning_rate": 5.932053305728824e-06,
+      "loss": 0.7983,
+      "step": 1320
+    },
+    {
+      "epoch": 0.45788561525129984,
+      "grad_norm": 2.860989427932612,
+      "learning_rate": 5.926537125108631e-06,
+      "loss": 0.7293,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4582322357019064,
+      "grad_norm": 4.27966892087902,
+      "learning_rate": 5.921019776423186e-06,
+      "loss": 0.7756,
+      "step": 1322
+    },
+    {
+      "epoch": 0.458578856152513,
+      "grad_norm": 3.1467396502077998,
+      "learning_rate": 5.915501266628087e-06,
+      "loss": 0.7732,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4589254766031196,
+      "grad_norm": 4.448825466077771,
+      "learning_rate": 5.909981602680405e-06,
+      "loss": 0.651,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4592720970537262,
+      "grad_norm": 4.63672698571625,
+      "learning_rate": 5.904460791538655e-06,
+      "loss": 0.917,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45961871750433275,
+      "grad_norm": 5.120385731654488,
+      "learning_rate": 5.898938840162806e-06,
+      "loss": 0.7093,
+      "step": 1326
+    },
+    {
+      "epoch": 0.4599653379549393,
+      "grad_norm": 4.522712230445281,
+      "learning_rate": 5.893415755514267e-06,
+      "loss": 0.8956,
+      "step": 1327
+    },
+    {
+      "epoch": 0.46031195840554595,
+      "grad_norm": 6.249988627883992,
+      "learning_rate": 5.8878915445558635e-06,
+      "loss": 0.7449,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4606585788561525,
+      "grad_norm": 3.2657824577120462,
+      "learning_rate": 5.882366214251854e-06,
+      "loss": 0.7875,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4610051993067591,
+      "grad_norm": 4.0176006149051995,
+      "learning_rate": 5.876839771567898e-06,
+      "loss": 0.808,
+      "step": 1330
+    },
+    {
+      "epoch": 0.46135181975736567,
+      "grad_norm": 4.4210659148257925,
+      "learning_rate": 5.8713122234710655e-06,
+      "loss": 0.6992,
+      "step": 1331
+    },
+    {
+      "epoch": 0.4616984402079723,
+      "grad_norm": 3.2009887730460047,
+      "learning_rate": 5.8657835769298155e-06,
+      "loss": 0.7594,
+      "step": 1332
+    },
+    {
+      "epoch": 0.46204506065857887,
+      "grad_norm": 5.127484042828931,
+      "learning_rate": 5.860253838913991e-06,
+      "loss": 0.821,
+      "step": 1333
+    },
+    {
+      "epoch": 0.46239168110918544,
+      "grad_norm": 8.477100596025329,
+      "learning_rate": 5.854723016394815e-06,
+      "loss": 0.8051,
+      "step": 1334
+    },
+    {
+      "epoch": 0.462738301559792,
+      "grad_norm": 4.3975326404680235,
+      "learning_rate": 5.849191116344875e-06,
+      "loss": 0.7065,
+      "step": 1335
+    },
+    {
+      "epoch": 0.46308492201039864,
+      "grad_norm": 3.4112461899465396,
+      "learning_rate": 5.843658145738117e-06,
+      "loss": 0.922,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4634315424610052,
+      "grad_norm": 2.9860272902117373,
+      "learning_rate": 5.838124111549837e-06,
+      "loss": 0.8336,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4637781629116118,
+      "grad_norm": 4.7513933818039975,
+      "learning_rate": 5.83258902075667e-06,
+      "loss": 0.8495,
+      "step": 1338
+    },
+    {
+      "epoch": 0.46412478336221835,
+      "grad_norm": 2.708268864932647,
+      "learning_rate": 5.827052880336589e-06,
+      "loss": 0.7973,
+      "step": 1339
+    },
+    {
+      "epoch": 0.464471403812825,
+      "grad_norm": 2.974383982136568,
+      "learning_rate": 5.821515697268883e-06,
+      "loss": 0.8581,
+      "step": 1340
+    },
+    {
+      "epoch": 0.46481802426343155,
+      "grad_norm": 5.968861720667599,
+      "learning_rate": 5.815977478534159e-06,
+      "loss": 0.7803,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4651646447140381,
+      "grad_norm": 3.821089152657383,
+      "learning_rate": 5.810438231114332e-06,
+      "loss": 0.8733,
+      "step": 1342
+    },
+    {
+      "epoch": 0.4655112651646447,
+      "grad_norm": 9.241769584321643,
+      "learning_rate": 5.804897961992606e-06,
+      "loss": 0.8214,
+      "step": 1343
+    },
+    {
+      "epoch": 0.4658578856152513,
+      "grad_norm": 3.6364067813021967,
+      "learning_rate": 5.799356678153481e-06,
+      "loss": 0.7713,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4662045060658579,
+      "grad_norm": 3.6330979725528323,
+      "learning_rate": 5.793814386582736e-06,
+      "loss": 0.7729,
+      "step": 1345
+    },
+    {
+      "epoch": 0.46655112651646446,
+      "grad_norm": 2.8388844652923146,
+      "learning_rate": 5.788271094267412e-06,
+      "loss": 0.6774,
+      "step": 1346
+    },
+    {
+      "epoch": 0.46689774696707104,
+      "grad_norm": 3.8731347142279517,
+      "learning_rate": 5.782726808195822e-06,
+      "loss": 0.7637,
+      "step": 1347
+    },
+    {
+      "epoch": 0.46724436741767766,
+      "grad_norm": 2.8590897594737816,
+      "learning_rate": 5.777181535357524e-06,
+      "loss": 0.7363,
+      "step": 1348
+    },
+    {
+      "epoch": 0.46759098786828424,
+      "grad_norm": 4.9006240527030664,
+      "learning_rate": 5.7716352827433254e-06,
+      "loss": 0.7288,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4679376083188908,
+      "grad_norm": 2.991799976396695,
+      "learning_rate": 5.766088057345266e-06,
+      "loss": 0.7346,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4682842287694974,
+      "grad_norm": 4.764077579823415,
+      "learning_rate": 5.760539866156611e-06,
+      "loss": 0.7546,
+      "step": 1351
+    },
+    {
+      "epoch": 0.468630849220104,
+      "grad_norm": 24.984222537795613,
+      "learning_rate": 5.754990716171849e-06,
+      "loss": 0.7573,
+      "step": 1352
+    },
+    {
+      "epoch": 0.4689774696707106,
+      "grad_norm": 5.328567531878857,
+      "learning_rate": 5.749440614386665e-06,
+      "loss": 0.7333,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46932409012131715,
+      "grad_norm": 5.683273762027506,
+      "learning_rate": 5.74388956779796e-06,
+      "loss": 0.6478,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4696707105719237,
+      "grad_norm": 3.194925810297735,
+      "learning_rate": 5.738337583403816e-06,
+      "loss": 0.6925,
+      "step": 1355
+    },
+    {
+      "epoch": 0.47001733102253035,
+      "grad_norm": 3.9761697766731654,
+      "learning_rate": 5.732784668203496e-06,
+      "loss": 0.7615,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4703639514731369,
+      "grad_norm": 7.403732428101122,
+      "learning_rate": 5.7272308291974424e-06,
+      "loss": 0.7823,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4707105719237435,
+      "grad_norm": 2.9317367986491165,
+      "learning_rate": 5.721676073387259e-06,
+      "loss": 0.6999,
+      "step": 1358
+    },
+    {
+      "epoch": 0.47105719237435006,
+      "grad_norm": 2.906073919903494,
+      "learning_rate": 5.716120407775707e-06,
+      "loss": 0.7454,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4714038128249567,
+      "grad_norm": 5.845981184310017,
+      "learning_rate": 5.7105638393666905e-06,
+      "loss": 0.7903,
+      "step": 1360
+    },
+    {
+      "epoch": 0.47175043327556326,
+      "grad_norm": 3.661608973115144,
+      "learning_rate": 5.705006375165258e-06,
+      "loss": 0.8034,
+      "step": 1361
+    },
+    {
+      "epoch": 0.47209705372616984,
+      "grad_norm": 2.6645018421062177,
+      "learning_rate": 5.699448022177583e-06,
+      "loss": 0.777,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4724436741767764,
+      "grad_norm": 2.4502875262062163,
+      "learning_rate": 5.693888787410959e-06,
+      "loss": 0.7599,
+      "step": 1363
+    },
+    {
+      "epoch": 0.47279029462738303,
+      "grad_norm": 6.570627353988905,
+      "learning_rate": 5.688328677873796e-06,
+      "loss": 0.699,
+      "step": 1364
+    },
+    {
+      "epoch": 0.4731369150779896,
+      "grad_norm": 8.146686923382003,
+      "learning_rate": 5.682767700575601e-06,
+      "loss": 0.81,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4734835355285962,
+      "grad_norm": 2.7646995858734438,
+      "learning_rate": 5.677205862526976e-06,
+      "loss": 0.7517,
+      "step": 1366
+    },
+    {
+      "epoch": 0.47383015597920275,
+      "grad_norm": 4.285832809746341,
+      "learning_rate": 5.671643170739613e-06,
+      "loss": 0.7362,
+      "step": 1367
+    },
+    {
+      "epoch": 0.4741767764298094,
+      "grad_norm": 32.78870228963693,
+      "learning_rate": 5.666079632226276e-06,
+      "loss": 0.7277,
+      "step": 1368
+    },
+    {
+      "epoch": 0.47452339688041595,
+      "grad_norm": 2.9947587596369374,
+      "learning_rate": 5.6605152540007944e-06,
+      "loss": 0.7862,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4748700173310225,
+      "grad_norm": 6.805592835650406,
+      "learning_rate": 5.654950043078063e-06,
+      "loss": 0.7914,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4752166377816291,
+      "grad_norm": 3.113930142130086,
+      "learning_rate": 5.64938400647402e-06,
+      "loss": 0.7736,
+      "step": 1371
+    },
+    {
+      "epoch": 0.4755632582322357,
+      "grad_norm": 10.361559634589991,
+      "learning_rate": 5.643817151205649e-06,
+      "loss": 0.7434,
+      "step": 1372
+    },
+    {
+      "epoch": 0.4759098786828423,
+      "grad_norm": 3.1872306909840398,
+      "learning_rate": 5.638249484290962e-06,
+      "loss": 0.717,
+      "step": 1373
+    },
+    {
+      "epoch": 0.47625649913344886,
+      "grad_norm": 4.968339054690989,
+      "learning_rate": 5.632681012748995e-06,
+      "loss": 0.8254,
+      "step": 1374
+    },
+    {
+      "epoch": 0.47660311958405543,
+      "grad_norm": 3.2267301139965023,
+      "learning_rate": 5.627111743599802e-06,
+      "loss": 0.7684,
+      "step": 1375
+    },
+    {
+      "epoch": 0.47694974003466206,
+      "grad_norm": 3.4452713172719682,
+      "learning_rate": 5.621541683864439e-06,
+      "loss": 0.8271,
+      "step": 1376
+    },
+    {
+      "epoch": 0.47729636048526863,
+      "grad_norm": 4.97926964183782,
+      "learning_rate": 5.615970840564956e-06,
+      "loss": 0.6684,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4776429809358752,
+      "grad_norm": 3.6077676932098655,
+      "learning_rate": 5.610399220724401e-06,
+      "loss": 0.7379,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4779896013864818,
+      "grad_norm": 2.797148693948706,
+      "learning_rate": 5.604826831366789e-06,
+      "loss": 0.7998,
+      "step": 1379
+    },
+    {
+      "epoch": 0.4783362218370884,
+      "grad_norm": 8.368518235215582,
+      "learning_rate": 5.59925367951711e-06,
+      "loss": 0.6821,
+      "step": 1380
+    },
+    {
+      "epoch": 0.478682842287695,
+      "grad_norm": 3.132721888922046,
+      "learning_rate": 5.5936797722013205e-06,
+      "loss": 0.8666,
+      "step": 1381
+    },
+    {
+      "epoch": 0.47902946273830155,
+      "grad_norm": 2.5753721317960103,
+      "learning_rate": 5.58810511644632e-06,
+      "loss": 0.8656,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4793760831889081,
+      "grad_norm": 4.90023764615192,
+      "learning_rate": 5.582529719279957e-06,
+      "loss": 0.7778,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47972270363951475,
+      "grad_norm": 23.86909242980265,
+      "learning_rate": 5.576953587731015e-06,
+      "loss": 0.7977,
+      "step": 1384
+    },
+    {
+      "epoch": 0.4800693240901213,
+      "grad_norm": 3.6558621700922473,
+      "learning_rate": 5.571376728829201e-06,
+      "loss": 0.6803,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4804159445407279,
+      "grad_norm": 3.5070279059305896,
+      "learning_rate": 5.565799149605142e-06,
+      "loss": 0.8196,
+      "step": 1386
+    },
+    {
+      "epoch": 0.48076256499133446,
+      "grad_norm": 2.4151810116680967,
+      "learning_rate": 5.560220857090367e-06,
+      "loss": 0.7088,
+      "step": 1387
+    },
+    {
+      "epoch": 0.4811091854419411,
+      "grad_norm": 11.449961532325025,
+      "learning_rate": 5.554641858317313e-06,
+      "loss": 0.7475,
+      "step": 1388
+    },
+    {
+      "epoch": 0.48145580589254766,
+      "grad_norm": 3.783441423812622,
+      "learning_rate": 5.549062160319301e-06,
+      "loss": 0.7651,
+      "step": 1389
+    },
+    {
+      "epoch": 0.48180242634315423,
+      "grad_norm": 3.665181443322686,
+      "learning_rate": 5.543481770130533e-06,
+      "loss": 0.7929,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4821490467937608,
+      "grad_norm": 15.024185274459805,
+      "learning_rate": 5.537900694786089e-06,
+      "loss": 0.7508,
+      "step": 1391
+    },
+    {
+      "epoch": 0.48249566724436743,
+      "grad_norm": 5.440007536365913,
+      "learning_rate": 5.5323189413219085e-06,
+      "loss": 0.7301,
+      "step": 1392
+    },
+    {
+      "epoch": 0.482842287694974,
+      "grad_norm": 12.975231629057202,
+      "learning_rate": 5.526736516774785e-06,
+      "loss": 0.757,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4831889081455806,
+      "grad_norm": 5.060815832691072,
+      "learning_rate": 5.521153428182364e-06,
+      "loss": 0.7349,
+      "step": 1394
+    },
+    {
+      "epoch": 0.48353552859618715,
+      "grad_norm": 5.323546187884733,
+      "learning_rate": 5.5155696825831205e-06,
+      "loss": 0.7281,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4838821490467938,
+      "grad_norm": 10.41276223647833,
+      "learning_rate": 5.509985287016365e-06,
+      "loss": 0.7902,
+      "step": 1396
+    },
+    {
+      "epoch": 0.48422876949740035,
+      "grad_norm": 4.2362884912879615,
+      "learning_rate": 5.504400248522221e-06,
+      "loss": 0.6743,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4845753899480069,
+      "grad_norm": 4.829409965138924,
+      "learning_rate": 5.498814574141624e-06,
+      "loss": 0.8019,
+      "step": 1398
+    },
+    {
+      "epoch": 0.48492201039861355,
+      "grad_norm": 3.511580954477383,
+      "learning_rate": 5.493228270916317e-06,
+      "loss": 0.6674,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4852686308492201,
+      "grad_norm": 3.868512532879684,
+      "learning_rate": 5.487641345888829e-06,
+      "loss": 0.8545,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4856152512998267,
+      "grad_norm": 2.5689660054585057,
+      "learning_rate": 5.482053806102474e-06,
+      "loss": 0.8017,
+      "step": 1401
+    },
+    {
+      "epoch": 0.48596187175043326,
+      "grad_norm": 7.224969571439596,
+      "learning_rate": 5.476465658601344e-06,
+      "loss": 0.7715,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4863084922010399,
+      "grad_norm": 3.632682635823041,
+      "learning_rate": 5.470876910430294e-06,
+      "loss": 0.6732,
+      "step": 1403
+    },
+    {
+      "epoch": 0.48665511265164646,
+      "grad_norm": 7.28448154781623,
+      "learning_rate": 5.465287568634939e-06,
+      "loss": 0.7928,
+      "step": 1404
+    },
+    {
+      "epoch": 0.48700173310225303,
+      "grad_norm": 6.347554773300054,
+      "learning_rate": 5.4596976402616395e-06,
+      "loss": 0.819,
+      "step": 1405
+    },
+    {
+      "epoch": 0.4873483535528596,
+      "grad_norm": 4.0906398205962295,
+      "learning_rate": 5.4541071323575005e-06,
+      "loss": 0.7013,
+      "step": 1406
+    },
+    {
+      "epoch": 0.48769497400346623,
+      "grad_norm": 3.3354851759804838,
+      "learning_rate": 5.4485160519703486e-06,
+      "loss": 0.8167,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4880415944540728,
+      "grad_norm": 3.6128612185816293,
+      "learning_rate": 5.442924406148739e-06,
+      "loss": 0.7427,
+      "step": 1408
+    },
+    {
+      "epoch": 0.4883882149046794,
+      "grad_norm": 4.055225617791424,
+      "learning_rate": 5.437332201941942e-06,
+      "loss": 0.7612,
+      "step": 1409
+    },
+    {
+      "epoch": 0.48873483535528595,
+      "grad_norm": 4.59255725656241,
+      "learning_rate": 5.4317394463999225e-06,
+      "loss": 0.8267,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4890814558058926,
+      "grad_norm": 4.4723898036258625,
+      "learning_rate": 5.42614614657335e-06,
+      "loss": 0.7045,
+      "step": 1411
+    },
+    {
+      "epoch": 0.48942807625649914,
+      "grad_norm": 4.1493799526853055,
+      "learning_rate": 5.420552309513578e-06,
+      "loss": 0.6546,
+      "step": 1412
+    },
+    {
+      "epoch": 0.4897746967071057,
+      "grad_norm": 5.277105142779789,
+      "learning_rate": 5.41495794227263e-06,
+      "loss": 0.7639,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4901213171577123,
+      "grad_norm": 4.89929725702432,
+      "learning_rate": 5.409363051903208e-06,
+      "loss": 0.782,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4904679376083189,
+      "grad_norm": 3.4612124021540693,
+      "learning_rate": 5.403767645458666e-06,
+      "loss": 0.7837,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4908145580589255,
+      "grad_norm": 2.424786673966588,
+      "learning_rate": 5.398171729993014e-06,
+      "loss": 0.7951,
+      "step": 1416
+    },
+    {
+      "epoch": 0.49116117850953206,
+      "grad_norm": 4.287462111852363,
+      "learning_rate": 5.3925753125609e-06,
+      "loss": 0.7403,
+      "step": 1417
+    },
+    {
+      "epoch": 0.49150779896013863,
+      "grad_norm": 3.573014501255699,
+      "learning_rate": 5.386978400217604e-06,
+      "loss": 0.767,
+      "step": 1418
+    },
+    {
+      "epoch": 0.49185441941074526,
+      "grad_norm": 5.5907117712589836,
+      "learning_rate": 5.381381000019037e-06,
+      "loss": 0.779,
+      "step": 1419
+    },
+    {
+      "epoch": 0.49220103986135183,
+      "grad_norm": 11.571161263066982,
+      "learning_rate": 5.375783119021715e-06,
+      "loss": 0.6903,
+      "step": 1420
+    },
+    {
+      "epoch": 0.4925476603119584,
+      "grad_norm": 11.777553109595853,
+      "learning_rate": 5.370184764282767e-06,
+      "loss": 0.7643,
+      "step": 1421
+    },
+    {
+      "epoch": 0.492894280762565,
+      "grad_norm": 5.9452542983403225,
+      "learning_rate": 5.3645859428599175e-06,
+      "loss": 0.6658,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4932409012131716,
+      "grad_norm": 2.543100449297375,
+      "learning_rate": 5.358986661811479e-06,
+      "loss": 0.7132,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4935875216637782,
+      "grad_norm": 9.55932974698817,
+      "learning_rate": 5.353386928196344e-06,
+      "loss": 0.6688,
+      "step": 1424
+    },
+    {
+      "epoch": 0.49393414211438474,
+      "grad_norm": 2.7107583770230557,
+      "learning_rate": 5.3477867490739764e-06,
+      "loss": 0.8557,
+      "step": 1425
+    },
+    {
+      "epoch": 0.4942807625649913,
+      "grad_norm": 4.041103483947281,
+      "learning_rate": 5.342186131504398e-06,
+      "loss": 0.8205,
+      "step": 1426
+    },
+    {
+      "epoch": 0.49462738301559794,
+      "grad_norm": 3.7847061901381718,
+      "learning_rate": 5.336585082548188e-06,
+      "loss": 0.7312,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4949740034662045,
+      "grad_norm": 3.066340515415406,
+      "learning_rate": 5.330983609266464e-06,
+      "loss": 0.7654,
+      "step": 1428
+    },
+    {
+      "epoch": 0.4953206239168111,
+      "grad_norm": 12.233802638521116,
+      "learning_rate": 5.325381718720886e-06,
+      "loss": 0.7611,
+      "step": 1429
+    },
+    {
+      "epoch": 0.49566724436741766,
+      "grad_norm": 4.588057084791653,
+      "learning_rate": 5.319779417973635e-06,
+      "loss": 0.7099,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4960138648180243,
+      "grad_norm": 3.9624163893800133,
+      "learning_rate": 5.314176714087408e-06,
+      "loss": 0.7059,
+      "step": 1431
+    },
+    {
+      "epoch": 0.49636048526863086,
+      "grad_norm": 3.4419701413200245,
+      "learning_rate": 5.308573614125413e-06,
+      "loss": 0.6716,
+      "step": 1432
+    },
+    {
+      "epoch": 0.49670710571923743,
+      "grad_norm": 5.4039873216539105,
+      "learning_rate": 5.302970125151356e-06,
+      "loss": 0.7193,
+      "step": 1433
+    },
+    {
+      "epoch": 0.497053726169844,
+      "grad_norm": 8.79057913328972,
+      "learning_rate": 5.297366254229436e-06,
+      "loss": 0.824,
+      "step": 1434
+    },
+    {
+      "epoch": 0.49740034662045063,
+      "grad_norm": 7.129091325916357,
+      "learning_rate": 5.291762008424331e-06,
+      "loss": 0.7309,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4977469670710572,
+      "grad_norm": 76.5656376278366,
+      "learning_rate": 5.28615739480119e-06,
+      "loss": 0.8026,
+      "step": 1436
+    },
+    {
+      "epoch": 0.49809358752166377,
+      "grad_norm": 5.421795959459407,
+      "learning_rate": 5.280552420425631e-06,
+      "loss": 0.6613,
+      "step": 1437
+    },
+    {
+      "epoch": 0.49844020797227034,
+      "grad_norm": 3.05853837121045,
+      "learning_rate": 5.2749470923637204e-06,
+      "loss": 0.8199,
+      "step": 1438
+    },
+    {
+      "epoch": 0.49878682842287697,
+      "grad_norm": 3.026955928652836,
+      "learning_rate": 5.2693414176819745e-06,
+      "loss": 0.7078,
+      "step": 1439
+    },
+    {
+      "epoch": 0.49913344887348354,
+      "grad_norm": 2.9540256706249926,
+      "learning_rate": 5.2637354034473485e-06,
+      "loss": 0.7748,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4994800693240901,
+      "grad_norm": 5.8385133109016945,
+      "learning_rate": 5.25812905672722e-06,
+      "loss": 0.801,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4998266897746967,
+      "grad_norm": 4.402738438036581,
+      "learning_rate": 5.25252238458939e-06,
+      "loss": 0.7799,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5001733102253033,
+      "grad_norm": 4.3159780648877435,
+      "learning_rate": 5.246915394102068e-06,
+      "loss": 0.7566,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5005199306759098,
+      "grad_norm": 2.974296876307307,
+      "learning_rate": 5.241308092333864e-06,
+      "loss": 0.7075,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5008665511265165,
+      "grad_norm": 3.9109177725633955,
+      "learning_rate": 5.2357004863537845e-06,
+      "loss": 0.7371,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5012131715771231,
+      "grad_norm": 4.9363291728073655,
+      "learning_rate": 5.230092583231213e-06,
+      "loss": 0.7367,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5015597920277296,
+      "grad_norm": 3.7491732595326344,
+      "learning_rate": 5.224484390035915e-06,
+      "loss": 0.6432,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5019064124783362,
+      "grad_norm": 6.203703334101342,
+      "learning_rate": 5.218875913838017e-06,
+      "loss": 0.778,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5022530329289429,
+      "grad_norm": 2.7290941981214814,
+      "learning_rate": 5.213267161708002e-06,
+      "loss": 0.7957,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5025996533795494,
+      "grad_norm": 3.50640796763192,
+      "learning_rate": 5.207658140716704e-06,
+      "loss": 0.7979,
+      "step": 1450
+    },
+    {
+      "epoch": 0.502946273830156,
+      "grad_norm": 9.45771343056091,
+      "learning_rate": 5.202048857935291e-06,
+      "loss": 0.7936,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5032928942807625,
+      "grad_norm": 3.724964899849946,
+      "learning_rate": 5.196439320435266e-06,
+      "loss": 0.7084,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5036395147313691,
+      "grad_norm": 9.71046077668804,
+      "learning_rate": 5.190829535288453e-06,
+      "loss": 0.6699,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5039861351819758,
+      "grad_norm": 3.399093699926791,
+      "learning_rate": 5.185219509566985e-06,
+      "loss": 0.7769,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5043327556325823,
+      "grad_norm": 3.6867557022517063,
+      "learning_rate": 5.179609250343298e-06,
+      "loss": 0.8166,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5046793760831889,
+      "grad_norm": 4.216189027034735,
+      "learning_rate": 5.173998764690126e-06,
+      "loss": 0.759,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5050259965337955,
+      "grad_norm": 4.220272070627803,
+      "learning_rate": 5.1683880596804854e-06,
+      "loss": 0.7958,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5053726169844021,
+      "grad_norm": 3.3948311158416393,
+      "learning_rate": 5.162777142387672e-06,
+      "loss": 0.7015,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5057192374350087,
+      "grad_norm": 3.012988148149591,
+      "learning_rate": 5.157166019885246e-06,
+      "loss": 0.7203,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5060658578856152,
+      "grad_norm": 3.0736255486944732,
+      "learning_rate": 5.151554699247029e-06,
+      "loss": 0.6771,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5064124783362218,
+      "grad_norm": 3.9665712953727232,
+      "learning_rate": 5.14594318754709e-06,
+      "loss": 0.7576,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5067590987868285,
+      "grad_norm": 5.141459288895079,
+      "learning_rate": 5.140331491859739e-06,
+      "loss": 0.7058,
+      "step": 1462
+    },
+    {
+      "epoch": 0.507105719237435,
+      "grad_norm": 3.449950408417885,
+      "learning_rate": 5.1347196192595235e-06,
+      "loss": 0.7585,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5074523396880416,
+      "grad_norm": 2.7477688428525133,
+      "learning_rate": 5.129107576821205e-06,
+      "loss": 0.6145,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5077989601386482,
+      "grad_norm": 4.757699665984696,
+      "learning_rate": 5.123495371619765e-06,
+      "loss": 0.7536,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5081455805892547,
+      "grad_norm": 20.045200994571445,
+      "learning_rate": 5.11788301073039e-06,
+      "loss": 0.753,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5084922010398614,
+      "grad_norm": 4.039864263445985,
+      "learning_rate": 5.11227050122846e-06,
+      "loss": 0.8401,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5088388214904679,
+      "grad_norm": 3.1133050177249633,
+      "learning_rate": 5.106657850189546e-06,
+      "loss": 0.7091,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5091854419410745,
+      "grad_norm": 2.5310294612167636,
+      "learning_rate": 5.101045064689392e-06,
+      "loss": 0.6718,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5095320623916811,
+      "grad_norm": 2.6842827380398906,
+      "learning_rate": 5.095432151803919e-06,
+      "loss": 0.7028,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5098786828422877,
+      "grad_norm": 38.199579287649684,
+      "learning_rate": 5.089819118609201e-06,
+      "loss": 0.779,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5102253032928943,
+      "grad_norm": 9.53987735630407,
+      "learning_rate": 5.084205972181469e-06,
+      "loss": 0.8517,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5105719237435009,
+      "grad_norm": 4.384824887093704,
+      "learning_rate": 5.078592719597097e-06,
+      "loss": 0.7266,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5109185441941074,
+      "grad_norm": 13.496040201459438,
+      "learning_rate": 5.072979367932586e-06,
+      "loss": 0.8354,
+      "step": 1474
+    },
+    {
+      "epoch": 0.511265164644714,
+      "grad_norm": 2.8090246767778386,
+      "learning_rate": 5.067365924264569e-06,
+      "loss": 0.743,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5116117850953206,
+      "grad_norm": 4.988428543423536,
+      "learning_rate": 5.061752395669794e-06,
+      "loss": 0.8312,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5119584055459272,
+      "grad_norm": 28.480439450135414,
+      "learning_rate": 5.0561387892251115e-06,
+      "loss": 0.8651,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5123050259965338,
+      "grad_norm": 3.4806189178445974,
+      "learning_rate": 5.050525112007477e-06,
+      "loss": 0.7322,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5126516464471403,
+      "grad_norm": 3.528363833067646,
+      "learning_rate": 5.044911371093929e-06,
+      "loss": 0.7706,
+      "step": 1479
+    },
+    {
+      "epoch": 0.512998266897747,
+      "grad_norm": 3.8057043332604623,
+      "learning_rate": 5.039297573561589e-06,
+      "loss": 0.8106,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5133448873483536,
+      "grad_norm": 4.141105290525823,
+      "learning_rate": 5.033683726487654e-06,
+      "loss": 0.8494,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5136915077989601,
+      "grad_norm": 4.097185912114074,
+      "learning_rate": 5.028069836949372e-06,
+      "loss": 0.8063,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5140381282495667,
+      "grad_norm": 3.422460930349162,
+      "learning_rate": 5.022455912024056e-06,
+      "loss": 0.7925,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5143847487001733,
+      "grad_norm": 4.3047597313701305,
+      "learning_rate": 5.01684195878906e-06,
+      "loss": 0.7052,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5147313691507799,
+      "grad_norm": 3.738319937689277,
+      "learning_rate": 5.011227984321773e-06,
+      "loss": 0.7114,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5150779896013865,
+      "grad_norm": 2.815086139370156,
+      "learning_rate": 5.005613995699608e-06,
+      "loss": 0.6529,
+      "step": 1486
+    },
+    {
+      "epoch": 0.515424610051993,
+      "grad_norm": 4.186409535359648,
+      "learning_rate": 5e-06,
+      "loss": 0.7711,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5157712305025997,
+      "grad_norm": 3.273741047836348,
+      "learning_rate": 4.994386004300393e-06,
+      "loss": 0.8114,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5161178509532063,
+      "grad_norm": 3.0078960016882466,
+      "learning_rate": 4.988772015678229e-06,
+      "loss": 0.6947,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5164644714038128,
+      "grad_norm": 3.398608395315642,
+      "learning_rate": 4.983158041210941e-06,
+      "loss": 0.7987,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5168110918544194,
+      "grad_norm": 5.196667482986137,
+      "learning_rate": 4.977544087975945e-06,
+      "loss": 0.6969,
+      "step": 1491
+    },
+    {
+      "epoch": 0.517157712305026,
+      "grad_norm": 3.643067942795144,
+      "learning_rate": 4.97193016305063e-06,
+      "loss": 0.7391,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5175043327556326,
+      "grad_norm": 3.304993928420365,
+      "learning_rate": 4.9663162735123485e-06,
+      "loss": 0.7636,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5178509532062392,
+      "grad_norm": 2.7619955295863723,
+      "learning_rate": 4.960702426438411e-06,
+      "loss": 0.7448,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5181975736568457,
+      "grad_norm": 3.993273772134562,
+      "learning_rate": 4.955088628906073e-06,
+      "loss": 0.8345,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5185441941074523,
+      "grad_norm": 3.579731134515395,
+      "learning_rate": 4.949474887992525e-06,
+      "loss": 0.7506,
+      "step": 1496
+    },
+    {
+      "epoch": 0.518890814558059,
+      "grad_norm": 2.6177244473525594,
+      "learning_rate": 4.94386121077489e-06,
+      "loss": 0.6596,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5192374350086655,
+      "grad_norm": 5.842108454695518,
+      "learning_rate": 4.938247604330209e-06,
+      "loss": 0.8124,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5195840554592721,
+      "grad_norm": 6.678135735898391,
+      "learning_rate": 4.932634075735433e-06,
+      "loss": 0.7442,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5199306759098787,
+      "grad_norm": 13.822531320667277,
+      "learning_rate": 4.927020632067418e-06,
+      "loss": 0.747,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5202772963604853,
+      "grad_norm": 2.2334007873256505,
+      "learning_rate": 4.921407280402907e-06,
+      "loss": 0.7115,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5206239168110919,
+      "grad_norm": 3.546136306126183,
+      "learning_rate": 4.915794027818532e-06,
+      "loss": 0.6843,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5209705372616984,
+      "grad_norm": 6.3399613571488675,
+      "learning_rate": 4.910180881390799e-06,
+      "loss": 0.8304,
+      "step": 1503
+    },
+    {
+      "epoch": 0.521317157712305,
+      "grad_norm": 3.5224193071897894,
+      "learning_rate": 4.904567848196082e-06,
+      "loss": 0.7721,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5216637781629117,
+      "grad_norm": 7.715400109255952,
+      "learning_rate": 4.898954935310608e-06,
+      "loss": 0.8309,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5220103986135182,
+      "grad_norm": 7.533901378637971,
+      "learning_rate": 4.893342149810457e-06,
+      "loss": 0.7862,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5223570190641248,
+      "grad_norm": 3.698808241628836,
+      "learning_rate": 4.887729498771541e-06,
+      "loss": 0.7465,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5227036395147314,
+      "grad_norm": 4.062447048252495,
+      "learning_rate": 4.882116989269611e-06,
+      "loss": 0.6732,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5230502599653379,
+      "grad_norm": 5.795519023814258,
+      "learning_rate": 4.876504628380236e-06,
+      "loss": 0.8011,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5233968804159446,
+      "grad_norm": 4.733816455776274,
+      "learning_rate": 4.8708924231787955e-06,
+      "loss": 0.6877,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5237435008665511,
+      "grad_norm": 7.318929998482837,
+      "learning_rate": 4.865280380740479e-06,
+      "loss": 0.8137,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5240901213171577,
+      "grad_norm": 3.9234179366380166,
+      "learning_rate": 4.859668508140263e-06,
+      "loss": 0.654,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5244367417677643,
+      "grad_norm": 16.47604363706244,
+      "learning_rate": 4.854056812452911e-06,
+      "loss": 0.8031,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5247833622183709,
+      "grad_norm": 4.505014841457624,
+      "learning_rate": 4.848445300752972e-06,
+      "loss": 0.7194,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5251299826689775,
+      "grad_norm": 3.0524255175837363,
+      "learning_rate": 4.842833980114755e-06,
+      "loss": 0.7379,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5254766031195841,
+      "grad_norm": 5.461061485825344,
+      "learning_rate": 4.83722285761233e-06,
+      "loss": 0.7902,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5258232235701906,
+      "grad_norm": 2.801021095806628,
+      "learning_rate": 4.831611940319516e-06,
+      "loss": 0.7286,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5261698440207973,
+      "grad_norm": 4.0964728528878425,
+      "learning_rate": 4.8260012353098755e-06,
+      "loss": 0.8123,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5265164644714038,
+      "grad_norm": 2.7573380126028337,
+      "learning_rate": 4.820390749656704e-06,
+      "loss": 0.6866,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5268630849220104,
+      "grad_norm": 6.689628459085869,
+      "learning_rate": 4.814780490433016e-06,
+      "loss": 0.7438,
+      "step": 1520
+    },
+    {
+      "epoch": 0.527209705372617,
+      "grad_norm": 4.523727389099235,
+      "learning_rate": 4.809170464711549e-06,
+      "loss": 0.7961,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5275563258232235,
+      "grad_norm": 4.100367246706533,
+      "learning_rate": 4.803560679564735e-06,
+      "loss": 0.7048,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5279029462738302,
+      "grad_norm": 3.3975504085523847,
+      "learning_rate": 4.79795114206471e-06,
+      "loss": 0.8105,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5282495667244368,
+      "grad_norm": 2.6701984543262016,
+      "learning_rate": 4.792341859283298e-06,
+      "loss": 0.8808,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5285961871750433,
+      "grad_norm": 7.93717875493004,
+      "learning_rate": 4.786732838291999e-06,
+      "loss": 0.6118,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5289428076256499,
+      "grad_norm": 3.426619048651426,
+      "learning_rate": 4.781124086161984e-06,
+      "loss": 0.7988,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5292894280762565,
+      "grad_norm": 3.318411670505576,
+      "learning_rate": 4.775515609964086e-06,
+      "loss": 0.8389,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5296360485268631,
+      "grad_norm": 4.583473615681366,
+      "learning_rate": 4.769907416768788e-06,
+      "loss": 0.7235,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5299826689774697,
+      "grad_norm": 4.417534404517796,
+      "learning_rate": 4.764299513646217e-06,
+      "loss": 0.7661,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5303292894280762,
+      "grad_norm": 3.8945072363348987,
+      "learning_rate": 4.758691907666136e-06,
+      "loss": 0.8293,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5306759098786828,
+      "grad_norm": 3.4131446940453882,
+      "learning_rate": 4.753084605897935e-06,
+      "loss": 0.8715,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5310225303292895,
+      "grad_norm": 3.9937135465275855,
+      "learning_rate": 4.7474776154106124e-06,
+      "loss": 0.7542,
+      "step": 1532
+    },
+    {
+      "epoch": 0.531369150779896,
+      "grad_norm": 3.0772330725482298,
+      "learning_rate": 4.741870943272781e-06,
+      "loss": 0.7861,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5317157712305026,
+      "grad_norm": 4.824191862589887,
+      "learning_rate": 4.736264596552652e-06,
+      "loss": 0.6955,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5320623916811091,
+      "grad_norm": 3.6225232955318827,
+      "learning_rate": 4.7306585823180255e-06,
+      "loss": 0.6975,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5324090121317158,
+      "grad_norm": 6.186749559835109,
+      "learning_rate": 4.725052907636282e-06,
+      "loss": 0.6682,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5327556325823224,
+      "grad_norm": 4.721261526570961,
+      "learning_rate": 4.719447579574371e-06,
+      "loss": 0.6122,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5331022530329289,
+      "grad_norm": 3.7721165938174885,
+      "learning_rate": 4.713842605198811e-06,
+      "loss": 0.8161,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5334488734835355,
+      "grad_norm": 4.202595725524968,
+      "learning_rate": 4.708237991575671e-06,
+      "loss": 0.794,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5337954939341422,
+      "grad_norm": 7.259663576322378,
+      "learning_rate": 4.702633745770565e-06,
+      "loss": 0.684,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5341421143847487,
+      "grad_norm": 5.357834978016881,
+      "learning_rate": 4.697029874848645e-06,
+      "loss": 0.9267,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5344887348353553,
+      "grad_norm": 9.679693134412382,
+      "learning_rate": 4.69142638587459e-06,
+      "loss": 0.7589,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5348353552859618,
+      "grad_norm": 3.5927464996743743,
+      "learning_rate": 4.685823285912594e-06,
+      "loss": 0.7805,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5351819757365684,
+      "grad_norm": 4.002430094410755,
+      "learning_rate": 4.6802205820263665e-06,
+      "loss": 0.7866,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5355285961871751,
+      "grad_norm": 7.416381977642875,
+      "learning_rate": 4.674618281279114e-06,
+      "loss": 0.7937,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5358752166377816,
+      "grad_norm": 12.079404027704957,
+      "learning_rate": 4.6690163907335365e-06,
+      "loss": 0.6906,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5362218370883882,
+      "grad_norm": 3.644673567345844,
+      "learning_rate": 4.663414917451815e-06,
+      "loss": 0.6891,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5365684575389948,
+      "grad_norm": 21.009652264656314,
+      "learning_rate": 4.657813868495604e-06,
+      "loss": 0.7603,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5369150779896014,
+      "grad_norm": 4.759430121203286,
+      "learning_rate": 4.652213250926025e-06,
+      "loss": 0.8442,
+      "step": 1549
+    },
+    {
+      "epoch": 0.537261698440208,
+      "grad_norm": 4.370610666222347,
+      "learning_rate": 4.646613071803657e-06,
+      "loss": 0.7184,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5376083188908145,
+      "grad_norm": 4.886252971791567,
+      "learning_rate": 4.641013338188521e-06,
+      "loss": 0.8257,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5379549393414211,
+      "grad_norm": 5.87916265628562,
+      "learning_rate": 4.635414057140085e-06,
+      "loss": 0.8217,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5383015597920278,
+      "grad_norm": 4.81919431417475,
+      "learning_rate": 4.629815235717234e-06,
+      "loss": 0.7552,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5386481802426343,
+      "grad_norm": 3.411332355606129,
+      "learning_rate": 4.624216880978287e-06,
+      "loss": 0.7085,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5389948006932409,
+      "grad_norm": 5.751138256803439,
+      "learning_rate": 4.618618999980965e-06,
+      "loss": 0.8791,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5393414211438475,
+      "grad_norm": 5.828625053319229,
+      "learning_rate": 4.6130215997823965e-06,
+      "loss": 0.6679,
+      "step": 1556
+    },
+    {
+      "epoch": 0.539688041594454,
+      "grad_norm": 3.266542058952399,
+      "learning_rate": 4.607424687439102e-06,
+      "loss": 0.7056,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5400346620450607,
+      "grad_norm": 3.3201862272808556,
+      "learning_rate": 4.601828270006988e-06,
+      "loss": 0.7192,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5403812824956672,
+      "grad_norm": 3.6398256505497177,
+      "learning_rate": 4.596232354541335e-06,
+      "loss": 0.6548,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5407279029462738,
+      "grad_norm": 4.306812259250759,
+      "learning_rate": 4.590636948096793e-06,
+      "loss": 0.7296,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5410745233968804,
+      "grad_norm": 4.6667616253046615,
+      "learning_rate": 4.58504205772737e-06,
+      "loss": 0.7415,
+      "step": 1561
+    },
+    {
+      "epoch": 0.541421143847487,
+      "grad_norm": 6.40909645996888,
+      "learning_rate": 4.5794476904864245e-06,
+      "loss": 0.7562,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5417677642980936,
+      "grad_norm": 8.205579906113217,
+      "learning_rate": 4.573853853426651e-06,
+      "loss": 0.7559,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5421143847487002,
+      "grad_norm": 4.759627284155818,
+      "learning_rate": 4.568260553600078e-06,
+      "loss": 0.8526,
+      "step": 1564
+    },
+    {
+      "epoch": 0.5424610051993067,
+      "grad_norm": 8.90834888517929,
+      "learning_rate": 4.5626677980580605e-06,
+      "loss": 0.8207,
+      "step": 1565
+    },
+    {
+      "epoch": 0.5428076256499134,
+      "grad_norm": 3.5553102059388726,
+      "learning_rate": 4.557075593851262e-06,
+      "loss": 0.7704,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5431542461005199,
+      "grad_norm": 4.0880603484799245,
+      "learning_rate": 4.551483948029655e-06,
+      "loss": 0.6514,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5435008665511265,
+      "grad_norm": 4.327372955782528,
+      "learning_rate": 4.545892867642503e-06,
+      "loss": 0.7581,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5438474870017331,
+      "grad_norm": 7.257996911593342,
+      "learning_rate": 4.540302359738361e-06,
+      "loss": 0.7296,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5441941074523396,
+      "grad_norm": 2.641068354651477,
+      "learning_rate": 4.534712431365062e-06,
+      "loss": 0.6683,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5445407279029463,
+      "grad_norm": 2.9644203231207022,
+      "learning_rate": 4.529123089569706e-06,
+      "loss": 0.8022,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5448873483535529,
+      "grad_norm": 14.922840838988717,
+      "learning_rate": 4.5235343413986585e-06,
+      "loss": 0.7911,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5452339688041594,
+      "grad_norm": 4.8922386368337385,
+      "learning_rate": 4.5179461938975285e-06,
+      "loss": 0.8069,
+      "step": 1573
+    },
+    {
+      "epoch": 0.545580589254766,
+      "grad_norm": 4.00377861389218,
+      "learning_rate": 4.512358654111173e-06,
+      "loss": 0.7226,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5459272097053726,
+      "grad_norm": 5.796331871343918,
+      "learning_rate": 4.5067717290836835e-06,
+      "loss": 0.7453,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5462738301559792,
+      "grad_norm": 4.306860721162211,
+      "learning_rate": 4.501185425858376e-06,
+      "loss": 0.7225,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5466204506065858,
+      "grad_norm": 3.1224906166388298,
+      "learning_rate": 4.49559975147778e-06,
+      "loss": 0.7487,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5469670710571923,
+      "grad_norm": 4.219805014518954,
+      "learning_rate": 4.4900147129836376e-06,
+      "loss": 0.7266,
+      "step": 1578
+    },
+    {
+      "epoch": 0.547313691507799,
+      "grad_norm": 4.171518711899639,
+      "learning_rate": 4.48443031741688e-06,
+      "loss": 0.7646,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5476603119584056,
+      "grad_norm": 5.478583432055409,
+      "learning_rate": 4.478846571817637e-06,
+      "loss": 0.7805,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5480069324090121,
+      "grad_norm": 4.268709561524463,
+      "learning_rate": 4.473263483225214e-06,
+      "loss": 0.7928,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5483535528596187,
+      "grad_norm": 5.216328871281038,
+      "learning_rate": 4.467681058678092e-06,
+      "loss": 0.7595,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5487001733102254,
+      "grad_norm": 5.058400781319662,
+      "learning_rate": 4.4620993052139134e-06,
+      "loss": 0.7359,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5490467937608319,
+      "grad_norm": 5.277670723337451,
+      "learning_rate": 4.456518229869468e-06,
+      "loss": 0.7531,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5493934142114385,
+      "grad_norm": 4.736728126214752,
+      "learning_rate": 4.450937839680701e-06,
+      "loss": 0.7351,
+      "step": 1585
+    },
+    {
+      "epoch": 0.549740034662045,
+      "grad_norm": 11.519492983089163,
+      "learning_rate": 4.445358141682688e-06,
+      "loss": 0.8052,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5500866551126516,
+      "grad_norm": 5.3557887091532175,
+      "learning_rate": 4.439779142909632e-06,
+      "loss": 0.8251,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5504332755632583,
+      "grad_norm": 2.3179563453619636,
+      "learning_rate": 4.434200850394861e-06,
+      "loss": 0.7613,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5507798960138648,
+      "grad_norm": 11.822401039138255,
+      "learning_rate": 4.4286232711708e-06,
+      "loss": 0.7549,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5511265164644714,
+      "grad_norm": 3.5993045850805188,
+      "learning_rate": 4.423046412268986e-06,
+      "loss": 0.8324,
+      "step": 1590
+    },
+    {
+      "epoch": 0.551473136915078,
+      "grad_norm": 2.8871955768141917,
+      "learning_rate": 4.4174702807200435e-06,
+      "loss": 0.7356,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5518197573656846,
+      "grad_norm": 7.939647155989422,
+      "learning_rate": 4.41189488355368e-06,
+      "loss": 0.7495,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5521663778162912,
+      "grad_norm": 3.6330060657882397,
+      "learning_rate": 4.406320227798683e-06,
+      "loss": 0.8227,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5525129982668977,
+      "grad_norm": 5.705070688469803,
+      "learning_rate": 4.4007463204828905e-06,
+      "loss": 0.7043,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5528596187175043,
+      "grad_norm": 4.8298555723167125,
+      "learning_rate": 4.395173168633213e-06,
+      "loss": 0.8219,
+      "step": 1595
+    },
+    {
+      "epoch": 0.553206239168111,
+      "grad_norm": 3.903561120566001,
+      "learning_rate": 4.389600779275601e-06,
+      "loss": 0.6315,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5535528596187175,
+      "grad_norm": 5.795271394033595,
+      "learning_rate": 4.384029159435044e-06,
+      "loss": 0.7273,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5538994800693241,
+      "grad_norm": 3.0834436840744988,
+      "learning_rate": 4.378458316135564e-06,
+      "loss": 0.7781,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5542461005199307,
+      "grad_norm": 4.545033283308228,
+      "learning_rate": 4.3728882564002e-06,
+      "loss": 0.7092,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5545927209705372,
+      "grad_norm": 4.590036140032168,
+      "learning_rate": 4.367318987251007e-06,
+      "loss": 0.7847,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5549393414211439,
+      "grad_norm": 4.277151080637421,
+      "learning_rate": 4.36175051570904e-06,
+      "loss": 0.7094,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5552859618717504,
+      "grad_norm": 2.9789781937402453,
+      "learning_rate": 4.356182848794352e-06,
+      "loss": 0.6541,
+      "step": 1602
+    },
+    {
+      "epoch": 0.555632582322357,
+      "grad_norm": 4.567678612056862,
+      "learning_rate": 4.350615993525982e-06,
+      "loss": 0.7041,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5559792027729636,
+      "grad_norm": 4.468953750935942,
+      "learning_rate": 4.345049956921938e-06,
+      "loss": 0.8039,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5563258232235702,
+      "grad_norm": 3.634741915767966,
+      "learning_rate": 4.339484745999206e-06,
+      "loss": 0.8491,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5566724436741768,
+      "grad_norm": 5.1370392855219285,
+      "learning_rate": 4.3339203677737265e-06,
+      "loss": 0.7052,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5570190641247834,
+      "grad_norm": 7.594082085859103,
+      "learning_rate": 4.328356829260388e-06,
+      "loss": 0.666,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5573656845753899,
+      "grad_norm": 8.41162376034449,
+      "learning_rate": 4.322794137473026e-06,
+      "loss": 0.7555,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5577123050259966,
+      "grad_norm": 3.0789572727417065,
+      "learning_rate": 4.317232299424402e-06,
+      "loss": 0.8117,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5580589254766031,
+      "grad_norm": 4.269801774105763,
+      "learning_rate": 4.311671322126207e-06,
+      "loss": 0.7659,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5584055459272097,
+      "grad_norm": 2.69644667438654,
+      "learning_rate": 4.306111212589042e-06,
+      "loss": 0.6413,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5587521663778163,
+      "grad_norm": 7.356643076607626,
+      "learning_rate": 4.300551977822418e-06,
+      "loss": 0.7391,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5590987868284228,
+      "grad_norm": 3.587701158637125,
+      "learning_rate": 4.294993624834743e-06,
+      "loss": 0.6534,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5594454072790295,
+      "grad_norm": 2.93195764160278,
+      "learning_rate": 4.28943616063331e-06,
+      "loss": 0.7011,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5597920277296361,
+      "grad_norm": 3.9906468018753856,
+      "learning_rate": 4.283879592224295e-06,
+      "loss": 0.6938,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5601386481802426,
+      "grad_norm": 3.8661119595373985,
+      "learning_rate": 4.278323926612742e-06,
+      "loss": 0.6901,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5604852686308492,
+      "grad_norm": 5.086415866462342,
+      "learning_rate": 4.272769170802558e-06,
+      "loss": 0.8251,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5608318890814558,
+      "grad_norm": 5.234614545254528,
+      "learning_rate": 4.267215331796505e-06,
+      "loss": 0.7395,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5611785095320624,
+      "grad_norm": 7.445838616585659,
+      "learning_rate": 4.261662416596187e-06,
+      "loss": 0.7055,
+      "step": 1619
+    },
+    {
+      "epoch": 0.561525129982669,
+      "grad_norm": 2.7442362596243126,
+      "learning_rate": 4.256110432202041e-06,
+      "loss": 0.647,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5618717504332755,
+      "grad_norm": 3.698073292217916,
+      "learning_rate": 4.250559385613336e-06,
+      "loss": 0.722,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5622183708838822,
+      "grad_norm": 3.3913826693914713,
+      "learning_rate": 4.245009283828153e-06,
+      "loss": 0.7552,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5625649913344888,
+      "grad_norm": 2.8886262568105594,
+      "learning_rate": 4.239460133843389e-06,
+      "loss": 0.8559,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5629116117850953,
+      "grad_norm": 5.42818892744633,
+      "learning_rate": 4.233911942654736e-06,
+      "loss": 0.64,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5632582322357019,
+      "grad_norm": 3.728499794271646,
+      "learning_rate": 4.228364717256675e-06,
+      "loss": 0.7513,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5636048526863084,
+      "grad_norm": 3.91189555892949,
+      "learning_rate": 4.222818464642477e-06,
+      "loss": 0.7048,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5639514731369151,
+      "grad_norm": 2.6293825314414865,
+      "learning_rate": 4.21727319180418e-06,
+      "loss": 0.6894,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5642980935875217,
+      "grad_norm": 7.687904592116894,
+      "learning_rate": 4.211728905732589e-06,
+      "loss": 0.6901,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5646447140381282,
+      "grad_norm": 5.479987533928917,
+      "learning_rate": 4.206185613417267e-06,
+      "loss": 0.6696,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5649913344887348,
+      "grad_norm": 4.257227296662813,
+      "learning_rate": 4.20064332184652e-06,
+      "loss": 0.7771,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5653379549393415,
+      "grad_norm": 5.488378856377371,
+      "learning_rate": 4.195102038007395e-06,
+      "loss": 0.7269,
+      "step": 1631
+    },
+    {
+      "epoch": 0.565684575389948,
+      "grad_norm": 3.5394886588787164,
+      "learning_rate": 4.189561768885669e-06,
+      "loss": 0.7522,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5660311958405546,
+      "grad_norm": 4.236691123359453,
+      "learning_rate": 4.184022521465841e-06,
+      "loss": 0.7483,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5663778162911611,
+      "grad_norm": 29.9384653562957,
+      "learning_rate": 4.17848430273112e-06,
+      "loss": 0.7422,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5667244367417678,
+      "grad_norm": 8.338002592070307,
+      "learning_rate": 4.172947119663412e-06,
+      "loss": 0.9081,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5670710571923744,
+      "grad_norm": 3.910576599389458,
+      "learning_rate": 4.167410979243331e-06,
+      "loss": 0.7213,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5674176776429809,
+      "grad_norm": 12.014125240014215,
+      "learning_rate": 4.1618758884501655e-06,
+      "loss": 0.7174,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5677642980935875,
+      "grad_norm": 4.56211536665226,
+      "learning_rate": 4.156341854261885e-06,
+      "loss": 0.8903,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5681109185441942,
+      "grad_norm": 6.118492058568657,
+      "learning_rate": 4.150808883655128e-06,
+      "loss": 0.8325,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5684575389948007,
+      "grad_norm": 4.735667860052605,
+      "learning_rate": 4.145276983605187e-06,
+      "loss": 0.7148,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5688041594454073,
+      "grad_norm": 3.0549170338534326,
+      "learning_rate": 4.139746161086011e-06,
+      "loss": 0.6764,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5691507798960138,
+      "grad_norm": 2.5188476313751447,
+      "learning_rate": 4.134216423070186e-06,
+      "loss": 0.8259,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5694974003466204,
+      "grad_norm": 3.46626287423553,
+      "learning_rate": 4.128687776528935e-06,
+      "loss": 0.673,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5698440207972271,
+      "grad_norm": 5.977950221507219,
+      "learning_rate": 4.1231602284321044e-06,
+      "loss": 0.6774,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5701906412478336,
+      "grad_norm": 10.9924477128594,
+      "learning_rate": 4.117633785748148e-06,
+      "loss": 0.7719,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5705372616984402,
+      "grad_norm": 3.647660677714126,
+      "learning_rate": 4.112108455444137e-06,
+      "loss": 0.6767,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5708838821490468,
+      "grad_norm": 5.56113946127857,
+      "learning_rate": 4.106584244485734e-06,
+      "loss": 0.8207,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5712305025996534,
+      "grad_norm": 6.89100428603212,
+      "learning_rate": 4.101061159837193e-06,
+      "loss": 0.661,
+      "step": 1648
+    },
+    {
+      "epoch": 0.57157712305026,
+      "grad_norm": 3.271401270619684,
+      "learning_rate": 4.095539208461346e-06,
+      "loss": 0.7313,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5719237435008665,
+      "grad_norm": 12.80187361691596,
+      "learning_rate": 4.090018397319598e-06,
+      "loss": 0.7081,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5722703639514731,
+      "grad_norm": 4.281870438955778,
+      "learning_rate": 4.084498733371914e-06,
+      "loss": 0.7476,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5726169844020798,
+      "grad_norm": 2.9666765546688705,
+      "learning_rate": 4.078980223576815e-06,
+      "loss": 0.7351,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5729636048526863,
+      "grad_norm": 2.9641176285187156,
+      "learning_rate": 4.073462874891369e-06,
+      "loss": 0.7727,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5733102253032929,
+      "grad_norm": 3.0684069276678434,
+      "learning_rate": 4.0679466942711766e-06,
+      "loss": 0.786,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5736568457538995,
+      "grad_norm": 4.923205618220261,
+      "learning_rate": 4.0624316886703665e-06,
+      "loss": 0.6805,
+      "step": 1655
+    },
+    {
+      "epoch": 0.574003466204506,
+      "grad_norm": 3.405057048568034,
+      "learning_rate": 4.056917865041587e-06,
+      "loss": 0.8025,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5743500866551127,
+      "grad_norm": 3.039471330849355,
+      "learning_rate": 4.0514052303359955e-06,
+      "loss": 0.9417,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5746967071057192,
+      "grad_norm": 4.795413466759691,
+      "learning_rate": 4.04589379150325e-06,
+      "loss": 0.702,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5750433275563258,
+      "grad_norm": 2.7334551959676534,
+      "learning_rate": 4.0403835554915045e-06,
+      "loss": 0.6925,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5753899480069324,
+      "grad_norm": 3.8292081254375607,
+      "learning_rate": 4.034874529247393e-06,
+      "loss": 0.7259,
+      "step": 1660
+    },
+    {
+      "epoch": 0.575736568457539,
+      "grad_norm": 3.2604562182786623,
+      "learning_rate": 4.029366719716025e-06,
+      "loss": 0.7239,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5760831889081456,
+      "grad_norm": 43.772635810517976,
+      "learning_rate": 4.023860133840974e-06,
+      "loss": 0.7026,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5764298093587522,
+      "grad_norm": 4.339690355248921,
+      "learning_rate": 4.018354778564278e-06,
+      "loss": 0.7723,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5767764298093587,
+      "grad_norm": 2.8549227440419216,
+      "learning_rate": 4.012850660826416e-06,
+      "loss": 0.7307,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5771230502599654,
+      "grad_norm": 3.2561208937983794,
+      "learning_rate": 4.007347787566311e-06,
+      "loss": 0.8605,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5774696707105719,
+      "grad_norm": 5.194104083110642,
+      "learning_rate": 4.0018461657213155e-06,
+      "loss": 0.7498,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5778162911611785,
+      "grad_norm": 3.806554702217227,
+      "learning_rate": 3.996345802227205e-06,
+      "loss": 0.6738,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5781629116117851,
+      "grad_norm": 3.761118865102344,
+      "learning_rate": 3.990846704018169e-06,
+      "loss": 0.8063,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5785095320623916,
+      "grad_norm": 3.559179931458347,
+      "learning_rate": 3.985348878026799e-06,
+      "loss": 0.6786,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5788561525129983,
+      "grad_norm": 3.8918479510761292,
+      "learning_rate": 3.979852331184089e-06,
+      "loss": 0.7636,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5792027729636049,
+      "grad_norm": 7.1867811310045235,
+      "learning_rate": 3.974357070419412e-06,
+      "loss": 0.777,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5795493934142114,
+      "grad_norm": 7.5006160032961,
+      "learning_rate": 3.968863102660526e-06,
+      "loss": 0.6835,
+      "step": 1672
+    },
+    {
+      "epoch": 0.579896013864818,
+      "grad_norm": 2.694344428849171,
+      "learning_rate": 3.963370434833556e-06,
+      "loss": 0.7582,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5802426343154246,
+      "grad_norm": 7.856527011474167,
+      "learning_rate": 3.957879073862991e-06,
+      "loss": 0.7987,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5805892547660312,
+      "grad_norm": 6.638262571630389,
+      "learning_rate": 3.952389026671671e-06,
+      "loss": 0.7637,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5809358752166378,
+      "grad_norm": 2.5897707845146662,
+      "learning_rate": 3.946900300180777e-06,
+      "loss": 0.6958,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5812824956672443,
+      "grad_norm": 3.1737601998967486,
+      "learning_rate": 3.941412901309829e-06,
+      "loss": 0.8337,
+      "step": 1677
+    },
+    {
+      "epoch": 0.581629116117851,
+      "grad_norm": 5.776399501625259,
+      "learning_rate": 3.93592683697667e-06,
+      "loss": 0.7723,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5819757365684576,
+      "grad_norm": 4.8418338834181505,
+      "learning_rate": 3.9304421140974645e-06,
+      "loss": 0.8178,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5823223570190641,
+      "grad_norm": 3.746719476895542,
+      "learning_rate": 3.924958739586684e-06,
+      "loss": 0.802,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5826689774696707,
+      "grad_norm": 6.041883852064335,
+      "learning_rate": 3.919476720357097e-06,
+      "loss": 0.7008,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5830155979202774,
+      "grad_norm": 6.578636216897437,
+      "learning_rate": 3.913996063319768e-06,
+      "loss": 0.6967,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5833622183708839,
+      "grad_norm": 17.91560688391045,
+      "learning_rate": 3.908516775384041e-06,
+      "loss": 0.6824,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5837088388214905,
+      "grad_norm": 5.441688913027476,
+      "learning_rate": 3.903038863457537e-06,
+      "loss": 0.8405,
+      "step": 1684
+    },
+    {
+      "epoch": 0.584055459272097,
+      "grad_norm": 4.510390929684881,
+      "learning_rate": 3.897562334446141e-06,
+      "loss": 0.7021,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5844020797227036,
+      "grad_norm": 2.7123982343689854,
+      "learning_rate": 3.8920871952539945e-06,
+      "loss": 0.824,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5847487001733103,
+      "grad_norm": 11.836195806032915,
+      "learning_rate": 3.886613452783485e-06,
+      "loss": 0.6661,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5850953206239168,
+      "grad_norm": 5.115330794274779,
+      "learning_rate": 3.881141113935241e-06,
+      "loss": 0.8369,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5854419410745234,
+      "grad_norm": 29.508309443800904,
+      "learning_rate": 3.875670185608123e-06,
+      "loss": 0.6764,
+      "step": 1689
+    },
+    {
+      "epoch": 0.58578856152513,
+      "grad_norm": 3.441118046906484,
+      "learning_rate": 3.870200674699208e-06,
+      "loss": 0.6533,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5861351819757366,
+      "grad_norm": 10.07089732615343,
+      "learning_rate": 3.864732588103798e-06,
+      "loss": 0.6755,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5864818024263432,
+      "grad_norm": 6.600273643450389,
+      "learning_rate": 3.859265932715381e-06,
+      "loss": 0.6638,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5868284228769497,
+      "grad_norm": 4.85974015645799,
+      "learning_rate": 3.8538007154256565e-06,
+      "loss": 0.686,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5871750433275563,
+      "grad_norm": 3.6685920743373037,
+      "learning_rate": 3.8483369431245045e-06,
+      "loss": 0.7933,
+      "step": 1694
+    },
+    {
+      "epoch": 0.587521663778163,
+      "grad_norm": 5.96590191199413,
+      "learning_rate": 3.842874622699984e-06,
+      "loss": 0.734,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5878682842287695,
+      "grad_norm": 10.730820143094912,
+      "learning_rate": 3.837413761038325e-06,
+      "loss": 0.7419,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5882149046793761,
+      "grad_norm": 2.7569673654236806,
+      "learning_rate": 3.831954365023915e-06,
+      "loss": 0.7536,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5885615251299827,
+      "grad_norm": 3.380493235760589,
+      "learning_rate": 3.826496441539298e-06,
+      "loss": 0.7559,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5889081455805892,
+      "grad_norm": 8.950283114189117,
+      "learning_rate": 3.821039997465159e-06,
+      "loss": 0.7603,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5892547660311959,
+      "grad_norm": 4.205252429073048,
+      "learning_rate": 3.815585039680317e-06,
+      "loss": 0.8239,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5896013864818024,
+      "grad_norm": 5.179977899226733,
+      "learning_rate": 3.8101315750617246e-06,
+      "loss": 0.7173,
+      "step": 1701
+    },
+    {
+      "epoch": 0.589948006932409,
+      "grad_norm": 3.702228949602466,
+      "learning_rate": 3.80467961048444e-06,
+      "loss": 0.6919,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5902946273830156,
+      "grad_norm": 2.8570179823974646,
+      "learning_rate": 3.799229152821638e-06,
+      "loss": 0.6997,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5906412478336222,
+      "grad_norm": 3.02493704740985,
+      "learning_rate": 3.793780208944595e-06,
+      "loss": 0.6974,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5909878682842288,
+      "grad_norm": 8.067294751934071,
+      "learning_rate": 3.7883327857226737e-06,
+      "loss": 0.7414,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5913344887348354,
+      "grad_norm": 3.5370575155652397,
+      "learning_rate": 3.782886890023325e-06,
+      "loss": 0.7135,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5916811091854419,
+      "grad_norm": 3.229254068198954,
+      "learning_rate": 3.77744252871207e-06,
+      "loss": 0.8519,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5920277296360485,
+      "grad_norm": 3.73274279611886,
+      "learning_rate": 3.771999708652495e-06,
+      "loss": 0.8461,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5923743500866551,
+      "grad_norm": 3.1277522366785635,
+      "learning_rate": 3.766558436706248e-06,
+      "loss": 0.7692,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5927209705372617,
+      "grad_norm": 11.9998954665211,
+      "learning_rate": 3.7611187197330203e-06,
+      "loss": 0.7436,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5930675909878683,
+      "grad_norm": 3.704428554033229,
+      "learning_rate": 3.7556805645905494e-06,
+      "loss": 0.7775,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5934142114384748,
+      "grad_norm": 3.127904152907681,
+      "learning_rate": 3.7502439781345946e-06,
+      "loss": 0.7869,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5937608318890815,
+      "grad_norm": 13.168747596476715,
+      "learning_rate": 3.7448089672189425e-06,
+      "loss": 0.7739,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5941074523396881,
+      "grad_norm": 2.9498689042733917,
+      "learning_rate": 3.739375538695397e-06,
+      "loss": 0.6825,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5944540727902946,
+      "grad_norm": 3.0827928904676436,
+      "learning_rate": 3.7339436994137614e-06,
+      "loss": 0.7664,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5948006932409012,
+      "grad_norm": 4.693405267773241,
+      "learning_rate": 3.728513456221839e-06,
+      "loss": 0.7435,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5951473136915078,
+      "grad_norm": 3.9325609386487805,
+      "learning_rate": 3.7230848159654177e-06,
+      "loss": 0.7414,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5954939341421144,
+      "grad_norm": 3.7097123333698625,
+      "learning_rate": 3.717657785488267e-06,
+      "loss": 0.8034,
+      "step": 1718
+    },
+    {
+      "epoch": 0.595840554592721,
+      "grad_norm": 5.712417794611662,
+      "learning_rate": 3.712232371632127e-06,
+      "loss": 0.7227,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5961871750433275,
+      "grad_norm": 2.797283916104411,
+      "learning_rate": 3.7068085812367003e-06,
+      "loss": 0.7362,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5965337954939341,
+      "grad_norm": 3.7465952584692666,
+      "learning_rate": 3.701386421139639e-06,
+      "loss": 0.8781,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5968804159445408,
+      "grad_norm": 10.727951448689478,
+      "learning_rate": 3.6959658981765467e-06,
+      "loss": 0.8779,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5972270363951473,
+      "grad_norm": 3.3994986984186144,
+      "learning_rate": 3.6905470191809547e-06,
+      "loss": 0.7465,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5975736568457539,
+      "grad_norm": 4.079518579685516,
+      "learning_rate": 3.6851297909843285e-06,
+      "loss": 0.7128,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5979202772963604,
+      "grad_norm": 5.98273072060238,
+      "learning_rate": 3.6797142204160507e-06,
+      "loss": 0.7909,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5982668977469671,
+      "grad_norm": 9.218433380200908,
+      "learning_rate": 3.674300314303412e-06,
+      "loss": 0.7552,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5986135181975737,
+      "grad_norm": 6.098685633188484,
+      "learning_rate": 3.668888079471609e-06,
+      "loss": 0.8285,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5989601386481802,
+      "grad_norm": 6.428635321016713,
+      "learning_rate": 3.663477522743726e-06,
+      "loss": 0.7271,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5993067590987868,
+      "grad_norm": 20.17096655007657,
+      "learning_rate": 3.658068650940735e-06,
+      "loss": 0.7818,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5996533795493935,
+      "grad_norm": 5.831116775958084,
+      "learning_rate": 3.652661470881484e-06,
+      "loss": 0.6964,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 6.615427660075534,
+      "learning_rate": 3.6472559893826866e-06,
+      "loss": 0.7535,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6003466204506066,
+      "grad_norm": 3.695646123138826,
+      "learning_rate": 3.6418522132589198e-06,
+      "loss": 0.7591,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6006932409012131,
+      "grad_norm": 2.497004157149348,
+      "learning_rate": 3.6364501493226007e-06,
+      "loss": 0.7691,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6010398613518197,
+      "grad_norm": 4.889527464656943,
+      "learning_rate": 3.631049804383998e-06,
+      "loss": 0.7773,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6013864818024264,
+      "grad_norm": 54.95562187656622,
+      "learning_rate": 3.6256511852512098e-06,
+      "loss": 0.7352,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6017331022530329,
+      "grad_norm": 3.9007252297167914,
+      "learning_rate": 3.620254298730157e-06,
+      "loss": 0.6842,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6020797227036395,
+      "grad_norm": 58.23315271611972,
+      "learning_rate": 3.614859151624578e-06,
+      "loss": 0.6516,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6024263431542461,
+      "grad_norm": 4.173099721806458,
+      "learning_rate": 3.6094657507360165e-06,
+      "loss": 0.7272,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6027729636048527,
+      "grad_norm": 4.515295681484046,
+      "learning_rate": 3.6040741028638167e-06,
+      "loss": 0.7727,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6031195840554593,
+      "grad_norm": 6.320743604163127,
+      "learning_rate": 3.598684214805112e-06,
+      "loss": 0.7257,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6034662045060658,
+      "grad_norm": 4.2003205353279744,
+      "learning_rate": 3.5932960933548167e-06,
+      "loss": 0.7886,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6038128249566724,
+      "grad_norm": 6.979412726925318,
+      "learning_rate": 3.5879097453056223e-06,
+      "loss": 0.7625,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6041594454072791,
+      "grad_norm": 3.6130023326433234,
+      "learning_rate": 3.582525177447976e-06,
+      "loss": 0.7566,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6045060658578856,
+      "grad_norm": 3.9353998628340134,
+      "learning_rate": 3.5771423965700863e-06,
+      "loss": 0.766,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6048526863084922,
+      "grad_norm": 2.786545879204279,
+      "learning_rate": 3.571761409457911e-06,
+      "loss": 0.8076,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6051993067590988,
+      "grad_norm": 2.5247585674876456,
+      "learning_rate": 3.5663822228951427e-06,
+      "loss": 0.8066,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6055459272097053,
+      "grad_norm": 3.677741202813628,
+      "learning_rate": 3.5610048436632062e-06,
+      "loss": 0.7363,
+      "step": 1747
+    },
+    {
+      "epoch": 0.605892547660312,
+      "grad_norm": 3.928520719735041,
+      "learning_rate": 3.555629278541244e-06,
+      "loss": 0.7374,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6062391681109185,
+      "grad_norm": 4.17162636477745,
+      "learning_rate": 3.5502555343061164e-06,
+      "loss": 0.7105,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6065857885615251,
+      "grad_norm": 3.5382255812418477,
+      "learning_rate": 3.5448836177323852e-06,
+      "loss": 0.7422,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6069324090121317,
+      "grad_norm": 4.603878543357415,
+      "learning_rate": 3.539513535592309e-06,
+      "loss": 0.6886,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6072790294627383,
+      "grad_norm": 3.209925730466833,
+      "learning_rate": 3.5341452946558385e-06,
+      "loss": 0.6807,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6076256499133449,
+      "grad_norm": 2.595798307899967,
+      "learning_rate": 3.5287789016905923e-06,
+      "loss": 0.6635,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6079722703639515,
+      "grad_norm": 8.870599740561438,
+      "learning_rate": 3.523414363461867e-06,
+      "loss": 0.7371,
+      "step": 1754
+    },
+    {
+      "epoch": 0.608318890814558,
+      "grad_norm": 3.4412564801461976,
+      "learning_rate": 3.518051686732621e-06,
+      "loss": 0.6909,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6086655112651647,
+      "grad_norm": 2.3765534425063586,
+      "learning_rate": 3.512690878263464e-06,
+      "loss": 0.7229,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6090121317157712,
+      "grad_norm": 3.4231390845742693,
+      "learning_rate": 3.5073319448126516e-06,
+      "loss": 0.7921,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6093587521663778,
+      "grad_norm": 22.301774833025174,
+      "learning_rate": 3.5019748931360754e-06,
+      "loss": 0.7564,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6097053726169844,
+      "grad_norm": 7.563416673009421,
+      "learning_rate": 3.4966197299872516e-06,
+      "loss": 0.8044,
+      "step": 1759
+    },
+    {
+      "epoch": 0.610051993067591,
+      "grad_norm": 4.337680223715413,
+      "learning_rate": 3.4912664621173196e-06,
+      "loss": 0.7694,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6103986135181976,
+      "grad_norm": 2.8677982981914556,
+      "learning_rate": 3.485915096275028e-06,
+      "loss": 0.7077,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6107452339688042,
+      "grad_norm": 11.750029593411469,
+      "learning_rate": 3.4805656392067287e-06,
+      "loss": 0.8768,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6110918544194107,
+      "grad_norm": 3.409919144558278,
+      "learning_rate": 3.475218097656367e-06,
+      "loss": 0.7515,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6114384748700173,
+      "grad_norm": 5.307335086467559,
+      "learning_rate": 3.469872478365468e-06,
+      "loss": 0.7053,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6117850953206239,
+      "grad_norm": 8.03672428201738,
+      "learning_rate": 3.464528788073142e-06,
+      "loss": 0.8108,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6121317157712305,
+      "grad_norm": 4.123925011387277,
+      "learning_rate": 3.459187033516062e-06,
+      "loss": 0.8401,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6124783362218371,
+      "grad_norm": 3.0467898339063946,
+      "learning_rate": 3.4538472214284623e-06,
+      "loss": 0.6969,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6128249566724436,
+      "grad_norm": 2.756154693160888,
+      "learning_rate": 3.44850935854213e-06,
+      "loss": 0.6791,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6131715771230503,
+      "grad_norm": 5.593966942965895,
+      "learning_rate": 3.4431734515863898e-06,
+      "loss": 0.7782,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6135181975736569,
+      "grad_norm": 3.664461767635982,
+      "learning_rate": 3.437839507288105e-06,
+      "loss": 0.7459,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6138648180242634,
+      "grad_norm": 17.36173502539667,
+      "learning_rate": 3.4325075323716628e-06,
+      "loss": 0.8082,
+      "step": 1771
+    },
+    {
+      "epoch": 0.61421143847487,
+      "grad_norm": 3.9793510089873623,
+      "learning_rate": 3.4271775335589703e-06,
+      "loss": 0.6913,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6145580589254767,
+      "grad_norm": 5.793814975904233,
+      "learning_rate": 3.421849517569442e-06,
+      "loss": 0.708,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6149046793760832,
+      "grad_norm": 6.7601761023949365,
+      "learning_rate": 3.4165234911199886e-06,
+      "loss": 0.8252,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6152512998266898,
+      "grad_norm": 4.669062043144735,
+      "learning_rate": 3.4111994609250178e-06,
+      "loss": 0.773,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6155979202772963,
+      "grad_norm": 5.799363227740167,
+      "learning_rate": 3.4058774336964195e-06,
+      "loss": 0.6802,
+      "step": 1776
+    },
+    {
+      "epoch": 0.615944540727903,
+      "grad_norm": 3.276063150418488,
+      "learning_rate": 3.4005574161435575e-06,
+      "loss": 0.7684,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6162911611785096,
+      "grad_norm": 11.113181902185076,
+      "learning_rate": 3.395239414973263e-06,
+      "loss": 0.7798,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6166377816291161,
+      "grad_norm": 3.1966063308940664,
+      "learning_rate": 3.3899234368898236e-06,
+      "loss": 0.7797,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6169844020797227,
+      "grad_norm": 3.368485087915924,
+      "learning_rate": 3.3846094885949787e-06,
+      "loss": 0.6292,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6173310225303293,
+      "grad_norm": 3.894211403816005,
+      "learning_rate": 3.3792975767879055e-06,
+      "loss": 0.6268,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6176776429809359,
+      "grad_norm": 2.6780222866627734,
+      "learning_rate": 3.3739877081652194e-06,
+      "loss": 0.7566,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6180242634315425,
+      "grad_norm": 13.704335786373438,
+      "learning_rate": 3.368679889420956e-06,
+      "loss": 0.7443,
+      "step": 1783
+    },
+    {
+      "epoch": 0.618370883882149,
+      "grad_norm": 3.0388208200008857,
+      "learning_rate": 3.363374127246564e-06,
+      "loss": 0.792,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6187175043327556,
+      "grad_norm": 16.680834807904237,
+      "learning_rate": 3.3580704283309044e-06,
+      "loss": 0.7138,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6190641247833623,
+      "grad_norm": 8.426417875846859,
+      "learning_rate": 3.3527687993602367e-06,
+      "loss": 0.8026,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6194107452339688,
+      "grad_norm": 9.022058107236731,
+      "learning_rate": 3.347469247018208e-06,
+      "loss": 0.6938,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6197573656845754,
+      "grad_norm": 3.4665970346391117,
+      "learning_rate": 3.342171777985851e-06,
+      "loss": 0.622,
+      "step": 1788
+    },
+    {
+      "epoch": 0.620103986135182,
+      "grad_norm": 4.667431966591001,
+      "learning_rate": 3.3368763989415664e-06,
+      "loss": 0.7368,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6204506065857885,
+      "grad_norm": 7.304269893894164,
+      "learning_rate": 3.331583116561127e-06,
+      "loss": 0.7911,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6207972270363952,
+      "grad_norm": 5.2319818130232045,
+      "learning_rate": 3.326291937517657e-06,
+      "loss": 0.7944,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6211438474870017,
+      "grad_norm": 3.203675843481919,
+      "learning_rate": 3.321002868481631e-06,
+      "loss": 0.7307,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6214904679376083,
+      "grad_norm": 3.267855693876197,
+      "learning_rate": 3.3157159161208653e-06,
+      "loss": 0.7852,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6218370883882149,
+      "grad_norm": 2.737931971112715,
+      "learning_rate": 3.3104310871005063e-06,
+      "loss": 0.7077,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6221837088388215,
+      "grad_norm": 3.432907609072897,
+      "learning_rate": 3.3051483880830216e-06,
+      "loss": 0.8019,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6225303292894281,
+      "grad_norm": 4.2449585505617495,
+      "learning_rate": 3.299867825728196e-06,
+      "loss": 0.7327,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6228769497400347,
+      "grad_norm": 2.6670138879159895,
+      "learning_rate": 3.2945894066931187e-06,
+      "loss": 0.7522,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6232235701906412,
+      "grad_norm": 3.0462685252845723,
+      "learning_rate": 3.2893131376321798e-06,
+      "loss": 0.7363,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6235701906412479,
+      "grad_norm": 3.2529028389610395,
+      "learning_rate": 3.2840390251970574e-06,
+      "loss": 0.8282,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6239168110918544,
+      "grad_norm": 4.679698824892218,
+      "learning_rate": 3.2787670760367086e-06,
+      "loss": 0.6927,
+      "step": 1800
+    },
+    {
+      "epoch": 0.624263431542461,
+      "grad_norm": 3.809977899385287,
+      "learning_rate": 3.2734972967973668e-06,
+      "loss": 0.7521,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6246100519930676,
+      "grad_norm": 4.22278146888325,
+      "learning_rate": 3.2682296941225265e-06,
+      "loss": 0.6444,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6249566724436741,
+      "grad_norm": 3.707217459622957,
+      "learning_rate": 3.262964274652943e-06,
+      "loss": 0.7939,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6253032928942808,
+      "grad_norm": 4.867282951250001,
+      "learning_rate": 3.2577010450266168e-06,
+      "loss": 0.645,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6256499133448874,
+      "grad_norm": 3.9570339634802525,
+      "learning_rate": 3.2524400118787835e-06,
+      "loss": 0.7252,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6259965337954939,
+      "grad_norm": 3.923364358604192,
+      "learning_rate": 3.2471811818419163e-06,
+      "loss": 0.7008,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6263431542461005,
+      "grad_norm": 3.2597997842991275,
+      "learning_rate": 3.2419245615457068e-06,
+      "loss": 0.7289,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6266897746967071,
+      "grad_norm": 3.3738132031071926,
+      "learning_rate": 3.2366701576170634e-06,
+      "loss": 0.7613,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6270363951473137,
+      "grad_norm": 5.08892080612494,
+      "learning_rate": 3.231417976680099e-06,
+      "loss": 0.7879,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6273830155979203,
+      "grad_norm": 7.957970240313329,
+      "learning_rate": 3.226168025356123e-06,
+      "loss": 0.7928,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6277296360485268,
+      "grad_norm": 4.3750048029147175,
+      "learning_rate": 3.2209203102636353e-06,
+      "loss": 0.8665,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6280762564991335,
+      "grad_norm": 4.169558318895332,
+      "learning_rate": 3.2156748380183166e-06,
+      "loss": 0.7468,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6284228769497401,
+      "grad_norm": 10.004877007531757,
+      "learning_rate": 3.2104316152330207e-06,
+      "loss": 0.7591,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6287694974003466,
+      "grad_norm": 2.88150787875958,
+      "learning_rate": 3.205190648517767e-06,
+      "loss": 0.7353,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6291161178509532,
+      "grad_norm": 4.993209714413688,
+      "learning_rate": 3.1999519444797226e-06,
+      "loss": 0.6376,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6294627383015597,
+      "grad_norm": 3.940930128164963,
+      "learning_rate": 3.194715509723213e-06,
+      "loss": 0.7469,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6298093587521664,
+      "grad_norm": 3.973238047538816,
+      "learning_rate": 3.1894813508496956e-06,
+      "loss": 0.782,
+      "step": 1817
+    },
+    {
+      "epoch": 0.630155979202773,
+      "grad_norm": 4.126224477211479,
+      "learning_rate": 3.1842494744577625e-06,
+      "loss": 0.7805,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6305025996533795,
+      "grad_norm": 4.691186874546647,
+      "learning_rate": 3.1790198871431268e-06,
+      "loss": 0.8445,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6308492201039861,
+      "grad_norm": 2.7977984948062273,
+      "learning_rate": 3.173792595498613e-06,
+      "loss": 0.7628,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6311958405545928,
+      "grad_norm": 7.952369082157315,
+      "learning_rate": 3.168567606114157e-06,
+      "loss": 0.7791,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6315424610051993,
+      "grad_norm": 5.1913820578793946,
+      "learning_rate": 3.163344925576787e-06,
+      "loss": 0.7823,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6318890814558059,
+      "grad_norm": 4.850406375014362,
+      "learning_rate": 3.158124560470627e-06,
+      "loss": 0.7213,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6322357019064124,
+      "grad_norm": 4.551109689404969,
+      "learning_rate": 3.152906517376877e-06,
+      "loss": 0.6698,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6325823223570191,
+      "grad_norm": 3.911814805359897,
+      "learning_rate": 3.1476908028738066e-06,
+      "loss": 0.7361,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6329289428076257,
+      "grad_norm": 4.528228876538064,
+      "learning_rate": 3.1424774235367584e-06,
+      "loss": 0.7575,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6332755632582322,
+      "grad_norm": 3.2821431080419643,
+      "learning_rate": 3.1372663859381245e-06,
+      "loss": 0.6957,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6336221837088388,
+      "grad_norm": 3.2296724936421315,
+      "learning_rate": 3.1320576966473482e-06,
+      "loss": 0.7735,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6339688041594455,
+      "grad_norm": 3.9838487442746455,
+      "learning_rate": 3.1268513622309093e-06,
+      "loss": 0.7444,
+      "step": 1829
+    },
+    {
+      "epoch": 0.634315424610052,
+      "grad_norm": 3.404383218404054,
+      "learning_rate": 3.1216473892523245e-06,
+      "loss": 0.7915,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6346620450606586,
+      "grad_norm": 3.7281851892411524,
+      "learning_rate": 3.116445784272125e-06,
+      "loss": 0.7752,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6350086655112651,
+      "grad_norm": 3.7497110935097364,
+      "learning_rate": 3.111246553847863e-06,
+      "loss": 0.686,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6353552859618717,
+      "grad_norm": 4.356522963685404,
+      "learning_rate": 3.1060497045340965e-06,
+      "loss": 0.7485,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6357019064124784,
+      "grad_norm": 6.155291837463046,
+      "learning_rate": 3.1008552428823795e-06,
+      "loss": 0.7512,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6360485268630849,
+      "grad_norm": 8.725054968419638,
+      "learning_rate": 3.095663175441259e-06,
+      "loss": 0.7987,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6363951473136915,
+      "grad_norm": 3.195994461602024,
+      "learning_rate": 3.0904735087562597e-06,
+      "loss": 0.6677,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6367417677642981,
+      "grad_norm": 4.928717574085122,
+      "learning_rate": 3.0852862493698823e-06,
+      "loss": 0.7026,
+      "step": 1837
+    },
+    {
+      "epoch": 0.6370883882149047,
+      "grad_norm": 5.506018062574888,
+      "learning_rate": 3.080101403821592e-06,
+      "loss": 0.7356,
+      "step": 1838
+    },
+    {
+      "epoch": 0.6374350086655113,
+      "grad_norm": 2.835775872785342,
+      "learning_rate": 3.0749189786478117e-06,
+      "loss": 0.7175,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6377816291161178,
+      "grad_norm": 4.900387892203197,
+      "learning_rate": 3.069738980381913e-06,
+      "loss": 0.8768,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6381282495667244,
+      "grad_norm": 3.226325796899993,
+      "learning_rate": 3.064561415554206e-06,
+      "loss": 0.7533,
+      "step": 1841
+    },
+    {
+      "epoch": 0.638474870017331,
+      "grad_norm": 2.889409932432054,
+      "learning_rate": 3.0593862906919343e-06,
+      "loss": 0.7001,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6388214904679376,
+      "grad_norm": 11.394423791317646,
+      "learning_rate": 3.0542136123192668e-06,
+      "loss": 0.7012,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6391681109185442,
+      "grad_norm": 8.380772586378333,
+      "learning_rate": 3.0490433869572857e-06,
+      "loss": 0.68,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6395147313691508,
+      "grad_norm": 3.1445911276821197,
+      "learning_rate": 3.0438756211239838e-06,
+      "loss": 0.8561,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6398613518197573,
+      "grad_norm": 2.879085710512253,
+      "learning_rate": 3.038710321334249e-06,
+      "loss": 0.7719,
+      "step": 1846
+    },
+    {
+      "epoch": 0.640207972270364,
+      "grad_norm": 3.7148805094349924,
+      "learning_rate": 3.033547494099864e-06,
+      "loss": 0.6761,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6405545927209705,
+      "grad_norm": 4.7507850864268395,
+      "learning_rate": 3.0283871459294924e-06,
+      "loss": 0.7922,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6409012131715771,
+      "grad_norm": 7.3256149539552045,
+      "learning_rate": 3.0232292833286735e-06,
+      "loss": 0.8471,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6412478336221837,
+      "grad_norm": 14.734513988307986,
+      "learning_rate": 3.018073912799815e-06,
+      "loss": 0.7526,
+      "step": 1850
+    },
+    {
+      "epoch": 0.6415944540727903,
+      "grad_norm": 4.130868307527957,
+      "learning_rate": 3.0129210408421766e-06,
+      "loss": 0.8711,
+      "step": 1851
+    },
+    {
+      "epoch": 0.6419410745233969,
+      "grad_norm": 6.448501671401746,
+      "learning_rate": 3.0077706739518732e-06,
+      "loss": 0.7037,
+      "step": 1852
+    },
+    {
+      "epoch": 0.6422876949740035,
+      "grad_norm": 3.4124308425742234,
+      "learning_rate": 3.002622818621863e-06,
+      "loss": 0.7365,
+      "step": 1853
+    },
+    {
+      "epoch": 0.64263431542461,
+      "grad_norm": 6.122614803379732,
+      "learning_rate": 2.9974774813419337e-06,
+      "loss": 0.7048,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6429809358752167,
+      "grad_norm": 2.835950352293005,
+      "learning_rate": 2.992334668598702e-06,
+      "loss": 0.7406,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6433275563258232,
+      "grad_norm": 16.676155429625044,
+      "learning_rate": 2.987194386875599e-06,
+      "loss": 0.6766,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6436741767764298,
+      "grad_norm": 4.376409213384368,
+      "learning_rate": 2.9820566426528653e-06,
+      "loss": 0.7399,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6440207972270364,
+      "grad_norm": 7.171286024538215,
+      "learning_rate": 2.9769214424075456e-06,
+      "loss": 0.8249,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6443674176776429,
+      "grad_norm": 3.49269680394789,
+      "learning_rate": 2.9717887926134747e-06,
+      "loss": 0.9029,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6447140381282496,
+      "grad_norm": 6.625889009177414,
+      "learning_rate": 2.9666586997412734e-06,
+      "loss": 0.7778,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6450606585788562,
+      "grad_norm": 4.563446564831484,
+      "learning_rate": 2.961531170258337e-06,
+      "loss": 0.7906,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6454072790294627,
+      "grad_norm": 3.184821099038022,
+      "learning_rate": 2.9564062106288314e-06,
+      "loss": 0.7292,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6457538994800693,
+      "grad_norm": 5.18085230879075,
+      "learning_rate": 2.951283827313681e-06,
+      "loss": 0.7679,
+      "step": 1863
+    },
+    {
+      "epoch": 0.646100519930676,
+      "grad_norm": 4.924043128849779,
+      "learning_rate": 2.946164026770566e-06,
+      "loss": 0.895,
+      "step": 1864
+    },
+    {
+      "epoch": 0.6464471403812825,
+      "grad_norm": 2.7369593782313375,
+      "learning_rate": 2.9410468154539063e-06,
+      "loss": 0.7651,
+      "step": 1865
+    },
+    {
+      "epoch": 0.6467937608318891,
+      "grad_norm": 2.859508707592912,
+      "learning_rate": 2.9359321998148602e-06,
+      "loss": 0.6336,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6471403812824956,
+      "grad_norm": 3.2444307639634906,
+      "learning_rate": 2.9308201863013113e-06,
+      "loss": 0.8516,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6474870017331023,
+      "grad_norm": 2.653503726244898,
+      "learning_rate": 2.9257107813578645e-06,
+      "loss": 0.6733,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6478336221837089,
+      "grad_norm": 4.999572364871843,
+      "learning_rate": 2.9206039914258354e-06,
+      "loss": 0.7203,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6481802426343154,
+      "grad_norm": 4.608728531520068,
+      "learning_rate": 2.915499822943244e-06,
+      "loss": 0.8428,
+      "step": 1870
+    },
+    {
+      "epoch": 0.648526863084922,
+      "grad_norm": 2.8850795089854393,
+      "learning_rate": 2.910398282344804e-06,
+      "loss": 0.799,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6488734835355287,
+      "grad_norm": 9.446120323921834,
+      "learning_rate": 2.9052993760619188e-06,
+      "loss": 0.7651,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6492201039861352,
+      "grad_norm": 4.745926137198915,
+      "learning_rate": 2.900203110522666e-06,
+      "loss": 0.8567,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6495667244367418,
+      "grad_norm": 4.555570097855874,
+      "learning_rate": 2.8951094921518e-06,
+      "loss": 0.7347,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6499133448873483,
+      "grad_norm": 4.710145737540706,
+      "learning_rate": 2.8900185273707326e-06,
+      "loss": 0.7151,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6502599653379549,
+      "grad_norm": 2.724900589829276,
+      "learning_rate": 2.8849302225975363e-06,
+      "loss": 0.7126,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6506065857885616,
+      "grad_norm": 4.368127106443266,
+      "learning_rate": 2.8798445842469214e-06,
+      "loss": 0.727,
+      "step": 1877
+    },
+    {
+      "epoch": 0.6509532062391681,
+      "grad_norm": 3.896391906750703,
+      "learning_rate": 2.874761618730243e-06,
+      "loss": 0.7253,
+      "step": 1878
+    },
+    {
+      "epoch": 0.6512998266897747,
+      "grad_norm": 3.990280604386818,
+      "learning_rate": 2.8696813324554884e-06,
+      "loss": 0.759,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6516464471403813,
+      "grad_norm": 4.2713625873326855,
+      "learning_rate": 2.864603731827263e-06,
+      "loss": 0.7187,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6519930675909879,
+      "grad_norm": 3.843678736142174,
+      "learning_rate": 2.859528823246789e-06,
+      "loss": 0.7428,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6523396880415945,
+      "grad_norm": 3.053894728473904,
+      "learning_rate": 2.85445661311189e-06,
+      "loss": 0.6576,
+      "step": 1882
+    },
+    {
+      "epoch": 0.652686308492201,
+      "grad_norm": 3.5549339599311454,
+      "learning_rate": 2.849387107816994e-06,
+      "loss": 0.7482,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6530329289428076,
+      "grad_norm": 6.326070735544303,
+      "learning_rate": 2.844320313753115e-06,
+      "loss": 0.7435,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6533795493934142,
+      "grad_norm": 4.282162437507706,
+      "learning_rate": 2.839256237307849e-06,
+      "loss": 0.7236,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6537261698440208,
+      "grad_norm": 4.39001477154441,
+      "learning_rate": 2.8341948848653737e-06,
+      "loss": 0.7616,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6540727902946274,
+      "grad_norm": 2.6980881974177953,
+      "learning_rate": 2.82913626280642e-06,
+      "loss": 0.7039,
+      "step": 1887
+    },
+    {
+      "epoch": 0.654419410745234,
+      "grad_norm": 3.550269716392352,
+      "learning_rate": 2.8240803775082846e-06,
+      "loss": 0.7157,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6547660311958405,
+      "grad_norm": 3.0687659631227318,
+      "learning_rate": 2.819027235344813e-06,
+      "loss": 0.8941,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6551126516464472,
+      "grad_norm": 2.9204631549111673,
+      "learning_rate": 2.813976842686391e-06,
+      "loss": 0.6424,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6554592720970537,
+      "grad_norm": 7.867062781801362,
+      "learning_rate": 2.8089292058999406e-06,
+      "loss": 0.7728,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6558058925476603,
+      "grad_norm": 4.694599681380446,
+      "learning_rate": 2.803884331348906e-06,
+      "loss": 0.7856,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6561525129982669,
+      "grad_norm": 7.264441854664844,
+      "learning_rate": 2.7988422253932536e-06,
+      "loss": 0.7488,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6564991334488735,
+      "grad_norm": 3.412109309780588,
+      "learning_rate": 2.7938028943894547e-06,
+      "loss": 0.7065,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6568457538994801,
+      "grad_norm": 3.407506635970828,
+      "learning_rate": 2.7887663446904858e-06,
+      "loss": 0.7742,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6571923743500867,
+      "grad_norm": 3.0128035277203575,
+      "learning_rate": 2.783732582645817e-06,
+      "loss": 0.8764,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6575389948006932,
+      "grad_norm": 6.4431653944358045,
+      "learning_rate": 2.7787016146013996e-06,
+      "loss": 0.7505,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6578856152512998,
+      "grad_norm": 3.160033488529259,
+      "learning_rate": 2.7736734468996657e-06,
+      "loss": 0.7099,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6582322357019064,
+      "grad_norm": 2.4261379984760167,
+      "learning_rate": 2.768648085879521e-06,
+      "loss": 0.6355,
+      "step": 1899
+    },
+    {
+      "epoch": 0.658578856152513,
+      "grad_norm": 2.5462379384316076,
+      "learning_rate": 2.763625537876328e-06,
+      "loss": 0.764,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6589254766031196,
+      "grad_norm": 2.9483368948695556,
+      "learning_rate": 2.7586058092219027e-06,
+      "loss": 0.7167,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6592720970537261,
+      "grad_norm": 18.27098222069936,
+      "learning_rate": 2.753588906244511e-06,
+      "loss": 0.7271,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6596187175043328,
+      "grad_norm": 4.597561974647824,
+      "learning_rate": 2.7485748352688497e-06,
+      "loss": 0.6223,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6599653379549394,
+      "grad_norm": 6.761920517439788,
+      "learning_rate": 2.7435636026160506e-06,
+      "loss": 0.6943,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6603119584055459,
+      "grad_norm": 4.90317158798711,
+      "learning_rate": 2.7385552146036663e-06,
+      "loss": 0.722,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6606585788561525,
+      "grad_norm": 10.045603240604777,
+      "learning_rate": 2.7335496775456614e-06,
+      "loss": 0.7091,
+      "step": 1906
+    },
+    {
+      "epoch": 0.661005199306759,
+      "grad_norm": 5.672847102332802,
+      "learning_rate": 2.7285469977524138e-06,
+      "loss": 0.7136,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6613518197573657,
+      "grad_norm": 3.487399433020322,
+      "learning_rate": 2.7235471815306867e-06,
+      "loss": 0.8013,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6616984402079723,
+      "grad_norm": 3.4091766919516195,
+      "learning_rate": 2.7185502351836424e-06,
+      "loss": 0.7054,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6620450606585788,
+      "grad_norm": 3.4193949130392367,
+      "learning_rate": 2.7135561650108224e-06,
+      "loss": 0.825,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6623916811091854,
+      "grad_norm": 4.309084139855552,
+      "learning_rate": 2.708564977308143e-06,
+      "loss": 0.6645,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6627383015597921,
+      "grad_norm": 8.090610612371654,
+      "learning_rate": 2.7035766783678865e-06,
+      "loss": 0.7295,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6630849220103986,
+      "grad_norm": 5.521516173522502,
+      "learning_rate": 2.6985912744786924e-06,
+      "loss": 0.6779,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6634315424610052,
+      "grad_norm": 3.874508336006056,
+      "learning_rate": 2.693608771925552e-06,
+      "loss": 0.7309,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6637781629116117,
+      "grad_norm": 5.066691078482635,
+      "learning_rate": 2.6886291769897977e-06,
+      "loss": 0.7806,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6641247833622184,
+      "grad_norm": 3.2617449709321624,
+      "learning_rate": 2.6836524959490966e-06,
+      "loss": 0.6756,
+      "step": 1916
+    },
+    {
+      "epoch": 0.664471403812825,
+      "grad_norm": 4.103046789407015,
+      "learning_rate": 2.678678735077446e-06,
+      "loss": 0.8074,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6648180242634315,
+      "grad_norm": 3.7412933380425804,
+      "learning_rate": 2.673707900645149e-06,
+      "loss": 0.8317,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6651646447140381,
+      "grad_norm": 6.772561102171276,
+      "learning_rate": 2.668739998918839e-06,
+      "loss": 0.7593,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6655112651646448,
+      "grad_norm": 2.5486599004645076,
+      "learning_rate": 2.663775036161437e-06,
+      "loss": 0.7031,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6658578856152513,
+      "grad_norm": 2.7691173763522725,
+      "learning_rate": 2.6588130186321655e-06,
+      "loss": 0.7094,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6662045060658579,
+      "grad_norm": 4.446103517306824,
+      "learning_rate": 2.6538539525865326e-06,
+      "loss": 0.7474,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6665511265164644,
+      "grad_norm": 3.4112497927885426,
+      "learning_rate": 2.6488978442763224e-06,
+      "loss": 0.7075,
+      "step": 1923
+    },
+    {
+      "epoch": 0.666897746967071,
+      "grad_norm": 3.7012507123934912,
+      "learning_rate": 2.6439446999495953e-06,
+      "loss": 0.6682,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6672443674176777,
+      "grad_norm": 4.0239420113226165,
+      "learning_rate": 2.6389945258506722e-06,
+      "loss": 0.6023,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6675909878682842,
+      "grad_norm": 5.66307462446542,
+      "learning_rate": 2.634047328220129e-06,
+      "loss": 0.7158,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6679376083188908,
+      "grad_norm": 3.1642582746391885,
+      "learning_rate": 2.629103113294795e-06,
+      "loss": 0.7346,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6682842287694974,
+      "grad_norm": 12.03695484613002,
+      "learning_rate": 2.6241618873077314e-06,
+      "loss": 0.6759,
+      "step": 1928
+    },
+    {
+      "epoch": 0.668630849220104,
+      "grad_norm": 5.540409899180265,
+      "learning_rate": 2.6192236564882333e-06,
+      "loss": 0.6491,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6689774696707106,
+      "grad_norm": 2.993480759710009,
+      "learning_rate": 2.6142884270618234e-06,
+      "loss": 0.7498,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6693240901213171,
+      "grad_norm": 2.970937335672007,
+      "learning_rate": 2.6093562052502375e-06,
+      "loss": 0.6737,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6696707105719237,
+      "grad_norm": 2.683014158609768,
+      "learning_rate": 2.6044269972714207e-06,
+      "loss": 0.7691,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6700173310225304,
+      "grad_norm": 4.027110868955184,
+      "learning_rate": 2.5995008093395197e-06,
+      "loss": 0.676,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6703639514731369,
+      "grad_norm": 3.416247149000949,
+      "learning_rate": 2.594577647664871e-06,
+      "loss": 0.7019,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6707105719237435,
+      "grad_norm": 3.6998971316936484,
+      "learning_rate": 2.5896575184539984e-06,
+      "loss": 0.8289,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6710571923743501,
+      "grad_norm": 3.398955484889919,
+      "learning_rate": 2.5847404279096026e-06,
+      "loss": 0.7016,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6714038128249566,
+      "grad_norm": 3.0457234011620162,
+      "learning_rate": 2.5798263822305527e-06,
+      "loss": 0.8051,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6717504332755633,
+      "grad_norm": 8.758380977851193,
+      "learning_rate": 2.5749153876118802e-06,
+      "loss": 0.7362,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6720970537261698,
+      "grad_norm": 25.621930286682318,
+      "learning_rate": 2.570007450244766e-06,
+      "loss": 0.8348,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6724436741767764,
+      "grad_norm": 8.385777649316836,
+      "learning_rate": 2.5651025763165437e-06,
+      "loss": 0.7798,
+      "step": 1940
+    },
+    {
+      "epoch": 0.672790294627383,
+      "grad_norm": 2.8648541339941,
+      "learning_rate": 2.5602007720106814e-06,
+      "loss": 0.6784,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6731369150779896,
+      "grad_norm": 13.460810596209662,
+      "learning_rate": 2.5553020435067767e-06,
+      "loss": 0.833,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6734835355285962,
+      "grad_norm": 25.016746599886087,
+      "learning_rate": 2.5504063969805517e-06,
+      "loss": 0.6941,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6738301559792028,
+      "grad_norm": 2.777221221080841,
+      "learning_rate": 2.545513838603839e-06,
+      "loss": 0.8468,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6741767764298093,
+      "grad_norm": 3.9933527692333173,
+      "learning_rate": 2.540624374544581e-06,
+      "loss": 0.6394,
+      "step": 1945
+    },
+    {
+      "epoch": 0.674523396880416,
+      "grad_norm": 5.629128577743265,
+      "learning_rate": 2.5357380109668197e-06,
+      "loss": 0.7629,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6748700173310225,
+      "grad_norm": 4.155686379790431,
+      "learning_rate": 2.5308547540306895e-06,
+      "loss": 0.8684,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6752166377816291,
+      "grad_norm": 4.685654356771173,
+      "learning_rate": 2.5259746098924065e-06,
+      "loss": 0.8101,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6755632582322357,
+      "grad_norm": 18.078506162248583,
+      "learning_rate": 2.52109758470426e-06,
+      "loss": 0.8207,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6759098786828422,
+      "grad_norm": 3.1674464596504524,
+      "learning_rate": 2.5162236846146106e-06,
+      "loss": 0.672,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6762564991334489,
+      "grad_norm": 4.385082032022346,
+      "learning_rate": 2.511352915767879e-06,
+      "loss": 0.7607,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6766031195840555,
+      "grad_norm": 3.8031000706249576,
+      "learning_rate": 2.506485284304537e-06,
+      "loss": 0.6529,
+      "step": 1952
+    },
+    {
+      "epoch": 0.676949740034662,
+      "grad_norm": 2.7774211678834106,
+      "learning_rate": 2.501620796361104e-06,
+      "loss": 0.7211,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6772963604852686,
+      "grad_norm": 2.5206760819268683,
+      "learning_rate": 2.4967594580701323e-06,
+      "loss": 0.7288,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6776429809358753,
+      "grad_norm": 3.07393088205366,
+      "learning_rate": 2.491901275560206e-06,
+      "loss": 0.6912,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6779896013864818,
+      "grad_norm": 4.626346991232572,
+      "learning_rate": 2.487046254955931e-06,
+      "loss": 0.6994,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6783362218370884,
+      "grad_norm": 6.880577329477428,
+      "learning_rate": 2.482194402377926e-06,
+      "loss": 0.6685,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6786828422876949,
+      "grad_norm": 8.092085745774645,
+      "learning_rate": 2.4773457239428183e-06,
+      "loss": 0.7666,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6790294627383016,
+      "grad_norm": 2.4759810548980115,
+      "learning_rate": 2.472500225763226e-06,
+      "loss": 0.7195,
+      "step": 1959
+    },
+    {
+      "epoch": 0.6793760831889082,
+      "grad_norm": 3.725106129354273,
+      "learning_rate": 2.4676579139477695e-06,
+      "loss": 0.6293,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6797227036395147,
+      "grad_norm": 5.042814403132438,
+      "learning_rate": 2.4628187946010446e-06,
+      "loss": 0.8967,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6800693240901213,
+      "grad_norm": 3.4845720528256927,
+      "learning_rate": 2.4579828738236227e-06,
+      "loss": 0.7866,
+      "step": 1962
+    },
+    {
+      "epoch": 0.680415944540728,
+      "grad_norm": 2.5346849513973986,
+      "learning_rate": 2.4531501577120477e-06,
+      "loss": 0.736,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6807625649913345,
+      "grad_norm": 6.094988387727664,
+      "learning_rate": 2.448320652358815e-06,
+      "loss": 0.7818,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6811091854419411,
+      "grad_norm": 5.825437957062341,
+      "learning_rate": 2.4434943638523794e-06,
+      "loss": 0.8043,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6814558058925476,
+      "grad_norm": 6.7952521184707315,
+      "learning_rate": 2.4386712982771364e-06,
+      "loss": 0.7146,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6818024263431542,
+      "grad_norm": 4.782467641160564,
+      "learning_rate": 2.4338514617134246e-06,
+      "loss": 0.7317,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6821490467937609,
+      "grad_norm": 9.391660210362323,
+      "learning_rate": 2.4290348602375065e-06,
+      "loss": 0.7476,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6824956672443674,
+      "grad_norm": 2.9858839769799865,
+      "learning_rate": 2.424221499921563e-06,
+      "loss": 0.7154,
+      "step": 1969
+    },
+    {
+      "epoch": 0.682842287694974,
+      "grad_norm": 3.1945524489094748,
+      "learning_rate": 2.419411386833696e-06,
+      "loss": 0.7015,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6831889081455806,
+      "grad_norm": 4.386243707490256,
+      "learning_rate": 2.41460452703791e-06,
+      "loss": 0.7101,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6835355285961872,
+      "grad_norm": 3.915190336987793,
+      "learning_rate": 2.4098009265941084e-06,
+      "loss": 0.7058,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6838821490467938,
+      "grad_norm": 3.619393949561544,
+      "learning_rate": 2.4050005915580875e-06,
+      "loss": 0.7995,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6842287694974003,
+      "grad_norm": 3.6477480791330676,
+      "learning_rate": 2.4002035279815245e-06,
+      "loss": 0.8087,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6845753899480069,
+      "grad_norm": 3.409611819002711,
+      "learning_rate": 2.3954097419119737e-06,
+      "loss": 0.7425,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6849220103986136,
+      "grad_norm": 5.5923561748120845,
+      "learning_rate": 2.390619239392857e-06,
+      "loss": 0.6975,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6852686308492201,
+      "grad_norm": 4.959177789564345,
+      "learning_rate": 2.385832026463457e-06,
+      "loss": 0.6978,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6856152512998267,
+      "grad_norm": 7.801558150777711,
+      "learning_rate": 2.38104810915891e-06,
+      "loss": 0.8247,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6859618717504333,
+      "grad_norm": 4.635019552121384,
+      "learning_rate": 2.376267493510195e-06,
+      "loss": 0.8326,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6863084922010398,
+      "grad_norm": 5.872178606043057,
+      "learning_rate": 2.3714901855441326e-06,
+      "loss": 0.6771,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6866551126516465,
+      "grad_norm": 3.876899633044994,
+      "learning_rate": 2.366716191283369e-06,
+      "loss": 0.686,
+      "step": 1981
+    },
+    {
+      "epoch": 0.687001733102253,
+      "grad_norm": 14.190466187898302,
+      "learning_rate": 2.361945516746377e-06,
+      "loss": 0.8044,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6873483535528596,
+      "grad_norm": 2.95318581803744,
+      "learning_rate": 2.3571781679474425e-06,
+      "loss": 0.7695,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6876949740034662,
+      "grad_norm": 5.862536724541926,
+      "learning_rate": 2.3524141508966598e-06,
+      "loss": 0.7381,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6880415944540728,
+      "grad_norm": 3.1319078923426718,
+      "learning_rate": 2.3476534715999177e-06,
+      "loss": 0.7978,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6883882149046794,
+      "grad_norm": 6.651664806401078,
+      "learning_rate": 2.3428961360589044e-06,
+      "loss": 0.7805,
+      "step": 1986
+    },
+    {
+      "epoch": 0.688734835355286,
+      "grad_norm": 3.6149106577221084,
+      "learning_rate": 2.3381421502710865e-06,
+      "loss": 0.8124,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6890814558058925,
+      "grad_norm": 4.07139535788142,
+      "learning_rate": 2.3333915202297155e-06,
+      "loss": 0.7153,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6894280762564992,
+      "grad_norm": 2.9153628467574095,
+      "learning_rate": 2.3286442519238073e-06,
+      "loss": 0.7156,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6897746967071057,
+      "grad_norm": 6.074504077480088,
+      "learning_rate": 2.323900351338136e-06,
+      "loss": 0.7005,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6901213171577123,
+      "grad_norm": 4.001019905791444,
+      "learning_rate": 2.3191598244532364e-06,
+      "loss": 0.7471,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6904679376083189,
+      "grad_norm": 4.115567118360468,
+      "learning_rate": 2.314422677245388e-06,
+      "loss": 0.7716,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6908145580589254,
+      "grad_norm": 3.5950145998426803,
+      "learning_rate": 2.3096889156866085e-06,
+      "loss": 0.6414,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6911611785095321,
+      "grad_norm": 4.1223375370770405,
+      "learning_rate": 2.3049585457446498e-06,
+      "loss": 0.7301,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6915077989601387,
+      "grad_norm": 8.9932111509883,
+      "learning_rate": 2.3002315733829865e-06,
+      "loss": 0.6991,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6918544194107452,
+      "grad_norm": 2.7306172478955673,
+      "learning_rate": 2.295508004560809e-06,
+      "loss": 0.7512,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6922010398613518,
+      "grad_norm": 8.640021133217823,
+      "learning_rate": 2.290787845233019e-06,
+      "loss": 0.6562,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6925476603119584,
+      "grad_norm": 3.317925409217376,
+      "learning_rate": 2.286071101350219e-06,
+      "loss": 0.6946,
+      "step": 1998
+    },
+    {
+      "epoch": 0.692894280762565,
+      "grad_norm": 5.559691110416209,
+      "learning_rate": 2.2813577788587066e-06,
+      "loss": 0.745,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6932409012131716,
+      "grad_norm": 4.024431869365518,
+      "learning_rate": 2.2766478837004608e-06,
+      "loss": 0.7022,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6935875216637781,
+      "grad_norm": 3.7210482614127813,
+      "learning_rate": 2.271941421813149e-06,
+      "loss": 0.8648,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6939341421143848,
+      "grad_norm": 4.08580176941678,
+      "learning_rate": 2.2672383991301044e-06,
+      "loss": 0.8311,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6942807625649914,
+      "grad_norm": 2.2560036273183273,
+      "learning_rate": 2.262538821580325e-06,
+      "loss": 0.6786,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6946273830155979,
+      "grad_norm": 3.365936210691333,
+      "learning_rate": 2.2578426950884686e-06,
+      "loss": 0.6165,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6949740034662045,
+      "grad_norm": 2.8957632925492587,
+      "learning_rate": 2.253150025574836e-06,
+      "loss": 0.7756,
+      "step": 2005
+    },
+    {
+      "epoch": 0.695320623916811,
+      "grad_norm": 7.169302509044313,
+      "learning_rate": 2.2484608189553758e-06,
+      "loss": 0.7265,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6956672443674177,
+      "grad_norm": 4.603797221862495,
+      "learning_rate": 2.243775081141667e-06,
+      "loss": 0.6726,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6960138648180243,
+      "grad_norm": 3.496775775045965,
+      "learning_rate": 2.239092818040922e-06,
+      "loss": 0.8867,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6963604852686308,
+      "grad_norm": 3.7569439979530164,
+      "learning_rate": 2.2344140355559657e-06,
+      "loss": 0.7384,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6967071057192374,
+      "grad_norm": 3.44025445545069,
+      "learning_rate": 2.22973873958524e-06,
+      "loss": 0.6428,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6970537261698441,
+      "grad_norm": 2.8207069806973735,
+      "learning_rate": 2.225066936022785e-06,
+      "loss": 0.6426,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6974003466204506,
+      "grad_norm": 3.626231134617548,
+      "learning_rate": 2.2203986307582443e-06,
+      "loss": 0.7156,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6977469670710572,
+      "grad_norm": 3.1484561263376527,
+      "learning_rate": 2.2157338296768492e-06,
+      "loss": 0.7745,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6980935875216637,
+      "grad_norm": 3.665009147741684,
+      "learning_rate": 2.2110725386594105e-06,
+      "loss": 0.7,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6984402079722704,
+      "grad_norm": 8.158070008520678,
+      "learning_rate": 2.2064147635823233e-06,
+      "loss": 0.724,
+      "step": 2015
+    },
+    {
+      "epoch": 0.698786828422877,
+      "grad_norm": 2.872530205762222,
+      "learning_rate": 2.2017605103175378e-06,
+      "loss": 0.6938,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6991334488734835,
+      "grad_norm": 13.005729985650191,
+      "learning_rate": 2.1971097847325717e-06,
+      "loss": 0.8055,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6994800693240901,
+      "grad_norm": 43.67479544683842,
+      "learning_rate": 2.1924625926904936e-06,
+      "loss": 0.6425,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6998266897746968,
+      "grad_norm": 4.856910842961655,
+      "learning_rate": 2.187818940049918e-06,
+      "loss": 0.676,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7001733102253033,
+      "grad_norm": 4.055234375097212,
+      "learning_rate": 2.183178832664997e-06,
+      "loss": 0.7298,
+      "step": 2020
+    },
+    {
+      "epoch": 0.7005199306759099,
+      "grad_norm": 3.2507292320717647,
+      "learning_rate": 2.1785422763854136e-06,
+      "loss": 0.6658,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7008665511265164,
+      "grad_norm": 2.881657916523818,
+      "learning_rate": 2.1739092770563736e-06,
+      "loss": 0.7533,
+      "step": 2022
+    },
+    {
+      "epoch": 0.701213171577123,
+      "grad_norm": 4.1629073629870526,
+      "learning_rate": 2.1692798405185983e-06,
+      "loss": 0.6916,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7015597920277297,
+      "grad_norm": 3.46389930872307,
+      "learning_rate": 2.1646539726083184e-06,
+      "loss": 0.7528,
+      "step": 2024
+    },
+    {
+      "epoch": 0.7019064124783362,
+      "grad_norm": 7.597320557861795,
+      "learning_rate": 2.160031679157267e-06,
+      "loss": 0.7934,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7022530329289428,
+      "grad_norm": 5.062153772678747,
+      "learning_rate": 2.1554129659926647e-06,
+      "loss": 0.6672,
+      "step": 2026
+    },
+    {
+      "epoch": 0.7025996533795494,
+      "grad_norm": 2.3371923859432293,
+      "learning_rate": 2.150797838937224e-06,
+      "loss": 0.7591,
+      "step": 2027
+    },
+    {
+      "epoch": 0.702946273830156,
+      "grad_norm": 4.019097185436826,
+      "learning_rate": 2.1461863038091395e-06,
+      "loss": 0.7246,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7032928942807626,
+      "grad_norm": 5.9280440717708185,
+      "learning_rate": 2.14157836642207e-06,
+      "loss": 0.7456,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7036395147313691,
+      "grad_norm": 4.470033962809355,
+      "learning_rate": 2.136974032585146e-06,
+      "loss": 0.7778,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7039861351819757,
+      "grad_norm": 4.14627147011323,
+      "learning_rate": 2.1323733081029457e-06,
+      "loss": 0.7534,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7043327556325824,
+      "grad_norm": 10.075220705190715,
+      "learning_rate": 2.1277761987755057e-06,
+      "loss": 0.938,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7046793760831889,
+      "grad_norm": 2.990542128979283,
+      "learning_rate": 2.123182710398302e-06,
+      "loss": 0.8214,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7050259965337955,
+      "grad_norm": 3.277574141766039,
+      "learning_rate": 2.118592848762245e-06,
+      "loss": 0.7666,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7053726169844021,
+      "grad_norm": 4.100161805071425,
+      "learning_rate": 2.1140066196536748e-06,
+      "loss": 0.6983,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7057192374350086,
+      "grad_norm": 3.239701403616387,
+      "learning_rate": 2.10942402885435e-06,
+      "loss": 0.8944,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7060658578856153,
+      "grad_norm": 3.423321697410569,
+      "learning_rate": 2.1048450821414435e-06,
+      "loss": 0.7395,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7064124783362218,
+      "grad_norm": 5.845612668507078,
+      "learning_rate": 2.1002697852875342e-06,
+      "loss": 0.7441,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7067590987868284,
+      "grad_norm": 5.4283369052763755,
+      "learning_rate": 2.0956981440605994e-06,
+      "loss": 0.693,
+      "step": 2039
+    },
+    {
+      "epoch": 0.707105719237435,
+      "grad_norm": 4.799035768498359,
+      "learning_rate": 2.091130164224009e-06,
+      "loss": 0.7683,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7074523396880416,
+      "grad_norm": 5.688713972927192,
+      "learning_rate": 2.0865658515365144e-06,
+      "loss": 0.6849,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7077989601386482,
+      "grad_norm": 2.5563878247577203,
+      "learning_rate": 2.0820052117522475e-06,
+      "loss": 0.8415,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7081455805892548,
+      "grad_norm": 7.117469989051123,
+      "learning_rate": 2.0774482506207067e-06,
+      "loss": 0.6373,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7084922010398613,
+      "grad_norm": 16.320667001403596,
+      "learning_rate": 2.0728949738867537e-06,
+      "loss": 0.7676,
+      "step": 2044
+    },
+    {
+      "epoch": 0.708838821490468,
+      "grad_norm": 3.406200599366482,
+      "learning_rate": 2.0683453872906074e-06,
+      "loss": 0.782,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7091854419410745,
+      "grad_norm": 2.9408635587785126,
+      "learning_rate": 2.0637994965678326e-06,
+      "loss": 0.8226,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7095320623916811,
+      "grad_norm": 10.643190622375753,
+      "learning_rate": 2.0592573074493334e-06,
+      "loss": 0.7566,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7098786828422877,
+      "grad_norm": 2.3666211639540817,
+      "learning_rate": 2.0547188256613476e-06,
+      "loss": 0.6814,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7102253032928942,
+      "grad_norm": 3.968025931498454,
+      "learning_rate": 2.0501840569254454e-06,
+      "loss": 0.7624,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7105719237435009,
+      "grad_norm": 3.181203894459752,
+      "learning_rate": 2.045653006958509e-06,
+      "loss": 0.7638,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7109185441941075,
+      "grad_norm": 2.4379352274727513,
+      "learning_rate": 2.0411256814727364e-06,
+      "loss": 0.8405,
+      "step": 2051
+    },
+    {
+      "epoch": 0.711265164644714,
+      "grad_norm": 3.2695500130210062,
+      "learning_rate": 2.0366020861756258e-06,
+      "loss": 0.796,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7116117850953206,
+      "grad_norm": 3.2158829259077715,
+      "learning_rate": 2.0320822267699768e-06,
+      "loss": 0.7329,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7119584055459273,
+      "grad_norm": 2.514404475993761,
+      "learning_rate": 2.0275661089538783e-06,
+      "loss": 0.7239,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7123050259965338,
+      "grad_norm": 3.991311307676216,
+      "learning_rate": 2.0230537384207004e-06,
+      "loss": 0.5926,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7126516464471404,
+      "grad_norm": 5.179531693976405,
+      "learning_rate": 2.018545120859096e-06,
+      "loss": 0.8398,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7129982668977469,
+      "grad_norm": 2.8161740363777192,
+      "learning_rate": 2.014040261952976e-06,
+      "loss": 0.7436,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7133448873483536,
+      "grad_norm": 5.58671457873459,
+      "learning_rate": 2.00953916738152e-06,
+      "loss": 0.8558,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7136915077989602,
+      "grad_norm": 64.27408379172164,
+      "learning_rate": 2.0050418428191607e-06,
+      "loss": 0.6861,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7140381282495667,
+      "grad_norm": 2.716195170636954,
+      "learning_rate": 2.0005482939355774e-06,
+      "loss": 0.8415,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7143847487001733,
+      "grad_norm": 3.224753021865301,
+      "learning_rate": 1.9960585263956893e-06,
+      "loss": 0.7456,
+      "step": 2061
+    },
+    {
+      "epoch": 0.71473136915078,
+      "grad_norm": 5.061724896238352,
+      "learning_rate": 1.9915725458596495e-06,
+      "loss": 0.6897,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7150779896013865,
+      "grad_norm": 4.041100940583403,
+      "learning_rate": 1.987090357982836e-06,
+      "loss": 0.85,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7154246100519931,
+      "grad_norm": 4.449200538131579,
+      "learning_rate": 1.9826119684158463e-06,
+      "loss": 0.7197,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7157712305025996,
+      "grad_norm": 11.021600865347017,
+      "learning_rate": 1.9781373828044882e-06,
+      "loss": 0.7104,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7161178509532062,
+      "grad_norm": 6.381448029282267,
+      "learning_rate": 1.973666606789777e-06,
+      "loss": 0.6633,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7164644714038129,
+      "grad_norm": 2.9051942504305766,
+      "learning_rate": 1.9691996460079192e-06,
+      "loss": 0.7312,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7168110918544194,
+      "grad_norm": 12.178796815326656,
+      "learning_rate": 1.9647365060903162e-06,
+      "loss": 0.8066,
+      "step": 2068
+    },
+    {
+      "epoch": 0.717157712305026,
+      "grad_norm": 7.170378919390571,
+      "learning_rate": 1.9602771926635537e-06,
+      "loss": 0.8202,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7175043327556326,
+      "grad_norm": 4.797961015526305,
+      "learning_rate": 1.955821711349391e-06,
+      "loss": 0.7231,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7178509532062392,
+      "grad_norm": 3.5783904626490335,
+      "learning_rate": 1.951370067764758e-06,
+      "loss": 0.7571,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7181975736568458,
+      "grad_norm": 4.052840202833268,
+      "learning_rate": 1.946922267521741e-06,
+      "loss": 0.7038,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7185441941074523,
+      "grad_norm": 3.1530034584159203,
+      "learning_rate": 1.9424783162275886e-06,
+      "loss": 0.6619,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7188908145580589,
+      "grad_norm": 2.6398614049310125,
+      "learning_rate": 1.938038219484692e-06,
+      "loss": 0.7262,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7192374350086655,
+      "grad_norm": 8.57081025744907,
+      "learning_rate": 1.9336019828905846e-06,
+      "loss": 0.9426,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7195840554592721,
+      "grad_norm": 5.503740480148244,
+      "learning_rate": 1.9291696120379377e-06,
+      "loss": 0.744,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7199306759098787,
+      "grad_norm": 5.142698714734999,
+      "learning_rate": 1.9247411125145404e-06,
+      "loss": 0.7304,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7202772963604853,
+      "grad_norm": 6.310919685738828,
+      "learning_rate": 1.9203164899033078e-06,
+      "loss": 0.6741,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7206239168110918,
+      "grad_norm": 2.873779922835794,
+      "learning_rate": 1.9158957497822657e-06,
+      "loss": 0.7377,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7209705372616985,
+      "grad_norm": 3.663366629657131,
+      "learning_rate": 1.911478897724545e-06,
+      "loss": 0.7648,
+      "step": 2080
+    },
+    {
+      "epoch": 0.721317157712305,
+      "grad_norm": 6.911136705492137,
+      "learning_rate": 1.9070659392983753e-06,
+      "loss": 0.7175,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7216637781629116,
+      "grad_norm": 4.2547772996855135,
+      "learning_rate": 1.902656880067078e-06,
+      "loss": 0.7197,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7220103986135182,
+      "grad_norm": 8.788026112800283,
+      "learning_rate": 1.8982517255890576e-06,
+      "loss": 0.7887,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7223570190641248,
+      "grad_norm": 5.664709709655114,
+      "learning_rate": 1.8938504814177971e-06,
+      "loss": 0.8109,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7227036395147314,
+      "grad_norm": 6.198558359677846,
+      "learning_rate": 1.88945315310185e-06,
+      "loss": 0.7233,
+      "step": 2085
+    },
+    {
+      "epoch": 0.723050259965338,
+      "grad_norm": 7.048348233076811,
+      "learning_rate": 1.8850597461848324e-06,
+      "loss": 0.7457,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7233968804159445,
+      "grad_norm": 3.4796514541388905,
+      "learning_rate": 1.8806702662054188e-06,
+      "loss": 0.7636,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7237435008665511,
+      "grad_norm": 4.047768783119565,
+      "learning_rate": 1.876284718697326e-06,
+      "loss": 0.739,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7240901213171577,
+      "grad_norm": 3.2986459330334337,
+      "learning_rate": 1.8719031091893241e-06,
+      "loss": 0.6536,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7244367417677643,
+      "grad_norm": 4.273987967249043,
+      "learning_rate": 1.8675254432052115e-06,
+      "loss": 0.5964,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7247833622183709,
+      "grad_norm": 4.930024743308132,
+      "learning_rate": 1.863151726263816e-06,
+      "loss": 0.7237,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7251299826689774,
+      "grad_norm": 3.7484754233573585,
+      "learning_rate": 1.8587819638789907e-06,
+      "loss": 0.7434,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7254766031195841,
+      "grad_norm": 3.440103445174199,
+      "learning_rate": 1.8544161615595951e-06,
+      "loss": 0.7173,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7258232235701907,
+      "grad_norm": 2.5553415244791813,
+      "learning_rate": 1.850054324809505e-06,
+      "loss": 0.7244,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7261698440207972,
+      "grad_norm": 3.2177865126823506,
+      "learning_rate": 1.8456964591275923e-06,
+      "loss": 0.6833,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7265164644714038,
+      "grad_norm": 10.499928624402457,
+      "learning_rate": 1.8413425700077214e-06,
+      "loss": 0.7401,
+      "step": 2096
+    },
+    {
+      "epoch": 0.7268630849220103,
+      "grad_norm": 2.9051930264855894,
+      "learning_rate": 1.836992662938753e-06,
+      "loss": 0.7109,
+      "step": 2097
+    },
+    {
+      "epoch": 0.727209705372617,
+      "grad_norm": 2.9643792821671653,
+      "learning_rate": 1.8326467434045136e-06,
+      "loss": 0.7742,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7275563258232236,
+      "grad_norm": 3.0357759170922525,
+      "learning_rate": 1.828304816883812e-06,
+      "loss": 0.7179,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7279029462738301,
+      "grad_norm": 6.149430581407018,
+      "learning_rate": 1.8239668888504208e-06,
+      "loss": 0.7558,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7282495667244367,
+      "grad_norm": 9.739613738558152,
+      "learning_rate": 1.8196329647730716e-06,
+      "loss": 0.7574,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7285961871750434,
+      "grad_norm": 6.176613081593087,
+      "learning_rate": 1.815303050115449e-06,
+      "loss": 0.6253,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7289428076256499,
+      "grad_norm": 2.836321653780645,
+      "learning_rate": 1.8109771503361817e-06,
+      "loss": 0.859,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7292894280762565,
+      "grad_norm": 2.94382055985651,
+      "learning_rate": 1.8066552708888384e-06,
+      "loss": 0.8181,
+      "step": 2104
+    },
+    {
+      "epoch": 0.729636048526863,
+      "grad_norm": 127.60608715950802,
+      "learning_rate": 1.8023374172219188e-06,
+      "loss": 0.5806,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7299826689774697,
+      "grad_norm": 2.5627239736162672,
+      "learning_rate": 1.7980235947788466e-06,
+      "loss": 0.6681,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7303292894280763,
+      "grad_norm": 3.8308989800958857,
+      "learning_rate": 1.7937138089979662e-06,
+      "loss": 0.7636,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7306759098786828,
+      "grad_norm": 7.535878544489929,
+      "learning_rate": 1.789408065312528e-06,
+      "loss": 0.8211,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7310225303292894,
+      "grad_norm": 3.538223864065933,
+      "learning_rate": 1.7851063691506893e-06,
+      "loss": 0.6537,
+      "step": 2109
+    },
+    {
+      "epoch": 0.7313691507798961,
+      "grad_norm": 6.2809689299436755,
+      "learning_rate": 1.7808087259355085e-06,
+      "loss": 0.7597,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7317157712305026,
+      "grad_norm": 39.63433020619583,
+      "learning_rate": 1.7765151410849297e-06,
+      "loss": 0.7373,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7320623916811092,
+      "grad_norm": 3.6817070271110595,
+      "learning_rate": 1.7722256200117832e-06,
+      "loss": 0.761,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7324090121317157,
+      "grad_norm": 2.774386150951619,
+      "learning_rate": 1.767940168123773e-06,
+      "loss": 0.7732,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7327556325823223,
+      "grad_norm": 4.366575558983553,
+      "learning_rate": 1.7636587908234759e-06,
+      "loss": 0.6806,
+      "step": 2114
+    },
+    {
+      "epoch": 0.733102253032929,
+      "grad_norm": 4.336612813324243,
+      "learning_rate": 1.759381493508332e-06,
+      "loss": 0.7506,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7334488734835355,
+      "grad_norm": 7.26242702999462,
+      "learning_rate": 1.7551082815706355e-06,
+      "loss": 0.7762,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7337954939341421,
+      "grad_norm": 3.72145957851131,
+      "learning_rate": 1.750839160397535e-06,
+      "loss": 0.7298,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7341421143847487,
+      "grad_norm": 2.991494109694439,
+      "learning_rate": 1.746574135371019e-06,
+      "loss": 0.8763,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7344887348353553,
+      "grad_norm": 2.7430625866411815,
+      "learning_rate": 1.7423132118679093e-06,
+      "loss": 0.684,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7348353552859619,
+      "grad_norm": 12.60099636434419,
+      "learning_rate": 1.7380563952598606e-06,
+      "loss": 0.728,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7351819757365684,
+      "grad_norm": 2.6140275315716393,
+      "learning_rate": 1.73380369091335e-06,
+      "loss": 0.7303,
+      "step": 2121
+    },
+    {
+      "epoch": 0.735528596187175,
+      "grad_norm": 4.441117904388193,
+      "learning_rate": 1.7295551041896686e-06,
+      "loss": 0.7452,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7358752166377817,
+      "grad_norm": 5.305648126353232,
+      "learning_rate": 1.7253106404449183e-06,
+      "loss": 0.654,
+      "step": 2123
+    },
+    {
+      "epoch": 0.7362218370883882,
+      "grad_norm": 3.2811270562110733,
+      "learning_rate": 1.721070305030002e-06,
+      "loss": 0.773,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7365684575389948,
+      "grad_norm": 3.119912481912095,
+      "learning_rate": 1.7168341032906193e-06,
+      "loss": 0.7238,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7369150779896014,
+      "grad_norm": 15.51177410991587,
+      "learning_rate": 1.7126020405672572e-06,
+      "loss": 0.775,
+      "step": 2126
+    },
+    {
+      "epoch": 0.737261698440208,
+      "grad_norm": 15.502683639965054,
+      "learning_rate": 1.708374122195186e-06,
+      "loss": 0.7156,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7376083188908146,
+      "grad_norm": 13.030429031915022,
+      "learning_rate": 1.7041503535044512e-06,
+      "loss": 0.8722,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7379549393414211,
+      "grad_norm": 2.547269793727284,
+      "learning_rate": 1.6999307398198623e-06,
+      "loss": 0.7638,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7383015597920277,
+      "grad_norm": 2.719088919032949,
+      "learning_rate": 1.6957152864609993e-06,
+      "loss": 0.8482,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7386481802426343,
+      "grad_norm": 4.249091456348031,
+      "learning_rate": 1.6915039987421916e-06,
+      "loss": 0.691,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7389948006932409,
+      "grad_norm": 8.657168849121074,
+      "learning_rate": 1.6872968819725176e-06,
+      "loss": 0.6477,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7393414211438475,
+      "grad_norm": 3.1706854936918933,
+      "learning_rate": 1.6830939414558007e-06,
+      "loss": 0.9169,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7396880415944541,
+      "grad_norm": 3.3052532301050332,
+      "learning_rate": 1.6788951824905924e-06,
+      "loss": 0.7683,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7400346620450606,
+      "grad_norm": 3.3290869264184573,
+      "learning_rate": 1.6747006103701796e-06,
+      "loss": 0.7037,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7403812824956673,
+      "grad_norm": 4.380547794526149,
+      "learning_rate": 1.6705102303825666e-06,
+      "loss": 0.8369,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7407279029462738,
+      "grad_norm": 3.2236739390852662,
+      "learning_rate": 1.666324047810477e-06,
+      "loss": 0.8329,
+      "step": 2137
+    },
+    {
+      "epoch": 0.7410745233968804,
+      "grad_norm": 5.500300178559343,
+      "learning_rate": 1.6621420679313415e-06,
+      "loss": 0.6877,
+      "step": 2138
+    },
+    {
+      "epoch": 0.741421143847487,
+      "grad_norm": 4.636493229462746,
+      "learning_rate": 1.6579642960172876e-06,
+      "loss": 0.8323,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7417677642980935,
+      "grad_norm": 2.7906651076605464,
+      "learning_rate": 1.653790737335143e-06,
+      "loss": 0.7852,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7421143847487002,
+      "grad_norm": 5.06963394045271,
+      "learning_rate": 1.649621397146423e-06,
+      "loss": 0.7117,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7424610051993068,
+      "grad_norm": 3.7733330113026944,
+      "learning_rate": 1.6454562807073248e-06,
+      "loss": 0.7526,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7428076256499133,
+      "grad_norm": 6.98600805760601,
+      "learning_rate": 1.6412953932687193e-06,
+      "loss": 0.8024,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7431542461005199,
+      "grad_norm": 4.993786328711226,
+      "learning_rate": 1.637138740076148e-06,
+      "loss": 0.6455,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7435008665511266,
+      "grad_norm": 28.258078998102242,
+      "learning_rate": 1.6329863263698125e-06,
+      "loss": 0.7963,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7438474870017331,
+      "grad_norm": 2.260561429737521,
+      "learning_rate": 1.6288381573845714e-06,
+      "loss": 0.8012,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7441941074523397,
+      "grad_norm": 4.71766784573236,
+      "learning_rate": 1.6246942383499308e-06,
+      "loss": 0.7384,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7445407279029462,
+      "grad_norm": 6.81992122193546,
+      "learning_rate": 1.620554574490042e-06,
+      "loss": 0.6747,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7448873483535529,
+      "grad_norm": 3.9250885034503002,
+      "learning_rate": 1.6164191710236838e-06,
+      "loss": 0.6697,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7452339688041595,
+      "grad_norm": 8.867913051635288,
+      "learning_rate": 1.6122880331642743e-06,
+      "loss": 0.7739,
+      "step": 2150
+    },
+    {
+      "epoch": 0.745580589254766,
+      "grad_norm": 3.592219822053752,
+      "learning_rate": 1.6081611661198482e-06,
+      "loss": 0.7706,
+      "step": 2151
+    },
+    {
+      "epoch": 0.7459272097053726,
+      "grad_norm": 3.389246461208805,
+      "learning_rate": 1.6040385750930577e-06,
+      "loss": 0.7289,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7462738301559793,
+      "grad_norm": 3.1803804952175914,
+      "learning_rate": 1.599920265281163e-06,
+      "loss": 0.6807,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7466204506065858,
+      "grad_norm": 3.3741494665901617,
+      "learning_rate": 1.59580624187603e-06,
+      "loss": 0.7577,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7469670710571924,
+      "grad_norm": 8.259117517650983,
+      "learning_rate": 1.591696510064115e-06,
+      "loss": 0.8031,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7473136915077989,
+      "grad_norm": 2.8517102571554314,
+      "learning_rate": 1.5875910750264706e-06,
+      "loss": 0.7737,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7476603119584055,
+      "grad_norm": 6.811080983794836,
+      "learning_rate": 1.5834899419387268e-06,
+      "loss": 0.7594,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7480069324090122,
+      "grad_norm": 2.2677119575358087,
+      "learning_rate": 1.5793931159710974e-06,
+      "loss": 0.6376,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7483535528596187,
+      "grad_norm": 4.166875865959536,
+      "learning_rate": 1.5753006022883606e-06,
+      "loss": 0.6716,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7487001733102253,
+      "grad_norm": 2.812325909083638,
+      "learning_rate": 1.5712124060498578e-06,
+      "loss": 0.6931,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7490467937608319,
+      "grad_norm": 2.3462020970878923,
+      "learning_rate": 1.5671285324094904e-06,
+      "loss": 0.7216,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7493934142114385,
+      "grad_norm": 2.6001024067874354,
+      "learning_rate": 1.563048986515709e-06,
+      "loss": 0.7537,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7497400346620451,
+      "grad_norm": 5.363133628080194,
+      "learning_rate": 1.5589737735115084e-06,
+      "loss": 0.7106,
+      "step": 2163
+    },
+    {
+      "epoch": 0.7500866551126516,
+      "grad_norm": 2.8733801270658264,
+      "learning_rate": 1.5549028985344228e-06,
+      "loss": 0.6823,
+      "step": 2164
+    },
+    {
+      "epoch": 0.7504332755632582,
+      "grad_norm": 2.7020837365141936,
+      "learning_rate": 1.5508363667165139e-06,
+      "loss": 0.808,
+      "step": 2165
+    },
+    {
+      "epoch": 0.7507798960138649,
+      "grad_norm": 9.065102646250395,
+      "learning_rate": 1.5467741831843713e-06,
+      "loss": 0.7308,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7511265164644714,
+      "grad_norm": 3.7837360884584217,
+      "learning_rate": 1.5427163530591017e-06,
+      "loss": 0.7343,
+      "step": 2167
+    },
+    {
+      "epoch": 0.751473136915078,
+      "grad_norm": 3.376852093021836,
+      "learning_rate": 1.5386628814563232e-06,
+      "loss": 0.7803,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7518197573656846,
+      "grad_norm": 5.807696988001566,
+      "learning_rate": 1.5346137734861598e-06,
+      "loss": 0.7084,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7521663778162911,
+      "grad_norm": 3.0051802231547224,
+      "learning_rate": 1.5305690342532338e-06,
+      "loss": 0.6836,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7525129982668978,
+      "grad_norm": 21.904365767066245,
+      "learning_rate": 1.52652866885666e-06,
+      "loss": 0.7547,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7528596187175043,
+      "grad_norm": 5.4977455421771335,
+      "learning_rate": 1.5224926823900394e-06,
+      "loss": 0.6987,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7532062391681109,
+      "grad_norm": 3.0476450474242465,
+      "learning_rate": 1.5184610799414535e-06,
+      "loss": 0.7378,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7535528596187175,
+      "grad_norm": 9.640462277747183,
+      "learning_rate": 1.5144338665934566e-06,
+      "loss": 0.7008,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7538994800693241,
+      "grad_norm": 12.73441799899826,
+      "learning_rate": 1.5104110474230666e-06,
+      "loss": 0.6706,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7542461005199307,
+      "grad_norm": 7.644095016227613,
+      "learning_rate": 1.5063926275017653e-06,
+      "loss": 0.7413,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7545927209705373,
+      "grad_norm": 37.88058240469977,
+      "learning_rate": 1.5023786118954869e-06,
+      "loss": 0.6996,
+      "step": 2177
+    },
+    {
+      "epoch": 0.7549393414211438,
+      "grad_norm": 4.141076784850181,
+      "learning_rate": 1.4983690056646162e-06,
+      "loss": 0.7983,
+      "step": 2178
+    },
+    {
+      "epoch": 0.7552859618717505,
+      "grad_norm": 2.545469495249976,
+      "learning_rate": 1.4943638138639772e-06,
+      "loss": 0.7757,
+      "step": 2179
+    },
+    {
+      "epoch": 0.755632582322357,
+      "grad_norm": 4.8285888505438646,
+      "learning_rate": 1.4903630415428245e-06,
+      "loss": 0.7624,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7559792027729636,
+      "grad_norm": 4.4971571752824735,
+      "learning_rate": 1.4863666937448474e-06,
+      "loss": 0.7821,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7563258232235702,
+      "grad_norm": 3.825585374854395,
+      "learning_rate": 1.4823747755081546e-06,
+      "loss": 0.7827,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7566724436741767,
+      "grad_norm": 27.3423164140503,
+      "learning_rate": 1.478387291865271e-06,
+      "loss": 0.8082,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7570190641247834,
+      "grad_norm": 4.794655029499163,
+      "learning_rate": 1.4744042478431297e-06,
+      "loss": 0.6419,
+      "step": 2184
+    },
+    {
+      "epoch": 0.75736568457539,
+      "grad_norm": 4.07529273005088,
+      "learning_rate": 1.4704256484630691e-06,
+      "loss": 0.7109,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7577123050259965,
+      "grad_norm": 3.6061246040038006,
+      "learning_rate": 1.4664514987408223e-06,
+      "loss": 0.6757,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7580589254766031,
+      "grad_norm": 3.432813302028276,
+      "learning_rate": 1.4624818036865146e-06,
+      "loss": 0.6994,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7584055459272097,
+      "grad_norm": 4.615342921707863,
+      "learning_rate": 1.4585165683046537e-06,
+      "loss": 0.7567,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7587521663778163,
+      "grad_norm": 9.583369138708024,
+      "learning_rate": 1.4545557975941255e-06,
+      "loss": 0.7146,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7590987868284229,
+      "grad_norm": 2.8239400265545984,
+      "learning_rate": 1.4505994965481885e-06,
+      "loss": 0.733,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7594454072790294,
+      "grad_norm": 4.522298250533864,
+      "learning_rate": 1.4466476701544657e-06,
+      "loss": 0.6783,
+      "step": 2191
+    },
+    {
+      "epoch": 0.759792027729636,
+      "grad_norm": 31.525684452875705,
+      "learning_rate": 1.4427003233949378e-06,
+      "loss": 0.7223,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7601386481802427,
+      "grad_norm": 2.3203851136351332,
+      "learning_rate": 1.4387574612459403e-06,
+      "loss": 0.6376,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7604852686308492,
+      "grad_norm": 2.5028701998662024,
+      "learning_rate": 1.4348190886781532e-06,
+      "loss": 0.7637,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7608318890814558,
+      "grad_norm": 3.3647367210739505,
+      "learning_rate": 1.4308852106565997e-06,
+      "loss": 0.7314,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7611785095320623,
+      "grad_norm": 3.2670390569438155,
+      "learning_rate": 1.4269558321406307e-06,
+      "loss": 0.7385,
+      "step": 2196
+    },
+    {
+      "epoch": 0.761525129982669,
+      "grad_norm": 8.205517707979675,
+      "learning_rate": 1.4230309580839291e-06,
+      "loss": 0.7298,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7618717504332756,
+      "grad_norm": 5.2512894995400785,
+      "learning_rate": 1.4191105934345007e-06,
+      "loss": 0.8169,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7622183708838821,
+      "grad_norm": 4.078508563216686,
+      "learning_rate": 1.4151947431346635e-06,
+      "loss": 0.7048,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7625649913344887,
+      "grad_norm": 3.426295471276162,
+      "learning_rate": 1.4112834121210462e-06,
+      "loss": 0.7364,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7629116117850954,
+      "grad_norm": 3.0669609164826466,
+      "learning_rate": 1.4073766053245752e-06,
+      "loss": 0.7625,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7632582322357019,
+      "grad_norm": 8.831771934710316,
+      "learning_rate": 1.4034743276704799e-06,
+      "loss": 0.756,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7636048526863085,
+      "grad_norm": 5.703967187873182,
+      "learning_rate": 1.3995765840782765e-06,
+      "loss": 0.8647,
+      "step": 2203
+    },
+    {
+      "epoch": 0.763951473136915,
+      "grad_norm": 7.755158576655444,
+      "learning_rate": 1.3956833794617653e-06,
+      "loss": 0.707,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7642980935875217,
+      "grad_norm": 12.848653572203988,
+      "learning_rate": 1.3917947187290254e-06,
+      "loss": 0.7527,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7646447140381283,
+      "grad_norm": 2.676597044017325,
+      "learning_rate": 1.387910606782407e-06,
+      "loss": 0.7296,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7649913344887348,
+      "grad_norm": 3.5478981392457216,
+      "learning_rate": 1.384031048518526e-06,
+      "loss": 0.7963,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7653379549393414,
+      "grad_norm": 5.485513690522526,
+      "learning_rate": 1.3801560488282574e-06,
+      "loss": 0.7081,
+      "step": 2208
+    },
+    {
+      "epoch": 0.765684575389948,
+      "grad_norm": 2.990405307294021,
+      "learning_rate": 1.3762856125967283e-06,
+      "loss": 0.728,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7660311958405546,
+      "grad_norm": 5.143711063982162,
+      "learning_rate": 1.3724197447033144e-06,
+      "loss": 0.7596,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7663778162911612,
+      "grad_norm": 3.800214498674543,
+      "learning_rate": 1.3685584500216315e-06,
+      "loss": 0.7104,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7667244367417677,
+      "grad_norm": 6.514095993108665,
+      "learning_rate": 1.3647017334195296e-06,
+      "loss": 0.747,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7670710571923743,
+      "grad_norm": 4.56458634919514,
+      "learning_rate": 1.360849599759088e-06,
+      "loss": 0.7331,
+      "step": 2213
+    },
+    {
+      "epoch": 0.767417677642981,
+      "grad_norm": 3.068061401928201,
+      "learning_rate": 1.3570020538966083e-06,
+      "loss": 0.7639,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7677642980935875,
+      "grad_norm": 2.9510569756232363,
+      "learning_rate": 1.3531591006826096e-06,
+      "loss": 0.7753,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7681109185441941,
+      "grad_norm": 7.2258870254730665,
+      "learning_rate": 1.3493207449618163e-06,
+      "loss": 0.7083,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7684575389948007,
+      "grad_norm": 2.9501069972752276,
+      "learning_rate": 1.3454869915731595e-06,
+      "loss": 0.6887,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7688041594454073,
+      "grad_norm": 3.474340556760777,
+      "learning_rate": 1.3416578453497737e-06,
+      "loss": 0.6766,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7691507798960139,
+      "grad_norm": 4.1335252941889165,
+      "learning_rate": 1.3378333111189774e-06,
+      "loss": 0.695,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7694974003466204,
+      "grad_norm": 4.071985659971527,
+      "learning_rate": 1.3340133937022797e-06,
+      "loss": 0.7741,
+      "step": 2220
+    },
+    {
+      "epoch": 0.769844020797227,
+      "grad_norm": 2.212784952376869,
+      "learning_rate": 1.3301980979153656e-06,
+      "loss": 0.6624,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7701906412478337,
+      "grad_norm": 4.084709216112869,
+      "learning_rate": 1.3263874285680962e-06,
+      "loss": 0.7261,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7705372616984402,
+      "grad_norm": 5.660616985351659,
+      "learning_rate": 1.3225813904644996e-06,
+      "loss": 0.7368,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7708838821490468,
+      "grad_norm": 3.12421782179147,
+      "learning_rate": 1.3187799884027663e-06,
+      "loss": 0.7912,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7712305025996534,
+      "grad_norm": 2.323403855528503,
+      "learning_rate": 1.3149832271752394e-06,
+      "loss": 0.7873,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7715771230502599,
+      "grad_norm": 9.097400842786318,
+      "learning_rate": 1.3111911115684185e-06,
+      "loss": 0.6797,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7719237435008666,
+      "grad_norm": 3.7282852006847382,
+      "learning_rate": 1.307403646362936e-06,
+      "loss": 0.7568,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7722703639514731,
+      "grad_norm": 2.952483022248241,
+      "learning_rate": 1.3036208363335689e-06,
+      "loss": 0.6708,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7726169844020797,
+      "grad_norm": 2.5538565406124114,
+      "learning_rate": 1.2998426862492247e-06,
+      "loss": 0.675,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7729636048526863,
+      "grad_norm": 3.4020486251151087,
+      "learning_rate": 1.2960692008729336e-06,
+      "loss": 0.7352,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7733102253032929,
+      "grad_norm": 4.815895696137967,
+      "learning_rate": 1.2923003849618483e-06,
+      "loss": 0.7475,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7736568457538995,
+      "grad_norm": 3.19877657522224,
+      "learning_rate": 1.2885362432672322e-06,
+      "loss": 0.745,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7740034662045061,
+      "grad_norm": 4.0487322126318395,
+      "learning_rate": 1.284776780534458e-06,
+      "loss": 0.7693,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7743500866551126,
+      "grad_norm": 4.782175287127135,
+      "learning_rate": 1.2810220015029978e-06,
+      "loss": 0.7144,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7746967071057193,
+      "grad_norm": 7.891212101046288,
+      "learning_rate": 1.2772719109064208e-06,
+      "loss": 0.7418,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7750433275563259,
+      "grad_norm": 12.033349079048854,
+      "learning_rate": 1.2735265134723861e-06,
+      "loss": 0.7641,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7753899480069324,
+      "grad_norm": 4.076946943235272,
+      "learning_rate": 1.2697858139226315e-06,
+      "loss": 0.6745,
+      "step": 2237
+    },
+    {
+      "epoch": 0.775736568457539,
+      "grad_norm": 7.7536478407365985,
+      "learning_rate": 1.2660498169729773e-06,
+      "loss": 0.7609,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7760831889081455,
+      "grad_norm": 5.334876154263405,
+      "learning_rate": 1.262318527333315e-06,
+      "loss": 0.6219,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7764298093587522,
+      "grad_norm": 4.0791616767204,
+      "learning_rate": 1.2585919497076e-06,
+      "loss": 0.8153,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7767764298093588,
+      "grad_norm": 2.533584424348394,
+      "learning_rate": 1.2548700887938492e-06,
+      "loss": 0.7008,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7771230502599653,
+      "grad_norm": 3.7551703538189383,
+      "learning_rate": 1.251152949284129e-06,
+      "loss": 0.7266,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7774696707105719,
+      "grad_norm": 2.6876344815809956,
+      "learning_rate": 1.2474405358645575e-06,
+      "loss": 0.5928,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7778162911611786,
+      "grad_norm": 1.9155560591537628,
+      "learning_rate": 1.2437328532152943e-06,
+      "loss": 0.7459,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7781629116117851,
+      "grad_norm": 3.153210760576075,
+      "learning_rate": 1.2400299060105331e-06,
+      "loss": 0.7874,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7785095320623917,
+      "grad_norm": 5.4956106789528665,
+      "learning_rate": 1.2363316989185032e-06,
+      "loss": 0.6474,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7788561525129982,
+      "grad_norm": 4.676942661805003,
+      "learning_rate": 1.23263823660145e-06,
+      "loss": 0.8937,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7792027729636049,
+      "grad_norm": 2.7832763424390894,
+      "learning_rate": 1.2289495237156418e-06,
+      "loss": 0.7298,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7795493934142115,
+      "grad_norm": 2.923119602667243,
+      "learning_rate": 1.2252655649113604e-06,
+      "loss": 0.6646,
+      "step": 2249
+    },
+    {
+      "epoch": 0.779896013864818,
+      "grad_norm": 4.337089892759537,
+      "learning_rate": 1.221586364832892e-06,
+      "loss": 0.6698,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7802426343154246,
+      "grad_norm": 3.1823294661853128,
+      "learning_rate": 1.2179119281185237e-06,
+      "loss": 0.8273,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7805892547660312,
+      "grad_norm": 2.7631722812973467,
+      "learning_rate": 1.2142422594005388e-06,
+      "loss": 0.7399,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7809358752166378,
+      "grad_norm": 4.7696698011109,
+      "learning_rate": 1.2105773633052093e-06,
+      "loss": 0.7664,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7812824956672444,
+      "grad_norm": 2.9610210442918707,
+      "learning_rate": 1.2069172444527898e-06,
+      "loss": 0.682,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7816291161178509,
+      "grad_norm": 3.5804841744452425,
+      "learning_rate": 1.203261907457513e-06,
+      "loss": 0.6451,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7819757365684575,
+      "grad_norm": 3.6035968327925176,
+      "learning_rate": 1.1996113569275841e-06,
+      "loss": 0.7309,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7823223570190642,
+      "grad_norm": 3.5136815605049083,
+      "learning_rate": 1.19596559746517e-06,
+      "loss": 0.6949,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7826689774696707,
+      "grad_norm": 5.826300003184325,
+      "learning_rate": 1.1923246336664014e-06,
+      "loss": 0.6525,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7830155979202773,
+      "grad_norm": 3.727379092074676,
+      "learning_rate": 1.1886884701213642e-06,
+      "loss": 0.6666,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7833622183708839,
+      "grad_norm": 2.798581481932015,
+      "learning_rate": 1.1850571114140897e-06,
+      "loss": 0.7546,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7837088388214904,
+      "grad_norm": 2.755694125475684,
+      "learning_rate": 1.1814305621225535e-06,
+      "loss": 0.8356,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7840554592720971,
+      "grad_norm": 10.4095940782413,
+      "learning_rate": 1.1778088268186688e-06,
+      "loss": 0.7591,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7844020797227036,
+      "grad_norm": 3.689245168767778,
+      "learning_rate": 1.1741919100682757e-06,
+      "loss": 0.818,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7847487001733102,
+      "grad_norm": 4.333219828880789,
+      "learning_rate": 1.1705798164311443e-06,
+      "loss": 0.7485,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7850953206239168,
+      "grad_norm": 3.4640956634608457,
+      "learning_rate": 1.1669725504609614e-06,
+      "loss": 0.7501,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7854419410745234,
+      "grad_norm": 2.9718813206557213,
+      "learning_rate": 1.1633701167053313e-06,
+      "loss": 0.8064,
+      "step": 2266
+    },
+    {
+      "epoch": 0.78578856152513,
+      "grad_norm": 7.315998984204347,
+      "learning_rate": 1.1597725197057646e-06,
+      "loss": 0.8082,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7861351819757366,
+      "grad_norm": 3.116854416455241,
+      "learning_rate": 1.1561797639976708e-06,
+      "loss": 0.7083,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7864818024263431,
+      "grad_norm": 17.895263607293362,
+      "learning_rate": 1.1525918541103598e-06,
+      "loss": 0.8394,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7868284228769498,
+      "grad_norm": 4.027770263792255,
+      "learning_rate": 1.1490087945670331e-06,
+      "loss": 0.7766,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7871750433275563,
+      "grad_norm": 2.823827573689336,
+      "learning_rate": 1.145430589884775e-06,
+      "loss": 0.7855,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7875216637781629,
+      "grad_norm": 2.9882624533489524,
+      "learning_rate": 1.1418572445745513e-06,
+      "loss": 0.7242,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7878682842287695,
+      "grad_norm": 2.7569161430488216,
+      "learning_rate": 1.1382887631412004e-06,
+      "loss": 0.784,
+      "step": 2273
+    },
+    {
+      "epoch": 0.788214904679376,
+      "grad_norm": 5.978471833130474,
+      "learning_rate": 1.13472515008343e-06,
+      "loss": 0.5785,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7885615251299827,
+      "grad_norm": 3.3106402886058026,
+      "learning_rate": 1.1311664098938092e-06,
+      "loss": 0.767,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7889081455805893,
+      "grad_norm": 5.248090617663467,
+      "learning_rate": 1.1276125470587645e-06,
+      "loss": 0.7178,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7892547660311958,
+      "grad_norm": 20.378997592816443,
+      "learning_rate": 1.1240635660585752e-06,
+      "loss": 0.7876,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7896013864818024,
+      "grad_norm": 4.425831480712005,
+      "learning_rate": 1.1205194713673606e-06,
+      "loss": 0.7867,
+      "step": 2278
+    },
+    {
+      "epoch": 0.789948006932409,
+      "grad_norm": 12.516515394451197,
+      "learning_rate": 1.1169802674530882e-06,
+      "loss": 0.8651,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7902946273830156,
+      "grad_norm": 3.374751865893061,
+      "learning_rate": 1.1134459587775543e-06,
+      "loss": 0.7279,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7906412478336222,
+      "grad_norm": 2.356025285414891,
+      "learning_rate": 1.1099165497963848e-06,
+      "loss": 0.7163,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7909878682842287,
+      "grad_norm": 3.8920458248863525,
+      "learning_rate": 1.1063920449590309e-06,
+      "loss": 0.6972,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7913344887348354,
+      "grad_norm": 9.839528414461073,
+      "learning_rate": 1.1028724487087562e-06,
+      "loss": 0.7752,
+      "step": 2283
+    },
+    {
+      "epoch": 0.791681109185442,
+      "grad_norm": 3.1803926837169048,
+      "learning_rate": 1.0993577654826409e-06,
+      "loss": 0.7781,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7920277296360485,
+      "grad_norm": 3.2679067645731097,
+      "learning_rate": 1.09584799971157e-06,
+      "loss": 0.6779,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7923743500866551,
+      "grad_norm": 3.178533858767169,
+      "learning_rate": 1.0923431558202274e-06,
+      "loss": 0.7845,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7927209705372616,
+      "grad_norm": 4.411989919349972,
+      "learning_rate": 1.088843238227098e-06,
+      "loss": 0.6865,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7930675909878683,
+      "grad_norm": 4.466974961485861,
+      "learning_rate": 1.085348251344448e-06,
+      "loss": 0.7601,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7934142114384749,
+      "grad_norm": 3.617618640869854,
+      "learning_rate": 1.0818581995783328e-06,
+      "loss": 0.8973,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7937608318890814,
+      "grad_norm": 5.377853404314917,
+      "learning_rate": 1.0783730873285846e-06,
+      "loss": 0.8167,
+      "step": 2290
+    },
+    {
+      "epoch": 0.794107452339688,
+      "grad_norm": 4.451206901265293,
+      "learning_rate": 1.0748929189888098e-06,
+      "loss": 0.6485,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7944540727902947,
+      "grad_norm": 3.50943355769039,
+      "learning_rate": 1.0714176989463804e-06,
+      "loss": 0.664,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7948006932409012,
+      "grad_norm": 2.869321420848003,
+      "learning_rate": 1.0679474315824306e-06,
+      "loss": 0.7545,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7951473136915078,
+      "grad_norm": 2.9062502586465557,
+      "learning_rate": 1.064482121271852e-06,
+      "loss": 0.6775,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7954939341421143,
+      "grad_norm": 2.943726380188781,
+      "learning_rate": 1.0610217723832855e-06,
+      "loss": 0.7508,
+      "step": 2295
+    },
+    {
+      "epoch": 0.795840554592721,
+      "grad_norm": 6.8131746665533806,
+      "learning_rate": 1.0575663892791183e-06,
+      "loss": 0.8413,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7961871750433276,
+      "grad_norm": 1.967585079136128,
+      "learning_rate": 1.0541159763154767e-06,
+      "loss": 0.7436,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7965337954939341,
+      "grad_norm": 4.053126200669751,
+      "learning_rate": 1.0506705378422228e-06,
+      "loss": 0.7684,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7968804159445407,
+      "grad_norm": 3.009056665885097,
+      "learning_rate": 1.047230078202942e-06,
+      "loss": 0.7718,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7972270363951474,
+      "grad_norm": 12.076727481967009,
+      "learning_rate": 1.0437946017349515e-06,
+      "loss": 0.8129,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7975736568457539,
+      "grad_norm": 3.200674799512866,
+      "learning_rate": 1.0403641127692805e-06,
+      "loss": 0.8276,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7979202772963605,
+      "grad_norm": 5.907401947625137,
+      "learning_rate": 1.0369386156306721e-06,
+      "loss": 0.7834,
+      "step": 2302
+    },
+    {
+      "epoch": 0.798266897746967,
+      "grad_norm": 3.1987447074248916,
+      "learning_rate": 1.0335181146375767e-06,
+      "loss": 0.691,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7986135181975736,
+      "grad_norm": 2.8027363372745326,
+      "learning_rate": 1.0301026141021442e-06,
+      "loss": 0.7752,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7989601386481803,
+      "grad_norm": 4.403782203759442,
+      "learning_rate": 1.026692118330223e-06,
+      "loss": 0.9032,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7993067590987868,
+      "grad_norm": 3.117718046153743,
+      "learning_rate": 1.0232866316213496e-06,
+      "loss": 0.682,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7996533795493934,
+      "grad_norm": 2.7508787700120965,
+      "learning_rate": 1.0198861582687497e-06,
+      "loss": 0.6975,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 4.19546558308559,
+      "learning_rate": 1.0164907025593268e-06,
+      "loss": 0.7255,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8003466204506066,
+      "grad_norm": 3.5583956146688487,
+      "learning_rate": 1.0131002687736551e-06,
+      "loss": 0.7331,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8006932409012132,
+      "grad_norm": 6.363477896963617,
+      "learning_rate": 1.009714861185983e-06,
+      "loss": 0.7297,
+      "step": 2310
+    },
+    {
+      "epoch": 0.8010398613518197,
+      "grad_norm": 3.8666922850108234,
+      "learning_rate": 1.0063344840642197e-06,
+      "loss": 0.8167,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8013864818024263,
+      "grad_norm": 4.34103281123582,
+      "learning_rate": 1.002959141669933e-06,
+      "loss": 0.74,
+      "step": 2312
+    },
+    {
+      "epoch": 0.801733102253033,
+      "grad_norm": 3.1530902931769007,
+      "learning_rate": 9.995888382583451e-07,
+      "loss": 0.7165,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8020797227036395,
+      "grad_norm": 24.12267007591921,
+      "learning_rate": 9.96223578078323e-07,
+      "loss": 0.7993,
+      "step": 2314
+    },
+    {
+      "epoch": 0.8024263431542461,
+      "grad_norm": 3.7774852259676788,
+      "learning_rate": 9.92863365372378e-07,
+      "loss": 0.7021,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8027729636048527,
+      "grad_norm": 3.8797416854940496,
+      "learning_rate": 9.895082043766575e-07,
+      "loss": 0.8225,
+      "step": 2316
+    },
+    {
+      "epoch": 0.8031195840554592,
+      "grad_norm": 9.185842645177056,
+      "learning_rate": 9.861580993209396e-07,
+      "loss": 0.7989,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8034662045060659,
+      "grad_norm": 4.734018803507718,
+      "learning_rate": 9.828130544286307e-07,
+      "loss": 0.728,
+      "step": 2318
+    },
+    {
+      "epoch": 0.8038128249566724,
+      "grad_norm": 3.32632723798084,
+      "learning_rate": 9.794730739167524e-07,
+      "loss": 0.7778,
+      "step": 2319
+    },
+    {
+      "epoch": 0.804159445407279,
+      "grad_norm": 5.381877772678218,
+      "learning_rate": 9.761381619959497e-07,
+      "loss": 0.6582,
+      "step": 2320
+    },
+    {
+      "epoch": 0.8045060658578856,
+      "grad_norm": 5.66538112276859,
+      "learning_rate": 9.728083228704722e-07,
+      "loss": 0.7451,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8048526863084922,
+      "grad_norm": 4.965576306228912,
+      "learning_rate": 9.694835607381753e-07,
+      "loss": 0.6163,
+      "step": 2322
+    },
+    {
+      "epoch": 0.8051993067590988,
+      "grad_norm": 6.393939995166974,
+      "learning_rate": 9.661638797905171e-07,
+      "loss": 0.6702,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8055459272097054,
+      "grad_norm": 6.93287252010143,
+      "learning_rate": 9.628492842125436e-07,
+      "loss": 0.7658,
+      "step": 2324
+    },
+    {
+      "epoch": 0.8058925476603119,
+      "grad_norm": 12.634377404360976,
+      "learning_rate": 9.595397781828958e-07,
+      "loss": 0.6047,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8062391681109186,
+      "grad_norm": 11.771510534514885,
+      "learning_rate": 9.562353658737938e-07,
+      "loss": 0.6767,
+      "step": 2326
+    },
+    {
+      "epoch": 0.8065857885615252,
+      "grad_norm": 4.816012241903835,
+      "learning_rate": 9.529360514510422e-07,
+      "loss": 0.7993,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8069324090121317,
+      "grad_norm": 3.3367238080616017,
+      "learning_rate": 9.496418390740142e-07,
+      "loss": 0.6979,
+      "step": 2328
+    },
+    {
+      "epoch": 0.8072790294627383,
+      "grad_norm": 3.370937279349196,
+      "learning_rate": 9.463527328956496e-07,
+      "loss": 0.8053,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8076256499133448,
+      "grad_norm": 4.338529361404702,
+      "learning_rate": 9.430687370624547e-07,
+      "loss": 0.736,
+      "step": 2330
+    },
+    {
+      "epoch": 0.8079722703639515,
+      "grad_norm": 5.76877296582566,
+      "learning_rate": 9.397898557144919e-07,
+      "loss": 0.7986,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8083188908145581,
+      "grad_norm": 4.641259030914428,
+      "learning_rate": 9.36516092985375e-07,
+      "loss": 0.7557,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8086655112651646,
+      "grad_norm": 3.9347113679257566,
+      "learning_rate": 9.332474530022668e-07,
+      "loss": 0.785,
+      "step": 2333
+    },
+    {
+      "epoch": 0.8090121317157712,
+      "grad_norm": 2.611649719045231,
+      "learning_rate": 9.299839398858696e-07,
+      "loss": 0.7204,
+      "step": 2334
+    },
+    {
+      "epoch": 0.8093587521663779,
+      "grad_norm": 5.774818129995877,
+      "learning_rate": 9.267255577504247e-07,
+      "loss": 0.7873,
+      "step": 2335
+    },
+    {
+      "epoch": 0.8097053726169844,
+      "grad_norm": 4.399733267261132,
+      "learning_rate": 9.234723107037031e-07,
+      "loss": 0.7725,
+      "step": 2336
+    },
+    {
+      "epoch": 0.810051993067591,
+      "grad_norm": 5.57787339681974,
+      "learning_rate": 9.202242028470027e-07,
+      "loss": 0.7308,
+      "step": 2337
+    },
+    {
+      "epoch": 0.8103986135181975,
+      "grad_norm": 24.028920955058236,
+      "learning_rate": 9.169812382751431e-07,
+      "loss": 0.7507,
+      "step": 2338
+    },
+    {
+      "epoch": 0.8107452339688042,
+      "grad_norm": 2.8081655622912596,
+      "learning_rate": 9.13743421076459e-07,
+      "loss": 0.7249,
+      "step": 2339
+    },
+    {
+      "epoch": 0.8110918544194108,
+      "grad_norm": 4.656544153841659,
+      "learning_rate": 9.105107553327963e-07,
+      "loss": 0.9217,
+      "step": 2340
+    },
+    {
+      "epoch": 0.8114384748700173,
+      "grad_norm": 5.350880306416844,
+      "learning_rate": 9.072832451195074e-07,
+      "loss": 0.8072,
+      "step": 2341
+    },
+    {
+      "epoch": 0.8117850953206239,
+      "grad_norm": 4.431800199087632,
+      "learning_rate": 9.04060894505443e-07,
+      "loss": 0.6766,
+      "step": 2342
+    },
+    {
+      "epoch": 0.8121317157712306,
+      "grad_norm": 4.208024923224915,
+      "learning_rate": 9.008437075529513e-07,
+      "loss": 0.854,
+      "step": 2343
+    },
+    {
+      "epoch": 0.8124783362218371,
+      "grad_norm": 10.49748204755042,
+      "learning_rate": 8.97631688317872e-07,
+      "loss": 0.796,
+      "step": 2344
+    },
+    {
+      "epoch": 0.8128249566724437,
+      "grad_norm": 4.597066196677089,
+      "learning_rate": 8.944248408495242e-07,
+      "loss": 0.7307,
+      "step": 2345
+    },
+    {
+      "epoch": 0.8131715771230502,
+      "grad_norm": 3.1655882385673526,
+      "learning_rate": 8.912231691907136e-07,
+      "loss": 0.6694,
+      "step": 2346
+    },
+    {
+      "epoch": 0.8135181975736568,
+      "grad_norm": 4.999385377789044,
+      "learning_rate": 8.880266773777157e-07,
+      "loss": 0.8259,
+      "step": 2347
+    },
+    {
+      "epoch": 0.8138648180242635,
+      "grad_norm": 2.8099262495873267,
+      "learning_rate": 8.848353694402823e-07,
+      "loss": 0.5904,
+      "step": 2348
+    },
+    {
+      "epoch": 0.81421143847487,
+      "grad_norm": 4.265465116537009,
+      "learning_rate": 8.816492494016244e-07,
+      "loss": 0.7099,
+      "step": 2349
+    },
+    {
+      "epoch": 0.8145580589254766,
+      "grad_norm": 3.259920996579149,
+      "learning_rate": 8.784683212784134e-07,
+      "loss": 0.6307,
+      "step": 2350
+    },
+    {
+      "epoch": 0.8149046793760832,
+      "grad_norm": 6.525861174418144,
+      "learning_rate": 8.752925890807768e-07,
+      "loss": 0.7344,
+      "step": 2351
+    },
+    {
+      "epoch": 0.8152512998266898,
+      "grad_norm": 3.3839375701548273,
+      "learning_rate": 8.721220568122912e-07,
+      "loss": 0.7955,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8155979202772964,
+      "grad_norm": 3.423142022448829,
+      "learning_rate": 8.689567284699779e-07,
+      "loss": 0.6773,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8159445407279029,
+      "grad_norm": 3.6171376142892866,
+      "learning_rate": 8.657966080442987e-07,
+      "loss": 0.74,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8162911611785095,
+      "grad_norm": 2.3281206156838556,
+      "learning_rate": 8.626416995191472e-07,
+      "loss": 0.7434,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8166377816291162,
+      "grad_norm": 4.791831826341685,
+      "learning_rate": 8.594920068718498e-07,
+      "loss": 0.6701,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8169844020797227,
+      "grad_norm": 6.598872342954494,
+      "learning_rate": 8.563475340731542e-07,
+      "loss": 0.7191,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8173310225303293,
+      "grad_norm": 5.677113612095688,
+      "learning_rate": 8.532082850872303e-07,
+      "loss": 0.7804,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8176776429809359,
+      "grad_norm": 4.06036791133746,
+      "learning_rate": 8.50074263871662e-07,
+      "loss": 0.6776,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8180242634315424,
+      "grad_norm": 8.44300970426741,
+      "learning_rate": 8.469454743774374e-07,
+      "loss": 0.733,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8183708838821491,
+      "grad_norm": 7.371531691065908,
+      "learning_rate": 8.438219205489578e-07,
+      "loss": 0.7889,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8187175043327556,
+      "grad_norm": 6.3268181023650785,
+      "learning_rate": 8.407036063240192e-07,
+      "loss": 0.8121,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8190641247833622,
+      "grad_norm": 2.8496163846457914,
+      "learning_rate": 8.375905356338116e-07,
+      "loss": 0.6595,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8194107452339688,
+      "grad_norm": 3.2513171170853505,
+      "learning_rate": 8.344827124029176e-07,
+      "loss": 0.8391,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8197573656845754,
+      "grad_norm": 3.565242771087725,
+      "learning_rate": 8.313801405492993e-07,
+      "loss": 0.6865,
+      "step": 2365
+    },
+    {
+      "epoch": 0.820103986135182,
+      "grad_norm": 3.746475381044354,
+      "learning_rate": 8.282828239843044e-07,
+      "loss": 0.6764,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8204506065857886,
+      "grad_norm": 10.019520714684065,
+      "learning_rate": 8.251907666126502e-07,
+      "loss": 0.6823,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8207972270363951,
+      "grad_norm": 8.074958516542756,
+      "learning_rate": 8.221039723324287e-07,
+      "loss": 0.8237,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8211438474870018,
+      "grad_norm": 2.475640414228001,
+      "learning_rate": 8.190224450350942e-07,
+      "loss": 0.7108,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8214904679376083,
+      "grad_norm": 3.6627814887676307,
+      "learning_rate": 8.15946188605462e-07,
+      "loss": 0.7472,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8218370883882149,
+      "grad_norm": 3.1994801980757535,
+      "learning_rate": 8.128752069216994e-07,
+      "loss": 0.6462,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8221837088388215,
+      "grad_norm": 3.051070213426927,
+      "learning_rate": 8.098095038553283e-07,
+      "loss": 0.6433,
+      "step": 2372
+    },
+    {
+      "epoch": 0.822530329289428,
+      "grad_norm": 3.659647390701968,
+      "learning_rate": 8.067490832712128e-07,
+      "loss": 0.7176,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8228769497400347,
+      "grad_norm": 4.724615509885641,
+      "learning_rate": 8.03693949027558e-07,
+      "loss": 0.8113,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8232235701906413,
+      "grad_norm": 2.3765344423161743,
+      "learning_rate": 8.006441049759094e-07,
+      "loss": 0.831,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8235701906412478,
+      "grad_norm": 2.6635308108935045,
+      "learning_rate": 7.975995549611353e-07,
+      "loss": 0.6962,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8239168110918544,
+      "grad_norm": 3.5910286486518257,
+      "learning_rate": 7.945603028214355e-07,
+      "loss": 0.7218,
+      "step": 2377
+    },
+    {
+      "epoch": 0.824263431542461,
+      "grad_norm": 3.815544607746058,
+      "learning_rate": 7.915263523883282e-07,
+      "loss": 0.7048,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8246100519930676,
+      "grad_norm": 4.631552886261216,
+      "learning_rate": 7.884977074866501e-07,
+      "loss": 0.7414,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8249566724436742,
+      "grad_norm": 3.6608110257313116,
+      "learning_rate": 7.85474371934547e-07,
+      "loss": 0.621,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8253032928942807,
+      "grad_norm": 4.402312536302008,
+      "learning_rate": 7.824563495434734e-07,
+      "loss": 0.6373,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8256499133448874,
+      "grad_norm": 3.9554941940022386,
+      "learning_rate": 7.794436441181846e-07,
+      "loss": 0.6456,
+      "step": 2382
+    },
+    {
+      "epoch": 0.825996533795494,
+      "grad_norm": 4.436596799192899,
+      "learning_rate": 7.76436259456732e-07,
+      "loss": 0.7239,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8263431542461005,
+      "grad_norm": 6.116678970633913,
+      "learning_rate": 7.734341993504613e-07,
+      "loss": 0.736,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8266897746967071,
+      "grad_norm": 3.702019195536418,
+      "learning_rate": 7.704374675840048e-07,
+      "loss": 0.6339,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8270363951473136,
+      "grad_norm": 4.459888690794445,
+      "learning_rate": 7.674460679352758e-07,
+      "loss": 0.8135,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8273830155979203,
+      "grad_norm": 3.2315619316495505,
+      "learning_rate": 7.644600041754663e-07,
+      "loss": 0.7252,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8277296360485269,
+      "grad_norm": 2.9030014541519593,
+      "learning_rate": 7.614792800690446e-07,
+      "loss": 0.7755,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8280762564991334,
+      "grad_norm": 2.5934074202840054,
+      "learning_rate": 7.585038993737437e-07,
+      "loss": 0.7303,
+      "step": 2389
+    },
+    {
+      "epoch": 0.82842287694974,
+      "grad_norm": 2.9915544651071544,
+      "learning_rate": 7.555338658405625e-07,
+      "loss": 0.7684,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8287694974003467,
+      "grad_norm": 28.0924940254878,
+      "learning_rate": 7.525691832137566e-07,
+      "loss": 0.7573,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8291161178509532,
+      "grad_norm": 3.988683833281145,
+      "learning_rate": 7.496098552308368e-07,
+      "loss": 0.7403,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8294627383015598,
+      "grad_norm": 3.198168419613615,
+      "learning_rate": 7.466558856225642e-07,
+      "loss": 0.7631,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8298093587521663,
+      "grad_norm": 4.108544593226742,
+      "learning_rate": 7.437072781129445e-07,
+      "loss": 0.7203,
+      "step": 2394
+    },
+    {
+      "epoch": 0.830155979202773,
+      "grad_norm": 3.6950622259276997,
+      "learning_rate": 7.407640364192226e-07,
+      "loss": 0.6945,
+      "step": 2395
+    },
+    {
+      "epoch": 0.8305025996533796,
+      "grad_norm": 10.164725073955578,
+      "learning_rate": 7.378261642518803e-07,
+      "loss": 0.6817,
+      "step": 2396
+    },
+    {
+      "epoch": 0.8308492201039861,
+      "grad_norm": 3.79269460517039,
+      "learning_rate": 7.348936653146288e-07,
+      "loss": 0.8187,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8311958405545927,
+      "grad_norm": 2.3243534755778352,
+      "learning_rate": 7.31966543304406e-07,
+      "loss": 0.7558,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8315424610051994,
+      "grad_norm": 4.533492722150875,
+      "learning_rate": 7.290448019113705e-07,
+      "loss": 0.826,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8318890814558059,
+      "grad_norm": 4.11634374450872,
+      "learning_rate": 7.261284448188994e-07,
+      "loss": 0.7873,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8322357019064125,
+      "grad_norm": 3.723689822007307,
+      "learning_rate": 7.232174757035804e-07,
+      "loss": 0.7463,
+      "step": 2401
+    },
+    {
+      "epoch": 0.832582322357019,
+      "grad_norm": 3.945831839128604,
+      "learning_rate": 7.203118982352086e-07,
+      "loss": 0.7956,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8329289428076256,
+      "grad_norm": 3.0008631555825485,
+      "learning_rate": 7.17411716076783e-07,
+      "loss": 0.747,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8332755632582323,
+      "grad_norm": 2.3260419953257125,
+      "learning_rate": 7.145169328845003e-07,
+      "loss": 0.6953,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8336221837088388,
+      "grad_norm": 7.119223989644115,
+      "learning_rate": 7.116275523077504e-07,
+      "loss": 0.7626,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8339688041594454,
+      "grad_norm": 3.238997588492064,
+      "learning_rate": 7.087435779891145e-07,
+      "loss": 0.7035,
+      "step": 2406
+    },
+    {
+      "epoch": 0.834315424610052,
+      "grad_norm": 3.910813911010553,
+      "learning_rate": 7.058650135643541e-07,
+      "loss": 0.7654,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8346620450606586,
+      "grad_norm": 3.4977422590709786,
+      "learning_rate": 7.029918626624122e-07,
+      "loss": 0.667,
+      "step": 2408
+    },
+    {
+      "epoch": 0.8350086655112652,
+      "grad_norm": 4.06671596419918,
+      "learning_rate": 7.001241289054111e-07,
+      "loss": 0.6274,
+      "step": 2409
+    },
+    {
+      "epoch": 0.8353552859618717,
+      "grad_norm": 24.782328441264667,
+      "learning_rate": 6.972618159086387e-07,
+      "loss": 0.734,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8357019064124783,
+      "grad_norm": 3.657154578207864,
+      "learning_rate": 6.94404927280552e-07,
+      "loss": 0.842,
+      "step": 2411
+    },
+    {
+      "epoch": 0.836048526863085,
+      "grad_norm": 22.440060071561106,
+      "learning_rate": 6.915534666227669e-07,
+      "loss": 0.7069,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8363951473136915,
+      "grad_norm": 4.206127282145511,
+      "learning_rate": 6.887074375300584e-07,
+      "loss": 0.6797,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8367417677642981,
+      "grad_norm": 3.0934868503302786,
+      "learning_rate": 6.858668435903543e-07,
+      "loss": 0.7352,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8370883882149047,
+      "grad_norm": 10.067241186629255,
+      "learning_rate": 6.830316883847282e-07,
+      "loss": 0.6838,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8374350086655112,
+      "grad_norm": 2.88821481190042,
+      "learning_rate": 6.802019754874017e-07,
+      "loss": 0.7787,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8377816291161179,
+      "grad_norm": 5.1593672153624714,
+      "learning_rate": 6.773777084657302e-07,
+      "loss": 0.8363,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8381282495667244,
+      "grad_norm": 2.306875956209282,
+      "learning_rate": 6.745588908802064e-07,
+      "loss": 0.6886,
+      "step": 2418
+    },
+    {
+      "epoch": 0.838474870017331,
+      "grad_norm": 3.0259978986900884,
+      "learning_rate": 6.717455262844519e-07,
+      "loss": 0.7549,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8388214904679376,
+      "grad_norm": 3.906330692320804,
+      "learning_rate": 6.689376182252155e-07,
+      "loss": 0.7391,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8391681109185442,
+      "grad_norm": 3.8867152495455515,
+      "learning_rate": 6.661351702423652e-07,
+      "loss": 0.7061,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8395147313691508,
+      "grad_norm": 4.620441145252264,
+      "learning_rate": 6.63338185868887e-07,
+      "loss": 0.7909,
+      "step": 2422
+    },
+    {
+      "epoch": 0.8398613518197574,
+      "grad_norm": 3.5659205100369773,
+      "learning_rate": 6.605466686308776e-07,
+      "loss": 0.7874,
+      "step": 2423
+    },
+    {
+      "epoch": 0.8402079722703639,
+      "grad_norm": 3.5570420623897276,
+      "learning_rate": 6.57760622047543e-07,
+      "loss": 0.7908,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8405545927209706,
+      "grad_norm": 3.3613302340610485,
+      "learning_rate": 6.549800496311909e-07,
+      "loss": 0.7387,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8409012131715772,
+      "grad_norm": 6.230286228539772,
+      "learning_rate": 6.522049548872305e-07,
+      "loss": 0.7739,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8412478336221837,
+      "grad_norm": 9.794565657636808,
+      "learning_rate": 6.494353413141613e-07,
+      "loss": 0.7269,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8415944540727903,
+      "grad_norm": 5.781733908093758,
+      "learning_rate": 6.466712124035745e-07,
+      "loss": 0.7281,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8419410745233968,
+      "grad_norm": 3.107230270479234,
+      "learning_rate": 6.439125716401501e-07,
+      "loss": 0.7731,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8422876949740035,
+      "grad_norm": 5.203385207849268,
+      "learning_rate": 6.411594225016449e-07,
+      "loss": 0.7744,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8426343154246101,
+      "grad_norm": 2.4438031472252892,
+      "learning_rate": 6.384117684588964e-07,
+      "loss": 0.7977,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8429809358752166,
+      "grad_norm": 3.7127140786251043,
+      "learning_rate": 6.356696129758094e-07,
+      "loss": 0.6956,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8433275563258232,
+      "grad_norm": 3.0763962025419764,
+      "learning_rate": 6.329329595093614e-07,
+      "loss": 0.7214,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8436741767764299,
+      "grad_norm": 4.121643831004832,
+      "learning_rate": 6.30201811509592e-07,
+      "loss": 0.7076,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8440207972270364,
+      "grad_norm": 3.9553465243774144,
+      "learning_rate": 6.274761724195994e-07,
+      "loss": 0.8171,
+      "step": 2435
+    },
+    {
+      "epoch": 0.844367417677643,
+      "grad_norm": 4.027469900882762,
+      "learning_rate": 6.247560456755397e-07,
+      "loss": 0.7207,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8447140381282495,
+      "grad_norm": 4.145222176247925,
+      "learning_rate": 6.220414347066162e-07,
+      "loss": 0.689,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8450606585788561,
+      "grad_norm": 3.9019231825423644,
+      "learning_rate": 6.193323429350795e-07,
+      "loss": 0.8156,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8454072790294628,
+      "grad_norm": 4.7863525663134645,
+      "learning_rate": 6.166287737762239e-07,
+      "loss": 0.6447,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8457538994800693,
+      "grad_norm": 5.857748531346675,
+      "learning_rate": 6.139307306383802e-07,
+      "loss": 0.8802,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8461005199306759,
+      "grad_norm": 4.840581815109347,
+      "learning_rate": 6.112382169229125e-07,
+      "loss": 0.714,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8464471403812825,
+      "grad_norm": 5.737339308716882,
+      "learning_rate": 6.085512360242146e-07,
+      "loss": 0.6261,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8467937608318891,
+      "grad_norm": 5.388844980611621,
+      "learning_rate": 6.058697913297057e-07,
+      "loss": 0.7541,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8471403812824957,
+      "grad_norm": 2.167462436855007,
+      "learning_rate": 6.031938862198233e-07,
+      "loss": 0.7633,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8474870017331022,
+      "grad_norm": 2.607130476092118,
+      "learning_rate": 6.005235240680246e-07,
+      "loss": 0.8716,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8478336221837088,
+      "grad_norm": 2.8528182631730066,
+      "learning_rate": 5.978587082407766e-07,
+      "loss": 0.722,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8481802426343155,
+      "grad_norm": 3.8907052574739143,
+      "learning_rate": 5.95199442097556e-07,
+      "loss": 0.7156,
+      "step": 2447
+    },
+    {
+      "epoch": 0.848526863084922,
+      "grad_norm": 2.476263768916597,
+      "learning_rate": 5.925457289908382e-07,
+      "loss": 0.6841,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8488734835355286,
+      "grad_norm": 4.121793878799829,
+      "learning_rate": 5.898975722661054e-07,
+      "loss": 0.5723,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8492201039861352,
+      "grad_norm": 3.0338066876402645,
+      "learning_rate": 5.872549752618301e-07,
+      "loss": 0.7034,
+      "step": 2450
+    },
+    {
+      "epoch": 0.8495667244367417,
+      "grad_norm": 4.322766908610293,
+      "learning_rate": 5.846179413094771e-07,
+      "loss": 0.7597,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8499133448873484,
+      "grad_norm": 9.325932806855299,
+      "learning_rate": 5.81986473733499e-07,
+      "loss": 0.7514,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8502599653379549,
+      "grad_norm": 3.223238927050487,
+      "learning_rate": 5.793605758513282e-07,
+      "loss": 0.8197,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8506065857885615,
+      "grad_norm": 14.583305315004958,
+      "learning_rate": 5.767402509733771e-07,
+      "loss": 0.8325,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8509532062391681,
+      "grad_norm": 11.21254574583326,
+      "learning_rate": 5.74125502403034e-07,
+      "loss": 0.8028,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8512998266897747,
+      "grad_norm": 4.709124748294657,
+      "learning_rate": 5.71516333436653e-07,
+      "loss": 0.8839,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8516464471403813,
+      "grad_norm": 6.24989333918621,
+      "learning_rate": 5.689127473635609e-07,
+      "loss": 0.6998,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8519930675909879,
+      "grad_norm": 3.958822124483661,
+      "learning_rate": 5.663147474660391e-07,
+      "loss": 0.8401,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8523396880415944,
+      "grad_norm": 3.0531307290094367,
+      "learning_rate": 5.637223370193301e-07,
+      "loss": 0.7232,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8526863084922011,
+      "grad_norm": 3.43253130268789,
+      "learning_rate": 5.611355192916302e-07,
+      "loss": 0.6285,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8530329289428076,
+      "grad_norm": 4.017074624816902,
+      "learning_rate": 5.585542975440838e-07,
+      "loss": 0.6791,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8533795493934142,
+      "grad_norm": 6.203692996649151,
+      "learning_rate": 5.559786750307822e-07,
+      "loss": 0.7685,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8537261698440208,
+      "grad_norm": 4.0406788954652875,
+      "learning_rate": 5.534086549987555e-07,
+      "loss": 0.761,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8540727902946273,
+      "grad_norm": 5.813439582505362,
+      "learning_rate": 5.508442406879727e-07,
+      "loss": 0.6594,
+      "step": 2464
+    },
+    {
+      "epoch": 0.854419410745234,
+      "grad_norm": 4.159560667256528,
+      "learning_rate": 5.48285435331336e-07,
+      "loss": 0.5887,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8547660311958406,
+      "grad_norm": 3.7940889162166025,
+      "learning_rate": 5.457322421546746e-07,
+      "loss": 0.8125,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8551126516464471,
+      "grad_norm": 6.454729610072989,
+      "learning_rate": 5.431846643767458e-07,
+      "loss": 0.6474,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8554592720970537,
+      "grad_norm": 4.245068705315487,
+      "learning_rate": 5.406427052092217e-07,
+      "loss": 0.7734,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8558058925476603,
+      "grad_norm": 4.915742560522922,
+      "learning_rate": 5.381063678566989e-07,
+      "loss": 0.8748,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8561525129982669,
+      "grad_norm": 4.938903859915773,
+      "learning_rate": 5.355756555166813e-07,
+      "loss": 0.7267,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8564991334488735,
+      "grad_norm": 3.1168165028602264,
+      "learning_rate": 5.330505713795836e-07,
+      "loss": 0.7477,
+      "step": 2471
+    },
+    {
+      "epoch": 0.85684575389948,
+      "grad_norm": 3.13519696699903,
+      "learning_rate": 5.305311186287254e-07,
+      "loss": 0.783,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8571923743500867,
+      "grad_norm": 7.845747813448013,
+      "learning_rate": 5.280173004403239e-07,
+      "loss": 0.7099,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8575389948006933,
+      "grad_norm": 3.950068248825599,
+      "learning_rate": 5.255091199834955e-07,
+      "loss": 0.693,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8578856152512998,
+      "grad_norm": 3.2114998323362505,
+      "learning_rate": 5.230065804202506e-07,
+      "loss": 0.7049,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8582322357019064,
+      "grad_norm": 21.750860986876045,
+      "learning_rate": 5.205096849054842e-07,
+      "loss": 0.7436,
+      "step": 2476
+    },
+    {
+      "epoch": 0.858578856152513,
+      "grad_norm": 2.616922824171401,
+      "learning_rate": 5.180184365869806e-07,
+      "loss": 0.6725,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8589254766031196,
+      "grad_norm": 5.696467103470003,
+      "learning_rate": 5.155328386054026e-07,
+      "loss": 0.7424,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8592720970537262,
+      "grad_norm": 5.398297700542328,
+      "learning_rate": 5.13052894094288e-07,
+      "loss": 0.7335,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8596187175043327,
+      "grad_norm": 5.83490189316114,
+      "learning_rate": 5.105786061800494e-07,
+      "loss": 0.6671,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8599653379549393,
+      "grad_norm": 3.7609716281001457,
+      "learning_rate": 5.081099779819682e-07,
+      "loss": 0.7485,
+      "step": 2481
+    },
+    {
+      "epoch": 0.860311958405546,
+      "grad_norm": 11.07686327258226,
+      "learning_rate": 5.056470126121904e-07,
+      "loss": 0.6943,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8606585788561525,
+      "grad_norm": 2.3348525504879465,
+      "learning_rate": 5.03189713175723e-07,
+      "loss": 0.6712,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8610051993067591,
+      "grad_norm": 4.208842615327125,
+      "learning_rate": 5.0073808277043e-07,
+      "loss": 0.7029,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8613518197573656,
+      "grad_norm": 2.6850788993412094,
+      "learning_rate": 4.982921244870292e-07,
+      "loss": 0.7265,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8616984402079723,
+      "grad_norm": 2.4020713011041925,
+      "learning_rate": 4.958518414090863e-07,
+      "loss": 0.6757,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8620450606585789,
+      "grad_norm": 3.211039150321696,
+      "learning_rate": 4.934172366130141e-07,
+      "loss": 0.6832,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8623916811091854,
+      "grad_norm": 14.468279706585307,
+      "learning_rate": 4.90988313168066e-07,
+      "loss": 0.7306,
+      "step": 2488
+    },
+    {
+      "epoch": 0.862738301559792,
+      "grad_norm": 3.0083643120421217,
+      "learning_rate": 4.885650741363312e-07,
+      "loss": 0.6346,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8630849220103987,
+      "grad_norm": 2.533802864738133,
+      "learning_rate": 4.861475225727369e-07,
+      "loss": 0.6696,
+      "step": 2490
+    },
+    {
+      "epoch": 0.8634315424610052,
+      "grad_norm": 4.535688164917354,
+      "learning_rate": 4.837356615250372e-07,
+      "loss": 0.6647,
+      "step": 2491
+    },
+    {
+      "epoch": 0.8637781629116118,
+      "grad_norm": 4.050126307308121,
+      "learning_rate": 4.813294940338126e-07,
+      "loss": 0.723,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8641247833622183,
+      "grad_norm": 3.190403681476133,
+      "learning_rate": 4.789290231324678e-07,
+      "loss": 0.782,
+      "step": 2493
+    },
+    {
+      "epoch": 0.864471403812825,
+      "grad_norm": 4.400681079828735,
+      "learning_rate": 4.765342518472221e-07,
+      "loss": 0.8546,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8648180242634316,
+      "grad_norm": 3.560905393199567,
+      "learning_rate": 4.7414518319711276e-07,
+      "loss": 0.7229,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8651646447140381,
+      "grad_norm": 51.77238010271364,
+      "learning_rate": 4.7176182019398555e-07,
+      "loss": 0.6925,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8655112651646447,
+      "grad_norm": 14.79828622898179,
+      "learning_rate": 4.6938416584249693e-07,
+      "loss": 0.6906,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8658578856152513,
+      "grad_norm": 3.54957701637997,
+      "learning_rate": 4.6701222314010366e-07,
+      "loss": 0.7877,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8662045060658579,
+      "grad_norm": 2.467251554540918,
+      "learning_rate": 4.64645995077061e-07,
+      "loss": 0.7277,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8665511265164645,
+      "grad_norm": 2.25721994842977,
+      "learning_rate": 4.622854846364222e-07,
+      "loss": 0.6853,
+      "step": 2500
+    },
+    {
+      "epoch": 0.866897746967071,
+      "grad_norm": 4.633946429298019,
+      "learning_rate": 4.599306947940313e-07,
+      "loss": 0.8035,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8672443674176776,
+      "grad_norm": 5.7435136635191135,
+      "learning_rate": 4.5758162851852083e-07,
+      "loss": 0.8127,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8675909878682843,
+      "grad_norm": 7.104079333199232,
+      "learning_rate": 4.5523828877130807e-07,
+      "loss": 0.7656,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8679376083188908,
+      "grad_norm": 4.596576804734678,
+      "learning_rate": 4.5290067850659035e-07,
+      "loss": 0.7075,
+      "step": 2504
+    },
+    {
+      "epoch": 0.8682842287694974,
+      "grad_norm": 4.708714093883089,
+      "learning_rate": 4.5056880067134156e-07,
+      "loss": 0.7395,
+      "step": 2505
+    },
+    {
+      "epoch": 0.868630849220104,
+      "grad_norm": 3.97975777479811,
+      "learning_rate": 4.482426582053107e-07,
+      "loss": 0.702,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8689774696707105,
+      "grad_norm": 6.060099773746487,
+      "learning_rate": 4.4592225404101464e-07,
+      "loss": 0.7747,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8693240901213172,
+      "grad_norm": 5.704054361532903,
+      "learning_rate": 4.4360759110373685e-07,
+      "loss": 0.7186,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8696707105719237,
+      "grad_norm": 2.7360012497311814,
+      "learning_rate": 4.412986723115209e-07,
+      "loss": 0.6703,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8700173310225303,
+      "grad_norm": 5.742286291704321,
+      "learning_rate": 4.3899550057517225e-07,
+      "loss": 0.6885,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8703639514731369,
+      "grad_norm": 3.6261374569683897,
+      "learning_rate": 4.3669807879825e-07,
+      "loss": 0.7927,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8707105719237435,
+      "grad_norm": 3.889536597589798,
+      "learning_rate": 4.3440640987706284e-07,
+      "loss": 0.7491,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8710571923743501,
+      "grad_norm": 6.42887217019907,
+      "learning_rate": 4.321204967006687e-07,
+      "loss": 0.7189,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8714038128249567,
+      "grad_norm": 5.429662546341451,
+      "learning_rate": 4.2984034215086955e-07,
+      "loss": 0.7444,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8717504332755632,
+      "grad_norm": 5.395209301497287,
+      "learning_rate": 4.275659491022044e-07,
+      "loss": 0.6887,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8720970537261699,
+      "grad_norm": 2.911802352450472,
+      "learning_rate": 4.2529732042195184e-07,
+      "loss": 0.7959,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8724436741767765,
+      "grad_norm": 2.209075052605805,
+      "learning_rate": 4.2303445897012406e-07,
+      "loss": 0.586,
+      "step": 2517
+    },
+    {
+      "epoch": 0.872790294627383,
+      "grad_norm": 2.4313594964217535,
+      "learning_rate": 4.2077736759946143e-07,
+      "loss": 0.6653,
+      "step": 2518
+    },
+    {
+      "epoch": 0.8731369150779896,
+      "grad_norm": 3.4112902114908548,
+      "learning_rate": 4.185260491554305e-07,
+      "loss": 0.6612,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8734835355285961,
+      "grad_norm": 3.9067945520884115,
+      "learning_rate": 4.162805064762171e-07,
+      "loss": 0.7973,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8738301559792028,
+      "grad_norm": 3.8959086823988884,
+      "learning_rate": 4.1404074239273007e-07,
+      "loss": 0.8242,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8741767764298094,
+      "grad_norm": 3.3727453715493905,
+      "learning_rate": 4.118067597285907e-07,
+      "loss": 0.7632,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8745233968804159,
+      "grad_norm": 3.462667226787518,
+      "learning_rate": 4.0957856130013354e-07,
+      "loss": 0.7553,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8748700173310225,
+      "grad_norm": 3.394698534200617,
+      "learning_rate": 4.073561499163986e-07,
+      "loss": 0.7258,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8752166377816292,
+      "grad_norm": 3.534216400703507,
+      "learning_rate": 4.051395283791321e-07,
+      "loss": 0.8033,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8755632582322357,
+      "grad_norm": 6.474228149367421,
+      "learning_rate": 4.0292869948278137e-07,
+      "loss": 0.7708,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8759098786828423,
+      "grad_norm": 5.493578778966039,
+      "learning_rate": 4.007236660144903e-07,
+      "loss": 0.6646,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8762564991334488,
+      "grad_norm": 4.729686594293128,
+      "learning_rate": 3.985244307540959e-07,
+      "loss": 0.7524,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8766031195840555,
+      "grad_norm": 17.332919906433254,
+      "learning_rate": 3.9633099647412745e-07,
+      "loss": 0.7111,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8769497400346621,
+      "grad_norm": 2.9920179519086387,
+      "learning_rate": 3.941433659397992e-07,
+      "loss": 0.5412,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8772963604852686,
+      "grad_norm": 4.247369339561204,
+      "learning_rate": 3.9196154190901003e-07,
+      "loss": 0.6008,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8776429809358752,
+      "grad_norm": 5.232830877001447,
+      "learning_rate": 3.8978552713233754e-07,
+      "loss": 0.7383,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8779896013864819,
+      "grad_norm": 7.773032379663242,
+      "learning_rate": 3.876153243530367e-07,
+      "loss": 0.743,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8783362218370884,
+      "grad_norm": 2.609168313574987,
+      "learning_rate": 3.854509363070363e-07,
+      "loss": 0.7871,
+      "step": 2534
+    },
+    {
+      "epoch": 0.878682842287695,
+      "grad_norm": 3.9183548635584837,
+      "learning_rate": 3.8329236572293096e-07,
+      "loss": 0.6997,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8790294627383015,
+      "grad_norm": 3.034919883004722,
+      "learning_rate": 3.811396153219854e-07,
+      "loss": 0.5854,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8793760831889081,
+      "grad_norm": 20.294293085058577,
+      "learning_rate": 3.789926878181238e-07,
+      "loss": 0.7175,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8797227036395148,
+      "grad_norm": 3.415787377296968,
+      "learning_rate": 3.768515859179328e-07,
+      "loss": 0.7556,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8800693240901213,
+      "grad_norm": 3.1703754257626935,
+      "learning_rate": 3.7471631232065377e-07,
+      "loss": 0.6994,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8804159445407279,
+      "grad_norm": 5.2962287024763235,
+      "learning_rate": 3.725868697181778e-07,
+      "loss": 0.7863,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8807625649913345,
+      "grad_norm": 5.332605512873317,
+      "learning_rate": 3.704632607950476e-07,
+      "loss": 0.8074,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8811091854419411,
+      "grad_norm": 3.860882763663691,
+      "learning_rate": 3.683454882284515e-07,
+      "loss": 0.8635,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8814558058925477,
+      "grad_norm": 5.025214489764415,
+      "learning_rate": 3.662335546882184e-07,
+      "loss": 0.8562,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8818024263431542,
+      "grad_norm": 4.124897048871715,
+      "learning_rate": 3.6412746283681797e-07,
+      "loss": 0.8502,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8821490467937608,
+      "grad_norm": 6.547750015838163,
+      "learning_rate": 3.620272153293541e-07,
+      "loss": 0.693,
+      "step": 2545
+    },
+    {
+      "epoch": 0.8824956672443675,
+      "grad_norm": 4.254764871367222,
+      "learning_rate": 3.5993281481356335e-07,
+      "loss": 0.8073,
+      "step": 2546
+    },
+    {
+      "epoch": 0.882842287694974,
+      "grad_norm": 3.046296751736737,
+      "learning_rate": 3.578442639298113e-07,
+      "loss": 0.6603,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8831889081455806,
+      "grad_norm": 4.906069052126101,
+      "learning_rate": 3.5576156531108795e-07,
+      "loss": 0.768,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8835355285961872,
+      "grad_norm": 2.9510927407392646,
+      "learning_rate": 3.536847215830075e-07,
+      "loss": 0.7421,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8838821490467937,
+      "grad_norm": 3.0241529823044044,
+      "learning_rate": 3.516137353638005e-07,
+      "loss": 0.6533,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8842287694974004,
+      "grad_norm": 3.186307479562003,
+      "learning_rate": 3.4954860926431524e-07,
+      "loss": 0.8703,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8845753899480069,
+      "grad_norm": 2.7702867759425813,
+      "learning_rate": 3.474893458880113e-07,
+      "loss": 0.8009,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8849220103986135,
+      "grad_norm": 5.582939674535199,
+      "learning_rate": 3.454359478309566e-07,
+      "loss": 0.7553,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8852686308492201,
+      "grad_norm": 6.104742545620451,
+      "learning_rate": 3.433884176818264e-07,
+      "loss": 0.8428,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8856152512998267,
+      "grad_norm": 3.5296797057844387,
+      "learning_rate": 3.413467580218982e-07,
+      "loss": 0.6594,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8859618717504333,
+      "grad_norm": 3.435879561339312,
+      "learning_rate": 3.3931097142504586e-07,
+      "loss": 0.7966,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8863084922010399,
+      "grad_norm": 6.234559986723826,
+      "learning_rate": 3.37281060457742e-07,
+      "loss": 0.7503,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8866551126516464,
+      "grad_norm": 3.937684780195544,
+      "learning_rate": 3.3525702767905235e-07,
+      "loss": 0.7468,
+      "step": 2558
+    },
+    {
+      "epoch": 0.887001733102253,
+      "grad_norm": 5.677104946391772,
+      "learning_rate": 3.332388756406313e-07,
+      "loss": 0.6878,
+      "step": 2559
+    },
+    {
+      "epoch": 0.8873483535528596,
+      "grad_norm": 4.03139789363777,
+      "learning_rate": 3.312266068867187e-07,
+      "loss": 0.7924,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8876949740034662,
+      "grad_norm": 4.8499312721062875,
+      "learning_rate": 3.292202239541381e-07,
+      "loss": 0.8499,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8880415944540728,
+      "grad_norm": 6.084868760757525,
+      "learning_rate": 3.2721972937229227e-07,
+      "loss": 0.7266,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8883882149046793,
+      "grad_norm": 3.1061560123352154,
+      "learning_rate": 3.252251256631622e-07,
+      "loss": 0.7271,
+      "step": 2563
+    },
+    {
+      "epoch": 0.888734835355286,
+      "grad_norm": 5.6313912265923545,
+      "learning_rate": 3.232364153413009e-07,
+      "loss": 0.7654,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8890814558058926,
+      "grad_norm": 3.8043531697651067,
+      "learning_rate": 3.2125360091383397e-07,
+      "loss": 0.79,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8894280762564991,
+      "grad_norm": 4.783941830031844,
+      "learning_rate": 3.192766848804507e-07,
+      "loss": 0.85,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8897746967071057,
+      "grad_norm": 3.830871027475743,
+      "learning_rate": 3.1730566973340706e-07,
+      "loss": 0.6889,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8901213171577123,
+      "grad_norm": 2.3799617740110715,
+      "learning_rate": 3.1534055795751916e-07,
+      "loss": 0.7096,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8904679376083189,
+      "grad_norm": 4.005661115199528,
+      "learning_rate": 3.133813520301598e-07,
+      "loss": 0.6524,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8908145580589255,
+      "grad_norm": 3.775497363444591,
+      "learning_rate": 3.114280544212589e-07,
+      "loss": 0.7721,
+      "step": 2570
+    },
+    {
+      "epoch": 0.891161178509532,
+      "grad_norm": 3.1270144608004906,
+      "learning_rate": 3.094806675932949e-07,
+      "loss": 0.7134,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8915077989601387,
+      "grad_norm": 3.1306926595233877,
+      "learning_rate": 3.075391940012973e-07,
+      "loss": 0.7202,
+      "step": 2572
+    },
+    {
+      "epoch": 0.8918544194107453,
+      "grad_norm": 3.104619398394467,
+      "learning_rate": 3.056036360928388e-07,
+      "loss": 0.689,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8922010398613518,
+      "grad_norm": 3.0907173237910706,
+      "learning_rate": 3.036739963080365e-07,
+      "loss": 0.7333,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8925476603119584,
+      "grad_norm": 3.406607465098224,
+      "learning_rate": 3.01750277079545e-07,
+      "loss": 0.789,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8928942807625649,
+      "grad_norm": 2.979917780230461,
+      "learning_rate": 2.9983248083255424e-07,
+      "loss": 0.5928,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8932409012131716,
+      "grad_norm": 4.402581622136268,
+      "learning_rate": 2.9792060998478813e-07,
+      "loss": 0.6923,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8935875216637782,
+      "grad_norm": 2.8131252438297634,
+      "learning_rate": 2.9601466694650304e-07,
+      "loss": 0.8034,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8939341421143847,
+      "grad_norm": 4.567222873063261,
+      "learning_rate": 2.9411465412047866e-07,
+      "loss": 0.7962,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8942807625649913,
+      "grad_norm": 4.107172024956197,
+      "learning_rate": 2.9222057390202065e-07,
+      "loss": 0.6501,
+      "step": 2580
+    },
+    {
+      "epoch": 0.894627383015598,
+      "grad_norm": 4.1807276524780566,
+      "learning_rate": 2.903324286789533e-07,
+      "loss": 0.7174,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8949740034662045,
+      "grad_norm": 4.791677694246326,
+      "learning_rate": 2.88450220831622e-07,
+      "loss": 0.7014,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8953206239168111,
+      "grad_norm": 6.751789112784741,
+      "learning_rate": 2.865739527328837e-07,
+      "loss": 0.6893,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8956672443674176,
+      "grad_norm": 8.000138560341753,
+      "learning_rate": 2.847036267481107e-07,
+      "loss": 0.6728,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8960138648180243,
+      "grad_norm": 5.788889649268189,
+      "learning_rate": 2.8283924523518026e-07,
+      "loss": 0.7595,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8963604852686309,
+      "grad_norm": 6.2536891237997505,
+      "learning_rate": 2.8098081054448055e-07,
+      "loss": 0.6498,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8967071057192374,
+      "grad_norm": 3.8715277392441996,
+      "learning_rate": 2.791283250188975e-07,
+      "loss": 0.783,
+      "step": 2587
+    },
+    {
+      "epoch": 0.897053726169844,
+      "grad_norm": 4.799715657484661,
+      "learning_rate": 2.7728179099382025e-07,
+      "loss": 0.7187,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8974003466204507,
+      "grad_norm": 4.458007431101365,
+      "learning_rate": 2.754412107971344e-07,
+      "loss": 0.6981,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8977469670710572,
+      "grad_norm": 3.6665680764758997,
+      "learning_rate": 2.7360658674921883e-07,
+      "loss": 0.7969,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8980935875216638,
+      "grad_norm": 5.241324101340116,
+      "learning_rate": 2.717779211629451e-07,
+      "loss": 0.6756,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8984402079722703,
+      "grad_norm": 6.214882846007423,
+      "learning_rate": 2.6995521634367136e-07,
+      "loss": 0.8003,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8987868284228769,
+      "grad_norm": 4.861825627682455,
+      "learning_rate": 2.681384745892424e-07,
+      "loss": 0.6549,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8991334488734836,
+      "grad_norm": 4.669586957696573,
+      "learning_rate": 2.663276981899854e-07,
+      "loss": 0.8361,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8994800693240901,
+      "grad_norm": 3.9383594796783483,
+      "learning_rate": 2.645228894287072e-07,
+      "loss": 0.6969,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8998266897746967,
+      "grad_norm": 8.960011221634625,
+      "learning_rate": 2.627240505806905e-07,
+      "loss": 0.7729,
+      "step": 2596
+    },
+    {
+      "epoch": 0.9001733102253033,
+      "grad_norm": 10.185554222716833,
+      "learning_rate": 2.609311839136924e-07,
+      "loss": 0.7324,
+      "step": 2597
+    },
+    {
+      "epoch": 0.9005199306759099,
+      "grad_norm": 2.7934609327686415,
+      "learning_rate": 2.5914429168793997e-07,
+      "loss": 0.8468,
+      "step": 2598
+    },
+    {
+      "epoch": 0.9008665511265165,
+      "grad_norm": 5.858761756429254,
+      "learning_rate": 2.5736337615613114e-07,
+      "loss": 0.6702,
+      "step": 2599
+    },
+    {
+      "epoch": 0.901213171577123,
+      "grad_norm": 3.4267383986558766,
+      "learning_rate": 2.555884395634273e-07,
+      "loss": 0.7397,
+      "step": 2600
+    },
+    {
+      "epoch": 0.9015597920277296,
+      "grad_norm": 4.283703638812449,
+      "learning_rate": 2.5381948414745283e-07,
+      "loss": 0.742,
+      "step": 2601
+    },
+    {
+      "epoch": 0.9019064124783363,
+      "grad_norm": 29.14259539171556,
+      "learning_rate": 2.520565121382895e-07,
+      "loss": 0.7505,
+      "step": 2602
+    },
+    {
+      "epoch": 0.9022530329289428,
+      "grad_norm": 3.261513367025463,
+      "learning_rate": 2.5029952575848003e-07,
+      "loss": 0.8016,
+      "step": 2603
+    },
+    {
+      "epoch": 0.9025996533795494,
+      "grad_norm": 3.0756128129074716,
+      "learning_rate": 2.4854852722301757e-07,
+      "loss": 0.7417,
+      "step": 2604
+    },
+    {
+      "epoch": 0.902946273830156,
+      "grad_norm": 3.4934395908376565,
+      "learning_rate": 2.468035187393475e-07,
+      "loss": 0.6277,
+      "step": 2605
+    },
+    {
+      "epoch": 0.9032928942807625,
+      "grad_norm": 5.364803903157824,
+      "learning_rate": 2.4506450250736725e-07,
+      "loss": 0.8467,
+      "step": 2606
+    },
+    {
+      "epoch": 0.9036395147313692,
+      "grad_norm": 4.457014586474601,
+      "learning_rate": 2.433314807194137e-07,
+      "loss": 0.828,
+      "step": 2607
+    },
+    {
+      "epoch": 0.9039861351819758,
+      "grad_norm": 9.105790952047052,
+      "learning_rate": 2.4160445556027143e-07,
+      "loss": 0.7914,
+      "step": 2608
+    },
+    {
+      "epoch": 0.9043327556325823,
+      "grad_norm": 4.122984027027024,
+      "learning_rate": 2.3988342920716276e-07,
+      "loss": 0.7407,
+      "step": 2609
+    },
+    {
+      "epoch": 0.9046793760831889,
+      "grad_norm": 3.9399169064676394,
+      "learning_rate": 2.3816840382974826e-07,
+      "loss": 0.7953,
+      "step": 2610
+    },
+    {
+      "epoch": 0.9050259965337955,
+      "grad_norm": 3.166490981154585,
+      "learning_rate": 2.3645938159012393e-07,
+      "loss": 0.7274,
+      "step": 2611
+    },
+    {
+      "epoch": 0.9053726169844021,
+      "grad_norm": 3.294059505957102,
+      "learning_rate": 2.3475636464281691e-07,
+      "loss": 0.6605,
+      "step": 2612
+    },
+    {
+      "epoch": 0.9057192374350087,
+      "grad_norm": 3.6632269933789323,
+      "learning_rate": 2.3305935513478318e-07,
+      "loss": 0.7605,
+      "step": 2613
+    },
+    {
+      "epoch": 0.9060658578856152,
+      "grad_norm": 5.4697412468162065,
+      "learning_rate": 2.3136835520540634e-07,
+      "loss": 0.7923,
+      "step": 2614
+    },
+    {
+      "epoch": 0.9064124783362218,
+      "grad_norm": 3.2598272559587183,
+      "learning_rate": 2.2968336698649286e-07,
+      "loss": 0.8117,
+      "step": 2615
+    },
+    {
+      "epoch": 0.9067590987868285,
+      "grad_norm": 8.304637760857105,
+      "learning_rate": 2.2800439260227182e-07,
+      "loss": 0.6208,
+      "step": 2616
+    },
+    {
+      "epoch": 0.907105719237435,
+      "grad_norm": 5.907154001581199,
+      "learning_rate": 2.263314341693884e-07,
+      "loss": 0.7483,
+      "step": 2617
+    },
+    {
+      "epoch": 0.9074523396880416,
+      "grad_norm": 4.081734171776198,
+      "learning_rate": 2.2466449379690503e-07,
+      "loss": 0.8066,
+      "step": 2618
+    },
+    {
+      "epoch": 0.9077989601386481,
+      "grad_norm": 3.3274037354330246,
+      "learning_rate": 2.2300357358629843e-07,
+      "loss": 0.7202,
+      "step": 2619
+    },
+    {
+      "epoch": 0.9081455805892548,
+      "grad_norm": 4.319589236665431,
+      "learning_rate": 2.2134867563145424e-07,
+      "loss": 0.7164,
+      "step": 2620
+    },
+    {
+      "epoch": 0.9084922010398614,
+      "grad_norm": 3.820239855559258,
+      "learning_rate": 2.1969980201866637e-07,
+      "loss": 0.6861,
+      "step": 2621
+    },
+    {
+      "epoch": 0.9088388214904679,
+      "grad_norm": 3.076000166453436,
+      "learning_rate": 2.1805695482663425e-07,
+      "loss": 0.7236,
+      "step": 2622
+    },
+    {
+      "epoch": 0.9091854419410745,
+      "grad_norm": 4.642795068459365,
+      "learning_rate": 2.1642013612645897e-07,
+      "loss": 0.6892,
+      "step": 2623
+    },
+    {
+      "epoch": 0.9095320623916812,
+      "grad_norm": 2.5224168132604725,
+      "learning_rate": 2.147893479816432e-07,
+      "loss": 0.7992,
+      "step": 2624
+    },
+    {
+      "epoch": 0.9098786828422877,
+      "grad_norm": 4.605376199647157,
+      "learning_rate": 2.1316459244808463e-07,
+      "loss": 0.6414,
+      "step": 2625
+    },
+    {
+      "epoch": 0.9102253032928943,
+      "grad_norm": 3.220615475306782,
+      "learning_rate": 2.1154587157407924e-07,
+      "loss": 0.7042,
+      "step": 2626
+    },
+    {
+      "epoch": 0.9105719237435008,
+      "grad_norm": 5.000113903786568,
+      "learning_rate": 2.0993318740031298e-07,
+      "loss": 0.7786,
+      "step": 2627
+    },
+    {
+      "epoch": 0.9109185441941074,
+      "grad_norm": 3.9128526639747956,
+      "learning_rate": 2.0832654195986124e-07,
+      "loss": 0.8317,
+      "step": 2628
+    },
+    {
+      "epoch": 0.9112651646447141,
+      "grad_norm": 4.416138393916403,
+      "learning_rate": 2.0672593727818656e-07,
+      "loss": 0.6304,
+      "step": 2629
+    },
+    {
+      "epoch": 0.9116117850953206,
+      "grad_norm": 4.395072847915375,
+      "learning_rate": 2.0513137537313765e-07,
+      "loss": 0.686,
+      "step": 2630
+    },
+    {
+      "epoch": 0.9119584055459272,
+      "grad_norm": 11.186606002526581,
+      "learning_rate": 2.0354285825494313e-07,
+      "loss": 0.8096,
+      "step": 2631
+    },
+    {
+      "epoch": 0.9123050259965338,
+      "grad_norm": 3.918553090110496,
+      "learning_rate": 2.0196038792621276e-07,
+      "loss": 0.811,
+      "step": 2632
+    },
+    {
+      "epoch": 0.9126516464471404,
+      "grad_norm": 4.438957424237482,
+      "learning_rate": 2.0038396638193236e-07,
+      "loss": 0.6919,
+      "step": 2633
+    },
+    {
+      "epoch": 0.912998266897747,
+      "grad_norm": 7.610373152599236,
+      "learning_rate": 1.9881359560946224e-07,
+      "loss": 0.6914,
+      "step": 2634
+    },
+    {
+      "epoch": 0.9133448873483535,
+      "grad_norm": 14.175736869274267,
+      "learning_rate": 1.972492775885354e-07,
+      "loss": 0.6379,
+      "step": 2635
+    },
+    {
+      "epoch": 0.9136915077989601,
+      "grad_norm": 3.6206769017006954,
+      "learning_rate": 1.956910142912527e-07,
+      "loss": 0.7393,
+      "step": 2636
+    },
+    {
+      "epoch": 0.9140381282495668,
+      "grad_norm": 3.6962054619347904,
+      "learning_rate": 1.9413880768208382e-07,
+      "loss": 0.7469,
+      "step": 2637
+    },
+    {
+      "epoch": 0.9143847487001733,
+      "grad_norm": 4.136228485530534,
+      "learning_rate": 1.9259265971786068e-07,
+      "loss": 0.6096,
+      "step": 2638
+    },
+    {
+      "epoch": 0.9147313691507799,
+      "grad_norm": 3.831222653358788,
+      "learning_rate": 1.9105257234777907e-07,
+      "loss": 0.5942,
+      "step": 2639
+    },
+    {
+      "epoch": 0.9150779896013865,
+      "grad_norm": 4.006442239275451,
+      "learning_rate": 1.895185475133948e-07,
+      "loss": 0.687,
+      "step": 2640
+    },
+    {
+      "epoch": 0.915424610051993,
+      "grad_norm": 5.17387146455175,
+      "learning_rate": 1.8799058714861817e-07,
+      "loss": 0.7971,
+      "step": 2641
+    },
+    {
+      "epoch": 0.9157712305025997,
+      "grad_norm": 13.932877825545708,
+      "learning_rate": 1.8646869317971718e-07,
+      "loss": 0.7646,
+      "step": 2642
+    },
+    {
+      "epoch": 0.9161178509532062,
+      "grad_norm": 3.9911153971849913,
+      "learning_rate": 1.8495286752530873e-07,
+      "loss": 0.7087,
+      "step": 2643
+    },
+    {
+      "epoch": 0.9164644714038128,
+      "grad_norm": 6.969470535036437,
+      "learning_rate": 1.8344311209636256e-07,
+      "loss": 0.8025,
+      "step": 2644
+    },
+    {
+      "epoch": 0.9168110918544194,
+      "grad_norm": 2.922533420361791,
+      "learning_rate": 1.8193942879619397e-07,
+      "loss": 0.7856,
+      "step": 2645
+    },
+    {
+      "epoch": 0.917157712305026,
+      "grad_norm": 4.722486168368089,
+      "learning_rate": 1.804418195204638e-07,
+      "loss": 0.7406,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9175043327556326,
+      "grad_norm": 3.567518029092603,
+      "learning_rate": 1.789502861571768e-07,
+      "loss": 0.7824,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9178509532062392,
+      "grad_norm": 8.911993250017213,
+      "learning_rate": 1.774648305866755e-07,
+      "loss": 0.6454,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9181975736568457,
+      "grad_norm": 4.173601114421507,
+      "learning_rate": 1.7598545468164251e-07,
+      "loss": 0.704,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9185441941074524,
+      "grad_norm": 3.0824258409045715,
+      "learning_rate": 1.7451216030709372e-07,
+      "loss": 0.6282,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9188908145580589,
+      "grad_norm": 4.5650512375489605,
+      "learning_rate": 1.7304494932038007e-07,
+      "loss": 0.8136,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9192374350086655,
+      "grad_norm": 5.424216774699509,
+      "learning_rate": 1.7158382357118254e-07,
+      "loss": 0.6383,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9195840554592721,
+      "grad_norm": 4.401033011849953,
+      "learning_rate": 1.701287849015104e-07,
+      "loss": 0.7717,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9199306759098786,
+      "grad_norm": 8.388120030762613,
+      "learning_rate": 1.6867983514569973e-07,
+      "loss": 0.7172,
+      "step": 2654
+    },
+    {
+      "epoch": 0.9202772963604853,
+      "grad_norm": 11.49040569235851,
+      "learning_rate": 1.672369761304099e-07,
+      "loss": 0.8116,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9206239168110919,
+      "grad_norm": 3.701387507167072,
+      "learning_rate": 1.6580020967462252e-07,
+      "loss": 0.797,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9209705372616984,
+      "grad_norm": 4.287856435936848,
+      "learning_rate": 1.6436953758963702e-07,
+      "loss": 0.6755,
+      "step": 2657
+    },
+    {
+      "epoch": 0.921317157712305,
+      "grad_norm": 12.613898216580823,
+      "learning_rate": 1.6294496167907126e-07,
+      "loss": 0.6656,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9216637781629116,
+      "grad_norm": 4.0283758418281455,
+      "learning_rate": 1.6152648373885582e-07,
+      "loss": 0.6762,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9220103986135182,
+      "grad_norm": 34.75370584460937,
+      "learning_rate": 1.601141055572375e-07,
+      "loss": 0.8132,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9223570190641248,
+      "grad_norm": 4.727593283318156,
+      "learning_rate": 1.5870782891476867e-07,
+      "loss": 0.6992,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9227036395147313,
+      "grad_norm": 2.8350361787910994,
+      "learning_rate": 1.5730765558431283e-07,
+      "loss": 0.7319,
+      "step": 2662
+    },
+    {
+      "epoch": 0.923050259965338,
+      "grad_norm": 5.253241740043765,
+      "learning_rate": 1.5591358733103857e-07,
+      "loss": 0.6949,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9233968804159446,
+      "grad_norm": 8.53711913159159,
+      "learning_rate": 1.545256259124167e-07,
+      "loss": 0.7986,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9237435008665511,
+      "grad_norm": 4.371783155418492,
+      "learning_rate": 1.5314377307821982e-07,
+      "loss": 0.695,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9240901213171577,
+      "grad_norm": 3.9848331452237113,
+      "learning_rate": 1.5176803057051937e-07,
+      "loss": 0.7159,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9244367417677642,
+      "grad_norm": 5.9099562906969245,
+      "learning_rate": 1.5039840012368578e-07,
+      "loss": 0.699,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9247833622183709,
+      "grad_norm": 5.936234340429319,
+      "learning_rate": 1.4903488346438233e-07,
+      "loss": 0.6964,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9251299826689775,
+      "grad_norm": 12.060239437248494,
+      "learning_rate": 1.4767748231156287e-07,
+      "loss": 0.6627,
+      "step": 2669
+    },
+    {
+      "epoch": 0.925476603119584,
+      "grad_norm": 3.4441721251242194,
+      "learning_rate": 1.4632619837647577e-07,
+      "loss": 0.6685,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9258232235701906,
+      "grad_norm": 3.2389041621592085,
+      "learning_rate": 1.449810333626539e-07,
+      "loss": 0.7616,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9261698440207973,
+      "grad_norm": 2.811284818151649,
+      "learning_rate": 1.4364198896591851e-07,
+      "loss": 0.6818,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9265164644714038,
+      "grad_norm": 5.364605982573727,
+      "learning_rate": 1.4230906687437317e-07,
+      "loss": 0.6645,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9268630849220104,
+      "grad_norm": 8.693896368782752,
+      "learning_rate": 1.4098226876840426e-07,
+      "loss": 0.7853,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9272097053726169,
+      "grad_norm": 10.471017263379165,
+      "learning_rate": 1.396615963206771e-07,
+      "loss": 0.7793,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9275563258232236,
+      "grad_norm": 3.1619478291767997,
+      "learning_rate": 1.3834705119613544e-07,
+      "loss": 0.7806,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9279029462738302,
+      "grad_norm": 5.30100828929561,
+      "learning_rate": 1.3703863505199698e-07,
+      "loss": 0.6778,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9282495667244367,
+      "grad_norm": 6.2830988608758105,
+      "learning_rate": 1.357363495377545e-07,
+      "loss": 0.7655,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9285961871750433,
+      "grad_norm": 2.2504109996697124,
+      "learning_rate": 1.344401962951697e-07,
+      "loss": 0.7901,
+      "step": 2679
+    },
+    {
+      "epoch": 0.92894280762565,
+      "grad_norm": 3.6315611672500308,
+      "learning_rate": 1.3315017695827604e-07,
+      "loss": 0.7716,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9292894280762565,
+      "grad_norm": 3.845438006162765,
+      "learning_rate": 1.3186629315337207e-07,
+      "loss": 0.818,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9296360485268631,
+      "grad_norm": 2.9621981464517515,
+      "learning_rate": 1.3058854649902308e-07,
+      "loss": 0.7799,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9299826689774696,
+      "grad_norm": 3.4870502163049193,
+      "learning_rate": 1.2931693860605666e-07,
+      "loss": 0.6496,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9303292894280762,
+      "grad_norm": 7.720393223662143,
+      "learning_rate": 1.2805147107756045e-07,
+      "loss": 0.6402,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9306759098786829,
+      "grad_norm": 2.7234506571850656,
+      "learning_rate": 1.2679214550888219e-07,
+      "loss": 0.8364,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9310225303292894,
+      "grad_norm": 4.490117459089937,
+      "learning_rate": 1.255389634876264e-07,
+      "loss": 0.7385,
+      "step": 2686
+    },
+    {
+      "epoch": 0.931369150779896,
+      "grad_norm": 2.7587568010170145,
+      "learning_rate": 1.2429192659365208e-07,
+      "loss": 0.7666,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9317157712305026,
+      "grad_norm": 3.0259264429089687,
+      "learning_rate": 1.2305103639907333e-07,
+      "loss": 0.6932,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9320623916811092,
+      "grad_norm": 8.769684936750506,
+      "learning_rate": 1.2181629446825105e-07,
+      "loss": 0.8038,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9324090121317158,
+      "grad_norm": 4.276216296497035,
+      "learning_rate": 1.2058770235779892e-07,
+      "loss": 0.6601,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9327556325823223,
+      "grad_norm": 4.744677416444633,
+      "learning_rate": 1.193652616165758e-07,
+      "loss": 0.8628,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9331022530329289,
+      "grad_norm": 2.572254245102933,
+      "learning_rate": 1.1814897378568724e-07,
+      "loss": 0.6092,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9334488734835356,
+      "grad_norm": 2.81408825207413,
+      "learning_rate": 1.1693884039847946e-07,
+      "loss": 0.7472,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9337954939341421,
+      "grad_norm": 3.847835959493398,
+      "learning_rate": 1.157348629805427e-07,
+      "loss": 0.8522,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9341421143847487,
+      "grad_norm": 19.64572018737726,
+      "learning_rate": 1.1453704304970447e-07,
+      "loss": 0.898,
+      "step": 2695
+    },
+    {
+      "epoch": 0.9344887348353553,
+      "grad_norm": 4.8221484873460145,
+      "learning_rate": 1.1334538211603074e-07,
+      "loss": 0.6271,
+      "step": 2696
+    },
+    {
+      "epoch": 0.9348353552859618,
+      "grad_norm": 3.3389279329219743,
+      "learning_rate": 1.1215988168182201e-07,
+      "loss": 0.7217,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9351819757365685,
+      "grad_norm": 7.959962377601034,
+      "learning_rate": 1.1098054324161333e-07,
+      "loss": 0.7262,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9355285961871751,
+      "grad_norm": 5.3077338379456345,
+      "learning_rate": 1.0980736828217154e-07,
+      "loss": 0.7882,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9358752166377816,
+      "grad_norm": 4.881190279039482,
+      "learning_rate": 1.0864035828249187e-07,
+      "loss": 0.7296,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9362218370883882,
+      "grad_norm": 3.713719267192957,
+      "learning_rate": 1.0747951471379914e-07,
+      "loss": 0.837,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9365684575389948,
+      "grad_norm": 3.429790110294795,
+      "learning_rate": 1.0632483903954383e-07,
+      "loss": 0.7871,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9369150779896014,
+      "grad_norm": 5.756305530881598,
+      "learning_rate": 1.051763327154004e-07,
+      "loss": 0.8002,
+      "step": 2703
+    },
+    {
+      "epoch": 0.937261698440208,
+      "grad_norm": 4.49359177369374,
+      "learning_rate": 1.0403399718926565e-07,
+      "loss": 0.7166,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9376083188908145,
+      "grad_norm": 2.8209999311147076,
+      "learning_rate": 1.0289783390125707e-07,
+      "loss": 0.8,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9379549393414212,
+      "grad_norm": 4.1469683111244535,
+      "learning_rate": 1.0176784428371111e-07,
+      "loss": 0.7387,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9383015597920278,
+      "grad_norm": 5.770523521830978,
+      "learning_rate": 1.0064402976118103e-07,
+      "loss": 0.7634,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9386481802426343,
+      "grad_norm": 9.162497833862103,
+      "learning_rate": 9.952639175043576e-08,
+      "loss": 0.7747,
+      "step": 2708
+    },
+    {
+      "epoch": 0.9389948006932409,
+      "grad_norm": 4.093870587596725,
+      "learning_rate": 9.841493166045767e-08,
+      "loss": 0.715,
+      "step": 2709
+    },
+    {
+      "epoch": 0.9393414211438474,
+      "grad_norm": 3.5751396967863274,
+      "learning_rate": 9.730965089243982e-08,
+      "loss": 0.7621,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9396880415944541,
+      "grad_norm": 3.0461187380583445,
+      "learning_rate": 9.621055083978591e-08,
+      "loss": 0.7634,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9400346620450607,
+      "grad_norm": 3.0241072492261214,
+      "learning_rate": 9.511763288810705e-08,
+      "loss": 0.7557,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9403812824956672,
+      "grad_norm": 3.91719046218526,
+      "learning_rate": 9.403089841522217e-08,
+      "loss": 0.6793,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9407279029462738,
+      "grad_norm": 6.290736082564225,
+      "learning_rate": 9.295034879115261e-08,
+      "loss": 0.7487,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9410745233968805,
+      "grad_norm": 3.307883538963728,
+      "learning_rate": 9.18759853781248e-08,
+      "loss": 0.7778,
+      "step": 2715
+    },
+    {
+      "epoch": 0.941421143847487,
+      "grad_norm": 9.986976310841149,
+      "learning_rate": 9.080780953056479e-08,
+      "loss": 0.8025,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9417677642980936,
+      "grad_norm": 4.950545006948777,
+      "learning_rate": 8.974582259509923e-08,
+      "loss": 0.755,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9421143847487001,
+      "grad_norm": 5.092640184643661,
+      "learning_rate": 8.869002591055165e-08,
+      "loss": 0.7064,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9424610051993068,
+      "grad_norm": 3.2235610088591606,
+      "learning_rate": 8.764042080794233e-08,
+      "loss": 0.6262,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9428076256499134,
+      "grad_norm": 3.131617381585624,
+      "learning_rate": 8.659700861048503e-08,
+      "loss": 0.709,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9431542461005199,
+      "grad_norm": 2.9401147850557936,
+      "learning_rate": 8.555979063358755e-08,
+      "loss": 0.7695,
+      "step": 2721
+    },
+    {
+      "epoch": 0.9435008665511265,
+      "grad_norm": 3.6952789566765594,
+      "learning_rate": 8.452876818484778e-08,
+      "loss": 0.7578,
+      "step": 2722
+    },
+    {
+      "epoch": 0.9438474870017332,
+      "grad_norm": 10.910952243409755,
+      "learning_rate": 8.35039425640538e-08,
+      "loss": 0.7104,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9441941074523397,
+      "grad_norm": 7.260721453882653,
+      "learning_rate": 8.248531506318047e-08,
+      "loss": 0.8154,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9445407279029463,
+      "grad_norm": 5.189952654782747,
+      "learning_rate": 8.147288696638945e-08,
+      "loss": 0.693,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9448873483535528,
+      "grad_norm": 4.816622691657644,
+      "learning_rate": 8.046665955002697e-08,
+      "loss": 0.6363,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9452339688041594,
+      "grad_norm": 4.945985012395804,
+      "learning_rate": 7.946663408262167e-08,
+      "loss": 0.7117,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9455805892547661,
+      "grad_norm": 4.778443924645704,
+      "learning_rate": 7.847281182488508e-08,
+      "loss": 0.7768,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9459272097053726,
+      "grad_norm": 3.3710967345085976,
+      "learning_rate": 7.748519402970612e-08,
+      "loss": 0.693,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9462738301559792,
+      "grad_norm": 3.8675126512545477,
+      "learning_rate": 7.650378194215436e-08,
+      "loss": 0.7471,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9466204506065858,
+      "grad_norm": 2.912858670769692,
+      "learning_rate": 7.5528576799474e-08,
+      "loss": 0.6919,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9469670710571924,
+      "grad_norm": 4.273833540425412,
+      "learning_rate": 7.455957983108497e-08,
+      "loss": 0.5762,
+      "step": 2732
+    },
+    {
+      "epoch": 0.947313691507799,
+      "grad_norm": 4.5794686942290985,
+      "learning_rate": 7.359679225858062e-08,
+      "loss": 0.7656,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9476603119584055,
+      "grad_norm": 14.036191315245212,
+      "learning_rate": 7.264021529572618e-08,
+      "loss": 0.7632,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9480069324090121,
+      "grad_norm": 5.559774370758878,
+      "learning_rate": 7.168985014845864e-08,
+      "loss": 0.6856,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9483535528596188,
+      "grad_norm": 4.08411623320441,
+      "learning_rate": 7.074569801488129e-08,
+      "loss": 0.7512,
+      "step": 2736
+    },
+    {
+      "epoch": 0.9487001733102253,
+      "grad_norm": 3.661613373360454,
+      "learning_rate": 6.980776008526758e-08,
+      "loss": 0.804,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9490467937608319,
+      "grad_norm": 5.461841936849577,
+      "learning_rate": 6.887603754205441e-08,
+      "loss": 0.7073,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9493934142114385,
+      "grad_norm": 6.661683730643877,
+      "learning_rate": 6.795053155984443e-08,
+      "loss": 0.6767,
+      "step": 2739
+    },
+    {
+      "epoch": 0.949740034662045,
+      "grad_norm": 12.49693704385828,
+      "learning_rate": 6.703124330540323e-08,
+      "loss": 0.7192,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9500866551126517,
+      "grad_norm": 4.425621088224397,
+      "learning_rate": 6.611817393765763e-08,
+      "loss": 0.7952,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9504332755632582,
+      "grad_norm": 3.9120777007075858,
+      "learning_rate": 6.521132460769408e-08,
+      "loss": 0.8515,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9507798960138648,
+      "grad_norm": 3.8595093529384004,
+      "learning_rate": 6.43106964587581e-08,
+      "loss": 0.6934,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9511265164644714,
+      "grad_norm": 13.307543329943123,
+      "learning_rate": 6.341629062625144e-08,
+      "loss": 0.752,
+      "step": 2744
+    },
+    {
+      "epoch": 0.951473136915078,
+      "grad_norm": 2.952392732560199,
+      "learning_rate": 6.252810823773381e-08,
+      "loss": 0.7856,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9518197573656846,
+      "grad_norm": 3.3863170206348703,
+      "learning_rate": 6.164615041291511e-08,
+      "loss": 0.7433,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9521663778162912,
+      "grad_norm": 4.884573239737837,
+      "learning_rate": 6.077041826366148e-08,
+      "loss": 0.7858,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9525129982668977,
+      "grad_norm": 2.892780445657699,
+      "learning_rate": 5.990091289399036e-08,
+      "loss": 0.7754,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9528596187175044,
+      "grad_norm": 3.3904761883113204,
+      "learning_rate": 5.903763540006713e-08,
+      "loss": 0.78,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9532062391681109,
+      "grad_norm": 2.7325150746011984,
+      "learning_rate": 5.818058687020789e-08,
+      "loss": 0.743,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9535528596187175,
+      "grad_norm": 2.937033686869785,
+      "learning_rate": 5.7329768384874477e-08,
+      "loss": 0.6572,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9538994800693241,
+      "grad_norm": 8.083988319051185,
+      "learning_rate": 5.6485181016674996e-08,
+      "loss": 0.7612,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9542461005199306,
+      "grad_norm": 2.0564890317708544,
+      "learning_rate": 5.5646825830363295e-08,
+      "loss": 0.7494,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9545927209705373,
+      "grad_norm": 2.9468879739069314,
+      "learning_rate": 5.481470388283505e-08,
+      "loss": 0.6631,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9549393414211439,
+      "grad_norm": 4.4032493830087756,
+      "learning_rate": 5.3988816223128346e-08,
+      "loss": 0.8524,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9552859618717504,
+      "grad_norm": 9.323019984713818,
+      "learning_rate": 5.3169163892421995e-08,
+      "loss": 0.7236,
+      "step": 2756
+    },
+    {
+      "epoch": 0.955632582322357,
+      "grad_norm": 2.917176710644127,
+      "learning_rate": 5.2355747924033865e-08,
+      "loss": 0.6779,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9559792027729636,
+      "grad_norm": 3.4761256233411846,
+      "learning_rate": 5.1548569343419786e-08,
+      "loss": 0.6363,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9563258232235702,
+      "grad_norm": 2.810916877612425,
+      "learning_rate": 5.074762916817244e-08,
+      "loss": 0.7116,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9566724436741768,
+      "grad_norm": 4.115504513051146,
+      "learning_rate": 4.995292840801913e-08,
+      "loss": 0.719,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9570190641247833,
+      "grad_norm": 3.195950519988086,
+      "learning_rate": 4.916446806482289e-08,
+      "loss": 0.765,
+      "step": 2761
+    },
+    {
+      "epoch": 0.95736568457539,
+      "grad_norm": 3.439990884714393,
+      "learning_rate": 4.838224913257805e-08,
+      "loss": 0.7604,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9577123050259966,
+      "grad_norm": 3.2521251694602347,
+      "learning_rate": 4.76062725974108e-08,
+      "loss": 0.7388,
+      "step": 2763
+    },
+    {
+      "epoch": 0.9580589254766031,
+      "grad_norm": 7.590543653800222,
+      "learning_rate": 4.683653943757804e-08,
+      "loss": 0.6734,
+      "step": 2764
+    },
+    {
+      "epoch": 0.9584055459272097,
+      "grad_norm": 6.018568772826001,
+      "learning_rate": 4.607305062346579e-08,
+      "loss": 0.7266,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9587521663778162,
+      "grad_norm": 6.049202779300583,
+      "learning_rate": 4.531580711758799e-08,
+      "loss": 0.6959,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9590987868284229,
+      "grad_norm": 2.4700807008278773,
+      "learning_rate": 4.456480987458489e-08,
+      "loss": 0.6658,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9594454072790295,
+      "grad_norm": 8.249536538470066,
+      "learning_rate": 4.382005984122195e-08,
+      "loss": 0.675,
+      "step": 2768
+    },
+    {
+      "epoch": 0.959792027729636,
+      "grad_norm": 3.2439226744087177,
+      "learning_rate": 4.308155795639035e-08,
+      "loss": 0.781,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9601386481802426,
+      "grad_norm": 18.683191358219094,
+      "learning_rate": 4.234930515110314e-08,
+      "loss": 0.7434,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9604852686308493,
+      "grad_norm": 3.8979983814939816,
+      "learning_rate": 4.162330234849521e-08,
+      "loss": 0.7108,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9608318890814558,
+      "grad_norm": 4.2746153728594365,
+      "learning_rate": 4.090355046382277e-08,
+      "loss": 0.7857,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9611785095320624,
+      "grad_norm": 15.823365932269292,
+      "learning_rate": 4.0190050404460556e-08,
+      "loss": 0.8553,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9615251299826689,
+      "grad_norm": 5.4182827306117884,
+      "learning_rate": 3.948280306990404e-08,
+      "loss": 0.6739,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9618717504332756,
+      "grad_norm": 6.609119620013426,
+      "learning_rate": 3.8781809351763345e-08,
+      "loss": 0.6788,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9622183708838822,
+      "grad_norm": 3.4688051035189833,
+      "learning_rate": 3.808707013376767e-08,
+      "loss": 0.7109,
+      "step": 2776
+    },
+    {
+      "epoch": 0.9625649913344887,
+      "grad_norm": 3.4149479634080473,
+      "learning_rate": 3.7398586291757524e-08,
+      "loss": 0.7534,
+      "step": 2777
+    },
+    {
+      "epoch": 0.9629116117850953,
+      "grad_norm": 3.4182202061322964,
+      "learning_rate": 3.671635869369139e-08,
+      "loss": 0.72,
+      "step": 2778
+    },
+    {
+      "epoch": 0.963258232235702,
+      "grad_norm": 3.263683274374266,
+      "learning_rate": 3.60403881996374e-08,
+      "loss": 0.7805,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9636048526863085,
+      "grad_norm": 7.364548822148344,
+      "learning_rate": 3.537067566177832e-08,
+      "loss": 0.8549,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9639514731369151,
+      "grad_norm": 3.6782519657335024,
+      "learning_rate": 3.4707221924404896e-08,
+      "loss": 0.6865,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9642980935875216,
+      "grad_norm": 3.5111256276841485,
+      "learning_rate": 3.405002782391975e-08,
+      "loss": 0.7295,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9646447140381282,
+      "grad_norm": 4.217240831555684,
+      "learning_rate": 3.339909418883236e-08,
+      "loss": 0.6625,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9649913344887349,
+      "grad_norm": 2.1481656910273292,
+      "learning_rate": 3.27544218397613e-08,
+      "loss": 0.7987,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9653379549393414,
+      "grad_norm": 3.541882335138784,
+      "learning_rate": 3.211601158943145e-08,
+      "loss": 0.7269,
+      "step": 2785
+    },
+    {
+      "epoch": 0.965684575389948,
+      "grad_norm": 3.219000870649726,
+      "learning_rate": 3.1483864242671805e-08,
+      "loss": 0.8039,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9660311958405546,
+      "grad_norm": 4.49541257387211,
+      "learning_rate": 3.085798059641709e-08,
+      "loss": 0.6373,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9663778162911612,
+      "grad_norm": 7.1713376616111955,
+      "learning_rate": 3.02383614397056e-08,
+      "loss": 0.694,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9667244367417678,
+      "grad_norm": 3.5408010622314614,
+      "learning_rate": 2.9625007553677477e-08,
+      "loss": 0.7358,
+      "step": 2789
+    },
+    {
+      "epoch": 0.9670710571923743,
+      "grad_norm": 3.8680216219095693,
+      "learning_rate": 2.9017919711574194e-08,
+      "loss": 0.7062,
+      "step": 2790
+    },
+    {
+      "epoch": 0.9674176776429809,
+      "grad_norm": 3.6457147202615667,
+      "learning_rate": 2.841709867873965e-08,
+      "loss": 0.7625,
+      "step": 2791
+    },
+    {
+      "epoch": 0.9677642980935875,
+      "grad_norm": 4.865869510997766,
+      "learning_rate": 2.7822545212614072e-08,
+      "loss": 0.8215,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9681109185441941,
+      "grad_norm": 4.449412056241128,
+      "learning_rate": 2.723426006273955e-08,
+      "loss": 0.7594,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9684575389948007,
+      "grad_norm": 6.339542002046698,
+      "learning_rate": 2.6652243970753945e-08,
+      "loss": 0.7146,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9688041594454073,
+      "grad_norm": 3.6586813932564053,
+      "learning_rate": 2.6076497670391443e-08,
+      "loss": 0.6498,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9691507798960138,
+      "grad_norm": 3.6852050972867434,
+      "learning_rate": 2.5507021887484774e-08,
+      "loss": 0.8046,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9694974003466205,
+      "grad_norm": 5.663938537645889,
+      "learning_rate": 2.4943817339957986e-08,
+      "loss": 0.6904,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9698440207972271,
+      "grad_norm": 6.0364331564419125,
+      "learning_rate": 2.4386884737831463e-08,
+      "loss": 0.7856,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9701906412478336,
+      "grad_norm": 2.951581664134406,
+      "learning_rate": 2.3836224783218565e-08,
+      "loss": 0.685,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9705372616984402,
+      "grad_norm": 3.324317688114565,
+      "learning_rate": 2.3291838170323432e-08,
+      "loss": 0.7142,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9708838821490468,
+      "grad_norm": 4.112577320308709,
+      "learning_rate": 2.2753725585443198e-08,
+      "loss": 0.7206,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9712305025996534,
+      "grad_norm": 5.589455722088366,
+      "learning_rate": 2.2221887706964097e-08,
+      "loss": 0.6986,
+      "step": 2802
+    },
+    {
+      "epoch": 0.97157712305026,
+      "grad_norm": 4.400845904741504,
+      "learning_rate": 2.1696325205363134e-08,
+      "loss": 0.6392,
+      "step": 2803
+    },
+    {
+      "epoch": 0.9719237435008665,
+      "grad_norm": 3.6652946611615227,
+      "learning_rate": 2.1177038743205313e-08,
+      "loss": 0.7557,
+      "step": 2804
+    },
+    {
+      "epoch": 0.9722703639514731,
+      "grad_norm": 2.590458961620275,
+      "learning_rate": 2.0664028975143634e-08,
+      "loss": 0.6812,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9726169844020798,
+      "grad_norm": 3.0015632310665152,
+      "learning_rate": 2.0157296547918536e-08,
+      "loss": 0.6304,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9729636048526863,
+      "grad_norm": 3.8670860150120476,
+      "learning_rate": 1.965684210035734e-08,
+      "loss": 0.7197,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9733102253032929,
+      "grad_norm": 4.300541793859677,
+      "learning_rate": 1.9162666263370932e-08,
+      "loss": 0.7468,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9736568457538994,
+      "grad_norm": 4.540443597023693,
+      "learning_rate": 1.8674769659956515e-08,
+      "loss": 0.7657,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9740034662045061,
+      "grad_norm": 5.250387597679544,
+      "learning_rate": 1.8193152905194857e-08,
+      "loss": 0.7679,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9743500866551127,
+      "grad_norm": 6.514013324309402,
+      "learning_rate": 1.7717816606249715e-08,
+      "loss": 0.7113,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9746967071057192,
+      "grad_norm": 3.2868875267608533,
+      "learning_rate": 1.7248761362367307e-08,
+      "loss": 0.8259,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9750433275563258,
+      "grad_norm": 3.9451892351323345,
+      "learning_rate": 1.6785987764875167e-08,
+      "loss": 0.745,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9753899480069325,
+      "grad_norm": 2.66670423729789,
+      "learning_rate": 1.6329496397182176e-08,
+      "loss": 0.6948,
+      "step": 2814
+    },
+    {
+      "epoch": 0.975736568457539,
+      "grad_norm": 3.9553114488325103,
+      "learning_rate": 1.587928783477688e-08,
+      "loss": 0.7657,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9760831889081456,
+      "grad_norm": 21.655557344052497,
+      "learning_rate": 1.5435362645227492e-08,
+      "loss": 0.758,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9764298093587521,
+      "grad_norm": 4.409143152365286,
+      "learning_rate": 1.499772138818134e-08,
+      "loss": 0.7154,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9767764298093587,
+      "grad_norm": 3.0131704648242774,
+      "learning_rate": 1.456636461536265e-08,
+      "loss": 0.7934,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9771230502599654,
+      "grad_norm": 5.295097022272975,
+      "learning_rate": 1.4141292870574198e-08,
+      "loss": 0.7615,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9774696707105719,
+      "grad_norm": 4.034401244303767,
+      "learning_rate": 1.3722506689693439e-08,
+      "loss": 0.7356,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9778162911611785,
+      "grad_norm": 4.703538744978473,
+      "learning_rate": 1.3310006600676384e-08,
+      "loss": 0.7868,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9781629116117851,
+      "grad_norm": 6.253232825106559,
+      "learning_rate": 1.2903793123552611e-08,
+      "loss": 0.771,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9785095320623917,
+      "grad_norm": 3.7124774471044737,
+      "learning_rate": 1.2503866770426365e-08,
+      "loss": 0.7683,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9788561525129983,
+      "grad_norm": 4.308186506997935,
+      "learning_rate": 1.2110228045476013e-08,
+      "loss": 0.8686,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9792027729636048,
+      "grad_norm": 3.9439080594716307,
+      "learning_rate": 1.1722877444953484e-08,
+      "loss": 0.7422,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9795493934142114,
+      "grad_norm": 10.227268374176901,
+      "learning_rate": 1.1341815457183714e-08,
+      "loss": 0.653,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9798960138648181,
+      "grad_norm": 4.654200937950107,
+      "learning_rate": 1.0967042562562424e-08,
+      "loss": 0.7139,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9802426343154246,
+      "grad_norm": 2.850493691741343,
+      "learning_rate": 1.0598559233558348e-08,
+      "loss": 0.819,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9805892547660312,
+      "grad_norm": 3.524345293429381,
+      "learning_rate": 1.0236365934710446e-08,
+      "loss": 0.6272,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9809358752166378,
+      "grad_norm": 3.764980287834406,
+      "learning_rate": 9.880463122627915e-09,
+      "loss": 0.7368,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9812824956672443,
+      "grad_norm": 3.9064091848753195,
+      "learning_rate": 9.53085124598907e-09,
+      "loss": 0.6241,
+      "step": 2831
+    },
+    {
+      "epoch": 0.981629116117851,
+      "grad_norm": 5.085060916446495,
+      "learning_rate": 9.187530745543016e-09,
+      "loss": 0.7892,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9819757365684575,
+      "grad_norm": 10.743715508589514,
+      "learning_rate": 8.85050205410576e-09,
+      "loss": 0.6699,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9823223570190641,
+      "grad_norm": 3.75953133683317,
+      "learning_rate": 8.51976559656187e-09,
+      "loss": 0.7387,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9826689774696707,
+      "grad_norm": 6.4835995171211165,
+      "learning_rate": 8.195321789864485e-09,
+      "loss": 0.7848,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9830155979202773,
+      "grad_norm": 5.406734268807864,
+      "learning_rate": 7.877171043031983e-09,
+      "loss": 0.6881,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9833622183708839,
+      "grad_norm": 12.992418274716927,
+      "learning_rate": 7.56531375715075e-09,
+      "loss": 0.8397,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9837088388214905,
+      "grad_norm": 3.0148747055767138,
+      "learning_rate": 7.259750325372405e-09,
+      "loss": 0.7101,
+      "step": 2838
+    },
+    {
+      "epoch": 0.984055459272097,
+      "grad_norm": 3.8995889088006144,
+      "learning_rate": 6.96048113291381e-09,
+      "loss": 0.8888,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9844020797227037,
+      "grad_norm": 12.587987524729217,
+      "learning_rate": 6.667506557057057e-09,
+      "loss": 0.7081,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9847487001733102,
+      "grad_norm": 2.3632150761862016,
+      "learning_rate": 6.380826967149478e-09,
+      "loss": 0.7321,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9850953206239168,
+      "grad_norm": 4.183776766809023,
+      "learning_rate": 6.100442724601419e-09,
+      "loss": 0.7501,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9854419410745234,
+      "grad_norm": 3.1614852575626635,
+      "learning_rate": 5.82635418288735e-09,
+      "loss": 0.7085,
+      "step": 2843
+    },
+    {
+      "epoch": 0.98578856152513,
+      "grad_norm": 2.643627566245314,
+      "learning_rate": 5.558561687544206e-09,
+      "loss": 0.737,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9861351819757366,
+      "grad_norm": 3.6016911881139517,
+      "learning_rate": 5.297065576173044e-09,
+      "loss": 0.839,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9864818024263432,
+      "grad_norm": 4.352111665969631,
+      "learning_rate": 5.041866178435717e-09,
+      "loss": 0.673,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9868284228769497,
+      "grad_norm": 4.720740357538921,
+      "learning_rate": 4.792963816057095e-09,
+      "loss": 0.6863,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9871750433275563,
+      "grad_norm": 9.304192834361986,
+      "learning_rate": 4.550358802822285e-09,
+      "loss": 0.7824,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9875216637781629,
+      "grad_norm": 9.55268084374393,
+      "learning_rate": 4.314051444578305e-09,
+      "loss": 0.6506,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9878682842287695,
+      "grad_norm": 16.112224014792904,
+      "learning_rate": 4.084042039233515e-09,
+      "loss": 0.8012,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9882149046793761,
+      "grad_norm": 3.782778749921925,
+      "learning_rate": 3.860330876754859e-09,
+      "loss": 0.7981,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9885615251299826,
+      "grad_norm": 4.367938102916673,
+      "learning_rate": 3.642918239170623e-09,
+      "loss": 0.6776,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9889081455805893,
+      "grad_norm": 6.3690345492236515,
+      "learning_rate": 3.431804400567673e-09,
+      "loss": 0.7963,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9892547660311959,
+      "grad_norm": 2.4939315549717387,
+      "learning_rate": 3.2269896270936683e-09,
+      "loss": 0.8197,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9896013864818024,
+      "grad_norm": 4.187073702082158,
+      "learning_rate": 3.028474176953178e-09,
+      "loss": 0.8288,
+      "step": 2855
+    },
+    {
+      "epoch": 0.989948006932409,
+      "grad_norm": 5.725630667522824,
+      "learning_rate": 2.836258300410455e-09,
+      "loss": 0.7638,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9902946273830155,
+      "grad_norm": 6.073795581421976,
+      "learning_rate": 2.6503422397883284e-09,
+      "loss": 0.6733,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9906412478336222,
+      "grad_norm": 3.576733463903546,
+      "learning_rate": 2.4707262294670907e-09,
+      "loss": 0.6944,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9909878682842288,
+      "grad_norm": 3.0479661402908964,
+      "learning_rate": 2.2974104958845e-09,
+      "loss": 0.8476,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9913344887348353,
+      "grad_norm": 14.76509057896928,
+      "learning_rate": 2.1303952575357777e-09,
+      "loss": 0.7059,
+      "step": 2860
+    },
+    {
+      "epoch": 0.991681109185442,
+      "grad_norm": 10.945064196982477,
+      "learning_rate": 1.9696807249736104e-09,
+      "loss": 0.7638,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9920277296360486,
+      "grad_norm": 2.570097771691247,
+      "learning_rate": 1.815267100807594e-09,
+      "loss": 0.7838,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9923743500866551,
+      "grad_norm": 5.350094368941037,
+      "learning_rate": 1.6671545797036782e-09,
+      "loss": 0.7728,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9927209705372617,
+      "grad_norm": 3.167222914880599,
+      "learning_rate": 1.5253433483836122e-09,
+      "loss": 0.8075,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9930675909878682,
+      "grad_norm": 3.846089683897135,
+      "learning_rate": 1.3898335856254996e-09,
+      "loss": 0.6442,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9934142114384749,
+      "grad_norm": 20.965421089865544,
+      "learning_rate": 1.2606254622643532e-09,
+      "loss": 0.6278,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9937608318890815,
+      "grad_norm": 4.080466495189374,
+      "learning_rate": 1.1377191411898747e-09,
+      "loss": 0.7195,
+      "step": 2867
+    },
+    {
+      "epoch": 0.994107452339688,
+      "grad_norm": 3.995822926437377,
+      "learning_rate": 1.0211147773470099e-09,
+      "loss": 0.7216,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9944540727902946,
+      "grad_norm": 3.6483472622592656,
+      "learning_rate": 9.108125177370586e-10,
+      "loss": 0.6866,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9948006932409013,
+      "grad_norm": 5.581912963930684,
+      "learning_rate": 8.068125014143446e-10,
+      "loss": 0.6288,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9951473136915078,
+      "grad_norm": 4.287701969201671,
+      "learning_rate": 7.091148594912112e-10,
+      "loss": 0.6381,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9954939341421144,
+      "grad_norm": 2.898294786094583,
+      "learning_rate": 6.177197151313597e-10,
+      "loss": 0.8538,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9958405545927209,
+      "grad_norm": 6.0028670659651775,
+      "learning_rate": 5.326271835554008e-10,
+      "loss": 0.7723,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9961871750433275,
+      "grad_norm": 3.1967269199106076,
+      "learning_rate": 4.5383737203807953e-10,
+      "loss": 0.7604,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9965337954939342,
+      "grad_norm": 3.5421310056991273,
+      "learning_rate": 3.8135037990716383e-10,
+      "loss": 0.787,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9968804159445407,
+      "grad_norm": 4.331789783522943,
+      "learning_rate": 3.1516629854566606e-10,
+      "loss": 0.7578,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9972270363951473,
+      "grad_norm": 3.748289209541486,
+      "learning_rate": 2.5528521139073227e-10,
+      "loss": 0.8531,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9975736568457539,
+      "grad_norm": 4.7286920494852485,
+      "learning_rate": 2.0170719393308723e-10,
+      "loss": 0.8677,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9979202772963605,
+      "grad_norm": 6.698072192840444,
+      "learning_rate": 1.544323137170345e-10,
+      "loss": 0.7212,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9982668977469671,
+      "grad_norm": 2.2611982348398687,
+      "learning_rate": 1.1346063034101129e-10,
+      "loss": 0.7399,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9986135181975736,
+      "grad_norm": 6.899324317395354,
+      "learning_rate": 7.879219545758875e-11,
+      "loss": 0.6701,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9989601386481802,
+      "grad_norm": 3.0423660504094125,
+      "learning_rate": 5.0427052771806485e-11,
+      "loss": 0.7689,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9993067590987869,
+      "grad_norm": 6.738591658066531,
+      "learning_rate": 2.8365238043948085e-11,
+      "loss": 0.7375,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9996533795493934,
+      "grad_norm": 13.05829471833478,
+      "learning_rate": 1.2606779086210553e-11,
+      "loss": 0.7958,
+      "step": 2884
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.405770989929639,
+      "learning_rate": 3.1516957649246893e-12,
+      "loss": 0.7929,
+      "step": 2885
+    },
+    {
+      "epoch": 1.0,
+      "step": 2885,
+      "total_flos": 2811783190478848.0,
+      "train_loss": 0.8937024936932212,
+      "train_runtime": 265145.644,
+      "train_samples_per_second": 2.786,
+      "train_steps_per_second": 0.011
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2885,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 800,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2811783190478848.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}