diff --git "a/checkpoint-2041/trainer_state.json" "b/checkpoint-2041/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2041/trainer_state.json"
@@ -0,0 +1,14320 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998775260257196,
+  "eval_steps": 500,
+  "global_step": 2041,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004898958971218616,
+      "grad_norm": 1.668303370475769,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.9871,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009797917942437231,
+      "grad_norm": 1.5414912700653076,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.9553,
+      "step": 2
+    },
+    {
+      "epoch": 0.001469687691365585,
+      "grad_norm": 1.5704071521759033,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.9932,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019595835884874463,
+      "grad_norm": 1.5206725597381592,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.955,
+      "step": 4
+    },
+    {
+      "epoch": 0.002449479485609308,
+      "grad_norm": 1.5219823122024536,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.9255,
+      "step": 5
+    },
+    {
+      "epoch": 0.00293937538273117,
+      "grad_norm": 1.458319902420044,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.918,
+      "step": 6
+    },
+    {
+      "epoch": 0.003429271279853031,
+      "grad_norm": 1.6977298259735107,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 1.0083,
+      "step": 7
+    },
+    {
+      "epoch": 0.0039191671769748925,
+      "grad_norm": 1.5800329446792603,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.9472,
+      "step": 8
+    },
+    {
+      "epoch": 0.004409063074096754,
+      "grad_norm": 1.5464119911193848,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 0.9621,
+      "step": 9
+    },
+    {
+      "epoch": 0.004898958971218616,
+      "grad_norm": 1.5234100818634033,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.9296,
+      "step": 10
+    },
+    {
+      "epoch": 0.005388854868340478,
+      "grad_norm": 1.481718897819519,
+      "learning_rate": 5.5e-07,
+      "loss": 0.9571,
+      "step": 11
+    },
+    {
+      "epoch": 0.00587875076546234,
+      "grad_norm": 1.4497618675231934,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.9538,
+      "step": 12
+    },
+    {
+      "epoch": 0.0063686466625842006,
+      "grad_norm": 1.5114537477493286,
+      "learning_rate": 6.5e-07,
+      "loss": 0.9591,
+      "step": 13
+    },
+    {
+      "epoch": 0.006858542559706062,
+      "grad_norm": 1.4514930248260498,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.917,
+      "step": 14
+    },
+    {
+      "epoch": 0.007348438456827924,
+      "grad_norm": 1.3999227285385132,
+      "learning_rate": 7.5e-07,
+      "loss": 0.9543,
+      "step": 15
+    },
+    {
+      "epoch": 0.007838334353949785,
+      "grad_norm": 1.3517608642578125,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.9439,
+      "step": 16
+    },
+    {
+      "epoch": 0.008328230251071647,
+      "grad_norm": 1.2273361682891846,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.961,
+      "step": 17
+    },
+    {
+      "epoch": 0.008818126148193509,
+      "grad_norm": 1.157639980316162,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.9285,
+      "step": 18
+    },
+    {
+      "epoch": 0.00930802204531537,
+      "grad_norm": 1.0808814764022827,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.91,
+      "step": 19
+    },
+    {
+      "epoch": 0.009797917942437232,
+      "grad_norm": 1.0530990362167358,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.899,
+      "step": 20
+    },
+    {
+      "epoch": 0.010287813839559094,
+      "grad_norm": 1.0722999572753906,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.9384,
+      "step": 21
+    },
+    {
+      "epoch": 0.010777709736680956,
+      "grad_norm": 0.9200907945632935,
+      "learning_rate": 1.1e-06,
+      "loss": 0.9012,
+      "step": 22
+    },
+    {
+      "epoch": 0.011267605633802818,
+      "grad_norm": 0.8614085912704468,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.9176,
+      "step": 23
+    },
+    {
+      "epoch": 0.01175750153092468,
+      "grad_norm": 0.8791833519935608,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.9024,
+      "step": 24
+    },
+    {
+      "epoch": 0.01224739742804654,
+      "grad_norm": 0.7829999923706055,
+      "learning_rate": 1.25e-06,
+      "loss": 0.8956,
+      "step": 25
+    },
+    {
+      "epoch": 0.012737293325168401,
+      "grad_norm": 0.7279092073440552,
+      "learning_rate": 1.3e-06,
+      "loss": 0.8837,
+      "step": 26
+    },
+    {
+      "epoch": 0.013227189222290263,
+      "grad_norm": 0.7260591983795166,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.9255,
+      "step": 27
+    },
+    {
+      "epoch": 0.013717085119412125,
+      "grad_norm": 0.6384291648864746,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8652,
+      "step": 28
+    },
+    {
+      "epoch": 0.014206981016533986,
+      "grad_norm": 0.6459453701972961,
+      "learning_rate": 1.45e-06,
+      "loss": 0.8353,
+      "step": 29
+    },
+    {
+      "epoch": 0.014696876913655848,
+      "grad_norm": 0.6280319094657898,
+      "learning_rate": 1.5e-06,
+      "loss": 0.8961,
+      "step": 30
+    },
+    {
+      "epoch": 0.01518677281077771,
+      "grad_norm": 0.6156179308891296,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.8929,
+      "step": 31
+    },
+    {
+      "epoch": 0.01567666870789957,
+      "grad_norm": 0.5640994310379028,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.8689,
+      "step": 32
+    },
+    {
+      "epoch": 0.016166564605021434,
+      "grad_norm": 0.5428940057754517,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.8843,
+      "step": 33
+    },
+    {
+      "epoch": 0.016656460502143294,
+      "grad_norm": 0.5078619122505188,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.879,
+      "step": 34
+    },
+    {
+      "epoch": 0.017146356399265157,
+      "grad_norm": 0.4971136748790741,
+      "learning_rate": 1.75e-06,
+      "loss": 0.8815,
+      "step": 35
+    },
+    {
+      "epoch": 0.017636252296387017,
+      "grad_norm": 0.4693506062030792,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.8494,
+      "step": 36
+    },
+    {
+      "epoch": 0.01812614819350888,
+      "grad_norm": 0.4563111960887909,
+      "learning_rate": 1.85e-06,
+      "loss": 0.8379,
+      "step": 37
+    },
+    {
+      "epoch": 0.01861604409063074,
+      "grad_norm": 0.4457121193408966,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.8923,
+      "step": 38
+    },
+    {
+      "epoch": 0.019105939987752604,
+      "grad_norm": 0.4194672107696533,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.8229,
+      "step": 39
+    },
+    {
+      "epoch": 0.019595835884874464,
+      "grad_norm": 0.40843793749809265,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8483,
+      "step": 40
+    },
+    {
+      "epoch": 0.020085731781996324,
+      "grad_norm": 0.39125168323516846,
+      "learning_rate": 2.05e-06,
+      "loss": 0.8487,
+      "step": 41
+    },
+    {
+      "epoch": 0.020575627679118188,
+      "grad_norm": 0.39917248487472534,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.8859,
+      "step": 42
+    },
+    {
+      "epoch": 0.021065523576240048,
+      "grad_norm": 0.36128363013267517,
+      "learning_rate": 2.15e-06,
+      "loss": 0.7875,
+      "step": 43
+    },
+    {
+      "epoch": 0.02155541947336191,
+      "grad_norm": 0.37637510895729065,
+      "learning_rate": 2.2e-06,
+      "loss": 0.8377,
+      "step": 44
+    },
+    {
+      "epoch": 0.02204531537048377,
+      "grad_norm": 0.3807457387447357,
+      "learning_rate": 2.25e-06,
+      "loss": 0.848,
+      "step": 45
+    },
+    {
+      "epoch": 0.022535211267605635,
+      "grad_norm": 0.36621737480163574,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.8301,
+      "step": 46
+    },
+    {
+      "epoch": 0.023025107164727495,
+      "grad_norm": 0.35659459233283997,
+      "learning_rate": 2.35e-06,
+      "loss": 0.8429,
+      "step": 47
+    },
+    {
+      "epoch": 0.02351500306184936,
+      "grad_norm": 0.3621164858341217,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.8035,
+      "step": 48
+    },
+    {
+      "epoch": 0.02400489895897122,
+      "grad_norm": 0.3681345283985138,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.7814,
+      "step": 49
+    },
+    {
+      "epoch": 0.02449479485609308,
+      "grad_norm": 0.3703807294368744,
+      "learning_rate": 2.5e-06,
+      "loss": 0.8359,
+      "step": 50
+    },
+    {
+      "epoch": 0.024984690753214942,
+      "grad_norm": 0.3428044319152832,
+      "learning_rate": 2.55e-06,
+      "loss": 0.7875,
+      "step": 51
+    },
+    {
+      "epoch": 0.025474586650336802,
+      "grad_norm": 0.3258868455886841,
+      "learning_rate": 2.6e-06,
+      "loss": 0.7972,
+      "step": 52
+    },
+    {
+      "epoch": 0.025964482547458666,
+      "grad_norm": 0.30134114623069763,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.8126,
+      "step": 53
+    },
+    {
+      "epoch": 0.026454378444580526,
+      "grad_norm": 0.3181776702404022,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.8232,
+      "step": 54
+    },
+    {
+      "epoch": 0.02694427434170239,
+      "grad_norm": 0.3208143711090088,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.8325,
+      "step": 55
+    },
+    {
+      "epoch": 0.02743417023882425,
+      "grad_norm": 0.2844039499759674,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7862,
+      "step": 56
+    },
+    {
+      "epoch": 0.027924066135946113,
+      "grad_norm": 0.28659218549728394,
+      "learning_rate": 2.85e-06,
+      "loss": 0.8426,
+      "step": 57
+    },
+    {
+      "epoch": 0.028413962033067973,
+      "grad_norm": 0.2850748300552368,
+      "learning_rate": 2.9e-06,
+      "loss": 0.7922,
+      "step": 58
+    },
+    {
+      "epoch": 0.028903857930189833,
+      "grad_norm": 0.27821680903434753,
+      "learning_rate": 2.95e-06,
+      "loss": 0.8085,
+      "step": 59
+    },
+    {
+      "epoch": 0.029393753827311696,
+      "grad_norm": 0.2871108055114746,
+      "learning_rate": 3e-06,
+      "loss": 0.8015,
+      "step": 60
+    },
+    {
+      "epoch": 0.029883649724433557,
+      "grad_norm": 0.2768966853618622,
+      "learning_rate": 3.05e-06,
+      "loss": 0.8081,
+      "step": 61
+    },
+    {
+      "epoch": 0.03037354562155542,
+      "grad_norm": 0.27445188164711,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.7832,
+      "step": 62
+    },
+    {
+      "epoch": 0.03086344151867728,
+      "grad_norm": 0.2732076048851013,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.8103,
+      "step": 63
+    },
+    {
+      "epoch": 0.03135333741579914,
+      "grad_norm": 0.2777082026004791,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.838,
+      "step": 64
+    },
+    {
+      "epoch": 0.03184323331292101,
+      "grad_norm": 0.2745436728000641,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.792,
+      "step": 65
+    },
+    {
+      "epoch": 0.03233312921004287,
+      "grad_norm": 0.2597768306732178,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.8076,
+      "step": 66
+    },
+    {
+      "epoch": 0.03282302510716473,
+      "grad_norm": 0.25194868445396423,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.7815,
+      "step": 67
+    },
+    {
+      "epoch": 0.03331292100428659,
+      "grad_norm": 0.24808664619922638,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.7815,
+      "step": 68
+    },
+    {
+      "epoch": 0.03380281690140845,
+      "grad_norm": 0.25291213393211365,
+      "learning_rate": 3.45e-06,
+      "loss": 0.7831,
+      "step": 69
+    },
+    {
+      "epoch": 0.034292712798530314,
+      "grad_norm": 0.2491445690393448,
+      "learning_rate": 3.5e-06,
+      "loss": 0.7918,
+      "step": 70
+    },
+    {
+      "epoch": 0.034782608695652174,
+      "grad_norm": 0.24815823137760162,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.7925,
+      "step": 71
+    },
+    {
+      "epoch": 0.035272504592774034,
+      "grad_norm": 0.23526257276535034,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7971,
+      "step": 72
+    },
+    {
+      "epoch": 0.035762400489895894,
+      "grad_norm": 0.23858553171157837,
+      "learning_rate": 3.65e-06,
+      "loss": 0.8177,
+      "step": 73
+    },
+    {
+      "epoch": 0.03625229638701776,
+      "grad_norm": 0.23471508920192719,
+      "learning_rate": 3.7e-06,
+      "loss": 0.7927,
+      "step": 74
+    },
+    {
+      "epoch": 0.03674219228413962,
+      "grad_norm": 0.23156484961509705,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7824,
+      "step": 75
+    },
+    {
+      "epoch": 0.03723208818126148,
+      "grad_norm": 0.23146644234657288,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.7813,
+      "step": 76
+    },
+    {
+      "epoch": 0.03772198407838334,
+      "grad_norm": 0.23870326578617096,
+      "learning_rate": 3.85e-06,
+      "loss": 0.7871,
+      "step": 77
+    },
+    {
+      "epoch": 0.03821187997550521,
+      "grad_norm": 0.2287260890007019,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.7842,
+      "step": 78
+    },
+    {
+      "epoch": 0.03870177587262707,
+      "grad_norm": 0.22993454337120056,
+      "learning_rate": 3.95e-06,
+      "loss": 0.7554,
+      "step": 79
+    },
+    {
+      "epoch": 0.03919167176974893,
+      "grad_norm": 0.22278738021850586,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.7796,
+      "step": 80
+    },
+    {
+      "epoch": 0.03968156766687079,
+      "grad_norm": 0.22093527019023895,
+      "learning_rate": 4.05e-06,
+      "loss": 0.7546,
+      "step": 81
+    },
+    {
+      "epoch": 0.04017146356399265,
+      "grad_norm": 0.22237907350063324,
+      "learning_rate": 4.1e-06,
+      "loss": 0.7762,
+      "step": 82
+    },
+    {
+      "epoch": 0.040661359461114516,
+      "grad_norm": 0.23848189413547516,
+      "learning_rate": 4.15e-06,
+      "loss": 0.7694,
+      "step": 83
+    },
+    {
+      "epoch": 0.041151255358236376,
+      "grad_norm": 0.22380702197551727,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7829,
+      "step": 84
+    },
+    {
+      "epoch": 0.041641151255358236,
+      "grad_norm": 0.22918623685836792,
+      "learning_rate": 4.25e-06,
+      "loss": 0.7784,
+      "step": 85
+    },
+    {
+      "epoch": 0.042131047152480096,
+      "grad_norm": 0.21957220137119293,
+      "learning_rate": 4.3e-06,
+      "loss": 0.7444,
+      "step": 86
+    },
+    {
+      "epoch": 0.04262094304960196,
+      "grad_norm": 0.2195909172296524,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.791,
+      "step": 87
+    },
+    {
+      "epoch": 0.04311083894672382,
+      "grad_norm": 0.22012612223625183,
+      "learning_rate": 4.4e-06,
+      "loss": 0.7374,
+      "step": 88
+    },
+    {
+      "epoch": 0.04360073484384568,
+      "grad_norm": 0.22052907943725586,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.7833,
+      "step": 89
+    },
+    {
+      "epoch": 0.04409063074096754,
+      "grad_norm": 0.22584034502506256,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7803,
+      "step": 90
+    },
+    {
+      "epoch": 0.0445805266380894,
+      "grad_norm": 0.21903306245803833,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.7222,
+      "step": 91
+    },
+    {
+      "epoch": 0.04507042253521127,
+      "grad_norm": 0.21120664477348328,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.7955,
+      "step": 92
+    },
+    {
+      "epoch": 0.04556031843233313,
+      "grad_norm": 0.22016657888889313,
+      "learning_rate": 4.65e-06,
+      "loss": 0.7551,
+      "step": 93
+    },
+    {
+      "epoch": 0.04605021432945499,
+      "grad_norm": 0.2233135849237442,
+      "learning_rate": 4.7e-06,
+      "loss": 0.7665,
+      "step": 94
+    },
+    {
+      "epoch": 0.04654011022657685,
+      "grad_norm": 0.21228142082691193,
+      "learning_rate": 4.75e-06,
+      "loss": 0.7513,
+      "step": 95
+    },
+    {
+      "epoch": 0.04703000612369872,
+      "grad_norm": 0.2213594615459442,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.757,
+      "step": 96
+    },
+    {
+      "epoch": 0.04751990202082058,
+      "grad_norm": 0.21695221960544586,
+      "learning_rate": 4.85e-06,
+      "loss": 0.7641,
+      "step": 97
+    },
+    {
+      "epoch": 0.04800979791794244,
+      "grad_norm": 0.2285292148590088,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.7558,
+      "step": 98
+    },
+    {
+      "epoch": 0.0484996938150643,
+      "grad_norm": 0.21678751707077026,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7418,
+      "step": 99
+    },
+    {
+      "epoch": 0.04898958971218616,
+      "grad_norm": 0.20614083111286163,
+      "learning_rate": 5e-06,
+      "loss": 0.7477,
+      "step": 100
+    },
+    {
+      "epoch": 0.049479485609308024,
+      "grad_norm": 0.2086937427520752,
+      "learning_rate": 4.999999659917706e-06,
+      "loss": 0.7649,
+      "step": 101
+    },
+    {
+      "epoch": 0.049969381506429884,
+      "grad_norm": 0.2142128199338913,
+      "learning_rate": 4.999998639670915e-06,
+      "loss": 0.7333,
+      "step": 102
+    },
+    {
+      "epoch": 0.050459277403551744,
+      "grad_norm": 0.21820193529129028,
+      "learning_rate": 4.999996939259905e-06,
+      "loss": 0.7542,
+      "step": 103
+    },
+    {
+      "epoch": 0.050949173300673604,
+      "grad_norm": 0.21764390170574188,
+      "learning_rate": 4.999994558685137e-06,
+      "loss": 0.7514,
+      "step": 104
+    },
+    {
+      "epoch": 0.05143906919779547,
+      "grad_norm": 0.21882116794586182,
+      "learning_rate": 4.999991497947262e-06,
+      "loss": 0.7423,
+      "step": 105
+    },
+    {
+      "epoch": 0.05192896509491733,
+      "grad_norm": 0.21646051108837128,
+      "learning_rate": 4.99998775704711e-06,
+      "loss": 0.7604,
+      "step": 106
+    },
+    {
+      "epoch": 0.05241886099203919,
+      "grad_norm": 0.2149217575788498,
+      "learning_rate": 4.9999833359857005e-06,
+      "loss": 0.7573,
+      "step": 107
+    },
+    {
+      "epoch": 0.05290875688916105,
+      "grad_norm": 0.21114769577980042,
+      "learning_rate": 4.9999782347642355e-06,
+      "loss": 0.7664,
+      "step": 108
+    },
+    {
+      "epoch": 0.05339865278628291,
+      "grad_norm": 0.23437361419200897,
+      "learning_rate": 4.9999724533841035e-06,
+      "loss": 0.7561,
+      "step": 109
+    },
+    {
+      "epoch": 0.05388854868340478,
+      "grad_norm": 0.21103020012378693,
+      "learning_rate": 4.999965991846876e-06,
+      "loss": 0.7931,
+      "step": 110
+    },
+    {
+      "epoch": 0.05437844458052664,
+      "grad_norm": 0.22028645873069763,
+      "learning_rate": 4.999958850154312e-06,
+      "loss": 0.7392,
+      "step": 111
+    },
+    {
+      "epoch": 0.0548683404776485,
+      "grad_norm": 0.21418073773384094,
+      "learning_rate": 4.999951028308353e-06,
+      "loss": 0.7394,
+      "step": 112
+    },
+    {
+      "epoch": 0.05535823637477036,
+      "grad_norm": 0.2170812487602234,
+      "learning_rate": 4.99994252631113e-06,
+      "loss": 0.7649,
+      "step": 113
+    },
+    {
+      "epoch": 0.055848132271892226,
+      "grad_norm": 0.2240699976682663,
+      "learning_rate": 4.999933344164954e-06,
+      "loss": 0.7407,
+      "step": 114
+    },
+    {
+      "epoch": 0.056338028169014086,
+      "grad_norm": 0.2185816615819931,
+      "learning_rate": 4.999923481872324e-06,
+      "loss": 0.7705,
+      "step": 115
+    },
+    {
+      "epoch": 0.056827924066135946,
+      "grad_norm": 0.21891696751117706,
+      "learning_rate": 4.999912939435922e-06,
+      "loss": 0.7243,
+      "step": 116
+    },
+    {
+      "epoch": 0.057317819963257806,
+      "grad_norm": 0.21337160468101501,
+      "learning_rate": 4.999901716858617e-06,
+      "loss": 0.7281,
+      "step": 117
+    },
+    {
+      "epoch": 0.057807715860379666,
+      "grad_norm": 0.22411711513996124,
+      "learning_rate": 4.9998898141434635e-06,
+      "loss": 0.7407,
+      "step": 118
+    },
+    {
+      "epoch": 0.05829761175750153,
+      "grad_norm": 0.21748706698417664,
+      "learning_rate": 4.999877231293698e-06,
+      "loss": 0.7421,
+      "step": 119
+    },
+    {
+      "epoch": 0.05878750765462339,
+      "grad_norm": 0.2109653800725937,
+      "learning_rate": 4.999863968312744e-06,
+      "loss": 0.748,
+      "step": 120
+    },
+    {
+      "epoch": 0.05927740355174525,
+      "grad_norm": 0.21555471420288086,
+      "learning_rate": 4.999850025204211e-06,
+      "loss": 0.7199,
+      "step": 121
+    },
+    {
+      "epoch": 0.05976729944886711,
+      "grad_norm": 0.221576988697052,
+      "learning_rate": 4.999835401971892e-06,
+      "loss": 0.7292,
+      "step": 122
+    },
+    {
+      "epoch": 0.06025719534598898,
+      "grad_norm": 0.21922191977500916,
+      "learning_rate": 4.9998200986197645e-06,
+      "loss": 0.7493,
+      "step": 123
+    },
+    {
+      "epoch": 0.06074709124311084,
+      "grad_norm": 0.22416526079177856,
+      "learning_rate": 4.999804115151994e-06,
+      "loss": 0.7382,
+      "step": 124
+    },
+    {
+      "epoch": 0.0612369871402327,
+      "grad_norm": 0.21385036408901215,
+      "learning_rate": 4.9997874515729275e-06,
+      "loss": 0.7542,
+      "step": 125
+    },
+    {
+      "epoch": 0.06172688303735456,
+      "grad_norm": 0.21742364764213562,
+      "learning_rate": 4.9997701078870985e-06,
+      "loss": 0.7512,
+      "step": 126
+    },
+    {
+      "epoch": 0.06221677893447643,
+      "grad_norm": 0.21706967055797577,
+      "learning_rate": 4.999752084099227e-06,
+      "loss": 0.7517,
+      "step": 127
+    },
+    {
+      "epoch": 0.06270667483159828,
+      "grad_norm": 0.24107079207897186,
+      "learning_rate": 4.999733380214215e-06,
+      "loss": 0.7558,
+      "step": 128
+    },
+    {
+      "epoch": 0.06319657072872015,
+      "grad_norm": 0.21388758718967438,
+      "learning_rate": 4.999713996237152e-06,
+      "loss": 0.765,
+      "step": 129
+    },
+    {
+      "epoch": 0.06368646662584201,
+      "grad_norm": 0.2108607441186905,
+      "learning_rate": 4.999693932173312e-06,
+      "loss": 0.7365,
+      "step": 130
+    },
+    {
+      "epoch": 0.06417636252296387,
+      "grad_norm": 0.21274609863758087,
+      "learning_rate": 4.999673188028153e-06,
+      "loss": 0.7529,
+      "step": 131
+    },
+    {
+      "epoch": 0.06466625842008573,
+      "grad_norm": 0.21276482939720154,
+      "learning_rate": 4.999651763807321e-06,
+      "loss": 0.7573,
+      "step": 132
+    },
+    {
+      "epoch": 0.06515615431720759,
+      "grad_norm": 0.22406885027885437,
+      "learning_rate": 4.999629659516641e-06,
+      "loss": 0.724,
+      "step": 133
+    },
+    {
+      "epoch": 0.06564605021432945,
+      "grad_norm": 0.22327333688735962,
+      "learning_rate": 4.99960687516213e-06,
+      "loss": 0.7507,
+      "step": 134
+    },
+    {
+      "epoch": 0.06613594611145132,
+      "grad_norm": 0.21817858517169952,
+      "learning_rate": 4.999583410749984e-06,
+      "loss": 0.7329,
+      "step": 135
+    },
+    {
+      "epoch": 0.06662584200857317,
+      "grad_norm": 0.2202960103750229,
+      "learning_rate": 4.99955926628659e-06,
+      "loss": 0.7001,
+      "step": 136
+    },
+    {
+      "epoch": 0.06711573790569504,
+      "grad_norm": 0.22170418500900269,
+      "learning_rate": 4.999534441778515e-06,
+      "loss": 0.7083,
+      "step": 137
+    },
+    {
+      "epoch": 0.0676056338028169,
+      "grad_norm": 0.22274421155452728,
+      "learning_rate": 4.999508937232514e-06,
+      "loss": 0.7545,
+      "step": 138
+    },
+    {
+      "epoch": 0.06809552969993876,
+      "grad_norm": 0.22415056824684143,
+      "learning_rate": 4.999482752655524e-06,
+      "loss": 0.7451,
+      "step": 139
+    },
+    {
+      "epoch": 0.06858542559706063,
+      "grad_norm": 0.20987579226493835,
+      "learning_rate": 4.999455888054672e-06,
+      "loss": 0.7274,
+      "step": 140
+    },
+    {
+      "epoch": 0.06907532149418248,
+      "grad_norm": 0.2241065949201584,
+      "learning_rate": 4.999428343437264e-06,
+      "loss": 0.7422,
+      "step": 141
+    },
+    {
+      "epoch": 0.06956521739130435,
+      "grad_norm": 0.2241523563861847,
+      "learning_rate": 4.999400118810794e-06,
+      "loss": 0.73,
+      "step": 142
+    },
+    {
+      "epoch": 0.07005511328842622,
+      "grad_norm": 0.21995913982391357,
+      "learning_rate": 4.999371214182944e-06,
+      "loss": 0.7614,
+      "step": 143
+    },
+    {
+      "epoch": 0.07054500918554807,
+      "grad_norm": 0.22398804128170013,
+      "learning_rate": 4.999341629561575e-06,
+      "loss": 0.7263,
+      "step": 144
+    },
+    {
+      "epoch": 0.07103490508266994,
+      "grad_norm": 0.2156529575586319,
+      "learning_rate": 4.9993113649547364e-06,
+      "loss": 0.759,
+      "step": 145
+    },
+    {
+      "epoch": 0.07152480097979179,
+      "grad_norm": 0.22519437968730927,
+      "learning_rate": 4.999280420370664e-06,
+      "loss": 0.7556,
+      "step": 146
+    },
+    {
+      "epoch": 0.07201469687691366,
+      "grad_norm": 0.22073966264724731,
+      "learning_rate": 4.999248795817774e-06,
+      "loss": 0.7485,
+      "step": 147
+    },
+    {
+      "epoch": 0.07250459277403552,
+      "grad_norm": 0.2195131778717041,
+      "learning_rate": 4.999216491304673e-06,
+      "loss": 0.732,
+      "step": 148
+    },
+    {
+      "epoch": 0.07299448867115738,
+      "grad_norm": 0.23442167043685913,
+      "learning_rate": 4.999183506840149e-06,
+      "loss": 0.7593,
+      "step": 149
+    },
+    {
+      "epoch": 0.07348438456827924,
+      "grad_norm": 0.225722998380661,
+      "learning_rate": 4.999149842433175e-06,
+      "loss": 0.7417,
+      "step": 150
+    },
+    {
+      "epoch": 0.0739742804654011,
+      "grad_norm": 0.22025473415851593,
+      "learning_rate": 4.9991154980929104e-06,
+      "loss": 0.7449,
+      "step": 151
+    },
+    {
+      "epoch": 0.07446417636252296,
+      "grad_norm": 0.23160138726234436,
+      "learning_rate": 4.9990804738287005e-06,
+      "loss": 0.7409,
+      "step": 152
+    },
+    {
+      "epoch": 0.07495407225964483,
+      "grad_norm": 0.21675211191177368,
+      "learning_rate": 4.999044769650072e-06,
+      "loss": 0.7417,
+      "step": 153
+    },
+    {
+      "epoch": 0.07544396815676668,
+      "grad_norm": 0.22495831549167633,
+      "learning_rate": 4.999008385566741e-06,
+      "loss": 0.7398,
+      "step": 154
+    },
+    {
+      "epoch": 0.07593386405388855,
+      "grad_norm": 0.22678421437740326,
+      "learning_rate": 4.9989713215886036e-06,
+      "loss": 0.7217,
+      "step": 155
+    },
+    {
+      "epoch": 0.07642375995101042,
+      "grad_norm": 0.21712447702884674,
+      "learning_rate": 4.998933577725746e-06,
+      "loss": 0.7394,
+      "step": 156
+    },
+    {
+      "epoch": 0.07691365584813227,
+      "grad_norm": 0.2327713668346405,
+      "learning_rate": 4.998895153988437e-06,
+      "loss": 0.7475,
+      "step": 157
+    },
+    {
+      "epoch": 0.07740355174525414,
+      "grad_norm": 0.21809875965118408,
+      "learning_rate": 4.998856050387129e-06,
+      "loss": 0.7397,
+      "step": 158
+    },
+    {
+      "epoch": 0.07789344764237599,
+      "grad_norm": 0.2248068004846573,
+      "learning_rate": 4.998816266932462e-06,
+      "loss": 0.7504,
+      "step": 159
+    },
+    {
+      "epoch": 0.07838334353949786,
+      "grad_norm": 0.22332903742790222,
+      "learning_rate": 4.998775803635259e-06,
+      "loss": 0.7187,
+      "step": 160
+    },
+    {
+      "epoch": 0.07887323943661972,
+      "grad_norm": 0.23418764770030975,
+      "learning_rate": 4.998734660506529e-06,
+      "loss": 0.7269,
+      "step": 161
+    },
+    {
+      "epoch": 0.07936313533374158,
+      "grad_norm": 0.21334218978881836,
+      "learning_rate": 4.998692837557465e-06,
+      "loss": 0.7228,
+      "step": 162
+    },
+    {
+      "epoch": 0.07985303123086344,
+      "grad_norm": 0.2249280959367752,
+      "learning_rate": 4.998650334799446e-06,
+      "loss": 0.7569,
+      "step": 163
+    },
+    {
+      "epoch": 0.0803429271279853,
+      "grad_norm": 0.22758226096630096,
+      "learning_rate": 4.998607152244036e-06,
+      "loss": 0.7351,
+      "step": 164
+    },
+    {
+      "epoch": 0.08083282302510716,
+      "grad_norm": 0.21682260930538177,
+      "learning_rate": 4.998563289902984e-06,
+      "loss": 0.705,
+      "step": 165
+    },
+    {
+      "epoch": 0.08132271892222903,
+      "grad_norm": 0.22186994552612305,
+      "learning_rate": 4.998518747788221e-06,
+      "loss": 0.6869,
+      "step": 166
+    },
+    {
+      "epoch": 0.08181261481935088,
+      "grad_norm": 0.2218897044658661,
+      "learning_rate": 4.9984735259118675e-06,
+      "loss": 0.7516,
+      "step": 167
+    },
+    {
+      "epoch": 0.08230251071647275,
+      "grad_norm": 0.24786899983882904,
+      "learning_rate": 4.9984276242862265e-06,
+      "loss": 0.7292,
+      "step": 168
+    },
+    {
+      "epoch": 0.0827924066135946,
+      "grad_norm": 0.22796982526779175,
+      "learning_rate": 4.998381042923786e-06,
+      "loss": 0.708,
+      "step": 169
+    },
+    {
+      "epoch": 0.08328230251071647,
+      "grad_norm": 0.2350655496120453,
+      "learning_rate": 4.998333781837219e-06,
+      "loss": 0.7427,
+      "step": 170
+    },
+    {
+      "epoch": 0.08377219840783834,
+      "grad_norm": 0.21661469340324402,
+      "learning_rate": 4.998285841039384e-06,
+      "loss": 0.7173,
+      "step": 171
+    },
+    {
+      "epoch": 0.08426209430496019,
+      "grad_norm": 0.2350396364927292,
+      "learning_rate": 4.998237220543324e-06,
+      "loss": 0.7485,
+      "step": 172
+    },
+    {
+      "epoch": 0.08475199020208206,
+      "grad_norm": 0.2210429608821869,
+      "learning_rate": 4.998187920362268e-06,
+      "loss": 0.7455,
+      "step": 173
+    },
+    {
+      "epoch": 0.08524188609920393,
+      "grad_norm": 0.23016513884067535,
+      "learning_rate": 4.998137940509626e-06,
+      "loss": 0.7116,
+      "step": 174
+    },
+    {
+      "epoch": 0.08573178199632578,
+      "grad_norm": 0.21964678168296814,
+      "learning_rate": 4.998087280998999e-06,
+      "loss": 0.7356,
+      "step": 175
+    },
+    {
+      "epoch": 0.08622167789344765,
+      "grad_norm": 0.22432510554790497,
+      "learning_rate": 4.998035941844167e-06,
+      "loss": 0.7315,
+      "step": 176
+    },
+    {
+      "epoch": 0.0867115737905695,
+      "grad_norm": 0.2379545122385025,
+      "learning_rate": 4.9979839230591e-06,
+      "loss": 0.7165,
+      "step": 177
+    },
+    {
+      "epoch": 0.08720146968769137,
+      "grad_norm": 0.234178826212883,
+      "learning_rate": 4.99793122465795e-06,
+      "loss": 0.7378,
+      "step": 178
+    },
+    {
+      "epoch": 0.08769136558481323,
+      "grad_norm": 0.216177836060524,
+      "learning_rate": 4.9978778466550524e-06,
+      "loss": 0.6997,
+      "step": 179
+    },
+    {
+      "epoch": 0.08818126148193509,
+      "grad_norm": 0.2306894212961197,
+      "learning_rate": 4.9978237890649315e-06,
+      "loss": 0.7341,
+      "step": 180
+    },
+    {
+      "epoch": 0.08867115737905695,
+      "grad_norm": 0.21481899917125702,
+      "learning_rate": 4.9977690519022945e-06,
+      "loss": 0.7151,
+      "step": 181
+    },
+    {
+      "epoch": 0.0891610532761788,
+      "grad_norm": 0.2187681943178177,
+      "learning_rate": 4.997713635182033e-06,
+      "loss": 0.7497,
+      "step": 182
+    },
+    {
+      "epoch": 0.08965094917330067,
+      "grad_norm": 0.22013579308986664,
+      "learning_rate": 4.997657538919224e-06,
+      "loss": 0.7094,
+      "step": 183
+    },
+    {
+      "epoch": 0.09014084507042254,
+      "grad_norm": 0.22705145180225372,
+      "learning_rate": 4.99760076312913e-06,
+      "loss": 0.7298,
+      "step": 184
+    },
+    {
+      "epoch": 0.0906307409675444,
+      "grad_norm": 0.2299501895904541,
+      "learning_rate": 4.997543307827196e-06,
+      "loss": 0.7584,
+      "step": 185
+    },
+    {
+      "epoch": 0.09112063686466626,
+      "grad_norm": 0.22536176443099976,
+      "learning_rate": 4.997485173029056e-06,
+      "loss": 0.711,
+      "step": 186
+    },
+    {
+      "epoch": 0.09161053276178811,
+      "grad_norm": 0.22876884043216705,
+      "learning_rate": 4.997426358750525e-06,
+      "loss": 0.6999,
+      "step": 187
+    },
+    {
+      "epoch": 0.09210042865890998,
+      "grad_norm": 0.2220255434513092,
+      "learning_rate": 4.997366865007605e-06,
+      "loss": 0.7033,
+      "step": 188
+    },
+    {
+      "epoch": 0.09259032455603185,
+      "grad_norm": 0.22282293438911438,
+      "learning_rate": 4.9973066918164815e-06,
+      "loss": 0.7296,
+      "step": 189
+    },
+    {
+      "epoch": 0.0930802204531537,
+      "grad_norm": 0.2244081050157547,
+      "learning_rate": 4.997245839193526e-06,
+      "loss": 0.7217,
+      "step": 190
+    },
+    {
+      "epoch": 0.09357011635027557,
+      "grad_norm": 0.24491725862026215,
+      "learning_rate": 4.997184307155294e-06,
+      "loss": 0.6933,
+      "step": 191
+    },
+    {
+      "epoch": 0.09406001224739743,
+      "grad_norm": 0.2341097742319107,
+      "learning_rate": 4.997122095718527e-06,
+      "loss": 0.7156,
+      "step": 192
+    },
+    {
+      "epoch": 0.09454990814451929,
+      "grad_norm": 0.23293037712574005,
+      "learning_rate": 4.997059204900151e-06,
+      "loss": 0.7219,
+      "step": 193
+    },
+    {
+      "epoch": 0.09503980404164115,
+      "grad_norm": 0.224905863404274,
+      "learning_rate": 4.996995634717276e-06,
+      "loss": 0.6925,
+      "step": 194
+    },
+    {
+      "epoch": 0.09552969993876301,
+      "grad_norm": 0.23174867033958435,
+      "learning_rate": 4.996931385187195e-06,
+      "loss": 0.7354,
+      "step": 195
+    },
+    {
+      "epoch": 0.09601959583588487,
+      "grad_norm": 0.2354530245065689,
+      "learning_rate": 4.996866456327392e-06,
+      "loss": 0.7218,
+      "step": 196
+    },
+    {
+      "epoch": 0.09650949173300674,
+      "grad_norm": 0.22878359258174896,
+      "learning_rate": 4.996800848155529e-06,
+      "loss": 0.7289,
+      "step": 197
+    },
+    {
+      "epoch": 0.0969993876301286,
+      "grad_norm": 0.22221261262893677,
+      "learning_rate": 4.996734560689457e-06,
+      "loss": 0.7276,
+      "step": 198
+    },
+    {
+      "epoch": 0.09748928352725046,
+      "grad_norm": 0.23485805094242096,
+      "learning_rate": 4.9966675939472094e-06,
+      "loss": 0.7407,
+      "step": 199
+    },
+    {
+      "epoch": 0.09797917942437231,
+      "grad_norm": 0.2275373488664627,
+      "learning_rate": 4.996599947947009e-06,
+      "loss": 0.7187,
+      "step": 200
+    },
+    {
+      "epoch": 0.09846907532149418,
+      "grad_norm": 0.22819741070270538,
+      "learning_rate": 4.996531622707255e-06,
+      "loss": 0.717,
+      "step": 201
+    },
+    {
+      "epoch": 0.09895897121861605,
+      "grad_norm": 0.24706602096557617,
+      "learning_rate": 4.99646261824654e-06,
+      "loss": 0.7126,
+      "step": 202
+    },
+    {
+      "epoch": 0.0994488671157379,
+      "grad_norm": 0.2240106165409088,
+      "learning_rate": 4.996392934583636e-06,
+      "loss": 0.7269,
+      "step": 203
+    },
+    {
+      "epoch": 0.09993876301285977,
+      "grad_norm": 0.23807524144649506,
+      "learning_rate": 4.996322571737502e-06,
+      "loss": 0.719,
+      "step": 204
+    },
+    {
+      "epoch": 0.10042865890998164,
+      "grad_norm": 0.23285457491874695,
+      "learning_rate": 4.996251529727282e-06,
+      "loss": 0.7088,
+      "step": 205
+    },
+    {
+      "epoch": 0.10091855480710349,
+      "grad_norm": 0.227351576089859,
+      "learning_rate": 4.996179808572303e-06,
+      "loss": 0.74,
+      "step": 206
+    },
+    {
+      "epoch": 0.10140845070422536,
+      "grad_norm": 0.2282775193452835,
+      "learning_rate": 4.996107408292079e-06,
+      "loss": 0.7037,
+      "step": 207
+    },
+    {
+      "epoch": 0.10189834660134721,
+      "grad_norm": 0.23110702633857727,
+      "learning_rate": 4.996034328906308e-06,
+      "loss": 0.7153,
+      "step": 208
+    },
+    {
+      "epoch": 0.10238824249846908,
+      "grad_norm": 0.2378554791212082,
+      "learning_rate": 4.995960570434869e-06,
+      "loss": 0.7097,
+      "step": 209
+    },
+    {
+      "epoch": 0.10287813839559094,
+      "grad_norm": 0.2271486073732376,
+      "learning_rate": 4.995886132897833e-06,
+      "loss": 0.7112,
+      "step": 210
+    },
+    {
+      "epoch": 0.1033680342927128,
+      "grad_norm": 0.24688458442687988,
+      "learning_rate": 4.99581101631545e-06,
+      "loss": 0.6905,
+      "step": 211
+    },
+    {
+      "epoch": 0.10385793018983466,
+      "grad_norm": 0.23487895727157593,
+      "learning_rate": 4.995735220708158e-06,
+      "loss": 0.6851,
+      "step": 212
+    },
+    {
+      "epoch": 0.10434782608695652,
+      "grad_norm": 0.23712845146656036,
+      "learning_rate": 4.995658746096577e-06,
+      "loss": 0.7162,
+      "step": 213
+    },
+    {
+      "epoch": 0.10483772198407838,
+      "grad_norm": 0.23982517421245575,
+      "learning_rate": 4.995581592501514e-06,
+      "loss": 0.7172,
+      "step": 214
+    },
+    {
+      "epoch": 0.10532761788120025,
+      "grad_norm": 0.24245846271514893,
+      "learning_rate": 4.99550375994396e-06,
+      "loss": 0.7461,
+      "step": 215
+    },
+    {
+      "epoch": 0.1058175137783221,
+      "grad_norm": 0.25469428300857544,
+      "learning_rate": 4.995425248445089e-06,
+      "loss": 0.7231,
+      "step": 216
+    },
+    {
+      "epoch": 0.10630740967544397,
+      "grad_norm": 0.23790335655212402,
+      "learning_rate": 4.995346058026263e-06,
+      "loss": 0.6956,
+      "step": 217
+    },
+    {
+      "epoch": 0.10679730557256582,
+      "grad_norm": 0.23641760647296906,
+      "learning_rate": 4.995266188709027e-06,
+      "loss": 0.7264,
+      "step": 218
+    },
+    {
+      "epoch": 0.10728720146968769,
+      "grad_norm": 0.24292510747909546,
+      "learning_rate": 4.99518564051511e-06,
+      "loss": 0.6772,
+      "step": 219
+    },
+    {
+      "epoch": 0.10777709736680956,
+      "grad_norm": 0.2255912572145462,
+      "learning_rate": 4.995104413466426e-06,
+      "loss": 0.6805,
+      "step": 220
+    },
+    {
+      "epoch": 0.10826699326393141,
+      "grad_norm": 0.24550217390060425,
+      "learning_rate": 4.995022507585075e-06,
+      "loss": 0.7456,
+      "step": 221
+    },
+    {
+      "epoch": 0.10875688916105328,
+      "grad_norm": 0.24428114295005798,
+      "learning_rate": 4.99493992289334e-06,
+      "loss": 0.7485,
+      "step": 222
+    },
+    {
+      "epoch": 0.10924678505817514,
+      "grad_norm": 0.2401362955570221,
+      "learning_rate": 4.994856659413691e-06,
+      "loss": 0.7092,
+      "step": 223
+    },
+    {
+      "epoch": 0.109736680955297,
+      "grad_norm": 0.24536854028701782,
+      "learning_rate": 4.99477271716878e-06,
+      "loss": 0.6942,
+      "step": 224
+    },
+    {
+      "epoch": 0.11022657685241886,
+      "grad_norm": 0.23893441259860992,
+      "learning_rate": 4.994688096181444e-06,
+      "loss": 0.7108,
+      "step": 225
+    },
+    {
+      "epoch": 0.11071647274954072,
+      "grad_norm": 0.2415887713432312,
+      "learning_rate": 4.994602796474707e-06,
+      "loss": 0.6844,
+      "step": 226
+    },
+    {
+      "epoch": 0.11120636864666258,
+      "grad_norm": 0.2488085925579071,
+      "learning_rate": 4.994516818071775e-06,
+      "loss": 0.7047,
+      "step": 227
+    },
+    {
+      "epoch": 0.11169626454378445,
+      "grad_norm": 0.24722400307655334,
+      "learning_rate": 4.994430160996041e-06,
+      "loss": 0.7382,
+      "step": 228
+    },
+    {
+      "epoch": 0.1121861604409063,
+      "grad_norm": 0.24387723207473755,
+      "learning_rate": 4.99434282527108e-06,
+      "loss": 0.7109,
+      "step": 229
+    },
+    {
+      "epoch": 0.11267605633802817,
+      "grad_norm": 0.24834807217121124,
+      "learning_rate": 4.9942548109206545e-06,
+      "loss": 0.7077,
+      "step": 230
+    },
+    {
+      "epoch": 0.11316595223515002,
+      "grad_norm": 0.23779404163360596,
+      "learning_rate": 4.994166117968709e-06,
+      "loss": 0.6803,
+      "step": 231
+    },
+    {
+      "epoch": 0.11365584813227189,
+      "grad_norm": 0.24619951844215393,
+      "learning_rate": 4.994076746439375e-06,
+      "loss": 0.7001,
+      "step": 232
+    },
+    {
+      "epoch": 0.11414574402939376,
+      "grad_norm": 0.23684336245059967,
+      "learning_rate": 4.993986696356966e-06,
+      "loss": 0.7151,
+      "step": 233
+    },
+    {
+      "epoch": 0.11463563992651561,
+      "grad_norm": 0.26148277521133423,
+      "learning_rate": 4.993895967745983e-06,
+      "loss": 0.6959,
+      "step": 234
+    },
+    {
+      "epoch": 0.11512553582363748,
+      "grad_norm": 0.24303443729877472,
+      "learning_rate": 4.993804560631109e-06,
+      "loss": 0.7427,
+      "step": 235
+    },
+    {
+      "epoch": 0.11561543172075933,
+      "grad_norm": 0.23858515918254852,
+      "learning_rate": 4.993712475037213e-06,
+      "loss": 0.7167,
+      "step": 236
+    },
+    {
+      "epoch": 0.1161053276178812,
+      "grad_norm": 0.23997746407985687,
+      "learning_rate": 4.993619710989349e-06,
+      "loss": 0.7456,
+      "step": 237
+    },
+    {
+      "epoch": 0.11659522351500307,
+      "grad_norm": 0.24167482554912567,
+      "learning_rate": 4.993526268512755e-06,
+      "loss": 0.7072,
+      "step": 238
+    },
+    {
+      "epoch": 0.11708511941212492,
+      "grad_norm": 0.2411864548921585,
+      "learning_rate": 4.9934321476328515e-06,
+      "loss": 0.6987,
+      "step": 239
+    },
+    {
+      "epoch": 0.11757501530924679,
+      "grad_norm": 0.2430465817451477,
+      "learning_rate": 4.993337348375249e-06,
+      "loss": 0.7009,
+      "step": 240
+    },
+    {
+      "epoch": 0.11806491120636865,
+      "grad_norm": 0.2417476773262024,
+      "learning_rate": 4.993241870765736e-06,
+      "loss": 0.695,
+      "step": 241
+    },
+    {
+      "epoch": 0.1185548071034905,
+      "grad_norm": 0.24546459317207336,
+      "learning_rate": 4.99314571483029e-06,
+      "loss": 0.7091,
+      "step": 242
+    },
+    {
+      "epoch": 0.11904470300061237,
+      "grad_norm": 0.2377326339483261,
+      "learning_rate": 4.993048880595072e-06,
+      "loss": 0.7005,
+      "step": 243
+    },
+    {
+      "epoch": 0.11953459889773423,
+      "grad_norm": 0.2439325898885727,
+      "learning_rate": 4.992951368086427e-06,
+      "loss": 0.6956,
+      "step": 244
+    },
+    {
+      "epoch": 0.1200244947948561,
+      "grad_norm": 0.2318105548620224,
+      "learning_rate": 4.992853177330884e-06,
+      "loss": 0.708,
+      "step": 245
+    },
+    {
+      "epoch": 0.12051439069197796,
+      "grad_norm": 0.24266572296619415,
+      "learning_rate": 4.992754308355159e-06,
+      "loss": 0.7126,
+      "step": 246
+    },
+    {
+      "epoch": 0.12100428658909981,
+      "grad_norm": 0.24958163499832153,
+      "learning_rate": 4.99265476118615e-06,
+      "loss": 0.7165,
+      "step": 247
+    },
+    {
+      "epoch": 0.12149418248622168,
+      "grad_norm": 0.24117092788219452,
+      "learning_rate": 4.99255453585094e-06,
+      "loss": 0.6977,
+      "step": 248
+    },
+    {
+      "epoch": 0.12198407838334353,
+      "grad_norm": 0.24092306196689606,
+      "learning_rate": 4.992453632376797e-06,
+      "loss": 0.7179,
+      "step": 249
+    },
+    {
+      "epoch": 0.1224739742804654,
+      "grad_norm": 0.24704957008361816,
+      "learning_rate": 4.992352050791175e-06,
+      "loss": 0.6955,
+      "step": 250
+    },
+    {
+      "epoch": 0.12296387017758727,
+      "grad_norm": 0.23530827462673187,
+      "learning_rate": 4.992249791121709e-06,
+      "loss": 0.7051,
+      "step": 251
+    },
+    {
+      "epoch": 0.12345376607470912,
+      "grad_norm": 0.2405635416507721,
+      "learning_rate": 4.992146853396219e-06,
+      "loss": 0.7247,
+      "step": 252
+    },
+    {
+      "epoch": 0.12394366197183099,
+      "grad_norm": 0.23462170362472534,
+      "learning_rate": 4.992043237642715e-06,
+      "loss": 0.709,
+      "step": 253
+    },
+    {
+      "epoch": 0.12443355786895285,
+      "grad_norm": 0.24728280305862427,
+      "learning_rate": 4.991938943889384e-06,
+      "loss": 0.6904,
+      "step": 254
+    },
+    {
+      "epoch": 0.12492345376607471,
+      "grad_norm": 0.2390342354774475,
+      "learning_rate": 4.991833972164602e-06,
+      "loss": 0.7031,
+      "step": 255
+    },
+    {
+      "epoch": 0.12541334966319656,
+      "grad_norm": 0.25283288955688477,
+      "learning_rate": 4.991728322496928e-06,
+      "loss": 0.717,
+      "step": 256
+    },
+    {
+      "epoch": 0.12590324556031843,
+      "grad_norm": 0.23909400403499603,
+      "learning_rate": 4.991621994915105e-06,
+      "loss": 0.6916,
+      "step": 257
+    },
+    {
+      "epoch": 0.1263931414574403,
+      "grad_norm": 0.24953396618366241,
+      "learning_rate": 4.991514989448063e-06,
+      "loss": 0.6901,
+      "step": 258
+    },
+    {
+      "epoch": 0.12688303735456216,
+      "grad_norm": 0.24343158304691315,
+      "learning_rate": 4.991407306124914e-06,
+      "loss": 0.7018,
+      "step": 259
+    },
+    {
+      "epoch": 0.12737293325168403,
+      "grad_norm": 0.24725763499736786,
+      "learning_rate": 4.991298944974953e-06,
+      "loss": 0.6996,
+      "step": 260
+    },
+    {
+      "epoch": 0.12786282914880587,
+      "grad_norm": 0.25944840908050537,
+      "learning_rate": 4.991189906027663e-06,
+      "loss": 0.7083,
+      "step": 261
+    },
+    {
+      "epoch": 0.12835272504592773,
+      "grad_norm": 0.2454294115304947,
+      "learning_rate": 4.991080189312709e-06,
+      "loss": 0.7058,
+      "step": 262
+    },
+    {
+      "epoch": 0.1288426209430496,
+      "grad_norm": 0.2501434087753296,
+      "learning_rate": 4.990969794859941e-06,
+      "loss": 0.7102,
+      "step": 263
+    },
+    {
+      "epoch": 0.12933251684017147,
+      "grad_norm": 0.26826831698417664,
+      "learning_rate": 4.990858722699395e-06,
+      "loss": 0.7305,
+      "step": 264
+    },
+    {
+      "epoch": 0.12982241273729334,
+      "grad_norm": 0.23846790194511414,
+      "learning_rate": 4.990746972861289e-06,
+      "loss": 0.6752,
+      "step": 265
+    },
+    {
+      "epoch": 0.13031230863441517,
+      "grad_norm": 0.25017425417900085,
+      "learning_rate": 4.990634545376027e-06,
+      "loss": 0.7062,
+      "step": 266
+    },
+    {
+      "epoch": 0.13080220453153704,
+      "grad_norm": 0.2652483880519867,
+      "learning_rate": 4.990521440274195e-06,
+      "loss": 0.7117,
+      "step": 267
+    },
+    {
+      "epoch": 0.1312921004286589,
+      "grad_norm": 0.24066515266895294,
+      "learning_rate": 4.990407657586568e-06,
+      "loss": 0.7193,
+      "step": 268
+    },
+    {
+      "epoch": 0.13178199632578078,
+      "grad_norm": 0.2460177093744278,
+      "learning_rate": 4.990293197344098e-06,
+      "loss": 0.6644,
+      "step": 269
+    },
+    {
+      "epoch": 0.13227189222290264,
+      "grad_norm": 0.2507753372192383,
+      "learning_rate": 4.99017805957793e-06,
+      "loss": 0.7114,
+      "step": 270
+    },
+    {
+      "epoch": 0.13276178812002448,
+      "grad_norm": 0.25557348132133484,
+      "learning_rate": 4.990062244319387e-06,
+      "loss": 0.6895,
+      "step": 271
+    },
+    {
+      "epoch": 0.13325168401714635,
+      "grad_norm": 0.2597479522228241,
+      "learning_rate": 4.989945751599978e-06,
+      "loss": 0.7285,
+      "step": 272
+    },
+    {
+      "epoch": 0.13374157991426822,
+      "grad_norm": 0.2611940801143646,
+      "learning_rate": 4.989828581451398e-06,
+      "loss": 0.7296,
+      "step": 273
+    },
+    {
+      "epoch": 0.13423147581139008,
+      "grad_norm": 0.2552890479564667,
+      "learning_rate": 4.989710733905524e-06,
+      "loss": 0.7324,
+      "step": 274
+    },
+    {
+      "epoch": 0.13472137170851195,
+      "grad_norm": 0.2605132460594177,
+      "learning_rate": 4.9895922089944185e-06,
+      "loss": 0.6969,
+      "step": 275
+    },
+    {
+      "epoch": 0.1352112676056338,
+      "grad_norm": 0.25689518451690674,
+      "learning_rate": 4.989473006750328e-06,
+      "loss": 0.7032,
+      "step": 276
+    },
+    {
+      "epoch": 0.13570116350275566,
+      "grad_norm": 0.25465458631515503,
+      "learning_rate": 4.989353127205684e-06,
+      "loss": 0.6865,
+      "step": 277
+    },
+    {
+      "epoch": 0.13619105939987752,
+      "grad_norm": 0.25785622000694275,
+      "learning_rate": 4.989232570393101e-06,
+      "loss": 0.7007,
+      "step": 278
+    },
+    {
+      "epoch": 0.1366809552969994,
+      "grad_norm": 0.25546690821647644,
+      "learning_rate": 4.989111336345378e-06,
+      "loss": 0.7342,
+      "step": 279
+    },
+    {
+      "epoch": 0.13717085119412126,
+      "grad_norm": 0.25426116585731506,
+      "learning_rate": 4.9889894250955005e-06,
+      "loss": 0.6978,
+      "step": 280
+    },
+    {
+      "epoch": 0.13766074709124312,
+      "grad_norm": 0.25376084446907043,
+      "learning_rate": 4.988866836676635e-06,
+      "loss": 0.6856,
+      "step": 281
+    },
+    {
+      "epoch": 0.13815064298836496,
+      "grad_norm": 0.2518889307975769,
+      "learning_rate": 4.988743571122133e-06,
+      "loss": 0.6964,
+      "step": 282
+    },
+    {
+      "epoch": 0.13864053888548683,
+      "grad_norm": 0.24486613273620605,
+      "learning_rate": 4.9886196284655315e-06,
+      "loss": 0.7082,
+      "step": 283
+    },
+    {
+      "epoch": 0.1391304347826087,
+      "grad_norm": 0.26591217517852783,
+      "learning_rate": 4.988495008740552e-06,
+      "loss": 0.6737,
+      "step": 284
+    },
+    {
+      "epoch": 0.13962033067973056,
+      "grad_norm": 0.2720721960067749,
+      "learning_rate": 4.988369711981097e-06,
+      "loss": 0.7142,
+      "step": 285
+    },
+    {
+      "epoch": 0.14011022657685243,
+      "grad_norm": 0.2502274215221405,
+      "learning_rate": 4.988243738221258e-06,
+      "loss": 0.6928,
+      "step": 286
+    },
+    {
+      "epoch": 0.14060012247397427,
+      "grad_norm": 0.2472478747367859,
+      "learning_rate": 4.988117087495306e-06,
+      "loss": 0.7081,
+      "step": 287
+    },
+    {
+      "epoch": 0.14109001837109614,
+      "grad_norm": 0.2597702145576477,
+      "learning_rate": 4.9879897598377005e-06,
+      "loss": 0.7252,
+      "step": 288
+    },
+    {
+      "epoch": 0.141579914268218,
+      "grad_norm": 0.2554181218147278,
+      "learning_rate": 4.987861755283081e-06,
+      "loss": 0.6871,
+      "step": 289
+    },
+    {
+      "epoch": 0.14206981016533987,
+      "grad_norm": 0.2611641585826874,
+      "learning_rate": 4.9877330738662755e-06,
+      "loss": 0.6922,
+      "step": 290
+    },
+    {
+      "epoch": 0.14255970606246174,
+      "grad_norm": 0.2400081306695938,
+      "learning_rate": 4.987603715622291e-06,
+      "loss": 0.6572,
+      "step": 291
+    },
+    {
+      "epoch": 0.14304960195958358,
+      "grad_norm": 0.2615026831626892,
+      "learning_rate": 4.987473680586323e-06,
+      "loss": 0.7023,
+      "step": 292
+    },
+    {
+      "epoch": 0.14353949785670544,
+      "grad_norm": 0.25383347272872925,
+      "learning_rate": 4.98734296879375e-06,
+      "loss": 0.6958,
+      "step": 293
+    },
+    {
+      "epoch": 0.1440293937538273,
+      "grad_norm": 0.25849246978759766,
+      "learning_rate": 4.987211580280133e-06,
+      "loss": 0.7196,
+      "step": 294
+    },
+    {
+      "epoch": 0.14451928965094918,
+      "grad_norm": 0.26360154151916504,
+      "learning_rate": 4.987079515081219e-06,
+      "loss": 0.7121,
+      "step": 295
+    },
+    {
+      "epoch": 0.14500918554807105,
+      "grad_norm": 0.2684732973575592,
+      "learning_rate": 4.986946773232939e-06,
+      "loss": 0.7198,
+      "step": 296
+    },
+    {
+      "epoch": 0.14549908144519288,
+      "grad_norm": 0.2610439360141754,
+      "learning_rate": 4.986813354771406e-06,
+      "loss": 0.6924,
+      "step": 297
+    },
+    {
+      "epoch": 0.14598897734231475,
+      "grad_norm": 0.25493988394737244,
+      "learning_rate": 4.986679259732919e-06,
+      "loss": 0.7354,
+      "step": 298
+    },
+    {
+      "epoch": 0.14647887323943662,
+      "grad_norm": 0.2612632215023041,
+      "learning_rate": 4.986544488153963e-06,
+      "loss": 0.6878,
+      "step": 299
+    },
+    {
+      "epoch": 0.14696876913655849,
+      "grad_norm": 0.25811073184013367,
+      "learning_rate": 4.986409040071202e-06,
+      "loss": 0.6793,
+      "step": 300
+    },
+    {
+      "epoch": 0.14745866503368035,
+      "grad_norm": 0.2634781301021576,
+      "learning_rate": 4.986272915521487e-06,
+      "loss": 0.7061,
+      "step": 301
+    },
+    {
+      "epoch": 0.1479485609308022,
+      "grad_norm": 0.25826555490493774,
+      "learning_rate": 4.9861361145418545e-06,
+      "loss": 0.664,
+      "step": 302
+    },
+    {
+      "epoch": 0.14843845682792406,
+      "grad_norm": 0.24985066056251526,
+      "learning_rate": 4.985998637169522e-06,
+      "loss": 0.6932,
+      "step": 303
+    },
+    {
+      "epoch": 0.14892835272504593,
+      "grad_norm": 0.25596854090690613,
+      "learning_rate": 4.985860483441894e-06,
+      "loss": 0.6568,
+      "step": 304
+    },
+    {
+      "epoch": 0.1494182486221678,
+      "grad_norm": 0.2568427324295044,
+      "learning_rate": 4.985721653396555e-06,
+      "loss": 0.6889,
+      "step": 305
+    },
+    {
+      "epoch": 0.14990814451928966,
+      "grad_norm": 0.2716825604438782,
+      "learning_rate": 4.985582147071277e-06,
+      "loss": 0.6813,
+      "step": 306
+    },
+    {
+      "epoch": 0.1503980404164115,
+      "grad_norm": 0.2683699131011963,
+      "learning_rate": 4.985441964504015e-06,
+      "loss": 0.6805,
+      "step": 307
+    },
+    {
+      "epoch": 0.15088793631353337,
+      "grad_norm": 0.2588299810886383,
+      "learning_rate": 4.985301105732908e-06,
+      "loss": 0.6981,
+      "step": 308
+    },
+    {
+      "epoch": 0.15137783221065523,
+      "grad_norm": 0.24666503071784973,
+      "learning_rate": 4.985159570796279e-06,
+      "loss": 0.6898,
+      "step": 309
+    },
+    {
+      "epoch": 0.1518677281077771,
+      "grad_norm": 0.2546054720878601,
+      "learning_rate": 4.985017359732636e-06,
+      "loss": 0.6994,
+      "step": 310
+    },
+    {
+      "epoch": 0.15235762400489897,
+      "grad_norm": 0.24838140606880188,
+      "learning_rate": 4.9848744725806666e-06,
+      "loss": 0.7174,
+      "step": 311
+    },
+    {
+      "epoch": 0.15284751990202083,
+      "grad_norm": 0.2766893804073334,
+      "learning_rate": 4.984730909379248e-06,
+      "loss": 0.7009,
+      "step": 312
+    },
+    {
+      "epoch": 0.15333741579914267,
+      "grad_norm": 0.2599968910217285,
+      "learning_rate": 4.984586670167438e-06,
+      "loss": 0.687,
+      "step": 313
+    },
+    {
+      "epoch": 0.15382731169626454,
+      "grad_norm": 0.2503408193588257,
+      "learning_rate": 4.984441754984479e-06,
+      "loss": 0.732,
+      "step": 314
+    },
+    {
+      "epoch": 0.1543172075933864,
+      "grad_norm": 0.2544567883014679,
+      "learning_rate": 4.984296163869798e-06,
+      "loss": 0.6474,
+      "step": 315
+    },
+    {
+      "epoch": 0.15480710349050827,
+      "grad_norm": 0.25679001212120056,
+      "learning_rate": 4.9841498968630055e-06,
+      "loss": 0.7131,
+      "step": 316
+    },
+    {
+      "epoch": 0.15529699938763014,
+      "grad_norm": 0.2618502080440521,
+      "learning_rate": 4.984002954003895e-06,
+      "loss": 0.7174,
+      "step": 317
+    },
+    {
+      "epoch": 0.15578689528475198,
+      "grad_norm": 0.268387109041214,
+      "learning_rate": 4.9838553353324445e-06,
+      "loss": 0.7087,
+      "step": 318
+    },
+    {
+      "epoch": 0.15627679118187385,
+      "grad_norm": 0.25605982542037964,
+      "learning_rate": 4.9837070408888175e-06,
+      "loss": 0.6813,
+      "step": 319
+    },
+    {
+      "epoch": 0.15676668707899571,
+      "grad_norm": 0.26913389563560486,
+      "learning_rate": 4.983558070713359e-06,
+      "loss": 0.73,
+      "step": 320
+    },
+    {
+      "epoch": 0.15725658297611758,
+      "grad_norm": 0.26553523540496826,
+      "learning_rate": 4.983408424846597e-06,
+      "loss": 0.6842,
+      "step": 321
+    },
+    {
+      "epoch": 0.15774647887323945,
+      "grad_norm": 0.2679034471511841,
+      "learning_rate": 4.983258103329248e-06,
+      "loss": 0.681,
+      "step": 322
+    },
+    {
+      "epoch": 0.1582363747703613,
+      "grad_norm": 0.2610563039779663,
+      "learning_rate": 4.983107106202209e-06,
+      "loss": 0.6812,
+      "step": 323
+    },
+    {
+      "epoch": 0.15872627066748315,
+      "grad_norm": 0.25914549827575684,
+      "learning_rate": 4.982955433506558e-06,
+      "loss": 0.6934,
+      "step": 324
+    },
+    {
+      "epoch": 0.15921616656460502,
+      "grad_norm": 0.25528278946876526,
+      "learning_rate": 4.982803085283563e-06,
+      "loss": 0.7034,
+      "step": 325
+    },
+    {
+      "epoch": 0.1597060624617269,
+      "grad_norm": 0.2526107132434845,
+      "learning_rate": 4.982650061574672e-06,
+      "loss": 0.704,
+      "step": 326
+    },
+    {
+      "epoch": 0.16019595835884876,
+      "grad_norm": 0.25860223174095154,
+      "learning_rate": 4.982496362421517e-06,
+      "loss": 0.6896,
+      "step": 327
+    },
+    {
+      "epoch": 0.1606858542559706,
+      "grad_norm": 0.2585155665874481,
+      "learning_rate": 4.982341987865914e-06,
+      "loss": 0.7077,
+      "step": 328
+    },
+    {
+      "epoch": 0.16117575015309246,
+      "grad_norm": 0.25666436553001404,
+      "learning_rate": 4.982186937949864e-06,
+      "loss": 0.6887,
+      "step": 329
+    },
+    {
+      "epoch": 0.16166564605021433,
+      "grad_norm": 0.248607337474823,
+      "learning_rate": 4.9820312127155515e-06,
+      "loss": 0.7001,
+      "step": 330
+    },
+    {
+      "epoch": 0.1621555419473362,
+      "grad_norm": 0.28019267320632935,
+      "learning_rate": 4.981874812205341e-06,
+      "loss": 0.7299,
+      "step": 331
+    },
+    {
+      "epoch": 0.16264543784445806,
+      "grad_norm": 0.26801958680152893,
+      "learning_rate": 4.981717736461788e-06,
+      "loss": 0.6979,
+      "step": 332
+    },
+    {
+      "epoch": 0.1631353337415799,
+      "grad_norm": 0.2580174505710602,
+      "learning_rate": 4.981559985527624e-06,
+      "loss": 0.6632,
+      "step": 333
+    },
+    {
+      "epoch": 0.16362522963870177,
+      "grad_norm": 0.26502886414527893,
+      "learning_rate": 4.9814015594457686e-06,
+      "loss": 0.6887,
+      "step": 334
+    },
+    {
+      "epoch": 0.16411512553582364,
+      "grad_norm": 0.27120092511177063,
+      "learning_rate": 4.9812424582593246e-06,
+      "loss": 0.7047,
+      "step": 335
+    },
+    {
+      "epoch": 0.1646050214329455,
+      "grad_norm": 0.2717300057411194,
+      "learning_rate": 4.981082682011577e-06,
+      "loss": 0.7083,
+      "step": 336
+    },
+    {
+      "epoch": 0.16509491733006737,
+      "grad_norm": 0.2662920653820038,
+      "learning_rate": 4.9809222307459984e-06,
+      "loss": 0.718,
+      "step": 337
+    },
+    {
+      "epoch": 0.1655848132271892,
+      "grad_norm": 0.25591158866882324,
+      "learning_rate": 4.980761104506238e-06,
+      "loss": 0.6912,
+      "step": 338
+    },
+    {
+      "epoch": 0.16607470912431108,
+      "grad_norm": 0.25804129242897034,
+      "learning_rate": 4.980599303336135e-06,
+      "loss": 0.6812,
+      "step": 339
+    },
+    {
+      "epoch": 0.16656460502143294,
+      "grad_norm": 0.26774024963378906,
+      "learning_rate": 4.980436827279709e-06,
+      "loss": 0.7064,
+      "step": 340
+    },
+    {
+      "epoch": 0.1670545009185548,
+      "grad_norm": 0.26028621196746826,
+      "learning_rate": 4.980273676381165e-06,
+      "loss": 0.685,
+      "step": 341
+    },
+    {
+      "epoch": 0.16754439681567668,
+      "grad_norm": 0.2737005054950714,
+      "learning_rate": 4.980109850684891e-06,
+      "loss": 0.6825,
+      "step": 342
+    },
+    {
+      "epoch": 0.16803429271279854,
+      "grad_norm": 0.26772844791412354,
+      "learning_rate": 4.979945350235459e-06,
+      "loss": 0.6999,
+      "step": 343
+    },
+    {
+      "epoch": 0.16852418860992038,
+      "grad_norm": 0.2749219834804535,
+      "learning_rate": 4.979780175077621e-06,
+      "loss": 0.67,
+      "step": 344
+    },
+    {
+      "epoch": 0.16901408450704225,
+      "grad_norm": 0.2647887170314789,
+      "learning_rate": 4.979614325256318e-06,
+      "loss": 0.7059,
+      "step": 345
+    },
+    {
+      "epoch": 0.16950398040416412,
+      "grad_norm": 0.2660578787326813,
+      "learning_rate": 4.979447800816673e-06,
+      "loss": 0.6964,
+      "step": 346
+    },
+    {
+      "epoch": 0.16999387630128598,
+      "grad_norm": 0.2642233669757843,
+      "learning_rate": 4.979280601803988e-06,
+      "loss": 0.6668,
+      "step": 347
+    },
+    {
+      "epoch": 0.17048377219840785,
+      "grad_norm": 0.26047587394714355,
+      "learning_rate": 4.979112728263755e-06,
+      "loss": 0.7015,
+      "step": 348
+    },
+    {
+      "epoch": 0.1709736680955297,
+      "grad_norm": 0.26166731119155884,
+      "learning_rate": 4.9789441802416454e-06,
+      "loss": 0.6964,
+      "step": 349
+    },
+    {
+      "epoch": 0.17146356399265156,
+      "grad_norm": 0.26318004727363586,
+      "learning_rate": 4.978774957783517e-06,
+      "loss": 0.7015,
+      "step": 350
+    },
+    {
+      "epoch": 0.17195345988977342,
+      "grad_norm": 0.2654951512813568,
+      "learning_rate": 4.978605060935407e-06,
+      "loss": 0.6653,
+      "step": 351
+    },
+    {
+      "epoch": 0.1724433557868953,
+      "grad_norm": 0.2767031490802765,
+      "learning_rate": 4.9784344897435405e-06,
+      "loss": 0.6837,
+      "step": 352
+    },
+    {
+      "epoch": 0.17293325168401716,
+      "grad_norm": 0.2719745635986328,
+      "learning_rate": 4.978263244254324e-06,
+      "loss": 0.6784,
+      "step": 353
+    },
+    {
+      "epoch": 0.173423147581139,
+      "grad_norm": 0.2508717179298401,
+      "learning_rate": 4.978091324514346e-06,
+      "loss": 0.6805,
+      "step": 354
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 0.26405903697013855,
+      "learning_rate": 4.977918730570381e-06,
+      "loss": 0.6521,
+      "step": 355
+    },
+    {
+      "epoch": 0.17440293937538273,
+      "grad_norm": 0.26240110397338867,
+      "learning_rate": 4.977745462469386e-06,
+      "loss": 0.675,
+      "step": 356
+    },
+    {
+      "epoch": 0.1748928352725046,
+      "grad_norm": 0.2654706835746765,
+      "learning_rate": 4.9775715202585005e-06,
+      "loss": 0.6967,
+      "step": 357
+    },
+    {
+      "epoch": 0.17538273116962647,
+      "grad_norm": 0.26780185103416443,
+      "learning_rate": 4.97739690398505e-06,
+      "loss": 0.6808,
+      "step": 358
+    },
+    {
+      "epoch": 0.1758726270667483,
+      "grad_norm": 0.27980104088783264,
+      "learning_rate": 4.9772216136965405e-06,
+      "loss": 0.6901,
+      "step": 359
+    },
+    {
+      "epoch": 0.17636252296387017,
+      "grad_norm": 0.2825722396373749,
+      "learning_rate": 4.977045649440662e-06,
+      "loss": 0.7147,
+      "step": 360
+    },
+    {
+      "epoch": 0.17685241886099204,
+      "grad_norm": 0.28033769130706787,
+      "learning_rate": 4.97686901126529e-06,
+      "loss": 0.69,
+      "step": 361
+    },
+    {
+      "epoch": 0.1773423147581139,
+      "grad_norm": 0.2616003751754761,
+      "learning_rate": 4.976691699218479e-06,
+      "loss": 0.679,
+      "step": 362
+    },
+    {
+      "epoch": 0.17783221065523577,
+      "grad_norm": 0.26597169041633606,
+      "learning_rate": 4.976513713348472e-06,
+      "loss": 0.6817,
+      "step": 363
+    },
+    {
+      "epoch": 0.1783221065523576,
+      "grad_norm": 0.2727013826370239,
+      "learning_rate": 4.976335053703692e-06,
+      "loss": 0.6878,
+      "step": 364
+    },
+    {
+      "epoch": 0.17881200244947948,
+      "grad_norm": 0.2789647579193115,
+      "learning_rate": 4.976155720332747e-06,
+      "loss": 0.6942,
+      "step": 365
+    },
+    {
+      "epoch": 0.17930189834660135,
+      "grad_norm": 0.26109856367111206,
+      "learning_rate": 4.975975713284426e-06,
+      "loss": 0.6992,
+      "step": 366
+    },
+    {
+      "epoch": 0.1797917942437232,
+      "grad_norm": 0.2630729675292969,
+      "learning_rate": 4.975795032607703e-06,
+      "loss": 0.7101,
+      "step": 367
+    },
+    {
+      "epoch": 0.18028169014084508,
+      "grad_norm": 0.28116390109062195,
+      "learning_rate": 4.9756136783517365e-06,
+      "loss": 0.6843,
+      "step": 368
+    },
+    {
+      "epoch": 0.18077158603796692,
+      "grad_norm": 0.26151201128959656,
+      "learning_rate": 4.975431650565865e-06,
+      "loss": 0.6973,
+      "step": 369
+    },
+    {
+      "epoch": 0.1812614819350888,
+      "grad_norm": 0.2683990001678467,
+      "learning_rate": 4.975248949299613e-06,
+      "loss": 0.7055,
+      "step": 370
+    },
+    {
+      "epoch": 0.18175137783221065,
+      "grad_norm": 0.26030197739601135,
+      "learning_rate": 4.9750655746026875e-06,
+      "loss": 0.7008,
+      "step": 371
+    },
+    {
+      "epoch": 0.18224127372933252,
+      "grad_norm": 0.26876166462898254,
+      "learning_rate": 4.974881526524978e-06,
+      "loss": 0.694,
+      "step": 372
+    },
+    {
+      "epoch": 0.1827311696264544,
+      "grad_norm": 0.2684592604637146,
+      "learning_rate": 4.974696805116558e-06,
+      "loss": 0.7156,
+      "step": 373
+    },
+    {
+      "epoch": 0.18322106552357623,
+      "grad_norm": 0.27490681409835815,
+      "learning_rate": 4.974511410427683e-06,
+      "loss": 0.6843,
+      "step": 374
+    },
+    {
+      "epoch": 0.1837109614206981,
+      "grad_norm": 0.2698623239994049,
+      "learning_rate": 4.974325342508793e-06,
+      "loss": 0.6917,
+      "step": 375
+    },
+    {
+      "epoch": 0.18420085731781996,
+      "grad_norm": 0.27144962549209595,
+      "learning_rate": 4.974138601410513e-06,
+      "loss": 0.7152,
+      "step": 376
+    },
+    {
+      "epoch": 0.18469075321494183,
+      "grad_norm": 0.26694294810295105,
+      "learning_rate": 4.9739511871836454e-06,
+      "loss": 0.6715,
+      "step": 377
+    },
+    {
+      "epoch": 0.1851806491120637,
+      "grad_norm": 0.2714312672615051,
+      "learning_rate": 4.973763099879181e-06,
+      "loss": 0.6997,
+      "step": 378
+    },
+    {
+      "epoch": 0.18567054500918556,
+      "grad_norm": 0.27155402302742004,
+      "learning_rate": 4.973574339548291e-06,
+      "loss": 0.6609,
+      "step": 379
+    },
+    {
+      "epoch": 0.1861604409063074,
+      "grad_norm": 0.2687325179576874,
+      "learning_rate": 4.973384906242332e-06,
+      "loss": 0.7095,
+      "step": 380
+    },
+    {
+      "epoch": 0.18665033680342927,
+      "grad_norm": 0.2704453766345978,
+      "learning_rate": 4.973194800012841e-06,
+      "loss": 0.6728,
+      "step": 381
+    },
+    {
+      "epoch": 0.18714023270055113,
+      "grad_norm": 0.2697974145412445,
+      "learning_rate": 4.973004020911541e-06,
+      "loss": 0.6817,
+      "step": 382
+    },
+    {
+      "epoch": 0.187630128597673,
+      "grad_norm": 0.27038195729255676,
+      "learning_rate": 4.972812568990334e-06,
+      "loss": 0.69,
+      "step": 383
+    },
+    {
+      "epoch": 0.18812002449479487,
+      "grad_norm": 0.27607688307762146,
+      "learning_rate": 4.97262044430131e-06,
+      "loss": 0.6935,
+      "step": 384
+    },
+    {
+      "epoch": 0.1886099203919167,
+      "grad_norm": 0.261408269405365,
+      "learning_rate": 4.972427646896738e-06,
+      "loss": 0.6641,
+      "step": 385
+    },
+    {
+      "epoch": 0.18909981628903857,
+      "grad_norm": 0.2864983379840851,
+      "learning_rate": 4.972234176829072e-06,
+      "loss": 0.6989,
+      "step": 386
+    },
+    {
+      "epoch": 0.18958971218616044,
+      "grad_norm": 0.2765742838382721,
+      "learning_rate": 4.972040034150949e-06,
+      "loss": 0.6516,
+      "step": 387
+    },
+    {
+      "epoch": 0.1900796080832823,
+      "grad_norm": 0.27313947677612305,
+      "learning_rate": 4.9718452189151885e-06,
+      "loss": 0.722,
+      "step": 388
+    },
+    {
+      "epoch": 0.19056950398040418,
+      "grad_norm": 0.27086713910102844,
+      "learning_rate": 4.971649731174793e-06,
+      "loss": 0.6706,
+      "step": 389
+    },
+    {
+      "epoch": 0.19105939987752601,
+      "grad_norm": 0.2872767150402069,
+      "learning_rate": 4.9714535709829475e-06,
+      "loss": 0.7021,
+      "step": 390
+    },
+    {
+      "epoch": 0.19154929577464788,
+      "grad_norm": 0.2844785749912262,
+      "learning_rate": 4.971256738393021e-06,
+      "loss": 0.6867,
+      "step": 391
+    },
+    {
+      "epoch": 0.19203919167176975,
+      "grad_norm": 0.2807650864124298,
+      "learning_rate": 4.971059233458565e-06,
+      "loss": 0.6744,
+      "step": 392
+    },
+    {
+      "epoch": 0.19252908756889162,
+      "grad_norm": 0.26693230867385864,
+      "learning_rate": 4.970861056233314e-06,
+      "loss": 0.7266,
+      "step": 393
+    },
+    {
+      "epoch": 0.19301898346601348,
+      "grad_norm": 0.27306389808654785,
+      "learning_rate": 4.970662206771184e-06,
+      "loss": 0.6958,
+      "step": 394
+    },
+    {
+      "epoch": 0.19350887936313532,
+      "grad_norm": 0.2836274802684784,
+      "learning_rate": 4.9704626851262775e-06,
+      "loss": 0.699,
+      "step": 395
+    },
+    {
+      "epoch": 0.1939987752602572,
+      "grad_norm": 0.27907463908195496,
+      "learning_rate": 4.970262491352875e-06,
+      "loss": 0.6961,
+      "step": 396
+    },
+    {
+      "epoch": 0.19448867115737906,
+      "grad_norm": 0.2898431420326233,
+      "learning_rate": 4.970061625505443e-06,
+      "loss": 0.6883,
+      "step": 397
+    },
+    {
+      "epoch": 0.19497856705450092,
+      "grad_norm": 0.2687452435493469,
+      "learning_rate": 4.969860087638632e-06,
+      "loss": 0.6499,
+      "step": 398
+    },
+    {
+      "epoch": 0.1954684629516228,
+      "grad_norm": 0.2855985760688782,
+      "learning_rate": 4.969657877807271e-06,
+      "loss": 0.6937,
+      "step": 399
+    },
+    {
+      "epoch": 0.19595835884874463,
+      "grad_norm": 0.27004674077033997,
+      "learning_rate": 4.969454996066377e-06,
+      "loss": 0.6813,
+      "step": 400
+    },
+    {
+      "epoch": 0.1964482547458665,
+      "grad_norm": 0.2813244163990021,
+      "learning_rate": 4.969251442471145e-06,
+      "loss": 0.6907,
+      "step": 401
+    },
+    {
+      "epoch": 0.19693815064298836,
+      "grad_norm": 0.2692974805831909,
+      "learning_rate": 4.969047217076956e-06,
+      "loss": 0.668,
+      "step": 402
+    },
+    {
+      "epoch": 0.19742804654011023,
+      "grad_norm": 0.2798345983028412,
+      "learning_rate": 4.968842319939374e-06,
+      "loss": 0.6637,
+      "step": 403
+    },
+    {
+      "epoch": 0.1979179424372321,
+      "grad_norm": 0.2649277448654175,
+      "learning_rate": 4.968636751114141e-06,
+      "loss": 0.7005,
+      "step": 404
+    },
+    {
+      "epoch": 0.19840783833435394,
+      "grad_norm": 0.270865261554718,
+      "learning_rate": 4.968430510657189e-06,
+      "loss": 0.6985,
+      "step": 405
+    },
+    {
+      "epoch": 0.1988977342314758,
+      "grad_norm": 0.2828701138496399,
+      "learning_rate": 4.968223598624625e-06,
+      "loss": 0.6675,
+      "step": 406
+    },
+    {
+      "epoch": 0.19938763012859767,
+      "grad_norm": 0.28420430421829224,
+      "learning_rate": 4.968016015072747e-06,
+      "loss": 0.7123,
+      "step": 407
+    },
+    {
+      "epoch": 0.19987752602571954,
+      "grad_norm": 0.2720252573490143,
+      "learning_rate": 4.967807760058029e-06,
+      "loss": 0.6612,
+      "step": 408
+    },
+    {
+      "epoch": 0.2003674219228414,
+      "grad_norm": 0.276832640171051,
+      "learning_rate": 4.9675988336371305e-06,
+      "loss": 0.6776,
+      "step": 409
+    },
+    {
+      "epoch": 0.20085731781996327,
+      "grad_norm": 0.2641506791114807,
+      "learning_rate": 4.967389235866893e-06,
+      "loss": 0.6782,
+      "step": 410
+    },
+    {
+      "epoch": 0.2013472137170851,
+      "grad_norm": 0.2748897969722748,
+      "learning_rate": 4.967178966804341e-06,
+      "loss": 0.6945,
+      "step": 411
+    },
+    {
+      "epoch": 0.20183710961420698,
+      "grad_norm": 0.27730754017829895,
+      "learning_rate": 4.966968026506683e-06,
+      "loss": 0.6681,
+      "step": 412
+    },
+    {
+      "epoch": 0.20232700551132884,
+      "grad_norm": 0.2730098366737366,
+      "learning_rate": 4.966756415031307e-06,
+      "loss": 0.7097,
+      "step": 413
+    },
+    {
+      "epoch": 0.2028169014084507,
+      "grad_norm": 0.26940208673477173,
+      "learning_rate": 4.9665441324357834e-06,
+      "loss": 0.6963,
+      "step": 414
+    },
+    {
+      "epoch": 0.20330679730557258,
+      "grad_norm": 0.28830280900001526,
+      "learning_rate": 4.9663311787778715e-06,
+      "loss": 0.6455,
+      "step": 415
+    },
+    {
+      "epoch": 0.20379669320269442,
+      "grad_norm": 0.2707945704460144,
+      "learning_rate": 4.966117554115505e-06,
+      "loss": 0.6781,
+      "step": 416
+    },
+    {
+      "epoch": 0.20428658909981628,
+      "grad_norm": 0.28427818417549133,
+      "learning_rate": 4.965903258506806e-06,
+      "loss": 0.654,
+      "step": 417
+    },
+    {
+      "epoch": 0.20477648499693815,
+      "grad_norm": 0.26649975776672363,
+      "learning_rate": 4.965688292010077e-06,
+      "loss": 0.7078,
+      "step": 418
+    },
+    {
+      "epoch": 0.20526638089406002,
+      "grad_norm": 0.27190712094306946,
+      "learning_rate": 4.9654726546838015e-06,
+      "loss": 0.6648,
+      "step": 419
+    },
+    {
+      "epoch": 0.20575627679118189,
+      "grad_norm": 0.27658402919769287,
+      "learning_rate": 4.965256346586648e-06,
+      "loss": 0.7066,
+      "step": 420
+    },
+    {
+      "epoch": 0.20624617268830372,
+      "grad_norm": 0.2791120111942291,
+      "learning_rate": 4.965039367777466e-06,
+      "loss": 0.6856,
+      "step": 421
+    },
+    {
+      "epoch": 0.2067360685854256,
+      "grad_norm": 0.27996936440467834,
+      "learning_rate": 4.964821718315289e-06,
+      "loss": 0.6779,
+      "step": 422
+    },
+    {
+      "epoch": 0.20722596448254746,
+      "grad_norm": 0.27422329783439636,
+      "learning_rate": 4.9646033982593315e-06,
+      "loss": 0.6641,
+      "step": 423
+    },
+    {
+      "epoch": 0.20771586037966933,
+      "grad_norm": 0.28245022892951965,
+      "learning_rate": 4.964384407668991e-06,
+      "loss": 0.6649,
+      "step": 424
+    },
+    {
+      "epoch": 0.2082057562767912,
+      "grad_norm": 0.2692187428474426,
+      "learning_rate": 4.964164746603847e-06,
+      "loss": 0.6685,
+      "step": 425
+    },
+    {
+      "epoch": 0.20869565217391303,
+      "grad_norm": 0.27592557668685913,
+      "learning_rate": 4.963944415123662e-06,
+      "loss": 0.6436,
+      "step": 426
+    },
+    {
+      "epoch": 0.2091855480710349,
+      "grad_norm": 0.2831850051879883,
+      "learning_rate": 4.9637234132883805e-06,
+      "loss": 0.6738,
+      "step": 427
+    },
+    {
+      "epoch": 0.20967544396815677,
+      "grad_norm": 0.30106180906295776,
+      "learning_rate": 4.9635017411581295e-06,
+      "loss": 0.7092,
+      "step": 428
+    },
+    {
+      "epoch": 0.21016533986527863,
+      "grad_norm": 0.2780488431453705,
+      "learning_rate": 4.96327939879322e-06,
+      "loss": 0.6872,
+      "step": 429
+    },
+    {
+      "epoch": 0.2106552357624005,
+      "grad_norm": 0.2832487225532532,
+      "learning_rate": 4.96305638625414e-06,
+      "loss": 0.6791,
+      "step": 430
+    },
+    {
+      "epoch": 0.21114513165952234,
+      "grad_norm": 0.27021270990371704,
+      "learning_rate": 4.962832703601568e-06,
+      "loss": 0.7024,
+      "step": 431
+    },
+    {
+      "epoch": 0.2116350275566442,
+      "grad_norm": 0.2906636595726013,
+      "learning_rate": 4.962608350896357e-06,
+      "loss": 0.6916,
+      "step": 432
+    },
+    {
+      "epoch": 0.21212492345376607,
+      "grad_norm": 0.27886664867401123,
+      "learning_rate": 4.962383328199549e-06,
+      "loss": 0.6882,
+      "step": 433
+    },
+    {
+      "epoch": 0.21261481935088794,
+      "grad_norm": 0.2870326638221741,
+      "learning_rate": 4.962157635572362e-06,
+      "loss": 0.6775,
+      "step": 434
+    },
+    {
+      "epoch": 0.2131047152480098,
+      "grad_norm": 0.2815456688404083,
+      "learning_rate": 4.9619312730762e-06,
+      "loss": 0.7052,
+      "step": 435
+    },
+    {
+      "epoch": 0.21359461114513165,
+      "grad_norm": 0.2805708646774292,
+      "learning_rate": 4.961704240772649e-06,
+      "loss": 0.6697,
+      "step": 436
+    },
+    {
+      "epoch": 0.2140845070422535,
+      "grad_norm": 0.29036977887153625,
+      "learning_rate": 4.961476538723477e-06,
+      "loss": 0.6859,
+      "step": 437
+    },
+    {
+      "epoch": 0.21457440293937538,
+      "grad_norm": 0.28320616483688354,
+      "learning_rate": 4.9612481669906335e-06,
+      "loss": 0.6502,
+      "step": 438
+    },
+    {
+      "epoch": 0.21506429883649725,
+      "grad_norm": 0.2814081013202667,
+      "learning_rate": 4.961019125636251e-06,
+      "loss": 0.6802,
+      "step": 439
+    },
+    {
+      "epoch": 0.21555419473361911,
+      "grad_norm": 0.2973114252090454,
+      "learning_rate": 4.960789414722642e-06,
+      "loss": 0.6808,
+      "step": 440
+    },
+    {
+      "epoch": 0.21604409063074098,
+      "grad_norm": 0.28980064392089844,
+      "learning_rate": 4.960559034312306e-06,
+      "loss": 0.6619,
+      "step": 441
+    },
+    {
+      "epoch": 0.21653398652786282,
+      "grad_norm": 0.28697559237480164,
+      "learning_rate": 4.960327984467919e-06,
+      "loss": 0.6548,
+      "step": 442
+    },
+    {
+      "epoch": 0.2170238824249847,
+      "grad_norm": 0.2818733751773834,
+      "learning_rate": 4.9600962652523435e-06,
+      "loss": 0.6608,
+      "step": 443
+    },
+    {
+      "epoch": 0.21751377832210655,
+      "grad_norm": 0.28344979882240295,
+      "learning_rate": 4.959863876728622e-06,
+      "loss": 0.6818,
+      "step": 444
+    },
+    {
+      "epoch": 0.21800367421922842,
+      "grad_norm": 0.277803510427475,
+      "learning_rate": 4.959630818959979e-06,
+      "loss": 0.6632,
+      "step": 445
+    },
+    {
+      "epoch": 0.2184935701163503,
+      "grad_norm": 0.290330708026886,
+      "learning_rate": 4.9593970920098226e-06,
+      "loss": 0.7021,
+      "step": 446
+    },
+    {
+      "epoch": 0.21898346601347213,
+      "grad_norm": 0.2983878254890442,
+      "learning_rate": 4.9591626959417395e-06,
+      "loss": 0.645,
+      "step": 447
+    },
+    {
+      "epoch": 0.219473361910594,
+      "grad_norm": 0.3008386492729187,
+      "learning_rate": 4.958927630819503e-06,
+      "loss": 0.6722,
+      "step": 448
+    },
+    {
+      "epoch": 0.21996325780771586,
+      "grad_norm": 0.29174381494522095,
+      "learning_rate": 4.9586918967070666e-06,
+      "loss": 0.7,
+      "step": 449
+    },
+    {
+      "epoch": 0.22045315370483773,
+      "grad_norm": 0.2773303985595703,
+      "learning_rate": 4.958455493668565e-06,
+      "loss": 0.6606,
+      "step": 450
+    },
+    {
+      "epoch": 0.2209430496019596,
+      "grad_norm": 0.29781636595726013,
+      "learning_rate": 4.958218421768314e-06,
+      "loss": 0.673,
+      "step": 451
+    },
+    {
+      "epoch": 0.22143294549908143,
+      "grad_norm": 0.2792598605155945,
+      "learning_rate": 4.957980681070814e-06,
+      "loss": 0.7003,
+      "step": 452
+    },
+    {
+      "epoch": 0.2219228413962033,
+      "grad_norm": 0.2928924262523651,
+      "learning_rate": 4.957742271640746e-06,
+      "loss": 0.6825,
+      "step": 453
+    },
+    {
+      "epoch": 0.22241273729332517,
+      "grad_norm": 0.2799023985862732,
+      "learning_rate": 4.957503193542973e-06,
+      "loss": 0.6808,
+      "step": 454
+    },
+    {
+      "epoch": 0.22290263319044704,
+      "grad_norm": 0.2886837124824524,
+      "learning_rate": 4.957263446842541e-06,
+      "loss": 0.6846,
+      "step": 455
+    },
+    {
+      "epoch": 0.2233925290875689,
+      "grad_norm": 0.29160356521606445,
+      "learning_rate": 4.957023031604674e-06,
+      "loss": 0.6755,
+      "step": 456
+    },
+    {
+      "epoch": 0.22388242498469074,
+      "grad_norm": 0.29495176672935486,
+      "learning_rate": 4.956781947894784e-06,
+      "loss": 0.6744,
+      "step": 457
+    },
+    {
+      "epoch": 0.2243723208818126,
+      "grad_norm": 0.28703880310058594,
+      "learning_rate": 4.95654019577846e-06,
+      "loss": 0.7161,
+      "step": 458
+    },
+    {
+      "epoch": 0.22486221677893448,
+      "grad_norm": 0.29158130288124084,
+      "learning_rate": 4.956297775321475e-06,
+      "loss": 0.6943,
+      "step": 459
+    },
+    {
+      "epoch": 0.22535211267605634,
+      "grad_norm": 0.2803298234939575,
+      "learning_rate": 4.956054686589783e-06,
+      "loss": 0.6823,
+      "step": 460
+    },
+    {
+      "epoch": 0.2258420085731782,
+      "grad_norm": 0.2935578525066376,
+      "learning_rate": 4.95581092964952e-06,
+      "loss": 0.6856,
+      "step": 461
+    },
+    {
+      "epoch": 0.22633190447030005,
+      "grad_norm": 0.28209394216537476,
+      "learning_rate": 4.955566504567004e-06,
+      "loss": 0.693,
+      "step": 462
+    },
+    {
+      "epoch": 0.22682180036742192,
+      "grad_norm": 0.2914902865886688,
+      "learning_rate": 4.955321411408735e-06,
+      "loss": 0.6978,
+      "step": 463
+    },
+    {
+      "epoch": 0.22731169626454378,
+      "grad_norm": 0.2760421335697174,
+      "learning_rate": 4.955075650241396e-06,
+      "loss": 0.6912,
+      "step": 464
+    },
+    {
+      "epoch": 0.22780159216166565,
+      "grad_norm": 0.2878439724445343,
+      "learning_rate": 4.954829221131847e-06,
+      "loss": 0.6973,
+      "step": 465
+    },
+    {
+      "epoch": 0.22829148805878752,
+      "grad_norm": 0.2849530279636383,
+      "learning_rate": 4.954582124147135e-06,
+      "loss": 0.6682,
+      "step": 466
+    },
+    {
+      "epoch": 0.22878138395590936,
+      "grad_norm": 0.2800678610801697,
+      "learning_rate": 4.954334359354486e-06,
+      "loss": 0.6524,
+      "step": 467
+    },
+    {
+      "epoch": 0.22927127985303122,
+      "grad_norm": 0.2819807231426239,
+      "learning_rate": 4.954085926821308e-06,
+      "loss": 0.6899,
+      "step": 468
+    },
+    {
+      "epoch": 0.2297611757501531,
+      "grad_norm": 0.290926456451416,
+      "learning_rate": 4.953836826615193e-06,
+      "loss": 0.6909,
+      "step": 469
+    },
+    {
+      "epoch": 0.23025107164727496,
+      "grad_norm": 0.2750617265701294,
+      "learning_rate": 4.953587058803909e-06,
+      "loss": 0.666,
+      "step": 470
+    },
+    {
+      "epoch": 0.23074096754439682,
+      "grad_norm": 0.2927756607532501,
+      "learning_rate": 4.953336623455413e-06,
+      "loss": 0.6676,
+      "step": 471
+    },
+    {
+      "epoch": 0.23123086344151866,
+      "grad_norm": 0.2824176549911499,
+      "learning_rate": 4.953085520637839e-06,
+      "loss": 0.6479,
+      "step": 472
+    },
+    {
+      "epoch": 0.23172075933864053,
+      "grad_norm": 0.2902854084968567,
+      "learning_rate": 4.952833750419502e-06,
+      "loss": 0.6675,
+      "step": 473
+    },
+    {
+      "epoch": 0.2322106552357624,
+      "grad_norm": 0.2878444790840149,
+      "learning_rate": 4.952581312868901e-06,
+      "loss": 0.6707,
+      "step": 474
+    },
+    {
+      "epoch": 0.23270055113288426,
+      "grad_norm": 0.29057806730270386,
+      "learning_rate": 4.952328208054716e-06,
+      "loss": 0.6931,
+      "step": 475
+    },
+    {
+      "epoch": 0.23319044703000613,
+      "grad_norm": 0.287222683429718,
+      "learning_rate": 4.952074436045806e-06,
+      "loss": 0.6675,
+      "step": 476
+    },
+    {
+      "epoch": 0.233680342927128,
+      "grad_norm": 0.2918603718280792,
+      "learning_rate": 4.951819996911217e-06,
+      "loss": 0.6752,
+      "step": 477
+    },
+    {
+      "epoch": 0.23417023882424984,
+      "grad_norm": 0.2872229814529419,
+      "learning_rate": 4.951564890720172e-06,
+      "loss": 0.6731,
+      "step": 478
+    },
+    {
+      "epoch": 0.2346601347213717,
+      "grad_norm": 0.28795695304870605,
+      "learning_rate": 4.951309117542075e-06,
+      "loss": 0.6798,
+      "step": 479
+    },
+    {
+      "epoch": 0.23515003061849357,
+      "grad_norm": 0.27824899554252625,
+      "learning_rate": 4.951052677446515e-06,
+      "loss": 0.705,
+      "step": 480
+    },
+    {
+      "epoch": 0.23563992651561544,
+      "grad_norm": 0.2940081059932709,
+      "learning_rate": 4.95079557050326e-06,
+      "loss": 0.6721,
+      "step": 481
+    },
+    {
+      "epoch": 0.2361298224127373,
+      "grad_norm": 0.31432050466537476,
+      "learning_rate": 4.950537796782261e-06,
+      "loss": 0.71,
+      "step": 482
+    },
+    {
+      "epoch": 0.23661971830985915,
+      "grad_norm": 0.28667113184928894,
+      "learning_rate": 4.950279356353648e-06,
+      "loss": 0.6781,
+      "step": 483
+    },
+    {
+      "epoch": 0.237109614206981,
+      "grad_norm": 0.28186243772506714,
+      "learning_rate": 4.950020249287734e-06,
+      "loss": 0.689,
+      "step": 484
+    },
+    {
+      "epoch": 0.23759951010410288,
+      "grad_norm": 0.282751202583313,
+      "learning_rate": 4.9497604756550134e-06,
+      "loss": 0.6943,
+      "step": 485
+    },
+    {
+      "epoch": 0.23808940600122475,
+      "grad_norm": 0.29054197669029236,
+      "learning_rate": 4.9495000355261625e-06,
+      "loss": 0.689,
+      "step": 486
+    },
+    {
+      "epoch": 0.2385793018983466,
+      "grad_norm": 0.28855305910110474,
+      "learning_rate": 4.949238928972037e-06,
+      "loss": 0.6566,
+      "step": 487
+    },
+    {
+      "epoch": 0.23906919779546845,
+      "grad_norm": 0.2937643229961395,
+      "learning_rate": 4.948977156063675e-06,
+      "loss": 0.6669,
+      "step": 488
+    },
+    {
+      "epoch": 0.23955909369259032,
+      "grad_norm": 0.27907466888427734,
+      "learning_rate": 4.948714716872297e-06,
+      "loss": 0.6802,
+      "step": 489
+    },
+    {
+      "epoch": 0.2400489895897122,
+      "grad_norm": 0.29693931341171265,
+      "learning_rate": 4.948451611469304e-06,
+      "loss": 0.6874,
+      "step": 490
+    },
+    {
+      "epoch": 0.24053888548683405,
+      "grad_norm": 0.2904781401157379,
+      "learning_rate": 4.948187839926276e-06,
+      "loss": 0.6912,
+      "step": 491
+    },
+    {
+      "epoch": 0.24102878138395592,
+      "grad_norm": 0.28125911951065063,
+      "learning_rate": 4.947923402314978e-06,
+      "loss": 0.6663,
+      "step": 492
+    },
+    {
+      "epoch": 0.24151867728107776,
+      "grad_norm": 0.3010816276073456,
+      "learning_rate": 4.947658298707354e-06,
+      "loss": 0.6784,
+      "step": 493
+    },
+    {
+      "epoch": 0.24200857317819963,
+      "grad_norm": 0.29629260301589966,
+      "learning_rate": 4.9473925291755295e-06,
+      "loss": 0.7006,
+      "step": 494
+    },
+    {
+      "epoch": 0.2424984690753215,
+      "grad_norm": 0.29816797375679016,
+      "learning_rate": 4.947126093791812e-06,
+      "loss": 0.6519,
+      "step": 495
+    },
+    {
+      "epoch": 0.24298836497244336,
+      "grad_norm": 0.29432323575019836,
+      "learning_rate": 4.946858992628688e-06,
+      "loss": 0.7049,
+      "step": 496
+    },
+    {
+      "epoch": 0.24347826086956523,
+      "grad_norm": 0.2930796444416046,
+      "learning_rate": 4.946591225758828e-06,
+      "loss": 0.661,
+      "step": 497
+    },
+    {
+      "epoch": 0.24396815676668707,
+      "grad_norm": 0.2817651629447937,
+      "learning_rate": 4.946322793255081e-06,
+      "loss": 0.6622,
+      "step": 498
+    },
+    {
+      "epoch": 0.24445805266380893,
+      "grad_norm": 0.28571751713752747,
+      "learning_rate": 4.946053695190479e-06,
+      "loss": 0.6726,
+      "step": 499
+    },
+    {
+      "epoch": 0.2449479485609308,
+      "grad_norm": 0.28773561120033264,
+      "learning_rate": 4.945783931638235e-06,
+      "loss": 0.6766,
+      "step": 500
+    },
+    {
+      "epoch": 0.24543784445805267,
+      "grad_norm": 0.28509920835494995,
+      "learning_rate": 4.945513502671741e-06,
+      "loss": 0.6548,
+      "step": 501
+    },
+    {
+      "epoch": 0.24592774035517453,
+      "grad_norm": 0.28688615560531616,
+      "learning_rate": 4.9452424083645725e-06,
+      "loss": 0.6814,
+      "step": 502
+    },
+    {
+      "epoch": 0.24641763625229637,
+      "grad_norm": 0.3019798994064331,
+      "learning_rate": 4.9449706487904845e-06,
+      "loss": 0.6688,
+      "step": 503
+    },
+    {
+      "epoch": 0.24690753214941824,
+      "grad_norm": 0.2820969521999359,
+      "learning_rate": 4.944698224023414e-06,
+      "loss": 0.6696,
+      "step": 504
+    },
+    {
+      "epoch": 0.2473974280465401,
+      "grad_norm": 0.2863926887512207,
+      "learning_rate": 4.944425134137478e-06,
+      "loss": 0.6699,
+      "step": 505
+    },
+    {
+      "epoch": 0.24788732394366197,
+      "grad_norm": 0.28923314809799194,
+      "learning_rate": 4.944151379206976e-06,
+      "loss": 0.6594,
+      "step": 506
+    },
+    {
+      "epoch": 0.24837721984078384,
+      "grad_norm": 0.29417911171913147,
+      "learning_rate": 4.943876959306386e-06,
+      "loss": 0.6771,
+      "step": 507
+    },
+    {
+      "epoch": 0.2488671157379057,
+      "grad_norm": 0.3137289881706238,
+      "learning_rate": 4.943601874510369e-06,
+      "loss": 0.6995,
+      "step": 508
+    },
+    {
+      "epoch": 0.24935701163502755,
+      "grad_norm": 0.2851713001728058,
+      "learning_rate": 4.9433261248937655e-06,
+      "loss": 0.6397,
+      "step": 509
+    },
+    {
+      "epoch": 0.24984690753214941,
+      "grad_norm": 0.3067043125629425,
+      "learning_rate": 4.943049710531599e-06,
+      "loss": 0.6696,
+      "step": 510
+    },
+    {
+      "epoch": 0.2503368034292713,
+      "grad_norm": 0.28277942538261414,
+      "learning_rate": 4.942772631499071e-06,
+      "loss": 0.6852,
+      "step": 511
+    },
+    {
+      "epoch": 0.2508266993263931,
+      "grad_norm": 0.2837325930595398,
+      "learning_rate": 4.942494887871565e-06,
+      "loss": 0.7028,
+      "step": 512
+    },
+    {
+      "epoch": 0.251316595223515,
+      "grad_norm": 0.28493350744247437,
+      "learning_rate": 4.942216479724647e-06,
+      "loss": 0.6616,
+      "step": 513
+    },
+    {
+      "epoch": 0.25180649112063686,
+      "grad_norm": 0.28880465030670166,
+      "learning_rate": 4.941937407134061e-06,
+      "loss": 0.6609,
+      "step": 514
+    },
+    {
+      "epoch": 0.25229638701775875,
+      "grad_norm": 0.29091018438339233,
+      "learning_rate": 4.941657670175734e-06,
+      "loss": 0.6714,
+      "step": 515
+    },
+    {
+      "epoch": 0.2527862829148806,
+      "grad_norm": 0.32527679204940796,
+      "learning_rate": 4.941377268925773e-06,
+      "loss": 0.6783,
+      "step": 516
+    },
+    {
+      "epoch": 0.25327617881200243,
+      "grad_norm": 0.31313812732696533,
+      "learning_rate": 4.941096203460465e-06,
+      "loss": 0.7043,
+      "step": 517
+    },
+    {
+      "epoch": 0.2537660747091243,
+      "grad_norm": 0.29403164982795715,
+      "learning_rate": 4.940814473856278e-06,
+      "loss": 0.678,
+      "step": 518
+    },
+    {
+      "epoch": 0.25425597060624616,
+      "grad_norm": 0.2897968590259552,
+      "learning_rate": 4.9405320801898615e-06,
+      "loss": 0.6726,
+      "step": 519
+    },
+    {
+      "epoch": 0.25474586650336806,
+      "grad_norm": 0.3000484108924866,
+      "learning_rate": 4.940249022538045e-06,
+      "loss": 0.6736,
+      "step": 520
+    },
+    {
+      "epoch": 0.2552357624004899,
+      "grad_norm": 0.3161175847053528,
+      "learning_rate": 4.9399653009778394e-06,
+      "loss": 0.6713,
+      "step": 521
+    },
+    {
+      "epoch": 0.25572565829761174,
+      "grad_norm": 0.30063289403915405,
+      "learning_rate": 4.9396809155864354e-06,
+      "loss": 0.6839,
+      "step": 522
+    },
+    {
+      "epoch": 0.25621555419473363,
+      "grad_norm": 0.3286581337451935,
+      "learning_rate": 4.939395866441204e-06,
+      "loss": 0.6747,
+      "step": 523
+    },
+    {
+      "epoch": 0.25670545009185547,
+      "grad_norm": 0.3079366087913513,
+      "learning_rate": 4.9391101536196975e-06,
+      "loss": 0.691,
+      "step": 524
+    },
+    {
+      "epoch": 0.25719534598897736,
+      "grad_norm": 0.30219361186027527,
+      "learning_rate": 4.938823777199649e-06,
+      "loss": 0.7159,
+      "step": 525
+    },
+    {
+      "epoch": 0.2576852418860992,
+      "grad_norm": 0.3035431206226349,
+      "learning_rate": 4.938536737258972e-06,
+      "loss": 0.6961,
+      "step": 526
+    },
+    {
+      "epoch": 0.25817513778322104,
+      "grad_norm": 0.29658591747283936,
+      "learning_rate": 4.938249033875759e-06,
+      "loss": 0.6925,
+      "step": 527
+    },
+    {
+      "epoch": 0.25866503368034294,
+      "grad_norm": 0.3085661232471466,
+      "learning_rate": 4.937960667128285e-06,
+      "loss": 0.6447,
+      "step": 528
+    },
+    {
+      "epoch": 0.2591549295774648,
+      "grad_norm": 0.30884072184562683,
+      "learning_rate": 4.9376716370950054e-06,
+      "loss": 0.6899,
+      "step": 529
+    },
+    {
+      "epoch": 0.25964482547458667,
+      "grad_norm": 0.30775752663612366,
+      "learning_rate": 4.937381943854554e-06,
+      "loss": 0.6918,
+      "step": 530
+    },
+    {
+      "epoch": 0.2601347213717085,
+      "grad_norm": 0.2908138930797577,
+      "learning_rate": 4.937091587485749e-06,
+      "loss": 0.691,
+      "step": 531
+    },
+    {
+      "epoch": 0.26062461726883035,
+      "grad_norm": 0.31106460094451904,
+      "learning_rate": 4.9368005680675825e-06,
+      "loss": 0.6864,
+      "step": 532
+    },
+    {
+      "epoch": 0.26111451316595224,
+      "grad_norm": 0.33596447110176086,
+      "learning_rate": 4.936508885679234e-06,
+      "loss": 0.6663,
+      "step": 533
+    },
+    {
+      "epoch": 0.2616044090630741,
+      "grad_norm": 0.30692586302757263,
+      "learning_rate": 4.936216540400059e-06,
+      "loss": 0.6774,
+      "step": 534
+    },
+    {
+      "epoch": 0.262094304960196,
+      "grad_norm": 0.30321234464645386,
+      "learning_rate": 4.935923532309596e-06,
+      "loss": 0.6516,
+      "step": 535
+    },
+    {
+      "epoch": 0.2625842008573178,
+      "grad_norm": 0.29872727394104004,
+      "learning_rate": 4.935629861487561e-06,
+      "loss": 0.6467,
+      "step": 536
+    },
+    {
+      "epoch": 0.26307409675443966,
+      "grad_norm": 0.3029206097126007,
+      "learning_rate": 4.9353355280138525e-06,
+      "loss": 0.6683,
+      "step": 537
+    },
+    {
+      "epoch": 0.26356399265156155,
+      "grad_norm": 0.2922782897949219,
+      "learning_rate": 4.935040531968548e-06,
+      "loss": 0.6554,
+      "step": 538
+    },
+    {
+      "epoch": 0.2640538885486834,
+      "grad_norm": 0.3034957945346832,
+      "learning_rate": 4.934744873431907e-06,
+      "loss": 0.6924,
+      "step": 539
+    },
+    {
+      "epoch": 0.2645437844458053,
+      "grad_norm": 0.28574198484420776,
+      "learning_rate": 4.934448552484367e-06,
+      "loss": 0.6931,
+      "step": 540
+    },
+    {
+      "epoch": 0.2650336803429271,
+      "grad_norm": 0.31014350056648254,
+      "learning_rate": 4.934151569206548e-06,
+      "loss": 0.677,
+      "step": 541
+    },
+    {
+      "epoch": 0.26552357624004896,
+      "grad_norm": 0.29660624265670776,
+      "learning_rate": 4.933853923679247e-06,
+      "loss": 0.6425,
+      "step": 542
+    },
+    {
+      "epoch": 0.26601347213717086,
+      "grad_norm": 0.3027561604976654,
+      "learning_rate": 4.933555615983445e-06,
+      "loss": 0.706,
+      "step": 543
+    },
+    {
+      "epoch": 0.2665033680342927,
+      "grad_norm": 0.3137056827545166,
+      "learning_rate": 4.933256646200301e-06,
+      "loss": 0.6832,
+      "step": 544
+    },
+    {
+      "epoch": 0.2669932639314146,
+      "grad_norm": 0.300325483083725,
+      "learning_rate": 4.9329570144111525e-06,
+      "loss": 0.6648,
+      "step": 545
+    },
+    {
+      "epoch": 0.26748315982853643,
+      "grad_norm": 0.29390445351600647,
+      "learning_rate": 4.932656720697523e-06,
+      "loss": 0.688,
+      "step": 546
+    },
+    {
+      "epoch": 0.26797305572565827,
+      "grad_norm": 0.3029269874095917,
+      "learning_rate": 4.932355765141108e-06,
+      "loss": 0.6781,
+      "step": 547
+    },
+    {
+      "epoch": 0.26846295162278017,
+      "grad_norm": 0.30967509746551514,
+      "learning_rate": 4.932054147823791e-06,
+      "loss": 0.6786,
+      "step": 548
+    },
+    {
+      "epoch": 0.268952847519902,
+      "grad_norm": 0.30142655968666077,
+      "learning_rate": 4.9317518688276285e-06,
+      "loss": 0.6964,
+      "step": 549
+    },
+    {
+      "epoch": 0.2694427434170239,
+      "grad_norm": 0.3007182776927948,
+      "learning_rate": 4.931448928234862e-06,
+      "loss": 0.6662,
+      "step": 550
+    },
+    {
+      "epoch": 0.26993263931414574,
+      "grad_norm": 0.29471781849861145,
+      "learning_rate": 4.931145326127911e-06,
+      "loss": 0.698,
+      "step": 551
+    },
+    {
+      "epoch": 0.2704225352112676,
+      "grad_norm": 0.30415260791778564,
+      "learning_rate": 4.930841062589375e-06,
+      "loss": 0.7043,
+      "step": 552
+    },
+    {
+      "epoch": 0.2709124311083895,
+      "grad_norm": 0.3111327290534973,
+      "learning_rate": 4.930536137702034e-06,
+      "loss": 0.7016,
+      "step": 553
+    },
+    {
+      "epoch": 0.2714023270055113,
+      "grad_norm": 0.3104318082332611,
+      "learning_rate": 4.930230551548848e-06,
+      "loss": 0.7092,
+      "step": 554
+    },
+    {
+      "epoch": 0.2718922229026332,
+      "grad_norm": 0.3054961562156677,
+      "learning_rate": 4.929924304212956e-06,
+      "loss": 0.658,
+      "step": 555
+    },
+    {
+      "epoch": 0.27238211879975505,
+      "grad_norm": 0.31626299023628235,
+      "learning_rate": 4.929617395777678e-06,
+      "loss": 0.684,
+      "step": 556
+    },
+    {
+      "epoch": 0.27287201469687694,
+      "grad_norm": 0.30073830485343933,
+      "learning_rate": 4.929309826326512e-06,
+      "loss": 0.6749,
+      "step": 557
+    },
+    {
+      "epoch": 0.2733619105939988,
+      "grad_norm": 0.30198878049850464,
+      "learning_rate": 4.929001595943138e-06,
+      "loss": 0.659,
+      "step": 558
+    },
+    {
+      "epoch": 0.2738518064911206,
+      "grad_norm": 0.3132243752479553,
+      "learning_rate": 4.928692704711416e-06,
+      "loss": 0.6918,
+      "step": 559
+    },
+    {
+      "epoch": 0.2743417023882425,
+      "grad_norm": 0.3196859359741211,
+      "learning_rate": 4.928383152715383e-06,
+      "loss": 0.6748,
+      "step": 560
+    },
+    {
+      "epoch": 0.27483159828536435,
+      "grad_norm": 0.2999397814273834,
+      "learning_rate": 4.928072940039258e-06,
+      "loss": 0.6835,
+      "step": 561
+    },
+    {
+      "epoch": 0.27532149418248625,
+      "grad_norm": 0.3084956705570221,
+      "learning_rate": 4.92776206676744e-06,
+      "loss": 0.6781,
+      "step": 562
+    },
+    {
+      "epoch": 0.2758113900796081,
+      "grad_norm": 0.3060546815395355,
+      "learning_rate": 4.927450532984507e-06,
+      "loss": 0.6698,
+      "step": 563
+    },
+    {
+      "epoch": 0.2763012859767299,
+      "grad_norm": 0.2978016138076782,
+      "learning_rate": 4.9271383387752145e-06,
+      "loss": 0.6748,
+      "step": 564
+    },
+    {
+      "epoch": 0.2767911818738518,
+      "grad_norm": 0.3078594505786896,
+      "learning_rate": 4.926825484224503e-06,
+      "loss": 0.6964,
+      "step": 565
+    },
+    {
+      "epoch": 0.27728107777097366,
+      "grad_norm": 0.30537673830986023,
+      "learning_rate": 4.926511969417488e-06,
+      "loss": 0.6821,
+      "step": 566
+    },
+    {
+      "epoch": 0.27777097366809556,
+      "grad_norm": 0.2968599200248718,
+      "learning_rate": 4.926197794439466e-06,
+      "loss": 0.6922,
+      "step": 567
+    },
+    {
+      "epoch": 0.2782608695652174,
+      "grad_norm": 0.3004016876220703,
+      "learning_rate": 4.9258829593759135e-06,
+      "loss": 0.6586,
+      "step": 568
+    },
+    {
+      "epoch": 0.27875076546233923,
+      "grad_norm": 0.2979053854942322,
+      "learning_rate": 4.925567464312487e-06,
+      "loss": 0.6963,
+      "step": 569
+    },
+    {
+      "epoch": 0.27924066135946113,
+      "grad_norm": 0.309206485748291,
+      "learning_rate": 4.9252513093350206e-06,
+      "loss": 0.6728,
+      "step": 570
+    },
+    {
+      "epoch": 0.27973055725658297,
+      "grad_norm": 0.2939196228981018,
+      "learning_rate": 4.924934494529531e-06,
+      "loss": 0.6411,
+      "step": 571
+    },
+    {
+      "epoch": 0.28022045315370486,
+      "grad_norm": 0.30378228425979614,
+      "learning_rate": 4.92461701998221e-06,
+      "loss": 0.6868,
+      "step": 572
+    },
+    {
+      "epoch": 0.2807103490508267,
+      "grad_norm": 0.2925695478916168,
+      "learning_rate": 4.924298885779435e-06,
+      "loss": 0.6987,
+      "step": 573
+    },
+    {
+      "epoch": 0.28120024494794854,
+      "grad_norm": 0.29253894090652466,
+      "learning_rate": 4.923980092007757e-06,
+      "loss": 0.6548,
+      "step": 574
+    },
+    {
+      "epoch": 0.28169014084507044,
+      "grad_norm": 0.30530020594596863,
+      "learning_rate": 4.923660638753911e-06,
+      "loss": 0.6667,
+      "step": 575
+    },
+    {
+      "epoch": 0.2821800367421923,
+      "grad_norm": 0.2993137240409851,
+      "learning_rate": 4.9233405261048066e-06,
+      "loss": 0.657,
+      "step": 576
+    },
+    {
+      "epoch": 0.28266993263931417,
+      "grad_norm": 0.3004169464111328,
+      "learning_rate": 4.923019754147537e-06,
+      "loss": 0.6751,
+      "step": 577
+    },
+    {
+      "epoch": 0.283159828536436,
+      "grad_norm": 0.3030836880207062,
+      "learning_rate": 4.922698322969372e-06,
+      "loss": 0.6605,
+      "step": 578
+    },
+    {
+      "epoch": 0.28364972443355785,
+      "grad_norm": 0.3000795841217041,
+      "learning_rate": 4.922376232657765e-06,
+      "loss": 0.7048,
+      "step": 579
+    },
+    {
+      "epoch": 0.28413962033067974,
+      "grad_norm": 0.3051553964614868,
+      "learning_rate": 4.922053483300343e-06,
+      "loss": 0.6613,
+      "step": 580
+    },
+    {
+      "epoch": 0.2846295162278016,
+      "grad_norm": 0.2963907718658447,
+      "learning_rate": 4.9217300749849175e-06,
+      "loss": 0.6541,
+      "step": 581
+    },
+    {
+      "epoch": 0.2851194121249235,
+      "grad_norm": 0.3138301968574524,
+      "learning_rate": 4.921406007799474e-06,
+      "loss": 0.6408,
+      "step": 582
+    },
+    {
+      "epoch": 0.2856093080220453,
+      "grad_norm": 0.292310893535614,
+      "learning_rate": 4.9210812818321814e-06,
+      "loss": 0.6784,
+      "step": 583
+    },
+    {
+      "epoch": 0.28609920391916716,
+      "grad_norm": 0.30917978286743164,
+      "learning_rate": 4.920755897171388e-06,
+      "loss": 0.679,
+      "step": 584
+    },
+    {
+      "epoch": 0.28658909981628905,
+      "grad_norm": 0.30815809965133667,
+      "learning_rate": 4.920429853905617e-06,
+      "loss": 0.669,
+      "step": 585
+    },
+    {
+      "epoch": 0.2870789957134109,
+      "grad_norm": 0.3038373589515686,
+      "learning_rate": 4.920103152123576e-06,
+      "loss": 0.6818,
+      "step": 586
+    },
+    {
+      "epoch": 0.2875688916105328,
+      "grad_norm": 0.3149436414241791,
+      "learning_rate": 4.919775791914148e-06,
+      "loss": 0.6871,
+      "step": 587
+    },
+    {
+      "epoch": 0.2880587875076546,
+      "grad_norm": 0.29806646704673767,
+      "learning_rate": 4.919447773366397e-06,
+      "loss": 0.6504,
+      "step": 588
+    },
+    {
+      "epoch": 0.28854868340477646,
+      "grad_norm": 0.29603317379951477,
+      "learning_rate": 4.919119096569567e-06,
+      "loss": 0.6709,
+      "step": 589
+    },
+    {
+      "epoch": 0.28903857930189836,
+      "grad_norm": 0.29373258352279663,
+      "learning_rate": 4.918789761613076e-06,
+      "loss": 0.6673,
+      "step": 590
+    },
+    {
+      "epoch": 0.2895284751990202,
+      "grad_norm": 0.3051702380180359,
+      "learning_rate": 4.91845976858653e-06,
+      "loss": 0.652,
+      "step": 591
+    },
+    {
+      "epoch": 0.2900183710961421,
+      "grad_norm": 0.3097754120826721,
+      "learning_rate": 4.918129117579704e-06,
+      "loss": 0.6782,
+      "step": 592
+    },
+    {
+      "epoch": 0.29050826699326393,
+      "grad_norm": 0.29908713698387146,
+      "learning_rate": 4.91779780868256e-06,
+      "loss": 0.6467,
+      "step": 593
+    },
+    {
+      "epoch": 0.29099816289038577,
+      "grad_norm": 0.30273228883743286,
+      "learning_rate": 4.917465841985234e-06,
+      "loss": 0.6372,
+      "step": 594
+    },
+    {
+      "epoch": 0.29148805878750766,
+      "grad_norm": 0.32377082109451294,
+      "learning_rate": 4.917133217578044e-06,
+      "loss": 0.7058,
+      "step": 595
+    },
+    {
+      "epoch": 0.2919779546846295,
+      "grad_norm": 0.31425973773002625,
+      "learning_rate": 4.916799935551485e-06,
+      "loss": 0.6975,
+      "step": 596
+    },
+    {
+      "epoch": 0.2924678505817514,
+      "grad_norm": 0.2944108247756958,
+      "learning_rate": 4.916465995996232e-06,
+      "loss": 0.6974,
+      "step": 597
+    },
+    {
+      "epoch": 0.29295774647887324,
+      "grad_norm": 0.3176249861717224,
+      "learning_rate": 4.916131399003139e-06,
+      "loss": 0.6825,
+      "step": 598
+    },
+    {
+      "epoch": 0.2934476423759951,
+      "grad_norm": 0.3081783950328827,
+      "learning_rate": 4.915796144663236e-06,
+      "loss": 0.6493,
+      "step": 599
+    },
+    {
+      "epoch": 0.29393753827311697,
+      "grad_norm": 0.3036037087440491,
+      "learning_rate": 4.915460233067738e-06,
+      "loss": 0.6487,
+      "step": 600
+    },
+    {
+      "epoch": 0.2944274341702388,
+      "grad_norm": 0.30290353298187256,
+      "learning_rate": 4.915123664308032e-06,
+      "loss": 0.6779,
+      "step": 601
+    },
+    {
+      "epoch": 0.2949173300673607,
+      "grad_norm": 0.3099123239517212,
+      "learning_rate": 4.9147864384756895e-06,
+      "loss": 0.6716,
+      "step": 602
+    },
+    {
+      "epoch": 0.29540722596448254,
+      "grad_norm": 0.29497596621513367,
+      "learning_rate": 4.914448555662455e-06,
+      "loss": 0.6712,
+      "step": 603
+    },
+    {
+      "epoch": 0.2958971218616044,
+      "grad_norm": 0.30663248896598816,
+      "learning_rate": 4.914110015960258e-06,
+      "loss": 0.656,
+      "step": 604
+    },
+    {
+      "epoch": 0.2963870177587263,
+      "grad_norm": 0.3016858994960785,
+      "learning_rate": 4.913770819461201e-06,
+      "loss": 0.6637,
+      "step": 605
+    },
+    {
+      "epoch": 0.2968769136558481,
+      "grad_norm": 0.3061820864677429,
+      "learning_rate": 4.913430966257569e-06,
+      "loss": 0.7025,
+      "step": 606
+    },
+    {
+      "epoch": 0.29736680955297,
+      "grad_norm": 0.31017857789993286,
+      "learning_rate": 4.913090456441826e-06,
+      "loss": 0.6858,
+      "step": 607
+    },
+    {
+      "epoch": 0.29785670545009185,
+      "grad_norm": 0.3078562319278717,
+      "learning_rate": 4.9127492901066095e-06,
+      "loss": 0.6884,
+      "step": 608
+    },
+    {
+      "epoch": 0.2983466013472137,
+      "grad_norm": 0.29675593972206116,
+      "learning_rate": 4.912407467344742e-06,
+      "loss": 0.6452,
+      "step": 609
+    },
+    {
+      "epoch": 0.2988364972443356,
+      "grad_norm": 0.3073781430721283,
+      "learning_rate": 4.9120649882492195e-06,
+      "loss": 0.6797,
+      "step": 610
+    },
+    {
+      "epoch": 0.2993263931414574,
+      "grad_norm": 0.3066023588180542,
+      "learning_rate": 4.911721852913223e-06,
+      "loss": 0.6618,
+      "step": 611
+    },
+    {
+      "epoch": 0.2998162890385793,
+      "grad_norm": 0.3042829930782318,
+      "learning_rate": 4.911378061430103e-06,
+      "loss": 0.6489,
+      "step": 612
+    },
+    {
+      "epoch": 0.30030618493570116,
+      "grad_norm": 0.2988530695438385,
+      "learning_rate": 4.911033613893397e-06,
+      "loss": 0.6994,
+      "step": 613
+    },
+    {
+      "epoch": 0.300796080832823,
+      "grad_norm": 0.31692102551460266,
+      "learning_rate": 4.910688510396815e-06,
+      "loss": 0.6698,
+      "step": 614
+    },
+    {
+      "epoch": 0.3012859767299449,
+      "grad_norm": 0.30138760805130005,
+      "learning_rate": 4.91034275103425e-06,
+      "loss": 0.652,
+      "step": 615
+    },
+    {
+      "epoch": 0.30177587262706673,
+      "grad_norm": 0.3250851035118103,
+      "learning_rate": 4.909996335899768e-06,
+      "loss": 0.6551,
+      "step": 616
+    },
+    {
+      "epoch": 0.3022657685241886,
+      "grad_norm": 0.31239888072013855,
+      "learning_rate": 4.909649265087621e-06,
+      "loss": 0.6676,
+      "step": 617
+    },
+    {
+      "epoch": 0.30275566442131047,
+      "grad_norm": 0.30099815130233765,
+      "learning_rate": 4.909301538692233e-06,
+      "loss": 0.6777,
+      "step": 618
+    },
+    {
+      "epoch": 0.3032455603184323,
+      "grad_norm": 0.3122364282608032,
+      "learning_rate": 4.908953156808207e-06,
+      "loss": 0.658,
+      "step": 619
+    },
+    {
+      "epoch": 0.3037354562155542,
+      "grad_norm": 0.3116292357444763,
+      "learning_rate": 4.908604119530328e-06,
+      "loss": 0.6574,
+      "step": 620
+    },
+    {
+      "epoch": 0.30422535211267604,
+      "grad_norm": 0.30465927720069885,
+      "learning_rate": 4.908254426953557e-06,
+      "loss": 0.6238,
+      "step": 621
+    },
+    {
+      "epoch": 0.30471524800979793,
+      "grad_norm": 0.3098441958427429,
+      "learning_rate": 4.907904079173033e-06,
+      "loss": 0.6996,
+      "step": 622
+    },
+    {
+      "epoch": 0.3052051439069198,
+      "grad_norm": 0.3130321502685547,
+      "learning_rate": 4.907553076284073e-06,
+      "loss": 0.6849,
+      "step": 623
+    },
+    {
+      "epoch": 0.30569503980404167,
+      "grad_norm": 0.31285127997398376,
+      "learning_rate": 4.907201418382174e-06,
+      "loss": 0.6887,
+      "step": 624
+    },
+    {
+      "epoch": 0.3061849357011635,
+      "grad_norm": 0.3078165650367737,
+      "learning_rate": 4.906849105563008e-06,
+      "loss": 0.6516,
+      "step": 625
+    },
+    {
+      "epoch": 0.30667483159828535,
+      "grad_norm": 0.30509424209594727,
+      "learning_rate": 4.906496137922429e-06,
+      "loss": 0.6655,
+      "step": 626
+    },
+    {
+      "epoch": 0.30716472749540724,
+      "grad_norm": 0.30886662006378174,
+      "learning_rate": 4.906142515556469e-06,
+      "loss": 0.6742,
+      "step": 627
+    },
+    {
+      "epoch": 0.3076546233925291,
+      "grad_norm": 0.30892255902290344,
+      "learning_rate": 4.905788238561333e-06,
+      "loss": 0.6623,
+      "step": 628
+    },
+    {
+      "epoch": 0.308144519289651,
+      "grad_norm": 0.3067830502986908,
+      "learning_rate": 4.905433307033411e-06,
+      "loss": 0.6616,
+      "step": 629
+    },
+    {
+      "epoch": 0.3086344151867728,
+      "grad_norm": 0.3024778962135315,
+      "learning_rate": 4.905077721069265e-06,
+      "loss": 0.6747,
+      "step": 630
+    },
+    {
+      "epoch": 0.30912431108389465,
+      "grad_norm": 0.3080230951309204,
+      "learning_rate": 4.90472148076564e-06,
+      "loss": 0.6855,
+      "step": 631
+    },
+    {
+      "epoch": 0.30961420698101655,
+      "grad_norm": 0.3134588897228241,
+      "learning_rate": 4.9043645862194545e-06,
+      "loss": 0.6981,
+      "step": 632
+    },
+    {
+      "epoch": 0.3101041028781384,
+      "grad_norm": 0.3064582049846649,
+      "learning_rate": 4.904007037527809e-06,
+      "loss": 0.6902,
+      "step": 633
+    },
+    {
+      "epoch": 0.3105939987752603,
+      "grad_norm": 0.3072272837162018,
+      "learning_rate": 4.903648834787981e-06,
+      "loss": 0.6876,
+      "step": 634
+    },
+    {
+      "epoch": 0.3110838946723821,
+      "grad_norm": 0.31387409567832947,
+      "learning_rate": 4.903289978097423e-06,
+      "loss": 0.6981,
+      "step": 635
+    },
+    {
+      "epoch": 0.31157379056950396,
+      "grad_norm": 0.30770981311798096,
+      "learning_rate": 4.90293046755377e-06,
+      "loss": 0.6514,
+      "step": 636
+    },
+    {
+      "epoch": 0.31206368646662586,
+      "grad_norm": 0.2999817132949829,
+      "learning_rate": 4.90257030325483e-06,
+      "loss": 0.6645,
+      "step": 637
+    },
+    {
+      "epoch": 0.3125535823637477,
+      "grad_norm": 0.3214426636695862,
+      "learning_rate": 4.9022094852985925e-06,
+      "loss": 0.6754,
+      "step": 638
+    },
+    {
+      "epoch": 0.3130434782608696,
+      "grad_norm": 0.3216695487499237,
+      "learning_rate": 4.901848013783225e-06,
+      "loss": 0.6642,
+      "step": 639
+    },
+    {
+      "epoch": 0.31353337415799143,
+      "grad_norm": 0.3139570355415344,
+      "learning_rate": 4.901485888807069e-06,
+      "loss": 0.6535,
+      "step": 640
+    },
+    {
+      "epoch": 0.31402327005511327,
+      "grad_norm": 0.3131052851676941,
+      "learning_rate": 4.9011231104686495e-06,
+      "loss": 0.6743,
+      "step": 641
+    },
+    {
+      "epoch": 0.31451316595223516,
+      "grad_norm": 0.2983650863170624,
+      "learning_rate": 4.900759678866663e-06,
+      "loss": 0.6808,
+      "step": 642
+    },
+    {
+      "epoch": 0.315003061849357,
+      "grad_norm": 0.31281766295433044,
+      "learning_rate": 4.90039559409999e-06,
+      "loss": 0.6625,
+      "step": 643
+    },
+    {
+      "epoch": 0.3154929577464789,
+      "grad_norm": 0.3253592550754547,
+      "learning_rate": 4.9000308562676825e-06,
+      "loss": 0.6513,
+      "step": 644
+    },
+    {
+      "epoch": 0.31598285364360074,
+      "grad_norm": 0.3129440248012543,
+      "learning_rate": 4.8996654654689745e-06,
+      "loss": 0.6746,
+      "step": 645
+    },
+    {
+      "epoch": 0.3164727495407226,
+      "grad_norm": 0.3100399971008301,
+      "learning_rate": 4.899299421803277e-06,
+      "loss": 0.6788,
+      "step": 646
+    },
+    {
+      "epoch": 0.31696264543784447,
+      "grad_norm": 0.3134135901927948,
+      "learning_rate": 4.898932725370177e-06,
+      "loss": 0.667,
+      "step": 647
+    },
+    {
+      "epoch": 0.3174525413349663,
+      "grad_norm": 0.2989857792854309,
+      "learning_rate": 4.898565376269439e-06,
+      "loss": 0.6549,
+      "step": 648
+    },
+    {
+      "epoch": 0.3179424372320882,
+      "grad_norm": 0.307166188955307,
+      "learning_rate": 4.89819737460101e-06,
+      "loss": 0.6819,
+      "step": 649
+    },
+    {
+      "epoch": 0.31843233312921004,
+      "grad_norm": 0.3121511936187744,
+      "learning_rate": 4.897828720465007e-06,
+      "loss": 0.6865,
+      "step": 650
+    },
+    {
+      "epoch": 0.3189222290263319,
+      "grad_norm": 0.3052747845649719,
+      "learning_rate": 4.897459413961729e-06,
+      "loss": 0.6588,
+      "step": 651
+    },
+    {
+      "epoch": 0.3194121249234538,
+      "grad_norm": 0.3178597688674927,
+      "learning_rate": 4.897089455191653e-06,
+      "loss": 0.6671,
+      "step": 652
+    },
+    {
+      "epoch": 0.3199020208205756,
+      "grad_norm": 0.3084629476070404,
+      "learning_rate": 4.896718844255431e-06,
+      "loss": 0.6698,
+      "step": 653
+    },
+    {
+      "epoch": 0.3203919167176975,
+      "grad_norm": 0.32827791571617126,
+      "learning_rate": 4.896347581253893e-06,
+      "loss": 0.667,
+      "step": 654
+    },
+    {
+      "epoch": 0.32088181261481935,
+      "grad_norm": 0.3143371641635895,
+      "learning_rate": 4.895975666288048e-06,
+      "loss": 0.6646,
+      "step": 655
+    },
+    {
+      "epoch": 0.3213717085119412,
+      "grad_norm": 0.3030723035335541,
+      "learning_rate": 4.8956030994590805e-06,
+      "loss": 0.6569,
+      "step": 656
+    },
+    {
+      "epoch": 0.3218616044090631,
+      "grad_norm": 0.313111275434494,
+      "learning_rate": 4.8952298808683554e-06,
+      "loss": 0.6601,
+      "step": 657
+    },
+    {
+      "epoch": 0.3223515003061849,
+      "grad_norm": 0.33992889523506165,
+      "learning_rate": 4.894856010617411e-06,
+      "loss": 0.6966,
+      "step": 658
+    },
+    {
+      "epoch": 0.3228413962033068,
+      "grad_norm": 0.31487035751342773,
+      "learning_rate": 4.894481488807964e-06,
+      "loss": 0.6819,
+      "step": 659
+    },
+    {
+      "epoch": 0.32333129210042866,
+      "grad_norm": 0.30711787939071655,
+      "learning_rate": 4.894106315541909e-06,
+      "loss": 0.6497,
+      "step": 660
+    },
+    {
+      "epoch": 0.3238211879975505,
+      "grad_norm": 0.30664297938346863,
+      "learning_rate": 4.89373049092132e-06,
+      "loss": 0.6343,
+      "step": 661
+    },
+    {
+      "epoch": 0.3243110838946724,
+      "grad_norm": 0.32928934693336487,
+      "learning_rate": 4.893354015048444e-06,
+      "loss": 0.6819,
+      "step": 662
+    },
+    {
+      "epoch": 0.32480097979179423,
+      "grad_norm": 0.3158688545227051,
+      "learning_rate": 4.892976888025708e-06,
+      "loss": 0.671,
+      "step": 663
+    },
+    {
+      "epoch": 0.3252908756889161,
+      "grad_norm": 0.311565101146698,
+      "learning_rate": 4.892599109955715e-06,
+      "loss": 0.6619,
+      "step": 664
+    },
+    {
+      "epoch": 0.32578077158603796,
+      "grad_norm": 0.317327082157135,
+      "learning_rate": 4.892220680941247e-06,
+      "loss": 0.6821,
+      "step": 665
+    },
+    {
+      "epoch": 0.3262706674831598,
+      "grad_norm": 0.32189080119132996,
+      "learning_rate": 4.891841601085259e-06,
+      "loss": 0.6384,
+      "step": 666
+    },
+    {
+      "epoch": 0.3267605633802817,
+      "grad_norm": 0.3046722412109375,
+      "learning_rate": 4.891461870490887e-06,
+      "loss": 0.673,
+      "step": 667
+    },
+    {
+      "epoch": 0.32725045927740354,
+      "grad_norm": 0.31418314576148987,
+      "learning_rate": 4.891081489261443e-06,
+      "loss": 0.6604,
+      "step": 668
+    },
+    {
+      "epoch": 0.32774035517452543,
+      "grad_norm": 0.31496915221214294,
+      "learning_rate": 4.890700457500416e-06,
+      "loss": 0.6448,
+      "step": 669
+    },
+    {
+      "epoch": 0.32823025107164727,
+      "grad_norm": 0.3126322329044342,
+      "learning_rate": 4.890318775311471e-06,
+      "loss": 0.6533,
+      "step": 670
+    },
+    {
+      "epoch": 0.3287201469687691,
+      "grad_norm": 0.305670827627182,
+      "learning_rate": 4.889936442798451e-06,
+      "loss": 0.64,
+      "step": 671
+    },
+    {
+      "epoch": 0.329210042865891,
+      "grad_norm": 0.32083502411842346,
+      "learning_rate": 4.889553460065375e-06,
+      "loss": 0.6648,
+      "step": 672
+    },
+    {
+      "epoch": 0.32969993876301285,
+      "grad_norm": 0.30888432264328003,
+      "learning_rate": 4.88916982721644e-06,
+      "loss": 0.6468,
+      "step": 673
+    },
+    {
+      "epoch": 0.33018983466013474,
+      "grad_norm": 0.3246508836746216,
+      "learning_rate": 4.88878554435602e-06,
+      "loss": 0.6956,
+      "step": 674
+    },
+    {
+      "epoch": 0.3306797305572566,
+      "grad_norm": 0.31742051243782043,
+      "learning_rate": 4.888400611588664e-06,
+      "loss": 0.6581,
+      "step": 675
+    },
+    {
+      "epoch": 0.3311696264543784,
+      "grad_norm": 0.2962053716182709,
+      "learning_rate": 4.8880150290191e-06,
+      "loss": 0.6419,
+      "step": 676
+    },
+    {
+      "epoch": 0.3316595223515003,
+      "grad_norm": 0.302521675825119,
+      "learning_rate": 4.887628796752232e-06,
+      "loss": 0.6541,
+      "step": 677
+    },
+    {
+      "epoch": 0.33214941824862215,
+      "grad_norm": 0.3055395483970642,
+      "learning_rate": 4.88724191489314e-06,
+      "loss": 0.657,
+      "step": 678
+    },
+    {
+      "epoch": 0.33263931414574405,
+      "grad_norm": 0.3162010908126831,
+      "learning_rate": 4.886854383547081e-06,
+      "loss": 0.6679,
+      "step": 679
+    },
+    {
+      "epoch": 0.3331292100428659,
+      "grad_norm": 0.3197442591190338,
+      "learning_rate": 4.88646620281949e-06,
+      "loss": 0.6334,
+      "step": 680
+    },
+    {
+      "epoch": 0.3336191059399877,
+      "grad_norm": 0.3153987228870392,
+      "learning_rate": 4.8860773728159775e-06,
+      "loss": 0.6522,
+      "step": 681
+    },
+    {
+      "epoch": 0.3341090018371096,
+      "grad_norm": 0.32448455691337585,
+      "learning_rate": 4.885687893642329e-06,
+      "loss": 0.6632,
+      "step": 682
+    },
+    {
+      "epoch": 0.33459889773423146,
+      "grad_norm": 0.31581366062164307,
+      "learning_rate": 4.885297765404512e-06,
+      "loss": 0.6734,
+      "step": 683
+    },
+    {
+      "epoch": 0.33508879363135335,
+      "grad_norm": 0.32606983184814453,
+      "learning_rate": 4.884906988208664e-06,
+      "loss": 0.6559,
+      "step": 684
+    },
+    {
+      "epoch": 0.3355786895284752,
+      "grad_norm": 0.3181923031806946,
+      "learning_rate": 4.884515562161104e-06,
+      "loss": 0.6753,
+      "step": 685
+    },
+    {
+      "epoch": 0.3360685854255971,
+      "grad_norm": 0.29703840613365173,
+      "learning_rate": 4.884123487368324e-06,
+      "loss": 0.6682,
+      "step": 686
+    },
+    {
+      "epoch": 0.3365584813227189,
+      "grad_norm": 0.30182990431785583,
+      "learning_rate": 4.8837307639369966e-06,
+      "loss": 0.6391,
+      "step": 687
+    },
+    {
+      "epoch": 0.33704837721984077,
+      "grad_norm": 0.3200860917568207,
+      "learning_rate": 4.883337391973966e-06,
+      "loss": 0.6768,
+      "step": 688
+    },
+    {
+      "epoch": 0.33753827311696266,
+      "grad_norm": 0.29847651720046997,
+      "learning_rate": 4.882943371586256e-06,
+      "loss": 0.6144,
+      "step": 689
+    },
+    {
+      "epoch": 0.3380281690140845,
+      "grad_norm": 0.3135667145252228,
+      "learning_rate": 4.882548702881067e-06,
+      "loss": 0.6512,
+      "step": 690
+    },
+    {
+      "epoch": 0.3385180649112064,
+      "grad_norm": 0.2976519465446472,
+      "learning_rate": 4.882153385965774e-06,
+      "loss": 0.6572,
+      "step": 691
+    },
+    {
+      "epoch": 0.33900796080832823,
+      "grad_norm": 0.3228099048137665,
+      "learning_rate": 4.881757420947928e-06,
+      "loss": 0.6546,
+      "step": 692
+    },
+    {
+      "epoch": 0.3394978567054501,
+      "grad_norm": 0.3073636293411255,
+      "learning_rate": 4.88136080793526e-06,
+      "loss": 0.6427,
+      "step": 693
+    },
+    {
+      "epoch": 0.33998775260257197,
+      "grad_norm": 0.3015718162059784,
+      "learning_rate": 4.880963547035674e-06,
+      "loss": 0.6581,
+      "step": 694
+    },
+    {
+      "epoch": 0.3404776484996938,
+      "grad_norm": 0.30982354283332825,
+      "learning_rate": 4.88056563835725e-06,
+      "loss": 0.6349,
+      "step": 695
+    },
+    {
+      "epoch": 0.3409675443968157,
+      "grad_norm": 0.3180277645587921,
+      "learning_rate": 4.880167082008247e-06,
+      "loss": 0.6618,
+      "step": 696
+    },
+    {
+      "epoch": 0.34145744029393754,
+      "grad_norm": 0.3110656440258026,
+      "learning_rate": 4.879767878097098e-06,
+      "loss": 0.665,
+      "step": 697
+    },
+    {
+      "epoch": 0.3419473361910594,
+      "grad_norm": 0.3028208911418915,
+      "learning_rate": 4.879368026732412e-06,
+      "loss": 0.6707,
+      "step": 698
+    },
+    {
+      "epoch": 0.3424372320881813,
+      "grad_norm": 0.3225865066051483,
+      "learning_rate": 4.878967528022976e-06,
+      "loss": 0.6498,
+      "step": 699
+    },
+    {
+      "epoch": 0.3429271279853031,
+      "grad_norm": 0.3116249442100525,
+      "learning_rate": 4.87856638207775e-06,
+      "loss": 0.6627,
+      "step": 700
+    },
+    {
+      "epoch": 0.343417023882425,
+      "grad_norm": 0.32170018553733826,
+      "learning_rate": 4.878164589005876e-06,
+      "loss": 0.6571,
+      "step": 701
+    },
+    {
+      "epoch": 0.34390691977954685,
+      "grad_norm": 0.3243235647678375,
+      "learning_rate": 4.877762148916665e-06,
+      "loss": 0.6281,
+      "step": 702
+    },
+    {
+      "epoch": 0.3443968156766687,
+      "grad_norm": 0.30337801575660706,
+      "learning_rate": 4.8773590619196064e-06,
+      "loss": 0.6781,
+      "step": 703
+    },
+    {
+      "epoch": 0.3448867115737906,
+      "grad_norm": 0.30504652857780457,
+      "learning_rate": 4.87695532812437e-06,
+      "loss": 0.6737,
+      "step": 704
+    },
+    {
+      "epoch": 0.3453766074709124,
+      "grad_norm": 0.33436259627342224,
+      "learning_rate": 4.876550947640794e-06,
+      "loss": 0.6872,
+      "step": 705
+    },
+    {
+      "epoch": 0.3458665033680343,
+      "grad_norm": 0.3146010935306549,
+      "learning_rate": 4.8761459205789e-06,
+      "loss": 0.6522,
+      "step": 706
+    },
+    {
+      "epoch": 0.34635639926515616,
+      "grad_norm": 0.3099222779273987,
+      "learning_rate": 4.8757402470488795e-06,
+      "loss": 0.6409,
+      "step": 707
+    },
+    {
+      "epoch": 0.346846295162278,
+      "grad_norm": 0.30881136655807495,
+      "learning_rate": 4.875333927161104e-06,
+      "loss": 0.6406,
+      "step": 708
+    },
+    {
+      "epoch": 0.3473361910593999,
+      "grad_norm": 0.310293048620224,
+      "learning_rate": 4.874926961026118e-06,
+      "loss": 0.6476,
+      "step": 709
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 0.33084210753440857,
+      "learning_rate": 4.874519348754644e-06,
+      "loss": 0.6554,
+      "step": 710
+    },
+    {
+      "epoch": 0.3483159828536436,
+      "grad_norm": 0.3112454116344452,
+      "learning_rate": 4.874111090457579e-06,
+      "loss": 0.668,
+      "step": 711
+    },
+    {
+      "epoch": 0.34880587875076546,
+      "grad_norm": 0.32594504952430725,
+      "learning_rate": 4.873702186245996e-06,
+      "loss": 0.678,
+      "step": 712
+    },
+    {
+      "epoch": 0.3492957746478873,
+      "grad_norm": 0.3138704299926758,
+      "learning_rate": 4.873292636231145e-06,
+      "loss": 0.6708,
+      "step": 713
+    },
+    {
+      "epoch": 0.3497856705450092,
+      "grad_norm": 0.3248952627182007,
+      "learning_rate": 4.872882440524448e-06,
+      "loss": 0.6453,
+      "step": 714
+    },
+    {
+      "epoch": 0.35027556644213104,
+      "grad_norm": 0.30387961864471436,
+      "learning_rate": 4.872471599237508e-06,
+      "loss": 0.6293,
+      "step": 715
+    },
+    {
+      "epoch": 0.35076546233925293,
+      "grad_norm": 0.3210145831108093,
+      "learning_rate": 4.8720601124821e-06,
+      "loss": 0.6811,
+      "step": 716
+    },
+    {
+      "epoch": 0.35125535823637477,
+      "grad_norm": 0.3145703673362732,
+      "learning_rate": 4.8716479803701755e-06,
+      "loss": 0.6534,
+      "step": 717
+    },
+    {
+      "epoch": 0.3517452541334966,
+      "grad_norm": 0.32032543420791626,
+      "learning_rate": 4.871235203013861e-06,
+      "loss": 0.677,
+      "step": 718
+    },
+    {
+      "epoch": 0.3522351500306185,
+      "grad_norm": 0.311057984828949,
+      "learning_rate": 4.870821780525458e-06,
+      "loss": 0.6599,
+      "step": 719
+    },
+    {
+      "epoch": 0.35272504592774034,
+      "grad_norm": 0.31371673941612244,
+      "learning_rate": 4.8704077130174475e-06,
+      "loss": 0.6655,
+      "step": 720
+    },
+    {
+      "epoch": 0.35321494182486224,
+      "grad_norm": 0.31282398104667664,
+      "learning_rate": 4.869993000602482e-06,
+      "loss": 0.6588,
+      "step": 721
+    },
+    {
+      "epoch": 0.3537048377219841,
+      "grad_norm": 0.3058509826660156,
+      "learning_rate": 4.8695776433933895e-06,
+      "loss": 0.6554,
+      "step": 722
+    },
+    {
+      "epoch": 0.3541947336191059,
+      "grad_norm": 0.32043665647506714,
+      "learning_rate": 4.8691616415031765e-06,
+      "loss": 0.6419,
+      "step": 723
+    },
+    {
+      "epoch": 0.3546846295162278,
+      "grad_norm": 0.3181472718715668,
+      "learning_rate": 4.86874499504502e-06,
+      "loss": 0.6498,
+      "step": 724
+    },
+    {
+      "epoch": 0.35517452541334965,
+      "grad_norm": 0.31420233845710754,
+      "learning_rate": 4.868327704132277e-06,
+      "loss": 0.6833,
+      "step": 725
+    },
+    {
+      "epoch": 0.35566442131047155,
+      "grad_norm": 0.30752044916152954,
+      "learning_rate": 4.867909768878479e-06,
+      "loss": 0.6696,
+      "step": 726
+    },
+    {
+      "epoch": 0.3561543172075934,
+      "grad_norm": 0.2980179488658905,
+      "learning_rate": 4.867491189397331e-06,
+      "loss": 0.6181,
+      "step": 727
+    },
+    {
+      "epoch": 0.3566442131047152,
+      "grad_norm": 0.33069589734077454,
+      "learning_rate": 4.867071965802715e-06,
+      "loss": 0.6842,
+      "step": 728
+    },
+    {
+      "epoch": 0.3571341090018371,
+      "grad_norm": 0.31239742040634155,
+      "learning_rate": 4.866652098208684e-06,
+      "loss": 0.6794,
+      "step": 729
+    },
+    {
+      "epoch": 0.35762400489895896,
+      "grad_norm": 0.3049803674221039,
+      "learning_rate": 4.866231586729474e-06,
+      "loss": 0.6487,
+      "step": 730
+    },
+    {
+      "epoch": 0.35811390079608085,
+      "grad_norm": 0.32708168029785156,
+      "learning_rate": 4.86581043147949e-06,
+      "loss": 0.6732,
+      "step": 731
+    },
+    {
+      "epoch": 0.3586037966932027,
+      "grad_norm": 0.317973792552948,
+      "learning_rate": 4.865388632573313e-06,
+      "loss": 0.6676,
+      "step": 732
+    },
+    {
+      "epoch": 0.35909369259032453,
+      "grad_norm": 0.3271551728248596,
+      "learning_rate": 4.864966190125702e-06,
+      "loss": 0.6439,
+      "step": 733
+    },
+    {
+      "epoch": 0.3595835884874464,
+      "grad_norm": 0.3095121383666992,
+      "learning_rate": 4.864543104251587e-06,
+      "loss": 0.6134,
+      "step": 734
+    },
+    {
+      "epoch": 0.36007348438456827,
+      "grad_norm": 0.3160260319709778,
+      "learning_rate": 4.864119375066077e-06,
+      "loss": 0.6594,
+      "step": 735
+    },
+    {
+      "epoch": 0.36056338028169016,
+      "grad_norm": 0.3187366724014282,
+      "learning_rate": 4.8636950026844534e-06,
+      "loss": 0.6758,
+      "step": 736
+    },
+    {
+      "epoch": 0.361053276178812,
+      "grad_norm": 0.32910969853401184,
+      "learning_rate": 4.863269987222174e-06,
+      "loss": 0.6372,
+      "step": 737
+    },
+    {
+      "epoch": 0.36154317207593384,
+      "grad_norm": 0.3208555281162262,
+      "learning_rate": 4.862844328794871e-06,
+      "loss": 0.632,
+      "step": 738
+    },
+    {
+      "epoch": 0.36203306797305573,
+      "grad_norm": 0.31716328859329224,
+      "learning_rate": 4.862418027518351e-06,
+      "loss": 0.6717,
+      "step": 739
+    },
+    {
+      "epoch": 0.3625229638701776,
+      "grad_norm": 0.33008137345314026,
+      "learning_rate": 4.861991083508595e-06,
+      "loss": 0.6629,
+      "step": 740
+    },
+    {
+      "epoch": 0.36301285976729947,
+      "grad_norm": 0.32629767060279846,
+      "learning_rate": 4.861563496881763e-06,
+      "loss": 0.6517,
+      "step": 741
+    },
+    {
+      "epoch": 0.3635027556644213,
+      "grad_norm": 0.32302382588386536,
+      "learning_rate": 4.861135267754183e-06,
+      "loss": 0.6497,
+      "step": 742
+    },
+    {
+      "epoch": 0.36399265156154315,
+      "grad_norm": 0.31350627541542053,
+      "learning_rate": 4.860706396242364e-06,
+      "loss": 0.6607,
+      "step": 743
+    },
+    {
+      "epoch": 0.36448254745866504,
+      "grad_norm": 0.3175422251224518,
+      "learning_rate": 4.860276882462986e-06,
+      "loss": 0.6311,
+      "step": 744
+    },
+    {
+      "epoch": 0.3649724433557869,
+      "grad_norm": 0.31468355655670166,
+      "learning_rate": 4.8598467265329065e-06,
+      "loss": 0.6402,
+      "step": 745
+    },
+    {
+      "epoch": 0.3654623392529088,
+      "grad_norm": 0.3147619962692261,
+      "learning_rate": 4.859415928569154e-06,
+      "loss": 0.6882,
+      "step": 746
+    },
+    {
+      "epoch": 0.3659522351500306,
+      "grad_norm": 0.31198322772979736,
+      "learning_rate": 4.858984488688937e-06,
+      "loss": 0.6679,
+      "step": 747
+    },
+    {
+      "epoch": 0.36644213104715245,
+      "grad_norm": 0.3297489285469055,
+      "learning_rate": 4.858552407009633e-06,
+      "loss": 0.685,
+      "step": 748
+    },
+    {
+      "epoch": 0.36693202694427435,
+      "grad_norm": 0.307198166847229,
+      "learning_rate": 4.858119683648796e-06,
+      "loss": 0.6682,
+      "step": 749
+    },
+    {
+      "epoch": 0.3674219228413962,
+      "grad_norm": 0.3127272129058838,
+      "learning_rate": 4.857686318724159e-06,
+      "loss": 0.6529,
+      "step": 750
+    },
+    {
+      "epoch": 0.3679118187385181,
+      "grad_norm": 0.31503623723983765,
+      "learning_rate": 4.857252312353622e-06,
+      "loss": 0.6607,
+      "step": 751
+    },
+    {
+      "epoch": 0.3684017146356399,
+      "grad_norm": 0.3225836157798767,
+      "learning_rate": 4.856817664655266e-06,
+      "loss": 0.6539,
+      "step": 752
+    },
+    {
+      "epoch": 0.3688916105327618,
+      "grad_norm": 0.3189702332019806,
+      "learning_rate": 4.856382375747341e-06,
+      "loss": 0.68,
+      "step": 753
+    },
+    {
+      "epoch": 0.36938150642988365,
+      "grad_norm": 0.3274145722389221,
+      "learning_rate": 4.855946445748277e-06,
+      "loss": 0.6388,
+      "step": 754
+    },
+    {
+      "epoch": 0.3698714023270055,
+      "grad_norm": 0.3334755003452301,
+      "learning_rate": 4.855509874776674e-06,
+      "loss": 0.6597,
+      "step": 755
+    },
+    {
+      "epoch": 0.3703612982241274,
+      "grad_norm": 0.3221217095851898,
+      "learning_rate": 4.855072662951308e-06,
+      "loss": 0.6532,
+      "step": 756
+    },
+    {
+      "epoch": 0.37085119412124923,
+      "grad_norm": 0.32857850193977356,
+      "learning_rate": 4.85463481039113e-06,
+      "loss": 0.6585,
+      "step": 757
+    },
+    {
+      "epoch": 0.3713410900183711,
+      "grad_norm": 0.3258369266986847,
+      "learning_rate": 4.854196317215265e-06,
+      "loss": 0.673,
+      "step": 758
+    },
+    {
+      "epoch": 0.37183098591549296,
+      "grad_norm": 0.3420117497444153,
+      "learning_rate": 4.853757183543012e-06,
+      "loss": 0.6838,
+      "step": 759
+    },
+    {
+      "epoch": 0.3723208818126148,
+      "grad_norm": 0.30562669038772583,
+      "learning_rate": 4.853317409493844e-06,
+      "loss": 0.6298,
+      "step": 760
+    },
+    {
+      "epoch": 0.3728107777097367,
+      "grad_norm": 0.3201233148574829,
+      "learning_rate": 4.852876995187408e-06,
+      "loss": 0.6723,
+      "step": 761
+    },
+    {
+      "epoch": 0.37330067360685854,
+      "grad_norm": 0.2966959774494171,
+      "learning_rate": 4.852435940743526e-06,
+      "loss": 0.6491,
+      "step": 762
+    },
+    {
+      "epoch": 0.37379056950398043,
+      "grad_norm": 0.32500383257865906,
+      "learning_rate": 4.851994246282194e-06,
+      "loss": 0.6564,
+      "step": 763
+    },
+    {
+      "epoch": 0.37428046540110227,
+      "grad_norm": 0.3188881278038025,
+      "learning_rate": 4.851551911923582e-06,
+      "loss": 0.6505,
+      "step": 764
+    },
+    {
+      "epoch": 0.3747703612982241,
+      "grad_norm": 0.3221505284309387,
+      "learning_rate": 4.8511089377880334e-06,
+      "loss": 0.6634,
+      "step": 765
+    },
+    {
+      "epoch": 0.375260257195346,
+      "grad_norm": 0.3052166700363159,
+      "learning_rate": 4.850665323996068e-06,
+      "loss": 0.6639,
+      "step": 766
+    },
+    {
+      "epoch": 0.37575015309246784,
+      "grad_norm": 0.33018791675567627,
+      "learning_rate": 4.850221070668376e-06,
+      "loss": 0.6912,
+      "step": 767
+    },
+    {
+      "epoch": 0.37624004898958974,
+      "grad_norm": 0.32073283195495605,
+      "learning_rate": 4.849776177925825e-06,
+      "loss": 0.6493,
+      "step": 768
+    },
+    {
+      "epoch": 0.3767299448867116,
+      "grad_norm": 0.31678506731987,
+      "learning_rate": 4.849330645889454e-06,
+      "loss": 0.6803,
+      "step": 769
+    },
+    {
+      "epoch": 0.3772198407838334,
+      "grad_norm": 0.32724466919898987,
+      "learning_rate": 4.848884474680477e-06,
+      "loss": 0.6704,
+      "step": 770
+    },
+    {
+      "epoch": 0.3777097366809553,
+      "grad_norm": 0.3205874264240265,
+      "learning_rate": 4.8484376644202845e-06,
+      "loss": 0.6577,
+      "step": 771
+    },
+    {
+      "epoch": 0.37819963257807715,
+      "grad_norm": 0.30893105268478394,
+      "learning_rate": 4.847990215230435e-06,
+      "loss": 0.666,
+      "step": 772
+    },
+    {
+      "epoch": 0.37868952847519904,
+      "grad_norm": 0.31219717860221863,
+      "learning_rate": 4.847542127232665e-06,
+      "loss": 0.6565,
+      "step": 773
+    },
+    {
+      "epoch": 0.3791794243723209,
+      "grad_norm": 0.3261759579181671,
+      "learning_rate": 4.847093400548885e-06,
+      "loss": 0.677,
+      "step": 774
+    },
+    {
+      "epoch": 0.3796693202694427,
+      "grad_norm": 0.31503742933273315,
+      "learning_rate": 4.846644035301178e-06,
+      "loss": 0.6666,
+      "step": 775
+    },
+    {
+      "epoch": 0.3801592161665646,
+      "grad_norm": 0.31190887093544006,
+      "learning_rate": 4.8461940316118e-06,
+      "loss": 0.6574,
+      "step": 776
+    },
+    {
+      "epoch": 0.38064911206368646,
+      "grad_norm": 0.3217906355857849,
+      "learning_rate": 4.845743389603182e-06,
+      "loss": 0.6441,
+      "step": 777
+    },
+    {
+      "epoch": 0.38113900796080835,
+      "grad_norm": 0.33193954825401306,
+      "learning_rate": 4.845292109397929e-06,
+      "loss": 0.6248,
+      "step": 778
+    },
+    {
+      "epoch": 0.3816289038579302,
+      "grad_norm": 0.3300223648548126,
+      "learning_rate": 4.844840191118819e-06,
+      "loss": 0.6431,
+      "step": 779
+    },
+    {
+      "epoch": 0.38211879975505203,
+      "grad_norm": 0.33220401406288147,
+      "learning_rate": 4.844387634888802e-06,
+      "loss": 0.6605,
+      "step": 780
+    },
+    {
+      "epoch": 0.3826086956521739,
+      "grad_norm": 0.3159903585910797,
+      "learning_rate": 4.8439344408310044e-06,
+      "loss": 0.655,
+      "step": 781
+    },
+    {
+      "epoch": 0.38309859154929576,
+      "grad_norm": 0.32114678621292114,
+      "learning_rate": 4.843480609068724e-06,
+      "loss": 0.658,
+      "step": 782
+    },
+    {
+      "epoch": 0.38358848744641766,
+      "grad_norm": 0.3086536228656769,
+      "learning_rate": 4.843026139725433e-06,
+      "loss": 0.6794,
+      "step": 783
+    },
+    {
+      "epoch": 0.3840783833435395,
+      "grad_norm": 0.31365227699279785,
+      "learning_rate": 4.842571032924778e-06,
+      "loss": 0.6678,
+      "step": 784
+    },
+    {
+      "epoch": 0.38456827924066134,
+      "grad_norm": 0.3227309286594391,
+      "learning_rate": 4.842115288790577e-06,
+      "loss": 0.6379,
+      "step": 785
+    },
+    {
+      "epoch": 0.38505817513778323,
+      "grad_norm": 0.32405832409858704,
+      "learning_rate": 4.8416589074468224e-06,
+      "loss": 0.652,
+      "step": 786
+    },
+    {
+      "epoch": 0.38554807103490507,
+      "grad_norm": 0.31400805711746216,
+      "learning_rate": 4.841201889017681e-06,
+      "loss": 0.6566,
+      "step": 787
+    },
+    {
+      "epoch": 0.38603796693202697,
+      "grad_norm": 0.3314230144023895,
+      "learning_rate": 4.84074423362749e-06,
+      "loss": 0.6414,
+      "step": 788
+    },
+    {
+      "epoch": 0.3865278628291488,
+      "grad_norm": 0.3287082314491272,
+      "learning_rate": 4.840285941400765e-06,
+      "loss": 0.6401,
+      "step": 789
+    },
+    {
+      "epoch": 0.38701775872627064,
+      "grad_norm": 0.31949952244758606,
+      "learning_rate": 4.8398270124621886e-06,
+      "loss": 0.6583,
+      "step": 790
+    },
+    {
+      "epoch": 0.38750765462339254,
+      "grad_norm": 0.3409273028373718,
+      "learning_rate": 4.839367446936621e-06,
+      "loss": 0.651,
+      "step": 791
+    },
+    {
+      "epoch": 0.3879975505205144,
+      "grad_norm": 0.31103092432022095,
+      "learning_rate": 4.838907244949093e-06,
+      "loss": 0.6546,
+      "step": 792
+    },
+    {
+      "epoch": 0.3884874464176363,
+      "grad_norm": 0.31950950622558594,
+      "learning_rate": 4.8384464066248114e-06,
+      "loss": 0.6822,
+      "step": 793
+    },
+    {
+      "epoch": 0.3889773423147581,
+      "grad_norm": 0.3224134147167206,
+      "learning_rate": 4.8379849320891545e-06,
+      "loss": 0.6731,
+      "step": 794
+    },
+    {
+      "epoch": 0.38946723821187995,
+      "grad_norm": 0.31685829162597656,
+      "learning_rate": 4.8375228214676735e-06,
+      "loss": 0.6401,
+      "step": 795
+    },
+    {
+      "epoch": 0.38995713410900185,
+      "grad_norm": 0.3154316544532776,
+      "learning_rate": 4.837060074886092e-06,
+      "loss": 0.6542,
+      "step": 796
+    },
+    {
+      "epoch": 0.3904470300061237,
+      "grad_norm": 0.30656275153160095,
+      "learning_rate": 4.836596692470309e-06,
+      "loss": 0.6541,
+      "step": 797
+    },
+    {
+      "epoch": 0.3909369259032456,
+      "grad_norm": 0.30986401438713074,
+      "learning_rate": 4.836132674346393e-06,
+      "loss": 0.6513,
+      "step": 798
+    },
+    {
+      "epoch": 0.3914268218003674,
+      "grad_norm": 0.3240419328212738,
+      "learning_rate": 4.83566802064059e-06,
+      "loss": 0.6405,
+      "step": 799
+    },
+    {
+      "epoch": 0.39191671769748926,
+      "grad_norm": 0.3198719620704651,
+      "learning_rate": 4.8352027314793145e-06,
+      "loss": 0.633,
+      "step": 800
+    },
+    {
+      "epoch": 0.39240661359461115,
+      "grad_norm": 0.336039662361145,
+      "learning_rate": 4.834736806989156e-06,
+      "loss": 0.6608,
+      "step": 801
+    },
+    {
+      "epoch": 0.392896509491733,
+      "grad_norm": 0.3361132740974426,
+      "learning_rate": 4.834270247296878e-06,
+      "loss": 0.6726,
+      "step": 802
+    },
+    {
+      "epoch": 0.3933864053888549,
+      "grad_norm": 0.3405260741710663,
+      "learning_rate": 4.833803052529414e-06,
+      "loss": 0.6692,
+      "step": 803
+    },
+    {
+      "epoch": 0.3938763012859767,
+      "grad_norm": 0.3248771131038666,
+      "learning_rate": 4.8333352228138725e-06,
+      "loss": 0.6819,
+      "step": 804
+    },
+    {
+      "epoch": 0.39436619718309857,
+      "grad_norm": 0.3115767538547516,
+      "learning_rate": 4.8328667582775336e-06,
+      "loss": 0.6491,
+      "step": 805
+    },
+    {
+      "epoch": 0.39485609308022046,
+      "grad_norm": 0.31870561838150024,
+      "learning_rate": 4.83239765904785e-06,
+      "loss": 0.6663,
+      "step": 806
+    },
+    {
+      "epoch": 0.3953459889773423,
+      "grad_norm": 0.31889578700065613,
+      "learning_rate": 4.8319279252524495e-06,
+      "loss": 0.6528,
+      "step": 807
+    },
+    {
+      "epoch": 0.3958358848744642,
+      "grad_norm": 0.32603761553764343,
+      "learning_rate": 4.831457557019128e-06,
+      "loss": 0.6628,
+      "step": 808
+    },
+    {
+      "epoch": 0.39632578077158603,
+      "grad_norm": 0.3227848708629608,
+      "learning_rate": 4.830986554475859e-06,
+      "loss": 0.6717,
+      "step": 809
+    },
+    {
+      "epoch": 0.3968156766687079,
+      "grad_norm": 0.30794838070869446,
+      "learning_rate": 4.830514917750785e-06,
+      "loss": 0.6596,
+      "step": 810
+    },
+    {
+      "epoch": 0.39730557256582977,
+      "grad_norm": 0.34206315875053406,
+      "learning_rate": 4.830042646972221e-06,
+      "loss": 0.6542,
+      "step": 811
+    },
+    {
+      "epoch": 0.3977954684629516,
+      "grad_norm": 0.31525009870529175,
+      "learning_rate": 4.8295697422686586e-06,
+      "loss": 0.6531,
+      "step": 812
+    },
+    {
+      "epoch": 0.3982853643600735,
+      "grad_norm": 0.34714141488075256,
+      "learning_rate": 4.829096203768757e-06,
+      "loss": 0.6652,
+      "step": 813
+    },
+    {
+      "epoch": 0.39877526025719534,
+      "grad_norm": 0.3229881227016449,
+      "learning_rate": 4.828622031601351e-06,
+      "loss": 0.6705,
+      "step": 814
+    },
+    {
+      "epoch": 0.3992651561543172,
+      "grad_norm": 0.31995874643325806,
+      "learning_rate": 4.828147225895445e-06,
+      "loss": 0.6507,
+      "step": 815
+    },
+    {
+      "epoch": 0.3997550520514391,
+      "grad_norm": 0.326016902923584,
+      "learning_rate": 4.827671786780219e-06,
+      "loss": 0.6786,
+      "step": 816
+    },
+    {
+      "epoch": 0.4002449479485609,
+      "grad_norm": 0.318215548992157,
+      "learning_rate": 4.8271957143850235e-06,
+      "loss": 0.6425,
+      "step": 817
+    },
+    {
+      "epoch": 0.4007348438456828,
+      "grad_norm": 0.3327416777610779,
+      "learning_rate": 4.82671900883938e-06,
+      "loss": 0.6546,
+      "step": 818
+    },
+    {
+      "epoch": 0.40122473974280465,
+      "grad_norm": 0.33632010221481323,
+      "learning_rate": 4.826241670272986e-06,
+      "loss": 0.6582,
+      "step": 819
+    },
+    {
+      "epoch": 0.40171463563992654,
+      "grad_norm": 0.3371349275112152,
+      "learning_rate": 4.825763698815709e-06,
+      "loss": 0.6768,
+      "step": 820
+    },
+    {
+      "epoch": 0.4022045315370484,
+      "grad_norm": 0.332950621843338,
+      "learning_rate": 4.825285094597586e-06,
+      "loss": 0.6613,
+      "step": 821
+    },
+    {
+      "epoch": 0.4026944274341702,
+      "grad_norm": 0.31979843974113464,
+      "learning_rate": 4.824805857748831e-06,
+      "loss": 0.6281,
+      "step": 822
+    },
+    {
+      "epoch": 0.4031843233312921,
+      "grad_norm": 0.33180534839630127,
+      "learning_rate": 4.824325988399828e-06,
+      "loss": 0.6741,
+      "step": 823
+    },
+    {
+      "epoch": 0.40367421922841396,
+      "grad_norm": 0.32684949040412903,
+      "learning_rate": 4.8238454866811326e-06,
+      "loss": 0.6638,
+      "step": 824
+    },
+    {
+      "epoch": 0.40416411512553585,
+      "grad_norm": 0.3229905664920807,
+      "learning_rate": 4.823364352723472e-06,
+      "loss": 0.6545,
+      "step": 825
+    },
+    {
+      "epoch": 0.4046540110226577,
+      "grad_norm": 0.32018545269966125,
+      "learning_rate": 4.822882586657748e-06,
+      "loss": 0.6553,
+      "step": 826
+    },
+    {
+      "epoch": 0.40514390691977953,
+      "grad_norm": 0.32592976093292236,
+      "learning_rate": 4.8224001886150315e-06,
+      "loss": 0.6425,
+      "step": 827
+    },
+    {
+      "epoch": 0.4056338028169014,
+      "grad_norm": 0.3335873484611511,
+      "learning_rate": 4.821917158726568e-06,
+      "loss": 0.6419,
+      "step": 828
+    },
+    {
+      "epoch": 0.40612369871402326,
+      "grad_norm": 0.32976460456848145,
+      "learning_rate": 4.821433497123772e-06,
+      "loss": 0.6545,
+      "step": 829
+    },
+    {
+      "epoch": 0.40661359461114516,
+      "grad_norm": 0.3185843527317047,
+      "learning_rate": 4.820949203938231e-06,
+      "loss": 0.6607,
+      "step": 830
+    },
+    {
+      "epoch": 0.407103490508267,
+      "grad_norm": 0.3265590965747833,
+      "learning_rate": 4.820464279301705e-06,
+      "loss": 0.6779,
+      "step": 831
+    },
+    {
+      "epoch": 0.40759338640538884,
+      "grad_norm": 0.32094770669937134,
+      "learning_rate": 4.819978723346127e-06,
+      "loss": 0.6139,
+      "step": 832
+    },
+    {
+      "epoch": 0.40808328230251073,
+      "grad_norm": 0.34664106369018555,
+      "learning_rate": 4.819492536203598e-06,
+      "loss": 0.6486,
+      "step": 833
+    },
+    {
+      "epoch": 0.40857317819963257,
+      "grad_norm": 0.32802873849868774,
+      "learning_rate": 4.819005718006394e-06,
+      "loss": 0.6516,
+      "step": 834
+    },
+    {
+      "epoch": 0.40906307409675446,
+      "grad_norm": 0.324709415435791,
+      "learning_rate": 4.8185182688869615e-06,
+      "loss": 0.665,
+      "step": 835
+    },
+    {
+      "epoch": 0.4095529699938763,
+      "grad_norm": 0.31625524163246155,
+      "learning_rate": 4.818030188977918e-06,
+      "loss": 0.6571,
+      "step": 836
+    },
+    {
+      "epoch": 0.41004286589099814,
+      "grad_norm": 0.3230063021183014,
+      "learning_rate": 4.817541478412055e-06,
+      "loss": 0.6628,
+      "step": 837
+    },
+    {
+      "epoch": 0.41053276178812004,
+      "grad_norm": 0.3374910354614258,
+      "learning_rate": 4.817052137322334e-06,
+      "loss": 0.6773,
+      "step": 838
+    },
+    {
+      "epoch": 0.4110226576852419,
+      "grad_norm": 0.32604286074638367,
+      "learning_rate": 4.816562165841885e-06,
+      "loss": 0.6695,
+      "step": 839
+    },
+    {
+      "epoch": 0.41151255358236377,
+      "grad_norm": 0.3211324214935303,
+      "learning_rate": 4.816071564104015e-06,
+      "loss": 0.6546,
+      "step": 840
+    },
+    {
+      "epoch": 0.4120024494794856,
+      "grad_norm": 0.3172019124031067,
+      "learning_rate": 4.815580332242199e-06,
+      "loss": 0.661,
+      "step": 841
+    },
+    {
+      "epoch": 0.41249234537660745,
+      "grad_norm": 0.32369232177734375,
+      "learning_rate": 4.8150884703900855e-06,
+      "loss": 0.6643,
+      "step": 842
+    },
+    {
+      "epoch": 0.41298224127372934,
+      "grad_norm": 0.3227887749671936,
+      "learning_rate": 4.814595978681493e-06,
+      "loss": 0.6353,
+      "step": 843
+    },
+    {
+      "epoch": 0.4134721371708512,
+      "grad_norm": 0.32307952642440796,
+      "learning_rate": 4.814102857250411e-06,
+      "loss": 0.6701,
+      "step": 844
+    },
+    {
+      "epoch": 0.4139620330679731,
+      "grad_norm": 0.335151731967926,
+      "learning_rate": 4.813609106231002e-06,
+      "loss": 0.6335,
+      "step": 845
+    },
+    {
+      "epoch": 0.4144519289650949,
+      "grad_norm": 0.32878342270851135,
+      "learning_rate": 4.813114725757597e-06,
+      "loss": 0.6643,
+      "step": 846
+    },
+    {
+      "epoch": 0.41494182486221676,
+      "grad_norm": 0.3271647095680237,
+      "learning_rate": 4.812619715964702e-06,
+      "loss": 0.6337,
+      "step": 847
+    },
+    {
+      "epoch": 0.41543172075933865,
+      "grad_norm": 0.3339027166366577,
+      "learning_rate": 4.8121240769869905e-06,
+      "loss": 0.6458,
+      "step": 848
+    },
+    {
+      "epoch": 0.4159216166564605,
+      "grad_norm": 0.33395877480506897,
+      "learning_rate": 4.8116278089593104e-06,
+      "loss": 0.6487,
+      "step": 849
+    },
+    {
+      "epoch": 0.4164115125535824,
+      "grad_norm": 0.34332144260406494,
+      "learning_rate": 4.811130912016679e-06,
+      "loss": 0.6517,
+      "step": 850
+    },
+    {
+      "epoch": 0.4169014084507042,
+      "grad_norm": 0.37615567445755005,
+      "learning_rate": 4.810633386294285e-06,
+      "loss": 0.6458,
+      "step": 851
+    },
+    {
+      "epoch": 0.41739130434782606,
+      "grad_norm": 0.3249885141849518,
+      "learning_rate": 4.810135231927486e-06,
+      "loss": 0.6688,
+      "step": 852
+    },
+    {
+      "epoch": 0.41788120024494796,
+      "grad_norm": 0.32357704639434814,
+      "learning_rate": 4.8096364490518164e-06,
+      "loss": 0.6505,
+      "step": 853
+    },
+    {
+      "epoch": 0.4183710961420698,
+      "grad_norm": 0.3268754482269287,
+      "learning_rate": 4.809137037802975e-06,
+      "loss": 0.6305,
+      "step": 854
+    },
+    {
+      "epoch": 0.4188609920391917,
+      "grad_norm": 0.32692062854766846,
+      "learning_rate": 4.808636998316837e-06,
+      "loss": 0.6655,
+      "step": 855
+    },
+    {
+      "epoch": 0.41935088793631353,
+      "grad_norm": 0.33697402477264404,
+      "learning_rate": 4.808136330729444e-06,
+      "loss": 0.6642,
+      "step": 856
+    },
+    {
+      "epoch": 0.41984078383343537,
+      "grad_norm": 0.3350159227848053,
+      "learning_rate": 4.807635035177012e-06,
+      "loss": 0.6565,
+      "step": 857
+    },
+    {
+      "epoch": 0.42033067973055727,
+      "grad_norm": 0.319500595331192,
+      "learning_rate": 4.807133111795925e-06,
+      "loss": 0.6227,
+      "step": 858
+    },
+    {
+      "epoch": 0.4208205756276791,
+      "grad_norm": 0.31828346848487854,
+      "learning_rate": 4.806630560722741e-06,
+      "loss": 0.6408,
+      "step": 859
+    },
+    {
+      "epoch": 0.421310471524801,
+      "grad_norm": 0.3408709466457367,
+      "learning_rate": 4.806127382094184e-06,
+      "loss": 0.6513,
+      "step": 860
+    },
+    {
+      "epoch": 0.42180036742192284,
+      "grad_norm": 0.3314608335494995,
+      "learning_rate": 4.805623576047155e-06,
+      "loss": 0.6601,
+      "step": 861
+    },
+    {
+      "epoch": 0.4222902633190447,
+      "grad_norm": 0.31937554478645325,
+      "learning_rate": 4.805119142718722e-06,
+      "loss": 0.6375,
+      "step": 862
+    },
+    {
+      "epoch": 0.4227801592161666,
+      "grad_norm": 0.33292335271835327,
+      "learning_rate": 4.804614082246121e-06,
+      "loss": 0.6677,
+      "step": 863
+    },
+    {
+      "epoch": 0.4232700551132884,
+      "grad_norm": 0.33356305956840515,
+      "learning_rate": 4.804108394766766e-06,
+      "loss": 0.6489,
+      "step": 864
+    },
+    {
+      "epoch": 0.4237599510104103,
+      "grad_norm": 0.33318352699279785,
+      "learning_rate": 4.8036020804182335e-06,
+      "loss": 0.6214,
+      "step": 865
+    },
+    {
+      "epoch": 0.42424984690753215,
+      "grad_norm": 0.329732209444046,
+      "learning_rate": 4.803095139338276e-06,
+      "loss": 0.6433,
+      "step": 866
+    },
+    {
+      "epoch": 0.424739742804654,
+      "grad_norm": 0.3253389894962311,
+      "learning_rate": 4.802587571664815e-06,
+      "loss": 0.6813,
+      "step": 867
+    },
+    {
+      "epoch": 0.4252296387017759,
+      "grad_norm": 0.3183434009552002,
+      "learning_rate": 4.802079377535943e-06,
+      "loss": 0.656,
+      "step": 868
+    },
+    {
+      "epoch": 0.4257195345988977,
+      "grad_norm": 0.3274723291397095,
+      "learning_rate": 4.801570557089921e-06,
+      "loss": 0.5987,
+      "step": 869
+    },
+    {
+      "epoch": 0.4262094304960196,
+      "grad_norm": 0.3257862627506256,
+      "learning_rate": 4.8010611104651814e-06,
+      "loss": 0.6481,
+      "step": 870
+    },
+    {
+      "epoch": 0.42669932639314145,
+      "grad_norm": 0.3261187970638275,
+      "learning_rate": 4.800551037800329e-06,
+      "loss": 0.6453,
+      "step": 871
+    },
+    {
+      "epoch": 0.4271892222902633,
+      "grad_norm": 0.3393871784210205,
+      "learning_rate": 4.800040339234134e-06,
+      "loss": 0.6306,
+      "step": 872
+    },
+    {
+      "epoch": 0.4276791181873852,
+      "grad_norm": 0.3195154368877411,
+      "learning_rate": 4.799529014905544e-06,
+      "loss": 0.6537,
+      "step": 873
+    },
+    {
+      "epoch": 0.428169014084507,
+      "grad_norm": 0.3519408702850342,
+      "learning_rate": 4.79901706495367e-06,
+      "loss": 0.6727,
+      "step": 874
+    },
+    {
+      "epoch": 0.4286589099816289,
+      "grad_norm": 0.33387088775634766,
+      "learning_rate": 4.7985044895177976e-06,
+      "loss": 0.6556,
+      "step": 875
+    },
+    {
+      "epoch": 0.42914880587875076,
+      "grad_norm": 0.32974541187286377,
+      "learning_rate": 4.7979912887373805e-06,
+      "loss": 0.6579,
+      "step": 876
+    },
+    {
+      "epoch": 0.4296387017758726,
+      "grad_norm": 0.32956287264823914,
+      "learning_rate": 4.7974774627520424e-06,
+      "loss": 0.6522,
+      "step": 877
+    },
+    {
+      "epoch": 0.4301285976729945,
+      "grad_norm": 0.31489697098731995,
+      "learning_rate": 4.7969630117015785e-06,
+      "loss": 0.6613,
+      "step": 878
+    },
+    {
+      "epoch": 0.43061849357011633,
+      "grad_norm": 0.33859875798225403,
+      "learning_rate": 4.796447935725954e-06,
+      "loss": 0.6546,
+      "step": 879
+    },
+    {
+      "epoch": 0.43110838946723823,
+      "grad_norm": 0.33555036783218384,
+      "learning_rate": 4.795932234965302e-06,
+      "loss": 0.6083,
+      "step": 880
+    },
+    {
+      "epoch": 0.43159828536436007,
+      "grad_norm": 0.3280044496059418,
+      "learning_rate": 4.795415909559929e-06,
+      "loss": 0.6721,
+      "step": 881
+    },
+    {
+      "epoch": 0.43208818126148196,
+      "grad_norm": 0.331339567899704,
+      "learning_rate": 4.7948989596503085e-06,
+      "loss": 0.6529,
+      "step": 882
+    },
+    {
+      "epoch": 0.4325780771586038,
+      "grad_norm": 0.3267557919025421,
+      "learning_rate": 4.794381385377084e-06,
+      "loss": 0.6483,
+      "step": 883
+    },
+    {
+      "epoch": 0.43306797305572564,
+      "grad_norm": 0.3242381513118744,
+      "learning_rate": 4.79386318688107e-06,
+      "loss": 0.652,
+      "step": 884
+    },
+    {
+      "epoch": 0.43355786895284754,
+      "grad_norm": 0.35081592202186584,
+      "learning_rate": 4.793344364303252e-06,
+      "loss": 0.6859,
+      "step": 885
+    },
+    {
+      "epoch": 0.4340477648499694,
+      "grad_norm": 0.33770686388015747,
+      "learning_rate": 4.792824917784783e-06,
+      "loss": 0.6315,
+      "step": 886
+    },
+    {
+      "epoch": 0.43453766074709127,
+      "grad_norm": 0.33075547218322754,
+      "learning_rate": 4.792304847466986e-06,
+      "loss": 0.6512,
+      "step": 887
+    },
+    {
+      "epoch": 0.4350275566442131,
+      "grad_norm": 0.3429637551307678,
+      "learning_rate": 4.791784153491355e-06,
+      "loss": 0.6803,
+      "step": 888
+    },
+    {
+      "epoch": 0.43551745254133495,
+      "grad_norm": 0.3312215209007263,
+      "learning_rate": 4.791262835999554e-06,
+      "loss": 0.6292,
+      "step": 889
+    },
+    {
+      "epoch": 0.43600734843845684,
+      "grad_norm": 0.3361411690711975,
+      "learning_rate": 4.790740895133415e-06,
+      "loss": 0.6537,
+      "step": 890
+    },
+    {
+      "epoch": 0.4364972443355787,
+      "grad_norm": 0.320117324590683,
+      "learning_rate": 4.790218331034939e-06,
+      "loss": 0.669,
+      "step": 891
+    },
+    {
+      "epoch": 0.4369871402327006,
+      "grad_norm": 0.3243316113948822,
+      "learning_rate": 4.7896951438463e-06,
+      "loss": 0.6205,
+      "step": 892
+    },
+    {
+      "epoch": 0.4374770361298224,
+      "grad_norm": 0.3368995785713196,
+      "learning_rate": 4.789171333709838e-06,
+      "loss": 0.6719,
+      "step": 893
+    },
+    {
+      "epoch": 0.43796693202694426,
+      "grad_norm": 0.3193654716014862,
+      "learning_rate": 4.788646900768064e-06,
+      "loss": 0.657,
+      "step": 894
+    },
+    {
+      "epoch": 0.43845682792406615,
+      "grad_norm": 0.3282233476638794,
+      "learning_rate": 4.788121845163658e-06,
+      "loss": 0.6526,
+      "step": 895
+    },
+    {
+      "epoch": 0.438946723821188,
+      "grad_norm": 0.3279890716075897,
+      "learning_rate": 4.787596167039471e-06,
+      "loss": 0.6384,
+      "step": 896
+    },
+    {
+      "epoch": 0.4394366197183099,
+      "grad_norm": 0.3300457298755646,
+      "learning_rate": 4.787069866538521e-06,
+      "loss": 0.6702,
+      "step": 897
+    },
+    {
+      "epoch": 0.4399265156154317,
+      "grad_norm": 0.33194148540496826,
+      "learning_rate": 4.7865429438039955e-06,
+      "loss": 0.6595,
+      "step": 898
+    },
+    {
+      "epoch": 0.44041641151255356,
+      "grad_norm": 0.34703534841537476,
+      "learning_rate": 4.786015398979254e-06,
+      "loss": 0.6537,
+      "step": 899
+    },
+    {
+      "epoch": 0.44090630740967546,
+      "grad_norm": 0.3270828425884247,
+      "learning_rate": 4.7854872322078225e-06,
+      "loss": 0.6616,
+      "step": 900
+    },
+    {
+      "epoch": 0.4413962033067973,
+      "grad_norm": 0.3177931606769562,
+      "learning_rate": 4.784958443633398e-06,
+      "loss": 0.6359,
+      "step": 901
+    },
+    {
+      "epoch": 0.4418860992039192,
+      "grad_norm": 0.33886033296585083,
+      "learning_rate": 4.784429033399844e-06,
+      "loss": 0.6472,
+      "step": 902
+    },
+    {
+      "epoch": 0.44237599510104103,
+      "grad_norm": 0.3440721929073334,
+      "learning_rate": 4.783899001651196e-06,
+      "loss": 0.6417,
+      "step": 903
+    },
+    {
+      "epoch": 0.44286589099816287,
+      "grad_norm": 0.3315575420856476,
+      "learning_rate": 4.783368348531657e-06,
+      "loss": 0.6478,
+      "step": 904
+    },
+    {
+      "epoch": 0.44335578689528476,
+      "grad_norm": 0.32453784346580505,
+      "learning_rate": 4.782837074185602e-06,
+      "loss": 0.6605,
+      "step": 905
+    },
+    {
+      "epoch": 0.4438456827924066,
+      "grad_norm": 0.3341559171676636,
+      "learning_rate": 4.782305178757569e-06,
+      "loss": 0.6778,
+      "step": 906
+    },
+    {
+      "epoch": 0.4443355786895285,
+      "grad_norm": 0.338220477104187,
+      "learning_rate": 4.781772662392272e-06,
+      "loss": 0.6609,
+      "step": 907
+    },
+    {
+      "epoch": 0.44482547458665034,
+      "grad_norm": 0.34984090924263,
+      "learning_rate": 4.781239525234588e-06,
+      "loss": 0.668,
+      "step": 908
+    },
+    {
+      "epoch": 0.4453153704837722,
+      "grad_norm": 0.31468555331230164,
+      "learning_rate": 4.780705767429565e-06,
+      "loss": 0.5993,
+      "step": 909
+    },
+    {
+      "epoch": 0.44580526638089407,
+      "grad_norm": 0.327556312084198,
+      "learning_rate": 4.780171389122423e-06,
+      "loss": 0.6509,
+      "step": 910
+    },
+    {
+      "epoch": 0.4462951622780159,
+      "grad_norm": 0.3160576820373535,
+      "learning_rate": 4.779636390458546e-06,
+      "loss": 0.6445,
+      "step": 911
+    },
+    {
+      "epoch": 0.4467850581751378,
+      "grad_norm": 0.3446383476257324,
+      "learning_rate": 4.7791007715834906e-06,
+      "loss": 0.64,
+      "step": 912
+    },
+    {
+      "epoch": 0.44727495407225965,
+      "grad_norm": 0.32624417543411255,
+      "learning_rate": 4.778564532642978e-06,
+      "loss": 0.6636,
+      "step": 913
+    },
+    {
+      "epoch": 0.4477648499693815,
+      "grad_norm": 0.3289485275745392,
+      "learning_rate": 4.778027673782901e-06,
+      "loss": 0.656,
+      "step": 914
+    },
+    {
+      "epoch": 0.4482547458665034,
+      "grad_norm": 0.3322214186191559,
+      "learning_rate": 4.777490195149322e-06,
+      "loss": 0.6997,
+      "step": 915
+    },
+    {
+      "epoch": 0.4487446417636252,
+      "grad_norm": 0.3421451449394226,
+      "learning_rate": 4.776952096888471e-06,
+      "loss": 0.6594,
+      "step": 916
+    },
+    {
+      "epoch": 0.4492345376607471,
+      "grad_norm": 0.3431428372859955,
+      "learning_rate": 4.7764133791467434e-06,
+      "loss": 0.651,
+      "step": 917
+    },
+    {
+      "epoch": 0.44972443355786895,
+      "grad_norm": 0.32274821400642395,
+      "learning_rate": 4.7758740420707086e-06,
+      "loss": 0.6277,
+      "step": 918
+    },
+    {
+      "epoch": 0.4502143294549908,
+      "grad_norm": 0.3264138102531433,
+      "learning_rate": 4.7753340858071e-06,
+      "loss": 0.6352,
+      "step": 919
+    },
+    {
+      "epoch": 0.4507042253521127,
+      "grad_norm": 0.3209553062915802,
+      "learning_rate": 4.7747935105028224e-06,
+      "loss": 0.6557,
+      "step": 920
+    },
+    {
+      "epoch": 0.4511941212492345,
+      "grad_norm": 0.3410312831401825,
+      "learning_rate": 4.774252316304947e-06,
+      "loss": 0.6751,
+      "step": 921
+    },
+    {
+      "epoch": 0.4516840171463564,
+      "grad_norm": 0.35638895630836487,
+      "learning_rate": 4.773710503360714e-06,
+      "loss": 0.6669,
+      "step": 922
+    },
+    {
+      "epoch": 0.45217391304347826,
+      "grad_norm": 0.34015554189682007,
+      "learning_rate": 4.773168071817534e-06,
+      "loss": 0.6505,
+      "step": 923
+    },
+    {
+      "epoch": 0.4526638089406001,
+      "grad_norm": 0.32471373677253723,
+      "learning_rate": 4.772625021822983e-06,
+      "loss": 0.6682,
+      "step": 924
+    },
+    {
+      "epoch": 0.453153704837722,
+      "grad_norm": 0.31346777081489563,
+      "learning_rate": 4.7720813535248056e-06,
+      "loss": 0.643,
+      "step": 925
+    },
+    {
+      "epoch": 0.45364360073484383,
+      "grad_norm": 0.33824825286865234,
+      "learning_rate": 4.771537067070917e-06,
+      "loss": 0.6633,
+      "step": 926
+    },
+    {
+      "epoch": 0.4541334966319657,
+      "grad_norm": 0.3258407711982727,
+      "learning_rate": 4.770992162609397e-06,
+      "loss": 0.6483,
+      "step": 927
+    },
+    {
+      "epoch": 0.45462339252908757,
+      "grad_norm": 0.32574284076690674,
+      "learning_rate": 4.770446640288497e-06,
+      "loss": 0.6399,
+      "step": 928
+    },
+    {
+      "epoch": 0.4551132884262094,
+      "grad_norm": 0.31942424178123474,
+      "learning_rate": 4.769900500256634e-06,
+      "loss": 0.6746,
+      "step": 929
+    },
+    {
+      "epoch": 0.4556031843233313,
+      "grad_norm": 0.3261716365814209,
+      "learning_rate": 4.769353742662395e-06,
+      "loss": 0.6626,
+      "step": 930
+    },
+    {
+      "epoch": 0.45609308022045314,
+      "grad_norm": 0.32581284642219543,
+      "learning_rate": 4.768806367654534e-06,
+      "loss": 0.6647,
+      "step": 931
+    },
+    {
+      "epoch": 0.45658297611757503,
+      "grad_norm": 0.32622581720352173,
+      "learning_rate": 4.768258375381971e-06,
+      "loss": 0.6535,
+      "step": 932
+    },
+    {
+      "epoch": 0.4570728720146969,
+      "grad_norm": 0.3114738166332245,
+      "learning_rate": 4.767709765993799e-06,
+      "loss": 0.6342,
+      "step": 933
+    },
+    {
+      "epoch": 0.4575627679118187,
+      "grad_norm": 0.33273571729660034,
+      "learning_rate": 4.7671605396392735e-06,
+      "loss": 0.6558,
+      "step": 934
+    },
+    {
+      "epoch": 0.4580526638089406,
+      "grad_norm": 0.3257521688938141,
+      "learning_rate": 4.766610696467822e-06,
+      "loss": 0.6407,
+      "step": 935
+    },
+    {
+      "epoch": 0.45854255970606245,
+      "grad_norm": 0.3364790678024292,
+      "learning_rate": 4.766060236629037e-06,
+      "loss": 0.6393,
+      "step": 936
+    },
+    {
+      "epoch": 0.45903245560318434,
+      "grad_norm": 0.33170202374458313,
+      "learning_rate": 4.765509160272681e-06,
+      "loss": 0.6292,
+      "step": 937
+    },
+    {
+      "epoch": 0.4595223515003062,
+      "grad_norm": 0.31647032499313354,
+      "learning_rate": 4.764957467548681e-06,
+      "loss": 0.634,
+      "step": 938
+    },
+    {
+      "epoch": 0.460012247397428,
+      "grad_norm": 0.32740557193756104,
+      "learning_rate": 4.7644051586071345e-06,
+      "loss": 0.6115,
+      "step": 939
+    },
+    {
+      "epoch": 0.4605021432945499,
+      "grad_norm": 0.32127001881599426,
+      "learning_rate": 4.763852233598307e-06,
+      "loss": 0.6474,
+      "step": 940
+    },
+    {
+      "epoch": 0.46099203919167175,
+      "grad_norm": 0.3447674512863159,
+      "learning_rate": 4.76329869267263e-06,
+      "loss": 0.6845,
+      "step": 941
+    },
+    {
+      "epoch": 0.46148193508879365,
+      "grad_norm": 0.32889634370803833,
+      "learning_rate": 4.762744535980702e-06,
+      "loss": 0.6364,
+      "step": 942
+    },
+    {
+      "epoch": 0.4619718309859155,
+      "grad_norm": 0.34127506613731384,
+      "learning_rate": 4.762189763673291e-06,
+      "loss": 0.6549,
+      "step": 943
+    },
+    {
+      "epoch": 0.4624617268830373,
+      "grad_norm": 0.33296868205070496,
+      "learning_rate": 4.761634375901333e-06,
+      "loss": 0.6462,
+      "step": 944
+    },
+    {
+      "epoch": 0.4629516227801592,
+      "grad_norm": 0.3282979428768158,
+      "learning_rate": 4.761078372815927e-06,
+      "loss": 0.6408,
+      "step": 945
+    },
+    {
+      "epoch": 0.46344151867728106,
+      "grad_norm": 0.3336234986782074,
+      "learning_rate": 4.760521754568345e-06,
+      "loss": 0.6441,
+      "step": 946
+    },
+    {
+      "epoch": 0.46393141457440296,
+      "grad_norm": 0.3332432806491852,
+      "learning_rate": 4.759964521310021e-06,
+      "loss": 0.6592,
+      "step": 947
+    },
+    {
+      "epoch": 0.4644213104715248,
+      "grad_norm": 0.3263160288333893,
+      "learning_rate": 4.7594066731925624e-06,
+      "loss": 0.6621,
+      "step": 948
+    },
+    {
+      "epoch": 0.4649112063686467,
+      "grad_norm": 0.3333407938480377,
+      "learning_rate": 4.758848210367739e-06,
+      "loss": 0.6374,
+      "step": 949
+    },
+    {
+      "epoch": 0.46540110226576853,
+      "grad_norm": 0.3290969729423523,
+      "learning_rate": 4.7582891329874894e-06,
+      "loss": 0.6531,
+      "step": 950
+    },
+    {
+      "epoch": 0.46589099816289037,
+      "grad_norm": 0.32899218797683716,
+      "learning_rate": 4.7577294412039195e-06,
+      "loss": 0.6478,
+      "step": 951
+    },
+    {
+      "epoch": 0.46638089406001226,
+      "grad_norm": 0.32041114568710327,
+      "learning_rate": 4.757169135169304e-06,
+      "loss": 0.6581,
+      "step": 952
+    },
+    {
+      "epoch": 0.4668707899571341,
+      "grad_norm": 0.3273862302303314,
+      "learning_rate": 4.7566082150360795e-06,
+      "loss": 0.6173,
+      "step": 953
+    },
+    {
+      "epoch": 0.467360685854256,
+      "grad_norm": 0.3293192684650421,
+      "learning_rate": 4.7560466809568565e-06,
+      "loss": 0.6345,
+      "step": 954
+    },
+    {
+      "epoch": 0.46785058175137784,
+      "grad_norm": 0.3363111913204193,
+      "learning_rate": 4.755484533084407e-06,
+      "loss": 0.6461,
+      "step": 955
+    },
+    {
+      "epoch": 0.4683404776484997,
+      "grad_norm": 0.3306375741958618,
+      "learning_rate": 4.754921771571674e-06,
+      "loss": 0.6364,
+      "step": 956
+    },
+    {
+      "epoch": 0.46883037354562157,
+      "grad_norm": 0.3344152271747589,
+      "learning_rate": 4.754358396571764e-06,
+      "loss": 0.6205,
+      "step": 957
+    },
+    {
+      "epoch": 0.4693202694427434,
+      "grad_norm": 0.32841694355010986,
+      "learning_rate": 4.753794408237954e-06,
+      "loss": 0.662,
+      "step": 958
+    },
+    {
+      "epoch": 0.4698101653398653,
+      "grad_norm": 0.3455883860588074,
+      "learning_rate": 4.753229806723684e-06,
+      "loss": 0.6537,
+      "step": 959
+    },
+    {
+      "epoch": 0.47030006123698714,
+      "grad_norm": 0.3336228132247925,
+      "learning_rate": 4.752664592182564e-06,
+      "loss": 0.6564,
+      "step": 960
+    },
+    {
+      "epoch": 0.470789957134109,
+      "grad_norm": 0.32779762148857117,
+      "learning_rate": 4.75209876476837e-06,
+      "loss": 0.6124,
+      "step": 961
+    },
+    {
+      "epoch": 0.4712798530312309,
+      "grad_norm": 0.3376416862010956,
+      "learning_rate": 4.751532324635043e-06,
+      "loss": 0.6485,
+      "step": 962
+    },
+    {
+      "epoch": 0.4717697489283527,
+      "grad_norm": 0.33599647879600525,
+      "learning_rate": 4.750965271936692e-06,
+      "loss": 0.6489,
+      "step": 963
+    },
+    {
+      "epoch": 0.4722596448254746,
+      "grad_norm": 0.33952006697654724,
+      "learning_rate": 4.750397606827594e-06,
+      "loss": 0.6501,
+      "step": 964
+    },
+    {
+      "epoch": 0.47274954072259645,
+      "grad_norm": 0.3457185626029968,
+      "learning_rate": 4.74982932946219e-06,
+      "loss": 0.6554,
+      "step": 965
+    },
+    {
+      "epoch": 0.4732394366197183,
+      "grad_norm": 0.3317156732082367,
+      "learning_rate": 4.749260439995089e-06,
+      "loss": 0.6445,
+      "step": 966
+    },
+    {
+      "epoch": 0.4737293325168402,
+      "grad_norm": 0.32509130239486694,
+      "learning_rate": 4.748690938581068e-06,
+      "loss": 0.6303,
+      "step": 967
+    },
+    {
+      "epoch": 0.474219228413962,
+      "grad_norm": 0.3271726369857788,
+      "learning_rate": 4.748120825375067e-06,
+      "loss": 0.6544,
+      "step": 968
+    },
+    {
+      "epoch": 0.4747091243110839,
+      "grad_norm": 0.32812872529029846,
+      "learning_rate": 4.747550100532195e-06,
+      "loss": 0.6249,
+      "step": 969
+    },
+    {
+      "epoch": 0.47519902020820576,
+      "grad_norm": 0.33634963631629944,
+      "learning_rate": 4.746978764207727e-06,
+      "loss": 0.6416,
+      "step": 970
+    },
+    {
+      "epoch": 0.4756889161053276,
+      "grad_norm": 0.33176323771476746,
+      "learning_rate": 4.746406816557103e-06,
+      "loss": 0.6538,
+      "step": 971
+    },
+    {
+      "epoch": 0.4761788120024495,
+      "grad_norm": 0.3377355933189392,
+      "learning_rate": 4.7458342577359315e-06,
+      "loss": 0.6325,
+      "step": 972
+    },
+    {
+      "epoch": 0.47666870789957133,
+      "grad_norm": 0.3324216902256012,
+      "learning_rate": 4.745261087899987e-06,
+      "loss": 0.6774,
+      "step": 973
+    },
+    {
+      "epoch": 0.4771586037966932,
+      "grad_norm": 0.33282217383384705,
+      "learning_rate": 4.744687307205207e-06,
+      "loss": 0.6317,
+      "step": 974
+    },
+    {
+      "epoch": 0.47764849969381507,
+      "grad_norm": 0.3376564383506775,
+      "learning_rate": 4.7441129158077e-06,
+      "loss": 0.6658,
+      "step": 975
+    },
+    {
+      "epoch": 0.4781383955909369,
+      "grad_norm": 0.3456125855445862,
+      "learning_rate": 4.743537913863736e-06,
+      "loss": 0.6802,
+      "step": 976
+    },
+    {
+      "epoch": 0.4786282914880588,
+      "grad_norm": 0.3268921971321106,
+      "learning_rate": 4.742962301529755e-06,
+      "loss": 0.6463,
+      "step": 977
+    },
+    {
+      "epoch": 0.47911818738518064,
+      "grad_norm": 0.34568679332733154,
+      "learning_rate": 4.742386078962362e-06,
+      "loss": 0.6507,
+      "step": 978
+    },
+    {
+      "epoch": 0.47960808328230253,
+      "grad_norm": 0.3478900194168091,
+      "learning_rate": 4.741809246318326e-06,
+      "loss": 0.6582,
+      "step": 979
+    },
+    {
+      "epoch": 0.4800979791794244,
+      "grad_norm": 0.3300599753856659,
+      "learning_rate": 4.741231803754583e-06,
+      "loss": 0.6444,
+      "step": 980
+    },
+    {
+      "epoch": 0.4805878750765462,
+      "grad_norm": 0.3292956054210663,
+      "learning_rate": 4.740653751428238e-06,
+      "loss": 0.635,
+      "step": 981
+    },
+    {
+      "epoch": 0.4810777709736681,
+      "grad_norm": 0.3479529619216919,
+      "learning_rate": 4.740075089496557e-06,
+      "loss": 0.6671,
+      "step": 982
+    },
+    {
+      "epoch": 0.48156766687078995,
+      "grad_norm": 0.33764973282814026,
+      "learning_rate": 4.739495818116974e-06,
+      "loss": 0.6346,
+      "step": 983
+    },
+    {
+      "epoch": 0.48205756276791184,
+      "grad_norm": 0.3391423225402832,
+      "learning_rate": 4.738915937447091e-06,
+      "loss": 0.6376,
+      "step": 984
+    },
+    {
+      "epoch": 0.4825474586650337,
+      "grad_norm": 0.3254561126232147,
+      "learning_rate": 4.738335447644673e-06,
+      "loss": 0.6551,
+      "step": 985
+    },
+    {
+      "epoch": 0.4830373545621555,
+      "grad_norm": 0.32766976952552795,
+      "learning_rate": 4.7377543488676495e-06,
+      "loss": 0.6372,
+      "step": 986
+    },
+    {
+      "epoch": 0.4835272504592774,
+      "grad_norm": 0.340474009513855,
+      "learning_rate": 4.7371726412741205e-06,
+      "loss": 0.6697,
+      "step": 987
+    },
+    {
+      "epoch": 0.48401714635639925,
+      "grad_norm": 0.3602207601070404,
+      "learning_rate": 4.736590325022346e-06,
+      "loss": 0.6462,
+      "step": 988
+    },
+    {
+      "epoch": 0.48450704225352115,
+      "grad_norm": 0.332263708114624,
+      "learning_rate": 4.7360074002707556e-06,
+      "loss": 0.6545,
+      "step": 989
+    },
+    {
+      "epoch": 0.484996938150643,
+      "grad_norm": 0.3312827944755554,
+      "learning_rate": 4.735423867177944e-06,
+      "loss": 0.6291,
+      "step": 990
+    },
+    {
+      "epoch": 0.4854868340477648,
+      "grad_norm": 0.35063663125038147,
+      "learning_rate": 4.73483972590267e-06,
+      "loss": 0.6423,
+      "step": 991
+    },
+    {
+      "epoch": 0.4859767299448867,
+      "grad_norm": 0.3246268033981323,
+      "learning_rate": 4.734254976603858e-06,
+      "loss": 0.653,
+      "step": 992
+    },
+    {
+      "epoch": 0.48646662584200856,
+      "grad_norm": 0.33790072798728943,
+      "learning_rate": 4.7336696194405995e-06,
+      "loss": 0.6448,
+      "step": 993
+    },
+    {
+      "epoch": 0.48695652173913045,
+      "grad_norm": 0.3333311676979065,
+      "learning_rate": 4.733083654572148e-06,
+      "loss": 0.6458,
+      "step": 994
+    },
+    {
+      "epoch": 0.4874464176362523,
+      "grad_norm": 0.3357294499874115,
+      "learning_rate": 4.732497082157928e-06,
+      "loss": 0.6462,
+      "step": 995
+    },
+    {
+      "epoch": 0.48793631353337413,
+      "grad_norm": 0.3324768841266632,
+      "learning_rate": 4.731909902357522e-06,
+      "loss": 0.6442,
+      "step": 996
+    },
+    {
+      "epoch": 0.48842620943049603,
+      "grad_norm": 0.3282735049724579,
+      "learning_rate": 4.731322115330683e-06,
+      "loss": 0.5999,
+      "step": 997
+    },
+    {
+      "epoch": 0.48891610532761787,
+      "grad_norm": 0.3322250247001648,
+      "learning_rate": 4.730733721237329e-06,
+      "loss": 0.6564,
+      "step": 998
+    },
+    {
+      "epoch": 0.48940600122473976,
+      "grad_norm": 0.33998483419418335,
+      "learning_rate": 4.7301447202375405e-06,
+      "loss": 0.6403,
+      "step": 999
+    },
+    {
+      "epoch": 0.4898958971218616,
+      "grad_norm": 0.3261421024799347,
+      "learning_rate": 4.729555112491566e-06,
+      "loss": 0.6633,
+      "step": 1000
+    },
+    {
+      "epoch": 0.49038579301898344,
+      "grad_norm": 0.3308633267879486,
+      "learning_rate": 4.728964898159816e-06,
+      "loss": 0.6439,
+      "step": 1001
+    },
+    {
+      "epoch": 0.49087568891610533,
+      "grad_norm": 0.3286653459072113,
+      "learning_rate": 4.728374077402868e-06,
+      "loss": 0.6699,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4913655848132272,
+      "grad_norm": 0.34002241492271423,
+      "learning_rate": 4.727782650381464e-06,
+      "loss": 0.6804,
+      "step": 1003
+    },
+    {
+      "epoch": 0.49185548071034907,
+      "grad_norm": 0.3460935056209564,
+      "learning_rate": 4.727190617256513e-06,
+      "loss": 0.6293,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4923453766074709,
+      "grad_norm": 0.3384420871734619,
+      "learning_rate": 4.726597978189085e-06,
+      "loss": 0.6425,
+      "step": 1005
+    },
+    {
+      "epoch": 0.49283527250459275,
+      "grad_norm": 0.3322122097015381,
+      "learning_rate": 4.726004733340417e-06,
+      "loss": 0.624,
+      "step": 1006
+    },
+    {
+      "epoch": 0.49332516840171464,
+      "grad_norm": 0.33144181966781616,
+      "learning_rate": 4.725410882871911e-06,
+      "loss": 0.6088,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4938150642988365,
+      "grad_norm": 0.3404276967048645,
+      "learning_rate": 4.724816426945134e-06,
+      "loss": 0.6394,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4943049601959584,
+      "grad_norm": 0.3359997570514679,
+      "learning_rate": 4.724221365721816e-06,
+      "loss": 0.6123,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4947948560930802,
+      "grad_norm": 0.3339918255805969,
+      "learning_rate": 4.723625699363854e-06,
+      "loss": 0.6568,
+      "step": 1010
+    },
+    {
+      "epoch": 0.49528475199020205,
+      "grad_norm": 0.3534146547317505,
+      "learning_rate": 4.723029428033308e-06,
+      "loss": 0.6777,
+      "step": 1011
+    },
+    {
+      "epoch": 0.49577464788732395,
+      "grad_norm": 0.39140719175338745,
+      "learning_rate": 4.722432551892402e-06,
+      "loss": 0.6508,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4962645437844458,
+      "grad_norm": 0.3383652865886688,
+      "learning_rate": 4.721835071103527e-06,
+      "loss": 0.6109,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4967544396815677,
+      "grad_norm": 0.33597037196159363,
+      "learning_rate": 4.721236985829237e-06,
+      "loss": 0.6525,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4972443355786895,
+      "grad_norm": 0.34626680612564087,
+      "learning_rate": 4.720638296232251e-06,
+      "loss": 0.6479,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4977342314758114,
+      "grad_norm": 0.3565669357776642,
+      "learning_rate": 4.720039002475449e-06,
+      "loss": 0.6626,
+      "step": 1016
+    },
+    {
+      "epoch": 0.49822412737293326,
+      "grad_norm": 0.3603387475013733,
+      "learning_rate": 4.719439104721883e-06,
+      "loss": 0.635,
+      "step": 1017
+    },
+    {
+      "epoch": 0.4987140232700551,
+      "grad_norm": 0.34352055191993713,
+      "learning_rate": 4.71883860313476e-06,
+      "loss": 0.6377,
+      "step": 1018
+    },
+    {
+      "epoch": 0.499203919167177,
+      "grad_norm": 0.3388681411743164,
+      "learning_rate": 4.71823749787746e-06,
+      "loss": 0.6413,
+      "step": 1019
+    },
+    {
+      "epoch": 0.49969381506429883,
+      "grad_norm": 0.34849029779434204,
+      "learning_rate": 4.717635789113521e-06,
+      "loss": 0.6617,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5001837109614207,
+      "grad_norm": 0.3474747836589813,
+      "learning_rate": 4.717033477006648e-06,
+      "loss": 0.6206,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5006736068585426,
+      "grad_norm": 0.3422054648399353,
+      "learning_rate": 4.716430561720709e-06,
+      "loss": 0.6646,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5011635027556645,
+      "grad_norm": 0.3335569202899933,
+      "learning_rate": 4.715827043419737e-06,
+      "loss": 0.633,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5016533986527862,
+      "grad_norm": 0.33768150210380554,
+      "learning_rate": 4.7152229222679285e-06,
+      "loss": 0.6749,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5021432945499081,
+      "grad_norm": 0.3343658447265625,
+      "learning_rate": 4.714618198429645e-06,
+      "loss": 0.6464,
+      "step": 1025
+    },
+    {
+      "epoch": 0.50263319044703,
+      "grad_norm": 0.33807894587516785,
+      "learning_rate": 4.71401287206941e-06,
+      "loss": 0.6605,
+      "step": 1026
+    },
+    {
+      "epoch": 0.5031230863441518,
+      "grad_norm": 0.3393549621105194,
+      "learning_rate": 4.713406943351913e-06,
+      "loss": 0.6217,
+      "step": 1027
+    },
+    {
+      "epoch": 0.5036129822412737,
+      "grad_norm": 0.33353668451309204,
+      "learning_rate": 4.7128004124420074e-06,
+      "loss": 0.6658,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5041028781383956,
+      "grad_norm": 0.34081828594207764,
+      "learning_rate": 4.712193279504707e-06,
+      "loss": 0.6708,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5045927740355175,
+      "grad_norm": 0.3360579311847687,
+      "learning_rate": 4.711585544705195e-06,
+      "loss": 0.6333,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5050826699326393,
+      "grad_norm": 0.34213027358055115,
+      "learning_rate": 4.710977208208812e-06,
+      "loss": 0.6771,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5055725658297612,
+      "grad_norm": 0.33724650740623474,
+      "learning_rate": 4.710368270181068e-06,
+      "loss": 0.6521,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5060624617268831,
+      "grad_norm": 0.348318487405777,
+      "learning_rate": 4.709758730787633e-06,
+      "loss": 0.6532,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5065523576240049,
+      "grad_norm": 0.3409515917301178,
+      "learning_rate": 4.709148590194343e-06,
+      "loss": 0.6564,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5070422535211268,
+      "grad_norm": 0.33423691987991333,
+      "learning_rate": 4.708537848567195e-06,
+      "loss": 0.625,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5075321494182486,
+      "grad_norm": 0.33926403522491455,
+      "learning_rate": 4.707926506072352e-06,
+      "loss": 0.6819,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5080220453153704,
+      "grad_norm": 0.32573071122169495,
+      "learning_rate": 4.7073145628761394e-06,
+      "loss": 0.6431,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5085119412124923,
+      "grad_norm": 0.3247716426849365,
+      "learning_rate": 4.706702019145045e-06,
+      "loss": 0.6184,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5090018371096142,
+      "grad_norm": 0.32566073536872864,
+      "learning_rate": 4.706088875045722e-06,
+      "loss": 0.6466,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5094917330067361,
+      "grad_norm": 0.34337013959884644,
+      "learning_rate": 4.705475130744986e-06,
+      "loss": 0.6602,
+      "step": 1040
+    },
+    {
+      "epoch": 0.5099816289038579,
+      "grad_norm": 0.33207571506500244,
+      "learning_rate": 4.704860786409815e-06,
+      "loss": 0.6485,
+      "step": 1041
+    },
+    {
+      "epoch": 0.5104715248009798,
+      "grad_norm": 0.3431706428527832,
+      "learning_rate": 4.704245842207353e-06,
+      "loss": 0.6468,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5109614206981017,
+      "grad_norm": 0.33269333839416504,
+      "learning_rate": 4.703630298304902e-06,
+      "loss": 0.6604,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5114513165952235,
+      "grad_norm": 0.33562201261520386,
+      "learning_rate": 4.7030141548699325e-06,
+      "loss": 0.6441,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5119412124923454,
+      "grad_norm": 0.32720717787742615,
+      "learning_rate": 4.702397412070076e-06,
+      "loss": 0.6311,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5124311083894673,
+      "grad_norm": 0.3368285298347473,
+      "learning_rate": 4.701780070073129e-06,
+      "loss": 0.6353,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5129210042865892,
+      "grad_norm": 0.34569135308265686,
+      "learning_rate": 4.701162129047045e-06,
+      "loss": 0.6613,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5134109001837109,
+      "grad_norm": 0.33015212416648865,
+      "learning_rate": 4.700543589159948e-06,
+      "loss": 0.6612,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5139007960808328,
+      "grad_norm": 0.3241855800151825,
+      "learning_rate": 4.699924450580121e-06,
+      "loss": 0.6207,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5143906919779547,
+      "grad_norm": 0.34834349155426025,
+      "learning_rate": 4.699304713476009e-06,
+      "loss": 0.6437,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5148805878750765,
+      "grad_norm": 0.3462347388267517,
+      "learning_rate": 4.698684378016223e-06,
+      "loss": 0.6784,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5153704837721984,
+      "grad_norm": 0.34267929196357727,
+      "learning_rate": 4.698063444369534e-06,
+      "loss": 0.6536,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5158603796693203,
+      "grad_norm": 0.3365679383277893,
+      "learning_rate": 4.697441912704878e-06,
+      "loss": 0.631,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5163502755664421,
+      "grad_norm": 0.3212380111217499,
+      "learning_rate": 4.69681978319135e-06,
+      "loss": 0.6266,
+      "step": 1054
+    },
+    {
+      "epoch": 0.516840171463564,
+      "grad_norm": 0.3165789544582367,
+      "learning_rate": 4.696197055998213e-06,
+      "loss": 0.6088,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5173300673606859,
+      "grad_norm": 0.33306971192359924,
+      "learning_rate": 4.695573731294888e-06,
+      "loss": 0.6357,
+      "step": 1056
+    },
+    {
+      "epoch": 0.5178199632578078,
+      "grad_norm": 0.3291991353034973,
+      "learning_rate": 4.694949809250962e-06,
+      "loss": 0.652,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5183098591549296,
+      "grad_norm": 0.33322209119796753,
+      "learning_rate": 4.694325290036181e-06,
+      "loss": 0.6504,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5187997550520514,
+      "grad_norm": 0.3383603096008301,
+      "learning_rate": 4.693700173820457e-06,
+      "loss": 0.6514,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5192896509491733,
+      "grad_norm": 0.31692832708358765,
+      "learning_rate": 4.693074460773861e-06,
+      "loss": 0.6002,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5197795468462951,
+      "grad_norm": 0.3406504690647125,
+      "learning_rate": 4.69244815106663e-06,
+      "loss": 0.6455,
+      "step": 1061
+    },
+    {
+      "epoch": 0.520269442743417,
+      "grad_norm": 0.33714982867240906,
+      "learning_rate": 4.691821244869161e-06,
+      "loss": 0.6195,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5207593386405389,
+      "grad_norm": 0.33816972374916077,
+      "learning_rate": 4.691193742352013e-06,
+      "loss": 0.6575,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5212492345376607,
+      "grad_norm": 0.3503166735172272,
+      "learning_rate": 4.690565643685908e-06,
+      "loss": 0.6625,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 0.3353184461593628,
+      "learning_rate": 4.689936949041731e-06,
+      "loss": 0.6535,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5222290263319045,
+      "grad_norm": 0.34093669056892395,
+      "learning_rate": 4.689307658590527e-06,
+      "loss": 0.6244,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5227189222290264,
+      "grad_norm": 0.3339327871799469,
+      "learning_rate": 4.688677772503506e-06,
+      "loss": 0.6777,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5232088181261482,
+      "grad_norm": 0.3263406455516815,
+      "learning_rate": 4.688047290952038e-06,
+      "loss": 0.6408,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5236987140232701,
+      "grad_norm": 0.3410097062587738,
+      "learning_rate": 4.687416214107655e-06,
+      "loss": 0.6394,
+      "step": 1069
+    },
+    {
+      "epoch": 0.524188609920392,
+      "grad_norm": 0.34035810828208923,
+      "learning_rate": 4.686784542142052e-06,
+      "loss": 0.6518,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5246785058175137,
+      "grad_norm": 0.33170121908187866,
+      "learning_rate": 4.686152275227085e-06,
+      "loss": 0.6382,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5251684017146356,
+      "grad_norm": 0.3450348973274231,
+      "learning_rate": 4.685519413534773e-06,
+      "loss": 0.6588,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5256582976117575,
+      "grad_norm": 0.34389328956604004,
+      "learning_rate": 4.684885957237295e-06,
+      "loss": 0.6438,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5261481935088793,
+      "grad_norm": 0.34908241033554077,
+      "learning_rate": 4.6842519065069924e-06,
+      "loss": 0.6369,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5266380894060012,
+      "grad_norm": 0.3395238220691681,
+      "learning_rate": 4.683617261516371e-06,
+      "loss": 0.6595,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5271279853031231,
+      "grad_norm": 0.3576270043849945,
+      "learning_rate": 4.682982022438093e-06,
+      "loss": 0.6664,
+      "step": 1076
+    },
+    {
+      "epoch": 0.527617881200245,
+      "grad_norm": 0.3306131064891815,
+      "learning_rate": 4.6823461894449885e-06,
+      "loss": 0.6365,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5281077770973668,
+      "grad_norm": 0.3483700156211853,
+      "learning_rate": 4.681709762710042e-06,
+      "loss": 0.6366,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5285976729944887,
+      "grad_norm": 0.3481180965900421,
+      "learning_rate": 4.681072742406408e-06,
+      "loss": 0.6414,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5290875688916106,
+      "grad_norm": 0.3466982841491699,
+      "learning_rate": 4.680435128707395e-06,
+      "loss": 0.6119,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5295774647887324,
+      "grad_norm": 0.35319292545318604,
+      "learning_rate": 4.679796921786477e-06,
+      "loss": 0.6763,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5300673606858542,
+      "grad_norm": 0.33957529067993164,
+      "learning_rate": 4.679158121817287e-06,
+      "loss": 0.644,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5305572565829761,
+      "grad_norm": 0.3421860337257385,
+      "learning_rate": 4.6785187289736215e-06,
+      "loss": 0.6345,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5310471524800979,
+      "grad_norm": 0.3383704721927643,
+      "learning_rate": 4.677878743429438e-06,
+      "loss": 0.6402,
+      "step": 1084
+    },
+    {
+      "epoch": 0.5315370483772198,
+      "grad_norm": 0.33505967259407043,
+      "learning_rate": 4.677238165358854e-06,
+      "loss": 0.6463,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5320269442743417,
+      "grad_norm": 0.33692505955696106,
+      "learning_rate": 4.676596994936149e-06,
+      "loss": 0.6614,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5325168401714636,
+      "grad_norm": 0.3422906994819641,
+      "learning_rate": 4.675955232335764e-06,
+      "loss": 0.684,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5330067360685854,
+      "grad_norm": 0.33133259415626526,
+      "learning_rate": 4.6753128777323e-06,
+      "loss": 0.6523,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5334966319657073,
+      "grad_norm": 0.3325001001358032,
+      "learning_rate": 4.67466993130052e-06,
+      "loss": 0.6484,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5339865278628292,
+      "grad_norm": 0.3328482508659363,
+      "learning_rate": 4.674026393215349e-06,
+      "loss": 0.6595,
+      "step": 1090
+    },
+    {
+      "epoch": 0.534476423759951,
+      "grad_norm": 0.3406400680541992,
+      "learning_rate": 4.673382263651869e-06,
+      "loss": 0.6405,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5349663196570729,
+      "grad_norm": 0.3300803005695343,
+      "learning_rate": 4.6727375427853285e-06,
+      "loss": 0.629,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5354562155541948,
+      "grad_norm": 0.34695202112197876,
+      "learning_rate": 4.672092230791132e-06,
+      "loss": 0.667,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5359461114513165,
+      "grad_norm": 0.3438461422920227,
+      "learning_rate": 4.6714463278448475e-06,
+      "loss": 0.6396,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5364360073484384,
+      "grad_norm": 0.33978861570358276,
+      "learning_rate": 4.670799834122204e-06,
+      "loss": 0.6649,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5369259032455603,
+      "grad_norm": 0.34469425678253174,
+      "learning_rate": 4.6701527497990886e-06,
+      "loss": 0.6625,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5374157991426822,
+      "grad_norm": 0.34310537576675415,
+      "learning_rate": 4.669505075051553e-06,
+      "loss": 0.6516,
+      "step": 1097
+    },
+    {
+      "epoch": 0.537905695039804,
+      "grad_norm": 0.34367427229881287,
+      "learning_rate": 4.668856810055805e-06,
+      "loss": 0.6483,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5383955909369259,
+      "grad_norm": 0.33170315623283386,
+      "learning_rate": 4.668207954988217e-06,
+      "loss": 0.6468,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5388854868340478,
+      "grad_norm": 0.3421383798122406,
+      "learning_rate": 4.66755851002532e-06,
+      "loss": 0.6819,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5393753827311696,
+      "grad_norm": 0.3380526304244995,
+      "learning_rate": 4.666908475343805e-06,
+      "loss": 0.6373,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5398652786282915,
+      "grad_norm": 0.35264331102371216,
+      "learning_rate": 4.666257851120525e-06,
+      "loss": 0.6699,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5403551745254134,
+      "grad_norm": 0.3282771408557892,
+      "learning_rate": 4.6656066375324924e-06,
+      "loss": 0.6482,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5408450704225352,
+      "grad_norm": 0.34882423281669617,
+      "learning_rate": 4.664954834756881e-06,
+      "loss": 0.665,
+      "step": 1104
+    },
+    {
+      "epoch": 0.541334966319657,
+      "grad_norm": 0.3674059808254242,
+      "learning_rate": 4.664302442971024e-06,
+      "loss": 0.6649,
+      "step": 1105
+    },
+    {
+      "epoch": 0.541824862216779,
+      "grad_norm": 0.3493945896625519,
+      "learning_rate": 4.663649462352413e-06,
+      "loss": 0.6531,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5423147581139008,
+      "grad_norm": 0.347479909658432,
+      "learning_rate": 4.662995893078702e-06,
+      "loss": 0.6499,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5428046540110226,
+      "grad_norm": 0.3363555669784546,
+      "learning_rate": 4.6623417353277074e-06,
+      "loss": 0.6465,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5432945499081445,
+      "grad_norm": 0.35221952199935913,
+      "learning_rate": 4.661686989277401e-06,
+      "loss": 0.6255,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5437844458052664,
+      "grad_norm": 0.3683142066001892,
+      "learning_rate": 4.661031655105917e-06,
+      "loss": 0.6582,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5442743417023882,
+      "grad_norm": 0.3454713225364685,
+      "learning_rate": 4.660375732991551e-06,
+      "loss": 0.6543,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5447642375995101,
+      "grad_norm": 0.33530518412590027,
+      "learning_rate": 4.659719223112755e-06,
+      "loss": 0.6654,
+      "step": 1112
+    },
+    {
+      "epoch": 0.545254133496632,
+      "grad_norm": 0.3434372544288635,
+      "learning_rate": 4.659062125648144e-06,
+      "loss": 0.6548,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5457440293937539,
+      "grad_norm": 0.33734411001205444,
+      "learning_rate": 4.658404440776491e-06,
+      "loss": 0.6547,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5462339252908757,
+      "grad_norm": 0.339223176240921,
+      "learning_rate": 4.657746168676731e-06,
+      "loss": 0.62,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5467238211879976,
+      "grad_norm": 0.3487626910209656,
+      "learning_rate": 4.657087309527957e-06,
+      "loss": 0.6413,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5472137170851195,
+      "grad_norm": 0.3410091996192932,
+      "learning_rate": 4.656427863509421e-06,
+      "loss": 0.6577,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5477036129822412,
+      "grad_norm": 0.34262603521347046,
+      "learning_rate": 4.655767830800536e-06,
+      "loss": 0.6196,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5481935088793631,
+      "grad_norm": 0.3456190526485443,
+      "learning_rate": 4.655107211580874e-06,
+      "loss": 0.6383,
+      "step": 1119
+    },
+    {
+      "epoch": 0.548683404776485,
+      "grad_norm": 0.3383759558200836,
+      "learning_rate": 4.654446006030168e-06,
+      "loss": 0.6492,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5491733006736068,
+      "grad_norm": 0.3437906801700592,
+      "learning_rate": 4.653784214328309e-06,
+      "loss": 0.6358,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5496631965707287,
+      "grad_norm": 0.34938323497772217,
+      "learning_rate": 4.653121836655348e-06,
+      "loss": 0.6887,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5501530924678506,
+      "grad_norm": 0.35022029280662537,
+      "learning_rate": 4.652458873191496e-06,
+      "loss": 0.6198,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5506429883649725,
+      "grad_norm": 0.36812126636505127,
+      "learning_rate": 4.651795324117121e-06,
+      "loss": 0.6725,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5511328842620943,
+      "grad_norm": 0.33869123458862305,
+      "learning_rate": 4.651131189612753e-06,
+      "loss": 0.6463,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5516227801592162,
+      "grad_norm": 0.3331858217716217,
+      "learning_rate": 4.6504664698590795e-06,
+      "loss": 0.6208,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5521126760563381,
+      "grad_norm": 0.37185487151145935,
+      "learning_rate": 4.64980116503695e-06,
+      "loss": 0.6047,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5526025719534599,
+      "grad_norm": 0.33874985575675964,
+      "learning_rate": 4.64913527532737e-06,
+      "loss": 0.6609,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5530924678505817,
+      "grad_norm": 0.3434850871562958,
+      "learning_rate": 4.648468800911506e-06,
+      "loss": 0.64,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5535823637477036,
+      "grad_norm": 0.3372526168823242,
+      "learning_rate": 4.64780174197068e-06,
+      "loss": 0.6444,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5540722596448254,
+      "grad_norm": 0.34443947672843933,
+      "learning_rate": 4.64713409868638e-06,
+      "loss": 0.6298,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5545621555419473,
+      "grad_norm": 0.3271819055080414,
+      "learning_rate": 4.646465871240246e-06,
+      "loss": 0.6408,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5550520514390692,
+      "grad_norm": 0.35289159417152405,
+      "learning_rate": 4.645797059814083e-06,
+      "loss": 0.6305,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5555419473361911,
+      "grad_norm": 0.3394930362701416,
+      "learning_rate": 4.645127664589848e-06,
+      "loss": 0.6392,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5560318432333129,
+      "grad_norm": 0.34543678164482117,
+      "learning_rate": 4.644457685749663e-06,
+      "loss": 0.6357,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5565217391304348,
+      "grad_norm": 0.34728914499282837,
+      "learning_rate": 4.643787123475806e-06,
+      "loss": 0.6703,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5570116350275567,
+      "grad_norm": 0.3504582643508911,
+      "learning_rate": 4.643115977950713e-06,
+      "loss": 0.6491,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5575015309246785,
+      "grad_norm": 0.34880971908569336,
+      "learning_rate": 4.642444249356981e-06,
+      "loss": 0.6385,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5579914268218004,
+      "grad_norm": 0.3401770293712616,
+      "learning_rate": 4.641771937877364e-06,
+      "loss": 0.6738,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5584813227189223,
+      "grad_norm": 0.3317437171936035,
+      "learning_rate": 4.641099043694775e-06,
+      "loss": 0.6203,
+      "step": 1140
+    },
+    {
+      "epoch": 0.558971218616044,
+      "grad_norm": 0.3573921322822571,
+      "learning_rate": 4.640425566992284e-06,
+      "loss": 0.657,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5594611145131659,
+      "grad_norm": 0.361086368560791,
+      "learning_rate": 4.639751507953124e-06,
+      "loss": 0.6261,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5599510104102878,
+      "grad_norm": 0.35282137989997864,
+      "learning_rate": 4.639076866760681e-06,
+      "loss": 0.6494,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5604409063074097,
+      "grad_norm": 0.3424915671348572,
+      "learning_rate": 4.638401643598504e-06,
+      "loss": 0.6582,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5609308022045315,
+      "grad_norm": 0.3451825976371765,
+      "learning_rate": 4.6377258386502956e-06,
+      "loss": 0.6469,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5614206981016534,
+      "grad_norm": 0.38539865612983704,
+      "learning_rate": 4.637049452099921e-06,
+      "loss": 0.6455,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5619105939987753,
+      "grad_norm": 0.34357380867004395,
+      "learning_rate": 4.636372484131402e-06,
+      "loss": 0.6174,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5624004898958971,
+      "grad_norm": 0.3448379337787628,
+      "learning_rate": 4.6356949349289164e-06,
+      "loss": 0.6623,
+      "step": 1148
+    },
+    {
+      "epoch": 0.562890385793019,
+      "grad_norm": 0.33698979020118713,
+      "learning_rate": 4.635016804676804e-06,
+      "loss": 0.6447,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5633802816901409,
+      "grad_norm": 0.3464658260345459,
+      "learning_rate": 4.634338093559561e-06,
+      "loss": 0.6389,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5638701775872627,
+      "grad_norm": 0.3458057940006256,
+      "learning_rate": 4.633658801761841e-06,
+      "loss": 0.6406,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5643600734843846,
+      "grad_norm": 0.33887600898742676,
+      "learning_rate": 4.6329789294684554e-06,
+      "loss": 0.639,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5648499693815064,
+      "grad_norm": 0.3495391309261322,
+      "learning_rate": 4.632298476864376e-06,
+      "loss": 0.6676,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5653398652786283,
+      "grad_norm": 0.34821441769599915,
+      "learning_rate": 4.631617444134729e-06,
+      "loss": 0.6338,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5658297611757501,
+      "grad_norm": 0.3480973243713379,
+      "learning_rate": 4.630935831464801e-06,
+      "loss": 0.6485,
+      "step": 1155
+    },
+    {
+      "epoch": 0.566319657072872,
+      "grad_norm": 0.34872427582740784,
+      "learning_rate": 4.630253639040035e-06,
+      "loss": 0.6717,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5668095529699939,
+      "grad_norm": 0.35412171483039856,
+      "learning_rate": 4.629570867046032e-06,
+      "loss": 0.6638,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5672994488671157,
+      "grad_norm": 0.3398905396461487,
+      "learning_rate": 4.628887515668552e-06,
+      "loss": 0.6222,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5677893447642376,
+      "grad_norm": 0.3369976878166199,
+      "learning_rate": 4.628203585093511e-06,
+      "loss": 0.6314,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5682792406613595,
+      "grad_norm": 0.3475426733493805,
+      "learning_rate": 4.627519075506983e-06,
+      "loss": 0.6509,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5687691365584813,
+      "grad_norm": 0.33894526958465576,
+      "learning_rate": 4.6268339870952e-06,
+      "loss": 0.6242,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5692590324556032,
+      "grad_norm": 0.33706197142601013,
+      "learning_rate": 4.626148320044551e-06,
+      "loss": 0.6349,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5697489283527251,
+      "grad_norm": 0.34657469391822815,
+      "learning_rate": 4.625462074541582e-06,
+      "loss": 0.6365,
+      "step": 1163
+    },
+    {
+      "epoch": 0.570238824249847,
+      "grad_norm": 0.33160915970802307,
+      "learning_rate": 4.624775250772999e-06,
+      "loss": 0.6154,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5707287201469687,
+      "grad_norm": 0.3429828882217407,
+      "learning_rate": 4.62408784892566e-06,
+      "loss": 0.6216,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5712186160440906,
+      "grad_norm": 0.3496789038181305,
+      "learning_rate": 4.623399869186587e-06,
+      "loss": 0.6116,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5717085119412125,
+      "grad_norm": 0.3350716531276703,
+      "learning_rate": 4.622711311742954e-06,
+      "loss": 0.6266,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5721984078383343,
+      "grad_norm": 0.3509974777698517,
+      "learning_rate": 4.6220221767820926e-06,
+      "loss": 0.6623,
+      "step": 1168
+    },
+    {
+      "epoch": 0.5726883037354562,
+      "grad_norm": 0.33819615840911865,
+      "learning_rate": 4.621332464491495e-06,
+      "loss": 0.6323,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5731781996325781,
+      "grad_norm": 0.3499974310398102,
+      "learning_rate": 4.620642175058808e-06,
+      "loss": 0.6694,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5736680955296999,
+      "grad_norm": 0.34968307614326477,
+      "learning_rate": 4.619951308671836e-06,
+      "loss": 0.6417,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5741579914268218,
+      "grad_norm": 0.34229612350463867,
+      "learning_rate": 4.6192598655185396e-06,
+      "loss": 0.6577,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5746478873239437,
+      "grad_norm": 0.35713833570480347,
+      "learning_rate": 4.618567845787036e-06,
+      "loss": 0.6494,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5751377832210656,
+      "grad_norm": 0.3432537913322449,
+      "learning_rate": 4.617875249665601e-06,
+      "loss": 0.6442,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5756276791181874,
+      "grad_norm": 0.34398871660232544,
+      "learning_rate": 4.617182077342668e-06,
+      "loss": 0.6111,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5761175750153092,
+      "grad_norm": 0.3581888675689697,
+      "learning_rate": 4.616488329006822e-06,
+      "loss": 0.6443,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5766074709124311,
+      "grad_norm": 0.34683945775032043,
+      "learning_rate": 4.615794004846811e-06,
+      "loss": 0.6402,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5770973668095529,
+      "grad_norm": 0.3287530541419983,
+      "learning_rate": 4.615099105051536e-06,
+      "loss": 0.6354,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5775872627066748,
+      "grad_norm": 0.33705562353134155,
+      "learning_rate": 4.614403629810055e-06,
+      "loss": 0.628,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5780771586037967,
+      "grad_norm": 0.3340762257575989,
+      "learning_rate": 4.613707579311584e-06,
+      "loss": 0.6575,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5785670545009186,
+      "grad_norm": 0.34472784399986267,
+      "learning_rate": 4.613010953745494e-06,
+      "loss": 0.6307,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5790569503980404,
+      "grad_norm": 0.3373768925666809,
+      "learning_rate": 4.612313753301313e-06,
+      "loss": 0.638,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5795468462951623,
+      "grad_norm": 0.35141313076019287,
+      "learning_rate": 4.611615978168725e-06,
+      "loss": 0.6398,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5800367421922842,
+      "grad_norm": 0.35514089465141296,
+      "learning_rate": 4.610917628537571e-06,
+      "loss": 0.6478,
+      "step": 1184
+    },
+    {
+      "epoch": 0.580526638089406,
+      "grad_norm": 0.3332272469997406,
+      "learning_rate": 4.61021870459785e-06,
+      "loss": 0.6333,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5810165339865279,
+      "grad_norm": 0.3489050269126892,
+      "learning_rate": 4.6095192065397125e-06,
+      "loss": 0.6534,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5815064298836498,
+      "grad_norm": 0.3449384272098541,
+      "learning_rate": 4.608819134553469e-06,
+      "loss": 0.6622,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5819963257807715,
+      "grad_norm": 0.3545195162296295,
+      "learning_rate": 4.608118488829587e-06,
+      "loss": 0.6835,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5824862216778934,
+      "grad_norm": 0.37098783254623413,
+      "learning_rate": 4.607417269558685e-06,
+      "loss": 0.6472,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5829761175750153,
+      "grad_norm": 0.32945647835731506,
+      "learning_rate": 4.606715476931543e-06,
+      "loss": 0.6581,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5834660134721372,
+      "grad_norm": 0.33627375960350037,
+      "learning_rate": 4.606013111139095e-06,
+      "loss": 0.6253,
+      "step": 1191
+    },
+    {
+      "epoch": 0.583955909369259,
+      "grad_norm": 0.34866514801979065,
+      "learning_rate": 4.605310172372429e-06,
+      "loss": 0.6474,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5844458052663809,
+      "grad_norm": 0.34531986713409424,
+      "learning_rate": 4.6046066608227925e-06,
+      "loss": 0.6454,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5849357011635028,
+      "grad_norm": 0.35326138138771057,
+      "learning_rate": 4.6039025766815855e-06,
+      "loss": 0.6415,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5854255970606246,
+      "grad_norm": 0.3371431529521942,
+      "learning_rate": 4.6031979201403655e-06,
+      "loss": 0.6329,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5859154929577465,
+      "grad_norm": 0.34482115507125854,
+      "learning_rate": 4.602492691390847e-06,
+      "loss": 0.6554,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5864053888548684,
+      "grad_norm": 0.3397802412509918,
+      "learning_rate": 4.601786890624896e-06,
+      "loss": 0.6374,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5868952847519902,
+      "grad_norm": 0.34569984674453735,
+      "learning_rate": 4.601080518034539e-06,
+      "loss": 0.6305,
+      "step": 1198
+    },
+    {
+      "epoch": 0.587385180649112,
+      "grad_norm": 0.34162670373916626,
+      "learning_rate": 4.6003735738119546e-06,
+      "loss": 0.6238,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5878750765462339,
+      "grad_norm": 0.32768315076828003,
+      "learning_rate": 4.59966605814948e-06,
+      "loss": 0.6465,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5883649724433558,
+      "grad_norm": 0.33572009205818176,
+      "learning_rate": 4.598957971239603e-06,
+      "loss": 0.6497,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5888548683404776,
+      "grad_norm": 0.34493955969810486,
+      "learning_rate": 4.598249313274972e-06,
+      "loss": 0.6389,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5893447642375995,
+      "grad_norm": 0.32795292139053345,
+      "learning_rate": 4.5975400844483885e-06,
+      "loss": 0.648,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5898346601347214,
+      "grad_norm": 0.34611472487449646,
+      "learning_rate": 4.596830284952809e-06,
+      "loss": 0.6359,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5903245560318432,
+      "grad_norm": 0.353735089302063,
+      "learning_rate": 4.596119914981346e-06,
+      "loss": 0.6448,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5908144519289651,
+      "grad_norm": 0.3331935703754425,
+      "learning_rate": 4.5954089747272665e-06,
+      "loss": 0.6347,
+      "step": 1206
+    },
+    {
+      "epoch": 0.591304347826087,
+      "grad_norm": 0.35150009393692017,
+      "learning_rate": 4.5946974643839934e-06,
+      "loss": 0.6409,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5917942437232088,
+      "grad_norm": 0.3617503345012665,
+      "learning_rate": 4.593985384145103e-06,
+      "loss": 0.665,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5922841396203307,
+      "grad_norm": 0.3458351492881775,
+      "learning_rate": 4.593272734204331e-06,
+      "loss": 0.6473,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5927740355174526,
+      "grad_norm": 0.35933324694633484,
+      "learning_rate": 4.592559514755562e-06,
+      "loss": 0.6561,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5932639314145745,
+      "grad_norm": 0.34097525477409363,
+      "learning_rate": 4.591845725992841e-06,
+      "loss": 0.6501,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5937538273116962,
+      "grad_norm": 0.3733212649822235,
+      "learning_rate": 4.591131368110363e-06,
+      "loss": 0.6705,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5942437232088181,
+      "grad_norm": 0.3457193374633789,
+      "learning_rate": 4.590416441302483e-06,
+      "loss": 0.641,
+      "step": 1213
+    },
+    {
+      "epoch": 0.59473361910594,
+      "grad_norm": 0.36639225482940674,
+      "learning_rate": 4.589700945763707e-06,
+      "loss": 0.6582,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5952235150030618,
+      "grad_norm": 0.3443004786968231,
+      "learning_rate": 4.588984881688696e-06,
+      "loss": 0.6359,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5957134109001837,
+      "grad_norm": 0.33908525109291077,
+      "learning_rate": 4.588268249272269e-06,
+      "loss": 0.5971,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5962033067973056,
+      "grad_norm": 0.3528711199760437,
+      "learning_rate": 4.5875510487093946e-06,
+      "loss": 0.6577,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5966932026944274,
+      "grad_norm": 0.3453948199748993,
+      "learning_rate": 4.5868332801952e-06,
+      "loss": 0.665,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5971830985915493,
+      "grad_norm": 0.3543250858783722,
+      "learning_rate": 4.586114943924964e-06,
+      "loss": 0.6449,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5976729944886712,
+      "grad_norm": 0.35478734970092773,
+      "learning_rate": 4.585396040094124e-06,
+      "loss": 0.634,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5981628903857931,
+      "grad_norm": 0.35524943470954895,
+      "learning_rate": 4.584676568898267e-06,
+      "loss": 0.624,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5986527862829149,
+      "grad_norm": 0.35198506712913513,
+      "learning_rate": 4.583956530533137e-06,
+      "loss": 0.6355,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5991426821800367,
+      "grad_norm": 0.355237752199173,
+      "learning_rate": 4.583235925194632e-06,
+      "loss": 0.6491,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5996325780771586,
+      "grad_norm": 0.348114550113678,
+      "learning_rate": 4.582514753078805e-06,
+      "loss": 0.6518,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6001224739742804,
+      "grad_norm": 0.351275235414505,
+      "learning_rate": 4.581793014381861e-06,
+      "loss": 0.6102,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6006123698714023,
+      "grad_norm": 0.3485579490661621,
+      "learning_rate": 4.58107070930016e-06,
+      "loss": 0.6296,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6011022657685242,
+      "grad_norm": 0.3615967929363251,
+      "learning_rate": 4.580347838030219e-06,
+      "loss": 0.6501,
+      "step": 1227
+    },
+    {
+      "epoch": 0.601592161665646,
+      "grad_norm": 0.33674904704093933,
+      "learning_rate": 4.579624400768704e-06,
+      "loss": 0.6393,
+      "step": 1228
+    },
+    {
+      "epoch": 0.6020820575627679,
+      "grad_norm": 0.3606870174407959,
+      "learning_rate": 4.578900397712439e-06,
+      "loss": 0.6315,
+      "step": 1229
+    },
+    {
+      "epoch": 0.6025719534598898,
+      "grad_norm": 0.3589491844177246,
+      "learning_rate": 4.5781758290584e-06,
+      "loss": 0.632,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6030618493570117,
+      "grad_norm": 0.3339066803455353,
+      "learning_rate": 4.577450695003717e-06,
+      "loss": 0.6391,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6035517452541335,
+      "grad_norm": 0.3310963213443756,
+      "learning_rate": 4.576724995745674e-06,
+      "loss": 0.6447,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6040416411512554,
+      "grad_norm": 0.3522682785987854,
+      "learning_rate": 4.5759987314817115e-06,
+      "loss": 0.6486,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6045315370483773,
+      "grad_norm": 0.3547263443470001,
+      "learning_rate": 4.575271902409418e-06,
+      "loss": 0.6352,
+      "step": 1234
+    },
+    {
+      "epoch": 0.605021432945499,
+      "grad_norm": 0.34916889667510986,
+      "learning_rate": 4.574544508726539e-06,
+      "loss": 0.6312,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6055113288426209,
+      "grad_norm": 0.33695587515830994,
+      "learning_rate": 4.573816550630977e-06,
+      "loss": 0.6345,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6060012247397428,
+      "grad_norm": 0.345257967710495,
+      "learning_rate": 4.57308802832078e-06,
+      "loss": 0.6353,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6064911206368646,
+      "grad_norm": 0.36001551151275635,
+      "learning_rate": 4.572358941994158e-06,
+      "loss": 0.6327,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6069810165339865,
+      "grad_norm": 0.3538808822631836,
+      "learning_rate": 4.571629291849467e-06,
+      "loss": 0.6475,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6074709124311084,
+      "grad_norm": 0.34200891852378845,
+      "learning_rate": 4.570899078085223e-06,
+      "loss": 0.6101,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6079608083282303,
+      "grad_norm": 0.34063735604286194,
+      "learning_rate": 4.57016830090009e-06,
+      "loss": 0.6288,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6084507042253521,
+      "grad_norm": 0.3665306866168976,
+      "learning_rate": 4.569436960492889e-06,
+      "loss": 0.6801,
+      "step": 1242
+    },
+    {
+      "epoch": 0.608940600122474,
+      "grad_norm": 0.35070276260375977,
+      "learning_rate": 4.5687050570625915e-06,
+      "loss": 0.6408,
+      "step": 1243
+    },
+    {
+      "epoch": 0.6094304960195959,
+      "grad_norm": 0.33832818269729614,
+      "learning_rate": 4.5679725908083236e-06,
+      "loss": 0.6426,
+      "step": 1244
+    },
+    {
+      "epoch": 0.6099203919167177,
+      "grad_norm": 0.34335771203041077,
+      "learning_rate": 4.5672395619293645e-06,
+      "loss": 0.6562,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6104102878138395,
+      "grad_norm": 0.3389594554901123,
+      "learning_rate": 4.566505970625147e-06,
+      "loss": 0.6519,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6109001837109614,
+      "grad_norm": 0.33421632647514343,
+      "learning_rate": 4.565771817095255e-06,
+      "loss": 0.6308,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6113900796080833,
+      "grad_norm": 0.3511391580104828,
+      "learning_rate": 4.565037101539428e-06,
+      "loss": 0.6095,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6118799755052051,
+      "grad_norm": 0.3483385741710663,
+      "learning_rate": 4.564301824157557e-06,
+      "loss": 0.6042,
+      "step": 1249
+    },
+    {
+      "epoch": 0.612369871402327,
+      "grad_norm": 0.33592280745506287,
+      "learning_rate": 4.563565985149684e-06,
+      "loss": 0.6507,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6128597672994489,
+      "grad_norm": 0.35174697637557983,
+      "learning_rate": 4.562829584716007e-06,
+      "loss": 0.6316,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6133496631965707,
+      "grad_norm": 0.34856846928596497,
+      "learning_rate": 4.562092623056875e-06,
+      "loss": 0.6196,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6138395590936926,
+      "grad_norm": 0.3431689143180847,
+      "learning_rate": 4.5613551003727905e-06,
+      "loss": 0.6273,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6143294549908145,
+      "grad_norm": 0.34159907698631287,
+      "learning_rate": 4.560617016864408e-06,
+      "loss": 0.6265,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6148193508879363,
+      "grad_norm": 0.3400711715221405,
+      "learning_rate": 4.5598783727325345e-06,
+      "loss": 0.6318,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6153092467850582,
+      "grad_norm": 0.3741343319416046,
+      "learning_rate": 4.559139168178131e-06,
+      "loss": 0.6409,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6157991426821801,
+      "grad_norm": 0.342777818441391,
+      "learning_rate": 4.558399403402308e-06,
+      "loss": 0.6472,
+      "step": 1257
+    },
+    {
+      "epoch": 0.616289038579302,
+      "grad_norm": 0.3362821936607361,
+      "learning_rate": 4.557659078606332e-06,
+      "loss": 0.6174,
+      "step": 1258
+    },
+    {
+      "epoch": 0.6167789344764237,
+      "grad_norm": 0.3407340347766876,
+      "learning_rate": 4.5569181939916195e-06,
+      "loss": 0.6475,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6172688303735456,
+      "grad_norm": 0.3264087736606598,
+      "learning_rate": 4.5561767497597385e-06,
+      "loss": 0.6583,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6177587262706675,
+      "grad_norm": 0.3396836221218109,
+      "learning_rate": 4.555434746112412e-06,
+      "loss": 0.656,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6182486221677893,
+      "grad_norm": 0.3620426654815674,
+      "learning_rate": 4.5546921832515145e-06,
+      "loss": 0.642,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6187385180649112,
+      "grad_norm": 0.3363342881202698,
+      "learning_rate": 4.5539490613790715e-06,
+      "loss": 0.6303,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6192284139620331,
+      "grad_norm": 0.3500632047653198,
+      "learning_rate": 4.55320538069726e-06,
+      "loss": 0.6407,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6197183098591549,
+      "grad_norm": 0.35342806577682495,
+      "learning_rate": 4.55246114140841e-06,
+      "loss": 0.6814,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6202082057562768,
+      "grad_norm": 0.336715966463089,
+      "learning_rate": 4.5517163437150056e-06,
+      "loss": 0.6444,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6206981016533987,
+      "grad_norm": 0.3470943868160248,
+      "learning_rate": 4.55097098781968e-06,
+      "loss": 0.6302,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6211879975505206,
+      "grad_norm": 0.3440588116645813,
+      "learning_rate": 4.5502250739252165e-06,
+      "loss": 0.637,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6216778934476423,
+      "grad_norm": 0.34795132279396057,
+      "learning_rate": 4.5494786022345565e-06,
+      "loss": 0.6319,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6221677893447642,
+      "grad_norm": 0.35112684965133667,
+      "learning_rate": 4.548731572950788e-06,
+      "loss": 0.6207,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6226576852418861,
+      "grad_norm": 0.34473106265068054,
+      "learning_rate": 4.547983986277151e-06,
+      "loss": 0.6297,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6231475811390079,
+      "grad_norm": 0.3364686667919159,
+      "learning_rate": 4.547235842417041e-06,
+      "loss": 0.6783,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6236374770361298,
+      "grad_norm": 0.3369466960430145,
+      "learning_rate": 4.546487141574e-06,
+      "loss": 0.6229,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6241273729332517,
+      "grad_norm": 0.3586369752883911,
+      "learning_rate": 4.545737883951724e-06,
+      "loss": 0.6675,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6246172688303735,
+      "grad_norm": 0.34704333543777466,
+      "learning_rate": 4.544988069754061e-06,
+      "loss": 0.6376,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6251071647274954,
+      "grad_norm": 0.3543643057346344,
+      "learning_rate": 4.544237699185012e-06,
+      "loss": 0.6614,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6255970606246173,
+      "grad_norm": 0.3370320796966553,
+      "learning_rate": 4.543486772448722e-06,
+      "loss": 0.6609,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6260869565217392,
+      "grad_norm": 0.35931840538978577,
+      "learning_rate": 4.542735289749498e-06,
+      "loss": 0.6458,
+      "step": 1278
+    },
+    {
+      "epoch": 0.626576852418861,
+      "grad_norm": 0.35878509283065796,
+      "learning_rate": 4.5419832512917895e-06,
+      "loss": 0.658,
+      "step": 1279
+    },
+    {
+      "epoch": 0.6270667483159829,
+      "grad_norm": 0.3401985168457031,
+      "learning_rate": 4.5412306572802014e-06,
+      "loss": 0.6241,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6275566442131048,
+      "grad_norm": 0.3442074656486511,
+      "learning_rate": 4.540477507919489e-06,
+      "loss": 0.6685,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6280465401102265,
+      "grad_norm": 0.3460010290145874,
+      "learning_rate": 4.539723803414557e-06,
+      "loss": 0.6107,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6285364360073484,
+      "grad_norm": 0.3613050580024719,
+      "learning_rate": 4.538969543970465e-06,
+      "loss": 0.6691,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6290263319044703,
+      "grad_norm": 0.3598838448524475,
+      "learning_rate": 4.53821472979242e-06,
+      "loss": 0.6615,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6295162278015921,
+      "grad_norm": 0.3557281792163849,
+      "learning_rate": 4.537459361085781e-06,
+      "loss": 0.6287,
+      "step": 1285
+    },
+    {
+      "epoch": 0.630006123698714,
+      "grad_norm": 0.33958539366722107,
+      "learning_rate": 4.5367034380560584e-06,
+      "loss": 0.6398,
+      "step": 1286
+    },
+    {
+      "epoch": 0.6304960195958359,
+      "grad_norm": 0.36244890093803406,
+      "learning_rate": 4.535946960908913e-06,
+      "loss": 0.6518,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6309859154929578,
+      "grad_norm": 0.3795658349990845,
+      "learning_rate": 4.535189929850156e-06,
+      "loss": 0.6558,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6314758113900796,
+      "grad_norm": 0.359651118516922,
+      "learning_rate": 4.534432345085751e-06,
+      "loss": 0.6182,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6319657072872015,
+      "grad_norm": 0.37145259976387024,
+      "learning_rate": 4.533674206821809e-06,
+      "loss": 0.6564,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6324556031843234,
+      "grad_norm": 0.3477351665496826,
+      "learning_rate": 4.532915515264595e-06,
+      "loss": 0.648,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6329454990814452,
+      "grad_norm": 0.33393770456314087,
+      "learning_rate": 4.532156270620522e-06,
+      "loss": 0.6628,
+      "step": 1292
+    },
+    {
+      "epoch": 0.633435394978567,
+      "grad_norm": 0.35315823554992676,
+      "learning_rate": 4.531396473096156e-06,
+      "loss": 0.6305,
+      "step": 1293
+    },
+    {
+      "epoch": 0.6339252908756889,
+      "grad_norm": 0.35018977522850037,
+      "learning_rate": 4.53063612289821e-06,
+      "loss": 0.645,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6344151867728107,
+      "grad_norm": 0.36220550537109375,
+      "learning_rate": 4.529875220233551e-06,
+      "loss": 0.6243,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6349050826699326,
+      "grad_norm": 0.35167255997657776,
+      "learning_rate": 4.5291137653091935e-06,
+      "loss": 0.6497,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6353949785670545,
+      "grad_norm": 0.36633917689323425,
+      "learning_rate": 4.528351758332303e-06,
+      "loss": 0.6378,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6358848744641764,
+      "grad_norm": 0.3628346621990204,
+      "learning_rate": 4.5275891995101975e-06,
+      "loss": 0.6283,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6363747703612982,
+      "grad_norm": 0.34469640254974365,
+      "learning_rate": 4.526826089050341e-06,
+      "loss": 0.6366,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6368646662584201,
+      "grad_norm": 0.3598330020904541,
+      "learning_rate": 4.526062427160353e-06,
+      "loss": 0.6479,
+      "step": 1300
+    },
+    {
+      "epoch": 0.637354562155542,
+      "grad_norm": 0.3442913591861725,
+      "learning_rate": 4.525298214047995e-06,
+      "loss": 0.6231,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6378444580526638,
+      "grad_norm": 0.3670358955860138,
+      "learning_rate": 4.5245334499211874e-06,
+      "loss": 0.6407,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6383343539497857,
+      "grad_norm": 0.36164307594299316,
+      "learning_rate": 4.523768134987995e-06,
+      "loss": 0.6431,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6388242498469076,
+      "grad_norm": 0.33182641863822937,
+      "learning_rate": 4.523002269456633e-06,
+      "loss": 0.6371,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6393141457440293,
+      "grad_norm": 0.3582742214202881,
+      "learning_rate": 4.522235853535468e-06,
+      "loss": 0.6212,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6398040416411512,
+      "grad_norm": 0.3448125123977661,
+      "learning_rate": 4.521468887433016e-06,
+      "loss": 0.6476,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6402939375382731,
+      "grad_norm": 0.355258584022522,
+      "learning_rate": 4.520701371357942e-06,
+      "loss": 0.6284,
+      "step": 1307
+    },
+    {
+      "epoch": 0.640783833435395,
+      "grad_norm": 0.3545528054237366,
+      "learning_rate": 4.519933305519061e-06,
+      "loss": 0.6225,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6412737293325168,
+      "grad_norm": 0.35198697447776794,
+      "learning_rate": 4.5191646901253375e-06,
+      "loss": 0.6257,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6417636252296387,
+      "grad_norm": 0.3620002567768097,
+      "learning_rate": 4.518395525385884e-06,
+      "loss": 0.642,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6422535211267606,
+      "grad_norm": 0.35877859592437744,
+      "learning_rate": 4.517625811509966e-06,
+      "loss": 0.66,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6427434170238824,
+      "grad_norm": 0.3415077328681946,
+      "learning_rate": 4.5168555487069955e-06,
+      "loss": 0.615,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6432333129210043,
+      "grad_norm": 0.34771469235420227,
+      "learning_rate": 4.516084737186534e-06,
+      "loss": 0.6361,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6437232088181262,
+      "grad_norm": 0.35810625553131104,
+      "learning_rate": 4.515313377158295e-06,
+      "loss": 0.6555,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6442131047152481,
+      "grad_norm": 0.36541587114334106,
+      "learning_rate": 4.514541468832137e-06,
+      "loss": 0.6216,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6447030006123698,
+      "grad_norm": 0.3559669554233551,
+      "learning_rate": 4.5137690124180714e-06,
+      "loss": 0.6304,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6451928965094917,
+      "grad_norm": 0.3511303663253784,
+      "learning_rate": 4.512996008126256e-06,
+      "loss": 0.6497,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6456827924066136,
+      "grad_norm": 0.3545239269733429,
+      "learning_rate": 4.512222456166999e-06,
+      "loss": 0.6696,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6461726883037354,
+      "grad_norm": 0.35808631777763367,
+      "learning_rate": 4.511448356750759e-06,
+      "loss": 0.6499,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6466625842008573,
+      "grad_norm": 0.34952282905578613,
+      "learning_rate": 4.51067371008814e-06,
+      "loss": 0.628,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6471524800979792,
+      "grad_norm": 0.33353301882743835,
+      "learning_rate": 4.509898516389899e-06,
+      "loss": 0.6132,
+      "step": 1321
+    },
+    {
+      "epoch": 0.647642375995101,
+      "grad_norm": 0.3683788478374481,
+      "learning_rate": 4.509122775866937e-06,
+      "loss": 0.6341,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6481322718922229,
+      "grad_norm": 0.3552803695201874,
+      "learning_rate": 4.50834648873031e-06,
+      "loss": 0.62,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6486221677893448,
+      "grad_norm": 0.35079067945480347,
+      "learning_rate": 4.507569655191216e-06,
+      "loss": 0.6079,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6491120636864667,
+      "grad_norm": 0.34289830923080444,
+      "learning_rate": 4.506792275461007e-06,
+      "loss": 0.6469,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6496019595835885,
+      "grad_norm": 0.3386910855770111,
+      "learning_rate": 4.506014349751179e-06,
+      "loss": 0.6557,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6500918554807104,
+      "grad_norm": 0.37173083424568176,
+      "learning_rate": 4.505235878273382e-06,
+      "loss": 0.6401,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6505817513778323,
+      "grad_norm": 0.37466418743133545,
+      "learning_rate": 4.504456861239409e-06,
+      "loss": 0.6153,
+      "step": 1328
+    },
+    {
+      "epoch": 0.651071647274954,
+      "grad_norm": 0.36054158210754395,
+      "learning_rate": 4.503677298861206e-06,
+      "loss": 0.6145,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6515615431720759,
+      "grad_norm": 0.3630097508430481,
+      "learning_rate": 4.502897191350863e-06,
+      "loss": 0.6296,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6520514390691978,
+      "grad_norm": 0.3524821400642395,
+      "learning_rate": 4.502116538920623e-06,
+      "loss": 0.6417,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6525413349663196,
+      "grad_norm": 0.3513556122779846,
+      "learning_rate": 4.501335341782874e-06,
+      "loss": 0.6269,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6530312308634415,
+      "grad_norm": 0.3473247289657593,
+      "learning_rate": 4.500553600150153e-06,
+      "loss": 0.6107,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6535211267605634,
+      "grad_norm": 0.35425952076911926,
+      "learning_rate": 4.4997713142351436e-06,
+      "loss": 0.649,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6540110226576853,
+      "grad_norm": 0.3341909945011139,
+      "learning_rate": 4.498988484250681e-06,
+      "loss": 0.6366,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6545009185548071,
+      "grad_norm": 0.340781569480896,
+      "learning_rate": 4.498205110409746e-06,
+      "loss": 0.6243,
+      "step": 1336
+    },
+    {
+      "epoch": 0.654990814451929,
+      "grad_norm": 0.36256301403045654,
+      "learning_rate": 4.497421192925468e-06,
+      "loss": 0.6265,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6554807103490509,
+      "grad_norm": 0.3434717357158661,
+      "learning_rate": 4.496636732011123e-06,
+      "loss": 0.63,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6559706062461726,
+      "grad_norm": 0.3541300296783447,
+      "learning_rate": 4.495851727880139e-06,
+      "loss": 0.6459,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6564605021432945,
+      "grad_norm": 0.3564087450504303,
+      "learning_rate": 4.4950661807460845e-06,
+      "loss": 0.6334,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6569503980404164,
+      "grad_norm": 0.34885209798812866,
+      "learning_rate": 4.494280090822684e-06,
+      "loss": 0.6317,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6574402939375382,
+      "grad_norm": 0.3564472496509552,
+      "learning_rate": 4.493493458323804e-06,
+      "loss": 0.6494,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6579301898346601,
+      "grad_norm": 0.35500457882881165,
+      "learning_rate": 4.492706283463459e-06,
+      "loss": 0.6453,
+      "step": 1343
+    },
+    {
+      "epoch": 0.658420085731782,
+      "grad_norm": 0.3623114824295044,
+      "learning_rate": 4.4919185664558155e-06,
+      "loss": 0.6336,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6589099816289039,
+      "grad_norm": 0.3338193893432617,
+      "learning_rate": 4.4911303075151815e-06,
+      "loss": 0.6408,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6593998775260257,
+      "grad_norm": 0.3407876193523407,
+      "learning_rate": 4.490341506856017e-06,
+      "loss": 0.6248,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6598897734231476,
+      "grad_norm": 0.3452967703342438,
+      "learning_rate": 4.489552164692926e-06,
+      "loss": 0.662,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6603796693202695,
+      "grad_norm": 0.35615143179893494,
+      "learning_rate": 4.488762281240664e-06,
+      "loss": 0.6369,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6608695652173913,
+      "grad_norm": 0.35876408219337463,
+      "learning_rate": 4.48797185671413e-06,
+      "loss": 0.6448,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6613594611145132,
+      "grad_norm": 0.3522142171859741,
+      "learning_rate": 4.487180891328371e-06,
+      "loss": 0.6286,
+      "step": 1350
+    },
+    {
+      "epoch": 0.661849357011635,
+      "grad_norm": 0.3411237299442291,
+      "learning_rate": 4.486389385298583e-06,
+      "loss": 0.6281,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6623392529087568,
+      "grad_norm": 0.3518471121788025,
+      "learning_rate": 4.485597338840106e-06,
+      "loss": 0.6391,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6628291488058787,
+      "grad_norm": 0.35424360632896423,
+      "learning_rate": 4.48480475216843e-06,
+      "loss": 0.6272,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6633190447030006,
+      "grad_norm": 0.35385259985923767,
+      "learning_rate": 4.48401162549919e-06,
+      "loss": 0.6423,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6638089406001225,
+      "grad_norm": 0.33582809567451477,
+      "learning_rate": 4.48321795904817e-06,
+      "loss": 0.6471,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6642988364972443,
+      "grad_norm": 0.35831162333488464,
+      "learning_rate": 4.4824237530312984e-06,
+      "loss": 0.6354,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6647887323943662,
+      "grad_norm": 0.33920028805732727,
+      "learning_rate": 4.481629007664652e-06,
+      "loss": 0.6028,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6652786282914881,
+      "grad_norm": 0.33606863021850586,
+      "learning_rate": 4.480833723164453e-06,
+      "loss": 0.6025,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6657685241886099,
+      "grad_norm": 0.3546634316444397,
+      "learning_rate": 4.480037899747073e-06,
+      "loss": 0.6427,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6662584200857318,
+      "grad_norm": 0.3600572347640991,
+      "learning_rate": 4.479241537629026e-06,
+      "loss": 0.6351,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6667483159828537,
+      "grad_norm": 0.37154367566108704,
+      "learning_rate": 4.4784446370269765e-06,
+      "loss": 0.6371,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6672382118799755,
+      "grad_norm": 0.3503185212612152,
+      "learning_rate": 4.477647198157733e-06,
+      "loss": 0.6452,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6677281077770973,
+      "grad_norm": 0.3488251566886902,
+      "learning_rate": 4.476849221238253e-06,
+      "loss": 0.6447,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6682180036742192,
+      "grad_norm": 0.3618388772010803,
+      "learning_rate": 4.476050706485637e-06,
+      "loss": 0.6293,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6687078995713411,
+      "grad_norm": 0.3459049165248871,
+      "learning_rate": 4.475251654117134e-06,
+      "loss": 0.6524,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6691977954684629,
+      "grad_norm": 0.33803868293762207,
+      "learning_rate": 4.4744520643501396e-06,
+      "loss": 0.6334,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6696876913655848,
+      "grad_norm": 0.3458579480648041,
+      "learning_rate": 4.473651937402194e-06,
+      "loss": 0.6286,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6701775872627067,
+      "grad_norm": 0.3435010612010956,
+      "learning_rate": 4.472851273490985e-06,
+      "loss": 0.6495,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6706674831598285,
+      "grad_norm": 0.3350188732147217,
+      "learning_rate": 4.472050072834345e-06,
+      "loss": 0.625,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6711573790569504,
+      "grad_norm": 0.3355303108692169,
+      "learning_rate": 4.471248335650255e-06,
+      "loss": 0.627,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6716472749540723,
+      "grad_norm": 0.3544834554195404,
+      "learning_rate": 4.470446062156838e-06,
+      "loss": 0.6629,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6721371708511942,
+      "grad_norm": 0.3765515089035034,
+      "learning_rate": 4.469643252572367e-06,
+      "loss": 0.6615,
+      "step": 1372
+    },
+    {
+      "epoch": 0.672627066748316,
+      "grad_norm": 0.36532217264175415,
+      "learning_rate": 4.468839907115259e-06,
+      "loss": 0.6492,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6731169626454379,
+      "grad_norm": 0.3455331325531006,
+      "learning_rate": 4.468036026004075e-06,
+      "loss": 0.6604,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6736068585425597,
+      "grad_norm": 0.34351903200149536,
+      "learning_rate": 4.4672316094575265e-06,
+      "loss": 0.6318,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6740967544396815,
+      "grad_norm": 0.37456777691841125,
+      "learning_rate": 4.466426657694466e-06,
+      "loss": 0.6362,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6745866503368034,
+      "grad_norm": 0.3516497015953064,
+      "learning_rate": 4.465621170933894e-06,
+      "loss": 0.6198,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6750765462339253,
+      "grad_norm": 0.34695130586624146,
+      "learning_rate": 4.464815149394954e-06,
+      "loss": 0.6305,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6755664421310471,
+      "grad_norm": 0.35416507720947266,
+      "learning_rate": 4.464008593296939e-06,
+      "loss": 0.6378,
+      "step": 1379
+    },
+    {
+      "epoch": 0.676056338028169,
+      "grad_norm": 0.3488617241382599,
+      "learning_rate": 4.463201502859286e-06,
+      "loss": 0.6306,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6765462339252909,
+      "grad_norm": 0.34790799021720886,
+      "learning_rate": 4.462393878301574e-06,
+      "loss": 0.6608,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6770361298224128,
+      "grad_norm": 0.3412380516529083,
+      "learning_rate": 4.461585719843533e-06,
+      "loss": 0.6133,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6775260257195346,
+      "grad_norm": 0.35063865780830383,
+      "learning_rate": 4.460777027705033e-06,
+      "loss": 0.6292,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6780159216166565,
+      "grad_norm": 0.33439743518829346,
+      "learning_rate": 4.459967802106092e-06,
+      "loss": 0.6287,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6785058175137784,
+      "grad_norm": 0.34881168603897095,
+      "learning_rate": 4.459158043266874e-06,
+      "loss": 0.6503,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6789957134109001,
+      "grad_norm": 0.3516039252281189,
+      "learning_rate": 4.458347751407686e-06,
+      "loss": 0.6667,
+      "step": 1386
+    },
+    {
+      "epoch": 0.679485609308022,
+      "grad_norm": 0.3718115985393524,
+      "learning_rate": 4.45753692674898e-06,
+      "loss": 0.6185,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6799755052051439,
+      "grad_norm": 0.3400097191333771,
+      "learning_rate": 4.456725569511353e-06,
+      "loss": 0.65,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6804654011022657,
+      "grad_norm": 0.35042303800582886,
+      "learning_rate": 4.45591367991555e-06,
+      "loss": 0.6319,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6809552969993876,
+      "grad_norm": 0.3450476825237274,
+      "learning_rate": 4.455101258182458e-06,
+      "loss": 0.6413,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6814451928965095,
+      "grad_norm": 0.3550189137458801,
+      "learning_rate": 4.454288304533107e-06,
+      "loss": 0.6349,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6819350887936314,
+      "grad_norm": 0.34984391927719116,
+      "learning_rate": 4.453474819188676e-06,
+      "loss": 0.6509,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6824249846907532,
+      "grad_norm": 0.35120296478271484,
+      "learning_rate": 4.452660802370485e-06,
+      "loss": 0.6078,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6829148805878751,
+      "grad_norm": 0.34375014901161194,
+      "learning_rate": 4.451846254300002e-06,
+      "loss": 0.6337,
+      "step": 1394
+    },
+    {
+      "epoch": 0.683404776484997,
+      "grad_norm": 0.3296529948711395,
+      "learning_rate": 4.451031175198836e-06,
+      "loss": 0.6264,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6838946723821188,
+      "grad_norm": 0.35613584518432617,
+      "learning_rate": 4.450215565288743e-06,
+      "loss": 0.6202,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6843845682792407,
+      "grad_norm": 0.3382437825202942,
+      "learning_rate": 4.449399424791624e-06,
+      "loss": 0.6263,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6848744641763626,
+      "grad_norm": 0.35394805669784546,
+      "learning_rate": 4.44858275392952e-06,
+      "loss": 0.6392,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6853643600734843,
+      "grad_norm": 0.3411145806312561,
+      "learning_rate": 4.447765552924621e-06,
+      "loss": 0.6506,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6858542559706062,
+      "grad_norm": 0.3441709876060486,
+      "learning_rate": 4.446947821999259e-06,
+      "loss": 0.6398,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6863441518677281,
+      "grad_norm": 0.35997113585472107,
+      "learning_rate": 4.446129561375911e-06,
+      "loss": 0.6614,
+      "step": 1401
+    },
+    {
+      "epoch": 0.68683404776485,
+      "grad_norm": 0.3488248288631439,
+      "learning_rate": 4.445310771277197e-06,
+      "loss": 0.6482,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6873239436619718,
+      "grad_norm": 0.3390911817550659,
+      "learning_rate": 4.444491451925883e-06,
+      "loss": 0.6138,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6878138395590937,
+      "grad_norm": 0.3483215272426605,
+      "learning_rate": 4.443671603544878e-06,
+      "loss": 0.6455,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6883037354562156,
+      "grad_norm": 0.36293405294418335,
+      "learning_rate": 4.442851226357232e-06,
+      "loss": 0.6562,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6887936313533374,
+      "grad_norm": 0.35973989963531494,
+      "learning_rate": 4.442030320586145e-06,
+      "loss": 0.6355,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6892835272504593,
+      "grad_norm": 0.35593295097351074,
+      "learning_rate": 4.441208886454956e-06,
+      "loss": 0.628,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6897734231475812,
+      "grad_norm": 0.3477623164653778,
+      "learning_rate": 4.440386924187148e-06,
+      "loss": 0.6308,
+      "step": 1408
+    },
+    {
+      "epoch": 0.690263319044703,
+      "grad_norm": 0.3570405840873718,
+      "learning_rate": 4.43956443400635e-06,
+      "loss": 0.6481,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6907532149418248,
+      "grad_norm": 0.3418077230453491,
+      "learning_rate": 4.438741416136335e-06,
+      "loss": 0.6331,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6912431108389467,
+      "grad_norm": 0.3490599989891052,
+      "learning_rate": 4.4379178708010155e-06,
+      "loss": 0.6256,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6917330067360686,
+      "grad_norm": 0.35988834500312805,
+      "learning_rate": 4.437093798224451e-06,
+      "loss": 0.6456,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6922229026331904,
+      "grad_norm": 0.37563356757164,
+      "learning_rate": 4.4362691986308435e-06,
+      "loss": 0.648,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6927127985303123,
+      "grad_norm": 0.37250471115112305,
+      "learning_rate": 4.435444072244539e-06,
+      "loss": 0.6773,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6932026944274342,
+      "grad_norm": 0.34101811051368713,
+      "learning_rate": 4.434618419290024e-06,
+      "loss": 0.6535,
+      "step": 1415
+    },
+    {
+      "epoch": 0.693692590324556,
+      "grad_norm": 0.3587406277656555,
+      "learning_rate": 4.433792239991934e-06,
+      "loss": 0.6416,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6941824862216779,
+      "grad_norm": 0.34490567445755005,
+      "learning_rate": 4.432965534575041e-06,
+      "loss": 0.6517,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6946723821187998,
+      "grad_norm": 0.3421916961669922,
+      "learning_rate": 4.432138303264265e-06,
+      "loss": 0.648,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6951622780159216,
+      "grad_norm": 0.34492430090904236,
+      "learning_rate": 4.431310546284668e-06,
+      "loss": 0.6534,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 0.33511802554130554,
+      "learning_rate": 4.430482263861451e-06,
+      "loss": 0.6188,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6961420698101654,
+      "grad_norm": 0.3665695786476135,
+      "learning_rate": 4.429653456219965e-06,
+      "loss": 0.637,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6966319657072872,
+      "grad_norm": 0.33063197135925293,
+      "learning_rate": 4.428824123585699e-06,
+      "loss": 0.6375,
+      "step": 1422
+    },
+    {
+      "epoch": 0.697121861604409,
+      "grad_norm": 0.3582562804222107,
+      "learning_rate": 4.427994266184285e-06,
+      "loss": 0.6294,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6976117575015309,
+      "grad_norm": 0.3703780770301819,
+      "learning_rate": 4.427163884241501e-06,
+      "loss": 0.6699,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6981016533986528,
+      "grad_norm": 0.35689881443977356,
+      "learning_rate": 4.426332977983264e-06,
+      "loss": 0.6439,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6985915492957746,
+      "grad_norm": 0.3539344072341919,
+      "learning_rate": 4.425501547635635e-06,
+      "loss": 0.6608,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6990814451928965,
+      "grad_norm": 0.3454554080963135,
+      "learning_rate": 4.424669593424819e-06,
+      "loss": 0.6364,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6995713410900184,
+      "grad_norm": 0.340656042098999,
+      "learning_rate": 4.4238371155771614e-06,
+      "loss": 0.6277,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7000612369871402,
+      "grad_norm": 0.35011354088783264,
+      "learning_rate": 4.423004114319151e-06,
+      "loss": 0.6621,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7005511328842621,
+      "grad_norm": 0.34832993149757385,
+      "learning_rate": 4.42217058987742e-06,
+      "loss": 0.6516,
+      "step": 1430
+    },
+    {
+      "epoch": 0.701041028781384,
+      "grad_norm": 0.3486762046813965,
+      "learning_rate": 4.421336542478741e-06,
+      "loss": 0.6138,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7015309246785059,
+      "grad_norm": 0.3553650975227356,
+      "learning_rate": 4.42050197235003e-06,
+      "loss": 0.6215,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7020208205756276,
+      "grad_norm": 0.3504565954208374,
+      "learning_rate": 4.419666879718344e-06,
+      "loss": 0.6377,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7025107164727495,
+      "grad_norm": 0.3645647466182709,
+      "learning_rate": 4.418831264810886e-06,
+      "loss": 0.6505,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7030006123698714,
+      "grad_norm": 0.355643093585968,
+      "learning_rate": 4.417995127854995e-06,
+      "loss": 0.6316,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7034905082669932,
+      "grad_norm": 0.34611374139785767,
+      "learning_rate": 4.417158469078156e-06,
+      "loss": 0.6224,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7039804041641151,
+      "grad_norm": 0.35592782497406006,
+      "learning_rate": 4.416321288707997e-06,
+      "loss": 0.6477,
+      "step": 1437
+    },
+    {
+      "epoch": 0.704470300061237,
+      "grad_norm": 0.3591293394565582,
+      "learning_rate": 4.415483586972286e-06,
+      "loss": 0.6266,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7049601959583589,
+      "grad_norm": 0.3570345342159271,
+      "learning_rate": 4.414645364098932e-06,
+      "loss": 0.6545,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7054500918554807,
+      "grad_norm": 0.36388275027275085,
+      "learning_rate": 4.413806620315986e-06,
+      "loss": 0.656,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7059399877526026,
+      "grad_norm": 0.3475639522075653,
+      "learning_rate": 4.4129673558516435e-06,
+      "loss": 0.6464,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7064298836497245,
+      "grad_norm": 0.34937432408332825,
+      "learning_rate": 4.412127570934238e-06,
+      "loss": 0.6417,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7069197795468463,
+      "grad_norm": 0.3504284918308258,
+      "learning_rate": 4.411287265792248e-06,
+      "loss": 0.6323,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7074096754439682,
+      "grad_norm": 0.3767411708831787,
+      "learning_rate": 4.41044644065429e-06,
+      "loss": 0.6448,
+      "step": 1444
+    },
+    {
+      "epoch": 0.70789957134109,
+      "grad_norm": 0.3602970838546753,
+      "learning_rate": 4.409605095749125e-06,
+      "loss": 0.6345,
+      "step": 1445
+    },
+    {
+      "epoch": 0.7083894672382118,
+      "grad_norm": 0.3349767327308655,
+      "learning_rate": 4.408763231305654e-06,
+      "loss": 0.6224,
+      "step": 1446
+    },
+    {
+      "epoch": 0.7088793631353337,
+      "grad_norm": 0.36673104763031006,
+      "learning_rate": 4.407920847552918e-06,
+      "loss": 0.6457,
+      "step": 1447
+    },
+    {
+      "epoch": 0.7093692590324556,
+      "grad_norm": 0.3664603531360626,
+      "learning_rate": 4.407077944720104e-06,
+      "loss": 0.6302,
+      "step": 1448
+    },
+    {
+      "epoch": 0.7098591549295775,
+      "grad_norm": 0.3508128821849823,
+      "learning_rate": 4.4062345230365345e-06,
+      "loss": 0.6434,
+      "step": 1449
+    },
+    {
+      "epoch": 0.7103490508266993,
+      "grad_norm": 0.3559228479862213,
+      "learning_rate": 4.4053905827316765e-06,
+      "loss": 0.6112,
+      "step": 1450
+    },
+    {
+      "epoch": 0.7108389467238212,
+      "grad_norm": 0.3495516777038574,
+      "learning_rate": 4.404546124035137e-06,
+      "loss": 0.6522,
+      "step": 1451
+    },
+    {
+      "epoch": 0.7113288426209431,
+      "grad_norm": 0.3622876703739166,
+      "learning_rate": 4.4037011471766656e-06,
+      "loss": 0.667,
+      "step": 1452
+    },
+    {
+      "epoch": 0.7118187385180649,
+      "grad_norm": 0.3653157949447632,
+      "learning_rate": 4.402855652386151e-06,
+      "loss": 0.6518,
+      "step": 1453
+    },
+    {
+      "epoch": 0.7123086344151868,
+      "grad_norm": 0.36109858751296997,
+      "learning_rate": 4.402009639893622e-06,
+      "loss": 0.6193,
+      "step": 1454
+    },
+    {
+      "epoch": 0.7127985303123087,
+      "grad_norm": 0.3428484797477722,
+      "learning_rate": 4.401163109929251e-06,
+      "loss": 0.6384,
+      "step": 1455
+    },
+    {
+      "epoch": 0.7132884262094304,
+      "grad_norm": 0.33944132924079895,
+      "learning_rate": 4.40031606272335e-06,
+      "loss": 0.6195,
+      "step": 1456
+    },
+    {
+      "epoch": 0.7137783221065523,
+      "grad_norm": 0.34885114431381226,
+      "learning_rate": 4.399468498506372e-06,
+      "loss": 0.6374,
+      "step": 1457
+    },
+    {
+      "epoch": 0.7142682180036742,
+      "grad_norm": 0.3379736840724945,
+      "learning_rate": 4.39862041750891e-06,
+      "loss": 0.6476,
+      "step": 1458
+    },
+    {
+      "epoch": 0.7147581139007961,
+      "grad_norm": 0.3457338511943817,
+      "learning_rate": 4.397771819961696e-06,
+      "loss": 0.6336,
+      "step": 1459
+    },
+    {
+      "epoch": 0.7152480097979179,
+      "grad_norm": 0.35912013053894043,
+      "learning_rate": 4.3969227060956065e-06,
+      "loss": 0.6425,
+      "step": 1460
+    },
+    {
+      "epoch": 0.7157379056950398,
+      "grad_norm": 0.34900614619255066,
+      "learning_rate": 4.3960730761416556e-06,
+      "loss": 0.6181,
+      "step": 1461
+    },
+    {
+      "epoch": 0.7162278015921617,
+      "grad_norm": 0.3592720925807953,
+      "learning_rate": 4.395222930330999e-06,
+      "loss": 0.6493,
+      "step": 1462
+    },
+    {
+      "epoch": 0.7167176974892835,
+      "grad_norm": 0.37626612186431885,
+      "learning_rate": 4.394372268894932e-06,
+      "loss": 0.6488,
+      "step": 1463
+    },
+    {
+      "epoch": 0.7172075933864054,
+      "grad_norm": 0.33996111154556274,
+      "learning_rate": 4.3935210920648906e-06,
+      "loss": 0.6002,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7176974892835273,
+      "grad_norm": 0.3402058482170105,
+      "learning_rate": 4.392669400072451e-06,
+      "loss": 0.6641,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7181873851806491,
+      "grad_norm": 0.3400043249130249,
+      "learning_rate": 4.39181719314933e-06,
+      "loss": 0.6471,
+      "step": 1466
+    },
+    {
+      "epoch": 0.718677281077771,
+      "grad_norm": 0.35515210032463074,
+      "learning_rate": 4.3909644715273826e-06,
+      "loss": 0.616,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7191671769748929,
+      "grad_norm": 0.3558686673641205,
+      "learning_rate": 4.390111235438606e-06,
+      "loss": 0.6475,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7196570728720147,
+      "grad_norm": 0.357598215341568,
+      "learning_rate": 4.3892574851151375e-06,
+      "loss": 0.6427,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7201469687691365,
+      "grad_norm": 0.3601952791213989,
+      "learning_rate": 4.388403220789252e-06,
+      "loss": 0.6324,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7206368646662584,
+      "grad_norm": 0.3500250279903412,
+      "learning_rate": 4.387548442693367e-06,
+      "loss": 0.6429,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7211267605633803,
+      "grad_norm": 0.3473350703716278,
+      "learning_rate": 4.386693151060036e-06,
+      "loss": 0.6341,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7216166564605021,
+      "grad_norm": 0.3497799038887024,
+      "learning_rate": 4.385837346121957e-06,
+      "loss": 0.6235,
+      "step": 1473
+    },
+    {
+      "epoch": 0.722106552357624,
+      "grad_norm": 0.349621444940567,
+      "learning_rate": 4.384981028111965e-06,
+      "loss": 0.6265,
+      "step": 1474
+    },
+    {
+      "epoch": 0.7225964482547459,
+      "grad_norm": 0.35969698429107666,
+      "learning_rate": 4.384124197263033e-06,
+      "loss": 0.6484,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7230863441518677,
+      "grad_norm": 0.3602536618709564,
+      "learning_rate": 4.383266853808278e-06,
+      "loss": 0.6515,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7235762400489896,
+      "grad_norm": 0.35244783759117126,
+      "learning_rate": 4.382408997980952e-06,
+      "loss": 0.6041,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7240661359461115,
+      "grad_norm": 0.3843582570552826,
+      "learning_rate": 4.38155063001445e-06,
+      "loss": 0.6119,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7245560318432334,
+      "grad_norm": 0.3423139154911041,
+      "learning_rate": 4.3806917501423026e-06,
+      "loss": 0.6578,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7250459277403551,
+      "grad_norm": 0.3562920391559601,
+      "learning_rate": 4.379832358598183e-06,
+      "loss": 0.644,
+      "step": 1480
+    },
+    {
+      "epoch": 0.725535823637477,
+      "grad_norm": 0.37188270688056946,
+      "learning_rate": 4.378972455615901e-06,
+      "loss": 0.6539,
+      "step": 1481
+    },
+    {
+      "epoch": 0.7260257195345989,
+      "grad_norm": 0.3575827479362488,
+      "learning_rate": 4.378112041429409e-06,
+      "loss": 0.6499,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7265156154317207,
+      "grad_norm": 0.35851728916168213,
+      "learning_rate": 4.3772511162727934e-06,
+      "loss": 0.6381,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7270055113288426,
+      "grad_norm": 0.35518619418144226,
+      "learning_rate": 4.376389680380286e-06,
+      "loss": 0.6302,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7274954072259645,
+      "grad_norm": 0.33862000703811646,
+      "learning_rate": 4.37552773398625e-06,
+      "loss": 0.6378,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7279853031230863,
+      "grad_norm": 0.3456350564956665,
+      "learning_rate": 4.374665277325195e-06,
+      "loss": 0.6266,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7284751990202082,
+      "grad_norm": 0.35843735933303833,
+      "learning_rate": 4.373802310631765e-06,
+      "loss": 0.6189,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7289650949173301,
+      "grad_norm": 0.35915929079055786,
+      "learning_rate": 4.372938834140742e-06,
+      "loss": 0.6305,
+      "step": 1488
+    },
+    {
+      "epoch": 0.729454990814452,
+      "grad_norm": 0.34460461139678955,
+      "learning_rate": 4.3720748480870515e-06,
+      "loss": 0.6325,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7299448867115738,
+      "grad_norm": 0.34665390849113464,
+      "learning_rate": 4.371210352705753e-06,
+      "loss": 0.6403,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7304347826086957,
+      "grad_norm": 0.34286579489707947,
+      "learning_rate": 4.370345348232046e-06,
+      "loss": 0.6241,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7309246785058175,
+      "grad_norm": 0.36061859130859375,
+      "learning_rate": 4.369479834901268e-06,
+      "loss": 0.647,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7314145744029393,
+      "grad_norm": 0.3598126173019409,
+      "learning_rate": 4.368613812948898e-06,
+      "loss": 0.6343,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7319044703000612,
+      "grad_norm": 0.3460848331451416,
+      "learning_rate": 4.367747282610548e-06,
+      "loss": 0.6104,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7323943661971831,
+      "grad_norm": 0.3402574360370636,
+      "learning_rate": 4.366880244121972e-06,
+      "loss": 0.6349,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7328842620943049,
+      "grad_norm": 0.37308433651924133,
+      "learning_rate": 4.366012697719064e-06,
+      "loss": 0.6387,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7333741579914268,
+      "grad_norm": 0.3392968773841858,
+      "learning_rate": 4.365144643637852e-06,
+      "loss": 0.6343,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7338640538885487,
+      "grad_norm": 0.3485359251499176,
+      "learning_rate": 4.364276082114504e-06,
+      "loss": 0.6412,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7343539497856706,
+      "grad_norm": 0.3518224358558655,
+      "learning_rate": 4.363407013385325e-06,
+      "loss": 0.6199,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7348438456827924,
+      "grad_norm": 0.35371819138526917,
+      "learning_rate": 4.36253743768676e-06,
+      "loss": 0.6297,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7353337415799143,
+      "grad_norm": 0.34904298186302185,
+      "learning_rate": 4.36166735525539e-06,
+      "loss": 0.5931,
+      "step": 1501
+    },
+    {
+      "epoch": 0.7358236374770362,
+      "grad_norm": 0.36501210927963257,
+      "learning_rate": 4.360796766327936e-06,
+      "loss": 0.6413,
+      "step": 1502
+    },
+    {
+      "epoch": 0.736313533374158,
+      "grad_norm": 0.356987327337265,
+      "learning_rate": 4.3599256711412555e-06,
+      "loss": 0.6476,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7368034292712798,
+      "grad_norm": 0.3617223799228668,
+      "learning_rate": 4.3590540699323415e-06,
+      "loss": 0.6283,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7372933251684017,
+      "grad_norm": 0.3590957820415497,
+      "learning_rate": 4.358181962938329e-06,
+      "loss": 0.633,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7377832210655236,
+      "grad_norm": 0.3524998426437378,
+      "learning_rate": 4.357309350396488e-06,
+      "loss": 0.6341,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7382731169626454,
+      "grad_norm": 0.3727060556411743,
+      "learning_rate": 4.356436232544227e-06,
+      "loss": 0.6502,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7387630128597673,
+      "grad_norm": 0.349979043006897,
+      "learning_rate": 4.35556260961909e-06,
+      "loss": 0.6538,
+      "step": 1508
+    },
+    {
+      "epoch": 0.7392529087568892,
+      "grad_norm": 0.3617153763771057,
+      "learning_rate": 4.354688481858762e-06,
+      "loss": 0.627,
+      "step": 1509
+    },
+    {
+      "epoch": 0.739742804654011,
+      "grad_norm": 0.36576417088508606,
+      "learning_rate": 4.353813849501062e-06,
+      "loss": 0.6394,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7402327005511329,
+      "grad_norm": 0.34674984216690063,
+      "learning_rate": 4.352938712783947e-06,
+      "loss": 0.6275,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7407225964482548,
+      "grad_norm": 0.3658475875854492,
+      "learning_rate": 4.352063071945514e-06,
+      "loss": 0.6237,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7412124923453766,
+      "grad_norm": 0.35605910420417786,
+      "learning_rate": 4.351186927223992e-06,
+      "loss": 0.6628,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7417023882424985,
+      "grad_norm": 0.3472442030906677,
+      "learning_rate": 4.350310278857754e-06,
+      "loss": 0.6531,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7421922841396204,
+      "grad_norm": 0.34654098749160767,
+      "learning_rate": 4.349433127085302e-06,
+      "loss": 0.6177,
+      "step": 1515
+    },
+    {
+      "epoch": 0.7426821800367422,
+      "grad_norm": 0.3483250141143799,
+      "learning_rate": 4.34855547214528e-06,
+      "loss": 0.5996,
+      "step": 1516
+    },
+    {
+      "epoch": 0.743172075933864,
+      "grad_norm": 0.3742973506450653,
+      "learning_rate": 4.34767731427647e-06,
+      "loss": 0.67,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7436619718309859,
+      "grad_norm": 0.36761683225631714,
+      "learning_rate": 4.346798653717787e-06,
+      "loss": 0.6544,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7441518677281078,
+      "grad_norm": 0.35679250955581665,
+      "learning_rate": 4.345919490708286e-06,
+      "loss": 0.627,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7446417636252296,
+      "grad_norm": 0.34713515639305115,
+      "learning_rate": 4.345039825487156e-06,
+      "loss": 0.6363,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7451316595223515,
+      "grad_norm": 0.3618803322315216,
+      "learning_rate": 4.344159658293723e-06,
+      "loss": 0.6641,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7456215554194734,
+      "grad_norm": 0.35535937547683716,
+      "learning_rate": 4.343278989367452e-06,
+      "loss": 0.6325,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7461114513165952,
+      "grad_norm": 0.3408360481262207,
+      "learning_rate": 4.342397818947943e-06,
+      "loss": 0.6324,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7466013472137171,
+      "grad_norm": 0.3660116493701935,
+      "learning_rate": 4.341516147274931e-06,
+      "loss": 0.6098,
+      "step": 1524
+    },
+    {
+      "epoch": 0.747091243110839,
+      "grad_norm": 0.3499428629875183,
+      "learning_rate": 4.34063397458829e-06,
+      "loss": 0.667,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7475811390079609,
+      "grad_norm": 0.3460999131202698,
+      "learning_rate": 4.339751301128028e-06,
+      "loss": 0.6339,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7480710349050826,
+      "grad_norm": 0.3715626895427704,
+      "learning_rate": 4.338868127134291e-06,
+      "loss": 0.6494,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7485609308022045,
+      "grad_norm": 0.35727787017822266,
+      "learning_rate": 4.3379844528473605e-06,
+      "loss": 0.6661,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7490508266993264,
+      "grad_norm": 0.34379470348358154,
+      "learning_rate": 4.337100278507654e-06,
+      "loss": 0.6311,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7495407225964482,
+      "grad_norm": 0.3414919972419739,
+      "learning_rate": 4.3362156043557246e-06,
+      "loss": 0.656,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7500306184935701,
+      "grad_norm": 0.3487052619457245,
+      "learning_rate": 4.335330430632263e-06,
+      "loss": 0.6433,
+      "step": 1531
+    },
+    {
+      "epoch": 0.750520514390692,
+      "grad_norm": 0.3398865759372711,
+      "learning_rate": 4.334444757578093e-06,
+      "loss": 0.6083,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7510104102878138,
+      "grad_norm": 0.3585461378097534,
+      "learning_rate": 4.3335585854341776e-06,
+      "loss": 0.6106,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7515003061849357,
+      "grad_norm": 0.3617606461048126,
+      "learning_rate": 4.332671914441614e-06,
+      "loss": 0.6305,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7519902020820576,
+      "grad_norm": 0.35764437913894653,
+      "learning_rate": 4.331784744841634e-06,
+      "loss": 0.6217,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7524800979791795,
+      "grad_norm": 0.3439844250679016,
+      "learning_rate": 4.330897076875606e-06,
+      "loss": 0.6109,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7529699938763013,
+      "grad_norm": 0.3562347888946533,
+      "learning_rate": 4.3300089107850365e-06,
+      "loss": 0.6455,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7534598897734232,
+      "grad_norm": 0.3550952076911926,
+      "learning_rate": 4.329120246811562e-06,
+      "loss": 0.6443,
+      "step": 1538
+    },
+    {
+      "epoch": 0.753949785670545,
+      "grad_norm": 0.35900068283081055,
+      "learning_rate": 4.328231085196959e-06,
+      "loss": 0.6399,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7544396815676668,
+      "grad_norm": 0.3578474819660187,
+      "learning_rate": 4.327341426183139e-06,
+      "loss": 0.6332,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7549295774647887,
+      "grad_norm": 0.37041589617729187,
+      "learning_rate": 4.326451270012146e-06,
+      "loss": 0.6453,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7554194733619106,
+      "grad_norm": 0.35958898067474365,
+      "learning_rate": 4.325560616926162e-06,
+      "loss": 0.6385,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7559093692590324,
+      "grad_norm": 0.3471086919307709,
+      "learning_rate": 4.324669467167504e-06,
+      "loss": 0.6272,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7563992651561543,
+      "grad_norm": 0.355865478515625,
+      "learning_rate": 4.323777820978622e-06,
+      "loss": 0.6502,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7568891610532762,
+      "grad_norm": 0.3631455898284912,
+      "learning_rate": 4.322885678602103e-06,
+      "loss": 0.6361,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7573790569503981,
+      "grad_norm": 0.3473595380783081,
+      "learning_rate": 4.321993040280669e-06,
+      "loss": 0.6271,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7578689528475199,
+      "grad_norm": 0.3650197684764862,
+      "learning_rate": 4.321099906257175e-06,
+      "loss": 0.6371,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7583588487446418,
+      "grad_norm": 0.3387010097503662,
+      "learning_rate": 4.320206276774613e-06,
+      "loss": 0.6436,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7588487446417637,
+      "grad_norm": 0.3699718713760376,
+      "learning_rate": 4.3193121520761105e-06,
+      "loss": 0.6585,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7593386405388854,
+      "grad_norm": 0.3527061343193054,
+      "learning_rate": 4.318417532404926e-06,
+      "loss": 0.6474,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7598285364360073,
+      "grad_norm": 0.35305193066596985,
+      "learning_rate": 4.3175224180044575e-06,
+      "loss": 0.6245,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7603184323331292,
+      "grad_norm": 0.352267861366272,
+      "learning_rate": 4.316626809118232e-06,
+      "loss": 0.6581,
+      "step": 1552
+    },
+    {
+      "epoch": 0.760808328230251,
+      "grad_norm": 0.35272592306137085,
+      "learning_rate": 4.3157307059899155e-06,
+      "loss": 0.631,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7612982241273729,
+      "grad_norm": 0.3434283137321472,
+      "learning_rate": 4.314834108863308e-06,
+      "loss": 0.6205,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7617881200244948,
+      "grad_norm": 0.35709479451179504,
+      "learning_rate": 4.313937017982342e-06,
+      "loss": 0.6205,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7622780159216167,
+      "grad_norm": 0.3480152189731598,
+      "learning_rate": 4.313039433591085e-06,
+      "loss": 0.6281,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7627679118187385,
+      "grad_norm": 0.3423363268375397,
+      "learning_rate": 4.3121413559337394e-06,
+      "loss": 0.6324,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7632578077158604,
+      "grad_norm": 0.3515855371952057,
+      "learning_rate": 4.311242785254641e-06,
+      "loss": 0.6476,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7637477036129823,
+      "grad_norm": 0.35674506425857544,
+      "learning_rate": 4.310343721798262e-06,
+      "loss": 0.6389,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7642375995101041,
+      "grad_norm": 0.3473707437515259,
+      "learning_rate": 4.309444165809204e-06,
+      "loss": 0.6393,
+      "step": 1560
+    },
+    {
+      "epoch": 0.764727495407226,
+      "grad_norm": 0.3437351584434509,
+      "learning_rate": 4.308544117532209e-06,
+      "loss": 0.6496,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7652173913043478,
+      "grad_norm": 0.35847991704940796,
+      "learning_rate": 4.307643577212145e-06,
+      "loss": 0.6237,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7657072872014696,
+      "grad_norm": 0.3546234667301178,
+      "learning_rate": 4.306742545094022e-06,
+      "loss": 0.6363,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7661971830985915,
+      "grad_norm": 0.3565414845943451,
+      "learning_rate": 4.305841021422978e-06,
+      "loss": 0.5923,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7666870789957134,
+      "grad_norm": 0.3588482737541199,
+      "learning_rate": 4.304939006444288e-06,
+      "loss": 0.6266,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7671769748928353,
+      "grad_norm": 0.3416237235069275,
+      "learning_rate": 4.3040365004033576e-06,
+      "loss": 0.6214,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7676668707899571,
+      "grad_norm": 0.3760249614715576,
+      "learning_rate": 4.30313350354573e-06,
+      "loss": 0.6285,
+      "step": 1567
+    },
+    {
+      "epoch": 0.768156766687079,
+      "grad_norm": 0.33408862352371216,
+      "learning_rate": 4.302230016117079e-06,
+      "loss": 0.6094,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7686466625842009,
+      "grad_norm": 0.3498431146144867,
+      "learning_rate": 4.301326038363213e-06,
+      "loss": 0.6187,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7691365584813227,
+      "grad_norm": 0.3693348467350006,
+      "learning_rate": 4.300421570530073e-06,
+      "loss": 0.6644,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7696264543784446,
+      "grad_norm": 0.3516463041305542,
+      "learning_rate": 4.299516612863734e-06,
+      "loss": 0.6295,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7701163502755665,
+      "grad_norm": 0.35464391112327576,
+      "learning_rate": 4.298611165610404e-06,
+      "loss": 0.6246,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7706062461726884,
+      "grad_norm": 0.3509035110473633,
+      "learning_rate": 4.297705229016423e-06,
+      "loss": 0.6263,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7710961420698101,
+      "grad_norm": 0.35598504543304443,
+      "learning_rate": 4.296798803328267e-06,
+      "loss": 0.6479,
+      "step": 1574
+    },
+    {
+      "epoch": 0.771586037966932,
+      "grad_norm": 0.37206852436065674,
+      "learning_rate": 4.295891888792543e-06,
+      "loss": 0.6134,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7720759338640539,
+      "grad_norm": 0.3516175448894501,
+      "learning_rate": 4.294984485655992e-06,
+      "loss": 0.6545,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7725658297611757,
+      "grad_norm": 0.36359652876853943,
+      "learning_rate": 4.294076594165487e-06,
+      "loss": 0.6332,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7730557256582976,
+      "grad_norm": 0.35042086243629456,
+      "learning_rate": 4.293168214568035e-06,
+      "loss": 0.6302,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7735456215554195,
+      "grad_norm": 0.35204851627349854,
+      "learning_rate": 4.292259347110772e-06,
+      "loss": 0.6727,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7740355174525413,
+      "grad_norm": 0.35548853874206543,
+      "learning_rate": 4.2913499920409735e-06,
+      "loss": 0.6345,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7745254133496632,
+      "grad_norm": 0.35779857635498047,
+      "learning_rate": 4.290440149606042e-06,
+      "loss": 0.6139,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7750153092467851,
+      "grad_norm": 0.3580545485019684,
+      "learning_rate": 4.289529820053515e-06,
+      "loss": 0.6419,
+      "step": 1582
+    },
+    {
+      "epoch": 0.775505205143907,
+      "grad_norm": 0.37659353017807007,
+      "learning_rate": 4.288619003631062e-06,
+      "loss": 0.6446,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7759951010410288,
+      "grad_norm": 0.3591335415840149,
+      "learning_rate": 4.287707700586486e-06,
+      "loss": 0.6224,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7764849969381507,
+      "grad_norm": 0.3432382345199585,
+      "learning_rate": 4.28679591116772e-06,
+      "loss": 0.6341,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7769748928352725,
+      "grad_norm": 0.35006436705589294,
+      "learning_rate": 4.285883635622831e-06,
+      "loss": 0.6045,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7774647887323943,
+      "grad_norm": 0.3341239094734192,
+      "learning_rate": 4.284970874200017e-06,
+      "loss": 0.6408,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7779546846295162,
+      "grad_norm": 0.34813687205314636,
+      "learning_rate": 4.284057627147612e-06,
+      "loss": 0.6298,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7784445805266381,
+      "grad_norm": 0.3582232594490051,
+      "learning_rate": 4.283143894714078e-06,
+      "loss": 0.6374,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7789344764237599,
+      "grad_norm": 0.35491639375686646,
+      "learning_rate": 4.28222967714801e-06,
+      "loss": 0.6355,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7794243723208818,
+      "grad_norm": 0.36043015122413635,
+      "learning_rate": 4.281314974698135e-06,
+      "loss": 0.6373,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7799142682180037,
+      "grad_norm": 0.3543197512626648,
+      "learning_rate": 4.280399787613314e-06,
+      "loss": 0.63,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7804041641151256,
+      "grad_norm": 0.3556041419506073,
+      "learning_rate": 4.279484116142536e-06,
+      "loss": 0.6066,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7808940600122474,
+      "grad_norm": 0.35114777088165283,
+      "learning_rate": 4.278567960534925e-06,
+      "loss": 0.6397,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7813839559093693,
+      "grad_norm": 0.35579121112823486,
+      "learning_rate": 4.277651321039736e-06,
+      "loss": 0.6348,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7818738518064912,
+      "grad_norm": 0.35150033235549927,
+      "learning_rate": 4.2767341979063535e-06,
+      "loss": 0.6416,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7823637477036129,
+      "grad_norm": 0.3624979853630066,
+      "learning_rate": 4.2758165913842975e-06,
+      "loss": 0.5853,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7828536436007348,
+      "grad_norm": 0.3627711236476898,
+      "learning_rate": 4.274898501723217e-06,
+      "loss": 0.6421,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7833435394978567,
+      "grad_norm": 0.36170175671577454,
+      "learning_rate": 4.273979929172892e-06,
+      "loss": 0.6371,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7838334353949785,
+      "grad_norm": 0.3567507863044739,
+      "learning_rate": 4.273060873983235e-06,
+      "loss": 0.6386,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7843233312921004,
+      "grad_norm": 0.3699304163455963,
+      "learning_rate": 4.27214133640429e-06,
+      "loss": 0.6442,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7848132271892223,
+      "grad_norm": 0.36581847071647644,
+      "learning_rate": 4.271221316686232e-06,
+      "loss": 0.6471,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7853031230863442,
+      "grad_norm": 0.35311195254325867,
+      "learning_rate": 4.270300815079365e-06,
+      "loss": 0.6282,
+      "step": 1603
+    },
+    {
+      "epoch": 0.785793018983466,
+      "grad_norm": 0.3268887400627136,
+      "learning_rate": 4.269379831834127e-06,
+      "loss": 0.6128,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7862829148805879,
+      "grad_norm": 0.3437340557575226,
+      "learning_rate": 4.268458367201088e-06,
+      "loss": 0.6396,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7867728107777098,
+      "grad_norm": 0.3491259813308716,
+      "learning_rate": 4.267536421430946e-06,
+      "loss": 0.6194,
+      "step": 1606
+    },
+    {
+      "epoch": 0.7872627066748316,
+      "grad_norm": 0.34736862778663635,
+      "learning_rate": 4.266613994774528e-06,
+      "loss": 0.6167,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7877526025719535,
+      "grad_norm": 0.35211804509162903,
+      "learning_rate": 4.265691087482799e-06,
+      "loss": 0.6222,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7882424984690753,
+      "grad_norm": 0.35701170563697815,
+      "learning_rate": 4.264767699806848e-06,
+      "loss": 0.6133,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7887323943661971,
+      "grad_norm": 0.33028626441955566,
+      "learning_rate": 4.263843831997899e-06,
+      "loss": 0.6248,
+      "step": 1610
+    },
+    {
+      "epoch": 0.789222290263319,
+      "grad_norm": 0.3628077208995819,
+      "learning_rate": 4.262919484307303e-06,
+      "loss": 0.6366,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7897121861604409,
+      "grad_norm": 0.3530922830104828,
+      "learning_rate": 4.2619946569865445e-06,
+      "loss": 0.6366,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7902020820575628,
+      "grad_norm": 0.3449934422969818,
+      "learning_rate": 4.261069350287237e-06,
+      "loss": 0.6086,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7906919779546846,
+      "grad_norm": 0.3515269458293915,
+      "learning_rate": 4.260143564461126e-06,
+      "loss": 0.5886,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7911818738518065,
+      "grad_norm": 0.3541713356971741,
+      "learning_rate": 4.259217299760083e-06,
+      "loss": 0.6526,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7916717697489284,
+      "grad_norm": 0.3405754268169403,
+      "learning_rate": 4.258290556436117e-06,
+      "loss": 0.6439,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7921616656460502,
+      "grad_norm": 0.35600221157073975,
+      "learning_rate": 4.257363334741361e-06,
+      "loss": 0.6257,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7926515615431721,
+      "grad_norm": 0.35318514704704285,
+      "learning_rate": 4.256435634928081e-06,
+      "loss": 0.6513,
+      "step": 1618
+    },
+    {
+      "epoch": 0.793141457440294,
+      "grad_norm": 0.34695348143577576,
+      "learning_rate": 4.255507457248672e-06,
+      "loss": 0.6401,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7936313533374157,
+      "grad_norm": 0.33103105425834656,
+      "learning_rate": 4.25457880195566e-06,
+      "loss": 0.6261,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7941212492345376,
+      "grad_norm": 0.35367587208747864,
+      "learning_rate": 4.2536496693016995e-06,
+      "loss": 0.6213,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7946111451316595,
+      "grad_norm": 0.35644322633743286,
+      "learning_rate": 4.252720059539577e-06,
+      "loss": 0.6512,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7951010410287814,
+      "grad_norm": 0.3602214753627777,
+      "learning_rate": 4.251789972922206e-06,
+      "loss": 0.6137,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7955909369259032,
+      "grad_norm": 0.36309823393821716,
+      "learning_rate": 4.250859409702632e-06,
+      "loss": 0.6002,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7960808328230251,
+      "grad_norm": 0.365477055311203,
+      "learning_rate": 4.24992837013403e-06,
+      "loss": 0.6633,
+      "step": 1625
+    },
+    {
+      "epoch": 0.796570728720147,
+      "grad_norm": 0.3455202877521515,
+      "learning_rate": 4.2489968544697045e-06,
+      "loss": 0.6502,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7970606246172688,
+      "grad_norm": 0.37106695771217346,
+      "learning_rate": 4.248064862963087e-06,
+      "loss": 0.6236,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7975505205143907,
+      "grad_norm": 0.3467376232147217,
+      "learning_rate": 4.247132395867742e-06,
+      "loss": 0.6214,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7980404164115126,
+      "grad_norm": 0.3580789268016815,
+      "learning_rate": 4.246199453437362e-06,
+      "loss": 0.6419,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7985303123086344,
+      "grad_norm": 0.3447851538658142,
+      "learning_rate": 4.2452660359257684e-06,
+      "loss": 0.6233,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7990202082057563,
+      "grad_norm": 0.36099758744239807,
+      "learning_rate": 4.244332143586912e-06,
+      "loss": 0.6256,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7995101041028781,
+      "grad_norm": 0.3572690188884735,
+      "learning_rate": 4.243397776674874e-06,
+      "loss": 0.5994,
+      "step": 1632
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35768821835517883,
+      "learning_rate": 4.242462935443862e-06,
+      "loss": 0.6054,
+      "step": 1633
+    },
+    {
+      "epoch": 0.8004898958971218,
+      "grad_norm": 0.3552550971508026,
+      "learning_rate": 4.241527620148216e-06,
+      "loss": 0.6365,
+      "step": 1634
+    },
+    {
+      "epoch": 0.8009797917942437,
+      "grad_norm": 0.37358149886131287,
+      "learning_rate": 4.240591831042402e-06,
+      "loss": 0.6517,
+      "step": 1635
+    },
+    {
+      "epoch": 0.8014696876913656,
+      "grad_norm": 0.34565654397010803,
+      "learning_rate": 4.239655568381018e-06,
+      "loss": 0.6153,
+      "step": 1636
+    },
+    {
+      "epoch": 0.8019595835884874,
+      "grad_norm": 0.3581133484840393,
+      "learning_rate": 4.238718832418787e-06,
+      "loss": 0.6264,
+      "step": 1637
+    },
+    {
+      "epoch": 0.8024494794856093,
+      "grad_norm": 0.3637845516204834,
+      "learning_rate": 4.237781623410565e-06,
+      "loss": 0.638,
+      "step": 1638
+    },
+    {
+      "epoch": 0.8029393753827312,
+      "grad_norm": 0.36229655146598816,
+      "learning_rate": 4.236843941611332e-06,
+      "loss": 0.6106,
+      "step": 1639
+    },
+    {
+      "epoch": 0.8034292712798531,
+      "grad_norm": 0.358633428812027,
+      "learning_rate": 4.2359057872762025e-06,
+      "loss": 0.6628,
+      "step": 1640
+    },
+    {
+      "epoch": 0.8039191671769749,
+      "grad_norm": 0.35790446400642395,
+      "learning_rate": 4.2349671606604125e-06,
+      "loss": 0.6229,
+      "step": 1641
+    },
+    {
+      "epoch": 0.8044090630740968,
+      "grad_norm": 0.37761494517326355,
+      "learning_rate": 4.234028062019333e-06,
+      "loss": 0.6671,
+      "step": 1642
+    },
+    {
+      "epoch": 0.8048989589712187,
+      "grad_norm": 0.341201514005661,
+      "learning_rate": 4.233088491608459e-06,
+      "loss": 0.631,
+      "step": 1643
+    },
+    {
+      "epoch": 0.8053888548683404,
+      "grad_norm": 0.34282517433166504,
+      "learning_rate": 4.232148449683418e-06,
+      "loss": 0.6492,
+      "step": 1644
+    },
+    {
+      "epoch": 0.8058787507654623,
+      "grad_norm": 0.35871371626853943,
+      "learning_rate": 4.231207936499959e-06,
+      "loss": 0.6257,
+      "step": 1645
+    },
+    {
+      "epoch": 0.8063686466625842,
+      "grad_norm": 0.3889302611351013,
+      "learning_rate": 4.230266952313968e-06,
+      "loss": 0.6251,
+      "step": 1646
+    },
+    {
+      "epoch": 0.806858542559706,
+      "grad_norm": 0.3646315336227417,
+      "learning_rate": 4.229325497381453e-06,
+      "loss": 0.6371,
+      "step": 1647
+    },
+    {
+      "epoch": 0.8073484384568279,
+      "grad_norm": 0.365292489528656,
+      "learning_rate": 4.22838357195855e-06,
+      "loss": 0.624,
+      "step": 1648
+    },
+    {
+      "epoch": 0.8078383343539498,
+      "grad_norm": 0.34523457288742065,
+      "learning_rate": 4.227441176301527e-06,
+      "loss": 0.6358,
+      "step": 1649
+    },
+    {
+      "epoch": 0.8083282302510717,
+      "grad_norm": 0.3475101590156555,
+      "learning_rate": 4.226498310666776e-06,
+      "loss": 0.6262,
+      "step": 1650
+    },
+    {
+      "epoch": 0.8088181261481935,
+      "grad_norm": 0.3511309027671814,
+      "learning_rate": 4.225554975310819e-06,
+      "loss": 0.6231,
+      "step": 1651
+    },
+    {
+      "epoch": 0.8093080220453154,
+      "grad_norm": 0.3588724732398987,
+      "learning_rate": 4.224611170490307e-06,
+      "loss": 0.6277,
+      "step": 1652
+    },
+    {
+      "epoch": 0.8097979179424373,
+      "grad_norm": 0.3514706790447235,
+      "learning_rate": 4.223666896462015e-06,
+      "loss": 0.622,
+      "step": 1653
+    },
+    {
+      "epoch": 0.8102878138395591,
+      "grad_norm": 0.3686499297618866,
+      "learning_rate": 4.222722153482849e-06,
+      "loss": 0.6774,
+      "step": 1654
+    },
+    {
+      "epoch": 0.810777709736681,
+      "grad_norm": 0.3523634374141693,
+      "learning_rate": 4.221776941809841e-06,
+      "loss": 0.6279,
+      "step": 1655
+    },
+    {
+      "epoch": 0.8112676056338028,
+      "grad_norm": 0.3380214273929596,
+      "learning_rate": 4.22083126170015e-06,
+      "loss": 0.6245,
+      "step": 1656
+    },
+    {
+      "epoch": 0.8117575015309246,
+      "grad_norm": 0.3579692840576172,
+      "learning_rate": 4.219885113411063e-06,
+      "loss": 0.614,
+      "step": 1657
+    },
+    {
+      "epoch": 0.8122473974280465,
+      "grad_norm": 0.3645648956298828,
+      "learning_rate": 4.218938497199996e-06,
+      "loss": 0.6338,
+      "step": 1658
+    },
+    {
+      "epoch": 0.8127372933251684,
+      "grad_norm": 0.34988459944725037,
+      "learning_rate": 4.21799141332449e-06,
+      "loss": 0.6407,
+      "step": 1659
+    },
+    {
+      "epoch": 0.8132271892222903,
+      "grad_norm": 0.35096099972724915,
+      "learning_rate": 4.217043862042215e-06,
+      "loss": 0.6379,
+      "step": 1660
+    },
+    {
+      "epoch": 0.8137170851194121,
+      "grad_norm": 0.37124431133270264,
+      "learning_rate": 4.216095843610967e-06,
+      "loss": 0.6242,
+      "step": 1661
+    },
+    {
+      "epoch": 0.814206981016534,
+      "grad_norm": 0.35415878891944885,
+      "learning_rate": 4.2151473582886686e-06,
+      "loss": 0.6206,
+      "step": 1662
+    },
+    {
+      "epoch": 0.8146968769136559,
+      "grad_norm": 0.3575012981891632,
+      "learning_rate": 4.214198406333371e-06,
+      "loss": 0.6119,
+      "step": 1663
+    },
+    {
+      "epoch": 0.8151867728107777,
+      "grad_norm": 0.34369057416915894,
+      "learning_rate": 4.213248988003251e-06,
+      "loss": 0.6221,
+      "step": 1664
+    },
+    {
+      "epoch": 0.8156766687078996,
+      "grad_norm": 0.36129137873649597,
+      "learning_rate": 4.212299103556614e-06,
+      "loss": 0.6347,
+      "step": 1665
+    },
+    {
+      "epoch": 0.8161665646050215,
+      "grad_norm": 0.35974010825157166,
+      "learning_rate": 4.211348753251891e-06,
+      "loss": 0.6499,
+      "step": 1666
+    },
+    {
+      "epoch": 0.8166564605021432,
+      "grad_norm": 0.3585021197795868,
+      "learning_rate": 4.210397937347638e-06,
+      "loss": 0.6611,
+      "step": 1667
+    },
+    {
+      "epoch": 0.8171463563992651,
+      "grad_norm": 0.36124205589294434,
+      "learning_rate": 4.2094466561025415e-06,
+      "loss": 0.6432,
+      "step": 1668
+    },
+    {
+      "epoch": 0.817636252296387,
+      "grad_norm": 0.3618478775024414,
+      "learning_rate": 4.208494909775413e-06,
+      "loss": 0.612,
+      "step": 1669
+    },
+    {
+      "epoch": 0.8181261481935089,
+      "grad_norm": 0.3520048260688782,
+      "learning_rate": 4.207542698625188e-06,
+      "loss": 0.6263,
+      "step": 1670
+    },
+    {
+      "epoch": 0.8186160440906307,
+      "grad_norm": 0.36058175563812256,
+      "learning_rate": 4.2065900229109315e-06,
+      "loss": 0.6556,
+      "step": 1671
+    },
+    {
+      "epoch": 0.8191059399877526,
+      "grad_norm": 0.34738948941230774,
+      "learning_rate": 4.205636882891835e-06,
+      "loss": 0.6356,
+      "step": 1672
+    },
+    {
+      "epoch": 0.8195958358848745,
+      "grad_norm": 0.3644973933696747,
+      "learning_rate": 4.204683278827214e-06,
+      "loss": 0.6252,
+      "step": 1673
+    },
+    {
+      "epoch": 0.8200857317819963,
+      "grad_norm": 0.3600853979587555,
+      "learning_rate": 4.203729210976513e-06,
+      "loss": 0.6384,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8205756276791182,
+      "grad_norm": 0.3561994135379791,
+      "learning_rate": 4.2027746795993e-06,
+      "loss": 0.6208,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8210655235762401,
+      "grad_norm": 0.3534051179885864,
+      "learning_rate": 4.20181968495527e-06,
+      "loss": 0.6327,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8215554194733619,
+      "grad_norm": 0.36424657702445984,
+      "learning_rate": 4.200864227304247e-06,
+      "loss": 0.6232,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8220453153704838,
+      "grad_norm": 0.35763734579086304,
+      "learning_rate": 4.199908306906176e-06,
+      "loss": 0.6281,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8225352112676056,
+      "grad_norm": 0.3533649146556854,
+      "learning_rate": 4.1989519240211305e-06,
+      "loss": 0.6262,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8230251071647275,
+      "grad_norm": 0.3638383746147156,
+      "learning_rate": 4.197995078909311e-06,
+      "loss": 0.6343,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8235150030618493,
+      "grad_norm": 0.348105788230896,
+      "learning_rate": 4.197037771831041e-06,
+      "loss": 0.6216,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8240048989589712,
+      "grad_norm": 0.33755728602409363,
+      "learning_rate": 4.1960800030467705e-06,
+      "loss": 0.6325,
+      "step": 1682
+    },
+    {
+      "epoch": 0.8244947948560931,
+      "grad_norm": 0.34193381667137146,
+      "learning_rate": 4.195121772817077e-06,
+      "loss": 0.6257,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8249846907532149,
+      "grad_norm": 0.3726844787597656,
+      "learning_rate": 4.194163081402662e-06,
+      "loss": 0.6393,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8254745866503368,
+      "grad_norm": 0.3632981777191162,
+      "learning_rate": 4.1932039290643534e-06,
+      "loss": 0.6377,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8259644825474587,
+      "grad_norm": 0.3556203544139862,
+      "learning_rate": 4.192244316063102e-06,
+      "loss": 0.621,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8264543784445805,
+      "grad_norm": 0.34743455052375793,
+      "learning_rate": 4.191284242659986e-06,
+      "loss": 0.6416,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8269442743417024,
+      "grad_norm": 0.3578084707260132,
+      "learning_rate": 4.190323709116211e-06,
+      "loss": 0.6362,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8274341702388243,
+      "grad_norm": 0.36085325479507446,
+      "learning_rate": 4.189362715693102e-06,
+      "loss": 0.6291,
+      "step": 1689
+    },
+    {
+      "epoch": 0.8279240661359462,
+      "grad_norm": 0.36349424719810486,
+      "learning_rate": 4.188401262652114e-06,
+      "loss": 0.5976,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8284139620330679,
+      "grad_norm": 0.3622356355190277,
+      "learning_rate": 4.187439350254826e-06,
+      "loss": 0.6348,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8289038579301898,
+      "grad_norm": 0.3604823052883148,
+      "learning_rate": 4.186476978762941e-06,
+      "loss": 0.6329,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8293937538273117,
+      "grad_norm": 0.36562973260879517,
+      "learning_rate": 4.185514148438288e-06,
+      "loss": 0.6238,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8298836497244335,
+      "grad_norm": 0.36667400598526,
+      "learning_rate": 4.184550859542819e-06,
+      "loss": 0.6411,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8303735456215554,
+      "grad_norm": 0.36434921622276306,
+      "learning_rate": 4.183587112338613e-06,
+      "loss": 0.6578,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8308634415186773,
+      "grad_norm": 0.3545973002910614,
+      "learning_rate": 4.182622907087872e-06,
+      "loss": 0.6012,
+      "step": 1696
+    },
+    {
+      "epoch": 0.8313533374157991,
+      "grad_norm": 0.3564360737800598,
+      "learning_rate": 4.181658244052924e-06,
+      "loss": 0.648,
+      "step": 1697
+    },
+    {
+      "epoch": 0.831843233312921,
+      "grad_norm": 0.3650110065937042,
+      "learning_rate": 4.180693123496221e-06,
+      "loss": 0.6444,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8323331292100429,
+      "grad_norm": 0.35943543910980225,
+      "learning_rate": 4.179727545680338e-06,
+      "loss": 0.6427,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8328230251071648,
+      "grad_norm": 0.3513593375682831,
+      "learning_rate": 4.178761510867977e-06,
+      "loss": 0.637,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8333129210042866,
+      "grad_norm": 0.3554759621620178,
+      "learning_rate": 4.177795019321963e-06,
+      "loss": 0.6416,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8338028169014085,
+      "grad_norm": 0.35935404896736145,
+      "learning_rate": 4.176828071305244e-06,
+      "loss": 0.636,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8342927127985303,
+      "grad_norm": 0.35404741764068604,
+      "learning_rate": 4.175860667080896e-06,
+      "loss": 0.6075,
+      "step": 1703
+    },
+    {
+      "epoch": 0.8347826086956521,
+      "grad_norm": 0.36948296427726746,
+      "learning_rate": 4.174892806912113e-06,
+      "loss": 0.6297,
+      "step": 1704
+    },
+    {
+      "epoch": 0.835272504592774,
+      "grad_norm": 0.34789204597473145,
+      "learning_rate": 4.1739244910622205e-06,
+      "loss": 0.6319,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8357624004898959,
+      "grad_norm": 0.372383713722229,
+      "learning_rate": 4.172955719794662e-06,
+      "loss": 0.6421,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8362522963870178,
+      "grad_norm": 0.37573274970054626,
+      "learning_rate": 4.171986493373008e-06,
+      "loss": 0.612,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8367421922841396,
+      "grad_norm": 0.3678482174873352,
+      "learning_rate": 4.171016812060951e-06,
+      "loss": 0.5881,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8372320881812615,
+      "grad_norm": 0.36841633915901184,
+      "learning_rate": 4.170046676122309e-06,
+      "loss": 0.5933,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8377219840783834,
+      "grad_norm": 0.33924609422683716,
+      "learning_rate": 4.169076085821021e-06,
+      "loss": 0.639,
+      "step": 1710
+    },
+    {
+      "epoch": 0.8382118799755052,
+      "grad_norm": 0.3626523017883301,
+      "learning_rate": 4.168105041421155e-06,
+      "loss": 0.6169,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8387017758726271,
+      "grad_norm": 0.3452063798904419,
+      "learning_rate": 4.167133543186896e-06,
+      "loss": 0.632,
+      "step": 1712
+    },
+    {
+      "epoch": 0.839191671769749,
+      "grad_norm": 0.34949636459350586,
+      "learning_rate": 4.166161591382557e-06,
+      "loss": 0.6253,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8396815676668707,
+      "grad_norm": 0.37437132000923157,
+      "learning_rate": 4.165189186272572e-06,
+      "loss": 0.6222,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8401714635639926,
+      "grad_norm": 0.35469987988471985,
+      "learning_rate": 4.164216328121499e-06,
+      "loss": 0.6246,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8406613594611145,
+      "grad_norm": 0.33854183554649353,
+      "learning_rate": 4.1632430171940196e-06,
+      "loss": 0.627,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8411512553582364,
+      "grad_norm": 0.357943058013916,
+      "learning_rate": 4.162269253754939e-06,
+      "loss": 0.6306,
+      "step": 1717
+    },
+    {
+      "epoch": 0.8416411512553582,
+      "grad_norm": 0.36788368225097656,
+      "learning_rate": 4.161295038069186e-06,
+      "loss": 0.6325,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8421310471524801,
+      "grad_norm": 0.35560229420661926,
+      "learning_rate": 4.160320370401809e-06,
+      "loss": 0.6382,
+      "step": 1719
+    },
+    {
+      "epoch": 0.842620943049602,
+      "grad_norm": 0.37370407581329346,
+      "learning_rate": 4.1593452510179834e-06,
+      "loss": 0.6626,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8431108389467238,
+      "grad_norm": 0.3486816883087158,
+      "learning_rate": 4.158369680183006e-06,
+      "loss": 0.6257,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8436007348438457,
+      "grad_norm": 0.38826611638069153,
+      "learning_rate": 4.157393658162296e-06,
+      "loss": 0.6456,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8440906307409676,
+      "grad_norm": 0.3426065146923065,
+      "learning_rate": 4.156417185221393e-06,
+      "loss": 0.6116,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8445805266380894,
+      "grad_norm": 0.34008917212486267,
+      "learning_rate": 4.155440261625966e-06,
+      "loss": 0.6092,
+      "step": 1724
+    },
+    {
+      "epoch": 0.8450704225352113,
+      "grad_norm": 0.36330732703208923,
+      "learning_rate": 4.154462887641801e-06,
+      "loss": 0.6345,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8455603184323331,
+      "grad_norm": 0.33931586146354675,
+      "learning_rate": 4.153485063534807e-06,
+      "loss": 0.6216,
+      "step": 1726
+    },
+    {
+      "epoch": 0.846050214329455,
+      "grad_norm": 0.35162895917892456,
+      "learning_rate": 4.152506789571018e-06,
+      "loss": 0.6104,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8465401102265768,
+      "grad_norm": 0.3457432985305786,
+      "learning_rate": 4.151528066016589e-06,
+      "loss": 0.6292,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8470300061236987,
+      "grad_norm": 0.3506641685962677,
+      "learning_rate": 4.150548893137796e-06,
+      "loss": 0.6454,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8475199020208206,
+      "grad_norm": 0.3497201204299927,
+      "learning_rate": 4.149569271201039e-06,
+      "loss": 0.6035,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8480097979179424,
+      "grad_norm": 0.36187872290611267,
+      "learning_rate": 4.148589200472839e-06,
+      "loss": 0.6305,
+      "step": 1731
+    },
+    {
+      "epoch": 0.8484996938150643,
+      "grad_norm": 0.3590366244316101,
+      "learning_rate": 4.147608681219842e-06,
+      "loss": 0.6045,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8489895897121862,
+      "grad_norm": 0.34981295466423035,
+      "learning_rate": 4.146627713708811e-06,
+      "loss": 0.6359,
+      "step": 1733
+    },
+    {
+      "epoch": 0.849479485609308,
+      "grad_norm": 0.35321328043937683,
+      "learning_rate": 4.145646298206636e-06,
+      "loss": 0.634,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8499693815064299,
+      "grad_norm": 0.36797085404396057,
+      "learning_rate": 4.144664434980325e-06,
+      "loss": 0.6321,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8504592774035518,
+      "grad_norm": 0.35578250885009766,
+      "learning_rate": 4.14368212429701e-06,
+      "loss": 0.6154,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8509491733006737,
+      "grad_norm": 0.33807629346847534,
+      "learning_rate": 4.142699366423945e-06,
+      "loss": 0.6026,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8514390691977954,
+      "grad_norm": 0.36523035168647766,
+      "learning_rate": 4.141716161628504e-06,
+      "loss": 0.5957,
+      "step": 1738
+    },
+    {
+      "epoch": 0.8519289650949173,
+      "grad_norm": 0.36455363035202026,
+      "learning_rate": 4.140732510178184e-06,
+      "loss": 0.6183,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8524188609920392,
+      "grad_norm": 0.3511872887611389,
+      "learning_rate": 4.1397484123406015e-06,
+      "loss": 0.6387,
+      "step": 1740
+    },
+    {
+      "epoch": 0.852908756889161,
+      "grad_norm": 0.34794583916664124,
+      "learning_rate": 4.1387638683834974e-06,
+      "loss": 0.6042,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8533986527862829,
+      "grad_norm": 0.36754393577575684,
+      "learning_rate": 4.137778878574732e-06,
+      "loss": 0.6099,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8538885486834048,
+      "grad_norm": 0.3553093373775482,
+      "learning_rate": 4.136793443182287e-06,
+      "loss": 0.6138,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8543784445805266,
+      "grad_norm": 0.35336676239967346,
+      "learning_rate": 4.135807562474267e-06,
+      "loss": 0.6279,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8548683404776485,
+      "grad_norm": 0.35361459851264954,
+      "learning_rate": 4.134821236718895e-06,
+      "loss": 0.6347,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8553582363747704,
+      "grad_norm": 0.35480567812919617,
+      "learning_rate": 4.133834466184516e-06,
+      "loss": 0.6186,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8558481322718923,
+      "grad_norm": 0.3604031801223755,
+      "learning_rate": 4.132847251139599e-06,
+      "loss": 0.6147,
+      "step": 1747
+    },
+    {
+      "epoch": 0.856338028169014,
+      "grad_norm": 0.35460978746414185,
+      "learning_rate": 4.131859591852729e-06,
+      "loss": 0.6355,
+      "step": 1748
+    },
+    {
+      "epoch": 0.856827924066136,
+      "grad_norm": 0.3537762761116028,
+      "learning_rate": 4.130871488592617e-06,
+      "loss": 0.603,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8573178199632578,
+      "grad_norm": 0.3467521071434021,
+      "learning_rate": 4.129882941628089e-06,
+      "loss": 0.6511,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8578077158603796,
+      "grad_norm": 0.3615700602531433,
+      "learning_rate": 4.128893951228098e-06,
+      "loss": 0.665,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8582976117575015,
+      "grad_norm": 0.36621513962745667,
+      "learning_rate": 4.1279045176617115e-06,
+      "loss": 0.6258,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8587875076546234,
+      "grad_norm": 0.3570707440376282,
+      "learning_rate": 4.126914641198123e-06,
+      "loss": 0.625,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8592774035517452,
+      "grad_norm": 0.3643128573894501,
+      "learning_rate": 4.125924322106643e-06,
+      "loss": 0.6517,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8597672994488671,
+      "grad_norm": 0.3491007387638092,
+      "learning_rate": 4.124933560656703e-06,
+      "loss": 0.6287,
+      "step": 1755
+    },
+    {
+      "epoch": 0.860257195345989,
+      "grad_norm": 0.3507339060306549,
+      "learning_rate": 4.123942357117856e-06,
+      "loss": 0.6517,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8607470912431109,
+      "grad_norm": 0.3513590097427368,
+      "learning_rate": 4.122950711759775e-06,
+      "loss": 0.635,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8612369871402327,
+      "grad_norm": 0.3566867709159851,
+      "learning_rate": 4.121958624852253e-06,
+      "loss": 0.6013,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8617268830373546,
+      "grad_norm": 0.35651153326034546,
+      "learning_rate": 4.120966096665201e-06,
+      "loss": 0.6442,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8622167789344765,
+      "grad_norm": 0.3672233819961548,
+      "learning_rate": 4.119973127468654e-06,
+      "loss": 0.6487,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8627066748315982,
+      "grad_norm": 0.35391879081726074,
+      "learning_rate": 4.118979717532764e-06,
+      "loss": 0.6348,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8631965707287201,
+      "grad_norm": 0.3441363573074341,
+      "learning_rate": 4.1179858671278044e-06,
+      "loss": 0.6458,
+      "step": 1762
+    },
+    {
+      "epoch": 0.863686466625842,
+      "grad_norm": 0.3333560824394226,
+      "learning_rate": 4.116991576524167e-06,
+      "loss": 0.6168,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8641763625229639,
+      "grad_norm": 0.34725016355514526,
+      "learning_rate": 4.115996845992366e-06,
+      "loss": 0.6328,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8646662584200857,
+      "grad_norm": 0.3482379913330078,
+      "learning_rate": 4.115001675803033e-06,
+      "loss": 0.6331,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8651561543172076,
+      "grad_norm": 0.3419699966907501,
+      "learning_rate": 4.114006066226919e-06,
+      "loss": 0.6225,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8656460502143295,
+      "grad_norm": 0.3558363914489746,
+      "learning_rate": 4.113010017534896e-06,
+      "loss": 0.6349,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8661359461114513,
+      "grad_norm": 0.3551356792449951,
+      "learning_rate": 4.112013529997954e-06,
+      "loss": 0.5967,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8666258420085732,
+      "grad_norm": 0.3523545265197754,
+      "learning_rate": 4.111016603887205e-06,
+      "loss": 0.6363,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8671157379056951,
+      "grad_norm": 0.3577764630317688,
+      "learning_rate": 4.1100192394738764e-06,
+      "loss": 0.621,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8676056338028169,
+      "grad_norm": 0.3645519018173218,
+      "learning_rate": 4.109021437029318e-06,
+      "loss": 0.6575,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8680955296999388,
+      "grad_norm": 0.34731167554855347,
+      "learning_rate": 4.108023196824998e-06,
+      "loss": 0.5998,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8685854255970606,
+      "grad_norm": 0.3572891056537628,
+      "learning_rate": 4.107024519132503e-06,
+      "loss": 0.6172,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8690753214941825,
+      "grad_norm": 0.3890896737575531,
+      "learning_rate": 4.10602540422354e-06,
+      "loss": 0.6429,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.3684241473674774,
+      "learning_rate": 4.105025852369933e-06,
+      "loss": 0.6395,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8700551132884262,
+      "grad_norm": 0.3604837656021118,
+      "learning_rate": 4.104025863843626e-06,
+      "loss": 0.6513,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8705450091855481,
+      "grad_norm": 0.35402780771255493,
+      "learning_rate": 4.103025438916682e-06,
+      "loss": 0.6296,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8710349050826699,
+      "grad_norm": 0.3665107786655426,
+      "learning_rate": 4.102024577861282e-06,
+      "loss": 0.6351,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8715248009797918,
+      "grad_norm": 0.3513564169406891,
+      "learning_rate": 4.101023280949726e-06,
+      "loss": 0.647,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8720146968769137,
+      "grad_norm": 0.3502311706542969,
+      "learning_rate": 4.100021548454434e-06,
+      "loss": 0.6501,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8725045927740355,
+      "grad_norm": 0.37091171741485596,
+      "learning_rate": 4.0990193806479426e-06,
+      "loss": 0.6542,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8729944886711574,
+      "grad_norm": 0.35671868920326233,
+      "learning_rate": 4.098016777802907e-06,
+      "loss": 0.6384,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8734843845682793,
+      "grad_norm": 0.3722166121006012,
+      "learning_rate": 4.097013740192101e-06,
+      "loss": 0.6407,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8739742804654012,
+      "grad_norm": 0.37666139006614685,
+      "learning_rate": 4.096010268088418e-06,
+      "loss": 0.6438,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8744641763625229,
+      "grad_norm": 0.3596649169921875,
+      "learning_rate": 4.095006361764867e-06,
+      "loss": 0.6393,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8749540722596448,
+      "grad_norm": 0.37628263235092163,
+      "learning_rate": 4.094002021494577e-06,
+      "loss": 0.6187,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8754439681567667,
+      "grad_norm": 0.3606310188770294,
+      "learning_rate": 4.092997247550796e-06,
+      "loss": 0.6279,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8759338640538885,
+      "grad_norm": 0.3513457179069519,
+      "learning_rate": 4.091992040206887e-06,
+      "loss": 0.6319,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8764237599510104,
+      "grad_norm": 0.34625834226608276,
+      "learning_rate": 4.090986399736333e-06,
+      "loss": 0.6373,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8769136558481323,
+      "grad_norm": 0.3485386371612549,
+      "learning_rate": 4.089980326412734e-06,
+      "loss": 0.6302,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8774035517452541,
+      "grad_norm": 0.35028499364852905,
+      "learning_rate": 4.088973820509811e-06,
+      "loss": 0.6411,
+      "step": 1791
+    },
+    {
+      "epoch": 0.877893447642376,
+      "grad_norm": 0.35805413126945496,
+      "learning_rate": 4.087966882301396e-06,
+      "loss": 0.6194,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8783833435394979,
+      "grad_norm": 0.35562488436698914,
+      "learning_rate": 4.086959512061444e-06,
+      "loss": 0.622,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8788732394366198,
+      "grad_norm": 0.33581987023353577,
+      "learning_rate": 4.085951710064027e-06,
+      "loss": 0.568,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8793631353337416,
+      "grad_norm": 0.34489452838897705,
+      "learning_rate": 4.084943476583332e-06,
+      "loss": 0.6419,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8798530312308634,
+      "grad_norm": 0.35095247626304626,
+      "learning_rate": 4.0839348118936665e-06,
+      "loss": 0.6247,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8803429271279853,
+      "grad_norm": 0.34744447469711304,
+      "learning_rate": 4.082925716269452e-06,
+      "loss": 0.6157,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8808328230251071,
+      "grad_norm": 0.35207876563072205,
+      "learning_rate": 4.08191618998523e-06,
+      "loss": 0.6346,
+      "step": 1798
+    },
+    {
+      "epoch": 0.881322718922229,
+      "grad_norm": 0.3440977931022644,
+      "learning_rate": 4.080906233315658e-06,
+      "loss": 0.6332,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8818126148193509,
+      "grad_norm": 0.3496447801589966,
+      "learning_rate": 4.0798958465355095e-06,
+      "loss": 0.6404,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8823025107164727,
+      "grad_norm": 0.3570871651172638,
+      "learning_rate": 4.078885029919678e-06,
+      "loss": 0.6242,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8827924066135946,
+      "grad_norm": 0.36829832196235657,
+      "learning_rate": 4.077873783743171e-06,
+      "loss": 0.6245,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8832823025107165,
+      "grad_norm": 0.36631497740745544,
+      "learning_rate": 4.076862108281116e-06,
+      "loss": 0.6343,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8837721984078384,
+      "grad_norm": 0.358917772769928,
+      "learning_rate": 4.075850003808752e-06,
+      "loss": 0.6369,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8842620943049602,
+      "grad_norm": 0.36538082361221313,
+      "learning_rate": 4.07483747060144e-06,
+      "loss": 0.6323,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8847519902020821,
+      "grad_norm": 0.3712175190448761,
+      "learning_rate": 4.073824508934656e-06,
+      "loss": 0.6186,
+      "step": 1806
+    },
+    {
+      "epoch": 0.885241886099204,
+      "grad_norm": 0.3590482771396637,
+      "learning_rate": 4.072811119083992e-06,
+      "loss": 0.6177,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8857317819963257,
+      "grad_norm": 0.3526355028152466,
+      "learning_rate": 4.0717973013251564e-06,
+      "loss": 0.6115,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8862216778934476,
+      "grad_norm": 0.35488682985305786,
+      "learning_rate": 4.070783055933975e-06,
+      "loss": 0.6342,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8867115737905695,
+      "grad_norm": 0.3698574900627136,
+      "learning_rate": 4.069768383186388e-06,
+      "loss": 0.5936,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8872014696876913,
+      "grad_norm": 0.3756300210952759,
+      "learning_rate": 4.068753283358455e-06,
+      "loss": 0.6303,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8876913655848132,
+      "grad_norm": 0.36526453495025635,
+      "learning_rate": 4.067737756726348e-06,
+      "loss": 0.6407,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8881812614819351,
+      "grad_norm": 0.3452131450176239,
+      "learning_rate": 4.066721803566358e-06,
+      "loss": 0.5891,
+      "step": 1813
+    },
+    {
+      "epoch": 0.888671157379057,
+      "grad_norm": 0.3645488917827606,
+      "learning_rate": 4.065705424154892e-06,
+      "loss": 0.6425,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8891610532761788,
+      "grad_norm": 0.3663477301597595,
+      "learning_rate": 4.064688618768472e-06,
+      "loss": 0.6577,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8896509491733007,
+      "grad_norm": 0.3654061257839203,
+      "learning_rate": 4.0636713876837345e-06,
+      "loss": 0.6422,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8901408450704226,
+      "grad_norm": 0.3733082711696625,
+      "learning_rate": 4.062653731177434e-06,
+      "loss": 0.6006,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8906307409675444,
+      "grad_norm": 0.3536593019962311,
+      "learning_rate": 4.06163564952644e-06,
+      "loss": 0.6134,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8911206368646662,
+      "grad_norm": 0.3528878092765808,
+      "learning_rate": 4.060617143007739e-06,
+      "loss": 0.6341,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8916105327617881,
+      "grad_norm": 0.35610130429267883,
+      "learning_rate": 4.05959821189843e-06,
+      "loss": 0.617,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8921004286589099,
+      "grad_norm": 0.35914796590805054,
+      "learning_rate": 4.058578856475731e-06,
+      "loss": 0.6271,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8925903245560318,
+      "grad_norm": 0.36907869577407837,
+      "learning_rate": 4.0575590770169724e-06,
+      "loss": 0.6567,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8930802204531537,
+      "grad_norm": 0.3563164174556732,
+      "learning_rate": 4.0565388737996016e-06,
+      "loss": 0.6443,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8935701163502756,
+      "grad_norm": 0.3560650944709778,
+      "learning_rate": 4.055518247101182e-06,
+      "loss": 0.6574,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8940600122473974,
+      "grad_norm": 0.3615463078022003,
+      "learning_rate": 4.0544971971993905e-06,
+      "loss": 0.625,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8945499081445193,
+      "grad_norm": 0.3564011752605438,
+      "learning_rate": 4.053475724372019e-06,
+      "loss": 0.6119,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8950398040416412,
+      "grad_norm": 0.3574683368206024,
+      "learning_rate": 4.052453828896978e-06,
+      "loss": 0.5942,
+      "step": 1827
+    },
+    {
+      "epoch": 0.895529699938763,
+      "grad_norm": 0.35468223690986633,
+      "learning_rate": 4.051431511052287e-06,
+      "loss": 0.6016,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8960195958358849,
+      "grad_norm": 0.3870941698551178,
+      "learning_rate": 4.0504087711160875e-06,
+      "loss": 0.6654,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8965094917330068,
+      "grad_norm": 0.3580857217311859,
+      "learning_rate": 4.049385609366628e-06,
+      "loss": 0.6343,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8969993876301287,
+      "grad_norm": 0.35600078105926514,
+      "learning_rate": 4.04836202608228e-06,
+      "loss": 0.6368,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8974892835272504,
+      "grad_norm": 0.3583797812461853,
+      "learning_rate": 4.047338021541522e-06,
+      "loss": 0.6369,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8979791794243723,
+      "grad_norm": 0.36189258098602295,
+      "learning_rate": 4.046313596022952e-06,
+      "loss": 0.6149,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8984690753214942,
+      "grad_norm": 0.3654255270957947,
+      "learning_rate": 4.0452887498052825e-06,
+      "loss": 0.6005,
+      "step": 1834
+    },
+    {
+      "epoch": 0.898958971218616,
+      "grad_norm": 0.36113041639328003,
+      "learning_rate": 4.0442634831673375e-06,
+      "loss": 0.6642,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8994488671157379,
+      "grad_norm": 0.3693462908267975,
+      "learning_rate": 4.043237796388056e-06,
+      "loss": 0.6199,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8999387630128598,
+      "grad_norm": 0.35426223278045654,
+      "learning_rate": 4.042211689746496e-06,
+      "loss": 0.6226,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9004286589099816,
+      "grad_norm": 0.3826047480106354,
+      "learning_rate": 4.041185163521822e-06,
+      "loss": 0.6345,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9009185548071035,
+      "grad_norm": 0.3614358603954315,
+      "learning_rate": 4.04015821799332e-06,
+      "loss": 0.6278,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9014084507042254,
+      "grad_norm": 0.35768815875053406,
+      "learning_rate": 4.039130853440385e-06,
+      "loss": 0.5985,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9018983466013473,
+      "grad_norm": 0.36194565892219543,
+      "learning_rate": 4.038103070142528e-06,
+      "loss": 0.6281,
+      "step": 1841
+    },
+    {
+      "epoch": 0.902388242498469,
+      "grad_norm": 0.3722645938396454,
+      "learning_rate": 4.037074868379374e-06,
+      "loss": 0.6285,
+      "step": 1842
+    },
+    {
+      "epoch": 0.902878138395591,
+      "grad_norm": 0.3487153947353363,
+      "learning_rate": 4.036046248430661e-06,
+      "loss": 0.6345,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9033680342927128,
+      "grad_norm": 0.35068756341934204,
+      "learning_rate": 4.035017210576242e-06,
+      "loss": 0.6214,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9038579301898346,
+      "grad_norm": 0.3582833409309387,
+      "learning_rate": 4.033987755096084e-06,
+      "loss": 0.642,
+      "step": 1845
+    },
+    {
+      "epoch": 0.9043478260869565,
+      "grad_norm": 0.3385012745857239,
+      "learning_rate": 4.032957882270264e-06,
+      "loss": 0.6125,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9048377219840784,
+      "grad_norm": 0.3503551781177521,
+      "learning_rate": 4.0319275923789765e-06,
+      "loss": 0.6447,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9053276178812002,
+      "grad_norm": 0.35716068744659424,
+      "learning_rate": 4.03089688570253e-06,
+      "loss": 0.6371,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9058175137783221,
+      "grad_norm": 0.36254581809043884,
+      "learning_rate": 4.029865762521342e-06,
+      "loss": 0.6388,
+      "step": 1849
+    },
+    {
+      "epoch": 0.906307409675444,
+      "grad_norm": 0.3596474528312683,
+      "learning_rate": 4.028834223115947e-06,
+      "loss": 0.6634,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9067973055725659,
+      "grad_norm": 0.35758960247039795,
+      "learning_rate": 4.027802267766991e-06,
+      "loss": 0.6292,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9072872014696877,
+      "grad_norm": 0.34626519680023193,
+      "learning_rate": 4.026769896755235e-06,
+      "loss": 0.6252,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9077770973668096,
+      "grad_norm": 0.34695932269096375,
+      "learning_rate": 4.025737110361551e-06,
+      "loss": 0.6054,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9082669932639315,
+      "grad_norm": 0.3792075216770172,
+      "learning_rate": 4.0247039088669255e-06,
+      "loss": 0.6513,
+      "step": 1854
+    },
+    {
+      "epoch": 0.9087568891610532,
+      "grad_norm": 0.36036989092826843,
+      "learning_rate": 4.023670292552456e-06,
+      "loss": 0.5844,
+      "step": 1855
+    },
+    {
+      "epoch": 0.9092467850581751,
+      "grad_norm": 0.3645704388618469,
+      "learning_rate": 4.022636261699356e-06,
+      "loss": 0.6277,
+      "step": 1856
+    },
+    {
+      "epoch": 0.909736680955297,
+      "grad_norm": 0.3849221169948578,
+      "learning_rate": 4.021601816588948e-06,
+      "loss": 0.6477,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9102265768524188,
+      "grad_norm": 0.356092244386673,
+      "learning_rate": 4.020566957502671e-06,
+      "loss": 0.6274,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9107164727495407,
+      "grad_norm": 0.37304049730300903,
+      "learning_rate": 4.019531684722074e-06,
+      "loss": 0.5975,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9112063686466626,
+      "grad_norm": 0.36938560009002686,
+      "learning_rate": 4.0184959985288194e-06,
+      "loss": 0.6473,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9116962645437845,
+      "grad_norm": 0.34542253613471985,
+      "learning_rate": 4.017459899204681e-06,
+      "loss": 0.6165,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9121861604409063,
+      "grad_norm": 0.3559544086456299,
+      "learning_rate": 4.016423387031548e-06,
+      "loss": 0.6188,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9126760563380282,
+      "grad_norm": 0.36169734597206116,
+      "learning_rate": 4.015386462291419e-06,
+      "loss": 0.5922,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9131659522351501,
+      "grad_norm": 0.396477609872818,
+      "learning_rate": 4.0143491252664055e-06,
+      "loss": 0.6297,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9136558481322719,
+      "grad_norm": 0.35498955845832825,
+      "learning_rate": 4.013311376238732e-06,
+      "loss": 0.624,
+      "step": 1865
+    },
+    {
+      "epoch": 0.9141457440293937,
+      "grad_norm": 0.3794139623641968,
+      "learning_rate": 4.012273215490734e-06,
+      "loss": 0.6572,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9146356399265156,
+      "grad_norm": 0.3604520261287689,
+      "learning_rate": 4.01123464330486e-06,
+      "loss": 0.63,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9151255358236374,
+      "grad_norm": 0.36176711320877075,
+      "learning_rate": 4.010195659963671e-06,
+      "loss": 0.6218,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9156154317207593,
+      "grad_norm": 0.360069215297699,
+      "learning_rate": 4.009156265749836e-06,
+      "loss": 0.6452,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9161053276178812,
+      "grad_norm": 0.3508584499359131,
+      "learning_rate": 4.008116460946141e-06,
+      "loss": 0.6217,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9165952235150031,
+      "grad_norm": 0.36335399746894836,
+      "learning_rate": 4.0070762458354805e-06,
+      "loss": 0.6274,
+      "step": 1871
+    },
+    {
+      "epoch": 0.9170851194121249,
+      "grad_norm": 0.3488854467868805,
+      "learning_rate": 4.0060356207008625e-06,
+      "loss": 0.6369,
+      "step": 1872
+    },
+    {
+      "epoch": 0.9175750153092468,
+      "grad_norm": 0.33648115396499634,
+      "learning_rate": 4.004994585825404e-06,
+      "loss": 0.5908,
+      "step": 1873
+    },
+    {
+      "epoch": 0.9180649112063687,
+      "grad_norm": 0.36617884039878845,
+      "learning_rate": 4.0039531414923365e-06,
+      "loss": 0.6186,
+      "step": 1874
+    },
+    {
+      "epoch": 0.9185548071034905,
+      "grad_norm": 0.3616359233856201,
+      "learning_rate": 4.0029112879850004e-06,
+      "loss": 0.6173,
+      "step": 1875
+    },
+    {
+      "epoch": 0.9190447030006124,
+      "grad_norm": 0.3591029644012451,
+      "learning_rate": 4.001869025586849e-06,
+      "loss": 0.6218,
+      "step": 1876
+    },
+    {
+      "epoch": 0.9195345988977343,
+      "grad_norm": 0.3484629988670349,
+      "learning_rate": 4.000826354581446e-06,
+      "loss": 0.624,
+      "step": 1877
+    },
+    {
+      "epoch": 0.920024494794856,
+      "grad_norm": 0.366221159696579,
+      "learning_rate": 3.9997832752524664e-06,
+      "loss": 0.6539,
+      "step": 1878
+    },
+    {
+      "epoch": 0.9205143906919779,
+      "grad_norm": 0.3641989529132843,
+      "learning_rate": 3.9987397878836976e-06,
+      "loss": 0.6072,
+      "step": 1879
+    },
+    {
+      "epoch": 0.9210042865890998,
+      "grad_norm": 0.3581095337867737,
+      "learning_rate": 3.997695892759035e-06,
+      "loss": 0.6071,
+      "step": 1880
+    },
+    {
+      "epoch": 0.9214941824862217,
+      "grad_norm": 0.3488609790802002,
+      "learning_rate": 3.996651590162488e-06,
+      "loss": 0.6015,
+      "step": 1881
+    },
+    {
+      "epoch": 0.9219840783833435,
+      "grad_norm": 0.3491750955581665,
+      "learning_rate": 3.995606880378176e-06,
+      "loss": 0.6407,
+      "step": 1882
+    },
+    {
+      "epoch": 0.9224739742804654,
+      "grad_norm": 0.3692910969257355,
+      "learning_rate": 3.994561763690328e-06,
+      "loss": 0.6435,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9229638701775873,
+      "grad_norm": 0.3535357117652893,
+      "learning_rate": 3.993516240383285e-06,
+      "loss": 0.6346,
+      "step": 1884
+    },
+    {
+      "epoch": 0.9234537660747091,
+      "grad_norm": 0.3640367090702057,
+      "learning_rate": 3.992470310741497e-06,
+      "loss": 0.6062,
+      "step": 1885
+    },
+    {
+      "epoch": 0.923943661971831,
+      "grad_norm": 0.35146427154541016,
+      "learning_rate": 3.9914239750495276e-06,
+      "loss": 0.6195,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9244335578689529,
+      "grad_norm": 0.36105847358703613,
+      "learning_rate": 3.990377233592048e-06,
+      "loss": 0.6403,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9249234537660747,
+      "grad_norm": 0.3821764886379242,
+      "learning_rate": 3.989330086653841e-06,
+      "loss": 0.6302,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9254133496631965,
+      "grad_norm": 0.3889336884021759,
+      "learning_rate": 3.988282534519799e-06,
+      "loss": 0.6222,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9259032455603184,
+      "grad_norm": 0.3606071174144745,
+      "learning_rate": 3.9872345774749255e-06,
+      "loss": 0.6098,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9263931414574403,
+      "grad_norm": 0.3542819619178772,
+      "learning_rate": 3.986186215804334e-06,
+      "loss": 0.6222,
+      "step": 1891
+    },
+    {
+      "epoch": 0.9268830373545621,
+      "grad_norm": 0.35070300102233887,
+      "learning_rate": 3.985137449793248e-06,
+      "loss": 0.6436,
+      "step": 1892
+    },
+    {
+      "epoch": 0.927372933251684,
+      "grad_norm": 0.35780608654022217,
+      "learning_rate": 3.984088279727e-06,
+      "loss": 0.6206,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9278628291488059,
+      "grad_norm": 0.35544446110725403,
+      "learning_rate": 3.9830387058910345e-06,
+      "loss": 0.6367,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9283527250459277,
+      "grad_norm": 0.34838807582855225,
+      "learning_rate": 3.981988728570904e-06,
+      "loss": 0.6406,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9288426209430496,
+      "grad_norm": 0.35056522488594055,
+      "learning_rate": 3.980938348052272e-06,
+      "loss": 0.6362,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9293325168401715,
+      "grad_norm": 0.36799871921539307,
+      "learning_rate": 3.97988756462091e-06,
+      "loss": 0.6235,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9298224127372934,
+      "grad_norm": 0.37848344445228577,
+      "learning_rate": 3.978836378562702e-06,
+      "loss": 0.6282,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9303123086344152,
+      "grad_norm": 0.36926591396331787,
+      "learning_rate": 3.977784790163638e-06,
+      "loss": 0.635,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9308022045315371,
+      "grad_norm": 0.35432732105255127,
+      "learning_rate": 3.976732799709821e-06,
+      "loss": 0.6322,
+      "step": 1900
+    },
+    {
+      "epoch": 0.931292100428659,
+      "grad_norm": 0.36038362979888916,
+      "learning_rate": 3.97568040748746e-06,
+      "loss": 0.6028,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9317819963257807,
+      "grad_norm": 0.3726102113723755,
+      "learning_rate": 3.974627613782876e-06,
+      "loss": 0.6231,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9322718922229026,
+      "grad_norm": 0.36531907320022583,
+      "learning_rate": 3.973574418882498e-06,
+      "loss": 0.6362,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9327617881200245,
+      "grad_norm": 0.3725537955760956,
+      "learning_rate": 3.9725208230728655e-06,
+      "loss": 0.6418,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9332516840171463,
+      "grad_norm": 0.3566921353340149,
+      "learning_rate": 3.971466826640623e-06,
+      "loss": 0.644,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9337415799142682,
+      "grad_norm": 0.35900595784187317,
+      "learning_rate": 3.970412429872529e-06,
+      "loss": 0.6519,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9342314758113901,
+      "grad_norm": 0.36603233218193054,
+      "learning_rate": 3.969357633055448e-06,
+      "loss": 0.6341,
+      "step": 1907
+    },
+    {
+      "epoch": 0.934721371708512,
+      "grad_norm": 0.3567545413970947,
+      "learning_rate": 3.968302436476356e-06,
+      "loss": 0.6285,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9352112676056338,
+      "grad_norm": 0.3554147183895111,
+      "learning_rate": 3.967246840422334e-06,
+      "loss": 0.6425,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9357011635027557,
+      "grad_norm": 0.36624595522880554,
+      "learning_rate": 3.966190845180574e-06,
+      "loss": 0.5982,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9361910593998776,
+      "grad_norm": 0.36004313826560974,
+      "learning_rate": 3.9651344510383775e-06,
+      "loss": 0.6278,
+      "step": 1911
+    },
+    {
+      "epoch": 0.9366809552969994,
+      "grad_norm": 0.35689929127693176,
+      "learning_rate": 3.964077658283151e-06,
+      "loss": 0.6367,
+      "step": 1912
+    },
+    {
+      "epoch": 0.9371708511941212,
+      "grad_norm": 0.35063254833221436,
+      "learning_rate": 3.963020467202413e-06,
+      "loss": 0.6005,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9376607470912431,
+      "grad_norm": 0.37015673518180847,
+      "learning_rate": 3.96196287808379e-06,
+      "loss": 0.6258,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9381506429883649,
+      "grad_norm": 0.3625577390193939,
+      "learning_rate": 3.960904891215015e-06,
+      "loss": 0.6365,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9386405388854868,
+      "grad_norm": 0.3661051392555237,
+      "learning_rate": 3.9598465068839285e-06,
+      "loss": 0.6044,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9391304347826087,
+      "grad_norm": 0.362802118062973,
+      "learning_rate": 3.958787725378483e-06,
+      "loss": 0.635,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9396203306797306,
+      "grad_norm": 0.37044766545295715,
+      "learning_rate": 3.957728546986735e-06,
+      "loss": 0.6189,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9401102265768524,
+      "grad_norm": 0.3538433611392975,
+      "learning_rate": 3.9566689719968535e-06,
+      "loss": 0.609,
+      "step": 1919
+    },
+    {
+      "epoch": 0.9406001224739743,
+      "grad_norm": 0.3680160343647003,
+      "learning_rate": 3.955609000697109e-06,
+      "loss": 0.6498,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9410900183710962,
+      "grad_norm": 0.37171587347984314,
+      "learning_rate": 3.954548633375887e-06,
+      "loss": 0.6175,
+      "step": 1921
+    },
+    {
+      "epoch": 0.941579914268218,
+      "grad_norm": 0.33701273798942566,
+      "learning_rate": 3.953487870321674e-06,
+      "loss": 0.5954,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9420698101653399,
+      "grad_norm": 0.3595403730869293,
+      "learning_rate": 3.95242671182307e-06,
+      "loss": 0.6421,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9425597060624618,
+      "grad_norm": 0.34721899032592773,
+      "learning_rate": 3.951365158168778e-06,
+      "loss": 0.639,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9430496019595835,
+      "grad_norm": 0.3664315640926361,
+      "learning_rate": 3.950303209647613e-06,
+      "loss": 0.6145,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9435394978567054,
+      "grad_norm": 0.3695625066757202,
+      "learning_rate": 3.949240866548492e-06,
+      "loss": 0.6511,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9440293937538273,
+      "grad_norm": 0.3734401762485504,
+      "learning_rate": 3.948178129160444e-06,
+      "loss": 0.627,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9445192896509492,
+      "grad_norm": 0.3736937642097473,
+      "learning_rate": 3.947114997772602e-06,
+      "loss": 0.6006,
+      "step": 1928
+    },
+    {
+      "epoch": 0.945009185548071,
+      "grad_norm": 0.366105318069458,
+      "learning_rate": 3.946051472674211e-06,
+      "loss": 0.6031,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9454990814451929,
+      "grad_norm": 0.3569345474243164,
+      "learning_rate": 3.944987554154616e-06,
+      "loss": 0.6117,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9459889773423148,
+      "grad_norm": 0.35967040061950684,
+      "learning_rate": 3.943923242503275e-06,
+      "loss": 0.6417,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9464788732394366,
+      "grad_norm": 0.365779846906662,
+      "learning_rate": 3.94285853800975e-06,
+      "loss": 0.6189,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9469687691365585,
+      "grad_norm": 0.35282260179519653,
+      "learning_rate": 3.941793440963712e-06,
+      "loss": 0.6236,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9474586650336804,
+      "grad_norm": 0.3678666055202484,
+      "learning_rate": 3.940727951654936e-06,
+      "loss": 0.6479,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9479485609308022,
+      "grad_norm": 0.35487714409828186,
+      "learning_rate": 3.939662070373306e-06,
+      "loss": 0.6344,
+      "step": 1935
+    },
+    {
+      "epoch": 0.948438456827924,
+      "grad_norm": 0.3745673894882202,
+      "learning_rate": 3.938595797408812e-06,
+      "loss": 0.6327,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9489283527250459,
+      "grad_norm": 0.34707069396972656,
+      "learning_rate": 3.93752913305155e-06,
+      "loss": 0.6264,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9494182486221678,
+      "grad_norm": 0.3721786141395569,
+      "learning_rate": 3.936462077591722e-06,
+      "loss": 0.6469,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9499081445192896,
+      "grad_norm": 0.36948245763778687,
+      "learning_rate": 3.93539463131964e-06,
+      "loss": 0.6323,
+      "step": 1939
+    },
+    {
+      "epoch": 0.9503980404164115,
+      "grad_norm": 0.3576546907424927,
+      "learning_rate": 3.934326794525718e-06,
+      "loss": 0.6413,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9508879363135334,
+      "grad_norm": 0.35429394245147705,
+      "learning_rate": 3.933258567500477e-06,
+      "loss": 0.6133,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9513778322106552,
+      "grad_norm": 0.34951600432395935,
+      "learning_rate": 3.932189950534548e-06,
+      "loss": 0.6254,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9518677281077771,
+      "grad_norm": 0.35437116026878357,
+      "learning_rate": 3.931120943918661e-06,
+      "loss": 0.6434,
+      "step": 1943
+    },
+    {
+      "epoch": 0.952357624004899,
+      "grad_norm": 0.35042843222618103,
+      "learning_rate": 3.9300515479436595e-06,
+      "loss": 0.6221,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9528475199020208,
+      "grad_norm": 0.34926679730415344,
+      "learning_rate": 3.928981762900489e-06,
+      "loss": 0.5836,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9533374157991427,
+      "grad_norm": 0.3715307414531708,
+      "learning_rate": 3.927911589080201e-06,
+      "loss": 0.6155,
+      "step": 1946
+    },
+    {
+      "epoch": 0.9538273116962646,
+      "grad_norm": 0.35882413387298584,
+      "learning_rate": 3.926841026773952e-06,
+      "loss": 0.6186,
+      "step": 1947
+    },
+    {
+      "epoch": 0.9543172075933865,
+      "grad_norm": 0.37663209438323975,
+      "learning_rate": 3.925770076273008e-06,
+      "loss": 0.6565,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9548071034905082,
+      "grad_norm": 0.36306554079055786,
+      "learning_rate": 3.924698737868737e-06,
+      "loss": 0.6346,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9552969993876301,
+      "grad_norm": 0.3902411162853241,
+      "learning_rate": 3.923627011852612e-06,
+      "loss": 0.584,
+      "step": 1950
+    },
+    {
+      "epoch": 0.955786895284752,
+      "grad_norm": 0.36925509572029114,
+      "learning_rate": 3.922554898516216e-06,
+      "loss": 0.6358,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9562767911818738,
+      "grad_norm": 0.3701590597629547,
+      "learning_rate": 3.9214823981512305e-06,
+      "loss": 0.6185,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9567666870789957,
+      "grad_norm": 0.36690884828567505,
+      "learning_rate": 3.920409511049451e-06,
+      "loss": 0.6318,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9572565829761176,
+      "grad_norm": 0.38318440318107605,
+      "learning_rate": 3.919336237502769e-06,
+      "loss": 0.6082,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9577464788732394,
+      "grad_norm": 0.37501823902130127,
+      "learning_rate": 3.918262577803188e-06,
+      "loss": 0.6186,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9582363747703613,
+      "grad_norm": 0.34495916962623596,
+      "learning_rate": 3.917188532242814e-06,
+      "loss": 0.5996,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9587262706674832,
+      "grad_norm": 0.35945406556129456,
+      "learning_rate": 3.916114101113857e-06,
+      "loss": 0.5956,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9592161665646051,
+      "grad_norm": 0.36524417996406555,
+      "learning_rate": 3.915039284708634e-06,
+      "loss": 0.6537,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9597060624617268,
+      "grad_norm": 0.3438582420349121,
+      "learning_rate": 3.913964083319566e-06,
+      "loss": 0.5874,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9601959583588487,
+      "grad_norm": 0.3767183721065521,
+      "learning_rate": 3.912888497239177e-06,
+      "loss": 0.6268,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9606858542559706,
+      "grad_norm": 0.3726462125778198,
+      "learning_rate": 3.911812526760099e-06,
+      "loss": 0.6273,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9611757501530924,
+      "grad_norm": 0.3737446367740631,
+      "learning_rate": 3.910736172175066e-06,
+      "loss": 0.6267,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9616656460502143,
+      "grad_norm": 0.3493219017982483,
+      "learning_rate": 3.909659433776918e-06,
+      "loss": 0.618,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9621555419473362,
+      "grad_norm": 0.37235260009765625,
+      "learning_rate": 3.908582311858597e-06,
+      "loss": 0.6447,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9626454378444581,
+      "grad_norm": 0.37288960814476013,
+      "learning_rate": 3.907504806713154e-06,
+      "loss": 0.6307,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9631353337415799,
+      "grad_norm": 0.3650355041027069,
+      "learning_rate": 3.906426918633738e-06,
+      "loss": 0.624,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9636252296387018,
+      "grad_norm": 0.34496167302131653,
+      "learning_rate": 3.905348647913607e-06,
+      "loss": 0.6037,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9641151255358237,
+      "grad_norm": 0.37034958600997925,
+      "learning_rate": 3.904269994846122e-06,
+      "loss": 0.608,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9646050214329455,
+      "grad_norm": 0.3597949743270874,
+      "learning_rate": 3.903190959724748e-06,
+      "loss": 0.5981,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9650949173300674,
+      "grad_norm": 0.3547936975955963,
+      "learning_rate": 3.902111542843052e-06,
+      "loss": 0.6195,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9655848132271893,
+      "grad_norm": 0.3734746277332306,
+      "learning_rate": 3.9010317444947074e-06,
+      "loss": 0.6278,
+      "step": 1971
+    },
+    {
+      "epoch": 0.966074709124311,
+      "grad_norm": 0.3595636487007141,
+      "learning_rate": 3.899951564973491e-06,
+      "loss": 0.6236,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9665646050214329,
+      "grad_norm": 0.3781930208206177,
+      "learning_rate": 3.898871004573282e-06,
+      "loss": 0.6224,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9670545009185548,
+      "grad_norm": 0.35438016057014465,
+      "learning_rate": 3.897790063588065e-06,
+      "loss": 0.6445,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9675443968156767,
+      "grad_norm": 0.361677348613739,
+      "learning_rate": 3.896708742311925e-06,
+      "loss": 0.6271,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9680342927127985,
+      "grad_norm": 0.3686682879924774,
+      "learning_rate": 3.895627041039055e-06,
+      "loss": 0.6232,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9685241886099204,
+      "grad_norm": 0.3676711320877075,
+      "learning_rate": 3.894544960063748e-06,
+      "loss": 0.6196,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9690140845070423,
+      "grad_norm": 0.3637702167034149,
+      "learning_rate": 3.893462499680399e-06,
+      "loss": 0.6257,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9695039804041641,
+      "grad_norm": 0.3684902489185333,
+      "learning_rate": 3.892379660183512e-06,
+      "loss": 0.6271,
+      "step": 1979
+    },
+    {
+      "epoch": 0.969993876301286,
+      "grad_norm": 0.35545337200164795,
+      "learning_rate": 3.891296441867689e-06,
+      "loss": 0.6133,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9704837721984079,
+      "grad_norm": 0.3473877012729645,
+      "learning_rate": 3.890212845027637e-06,
+      "loss": 0.6283,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9709736680955297,
+      "grad_norm": 0.3448870778083801,
+      "learning_rate": 3.8891288699581655e-06,
+      "loss": 0.6212,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9714635639926515,
+      "grad_norm": 0.37230542302131653,
+      "learning_rate": 3.888044516954187e-06,
+      "loss": 0.6195,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9719534598897734,
+      "grad_norm": 0.3552440106868744,
+      "learning_rate": 3.886959786310718e-06,
+      "loss": 0.663,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9724433557868953,
+      "grad_norm": 0.36294302344322205,
+      "learning_rate": 3.885874678322874e-06,
+      "loss": 0.6145,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9729332516840171,
+      "grad_norm": 0.3489057719707489,
+      "learning_rate": 3.884789193285879e-06,
+      "loss": 0.6149,
+      "step": 1986
+    },
+    {
+      "epoch": 0.973423147581139,
+      "grad_norm": 0.3702811002731323,
+      "learning_rate": 3.8837033314950544e-06,
+      "loss": 0.6065,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9739130434782609,
+      "grad_norm": 0.35552167892456055,
+      "learning_rate": 3.882617093245826e-06,
+      "loss": 0.6494,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9744029393753827,
+      "grad_norm": 0.3572336435317993,
+      "learning_rate": 3.881530478833724e-06,
+      "loss": 0.6326,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9748928352725046,
+      "grad_norm": 0.3529970645904541,
+      "learning_rate": 3.880443488554377e-06,
+      "loss": 0.6277,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9753827311696265,
+      "grad_norm": 0.37120121717453003,
+      "learning_rate": 3.879356122703518e-06,
+      "loss": 0.6376,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9758726270667483,
+      "grad_norm": 0.38676661252975464,
+      "learning_rate": 3.878268381576984e-06,
+      "loss": 0.6096,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9763625229638702,
+      "grad_norm": 0.3568069040775299,
+      "learning_rate": 3.87718026547071e-06,
+      "loss": 0.5951,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9768524188609921,
+      "grad_norm": 0.35809725522994995,
+      "learning_rate": 3.876091774680737e-06,
+      "loss": 0.5991,
+      "step": 1994
+    },
+    {
+      "epoch": 0.977342314758114,
+      "grad_norm": 0.367899090051651,
+      "learning_rate": 3.875002909503205e-06,
+      "loss": 0.5961,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9778322106552357,
+      "grad_norm": 0.36738499999046326,
+      "learning_rate": 3.873913670234358e-06,
+      "loss": 0.5969,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9783221065523576,
+      "grad_norm": 0.36432838439941406,
+      "learning_rate": 3.87282405717054e-06,
+      "loss": 0.6235,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9788120024494795,
+      "grad_norm": 0.3667569160461426,
+      "learning_rate": 3.8717340706081975e-06,
+      "loss": 0.6168,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9793018983466013,
+      "grad_norm": 0.3636808395385742,
+      "learning_rate": 3.870643710843878e-06,
+      "loss": 0.6294,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9797917942437232,
+      "grad_norm": 0.3527030646800995,
+      "learning_rate": 3.869552978174233e-06,
+      "loss": 0.5993,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9802816901408451,
+      "grad_norm": 0.35959500074386597,
+      "learning_rate": 3.868461872896011e-06,
+      "loss": 0.6486,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9807715860379669,
+      "grad_norm": 0.36117956042289734,
+      "learning_rate": 3.8673703953060685e-06,
+      "loss": 0.6335,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9812614819350888,
+      "grad_norm": 0.3546423017978668,
+      "learning_rate": 3.8662785457013544e-06,
+      "loss": 0.6193,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9817513778322107,
+      "grad_norm": 0.3523540496826172,
+      "learning_rate": 3.865186324378927e-06,
+      "loss": 0.6072,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9822412737293326,
+      "grad_norm": 0.3531913757324219,
+      "learning_rate": 3.8640937316359415e-06,
+      "loss": 0.627,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9827311696264543,
+      "grad_norm": 0.36261194944381714,
+      "learning_rate": 3.863000767769655e-06,
+      "loss": 0.6193,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9832210655235762,
+      "grad_norm": 0.37107813358306885,
+      "learning_rate": 3.861907433077425e-06,
+      "loss": 0.613,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9837109614206981,
+      "grad_norm": 0.3672920763492584,
+      "learning_rate": 3.860813727856712e-06,
+      "loss": 0.6213,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9842008573178199,
+      "grad_norm": 0.35620445013046265,
+      "learning_rate": 3.859719652405074e-06,
+      "loss": 0.6233,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9846907532149418,
+      "grad_norm": 0.380731999874115,
+      "learning_rate": 3.858625207020173e-06,
+      "loss": 0.5993,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9851806491120637,
+      "grad_norm": 0.3719562888145447,
+      "learning_rate": 3.85753039199977e-06,
+      "loss": 0.6203,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9856705450091855,
+      "grad_norm": 0.3641625940799713,
+      "learning_rate": 3.856435207641725e-06,
+      "loss": 0.6091,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9861604409063074,
+      "grad_norm": 0.3648725152015686,
+      "learning_rate": 3.855339654244003e-06,
+      "loss": 0.6306,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9866503368034293,
+      "grad_norm": 0.3548206090927124,
+      "learning_rate": 3.8542437321046654e-06,
+      "loss": 0.6358,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9871402327005512,
+      "grad_norm": 0.3661384880542755,
+      "learning_rate": 3.853147441521875e-06,
+      "loss": 0.6308,
+      "step": 2015
+    },
+    {
+      "epoch": 0.987630128597673,
+      "grad_norm": 0.35083141922950745,
+      "learning_rate": 3.8520507827938955e-06,
+      "loss": 0.6251,
+      "step": 2016
+    },
+    {
+      "epoch": 0.9881200244947949,
+      "grad_norm": 0.37429776787757874,
+      "learning_rate": 3.850953756219089e-06,
+      "loss": 0.6421,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9886099203919168,
+      "grad_norm": 0.3655894100666046,
+      "learning_rate": 3.849856362095921e-06,
+      "loss": 0.6497,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9890998162890385,
+      "grad_norm": 0.3669481575489044,
+      "learning_rate": 3.848758600722953e-06,
+      "loss": 0.6157,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9895897121861604,
+      "grad_norm": 0.3566364347934723,
+      "learning_rate": 3.847660472398849e-06,
+      "loss": 0.6148,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9900796080832823,
+      "grad_norm": 0.3615729808807373,
+      "learning_rate": 3.846561977422373e-06,
+      "loss": 0.6266,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9905695039804041,
+      "grad_norm": 0.37268856167793274,
+      "learning_rate": 3.845463116092387e-06,
+      "loss": 0.6189,
+      "step": 2022
+    },
+    {
+      "epoch": 0.991059399877526,
+      "grad_norm": 0.37262001633644104,
+      "learning_rate": 3.844363888707855e-06,
+      "loss": 0.6033,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9915492957746479,
+      "grad_norm": 0.3532877564430237,
+      "learning_rate": 3.843264295567836e-06,
+      "loss": 0.6293,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9920391916717698,
+      "grad_norm": 0.370309978723526,
+      "learning_rate": 3.842164336971496e-06,
+      "loss": 0.618,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9925290875688916,
+      "grad_norm": 0.3688422739505768,
+      "learning_rate": 3.841064013218093e-06,
+      "loss": 0.6375,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9930189834660135,
+      "grad_norm": 0.3712131977081299,
+      "learning_rate": 3.839963324606989e-06,
+      "loss": 0.637,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9935088793631354,
+      "grad_norm": 0.348637193441391,
+      "learning_rate": 3.838862271437643e-06,
+      "loss": 0.6249,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9939987752602572,
+      "grad_norm": 0.37826526165008545,
+      "learning_rate": 3.837760854009614e-06,
+      "loss": 0.613,
+      "step": 2029
+    },
+    {
+      "epoch": 0.994488671157379,
+      "grad_norm": 0.3614782989025116,
+      "learning_rate": 3.836659072622561e-06,
+      "loss": 0.6404,
+      "step": 2030
+    },
+    {
+      "epoch": 0.9949785670545009,
+      "grad_norm": 0.3549116849899292,
+      "learning_rate": 3.83555692757624e-06,
+      "loss": 0.6055,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9954684629516228,
+      "grad_norm": 0.34219223260879517,
+      "learning_rate": 3.834454419170508e-06,
+      "loss": 0.6064,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9959583588487446,
+      "grad_norm": 0.35754355788230896,
+      "learning_rate": 3.833351547705319e-06,
+      "loss": 0.6182,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9964482547458665,
+      "grad_norm": 0.36640578508377075,
+      "learning_rate": 3.832248313480727e-06,
+      "loss": 0.6177,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9969381506429884,
+      "grad_norm": 0.36644652485847473,
+      "learning_rate": 3.831144716796884e-06,
+      "loss": 0.6331,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9974280465401102,
+      "grad_norm": 0.35853859782218933,
+      "learning_rate": 3.830040757954041e-06,
+      "loss": 0.6277,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9979179424372321,
+      "grad_norm": 0.344949334859848,
+      "learning_rate": 3.828936437252549e-06,
+      "loss": 0.6289,
+      "step": 2037
+    },
+    {
+      "epoch": 0.998407838334354,
+      "grad_norm": 0.35342395305633545,
+      "learning_rate": 3.827831754992854e-06,
+      "loss": 0.6442,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9988977342314758,
+      "grad_norm": 0.36239710450172424,
+      "learning_rate": 3.826726711475502e-06,
+      "loss": 0.633,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9993876301285977,
+      "grad_norm": 0.35739865899086,
+      "learning_rate": 3.82562130700114e-06,
+      "loss": 0.6302,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9998775260257196,
+      "grad_norm": 0.3577716648578644,
+      "learning_rate": 3.8245155418705075e-06,
+      "loss": 0.6137,
+      "step": 2041
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6123,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2041,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.8288154127512895e+19,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}