diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,66533 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9994739610731194,
+  "eval_steps": 500,
+  "global_step": 9500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00010520778537611783,
+      "grad_norm": 1.57374906539917,
+      "learning_rate": 6.993006993006994e-07,
+      "loss": 2.7842,
+      "step": 1
+    },
+    {
+      "epoch": 0.00021041557075223566,
+      "grad_norm": 1.4931204319000244,
+      "learning_rate": 1.3986013986013987e-06,
+      "loss": 2.1824,
+      "step": 2
+    },
+    {
+      "epoch": 0.0003156233561283535,
+      "grad_norm": 1.372463583946228,
+      "learning_rate": 2.0979020979020983e-06,
+      "loss": 2.2936,
+      "step": 3
+    },
+    {
+      "epoch": 0.0004208311415044713,
+      "grad_norm": 1.475909948348999,
+      "learning_rate": 2.7972027972027974e-06,
+      "loss": 2.4173,
+      "step": 4
+    },
+    {
+      "epoch": 0.0005260389268805891,
+      "grad_norm": 2.1777091026306152,
+      "learning_rate": 3.496503496503497e-06,
+      "loss": 2.1757,
+      "step": 5
+    },
+    {
+      "epoch": 0.000631246712256707,
+      "grad_norm": 1.7631592750549316,
+      "learning_rate": 4.195804195804197e-06,
+      "loss": 2.4156,
+      "step": 6
+    },
+    {
+      "epoch": 0.0007364544976328248,
+      "grad_norm": 1.8002638816833496,
+      "learning_rate": 4.895104895104895e-06,
+      "loss": 2.3946,
+      "step": 7
+    },
+    {
+      "epoch": 0.0008416622830089426,
+      "grad_norm": 1.8707481622695923,
+      "learning_rate": 5.594405594405595e-06,
+      "loss": 2.1748,
+      "step": 8
+    },
+    {
+      "epoch": 0.0009468700683850605,
+      "grad_norm": 2.459454298019409,
+      "learning_rate": 6.2937062937062944e-06,
+      "loss": 2.4425,
+      "step": 9
+    },
+    {
+      "epoch": 0.0010520778537611783,
+      "grad_norm": 2.8795626163482666,
+      "learning_rate": 6.993006993006994e-06,
+      "loss": 2.0042,
+      "step": 10
+    },
+    {
+      "epoch": 0.0011572856391372961,
+      "grad_norm": 1.7612422704696655,
+      "learning_rate": 7.692307692307694e-06,
+      "loss": 2.2756,
+      "step": 11
+    },
+    {
+      "epoch": 0.001262493424513414,
+      "grad_norm": 1.7078272104263306,
+      "learning_rate": 8.391608391608393e-06,
+      "loss": 2.327,
+      "step": 12
+    },
+    {
+      "epoch": 0.0013677012098895318,
+      "grad_norm": 1.3450385332107544,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 2.1059,
+      "step": 13
+    },
+    {
+      "epoch": 0.0014729089952656496,
+      "grad_norm": 1.0579757690429688,
+      "learning_rate": 9.79020979020979e-06,
+      "loss": 2.3028,
+      "step": 14
+    },
+    {
+      "epoch": 0.0015781167806417674,
+      "grad_norm": 1.1774177551269531,
+      "learning_rate": 1.048951048951049e-05,
+      "loss": 2.1585,
+      "step": 15
+    },
+    {
+      "epoch": 0.0016833245660178853,
+      "grad_norm": 1.7530457973480225,
+      "learning_rate": 1.118881118881119e-05,
+      "loss": 2.4828,
+      "step": 16
+    },
+    {
+      "epoch": 0.001788532351394003,
+      "grad_norm": 1.65647554397583,
+      "learning_rate": 1.188811188811189e-05,
+      "loss": 1.976,
+      "step": 17
+    },
+    {
+      "epoch": 0.001893740136770121,
+      "grad_norm": 1.5993101596832275,
+      "learning_rate": 1.2587412587412589e-05,
+      "loss": 1.7929,
+      "step": 18
+    },
+    {
+      "epoch": 0.0019989479221462388,
+      "grad_norm": 1.1430275440216064,
+      "learning_rate": 1.3286713286713287e-05,
+      "loss": 2.0432,
+      "step": 19
+    },
+    {
+      "epoch": 0.0021041557075223566,
+      "grad_norm": 1.8768856525421143,
+      "learning_rate": 1.3986013986013988e-05,
+      "loss": 1.8982,
+      "step": 20
+    },
+    {
+      "epoch": 0.0022093634928984744,
+      "grad_norm": 1.7143903970718384,
+      "learning_rate": 1.4685314685314686e-05,
+      "loss": 2.0638,
+      "step": 21
+    },
+    {
+      "epoch": 0.0023145712782745922,
+      "grad_norm": 1.5235050916671753,
+      "learning_rate": 1.5384615384615387e-05,
+      "loss": 2.2032,
+      "step": 22
+    },
+    {
+      "epoch": 0.00241977906365071,
+      "grad_norm": 2.3741185665130615,
+      "learning_rate": 1.6083916083916083e-05,
+      "loss": 2.146,
+      "step": 23
+    },
+    {
+      "epoch": 0.002524986849026828,
+      "grad_norm": 1.3590655326843262,
+      "learning_rate": 1.6783216783216786e-05,
+      "loss": 2.2378,
+      "step": 24
+    },
+    {
+      "epoch": 0.0026301946344029457,
+      "grad_norm": 1.331407904624939,
+      "learning_rate": 1.7482517482517483e-05,
+      "loss": 1.82,
+      "step": 25
+    },
+    {
+      "epoch": 0.0027354024197790636,
+      "grad_norm": 1.4533194303512573,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 1.9261,
+      "step": 26
+    },
+    {
+      "epoch": 0.0028406102051551814,
+      "grad_norm": 1.5130401849746704,
+      "learning_rate": 1.888111888111888e-05,
+      "loss": 2.1417,
+      "step": 27
+    },
+    {
+      "epoch": 0.0029458179905312992,
+      "grad_norm": 1.1877672672271729,
+      "learning_rate": 1.958041958041958e-05,
+      "loss": 2.1062,
+      "step": 28
+    },
+    {
+      "epoch": 0.003051025775907417,
+      "grad_norm": 0.9648356437683105,
+      "learning_rate": 2.027972027972028e-05,
+      "loss": 2.0972,
+      "step": 29
+    },
+    {
+      "epoch": 0.003156233561283535,
+      "grad_norm": 1.1514739990234375,
+      "learning_rate": 2.097902097902098e-05,
+      "loss": 2.4502,
+      "step": 30
+    },
+    {
+      "epoch": 0.0032614413466596527,
+      "grad_norm": 0.9614094495773315,
+      "learning_rate": 2.1678321678321677e-05,
+      "loss": 1.953,
+      "step": 31
+    },
+    {
+      "epoch": 0.0033666491320357705,
+      "grad_norm": 0.970487117767334,
+      "learning_rate": 2.237762237762238e-05,
+      "loss": 1.9179,
+      "step": 32
+    },
+    {
+      "epoch": 0.0034718569174118884,
+      "grad_norm": 1.6860216856002808,
+      "learning_rate": 2.307692307692308e-05,
+      "loss": 2.0499,
+      "step": 33
+    },
+    {
+      "epoch": 0.003577064702788006,
+      "grad_norm": 1.1684486865997314,
+      "learning_rate": 2.377622377622378e-05,
+      "loss": 2.1496,
+      "step": 34
+    },
+    {
+      "epoch": 0.003682272488164124,
+      "grad_norm": 1.1669265031814575,
+      "learning_rate": 2.4475524475524478e-05,
+      "loss": 2.3943,
+      "step": 35
+    },
+    {
+      "epoch": 0.003787480273540242,
+      "grad_norm": 2.067823886871338,
+      "learning_rate": 2.5174825174825178e-05,
+      "loss": 2.2909,
+      "step": 36
+    },
+    {
+      "epoch": 0.0038926880589163597,
+      "grad_norm": 0.8208512663841248,
+      "learning_rate": 2.5874125874125877e-05,
+      "loss": 1.9907,
+      "step": 37
+    },
+    {
+      "epoch": 0.0039978958442924775,
+      "grad_norm": 1.7650972604751587,
+      "learning_rate": 2.6573426573426574e-05,
+      "loss": 2.1261,
+      "step": 38
+    },
+    {
+      "epoch": 0.004103103629668595,
+      "grad_norm": 1.0491235256195068,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 2.1774,
+      "step": 39
+    },
+    {
+      "epoch": 0.004208311415044713,
+      "grad_norm": 1.6899977922439575,
+      "learning_rate": 2.7972027972027976e-05,
+      "loss": 2.5479,
+      "step": 40
+    },
+    {
+      "epoch": 0.004313519200420831,
+      "grad_norm": 0.9535022974014282,
+      "learning_rate": 2.8671328671328672e-05,
+      "loss": 1.9773,
+      "step": 41
+    },
+    {
+      "epoch": 0.004418726985796949,
+      "grad_norm": 1.9684417247772217,
+      "learning_rate": 2.9370629370629372e-05,
+      "loss": 1.7069,
+      "step": 42
+    },
+    {
+      "epoch": 0.004523934771173067,
+      "grad_norm": 1.4697853326797485,
+      "learning_rate": 3.0069930069930068e-05,
+      "loss": 2.225,
+      "step": 43
+    },
+    {
+      "epoch": 0.0046291425565491845,
+      "grad_norm": 1.3747695684432983,
+      "learning_rate": 3.0769230769230774e-05,
+      "loss": 1.8592,
+      "step": 44
+    },
+    {
+      "epoch": 0.004734350341925302,
+      "grad_norm": 1.0535547733306885,
+      "learning_rate": 3.146853146853147e-05,
+      "loss": 1.8008,
+      "step": 45
+    },
+    {
+      "epoch": 0.00483955812730142,
+      "grad_norm": 0.7596752643585205,
+      "learning_rate": 3.216783216783217e-05,
+      "loss": 1.9742,
+      "step": 46
+    },
+    {
+      "epoch": 0.004944765912677538,
+      "grad_norm": 1.3590867519378662,
+      "learning_rate": 3.2867132867132866e-05,
+      "loss": 1.8842,
+      "step": 47
+    },
+    {
+      "epoch": 0.005049973698053656,
+      "grad_norm": 1.1350642442703247,
+      "learning_rate": 3.356643356643357e-05,
+      "loss": 1.6666,
+      "step": 48
+    },
+    {
+      "epoch": 0.005155181483429774,
+      "grad_norm": 1.2734442949295044,
+      "learning_rate": 3.4265734265734265e-05,
+      "loss": 2.0847,
+      "step": 49
+    },
+    {
+      "epoch": 0.0052603892688058915,
+      "grad_norm": 2.199073314666748,
+      "learning_rate": 3.4965034965034965e-05,
+      "loss": 2.0992,
+      "step": 50
+    },
+    {
+      "epoch": 0.005365597054182009,
+      "grad_norm": 1.4339218139648438,
+      "learning_rate": 3.566433566433567e-05,
+      "loss": 2.2127,
+      "step": 51
+    },
+    {
+      "epoch": 0.005470804839558127,
+      "grad_norm": 1.0405915975570679,
+      "learning_rate": 3.6363636363636364e-05,
+      "loss": 2.1656,
+      "step": 52
+    },
+    {
+      "epoch": 0.005576012624934245,
+      "grad_norm": 1.6510354280471802,
+      "learning_rate": 3.7062937062937064e-05,
+      "loss": 1.8267,
+      "step": 53
+    },
+    {
+      "epoch": 0.005681220410310363,
+      "grad_norm": 1.8749672174453735,
+      "learning_rate": 3.776223776223776e-05,
+      "loss": 2.3715,
+      "step": 54
+    },
+    {
+      "epoch": 0.005786428195686481,
+      "grad_norm": 1.5232490301132202,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 1.9894,
+      "step": 55
+    },
+    {
+      "epoch": 0.0058916359810625984,
+      "grad_norm": 1.09939706325531,
+      "learning_rate": 3.916083916083916e-05,
+      "loss": 2.1504,
+      "step": 56
+    },
+    {
+      "epoch": 0.005996843766438716,
+      "grad_norm": 0.9933468699455261,
+      "learning_rate": 3.986013986013986e-05,
+      "loss": 2.1873,
+      "step": 57
+    },
+    {
+      "epoch": 0.006102051551814834,
+      "grad_norm": 1.2622774839401245,
+      "learning_rate": 4.055944055944056e-05,
+      "loss": 1.9191,
+      "step": 58
+    },
+    {
+      "epoch": 0.006207259337190952,
+      "grad_norm": 1.2729672193527222,
+      "learning_rate": 4.125874125874126e-05,
+      "loss": 2.1369,
+      "step": 59
+    },
+    {
+      "epoch": 0.00631246712256707,
+      "grad_norm": 1.32735276222229,
+      "learning_rate": 4.195804195804196e-05,
+      "loss": 1.8028,
+      "step": 60
+    },
+    {
+      "epoch": 0.006417674907943188,
+      "grad_norm": 1.1334835290908813,
+      "learning_rate": 4.265734265734266e-05,
+      "loss": 2.1593,
+      "step": 61
+    },
+    {
+      "epoch": 0.006522882693319305,
+      "grad_norm": 1.2618657350540161,
+      "learning_rate": 4.335664335664335e-05,
+      "loss": 1.9589,
+      "step": 62
+    },
+    {
+      "epoch": 0.006628090478695423,
+      "grad_norm": 1.6038740873336792,
+      "learning_rate": 4.405594405594406e-05,
+      "loss": 2.2707,
+      "step": 63
+    },
+    {
+      "epoch": 0.006733298264071541,
+      "grad_norm": 1.2185710668563843,
+      "learning_rate": 4.475524475524476e-05,
+      "loss": 2.0879,
+      "step": 64
+    },
+    {
+      "epoch": 0.006838506049447659,
+      "grad_norm": 1.9370228052139282,
+      "learning_rate": 4.545454545454546e-05,
+      "loss": 2.0566,
+      "step": 65
+    },
+    {
+      "epoch": 0.006943713834823777,
+      "grad_norm": 1.3582186698913574,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 2.1007,
+      "step": 66
+    },
+    {
+      "epoch": 0.007048921620199895,
+      "grad_norm": 0.7574198246002197,
+      "learning_rate": 4.685314685314686e-05,
+      "loss": 2.0049,
+      "step": 67
+    },
+    {
+      "epoch": 0.007154129405576012,
+      "grad_norm": 1.2563366889953613,
+      "learning_rate": 4.755244755244756e-05,
+      "loss": 2.0965,
+      "step": 68
+    },
+    {
+      "epoch": 0.00725933719095213,
+      "grad_norm": 1.1161051988601685,
+      "learning_rate": 4.825174825174825e-05,
+      "loss": 1.5938,
+      "step": 69
+    },
+    {
+      "epoch": 0.007364544976328248,
+      "grad_norm": 2.441533088684082,
+      "learning_rate": 4.8951048951048956e-05,
+      "loss": 2.4897,
+      "step": 70
+    },
+    {
+      "epoch": 0.007469752761704366,
+      "grad_norm": 1.1693131923675537,
+      "learning_rate": 4.9650349650349656e-05,
+      "loss": 1.9202,
+      "step": 71
+    },
+    {
+      "epoch": 0.007574960547080484,
+      "grad_norm": 1.2114942073822021,
+      "learning_rate": 5.0349650349650356e-05,
+      "loss": 1.996,
+      "step": 72
+    },
+    {
+      "epoch": 0.0076801683324566015,
+      "grad_norm": 0.6417075991630554,
+      "learning_rate": 5.1048951048951055e-05,
+      "loss": 2.1455,
+      "step": 73
+    },
+    {
+      "epoch": 0.007785376117832719,
+      "grad_norm": 1.2000133991241455,
+      "learning_rate": 5.1748251748251755e-05,
+      "loss": 2.2027,
+      "step": 74
+    },
+    {
+      "epoch": 0.007890583903208837,
+      "grad_norm": 1.086300015449524,
+      "learning_rate": 5.244755244755245e-05,
+      "loss": 1.9952,
+      "step": 75
+    },
+    {
+      "epoch": 0.007995791688584955,
+      "grad_norm": 1.217038869857788,
+      "learning_rate": 5.314685314685315e-05,
+      "loss": 1.8331,
+      "step": 76
+    },
+    {
+      "epoch": 0.008100999473961073,
+      "grad_norm": 2.1391165256500244,
+      "learning_rate": 5.384615384615385e-05,
+      "loss": 1.7158,
+      "step": 77
+    },
+    {
+      "epoch": 0.00820620725933719,
+      "grad_norm": 1.2688722610473633,
+      "learning_rate": 5.4545454545454546e-05,
+      "loss": 1.8603,
+      "step": 78
+    },
+    {
+      "epoch": 0.008311415044713309,
+      "grad_norm": 1.1678074598312378,
+      "learning_rate": 5.524475524475524e-05,
+      "loss": 1.8435,
+      "step": 79
+    },
+    {
+      "epoch": 0.008416622830089426,
+      "grad_norm": 0.6559339165687561,
+      "learning_rate": 5.594405594405595e-05,
+      "loss": 2.1147,
+      "step": 80
+    },
+    {
+      "epoch": 0.008521830615465544,
+      "grad_norm": 1.0474909543991089,
+      "learning_rate": 5.664335664335665e-05,
+      "loss": 2.3672,
+      "step": 81
+    },
+    {
+      "epoch": 0.008627038400841662,
+      "grad_norm": 0.8283208012580872,
+      "learning_rate": 5.7342657342657345e-05,
+      "loss": 2.0395,
+      "step": 82
+    },
+    {
+      "epoch": 0.00873224618621778,
+      "grad_norm": 0.7440145015716553,
+      "learning_rate": 5.8041958041958044e-05,
+      "loss": 2.5889,
+      "step": 83
+    },
+    {
+      "epoch": 0.008837453971593898,
+      "grad_norm": 1.0744500160217285,
+      "learning_rate": 5.8741258741258744e-05,
+      "loss": 1.9353,
+      "step": 84
+    },
+    {
+      "epoch": 0.008942661756970016,
+      "grad_norm": 1.0884795188903809,
+      "learning_rate": 5.944055944055944e-05,
+      "loss": 1.8183,
+      "step": 85
+    },
+    {
+      "epoch": 0.009047869542346133,
+      "grad_norm": 1.3748575448989868,
+      "learning_rate": 6.0139860139860136e-05,
+      "loss": 2.0691,
+      "step": 86
+    },
+    {
+      "epoch": 0.009153077327722251,
+      "grad_norm": 0.7459085583686829,
+      "learning_rate": 6.083916083916085e-05,
+      "loss": 2.2508,
+      "step": 87
+    },
+    {
+      "epoch": 0.009258285113098369,
+      "grad_norm": 1.4073225259780884,
+      "learning_rate": 6.153846153846155e-05,
+      "loss": 1.9281,
+      "step": 88
+    },
+    {
+      "epoch": 0.009363492898474487,
+      "grad_norm": 0.988071620464325,
+      "learning_rate": 6.223776223776224e-05,
+      "loss": 1.7869,
+      "step": 89
+    },
+    {
+      "epoch": 0.009468700683850605,
+      "grad_norm": 1.1808884143829346,
+      "learning_rate": 6.293706293706293e-05,
+      "loss": 1.5635,
+      "step": 90
+    },
+    {
+      "epoch": 0.009573908469226722,
+      "grad_norm": 1.0313398838043213,
+      "learning_rate": 6.363636363636364e-05,
+      "loss": 2.0061,
+      "step": 91
+    },
+    {
+      "epoch": 0.00967911625460284,
+      "grad_norm": 0.7924686074256897,
+      "learning_rate": 6.433566433566433e-05,
+      "loss": 2.0563,
+      "step": 92
+    },
+    {
+      "epoch": 0.009784324039978958,
+      "grad_norm": 1.0491949319839478,
+      "learning_rate": 6.503496503496504e-05,
+      "loss": 1.8572,
+      "step": 93
+    },
+    {
+      "epoch": 0.009889531825355076,
+      "grad_norm": 0.8003563284873962,
+      "learning_rate": 6.573426573426573e-05,
+      "loss": 2.1663,
+      "step": 94
+    },
+    {
+      "epoch": 0.009994739610731194,
+      "grad_norm": 0.8011645078659058,
+      "learning_rate": 6.643356643356644e-05,
+      "loss": 2.1442,
+      "step": 95
+    },
+    {
+      "epoch": 0.010099947396107312,
+      "grad_norm": 1.4755507707595825,
+      "learning_rate": 6.713286713286715e-05,
+      "loss": 1.5279,
+      "step": 96
+    },
+    {
+      "epoch": 0.01020515518148343,
+      "grad_norm": 1.2490993738174438,
+      "learning_rate": 6.783216783216784e-05,
+      "loss": 1.7732,
+      "step": 97
+    },
+    {
+      "epoch": 0.010310362966859547,
+      "grad_norm": 0.8913796544075012,
+      "learning_rate": 6.853146853146853e-05,
+      "loss": 1.6675,
+      "step": 98
+    },
+    {
+      "epoch": 0.010415570752235665,
+      "grad_norm": 2.243081569671631,
+      "learning_rate": 6.923076923076924e-05,
+      "loss": 1.9422,
+      "step": 99
+    },
+    {
+      "epoch": 0.010520778537611783,
+      "grad_norm": 1.7832255363464355,
+      "learning_rate": 6.993006993006993e-05,
+      "loss": 1.9094,
+      "step": 100
+    },
+    {
+      "epoch": 0.0106259863229879,
+      "grad_norm": 1.1693270206451416,
+      "learning_rate": 7.062937062937062e-05,
+      "loss": 2.2014,
+      "step": 101
+    },
+    {
+      "epoch": 0.010731194108364019,
+      "grad_norm": 0.9510181546211243,
+      "learning_rate": 7.132867132867134e-05,
+      "loss": 2.0675,
+      "step": 102
+    },
+    {
+      "epoch": 0.010836401893740136,
+      "grad_norm": 1.1362980604171753,
+      "learning_rate": 7.202797202797204e-05,
+      "loss": 1.8922,
+      "step": 103
+    },
+    {
+      "epoch": 0.010941609679116254,
+      "grad_norm": 1.42923903465271,
+      "learning_rate": 7.272727272727273e-05,
+      "loss": 1.5876,
+      "step": 104
+    },
+    {
+      "epoch": 0.011046817464492372,
+      "grad_norm": 0.838302493095398,
+      "learning_rate": 7.342657342657343e-05,
+      "loss": 2.0269,
+      "step": 105
+    },
+    {
+      "epoch": 0.01115202524986849,
+      "grad_norm": 0.8059056401252747,
+      "learning_rate": 7.412587412587413e-05,
+      "loss": 2.2676,
+      "step": 106
+    },
+    {
+      "epoch": 0.011257233035244608,
+      "grad_norm": 1.8598095178604126,
+      "learning_rate": 7.482517482517482e-05,
+      "loss": 2.2518,
+      "step": 107
+    },
+    {
+      "epoch": 0.011362440820620726,
+      "grad_norm": 1.1147247552871704,
+      "learning_rate": 7.552447552447553e-05,
+      "loss": 1.8006,
+      "step": 108
+    },
+    {
+      "epoch": 0.011467648605996843,
+      "grad_norm": 2.8933708667755127,
+      "learning_rate": 7.622377622377622e-05,
+      "loss": 1.4782,
+      "step": 109
+    },
+    {
+      "epoch": 0.011572856391372961,
+      "grad_norm": 1.3423585891723633,
+      "learning_rate": 7.692307692307693e-05,
+      "loss": 1.6204,
+      "step": 110
+    },
+    {
+      "epoch": 0.011678064176749079,
+      "grad_norm": 1.0002747774124146,
+      "learning_rate": 7.762237762237763e-05,
+      "loss": 1.5048,
+      "step": 111
+    },
+    {
+      "epoch": 0.011783271962125197,
+      "grad_norm": 1.511420488357544,
+      "learning_rate": 7.832167832167832e-05,
+      "loss": 2.4107,
+      "step": 112
+    },
+    {
+      "epoch": 0.011888479747501315,
+      "grad_norm": 0.7754486799240112,
+      "learning_rate": 7.902097902097903e-05,
+      "loss": 2.1876,
+      "step": 113
+    },
+    {
+      "epoch": 0.011993687532877433,
+      "grad_norm": 0.6452352404594421,
+      "learning_rate": 7.972027972027972e-05,
+      "loss": 2.3313,
+      "step": 114
+    },
+    {
+      "epoch": 0.01209889531825355,
+      "grad_norm": 0.9890616536140442,
+      "learning_rate": 8.041958041958042e-05,
+      "loss": 2.172,
+      "step": 115
+    },
+    {
+      "epoch": 0.012204103103629668,
+      "grad_norm": 1.8413726091384888,
+      "learning_rate": 8.111888111888112e-05,
+      "loss": 1.969,
+      "step": 116
+    },
+    {
+      "epoch": 0.012309310889005786,
+      "grad_norm": 0.800186812877655,
+      "learning_rate": 8.181818181818183e-05,
+      "loss": 2.268,
+      "step": 117
+    },
+    {
+      "epoch": 0.012414518674381904,
+      "grad_norm": 1.1103674173355103,
+      "learning_rate": 8.251748251748252e-05,
+      "loss": 1.9077,
+      "step": 118
+    },
+    {
+      "epoch": 0.012519726459758022,
+      "grad_norm": 4.220036029815674,
+      "learning_rate": 8.321678321678323e-05,
+      "loss": 2.0661,
+      "step": 119
+    },
+    {
+      "epoch": 0.01262493424513414,
+      "grad_norm": 1.084676742553711,
+      "learning_rate": 8.391608391608392e-05,
+      "loss": 2.1746,
+      "step": 120
+    },
+    {
+      "epoch": 0.012730142030510257,
+      "grad_norm": 1.1023154258728027,
+      "learning_rate": 8.461538461538461e-05,
+      "loss": 1.8782,
+      "step": 121
+    },
+    {
+      "epoch": 0.012835349815886375,
+      "grad_norm": 0.8550626635551453,
+      "learning_rate": 8.531468531468532e-05,
+      "loss": 2.0319,
+      "step": 122
+    },
+    {
+      "epoch": 0.012940557601262493,
+      "grad_norm": 0.5834708213806152,
+      "learning_rate": 8.601398601398601e-05,
+      "loss": 2.1768,
+      "step": 123
+    },
+    {
+      "epoch": 0.01304576538663861,
+      "grad_norm": 0.9890002012252808,
+      "learning_rate": 8.67132867132867e-05,
+      "loss": 2.1604,
+      "step": 124
+    },
+    {
+      "epoch": 0.013150973172014729,
+      "grad_norm": 1.2125439643859863,
+      "learning_rate": 8.741258741258743e-05,
+      "loss": 2.0555,
+      "step": 125
+    },
+    {
+      "epoch": 0.013256180957390847,
+      "grad_norm": 0.8572925329208374,
+      "learning_rate": 8.811188811188812e-05,
+      "loss": 2.1043,
+      "step": 126
+    },
+    {
+      "epoch": 0.013361388742766964,
+      "grad_norm": 0.8613000512123108,
+      "learning_rate": 8.881118881118881e-05,
+      "loss": 2.1363,
+      "step": 127
+    },
+    {
+      "epoch": 0.013466596528143082,
+      "grad_norm": 1.0686365365982056,
+      "learning_rate": 8.951048951048952e-05,
+      "loss": 2.1241,
+      "step": 128
+    },
+    {
+      "epoch": 0.0135718043135192,
+      "grad_norm": 0.6561310291290283,
+      "learning_rate": 9.020979020979021e-05,
+      "loss": 2.037,
+      "step": 129
+    },
+    {
+      "epoch": 0.013677012098895318,
+      "grad_norm": 1.03804612159729,
+      "learning_rate": 9.090909090909092e-05,
+      "loss": 2.0004,
+      "step": 130
+    },
+    {
+      "epoch": 0.013782219884271436,
+      "grad_norm": 0.842249870300293,
+      "learning_rate": 9.160839160839161e-05,
+      "loss": 2.1774,
+      "step": 131
+    },
+    {
+      "epoch": 0.013887427669647553,
+      "grad_norm": 1.4844613075256348,
+      "learning_rate": 9.230769230769232e-05,
+      "loss": 2.0346,
+      "step": 132
+    },
+    {
+      "epoch": 0.013992635455023671,
+      "grad_norm": 0.7268441319465637,
+      "learning_rate": 9.300699300699301e-05,
+      "loss": 2.3364,
+      "step": 133
+    },
+    {
+      "epoch": 0.01409784324039979,
+      "grad_norm": 0.8013659715652466,
+      "learning_rate": 9.370629370629372e-05,
+      "loss": 2.2674,
+      "step": 134
+    },
+    {
+      "epoch": 0.014203051025775907,
+      "grad_norm": 0.6665470004081726,
+      "learning_rate": 9.440559440559441e-05,
+      "loss": 2.0605,
+      "step": 135
+    },
+    {
+      "epoch": 0.014308258811152025,
+      "grad_norm": 1.0592727661132812,
+      "learning_rate": 9.510489510489511e-05,
+      "loss": 1.8648,
+      "step": 136
+    },
+    {
+      "epoch": 0.014413466596528143,
+      "grad_norm": 0.9042508602142334,
+      "learning_rate": 9.580419580419581e-05,
+      "loss": 1.9756,
+      "step": 137
+    },
+    {
+      "epoch": 0.01451867438190426,
+      "grad_norm": 0.7357807755470276,
+      "learning_rate": 9.65034965034965e-05,
+      "loss": 1.9707,
+      "step": 138
+    },
+    {
+      "epoch": 0.014623882167280378,
+      "grad_norm": 0.8247336149215698,
+      "learning_rate": 9.72027972027972e-05,
+      "loss": 2.1052,
+      "step": 139
+    },
+    {
+      "epoch": 0.014729089952656496,
+      "grad_norm": 1.1244010925292969,
+      "learning_rate": 9.790209790209791e-05,
+      "loss": 1.8881,
+      "step": 140
+    },
+    {
+      "epoch": 0.014834297738032614,
+      "grad_norm": 0.7070969343185425,
+      "learning_rate": 9.86013986013986e-05,
+      "loss": 1.9978,
+      "step": 141
+    },
+    {
+      "epoch": 0.014939505523408732,
+      "grad_norm": 1.1827497482299805,
+      "learning_rate": 9.930069930069931e-05,
+      "loss": 1.708,
+      "step": 142
+    },
+    {
+      "epoch": 0.01504471330878485,
+      "grad_norm": 1.9365792274475098,
+      "learning_rate": 0.0001,
+      "loss": 1.8381,
+      "step": 143
+    },
+    {
+      "epoch": 0.015149921094160967,
+      "grad_norm": 0.8472030758857727,
+      "learning_rate": 0.00010069930069930071,
+      "loss": 2.1147,
+      "step": 144
+    },
+    {
+      "epoch": 0.015255128879537085,
+      "grad_norm": 0.7535983920097351,
+      "learning_rate": 0.0001013986013986014,
+      "loss": 2.3344,
+      "step": 145
+    },
+    {
+      "epoch": 0.015360336664913203,
+      "grad_norm": 1.1090881824493408,
+      "learning_rate": 0.00010209790209790211,
+      "loss": 1.7615,
+      "step": 146
+    },
+    {
+      "epoch": 0.015465544450289321,
+      "grad_norm": 0.7076790928840637,
+      "learning_rate": 0.00010279720279720279,
+      "loss": 2.0014,
+      "step": 147
+    },
+    {
+      "epoch": 0.015570752235665439,
+      "grad_norm": 0.7309038043022156,
+      "learning_rate": 0.00010349650349650351,
+      "loss": 2.5436,
+      "step": 148
+    },
+    {
+      "epoch": 0.015675960021041557,
+      "grad_norm": 0.9821389317512512,
+      "learning_rate": 0.00010419580419580419,
+      "loss": 2.0839,
+      "step": 149
+    },
+    {
+      "epoch": 0.015781167806417674,
+      "grad_norm": 0.5570242404937744,
+      "learning_rate": 0.0001048951048951049,
+      "loss": 2.1916,
+      "step": 150
+    },
+    {
+      "epoch": 0.015886375591793792,
+      "grad_norm": 1.2793196439743042,
+      "learning_rate": 0.00010559440559440561,
+      "loss": 2.1197,
+      "step": 151
+    },
+    {
+      "epoch": 0.01599158337716991,
+      "grad_norm": 1.9450187683105469,
+      "learning_rate": 0.0001062937062937063,
+      "loss": 2.5529,
+      "step": 152
+    },
+    {
+      "epoch": 0.016096791162546028,
+      "grad_norm": 2.063152313232422,
+      "learning_rate": 0.000106993006993007,
+      "loss": 1.9606,
+      "step": 153
+    },
+    {
+      "epoch": 0.016201998947922146,
+      "grad_norm": 1.4163166284561157,
+      "learning_rate": 0.0001076923076923077,
+      "loss": 2.002,
+      "step": 154
+    },
+    {
+      "epoch": 0.016307206733298264,
+      "grad_norm": 0.6336003541946411,
+      "learning_rate": 0.0001083916083916084,
+      "loss": 2.3557,
+      "step": 155
+    },
+    {
+      "epoch": 0.01641241451867438,
+      "grad_norm": 1.3749457597732544,
+      "learning_rate": 0.00010909090909090909,
+      "loss": 2.3967,
+      "step": 156
+    },
+    {
+      "epoch": 0.0165176223040505,
+      "grad_norm": 0.8000378608703613,
+      "learning_rate": 0.0001097902097902098,
+      "loss": 2.0812,
+      "step": 157
+    },
+    {
+      "epoch": 0.016622830089426617,
+      "grad_norm": 0.8916708827018738,
+      "learning_rate": 0.00011048951048951048,
+      "loss": 1.8927,
+      "step": 158
+    },
+    {
+      "epoch": 0.016728037874802735,
+      "grad_norm": 0.8637097477912903,
+      "learning_rate": 0.0001111888111888112,
+      "loss": 1.8131,
+      "step": 159
+    },
+    {
+      "epoch": 0.016833245660178853,
+      "grad_norm": 0.6609115600585938,
+      "learning_rate": 0.0001118881118881119,
+      "loss": 1.9201,
+      "step": 160
+    },
+    {
+      "epoch": 0.01693845344555497,
+      "grad_norm": 1.2824020385742188,
+      "learning_rate": 0.00011258741258741258,
+      "loss": 2.3487,
+      "step": 161
+    },
+    {
+      "epoch": 0.01704366123093109,
+      "grad_norm": 1.6525263786315918,
+      "learning_rate": 0.0001132867132867133,
+      "loss": 1.9996,
+      "step": 162
+    },
+    {
+      "epoch": 0.017148869016307206,
+      "grad_norm": 1.3814963102340698,
+      "learning_rate": 0.00011398601398601398,
+      "loss": 1.7609,
+      "step": 163
+    },
+    {
+      "epoch": 0.017254076801683324,
+      "grad_norm": 0.6455362439155579,
+      "learning_rate": 0.00011468531468531469,
+      "loss": 2.1223,
+      "step": 164
+    },
+    {
+      "epoch": 0.017359284587059442,
+      "grad_norm": 0.7798193097114563,
+      "learning_rate": 0.00011538461538461538,
+      "loss": 2.0314,
+      "step": 165
+    },
+    {
+      "epoch": 0.01746449237243556,
+      "grad_norm": 0.7223178744316101,
+      "learning_rate": 0.00011608391608391609,
+      "loss": 2.0679,
+      "step": 166
+    },
+    {
+      "epoch": 0.017569700157811678,
+      "grad_norm": 0.8387543559074402,
+      "learning_rate": 0.0001167832167832168,
+      "loss": 2.2392,
+      "step": 167
+    },
+    {
+      "epoch": 0.017674907943187795,
+      "grad_norm": 1.1753877401351929,
+      "learning_rate": 0.00011748251748251749,
+      "loss": 2.2198,
+      "step": 168
+    },
+    {
+      "epoch": 0.017780115728563913,
+      "grad_norm": 0.6246500015258789,
+      "learning_rate": 0.0001181818181818182,
+      "loss": 1.9576,
+      "step": 169
+    },
+    {
+      "epoch": 0.01788532351394003,
+      "grad_norm": 0.7104965448379517,
+      "learning_rate": 0.00011888111888111889,
+      "loss": 1.8494,
+      "step": 170
+    },
+    {
+      "epoch": 0.01799053129931615,
+      "grad_norm": 0.6617504954338074,
+      "learning_rate": 0.00011958041958041959,
+      "loss": 2.1685,
+      "step": 171
+    },
+    {
+      "epoch": 0.018095739084692267,
+      "grad_norm": 0.5817732810974121,
+      "learning_rate": 0.00012027972027972027,
+      "loss": 2.4582,
+      "step": 172
+    },
+    {
+      "epoch": 0.018200946870068384,
+      "grad_norm": 1.5956623554229736,
+      "learning_rate": 0.00012097902097902098,
+      "loss": 1.9319,
+      "step": 173
+    },
+    {
+      "epoch": 0.018306154655444502,
+      "grad_norm": 0.6000881195068359,
+      "learning_rate": 0.0001216783216783217,
+      "loss": 2.2725,
+      "step": 174
+    },
+    {
+      "epoch": 0.01841136244082062,
+      "grad_norm": 1.0373468399047852,
+      "learning_rate": 0.00012237762237762238,
+      "loss": 2.3838,
+      "step": 175
+    },
+    {
+      "epoch": 0.018516570226196738,
+      "grad_norm": 0.6030766367912292,
+      "learning_rate": 0.0001230769230769231,
+      "loss": 2.0372,
+      "step": 176
+    },
+    {
+      "epoch": 0.018621778011572856,
+      "grad_norm": 0.7469233274459839,
+      "learning_rate": 0.00012377622377622376,
+      "loss": 2.1564,
+      "step": 177
+    },
+    {
+      "epoch": 0.018726985796948974,
+      "grad_norm": 0.9262884259223938,
+      "learning_rate": 0.00012447552447552448,
+      "loss": 1.4632,
+      "step": 178
+    },
+    {
+      "epoch": 0.01883219358232509,
+      "grad_norm": 0.8348685503005981,
+      "learning_rate": 0.00012517482517482518,
+      "loss": 2.193,
+      "step": 179
+    },
+    {
+      "epoch": 0.01893740136770121,
+      "grad_norm": 0.6114562153816223,
+      "learning_rate": 0.00012587412587412587,
+      "loss": 2.0964,
+      "step": 180
+    },
+    {
+      "epoch": 0.019042609153077327,
+      "grad_norm": 1.039941668510437,
+      "learning_rate": 0.0001265734265734266,
+      "loss": 1.9306,
+      "step": 181
+    },
+    {
+      "epoch": 0.019147816938453445,
+      "grad_norm": 0.730789303779602,
+      "learning_rate": 0.00012727272727272728,
+      "loss": 1.7754,
+      "step": 182
+    },
+    {
+      "epoch": 0.019253024723829563,
+      "grad_norm": 0.76559978723526,
+      "learning_rate": 0.00012797202797202797,
+      "loss": 2.1377,
+      "step": 183
+    },
+    {
+      "epoch": 0.01935823250920568,
+      "grad_norm": 0.7957161068916321,
+      "learning_rate": 0.00012867132867132867,
+      "loss": 2.2704,
+      "step": 184
+    },
+    {
+      "epoch": 0.0194634402945818,
+      "grad_norm": 0.9248109459877014,
+      "learning_rate": 0.0001293706293706294,
+      "loss": 2.0038,
+      "step": 185
+    },
+    {
+      "epoch": 0.019568648079957916,
+      "grad_norm": 0.7987109422683716,
+      "learning_rate": 0.00013006993006993008,
+      "loss": 2.2301,
+      "step": 186
+    },
+    {
+      "epoch": 0.019673855865334034,
+      "grad_norm": 0.5288488864898682,
+      "learning_rate": 0.00013076923076923077,
+      "loss": 1.9312,
+      "step": 187
+    },
+    {
+      "epoch": 0.019779063650710152,
+      "grad_norm": 0.6533221006393433,
+      "learning_rate": 0.00013146853146853147,
+      "loss": 2.0679,
+      "step": 188
+    },
+    {
+      "epoch": 0.01988427143608627,
+      "grad_norm": 1.0460737943649292,
+      "learning_rate": 0.00013216783216783219,
+      "loss": 1.9375,
+      "step": 189
+    },
+    {
+      "epoch": 0.019989479221462388,
+      "grad_norm": 1.1394834518432617,
+      "learning_rate": 0.00013286713286713288,
+      "loss": 2.248,
+      "step": 190
+    },
+    {
+      "epoch": 0.020094687006838505,
+      "grad_norm": 0.6399918794631958,
+      "learning_rate": 0.00013356643356643357,
+      "loss": 2.1755,
+      "step": 191
+    },
+    {
+      "epoch": 0.020199894792214623,
+      "grad_norm": 1.1859235763549805,
+      "learning_rate": 0.0001342657342657343,
+      "loss": 1.8647,
+      "step": 192
+    },
+    {
+      "epoch": 0.02030510257759074,
+      "grad_norm": 0.7891494631767273,
+      "learning_rate": 0.00013496503496503496,
+      "loss": 2.1801,
+      "step": 193
+    },
+    {
+      "epoch": 0.02041031036296686,
+      "grad_norm": 0.8865561485290527,
+      "learning_rate": 0.00013566433566433568,
+      "loss": 1.9941,
+      "step": 194
+    },
+    {
+      "epoch": 0.020515518148342977,
+      "grad_norm": 0.6699333786964417,
+      "learning_rate": 0.00013636363636363637,
+      "loss": 1.9786,
+      "step": 195
+    },
+    {
+      "epoch": 0.020620725933719095,
+      "grad_norm": 0.6600997447967529,
+      "learning_rate": 0.00013706293706293706,
+      "loss": 2.2511,
+      "step": 196
+    },
+    {
+      "epoch": 0.020725933719095212,
+      "grad_norm": 0.6689944863319397,
+      "learning_rate": 0.00013776223776223778,
+      "loss": 2.0213,
+      "step": 197
+    },
+    {
+      "epoch": 0.02083114150447133,
+      "grad_norm": 0.8018683791160583,
+      "learning_rate": 0.00013846153846153847,
+      "loss": 2.0495,
+      "step": 198
+    },
+    {
+      "epoch": 0.020936349289847448,
+      "grad_norm": 1.0803660154342651,
+      "learning_rate": 0.00013916083916083917,
+      "loss": 1.1935,
+      "step": 199
+    },
+    {
+      "epoch": 0.021041557075223566,
+      "grad_norm": 2.967259168624878,
+      "learning_rate": 0.00013986013986013986,
+      "loss": 1.5776,
+      "step": 200
+    },
+    {
+      "epoch": 0.021146764860599684,
+      "grad_norm": 0.9621496200561523,
+      "learning_rate": 0.00014055944055944058,
+      "loss": 1.9666,
+      "step": 201
+    },
+    {
+      "epoch": 0.0212519726459758,
+      "grad_norm": 0.8522931933403015,
+      "learning_rate": 0.00014125874125874125,
+      "loss": 1.8651,
+      "step": 202
+    },
+    {
+      "epoch": 0.02135718043135192,
+      "grad_norm": 0.942253589630127,
+      "learning_rate": 0.00014195804195804197,
+      "loss": 2.1782,
+      "step": 203
+    },
+    {
+      "epoch": 0.021462388216728037,
+      "grad_norm": 0.6087969541549683,
+      "learning_rate": 0.00014265734265734269,
+      "loss": 2.165,
+      "step": 204
+    },
+    {
+      "epoch": 0.021567596002104155,
+      "grad_norm": 0.6847811937332153,
+      "learning_rate": 0.00014335664335664335,
+      "loss": 2.1116,
+      "step": 205
+    },
+    {
+      "epoch": 0.021672803787480273,
+      "grad_norm": 0.9084582328796387,
+      "learning_rate": 0.00014405594405594407,
+      "loss": 1.9343,
+      "step": 206
+    },
+    {
+      "epoch": 0.02177801157285639,
+      "grad_norm": 1.0526587963104248,
+      "learning_rate": 0.00014475524475524476,
+      "loss": 2.1309,
+      "step": 207
+    },
+    {
+      "epoch": 0.02188321935823251,
+      "grad_norm": 0.8651286363601685,
+      "learning_rate": 0.00014545454545454546,
+      "loss": 2.1103,
+      "step": 208
+    },
+    {
+      "epoch": 0.021988427143608626,
+      "grad_norm": 0.6131342053413391,
+      "learning_rate": 0.00014615384615384615,
+      "loss": 1.9468,
+      "step": 209
+    },
+    {
+      "epoch": 0.022093634928984744,
+      "grad_norm": 0.784864068031311,
+      "learning_rate": 0.00014685314685314687,
+      "loss": 2.1167,
+      "step": 210
+    },
+    {
+      "epoch": 0.022198842714360862,
+      "grad_norm": 1.1468371152877808,
+      "learning_rate": 0.00014755244755244756,
+      "loss": 1.8572,
+      "step": 211
+    },
+    {
+      "epoch": 0.02230405049973698,
+      "grad_norm": 0.9270665049552917,
+      "learning_rate": 0.00014825174825174825,
+      "loss": 1.9398,
+      "step": 212
+    },
+    {
+      "epoch": 0.022409258285113098,
+      "grad_norm": 0.7162615060806274,
+      "learning_rate": 0.00014895104895104897,
+      "loss": 2.0499,
+      "step": 213
+    },
+    {
+      "epoch": 0.022514466070489215,
+      "grad_norm": 0.6274422407150269,
+      "learning_rate": 0.00014965034965034964,
+      "loss": 1.8765,
+      "step": 214
+    },
+    {
+      "epoch": 0.022619673855865333,
+      "grad_norm": 1.4804967641830444,
+      "learning_rate": 0.00015034965034965036,
+      "loss": 1.7015,
+      "step": 215
+    },
+    {
+      "epoch": 0.02272488164124145,
+      "grad_norm": 0.5013079047203064,
+      "learning_rate": 0.00015104895104895105,
+      "loss": 2.0067,
+      "step": 216
+    },
+    {
+      "epoch": 0.02283008942661757,
+      "grad_norm": 0.7269556522369385,
+      "learning_rate": 0.00015174825174825175,
+      "loss": 1.934,
+      "step": 217
+    },
+    {
+      "epoch": 0.022935297211993687,
+      "grad_norm": 0.9243491291999817,
+      "learning_rate": 0.00015244755244755244,
+      "loss": 2.2308,
+      "step": 218
+    },
+    {
+      "epoch": 0.023040504997369805,
+      "grad_norm": 0.7595291137695312,
+      "learning_rate": 0.00015314685314685316,
+      "loss": 2.0564,
+      "step": 219
+    },
+    {
+      "epoch": 0.023145712782745922,
+      "grad_norm": 0.8647655248641968,
+      "learning_rate": 0.00015384615384615385,
+      "loss": 2.0716,
+      "step": 220
+    },
+    {
+      "epoch": 0.02325092056812204,
+      "grad_norm": 1.0030921697616577,
+      "learning_rate": 0.00015454545454545454,
+      "loss": 1.9771,
+      "step": 221
+    },
+    {
+      "epoch": 0.023356128353498158,
+      "grad_norm": 0.588725745677948,
+      "learning_rate": 0.00015524475524475526,
+      "loss": 2.2316,
+      "step": 222
+    },
+    {
+      "epoch": 0.023461336138874276,
+      "grad_norm": 1.029255747795105,
+      "learning_rate": 0.00015594405594405596,
+      "loss": 2.2892,
+      "step": 223
+    },
+    {
+      "epoch": 0.023566543924250394,
+      "grad_norm": 0.9255016446113586,
+      "learning_rate": 0.00015664335664335665,
+      "loss": 2.2594,
+      "step": 224
+    },
+    {
+      "epoch": 0.02367175170962651,
+      "grad_norm": 0.84700608253479,
+      "learning_rate": 0.00015734265734265734,
+      "loss": 1.8739,
+      "step": 225
+    },
+    {
+      "epoch": 0.02377695949500263,
+      "grad_norm": 0.8125012516975403,
+      "learning_rate": 0.00015804195804195806,
+      "loss": 2.0795,
+      "step": 226
+    },
+    {
+      "epoch": 0.023882167280378747,
+      "grad_norm": 0.8614683151245117,
+      "learning_rate": 0.00015874125874125876,
+      "loss": 2.4142,
+      "step": 227
+    },
+    {
+      "epoch": 0.023987375065754865,
+      "grad_norm": 1.1380670070648193,
+      "learning_rate": 0.00015944055944055945,
+      "loss": 2.1002,
+      "step": 228
+    },
+    {
+      "epoch": 0.024092582851130983,
+      "grad_norm": 0.4754476547241211,
+      "learning_rate": 0.00016013986013986014,
+      "loss": 2.0722,
+      "step": 229
+    },
+    {
+      "epoch": 0.0241977906365071,
+      "grad_norm": 0.861049473285675,
+      "learning_rate": 0.00016083916083916083,
+      "loss": 1.9022,
+      "step": 230
+    },
+    {
+      "epoch": 0.02430299842188322,
+      "grad_norm": 1.181152582168579,
+      "learning_rate": 0.00016153846153846155,
+      "loss": 2.0356,
+      "step": 231
+    },
+    {
+      "epoch": 0.024408206207259336,
+      "grad_norm": 0.8931273818016052,
+      "learning_rate": 0.00016223776223776225,
+      "loss": 1.7346,
+      "step": 232
+    },
+    {
+      "epoch": 0.024513413992635454,
+      "grad_norm": 0.7377164959907532,
+      "learning_rate": 0.00016293706293706294,
+      "loss": 2.0764,
+      "step": 233
+    },
+    {
+      "epoch": 0.024618621778011572,
+      "grad_norm": 1.085389494895935,
+      "learning_rate": 0.00016363636363636366,
+      "loss": 1.9141,
+      "step": 234
+    },
+    {
+      "epoch": 0.02472382956338769,
+      "grad_norm": 1.0622587203979492,
+      "learning_rate": 0.00016433566433566435,
+      "loss": 1.8455,
+      "step": 235
+    },
+    {
+      "epoch": 0.024829037348763808,
+      "grad_norm": 1.0064948797225952,
+      "learning_rate": 0.00016503496503496504,
+      "loss": 1.9076,
+      "step": 236
+    },
+    {
+      "epoch": 0.024934245134139926,
+      "grad_norm": 0.5665073394775391,
+      "learning_rate": 0.00016573426573426574,
+      "loss": 2.0581,
+      "step": 237
+    },
+    {
+      "epoch": 0.025039452919516043,
+      "grad_norm": 1.042952299118042,
+      "learning_rate": 0.00016643356643356646,
+      "loss": 1.9164,
+      "step": 238
+    },
+    {
+      "epoch": 0.02514466070489216,
+      "grad_norm": 1.4019465446472168,
+      "learning_rate": 0.00016713286713286712,
+      "loss": 1.8818,
+      "step": 239
+    },
+    {
+      "epoch": 0.02524986849026828,
+      "grad_norm": 1.0074125528335571,
+      "learning_rate": 0.00016783216783216784,
+      "loss": 1.971,
+      "step": 240
+    },
+    {
+      "epoch": 0.025355076275644397,
+      "grad_norm": 0.9089799523353577,
+      "learning_rate": 0.00016853146853146856,
+      "loss": 2.087,
+      "step": 241
+    },
+    {
+      "epoch": 0.025460284061020515,
+      "grad_norm": 0.7242212891578674,
+      "learning_rate": 0.00016923076923076923,
+      "loss": 2.1149,
+      "step": 242
+    },
+    {
+      "epoch": 0.025565491846396633,
+      "grad_norm": 1.1103755235671997,
+      "learning_rate": 0.00016993006993006995,
+      "loss": 2.1568,
+      "step": 243
+    },
+    {
+      "epoch": 0.02567069963177275,
+      "grad_norm": 1.03432297706604,
+      "learning_rate": 0.00017062937062937064,
+      "loss": 2.1965,
+      "step": 244
+    },
+    {
+      "epoch": 0.025775907417148868,
+      "grad_norm": 0.8843188881874084,
+      "learning_rate": 0.00017132867132867133,
+      "loss": 1.7656,
+      "step": 245
+    },
+    {
+      "epoch": 0.025881115202524986,
+      "grad_norm": 1.1456265449523926,
+      "learning_rate": 0.00017202797202797203,
+      "loss": 2.2859,
+      "step": 246
+    },
+    {
+      "epoch": 0.025986322987901104,
+      "grad_norm": 1.0545116662979126,
+      "learning_rate": 0.00017272727272727275,
+      "loss": 2.0337,
+      "step": 247
+    },
+    {
+      "epoch": 0.02609153077327722,
+      "grad_norm": 1.3290053606033325,
+      "learning_rate": 0.0001734265734265734,
+      "loss": 2.1952,
+      "step": 248
+    },
+    {
+      "epoch": 0.02619673855865334,
+      "grad_norm": 1.2462905645370483,
+      "learning_rate": 0.00017412587412587413,
+      "loss": 2.1667,
+      "step": 249
+    },
+    {
+      "epoch": 0.026301946344029457,
+      "grad_norm": 1.0230878591537476,
+      "learning_rate": 0.00017482517482517485,
+      "loss": 2.0687,
+      "step": 250
+    },
+    {
+      "epoch": 0.026407154129405575,
+      "grad_norm": 0.6971817016601562,
+      "learning_rate": 0.00017552447552447552,
+      "loss": 1.947,
+      "step": 251
+    },
+    {
+      "epoch": 0.026512361914781693,
+      "grad_norm": 1.1189419031143188,
+      "learning_rate": 0.00017622377622377624,
+      "loss": 1.9158,
+      "step": 252
+    },
+    {
+      "epoch": 0.02661756970015781,
+      "grad_norm": 0.9112401604652405,
+      "learning_rate": 0.00017692307692307693,
+      "loss": 2.2175,
+      "step": 253
+    },
+    {
+      "epoch": 0.02672277748553393,
+      "grad_norm": 0.6773673892021179,
+      "learning_rate": 0.00017762237762237762,
+      "loss": 2.0269,
+      "step": 254
+    },
+    {
+      "epoch": 0.026827985270910047,
+      "grad_norm": 1.021903157234192,
+      "learning_rate": 0.00017832167832167832,
+      "loss": 1.4599,
+      "step": 255
+    },
+    {
+      "epoch": 0.026933193056286164,
+      "grad_norm": 1.0119444131851196,
+      "learning_rate": 0.00017902097902097904,
+      "loss": 2.5356,
+      "step": 256
+    },
+    {
+      "epoch": 0.027038400841662282,
+      "grad_norm": 0.8079578280448914,
+      "learning_rate": 0.00017972027972027973,
+      "loss": 1.7094,
+      "step": 257
+    },
+    {
+      "epoch": 0.0271436086270384,
+      "grad_norm": 0.8587137460708618,
+      "learning_rate": 0.00018041958041958042,
+      "loss": 1.9044,
+      "step": 258
+    },
+    {
+      "epoch": 0.027248816412414518,
+      "grad_norm": 0.4971306324005127,
+      "learning_rate": 0.00018111888111888114,
+      "loss": 2.7055,
+      "step": 259
+    },
+    {
+      "epoch": 0.027354024197790636,
+      "grad_norm": 0.6644048094749451,
+      "learning_rate": 0.00018181818181818183,
+      "loss": 2.22,
+      "step": 260
+    },
+    {
+      "epoch": 0.027459231983166753,
+      "grad_norm": 0.8848311901092529,
+      "learning_rate": 0.00018251748251748253,
+      "loss": 2.1772,
+      "step": 261
+    },
+    {
+      "epoch": 0.02756443976854287,
+      "grad_norm": 0.8665701746940613,
+      "learning_rate": 0.00018321678321678322,
+      "loss": 2.0003,
+      "step": 262
+    },
+    {
+      "epoch": 0.02766964755391899,
+      "grad_norm": 0.9104865193367004,
+      "learning_rate": 0.0001839160839160839,
+      "loss": 2.2117,
+      "step": 263
+    },
+    {
+      "epoch": 0.027774855339295107,
+      "grad_norm": 0.7475347518920898,
+      "learning_rate": 0.00018461538461538463,
+      "loss": 2.0686,
+      "step": 264
+    },
+    {
+      "epoch": 0.027880063124671225,
+      "grad_norm": 0.9693028926849365,
+      "learning_rate": 0.00018531468531468533,
+      "loss": 2.1383,
+      "step": 265
+    },
+    {
+      "epoch": 0.027985270910047343,
+      "grad_norm": 0.6588386297225952,
+      "learning_rate": 0.00018601398601398602,
+      "loss": 1.8683,
+      "step": 266
+    },
+    {
+      "epoch": 0.02809047869542346,
+      "grad_norm": 1.076177716255188,
+      "learning_rate": 0.0001867132867132867,
+      "loss": 1.5809,
+      "step": 267
+    },
+    {
+      "epoch": 0.02819568648079958,
+      "grad_norm": 0.6180933713912964,
+      "learning_rate": 0.00018741258741258743,
+      "loss": 1.9758,
+      "step": 268
+    },
+    {
+      "epoch": 0.028300894266175696,
+      "grad_norm": 1.6872791051864624,
+      "learning_rate": 0.00018811188811188812,
+      "loss": 2.3591,
+      "step": 269
+    },
+    {
+      "epoch": 0.028406102051551814,
+      "grad_norm": 1.2484400272369385,
+      "learning_rate": 0.00018881118881118882,
+      "loss": 1.7551,
+      "step": 270
+    },
+    {
+      "epoch": 0.028511309836927932,
+      "grad_norm": 0.5355095863342285,
+      "learning_rate": 0.00018951048951048954,
+      "loss": 2.1445,
+      "step": 271
+    },
+    {
+      "epoch": 0.02861651762230405,
+      "grad_norm": 0.8856534361839294,
+      "learning_rate": 0.00019020979020979023,
+      "loss": 2.0174,
+      "step": 272
+    },
+    {
+      "epoch": 0.028721725407680167,
+      "grad_norm": 1.0029089450836182,
+      "learning_rate": 0.00019090909090909092,
+      "loss": 2.0354,
+      "step": 273
+    },
+    {
+      "epoch": 0.028826933193056285,
+      "grad_norm": 0.859724760055542,
+      "learning_rate": 0.00019160839160839161,
+      "loss": 2.0125,
+      "step": 274
+    },
+    {
+      "epoch": 0.028932140978432403,
+      "grad_norm": 1.281600832939148,
+      "learning_rate": 0.00019230769230769233,
+      "loss": 1.8134,
+      "step": 275
+    },
+    {
+      "epoch": 0.02903734876380852,
+      "grad_norm": 1.7904019355773926,
+      "learning_rate": 0.000193006993006993,
+      "loss": 1.7277,
+      "step": 276
+    },
+    {
+      "epoch": 0.02914255654918464,
+      "grad_norm": 0.7003609538078308,
+      "learning_rate": 0.00019370629370629372,
+      "loss": 2.0308,
+      "step": 277
+    },
+    {
+      "epoch": 0.029247764334560757,
+      "grad_norm": 0.8506051898002625,
+      "learning_rate": 0.0001944055944055944,
+      "loss": 1.8183,
+      "step": 278
+    },
+    {
+      "epoch": 0.029352972119936874,
+      "grad_norm": 0.7373519539833069,
+      "learning_rate": 0.0001951048951048951,
+      "loss": 2.0156,
+      "step": 279
+    },
+    {
+      "epoch": 0.029458179905312992,
+      "grad_norm": 0.8916333913803101,
+      "learning_rate": 0.00019580419580419583,
+      "loss": 1.9928,
+      "step": 280
+    },
+    {
+      "epoch": 0.02956338769068911,
+      "grad_norm": 0.7887770533561707,
+      "learning_rate": 0.00019650349650349652,
+      "loss": 2.0476,
+      "step": 281
+    },
+    {
+      "epoch": 0.029668595476065228,
+      "grad_norm": 1.0839449167251587,
+      "learning_rate": 0.0001972027972027972,
+      "loss": 2.457,
+      "step": 282
+    },
+    {
+      "epoch": 0.029773803261441346,
+      "grad_norm": 0.7722465395927429,
+      "learning_rate": 0.0001979020979020979,
+      "loss": 2.1979,
+      "step": 283
+    },
+    {
+      "epoch": 0.029879011046817464,
+      "grad_norm": 0.6390840411186218,
+      "learning_rate": 0.00019860139860139862,
+      "loss": 1.9645,
+      "step": 284
+    },
+    {
+      "epoch": 0.02998421883219358,
+      "grad_norm": 0.6858240365982056,
+      "learning_rate": 0.0001993006993006993,
+      "loss": 2.0411,
+      "step": 285
+    },
+    {
+      "epoch": 0.0300894266175697,
+      "grad_norm": 0.831282377243042,
+      "learning_rate": 0.0002,
+      "loss": 2.4271,
+      "step": 286
+    },
+    {
+      "epoch": 0.030194634402945817,
+      "grad_norm": 1.3080503940582275,
+      "learning_rate": 0.00019999999419366464,
+      "loss": 2.4988,
+      "step": 287
+    },
+    {
+      "epoch": 0.030299842188321935,
+      "grad_norm": 0.6604061126708984,
+      "learning_rate": 0.00019999997677465928,
+      "loss": 2.3058,
+      "step": 288
+    },
+    {
+      "epoch": 0.030405049973698053,
+      "grad_norm": 0.682809591293335,
+      "learning_rate": 0.00019999994774298586,
+      "loss": 2.3447,
+      "step": 289
+    },
+    {
+      "epoch": 0.03051025775907417,
+      "grad_norm": 0.7935196161270142,
+      "learning_rate": 0.00019999990709864784,
+      "loss": 1.9286,
+      "step": 290
+    },
+    {
+      "epoch": 0.03061546554445029,
+      "grad_norm": 0.7283986806869507,
+      "learning_rate": 0.00019999985484164988,
+      "loss": 2.1939,
+      "step": 291
+    },
+    {
+      "epoch": 0.030720673329826406,
+      "grad_norm": 0.6452785134315491,
+      "learning_rate": 0.00019999979097199807,
+      "loss": 2.186,
+      "step": 292
+    },
+    {
+      "epoch": 0.030825881115202524,
+      "grad_norm": 0.7383095622062683,
+      "learning_rate": 0.00019999971548969982,
+      "loss": 2.4815,
+      "step": 293
+    },
+    {
+      "epoch": 0.030931088900578642,
+      "grad_norm": 0.8694736361503601,
+      "learning_rate": 0.00019999962839476393,
+      "loss": 1.8029,
+      "step": 294
+    },
+    {
+      "epoch": 0.03103629668595476,
+      "grad_norm": 1.1409188508987427,
+      "learning_rate": 0.00019999952968720045,
+      "loss": 1.8402,
+      "step": 295
+    },
+    {
+      "epoch": 0.031141504471330878,
+      "grad_norm": 1.1897854804992676,
+      "learning_rate": 0.0001999994193670209,
+      "loss": 2.2032,
+      "step": 296
+    },
+    {
+      "epoch": 0.031246712256706995,
+      "grad_norm": 1.0531314611434937,
+      "learning_rate": 0.00019999929743423804,
+      "loss": 2.0809,
+      "step": 297
+    },
+    {
+      "epoch": 0.03135192004208311,
+      "grad_norm": 0.9149481654167175,
+      "learning_rate": 0.0001999991638888661,
+      "loss": 1.9994,
+      "step": 298
+    },
+    {
+      "epoch": 0.031457127827459234,
+      "grad_norm": 0.9323375821113586,
+      "learning_rate": 0.00019999901873092054,
+      "loss": 2.1517,
+      "step": 299
+    },
+    {
+      "epoch": 0.03156233561283535,
+      "grad_norm": 2.169660806655884,
+      "learning_rate": 0.0001999988619604182,
+      "loss": 1.935,
+      "step": 300
+    },
+    {
+      "epoch": 0.03166754339821147,
+      "grad_norm": 1.145294189453125,
+      "learning_rate": 0.0001999986935773773,
+      "loss": 2.2346,
+      "step": 301
+    },
+    {
+      "epoch": 0.031772751183587584,
+      "grad_norm": 0.9163552522659302,
+      "learning_rate": 0.00019999851358181746,
+      "loss": 1.757,
+      "step": 302
+    },
+    {
+      "epoch": 0.031877958968963706,
+      "grad_norm": 0.9327316284179688,
+      "learning_rate": 0.00019999832197375948,
+      "loss": 2.0945,
+      "step": 303
+    },
+    {
+      "epoch": 0.03198316675433982,
+      "grad_norm": 1.3414212465286255,
+      "learning_rate": 0.00019999811875322566,
+      "loss": 1.8407,
+      "step": 304
+    },
+    {
+      "epoch": 0.03208837453971594,
+      "grad_norm": 0.7236807346343994,
+      "learning_rate": 0.0001999979039202396,
+      "loss": 2.1481,
+      "step": 305
+    },
+    {
+      "epoch": 0.032193582325092056,
+      "grad_norm": 0.7519909143447876,
+      "learning_rate": 0.00019999767747482623,
+      "loss": 2.156,
+      "step": 306
+    },
+    {
+      "epoch": 0.03229879011046818,
+      "grad_norm": 1.3652244806289673,
+      "learning_rate": 0.00019999743941701188,
+      "loss": 1.8686,
+      "step": 307
+    },
+    {
+      "epoch": 0.03240399789584429,
+      "grad_norm": 0.5761735439300537,
+      "learning_rate": 0.00019999718974682417,
+      "loss": 2.0619,
+      "step": 308
+    },
+    {
+      "epoch": 0.03250920568122041,
+      "grad_norm": 1.1688545942306519,
+      "learning_rate": 0.0001999969284642921,
+      "loss": 2.0442,
+      "step": 309
+    },
+    {
+      "epoch": 0.03261441346659653,
+      "grad_norm": 0.6734681129455566,
+      "learning_rate": 0.000199996655569446,
+      "loss": 2.2581,
+      "step": 310
+    },
+    {
+      "epoch": 0.03271962125197265,
+      "grad_norm": 0.9942630529403687,
+      "learning_rate": 0.00019999637106231756,
+      "loss": 2.0594,
+      "step": 311
+    },
+    {
+      "epoch": 0.03282482903734876,
+      "grad_norm": 0.9288811087608337,
+      "learning_rate": 0.00019999607494293985,
+      "loss": 1.8957,
+      "step": 312
+    },
+    {
+      "epoch": 0.032930036822724884,
+      "grad_norm": 2.1428439617156982,
+      "learning_rate": 0.00019999576721134723,
+      "loss": 2.2481,
+      "step": 313
+    },
+    {
+      "epoch": 0.033035244608101,
+      "grad_norm": 1.1253790855407715,
+      "learning_rate": 0.00019999544786757545,
+      "loss": 2.0877,
+      "step": 314
+    },
+    {
+      "epoch": 0.03314045239347712,
+      "grad_norm": 0.9545801281929016,
+      "learning_rate": 0.00019999511691166157,
+      "loss": 2.2495,
+      "step": 315
+    },
+    {
+      "epoch": 0.033245660178853234,
+      "grad_norm": 0.9224940538406372,
+      "learning_rate": 0.00019999477434364405,
+      "loss": 2.0281,
+      "step": 316
+    },
+    {
+      "epoch": 0.033350867964229355,
+      "grad_norm": 0.9393844604492188,
+      "learning_rate": 0.00019999442016356266,
+      "loss": 1.7388,
+      "step": 317
+    },
+    {
+      "epoch": 0.03345607574960547,
+      "grad_norm": 0.8138694763183594,
+      "learning_rate": 0.00019999405437145856,
+      "loss": 1.8485,
+      "step": 318
+    },
+    {
+      "epoch": 0.03356128353498159,
+      "grad_norm": 0.9134123921394348,
+      "learning_rate": 0.00019999367696737415,
+      "loss": 1.9404,
+      "step": 319
+    },
+    {
+      "epoch": 0.033666491320357705,
+      "grad_norm": 0.7344478368759155,
+      "learning_rate": 0.0001999932879513533,
+      "loss": 2.4859,
+      "step": 320
+    },
+    {
+      "epoch": 0.03377169910573383,
+      "grad_norm": 1.5100687742233276,
+      "learning_rate": 0.00019999288732344122,
+      "loss": 1.9299,
+      "step": 321
+    },
+    {
+      "epoch": 0.03387690689110994,
+      "grad_norm": 0.7897876501083374,
+      "learning_rate": 0.0001999924750836844,
+      "loss": 2.0341,
+      "step": 322
+    },
+    {
+      "epoch": 0.03398211467648606,
+      "grad_norm": 0.8573527932167053,
+      "learning_rate": 0.00019999205123213073,
+      "loss": 1.8272,
+      "step": 323
+    },
+    {
+      "epoch": 0.03408732246186218,
+      "grad_norm": 0.6045930981636047,
+      "learning_rate": 0.00019999161576882938,
+      "loss": 2.3906,
+      "step": 324
+    },
+    {
+      "epoch": 0.0341925302472383,
+      "grad_norm": 1.1515878438949585,
+      "learning_rate": 0.000199991168693831,
+      "loss": 1.8901,
+      "step": 325
+    },
+    {
+      "epoch": 0.03429773803261441,
+      "grad_norm": 0.7033329606056213,
+      "learning_rate": 0.00019999071000718742,
+      "loss": 2.0389,
+      "step": 326
+    },
+    {
+      "epoch": 0.034402945817990534,
+      "grad_norm": 0.7773725986480713,
+      "learning_rate": 0.00019999023970895198,
+      "loss": 1.7427,
+      "step": 327
+    },
+    {
+      "epoch": 0.03450815360336665,
+      "grad_norm": 0.6639922261238098,
+      "learning_rate": 0.0001999897577991792,
+      "loss": 2.4179,
+      "step": 328
+    },
+    {
+      "epoch": 0.03461336138874277,
+      "grad_norm": 0.8306198716163635,
+      "learning_rate": 0.00019998926427792517,
+      "loss": 1.8829,
+      "step": 329
+    },
+    {
+      "epoch": 0.034718569174118884,
+      "grad_norm": 0.9085239768028259,
+      "learning_rate": 0.00019998875914524714,
+      "loss": 2.3608,
+      "step": 330
+    },
+    {
+      "epoch": 0.034823776959495005,
+      "grad_norm": 1.2059913873672485,
+      "learning_rate": 0.00019998824240120372,
+      "loss": 2.1859,
+      "step": 331
+    },
+    {
+      "epoch": 0.03492898474487112,
+      "grad_norm": 0.8465059399604797,
+      "learning_rate": 0.000199987714045855,
+      "loss": 1.9598,
+      "step": 332
+    },
+    {
+      "epoch": 0.03503419253024724,
+      "grad_norm": 1.1591929197311401,
+      "learning_rate": 0.00019998717407926228,
+      "loss": 1.7713,
+      "step": 333
+    },
+    {
+      "epoch": 0.035139400315623355,
+      "grad_norm": 1.2473613023757935,
+      "learning_rate": 0.0001999866225014883,
+      "loss": 2.0015,
+      "step": 334
+    },
+    {
+      "epoch": 0.035244608100999476,
+      "grad_norm": 1.0094128847122192,
+      "learning_rate": 0.0001999860593125971,
+      "loss": 2.3931,
+      "step": 335
+    },
+    {
+      "epoch": 0.03534981588637559,
+      "grad_norm": 0.587139368057251,
+      "learning_rate": 0.00019998548451265405,
+      "loss": 2.1993,
+      "step": 336
+    },
+    {
+      "epoch": 0.03545502367175171,
+      "grad_norm": 1.4602278470993042,
+      "learning_rate": 0.00019998489810172596,
+      "loss": 2.1095,
+      "step": 337
+    },
+    {
+      "epoch": 0.035560231457127826,
+      "grad_norm": 0.9542207717895508,
+      "learning_rate": 0.00019998430007988087,
+      "loss": 2.0584,
+      "step": 338
+    },
+    {
+      "epoch": 0.03566543924250395,
+      "grad_norm": 0.9044630527496338,
+      "learning_rate": 0.00019998369044718826,
+      "loss": 2.2235,
+      "step": 339
+    },
+    {
+      "epoch": 0.03577064702788006,
+      "grad_norm": 0.8812581300735474,
+      "learning_rate": 0.0001999830692037189,
+      "loss": 2.0134,
+      "step": 340
+    },
+    {
+      "epoch": 0.03587585481325618,
+      "grad_norm": 0.8419604301452637,
+      "learning_rate": 0.000199982436349545,
+      "loss": 2.0388,
+      "step": 341
+    },
+    {
+      "epoch": 0.0359810625986323,
+      "grad_norm": 0.7698401808738708,
+      "learning_rate": 0.00019998179188473997,
+      "loss": 1.9938,
+      "step": 342
+    },
+    {
+      "epoch": 0.03608627038400842,
+      "grad_norm": 0.808614194393158,
+      "learning_rate": 0.0001999811358093787,
+      "loss": 2.1887,
+      "step": 343
+    },
+    {
+      "epoch": 0.03619147816938453,
+      "grad_norm": 0.756734311580658,
+      "learning_rate": 0.00019998046812353732,
+      "loss": 2.1403,
+      "step": 344
+    },
+    {
+      "epoch": 0.036296685954760655,
+      "grad_norm": 0.8613669276237488,
+      "learning_rate": 0.00019997978882729345,
+      "loss": 1.4033,
+      "step": 345
+    },
+    {
+      "epoch": 0.03640189374013677,
+      "grad_norm": 0.8686724305152893,
+      "learning_rate": 0.0001999790979207259,
+      "loss": 2.2557,
+      "step": 346
+    },
+    {
+      "epoch": 0.03650710152551289,
+      "grad_norm": 1.0263909101486206,
+      "learning_rate": 0.00019997839540391495,
+      "loss": 2.1586,
+      "step": 347
+    },
+    {
+      "epoch": 0.036612309310889005,
+      "grad_norm": 1.0269930362701416,
+      "learning_rate": 0.00019997768127694214,
+      "loss": 2.3377,
+      "step": 348
+    },
+    {
+      "epoch": 0.036717517096265126,
+      "grad_norm": 1.0349308252334595,
+      "learning_rate": 0.00019997695553989042,
+      "loss": 1.8215,
+      "step": 349
+    },
+    {
+      "epoch": 0.03682272488164124,
+      "grad_norm": 0.8196139931678772,
+      "learning_rate": 0.0001999762181928441,
+      "loss": 2.0538,
+      "step": 350
+    },
+    {
+      "epoch": 0.03692793266701736,
+      "grad_norm": 0.9920759797096252,
+      "learning_rate": 0.00019997546923588875,
+      "loss": 2.2415,
+      "step": 351
+    },
+    {
+      "epoch": 0.037033140452393476,
+      "grad_norm": 1.101965069770813,
+      "learning_rate": 0.00019997470866911136,
+      "loss": 1.9107,
+      "step": 352
+    },
+    {
+      "epoch": 0.0371383482377696,
+      "grad_norm": 1.4456695318222046,
+      "learning_rate": 0.00019997393649260028,
+      "loss": 2.3002,
+      "step": 353
+    },
+    {
+      "epoch": 0.03724355602314571,
+      "grad_norm": 1.1259775161743164,
+      "learning_rate": 0.00019997315270644514,
+      "loss": 1.8325,
+      "step": 354
+    },
+    {
+      "epoch": 0.03734876380852183,
+      "grad_norm": 0.7673949003219604,
+      "learning_rate": 0.000199972357310737,
+      "loss": 2.3046,
+      "step": 355
+    },
+    {
+      "epoch": 0.03745397159389795,
+      "grad_norm": 0.8659745454788208,
+      "learning_rate": 0.00019997155030556822,
+      "loss": 2.1083,
+      "step": 356
+    },
+    {
+      "epoch": 0.03755917937927407,
+      "grad_norm": 0.6444668769836426,
+      "learning_rate": 0.0001999707316910325,
+      "loss": 2.1361,
+      "step": 357
+    },
+    {
+      "epoch": 0.03766438716465018,
+      "grad_norm": 0.978226900100708,
+      "learning_rate": 0.0001999699014672249,
+      "loss": 2.1801,
+      "step": 358
+    },
+    {
+      "epoch": 0.037769594950026304,
+      "grad_norm": 1.1167306900024414,
+      "learning_rate": 0.0001999690596342418,
+      "loss": 2.3531,
+      "step": 359
+    },
+    {
+      "epoch": 0.03787480273540242,
+      "grad_norm": 0.9561347365379333,
+      "learning_rate": 0.00019996820619218105,
+      "loss": 2.2767,
+      "step": 360
+    },
+    {
+      "epoch": 0.03798001052077854,
+      "grad_norm": 0.6484118700027466,
+      "learning_rate": 0.00019996734114114165,
+      "loss": 2.2618,
+      "step": 361
+    },
+    {
+      "epoch": 0.038085218306154654,
+      "grad_norm": 1.1894108057022095,
+      "learning_rate": 0.00019996646448122414,
+      "loss": 2.169,
+      "step": 362
+    },
+    {
+      "epoch": 0.038190426091530776,
+      "grad_norm": 0.7844385504722595,
+      "learning_rate": 0.00019996557621253027,
+      "loss": 1.8752,
+      "step": 363
+    },
+    {
+      "epoch": 0.03829563387690689,
+      "grad_norm": 0.7246622443199158,
+      "learning_rate": 0.00019996467633516326,
+      "loss": 2.1476,
+      "step": 364
+    },
+    {
+      "epoch": 0.03840084166228301,
+      "grad_norm": 0.7702452540397644,
+      "learning_rate": 0.0001999637648492275,
+      "loss": 2.038,
+      "step": 365
+    },
+    {
+      "epoch": 0.038506049447659126,
+      "grad_norm": 0.7791807055473328,
+      "learning_rate": 0.00019996284175482893,
+      "loss": 1.9363,
+      "step": 366
+    },
+    {
+      "epoch": 0.03861125723303525,
+      "grad_norm": 0.650177538394928,
+      "learning_rate": 0.00019996190705207475,
+      "loss": 2.0905,
+      "step": 367
+    },
+    {
+      "epoch": 0.03871646501841136,
+      "grad_norm": 1.108355164527893,
+      "learning_rate": 0.00019996096074107342,
+      "loss": 2.0743,
+      "step": 368
+    },
+    {
+      "epoch": 0.03882167280378748,
+      "grad_norm": 1.025335431098938,
+      "learning_rate": 0.0001999600028219349,
+      "loss": 1.895,
+      "step": 369
+    },
+    {
+      "epoch": 0.0389268805891636,
+      "grad_norm": 1.1372300386428833,
+      "learning_rate": 0.0001999590332947704,
+      "loss": 2.1421,
+      "step": 370
+    },
+    {
+      "epoch": 0.03903208837453972,
+      "grad_norm": 1.4822285175323486,
+      "learning_rate": 0.00019995805215969258,
+      "loss": 2.3544,
+      "step": 371
+    },
+    {
+      "epoch": 0.03913729615991583,
+      "grad_norm": 1.2210115194320679,
+      "learning_rate": 0.00019995705941681523,
+      "loss": 2.0829,
+      "step": 372
+    },
+    {
+      "epoch": 0.039242503945291954,
+      "grad_norm": 0.9157518148422241,
+      "learning_rate": 0.00019995605506625377,
+      "loss": 1.9328,
+      "step": 373
+    },
+    {
+      "epoch": 0.03934771173066807,
+      "grad_norm": 0.8865572214126587,
+      "learning_rate": 0.00019995503910812478,
+      "loss": 2.202,
+      "step": 374
+    },
+    {
+      "epoch": 0.03945291951604419,
+      "grad_norm": 1.0392677783966064,
+      "learning_rate": 0.00019995401154254626,
+      "loss": 2.1567,
+      "step": 375
+    },
+    {
+      "epoch": 0.039558127301420304,
+      "grad_norm": 1.0013548135757446,
+      "learning_rate": 0.00019995297236963749,
+      "loss": 1.9707,
+      "step": 376
+    },
+    {
+      "epoch": 0.039663335086796425,
+      "grad_norm": 0.6896701455116272,
+      "learning_rate": 0.00019995192158951919,
+      "loss": 1.8789,
+      "step": 377
+    },
+    {
+      "epoch": 0.03976854287217254,
+      "grad_norm": 0.6109434366226196,
+      "learning_rate": 0.00019995085920231336,
+      "loss": 2.2985,
+      "step": 378
+    },
+    {
+      "epoch": 0.03987375065754866,
+      "grad_norm": 0.7272301912307739,
+      "learning_rate": 0.00019994978520814337,
+      "loss": 1.8993,
+      "step": 379
+    },
+    {
+      "epoch": 0.039978958442924775,
+      "grad_norm": 1.2493723630905151,
+      "learning_rate": 0.00019994869960713397,
+      "loss": 2.2194,
+      "step": 380
+    },
+    {
+      "epoch": 0.040084166228300896,
+      "grad_norm": 0.7804626822471619,
+      "learning_rate": 0.0001999476023994112,
+      "loss": 2.0154,
+      "step": 381
+    },
+    {
+      "epoch": 0.04018937401367701,
+      "grad_norm": 0.863120973110199,
+      "learning_rate": 0.00019994649358510248,
+      "loss": 1.7918,
+      "step": 382
+    },
+    {
+      "epoch": 0.04029458179905313,
+      "grad_norm": 1.1610263586044312,
+      "learning_rate": 0.0001999453731643366,
+      "loss": 2.3425,
+      "step": 383
+    },
+    {
+      "epoch": 0.040399789584429247,
+      "grad_norm": 0.9565117359161377,
+      "learning_rate": 0.00019994424113724363,
+      "loss": 1.8774,
+      "step": 384
+    },
+    {
+      "epoch": 0.04050499736980537,
+      "grad_norm": 0.676775336265564,
+      "learning_rate": 0.00019994309750395506,
+      "loss": 2.066,
+      "step": 385
+    },
+    {
+      "epoch": 0.04061020515518148,
+      "grad_norm": 0.6846379637718201,
+      "learning_rate": 0.00019994194226460367,
+      "loss": 1.9599,
+      "step": 386
+    },
+    {
+      "epoch": 0.0407154129405576,
+      "grad_norm": 0.5663594603538513,
+      "learning_rate": 0.0001999407754193236,
+      "loss": 1.9791,
+      "step": 387
+    },
+    {
+      "epoch": 0.04082062072593372,
+      "grad_norm": 0.967096745967865,
+      "learning_rate": 0.0001999395969682504,
+      "loss": 2.1146,
+      "step": 388
+    },
+    {
+      "epoch": 0.04092582851130984,
+      "grad_norm": 0.7784907817840576,
+      "learning_rate": 0.00019993840691152093,
+      "loss": 2.0466,
+      "step": 389
+    },
+    {
+      "epoch": 0.041031036296685953,
+      "grad_norm": 0.7650101780891418,
+      "learning_rate": 0.0001999372052492733,
+      "loss": 2.2071,
+      "step": 390
+    },
+    {
+      "epoch": 0.041136244082062075,
+      "grad_norm": 0.6302483677864075,
+      "learning_rate": 0.00019993599198164715,
+      "loss": 2.0403,
+      "step": 391
+    },
+    {
+      "epoch": 0.04124145186743819,
+      "grad_norm": 0.6898879408836365,
+      "learning_rate": 0.00019993476710878332,
+      "loss": 1.7578,
+      "step": 392
+    },
+    {
+      "epoch": 0.04134665965281431,
+      "grad_norm": 1.1919569969177246,
+      "learning_rate": 0.00019993353063082404,
+      "loss": 1.9792,
+      "step": 393
+    },
+    {
+      "epoch": 0.041451867438190425,
+      "grad_norm": 0.6219474077224731,
+      "learning_rate": 0.00019993228254791293,
+      "loss": 2.4463,
+      "step": 394
+    },
+    {
+      "epoch": 0.041557075223566546,
+      "grad_norm": 0.9380045533180237,
+      "learning_rate": 0.00019993102286019495,
+      "loss": 1.8507,
+      "step": 395
+    },
+    {
+      "epoch": 0.04166228300894266,
+      "grad_norm": 1.5710420608520508,
+      "learning_rate": 0.0001999297515678163,
+      "loss": 2.1511,
+      "step": 396
+    },
+    {
+      "epoch": 0.04176749079431878,
+      "grad_norm": 0.6227583289146423,
+      "learning_rate": 0.00019992846867092473,
+      "loss": 2.0332,
+      "step": 397
+    },
+    {
+      "epoch": 0.041872698579694896,
+      "grad_norm": 0.7754364609718323,
+      "learning_rate": 0.0001999271741696691,
+      "loss": 1.6828,
+      "step": 398
+    },
+    {
+      "epoch": 0.04197790636507102,
+      "grad_norm": 1.0617812871932983,
+      "learning_rate": 0.0001999258680641998,
+      "loss": 2.0997,
+      "step": 399
+    },
+    {
+      "epoch": 0.04208311415044713,
+      "grad_norm": 0.8257030844688416,
+      "learning_rate": 0.00019992455035466847,
+      "loss": 1.948,
+      "step": 400
+    },
+    {
+      "epoch": 0.04218832193582325,
+      "grad_norm": 0.9446254968643188,
+      "learning_rate": 0.00019992322104122817,
+      "loss": 1.7266,
+      "step": 401
+    },
+    {
+      "epoch": 0.04229352972119937,
+      "grad_norm": 0.52854323387146,
+      "learning_rate": 0.00019992188012403324,
+      "loss": 2.2328,
+      "step": 402
+    },
+    {
+      "epoch": 0.04239873750657549,
+      "grad_norm": 0.9252466559410095,
+      "learning_rate": 0.00019992052760323941,
+      "loss": 2.1785,
+      "step": 403
+    },
+    {
+      "epoch": 0.0425039452919516,
+      "grad_norm": 0.7319321632385254,
+      "learning_rate": 0.00019991916347900378,
+      "loss": 2.2174,
+      "step": 404
+    },
+    {
+      "epoch": 0.042609153077327724,
+      "grad_norm": 0.5902657508850098,
+      "learning_rate": 0.00019991778775148465,
+      "loss": 2.1312,
+      "step": 405
+    },
+    {
+      "epoch": 0.04271436086270384,
+      "grad_norm": 0.8190617561340332,
+      "learning_rate": 0.0001999164004208419,
+      "loss": 1.658,
+      "step": 406
+    },
+    {
+      "epoch": 0.04281956864807996,
+      "grad_norm": 0.9919306635856628,
+      "learning_rate": 0.00019991500148723658,
+      "loss": 2.158,
+      "step": 407
+    },
+    {
+      "epoch": 0.042924776433456074,
+      "grad_norm": 1.2936168909072876,
+      "learning_rate": 0.00019991359095083112,
+      "loss": 1.8505,
+      "step": 408
+    },
+    {
+      "epoch": 0.043029984218832196,
+      "grad_norm": 0.947219729423523,
+      "learning_rate": 0.00019991216881178937,
+      "loss": 1.8942,
+      "step": 409
+    },
+    {
+      "epoch": 0.04313519200420831,
+      "grad_norm": 1.0280535221099854,
+      "learning_rate": 0.00019991073507027646,
+      "loss": 1.7948,
+      "step": 410
+    },
+    {
+      "epoch": 0.04324039978958443,
+      "grad_norm": 0.795024573802948,
+      "learning_rate": 0.00019990928972645887,
+      "loss": 1.9645,
+      "step": 411
+    },
+    {
+      "epoch": 0.043345607574960546,
+      "grad_norm": 0.9444951415061951,
+      "learning_rate": 0.00019990783278050448,
+      "loss": 2.0397,
+      "step": 412
+    },
+    {
+      "epoch": 0.04345081536033667,
+      "grad_norm": 0.7790425419807434,
+      "learning_rate": 0.00019990636423258246,
+      "loss": 2.0172,
+      "step": 413
+    },
+    {
+      "epoch": 0.04355602314571278,
+      "grad_norm": 0.8171620965003967,
+      "learning_rate": 0.00019990488408286333,
+      "loss": 1.9963,
+      "step": 414
+    },
+    {
+      "epoch": 0.0436612309310889,
+      "grad_norm": 1.3680206537246704,
+      "learning_rate": 0.000199903392331519,
+      "loss": 1.5671,
+      "step": 415
+    },
+    {
+      "epoch": 0.04376643871646502,
+      "grad_norm": 1.41587495803833,
+      "learning_rate": 0.00019990188897872266,
+      "loss": 2.2752,
+      "step": 416
+    },
+    {
+      "epoch": 0.04387164650184114,
+      "grad_norm": 0.888954222202301,
+      "learning_rate": 0.00019990037402464896,
+      "loss": 2.1677,
+      "step": 417
+    },
+    {
+      "epoch": 0.04397685428721725,
+      "grad_norm": 0.8410660028457642,
+      "learning_rate": 0.00019989884746947378,
+      "loss": 2.1372,
+      "step": 418
+    },
+    {
+      "epoch": 0.044082062072593374,
+      "grad_norm": 0.8259899616241455,
+      "learning_rate": 0.0001998973093133744,
+      "loss": 1.7659,
+      "step": 419
+    },
+    {
+      "epoch": 0.04418726985796949,
+      "grad_norm": 1.2934560775756836,
+      "learning_rate": 0.00019989575955652944,
+      "loss": 1.8465,
+      "step": 420
+    },
+    {
+      "epoch": 0.04429247764334561,
+      "grad_norm": 0.7665582299232483,
+      "learning_rate": 0.00019989419819911887,
+      "loss": 2.2067,
+      "step": 421
+    },
+    {
+      "epoch": 0.044397685428721724,
+      "grad_norm": 0.676923394203186,
+      "learning_rate": 0.000199892625241324,
+      "loss": 2.3273,
+      "step": 422
+    },
+    {
+      "epoch": 0.044502893214097845,
+      "grad_norm": 1.0820274353027344,
+      "learning_rate": 0.00019989104068332756,
+      "loss": 1.8123,
+      "step": 423
+    },
+    {
+      "epoch": 0.04460810099947396,
+      "grad_norm": 1.1282938718795776,
+      "learning_rate": 0.00019988944452531345,
+      "loss": 1.9473,
+      "step": 424
+    },
+    {
+      "epoch": 0.04471330878485008,
+      "grad_norm": 1.5150783061981201,
+      "learning_rate": 0.00019988783676746708,
+      "loss": 1.6027,
+      "step": 425
+    },
+    {
+      "epoch": 0.044818516570226195,
+      "grad_norm": 0.991240918636322,
+      "learning_rate": 0.00019988621740997512,
+      "loss": 1.9637,
+      "step": 426
+    },
+    {
+      "epoch": 0.04492372435560232,
+      "grad_norm": 0.8624051809310913,
+      "learning_rate": 0.00019988458645302568,
+      "loss": 1.9826,
+      "step": 427
+    },
+    {
+      "epoch": 0.04502893214097843,
+      "grad_norm": 0.9856376647949219,
+      "learning_rate": 0.00019988294389680812,
+      "loss": 2.3543,
+      "step": 428
+    },
+    {
+      "epoch": 0.04513413992635455,
+      "grad_norm": 0.5573300719261169,
+      "learning_rate": 0.0001998812897415132,
+      "loss": 1.955,
+      "step": 429
+    },
+    {
+      "epoch": 0.04523934771173067,
+      "grad_norm": 0.8506868481636047,
+      "learning_rate": 0.000199879623987333,
+      "loss": 2.0627,
+      "step": 430
+    },
+    {
+      "epoch": 0.04534455549710679,
+      "grad_norm": 1.4136837720870972,
+      "learning_rate": 0.00019987794663446095,
+      "loss": 2.0503,
+      "step": 431
+    },
+    {
+      "epoch": 0.0454497632824829,
+      "grad_norm": 0.9526495337486267,
+      "learning_rate": 0.0001998762576830919,
+      "loss": 1.7729,
+      "step": 432
+    },
+    {
+      "epoch": 0.045554971067859024,
+      "grad_norm": 0.8205630779266357,
+      "learning_rate": 0.00019987455713342187,
+      "loss": 2.3268,
+      "step": 433
+    },
+    {
+      "epoch": 0.04566017885323514,
+      "grad_norm": 0.9681246876716614,
+      "learning_rate": 0.0001998728449856484,
+      "loss": 1.862,
+      "step": 434
+    },
+    {
+      "epoch": 0.04576538663861126,
+      "grad_norm": 0.7441048622131348,
+      "learning_rate": 0.00019987112123997033,
+      "loss": 2.1119,
+      "step": 435
+    },
+    {
+      "epoch": 0.045870594423987374,
+      "grad_norm": 1.2735470533370972,
+      "learning_rate": 0.00019986938589658783,
+      "loss": 1.7813,
+      "step": 436
+    },
+    {
+      "epoch": 0.045975802209363495,
+      "grad_norm": 1.4802747964859009,
+      "learning_rate": 0.00019986763895570242,
+      "loss": 2.1473,
+      "step": 437
+    },
+    {
+      "epoch": 0.04608100999473961,
+      "grad_norm": 1.388631820678711,
+      "learning_rate": 0.0001998658804175169,
+      "loss": 1.8662,
+      "step": 438
+    },
+    {
+      "epoch": 0.04618621778011573,
+      "grad_norm": 0.5619102716445923,
+      "learning_rate": 0.00019986411028223558,
+      "loss": 2.0551,
+      "step": 439
+    },
+    {
+      "epoch": 0.046291425565491845,
+      "grad_norm": 0.7170664668083191,
+      "learning_rate": 0.000199862328550064,
+      "loss": 2.0751,
+      "step": 440
+    },
+    {
+      "epoch": 0.046396633350867966,
+      "grad_norm": 0.7204136252403259,
+      "learning_rate": 0.000199860535221209,
+      "loss": 2.0988,
+      "step": 441
+    },
+    {
+      "epoch": 0.04650184113624408,
+      "grad_norm": 1.1917775869369507,
+      "learning_rate": 0.0001998587302958789,
+      "loss": 1.8188,
+      "step": 442
+    },
+    {
+      "epoch": 0.0466070489216202,
+      "grad_norm": 0.861599862575531,
+      "learning_rate": 0.00019985691377428326,
+      "loss": 2.3283,
+      "step": 443
+    },
+    {
+      "epoch": 0.046712256706996316,
+      "grad_norm": 0.7343453168869019,
+      "learning_rate": 0.00019985508565663305,
+      "loss": 2.1423,
+      "step": 444
+    },
+    {
+      "epoch": 0.04681746449237244,
+      "grad_norm": 1.0733033418655396,
+      "learning_rate": 0.00019985324594314055,
+      "loss": 1.814,
+      "step": 445
+    },
+    {
+      "epoch": 0.04692267227774855,
+      "grad_norm": 0.7048352956771851,
+      "learning_rate": 0.00019985139463401944,
+      "loss": 2.2652,
+      "step": 446
+    },
+    {
+      "epoch": 0.04702788006312467,
+      "grad_norm": 0.9343500733375549,
+      "learning_rate": 0.00019984953172948465,
+      "loss": 2.1747,
+      "step": 447
+    },
+    {
+      "epoch": 0.04713308784850079,
+      "grad_norm": 0.9725675582885742,
+      "learning_rate": 0.00019984765722975254,
+      "loss": 2.0506,
+      "step": 448
+    },
+    {
+      "epoch": 0.04723829563387691,
+      "grad_norm": 1.3480443954467773,
+      "learning_rate": 0.00019984577113504076,
+      "loss": 2.4285,
+      "step": 449
+    },
+    {
+      "epoch": 0.04734350341925302,
+      "grad_norm": 1.5416226387023926,
+      "learning_rate": 0.0001998438734455684,
+      "loss": 1.7112,
+      "step": 450
+    },
+    {
+      "epoch": 0.047448711204629145,
+      "grad_norm": 0.7121959328651428,
+      "learning_rate": 0.0001998419641615558,
+      "loss": 1.9964,
+      "step": 451
+    },
+    {
+      "epoch": 0.04755391899000526,
+      "grad_norm": 0.905596911907196,
+      "learning_rate": 0.00019984004328322464,
+      "loss": 2.351,
+      "step": 452
+    },
+    {
+      "epoch": 0.04765912677538138,
+      "grad_norm": 0.9009513258934021,
+      "learning_rate": 0.00019983811081079807,
+      "loss": 2.0917,
+      "step": 453
+    },
+    {
+      "epoch": 0.047764334560757495,
+      "grad_norm": 1.4585151672363281,
+      "learning_rate": 0.0001998361667445004,
+      "loss": 1.8901,
+      "step": 454
+    },
+    {
+      "epoch": 0.047869542346133616,
+      "grad_norm": 1.1657588481903076,
+      "learning_rate": 0.00019983421108455746,
+      "loss": 1.8957,
+      "step": 455
+    },
+    {
+      "epoch": 0.04797475013150973,
+      "grad_norm": 0.7370551824569702,
+      "learning_rate": 0.00019983224383119633,
+      "loss": 1.9769,
+      "step": 456
+    },
+    {
+      "epoch": 0.04807995791688585,
+      "grad_norm": 0.9683654308319092,
+      "learning_rate": 0.00019983026498464546,
+      "loss": 1.6403,
+      "step": 457
+    },
+    {
+      "epoch": 0.048185165702261966,
+      "grad_norm": 0.7799041867256165,
+      "learning_rate": 0.00019982827454513466,
+      "loss": 1.9753,
+      "step": 458
+    },
+    {
+      "epoch": 0.04829037348763809,
+      "grad_norm": 1.0562982559204102,
+      "learning_rate": 0.00019982627251289504,
+      "loss": 2.2229,
+      "step": 459
+    },
+    {
+      "epoch": 0.0483955812730142,
+      "grad_norm": 0.7870259881019592,
+      "learning_rate": 0.00019982425888815915,
+      "loss": 2.025,
+      "step": 460
+    },
+    {
+      "epoch": 0.04850078905839032,
+      "grad_norm": 1.191759467124939,
+      "learning_rate": 0.00019982223367116076,
+      "loss": 2.3058,
+      "step": 461
+    },
+    {
+      "epoch": 0.04860599684376644,
+      "grad_norm": 1.1136091947555542,
+      "learning_rate": 0.0001998201968621351,
+      "loss": 1.824,
+      "step": 462
+    },
+    {
+      "epoch": 0.04871120462914256,
+      "grad_norm": 0.8942115306854248,
+      "learning_rate": 0.00019981814846131867,
+      "loss": 2.1555,
+      "step": 463
+    },
+    {
+      "epoch": 0.04881641241451867,
+      "grad_norm": 1.0691640377044678,
+      "learning_rate": 0.00019981608846894933,
+      "loss": 1.9872,
+      "step": 464
+    },
+    {
+      "epoch": 0.048921620199894794,
+      "grad_norm": 1.0228484869003296,
+      "learning_rate": 0.00019981401688526636,
+      "loss": 2.1705,
+      "step": 465
+    },
+    {
+      "epoch": 0.04902682798527091,
+      "grad_norm": 0.9551164507865906,
+      "learning_rate": 0.00019981193371051026,
+      "loss": 1.9308,
+      "step": 466
+    },
+    {
+      "epoch": 0.04913203577064703,
+      "grad_norm": 1.2936720848083496,
+      "learning_rate": 0.000199809838944923,
+      "loss": 2.2471,
+      "step": 467
+    },
+    {
+      "epoch": 0.049237243556023144,
+      "grad_norm": 0.9659333229064941,
+      "learning_rate": 0.00019980773258874778,
+      "loss": 1.9027,
+      "step": 468
+    },
+    {
+      "epoch": 0.049342451341399265,
+      "grad_norm": 0.814227819442749,
+      "learning_rate": 0.00019980561464222926,
+      "loss": 1.8804,
+      "step": 469
+    },
+    {
+      "epoch": 0.04944765912677538,
+      "grad_norm": 1.2761330604553223,
+      "learning_rate": 0.00019980348510561334,
+      "loss": 2.2427,
+      "step": 470
+    },
+    {
+      "epoch": 0.0495528669121515,
+      "grad_norm": 1.2825102806091309,
+      "learning_rate": 0.00019980134397914735,
+      "loss": 1.4184,
+      "step": 471
+    },
+    {
+      "epoch": 0.049658074697527615,
+      "grad_norm": 1.0994774103164673,
+      "learning_rate": 0.00019979919126307993,
+      "loss": 2.3454,
+      "step": 472
+    },
+    {
+      "epoch": 0.04976328248290374,
+      "grad_norm": 1.1263169050216675,
+      "learning_rate": 0.00019979702695766105,
+      "loss": 2.1226,
+      "step": 473
+    },
+    {
+      "epoch": 0.04986849026827985,
+      "grad_norm": 1.1242156028747559,
+      "learning_rate": 0.00019979485106314207,
+      "loss": 1.7008,
+      "step": 474
+    },
+    {
+      "epoch": 0.04997369805365597,
+      "grad_norm": 0.7600993514060974,
+      "learning_rate": 0.00019979266357977564,
+      "loss": 1.9561,
+      "step": 475
+    },
+    {
+      "epoch": 0.05007890583903209,
+      "grad_norm": 1.0761032104492188,
+      "learning_rate": 0.00019979046450781577,
+      "loss": 2.1057,
+      "step": 476
+    },
+    {
+      "epoch": 0.05018411362440821,
+      "grad_norm": 0.9772080779075623,
+      "learning_rate": 0.00019978825384751788,
+      "loss": 2.1573,
+      "step": 477
+    },
+    {
+      "epoch": 0.05028932140978432,
+      "grad_norm": 0.6181446313858032,
+      "learning_rate": 0.0001997860315991387,
+      "loss": 2.1602,
+      "step": 478
+    },
+    {
+      "epoch": 0.050394529195160444,
+      "grad_norm": 1.589167594909668,
+      "learning_rate": 0.0001997837977629362,
+      "loss": 2.1132,
+      "step": 479
+    },
+    {
+      "epoch": 0.05049973698053656,
+      "grad_norm": 0.6622167229652405,
+      "learning_rate": 0.0001997815523391699,
+      "loss": 2.331,
+      "step": 480
+    },
+    {
+      "epoch": 0.05060494476591268,
+      "grad_norm": 1.0673385858535767,
+      "learning_rate": 0.00019977929532810046,
+      "loss": 2.387,
+      "step": 481
+    },
+    {
+      "epoch": 0.050710152551288794,
+      "grad_norm": 0.897954523563385,
+      "learning_rate": 0.00019977702672999007,
+      "loss": 2.0587,
+      "step": 482
+    },
+    {
+      "epoch": 0.050815360336664915,
+      "grad_norm": 1.2834991216659546,
+      "learning_rate": 0.00019977474654510205,
+      "loss": 2.2961,
+      "step": 483
+    },
+    {
+      "epoch": 0.05092056812204103,
+      "grad_norm": 0.9525724649429321,
+      "learning_rate": 0.0001997724547737013,
+      "loss": 2.1939,
+      "step": 484
+    },
+    {
+      "epoch": 0.05102577590741715,
+      "grad_norm": 1.231590747833252,
+      "learning_rate": 0.00019977015141605392,
+      "loss": 2.0266,
+      "step": 485
+    },
+    {
+      "epoch": 0.051130983692793265,
+      "grad_norm": 2.677229642868042,
+      "learning_rate": 0.0001997678364724274,
+      "loss": 1.3896,
+      "step": 486
+    },
+    {
+      "epoch": 0.051236191478169386,
+      "grad_norm": 0.6867842674255371,
+      "learning_rate": 0.00019976550994309054,
+      "loss": 1.9468,
+      "step": 487
+    },
+    {
+      "epoch": 0.0513413992635455,
+      "grad_norm": 0.6714895963668823,
+      "learning_rate": 0.00019976317182831356,
+      "loss": 1.925,
+      "step": 488
+    },
+    {
+      "epoch": 0.05144660704892162,
+      "grad_norm": 1.246391773223877,
+      "learning_rate": 0.00019976082212836793,
+      "loss": 2.1784,
+      "step": 489
+    },
+    {
+      "epoch": 0.051551814834297736,
+      "grad_norm": 1.2437677383422852,
+      "learning_rate": 0.00019975846084352653,
+      "loss": 2.0721,
+      "step": 490
+    },
+    {
+      "epoch": 0.05165702261967386,
+      "grad_norm": 0.6305761933326721,
+      "learning_rate": 0.00019975608797406357,
+      "loss": 2.1311,
+      "step": 491
+    },
+    {
+      "epoch": 0.05176223040504997,
+      "grad_norm": 1.3557363748550415,
+      "learning_rate": 0.0001997537035202546,
+      "loss": 2.0776,
+      "step": 492
+    },
+    {
+      "epoch": 0.05186743819042609,
+      "grad_norm": 1.1781286001205444,
+      "learning_rate": 0.00019975130748237655,
+      "loss": 1.7604,
+      "step": 493
+    },
+    {
+      "epoch": 0.05197264597580221,
+      "grad_norm": 1.2240660190582275,
+      "learning_rate": 0.0001997488998607076,
+      "loss": 1.9454,
+      "step": 494
+    },
+    {
+      "epoch": 0.05207785376117833,
+      "grad_norm": 0.7471686601638794,
+      "learning_rate": 0.00019974648065552736,
+      "loss": 2.2171,
+      "step": 495
+    },
+    {
+      "epoch": 0.05218306154655444,
+      "grad_norm": 0.757163405418396,
+      "learning_rate": 0.0001997440498671168,
+      "loss": 2.1052,
+      "step": 496
+    },
+    {
+      "epoch": 0.052288269331930565,
+      "grad_norm": 0.6472220420837402,
+      "learning_rate": 0.00019974160749575818,
+      "loss": 1.8683,
+      "step": 497
+    },
+    {
+      "epoch": 0.05239347711730668,
+      "grad_norm": 1.0862151384353638,
+      "learning_rate": 0.00019973915354173515,
+      "loss": 1.9728,
+      "step": 498
+    },
+    {
+      "epoch": 0.0524986849026828,
+      "grad_norm": 0.7950903177261353,
+      "learning_rate": 0.00019973668800533264,
+      "loss": 1.7395,
+      "step": 499
+    },
+    {
+      "epoch": 0.052603892688058915,
+      "grad_norm": 0.8745871782302856,
+      "learning_rate": 0.00019973421088683696,
+      "loss": 1.7054,
+      "step": 500
+    },
+    {
+      "epoch": 0.052709100473435036,
+      "grad_norm": 0.8229191899299622,
+      "learning_rate": 0.00019973172218653578,
+      "loss": 1.9457,
+      "step": 501
+    },
+    {
+      "epoch": 0.05281430825881115,
+      "grad_norm": 1.065964698791504,
+      "learning_rate": 0.00019972922190471812,
+      "loss": 2.1223,
+      "step": 502
+    },
+    {
+      "epoch": 0.05291951604418727,
+      "grad_norm": 1.3047000169754028,
+      "learning_rate": 0.00019972671004167433,
+      "loss": 2.145,
+      "step": 503
+    },
+    {
+      "epoch": 0.053024723829563386,
+      "grad_norm": 0.7635353803634644,
+      "learning_rate": 0.00019972418659769606,
+      "loss": 2.4166,
+      "step": 504
+    },
+    {
+      "epoch": 0.05312993161493951,
+      "grad_norm": 0.7105801701545715,
+      "learning_rate": 0.00019972165157307643,
+      "loss": 2.3084,
+      "step": 505
+    },
+    {
+      "epoch": 0.05323513940031562,
+      "grad_norm": 1.485958218574524,
+      "learning_rate": 0.00019971910496810976,
+      "loss": 2.1874,
+      "step": 506
+    },
+    {
+      "epoch": 0.05334034718569174,
+      "grad_norm": 1.4032565355300903,
+      "learning_rate": 0.0001997165467830918,
+      "loss": 1.6684,
+      "step": 507
+    },
+    {
+      "epoch": 0.05344555497106786,
+      "grad_norm": 0.8490357398986816,
+      "learning_rate": 0.00019971397701831962,
+      "loss": 2.2013,
+      "step": 508
+    },
+    {
+      "epoch": 0.05355076275644398,
+      "grad_norm": 1.023547649383545,
+      "learning_rate": 0.00019971139567409165,
+      "loss": 1.885,
+      "step": 509
+    },
+    {
+      "epoch": 0.05365597054182009,
+      "grad_norm": 1.5144532918930054,
+      "learning_rate": 0.00019970880275070762,
+      "loss": 1.6336,
+      "step": 510
+    },
+    {
+      "epoch": 0.053761178327196214,
+      "grad_norm": 0.5086830258369446,
+      "learning_rate": 0.00019970619824846866,
+      "loss": 2.2339,
+      "step": 511
+    },
+    {
+      "epoch": 0.05386638611257233,
+      "grad_norm": 0.7949154376983643,
+      "learning_rate": 0.00019970358216767723,
+      "loss": 1.8825,
+      "step": 512
+    },
+    {
+      "epoch": 0.05397159389794845,
+      "grad_norm": 0.9173998832702637,
+      "learning_rate": 0.00019970095450863714,
+      "loss": 1.5642,
+      "step": 513
+    },
+    {
+      "epoch": 0.054076801683324564,
+      "grad_norm": 1.7043392658233643,
+      "learning_rate": 0.00019969831527165348,
+      "loss": 1.8851,
+      "step": 514
+    },
+    {
+      "epoch": 0.054182009468700686,
+      "grad_norm": 0.7839592099189758,
+      "learning_rate": 0.00019969566445703278,
+      "loss": 2.0123,
+      "step": 515
+    },
+    {
+      "epoch": 0.0542872172540768,
+      "grad_norm": 0.8127647638320923,
+      "learning_rate": 0.00019969300206508286,
+      "loss": 2.2239,
+      "step": 516
+    },
+    {
+      "epoch": 0.05439242503945292,
+      "grad_norm": 0.8179565668106079,
+      "learning_rate": 0.00019969032809611287,
+      "loss": 2.2176,
+      "step": 517
+    },
+    {
+      "epoch": 0.054497632824829036,
+      "grad_norm": 0.7557021975517273,
+      "learning_rate": 0.0001996876425504334,
+      "loss": 2.0041,
+      "step": 518
+    },
+    {
+      "epoch": 0.05460284061020516,
+      "grad_norm": 0.9029064178466797,
+      "learning_rate": 0.0001996849454283562,
+      "loss": 2.214,
+      "step": 519
+    },
+    {
+      "epoch": 0.05470804839558127,
+      "grad_norm": 0.778174638748169,
+      "learning_rate": 0.0001996822367301946,
+      "loss": 2.3064,
+      "step": 520
+    },
+    {
+      "epoch": 0.05481325618095739,
+      "grad_norm": 0.7716636061668396,
+      "learning_rate": 0.00019967951645626306,
+      "loss": 2.2683,
+      "step": 521
+    },
+    {
+      "epoch": 0.05491846396633351,
+      "grad_norm": 0.7747607231140137,
+      "learning_rate": 0.00019967678460687752,
+      "loss": 2.257,
+      "step": 522
+    },
+    {
+      "epoch": 0.05502367175170963,
+      "grad_norm": 1.1563743352890015,
+      "learning_rate": 0.00019967404118235521,
+      "loss": 1.6323,
+      "step": 523
+    },
+    {
+      "epoch": 0.05512887953708574,
+      "grad_norm": 0.7931325435638428,
+      "learning_rate": 0.0001996712861830147,
+      "loss": 1.4886,
+      "step": 524
+    },
+    {
+      "epoch": 0.055234087322461864,
+      "grad_norm": 1.187528133392334,
+      "learning_rate": 0.00019966851960917596,
+      "loss": 2.2017,
+      "step": 525
+    },
+    {
+      "epoch": 0.05533929510783798,
+      "grad_norm": 0.762750506401062,
+      "learning_rate": 0.00019966574146116023,
+      "loss": 1.953,
+      "step": 526
+    },
+    {
+      "epoch": 0.0554445028932141,
+      "grad_norm": 0.7893527150154114,
+      "learning_rate": 0.00019966295173929016,
+      "loss": 2.2539,
+      "step": 527
+    },
+    {
+      "epoch": 0.055549710678590214,
+      "grad_norm": 1.012707233428955,
+      "learning_rate": 0.00019966015044388966,
+      "loss": 2.0272,
+      "step": 528
+    },
+    {
+      "epoch": 0.055654918463966335,
+      "grad_norm": 0.7175372838973999,
+      "learning_rate": 0.00019965733757528405,
+      "loss": 2.2922,
+      "step": 529
+    },
+    {
+      "epoch": 0.05576012624934245,
+      "grad_norm": 1.1108931303024292,
+      "learning_rate": 0.0001996545131338,
+      "loss": 2.0014,
+      "step": 530
+    },
+    {
+      "epoch": 0.05586533403471857,
+      "grad_norm": 1.002982258796692,
+      "learning_rate": 0.00019965167711976552,
+      "loss": 1.9751,
+      "step": 531
+    },
+    {
+      "epoch": 0.055970541820094685,
+      "grad_norm": 1.2420021295547485,
+      "learning_rate": 0.00019964882953350989,
+      "loss": 1.8003,
+      "step": 532
+    },
+    {
+      "epoch": 0.05607574960547081,
+      "grad_norm": 0.8945255875587463,
+      "learning_rate": 0.00019964597037536383,
+      "loss": 2.0404,
+      "step": 533
+    },
+    {
+      "epoch": 0.05618095739084692,
+      "grad_norm": 0.8845522999763489,
+      "learning_rate": 0.00019964309964565937,
+      "loss": 1.8745,
+      "step": 534
+    },
+    {
+      "epoch": 0.05628616517622304,
+      "grad_norm": 0.826899528503418,
+      "learning_rate": 0.00019964021734472987,
+      "loss": 1.8893,
+      "step": 535
+    },
+    {
+      "epoch": 0.05639137296159916,
+      "grad_norm": 0.8589649200439453,
+      "learning_rate": 0.00019963732347291,
+      "loss": 2.2759,
+      "step": 536
+    },
+    {
+      "epoch": 0.05649658074697528,
+      "grad_norm": 0.9177697896957397,
+      "learning_rate": 0.00019963441803053588,
+      "loss": 1.9493,
+      "step": 537
+    },
+    {
+      "epoch": 0.05660178853235139,
+      "grad_norm": 1.1781582832336426,
+      "learning_rate": 0.0001996315010179449,
+      "loss": 1.8959,
+      "step": 538
+    },
+    {
+      "epoch": 0.056706996317727514,
+      "grad_norm": 1.119227647781372,
+      "learning_rate": 0.00019962857243547574,
+      "loss": 2.3161,
+      "step": 539
+    },
+    {
+      "epoch": 0.05681220410310363,
+      "grad_norm": 0.8425557017326355,
+      "learning_rate": 0.00019962563228346857,
+      "loss": 2.2769,
+      "step": 540
+    },
+    {
+      "epoch": 0.05691741188847975,
+      "grad_norm": 1.2644481658935547,
+      "learning_rate": 0.0001996226805622648,
+      "loss": 1.9871,
+      "step": 541
+    },
+    {
+      "epoch": 0.057022619673855864,
+      "grad_norm": 0.9367343187332153,
+      "learning_rate": 0.00019961971727220715,
+      "loss": 2.4598,
+      "step": 542
+    },
+    {
+      "epoch": 0.057127827459231985,
+      "grad_norm": 0.9345009326934814,
+      "learning_rate": 0.00019961674241363974,
+      "loss": 2.0872,
+      "step": 543
+    },
+    {
+      "epoch": 0.0572330352446081,
+      "grad_norm": 1.1807975769042969,
+      "learning_rate": 0.00019961375598690813,
+      "loss": 1.8824,
+      "step": 544
+    },
+    {
+      "epoch": 0.05733824302998422,
+      "grad_norm": 0.9418180584907532,
+      "learning_rate": 0.00019961075799235903,
+      "loss": 2.2195,
+      "step": 545
+    },
+    {
+      "epoch": 0.057443450815360335,
+      "grad_norm": 1.003548502922058,
+      "learning_rate": 0.0001996077484303406,
+      "loss": 1.5655,
+      "step": 546
+    },
+    {
+      "epoch": 0.057548658600736456,
+      "grad_norm": 0.9430425763130188,
+      "learning_rate": 0.00019960472730120237,
+      "loss": 1.9868,
+      "step": 547
+    },
+    {
+      "epoch": 0.05765386638611257,
+      "grad_norm": 0.7052996754646301,
+      "learning_rate": 0.00019960169460529515,
+      "loss": 2.1614,
+      "step": 548
+    },
+    {
+      "epoch": 0.05775907417148869,
+      "grad_norm": 0.9624035954475403,
+      "learning_rate": 0.0001995986503429711,
+      "loss": 1.8665,
+      "step": 549
+    },
+    {
+      "epoch": 0.057864281956864806,
+      "grad_norm": 0.8051895499229431,
+      "learning_rate": 0.00019959559451458375,
+      "loss": 2.1922,
+      "step": 550
+    },
+    {
+      "epoch": 0.05796948974224093,
+      "grad_norm": 5.186445713043213,
+      "learning_rate": 0.000199592527120488,
+      "loss": 2.3402,
+      "step": 551
+    },
+    {
+      "epoch": 0.05807469752761704,
+      "grad_norm": 0.7552983164787292,
+      "learning_rate": 0.00019958944816104,
+      "loss": 1.9901,
+      "step": 552
+    },
+    {
+      "epoch": 0.05817990531299316,
+      "grad_norm": 1.3142822980880737,
+      "learning_rate": 0.0001995863576365973,
+      "loss": 2.0759,
+      "step": 553
+    },
+    {
+      "epoch": 0.05828511309836928,
+      "grad_norm": 0.6937042474746704,
+      "learning_rate": 0.00019958325554751886,
+      "loss": 2.1585,
+      "step": 554
+    },
+    {
+      "epoch": 0.0583903208837454,
+      "grad_norm": 0.8620643615722656,
+      "learning_rate": 0.00019958014189416489,
+      "loss": 2.2186,
+      "step": 555
+    },
+    {
+      "epoch": 0.05849552866912151,
+      "grad_norm": 0.8564947843551636,
+      "learning_rate": 0.00019957701667689691,
+      "loss": 2.0109,
+      "step": 556
+    },
+    {
+      "epoch": 0.058600736454497634,
+      "grad_norm": 0.6378996968269348,
+      "learning_rate": 0.0001995738798960779,
+      "loss": 2.0941,
+      "step": 557
+    },
+    {
+      "epoch": 0.05870594423987375,
+      "grad_norm": 0.6846954822540283,
+      "learning_rate": 0.0001995707315520721,
+      "loss": 2.0447,
+      "step": 558
+    },
+    {
+      "epoch": 0.05881115202524987,
+      "grad_norm": 1.0269975662231445,
+      "learning_rate": 0.00019956757164524516,
+      "loss": 1.7775,
+      "step": 559
+    },
+    {
+      "epoch": 0.058916359810625984,
+      "grad_norm": 0.8012015223503113,
+      "learning_rate": 0.00019956440017596393,
+      "loss": 2.0707,
+      "step": 560
+    },
+    {
+      "epoch": 0.059021567596002106,
+      "grad_norm": 1.592552900314331,
+      "learning_rate": 0.0001995612171445968,
+      "loss": 1.9245,
+      "step": 561
+    },
+    {
+      "epoch": 0.05912677538137822,
+      "grad_norm": 0.7692115306854248,
+      "learning_rate": 0.00019955802255151338,
+      "loss": 2.3235,
+      "step": 562
+    },
+    {
+      "epoch": 0.05923198316675434,
+      "grad_norm": 0.8409086465835571,
+      "learning_rate": 0.00019955481639708463,
+      "loss": 2.0308,
+      "step": 563
+    },
+    {
+      "epoch": 0.059337190952130456,
+      "grad_norm": 0.7190937399864197,
+      "learning_rate": 0.00019955159868168288,
+      "loss": 2.2148,
+      "step": 564
+    },
+    {
+      "epoch": 0.05944239873750658,
+      "grad_norm": 1.3858685493469238,
+      "learning_rate": 0.00019954836940568177,
+      "loss": 2.1553,
+      "step": 565
+    },
+    {
+      "epoch": 0.05954760652288269,
+      "grad_norm": 0.7564934492111206,
+      "learning_rate": 0.00019954512856945632,
+      "loss": 1.713,
+      "step": 566
+    },
+    {
+      "epoch": 0.05965281430825881,
+      "grad_norm": 1.1952269077301025,
+      "learning_rate": 0.00019954187617338294,
+      "loss": 2.1569,
+      "step": 567
+    },
+    {
+      "epoch": 0.05975802209363493,
+      "grad_norm": 1.282727837562561,
+      "learning_rate": 0.00019953861221783922,
+      "loss": 2.0503,
+      "step": 568
+    },
+    {
+      "epoch": 0.05986322987901105,
+      "grad_norm": 0.8127844333648682,
+      "learning_rate": 0.00019953533670320422,
+      "loss": 1.7488,
+      "step": 569
+    },
+    {
+      "epoch": 0.05996843766438716,
+      "grad_norm": 1.0792090892791748,
+      "learning_rate": 0.00019953204962985837,
+      "loss": 1.9068,
+      "step": 570
+    },
+    {
+      "epoch": 0.060073645449763284,
+      "grad_norm": 0.9024432897567749,
+      "learning_rate": 0.00019952875099818332,
+      "loss": 2.2592,
+      "step": 571
+    },
+    {
+      "epoch": 0.0601788532351394,
+      "grad_norm": 1.1777130365371704,
+      "learning_rate": 0.0001995254408085622,
+      "loss": 2.4742,
+      "step": 572
+    },
+    {
+      "epoch": 0.06028406102051552,
+      "grad_norm": 0.7864671945571899,
+      "learning_rate": 0.00019952211906137932,
+      "loss": 2.0437,
+      "step": 573
+    },
+    {
+      "epoch": 0.060389268805891634,
+      "grad_norm": 0.9623721837997437,
+      "learning_rate": 0.00019951878575702047,
+      "loss": 2.3201,
+      "step": 574
+    },
+    {
+      "epoch": 0.060494476591267755,
+      "grad_norm": 0.7268714904785156,
+      "learning_rate": 0.00019951544089587278,
+      "loss": 1.9149,
+      "step": 575
+    },
+    {
+      "epoch": 0.06059968437664387,
+      "grad_norm": 1.013013243675232,
+      "learning_rate": 0.00019951208447832461,
+      "loss": 1.7478,
+      "step": 576
+    },
+    {
+      "epoch": 0.06070489216201999,
+      "grad_norm": 0.6717780232429504,
+      "learning_rate": 0.00019950871650476577,
+      "loss": 2.208,
+      "step": 577
+    },
+    {
+      "epoch": 0.060810099947396105,
+      "grad_norm": 2.0669307708740234,
+      "learning_rate": 0.00019950533697558732,
+      "loss": 2.185,
+      "step": 578
+    },
+    {
+      "epoch": 0.06091530773277223,
+      "grad_norm": 1.002172827720642,
+      "learning_rate": 0.0001995019458911818,
+      "loss": 2.0923,
+      "step": 579
+    },
+    {
+      "epoch": 0.06102051551814834,
+      "grad_norm": 1.053417444229126,
+      "learning_rate": 0.00019949854325194294,
+      "loss": 2.0336,
+      "step": 580
+    },
+    {
+      "epoch": 0.06112572330352446,
+      "grad_norm": 0.8169642686843872,
+      "learning_rate": 0.0001994951290582659,
+      "loss": 2.1974,
+      "step": 581
+    },
+    {
+      "epoch": 0.06123093108890058,
+      "grad_norm": 0.8666344881057739,
+      "learning_rate": 0.0001994917033105471,
+      "loss": 2.1284,
+      "step": 582
+    },
+    {
+      "epoch": 0.0613361388742767,
+      "grad_norm": 0.8064733743667603,
+      "learning_rate": 0.00019948826600918443,
+      "loss": 2.0592,
+      "step": 583
+    },
+    {
+      "epoch": 0.06144134665965281,
+      "grad_norm": 1.0283541679382324,
+      "learning_rate": 0.00019948481715457707,
+      "loss": 2.2653,
+      "step": 584
+    },
+    {
+      "epoch": 0.061546554445028934,
+      "grad_norm": 1.1109108924865723,
+      "learning_rate": 0.00019948135674712546,
+      "loss": 2.0874,
+      "step": 585
+    },
+    {
+      "epoch": 0.06165176223040505,
+      "grad_norm": 0.8930381536483765,
+      "learning_rate": 0.00019947788478723153,
+      "loss": 1.988,
+      "step": 586
+    },
+    {
+      "epoch": 0.06175697001578117,
+      "grad_norm": 0.8919090032577515,
+      "learning_rate": 0.00019947440127529836,
+      "loss": 2.3961,
+      "step": 587
+    },
+    {
+      "epoch": 0.061862177801157284,
+      "grad_norm": 0.9803879857063293,
+      "learning_rate": 0.00019947090621173053,
+      "loss": 2.2986,
+      "step": 588
+    },
+    {
+      "epoch": 0.061967385586533405,
+      "grad_norm": 2.3181352615356445,
+      "learning_rate": 0.00019946739959693393,
+      "loss": 2.3345,
+      "step": 589
+    },
+    {
+      "epoch": 0.06207259337190952,
+      "grad_norm": 1.0249263048171997,
+      "learning_rate": 0.00019946388143131575,
+      "loss": 1.9578,
+      "step": 590
+    },
+    {
+      "epoch": 0.06217780115728564,
+      "grad_norm": 1.8943885564804077,
+      "learning_rate": 0.00019946035171528455,
+      "loss": 1.5106,
+      "step": 591
+    },
+    {
+      "epoch": 0.062283008942661755,
+      "grad_norm": 1.0123568773269653,
+      "learning_rate": 0.0001994568104492502,
+      "loss": 2.1627,
+      "step": 592
+    },
+    {
+      "epoch": 0.062388216728037876,
+      "grad_norm": 0.9259027242660522,
+      "learning_rate": 0.00019945325763362398,
+      "loss": 1.6539,
+      "step": 593
+    },
+    {
+      "epoch": 0.06249342451341399,
+      "grad_norm": 0.9884849786758423,
+      "learning_rate": 0.00019944969326881845,
+      "loss": 2.0843,
+      "step": 594
+    },
+    {
+      "epoch": 0.0625986322987901,
+      "grad_norm": 0.735097348690033,
+      "learning_rate": 0.0001994461173552475,
+      "loss": 2.4694,
+      "step": 595
+    },
+    {
+      "epoch": 0.06270384008416623,
+      "grad_norm": 1.136003017425537,
+      "learning_rate": 0.0001994425298933264,
+      "loss": 1.7615,
+      "step": 596
+    },
+    {
+      "epoch": 0.06280904786954235,
+      "grad_norm": 1.316763162612915,
+      "learning_rate": 0.00019943893088347178,
+      "loss": 2.3421,
+      "step": 597
+    },
+    {
+      "epoch": 0.06291425565491847,
+      "grad_norm": 1.165870189666748,
+      "learning_rate": 0.00019943532032610156,
+      "loss": 2.354,
+      "step": 598
+    },
+    {
+      "epoch": 0.06301946344029458,
+      "grad_norm": 12.435591697692871,
+      "learning_rate": 0.00019943169822163502,
+      "loss": 2.0802,
+      "step": 599
+    },
+    {
+      "epoch": 0.0631246712256707,
+      "grad_norm": 1.253097414970398,
+      "learning_rate": 0.00019942806457049278,
+      "loss": 2.0145,
+      "step": 600
+    },
+    {
+      "epoch": 0.06322987901104682,
+      "grad_norm": 1.299344539642334,
+      "learning_rate": 0.00019942441937309684,
+      "loss": 1.7266,
+      "step": 601
+    },
+    {
+      "epoch": 0.06333508679642294,
+      "grad_norm": 0.9620880484580994,
+      "learning_rate": 0.00019942076262987043,
+      "loss": 2.2549,
+      "step": 602
+    },
+    {
+      "epoch": 0.06344029458179905,
+      "grad_norm": 1.1274274587631226,
+      "learning_rate": 0.00019941709434123826,
+      "loss": 1.9165,
+      "step": 603
+    },
+    {
+      "epoch": 0.06354550236717517,
+      "grad_norm": 0.8177148103713989,
+      "learning_rate": 0.00019941341450762629,
+      "loss": 2.0151,
+      "step": 604
+    },
+    {
+      "epoch": 0.06365071015255129,
+      "grad_norm": 1.068963885307312,
+      "learning_rate": 0.00019940972312946186,
+      "loss": 2.1957,
+      "step": 605
+    },
+    {
+      "epoch": 0.06375591793792741,
+      "grad_norm": 0.8297029733657837,
+      "learning_rate": 0.00019940602020717364,
+      "loss": 2.2105,
+      "step": 606
+    },
+    {
+      "epoch": 0.06386112572330352,
+      "grad_norm": 0.8417388796806335,
+      "learning_rate": 0.00019940230574119164,
+      "loss": 1.9136,
+      "step": 607
+    },
+    {
+      "epoch": 0.06396633350867964,
+      "grad_norm": 0.8712098598480225,
+      "learning_rate": 0.00019939857973194717,
+      "loss": 2.4713,
+      "step": 608
+    },
+    {
+      "epoch": 0.06407154129405576,
+      "grad_norm": 1.0298516750335693,
+      "learning_rate": 0.000199394842179873,
+      "loss": 1.8507,
+      "step": 609
+    },
+    {
+      "epoch": 0.06417674907943188,
+      "grad_norm": 0.9027281999588013,
+      "learning_rate": 0.00019939109308540304,
+      "loss": 2.0152,
+      "step": 610
+    },
+    {
+      "epoch": 0.06428195686480799,
+      "grad_norm": 1.213963270187378,
+      "learning_rate": 0.00019938733244897274,
+      "loss": 1.9294,
+      "step": 611
+    },
+    {
+      "epoch": 0.06438716465018411,
+      "grad_norm": 1.2193472385406494,
+      "learning_rate": 0.00019938356027101884,
+      "loss": 1.9002,
+      "step": 612
+    },
+    {
+      "epoch": 0.06449237243556023,
+      "grad_norm": 1.7133346796035767,
+      "learning_rate": 0.0001993797765519793,
+      "loss": 2.3218,
+      "step": 613
+    },
+    {
+      "epoch": 0.06459758022093635,
+      "grad_norm": 1.3086202144622803,
+      "learning_rate": 0.0001993759812922936,
+      "loss": 2.2043,
+      "step": 614
+    },
+    {
+      "epoch": 0.06470278800631246,
+      "grad_norm": 1.2867093086242676,
+      "learning_rate": 0.0001993721744924024,
+      "loss": 2.1243,
+      "step": 615
+    },
+    {
+      "epoch": 0.06480799579168858,
+      "grad_norm": 1.0671265125274658,
+      "learning_rate": 0.00019936835615274782,
+      "loss": 2.1034,
+      "step": 616
+    },
+    {
+      "epoch": 0.0649132035770647,
+      "grad_norm": 0.6220844984054565,
+      "learning_rate": 0.00019936452627377323,
+      "loss": 2.1934,
+      "step": 617
+    },
+    {
+      "epoch": 0.06501841136244083,
+      "grad_norm": 0.9203489422798157,
+      "learning_rate": 0.0001993606848559234,
+      "loss": 2.1497,
+      "step": 618
+    },
+    {
+      "epoch": 0.06512361914781693,
+      "grad_norm": 1.5812958478927612,
+      "learning_rate": 0.00019935683189964447,
+      "loss": 2.0074,
+      "step": 619
+    },
+    {
+      "epoch": 0.06522882693319305,
+      "grad_norm": 0.9502617120742798,
+      "learning_rate": 0.00019935296740538377,
+      "loss": 2.3318,
+      "step": 620
+    },
+    {
+      "epoch": 0.06533403471856918,
+      "grad_norm": 1.632648229598999,
+      "learning_rate": 0.00019934909137359018,
+      "loss": 1.8631,
+      "step": 621
+    },
+    {
+      "epoch": 0.0654392425039453,
+      "grad_norm": 1.4153451919555664,
+      "learning_rate": 0.00019934520380471372,
+      "loss": 1.6571,
+      "step": 622
+    },
+    {
+      "epoch": 0.0655444502893214,
+      "grad_norm": 0.8237829804420471,
+      "learning_rate": 0.00019934130469920588,
+      "loss": 1.9875,
+      "step": 623
+    },
+    {
+      "epoch": 0.06564965807469753,
+      "grad_norm": 0.8410417437553406,
+      "learning_rate": 0.00019933739405751945,
+      "loss": 2.0338,
+      "step": 624
+    },
+    {
+      "epoch": 0.06575486586007365,
+      "grad_norm": 0.9069095253944397,
+      "learning_rate": 0.00019933347188010858,
+      "loss": 2.1402,
+      "step": 625
+    },
+    {
+      "epoch": 0.06586007364544977,
+      "grad_norm": 1.0030279159545898,
+      "learning_rate": 0.0001993295381674287,
+      "loss": 2.1201,
+      "step": 626
+    },
+    {
+      "epoch": 0.06596528143082588,
+      "grad_norm": 1.0119670629501343,
+      "learning_rate": 0.00019932559291993665,
+      "loss": 1.9571,
+      "step": 627
+    },
+    {
+      "epoch": 0.066070489216202,
+      "grad_norm": 1.5967551469802856,
+      "learning_rate": 0.00019932163613809055,
+      "loss": 1.943,
+      "step": 628
+    },
+    {
+      "epoch": 0.06617569700157812,
+      "grad_norm": 0.7849034070968628,
+      "learning_rate": 0.0001993176678223499,
+      "loss": 2.3237,
+      "step": 629
+    },
+    {
+      "epoch": 0.06628090478695424,
+      "grad_norm": 0.9208400249481201,
+      "learning_rate": 0.00019931368797317553,
+      "loss": 2.4865,
+      "step": 630
+    },
+    {
+      "epoch": 0.06638611257233035,
+      "grad_norm": 1.232565999031067,
+      "learning_rate": 0.00019930969659102962,
+      "loss": 2.2367,
+      "step": 631
+    },
+    {
+      "epoch": 0.06649132035770647,
+      "grad_norm": 0.7089071273803711,
+      "learning_rate": 0.0001993056936763757,
+      "loss": 2.3407,
+      "step": 632
+    },
+    {
+      "epoch": 0.06659652814308259,
+      "grad_norm": 0.9543630480766296,
+      "learning_rate": 0.00019930167922967853,
+      "loss": 1.8678,
+      "step": 633
+    },
+    {
+      "epoch": 0.06670173592845871,
+      "grad_norm": 1.19600510597229,
+      "learning_rate": 0.00019929765325140436,
+      "loss": 1.9537,
+      "step": 634
+    },
+    {
+      "epoch": 0.06680694371383482,
+      "grad_norm": 0.7090086936950684,
+      "learning_rate": 0.0001992936157420207,
+      "loss": 1.946,
+      "step": 635
+    },
+    {
+      "epoch": 0.06691215149921094,
+      "grad_norm": 55.697574615478516,
+      "learning_rate": 0.0001992895667019964,
+      "loss": 2.0818,
+      "step": 636
+    },
+    {
+      "epoch": 0.06701735928458706,
+      "grad_norm": 0.9292970299720764,
+      "learning_rate": 0.00019928550613180164,
+      "loss": 2.256,
+      "step": 637
+    },
+    {
+      "epoch": 0.06712256706996318,
+      "grad_norm": 1.2075432538986206,
+      "learning_rate": 0.00019928143403190806,
+      "loss": 1.8407,
+      "step": 638
+    },
+    {
+      "epoch": 0.06722777485533929,
+      "grad_norm": 0.8975555300712585,
+      "learning_rate": 0.00019927735040278842,
+      "loss": 2.1384,
+      "step": 639
+    },
+    {
+      "epoch": 0.06733298264071541,
+      "grad_norm": 0.9104759097099304,
+      "learning_rate": 0.00019927325524491703,
+      "loss": 2.1348,
+      "step": 640
+    },
+    {
+      "epoch": 0.06743819042609153,
+      "grad_norm": 0.8688013553619385,
+      "learning_rate": 0.0001992691485587694,
+      "loss": 1.7637,
+      "step": 641
+    },
+    {
+      "epoch": 0.06754339821146765,
+      "grad_norm": 1.2788143157958984,
+      "learning_rate": 0.0001992650303448224,
+      "loss": 1.8858,
+      "step": 642
+    },
+    {
+      "epoch": 0.06764860599684376,
+      "grad_norm": 0.7544511556625366,
+      "learning_rate": 0.0001992609006035543,
+      "loss": 1.9076,
+      "step": 643
+    },
+    {
+      "epoch": 0.06775381378221988,
+      "grad_norm": 1.3862735033035278,
+      "learning_rate": 0.00019925675933544473,
+      "loss": 2.212,
+      "step": 644
+    },
+    {
+      "epoch": 0.067859021567596,
+      "grad_norm": 1.04462730884552,
+      "learning_rate": 0.00019925260654097448,
+      "loss": 1.9386,
+      "step": 645
+    },
+    {
+      "epoch": 0.06796422935297212,
+      "grad_norm": 0.9594143033027649,
+      "learning_rate": 0.0001992484422206259,
+      "loss": 2.0205,
+      "step": 646
+    },
+    {
+      "epoch": 0.06806943713834823,
+      "grad_norm": 1.3281629085540771,
+      "learning_rate": 0.00019924426637488252,
+      "loss": 1.6588,
+      "step": 647
+    },
+    {
+      "epoch": 0.06817464492372435,
+      "grad_norm": 1.1453989744186401,
+      "learning_rate": 0.0001992400790042293,
+      "loss": 1.8736,
+      "step": 648
+    },
+    {
+      "epoch": 0.06827985270910047,
+      "grad_norm": 1.5039159059524536,
+      "learning_rate": 0.0001992358801091525,
+      "loss": 1.3861,
+      "step": 649
+    },
+    {
+      "epoch": 0.0683850604944766,
+      "grad_norm": 1.353845477104187,
+      "learning_rate": 0.0001992316696901397,
+      "loss": 1.8725,
+      "step": 650
+    },
+    {
+      "epoch": 0.0684902682798527,
+      "grad_norm": 0.9831173419952393,
+      "learning_rate": 0.00019922744774767987,
+      "loss": 2.1601,
+      "step": 651
+    },
+    {
+      "epoch": 0.06859547606522882,
+      "grad_norm": 2.004272937774658,
+      "learning_rate": 0.0001992232142822633,
+      "loss": 1.7202,
+      "step": 652
+    },
+    {
+      "epoch": 0.06870068385060495,
+      "grad_norm": 0.7142511010169983,
+      "learning_rate": 0.00019921896929438158,
+      "loss": 2.1088,
+      "step": 653
+    },
+    {
+      "epoch": 0.06880589163598107,
+      "grad_norm": 0.8999900817871094,
+      "learning_rate": 0.00019921471278452768,
+      "loss": 1.9379,
+      "step": 654
+    },
+    {
+      "epoch": 0.06891109942135717,
+      "grad_norm": 1.3912204504013062,
+      "learning_rate": 0.00019921044475319585,
+      "loss": 1.6042,
+      "step": 655
+    },
+    {
+      "epoch": 0.0690163072067333,
+      "grad_norm": 0.8359988331794739,
+      "learning_rate": 0.0001992061652008818,
+      "loss": 2.0374,
+      "step": 656
+    },
+    {
+      "epoch": 0.06912151499210942,
+      "grad_norm": 0.8871389627456665,
+      "learning_rate": 0.00019920187412808248,
+      "loss": 2.088,
+      "step": 657
+    },
+    {
+      "epoch": 0.06922672277748554,
+      "grad_norm": 1.1207047700881958,
+      "learning_rate": 0.00019919757153529614,
+      "loss": 1.8852,
+      "step": 658
+    },
+    {
+      "epoch": 0.06933193056286165,
+      "grad_norm": 1.3903989791870117,
+      "learning_rate": 0.0001991932574230225,
+      "loss": 2.1249,
+      "step": 659
+    },
+    {
+      "epoch": 0.06943713834823777,
+      "grad_norm": 0.6716268062591553,
+      "learning_rate": 0.00019918893179176253,
+      "loss": 2.1855,
+      "step": 660
+    },
+    {
+      "epoch": 0.06954234613361389,
+      "grad_norm": 1.388175368309021,
+      "learning_rate": 0.0001991845946420185,
+      "loss": 2.5185,
+      "step": 661
+    },
+    {
+      "epoch": 0.06964755391899001,
+      "grad_norm": 1.7384852170944214,
+      "learning_rate": 0.0001991802459742941,
+      "loss": 1.2171,
+      "step": 662
+    },
+    {
+      "epoch": 0.06975276170436612,
+      "grad_norm": 1.0405939817428589,
+      "learning_rate": 0.0001991758857890943,
+      "loss": 2.0206,
+      "step": 663
+    },
+    {
+      "epoch": 0.06985796948974224,
+      "grad_norm": 0.9668706655502319,
+      "learning_rate": 0.0001991715140869255,
+      "loss": 2.1,
+      "step": 664
+    },
+    {
+      "epoch": 0.06996317727511836,
+      "grad_norm": 1.021179437637329,
+      "learning_rate": 0.00019916713086829533,
+      "loss": 1.8814,
+      "step": 665
+    },
+    {
+      "epoch": 0.07006838506049448,
+      "grad_norm": 0.6269777417182922,
+      "learning_rate": 0.0001991627361337128,
+      "loss": 2.5095,
+      "step": 666
+    },
+    {
+      "epoch": 0.07017359284587059,
+      "grad_norm": 1.236629605293274,
+      "learning_rate": 0.00019915832988368824,
+      "loss": 2.1106,
+      "step": 667
+    },
+    {
+      "epoch": 0.07027880063124671,
+      "grad_norm": 1.0450339317321777,
+      "learning_rate": 0.0001991539121187334,
+      "loss": 2.1438,
+      "step": 668
+    },
+    {
+      "epoch": 0.07038400841662283,
+      "grad_norm": 0.8368414044380188,
+      "learning_rate": 0.00019914948283936119,
+      "loss": 2.1143,
+      "step": 669
+    },
+    {
+      "epoch": 0.07048921620199895,
+      "grad_norm": 1.016280174255371,
+      "learning_rate": 0.0001991450420460861,
+      "loss": 2.1759,
+      "step": 670
+    },
+    {
+      "epoch": 0.07059442398737506,
+      "grad_norm": 1.2927619218826294,
+      "learning_rate": 0.00019914058973942368,
+      "loss": 1.9304,
+      "step": 671
+    },
+    {
+      "epoch": 0.07069963177275118,
+      "grad_norm": 1.0759140253067017,
+      "learning_rate": 0.0001991361259198911,
+      "loss": 1.7529,
+      "step": 672
+    },
+    {
+      "epoch": 0.0708048395581273,
+      "grad_norm": 0.802558183670044,
+      "learning_rate": 0.00019913165058800663,
+      "loss": 2.4059,
+      "step": 673
+    },
+    {
+      "epoch": 0.07091004734350342,
+      "grad_norm": 0.8148163557052612,
+      "learning_rate": 0.00019912716374429,
+      "loss": 2.1683,
+      "step": 674
+    },
+    {
+      "epoch": 0.07101525512887953,
+      "grad_norm": 1.259080171585083,
+      "learning_rate": 0.00019912266538926225,
+      "loss": 2.1657,
+      "step": 675
+    },
+    {
+      "epoch": 0.07112046291425565,
+      "grad_norm": 1.0984203815460205,
+      "learning_rate": 0.00019911815552344582,
+      "loss": 2.0398,
+      "step": 676
+    },
+    {
+      "epoch": 0.07122567069963177,
+      "grad_norm": 1.0263009071350098,
+      "learning_rate": 0.00019911363414736434,
+      "loss": 1.7294,
+      "step": 677
+    },
+    {
+      "epoch": 0.0713308784850079,
+      "grad_norm": 1.2345585823059082,
+      "learning_rate": 0.00019910910126154293,
+      "loss": 2.333,
+      "step": 678
+    },
+    {
+      "epoch": 0.071436086270384,
+      "grad_norm": 0.7678645253181458,
+      "learning_rate": 0.00019910455686650793,
+      "loss": 2.2339,
+      "step": 679
+    },
+    {
+      "epoch": 0.07154129405576012,
+      "grad_norm": 0.8823609948158264,
+      "learning_rate": 0.0001991000009627871,
+      "loss": 2.0109,
+      "step": 680
+    },
+    {
+      "epoch": 0.07164650184113625,
+      "grad_norm": 0.9095181822776794,
+      "learning_rate": 0.00019909543355090946,
+      "loss": 2.3621,
+      "step": 681
+    },
+    {
+      "epoch": 0.07175170962651237,
+      "grad_norm": 1.3118935823440552,
+      "learning_rate": 0.00019909085463140546,
+      "loss": 1.9576,
+      "step": 682
+    },
+    {
+      "epoch": 0.07185691741188847,
+      "grad_norm": 1.0895490646362305,
+      "learning_rate": 0.0001990862642048068,
+      "loss": 1.8436,
+      "step": 683
+    },
+    {
+      "epoch": 0.0719621251972646,
+      "grad_norm": 1.1863242387771606,
+      "learning_rate": 0.00019908166227164655,
+      "loss": 1.8009,
+      "step": 684
+    },
+    {
+      "epoch": 0.07206733298264072,
+      "grad_norm": 1.0298961400985718,
+      "learning_rate": 0.00019907704883245916,
+      "loss": 1.9716,
+      "step": 685
+    },
+    {
+      "epoch": 0.07217254076801684,
+      "grad_norm": 0.7246216535568237,
+      "learning_rate": 0.00019907242388778033,
+      "loss": 2.119,
+      "step": 686
+    },
+    {
+      "epoch": 0.07227774855339295,
+      "grad_norm": 1.172180414199829,
+      "learning_rate": 0.00019906778743814711,
+      "loss": 2.3032,
+      "step": 687
+    },
+    {
+      "epoch": 0.07238295633876907,
+      "grad_norm": 1.1342841386795044,
+      "learning_rate": 0.000199063139484098,
+      "loss": 2.4119,
+      "step": 688
+    },
+    {
+      "epoch": 0.07248816412414519,
+      "grad_norm": 1.160383701324463,
+      "learning_rate": 0.0001990584800261727,
+      "loss": 1.9328,
+      "step": 689
+    },
+    {
+      "epoch": 0.07259337190952131,
+      "grad_norm": 0.991884171962738,
+      "learning_rate": 0.00019905380906491232,
+      "loss": 2.1967,
+      "step": 690
+    },
+    {
+      "epoch": 0.07269857969489742,
+      "grad_norm": 1.0760473012924194,
+      "learning_rate": 0.00019904912660085927,
+      "loss": 1.9094,
+      "step": 691
+    },
+    {
+      "epoch": 0.07280378748027354,
+      "grad_norm": 1.6143467426300049,
+      "learning_rate": 0.00019904443263455728,
+      "loss": 1.8885,
+      "step": 692
+    },
+    {
+      "epoch": 0.07290899526564966,
+      "grad_norm": 1.200732707977295,
+      "learning_rate": 0.00019903972716655148,
+      "loss": 1.8561,
+      "step": 693
+    },
+    {
+      "epoch": 0.07301420305102578,
+      "grad_norm": 0.9255091547966003,
+      "learning_rate": 0.0001990350101973883,
+      "loss": 1.9438,
+      "step": 694
+    },
+    {
+      "epoch": 0.07311941083640189,
+      "grad_norm": 0.8052567839622498,
+      "learning_rate": 0.00019903028172761552,
+      "loss": 1.8901,
+      "step": 695
+    },
+    {
+      "epoch": 0.07322461862177801,
+      "grad_norm": 0.843561589717865,
+      "learning_rate": 0.00019902554175778222,
+      "loss": 2.1764,
+      "step": 696
+    },
+    {
+      "epoch": 0.07332982640715413,
+      "grad_norm": 1.0128999948501587,
+      "learning_rate": 0.0001990207902884388,
+      "loss": 1.6193,
+      "step": 697
+    },
+    {
+      "epoch": 0.07343503419253025,
+      "grad_norm": 1.1728532314300537,
+      "learning_rate": 0.00019901602732013709,
+      "loss": 1.8212,
+      "step": 698
+    },
+    {
+      "epoch": 0.07354024197790636,
+      "grad_norm": 0.6884058117866516,
+      "learning_rate": 0.00019901125285343022,
+      "loss": 2.1301,
+      "step": 699
+    },
+    {
+      "epoch": 0.07364544976328248,
+      "grad_norm": 0.8318617343902588,
+      "learning_rate": 0.00019900646688887253,
+      "loss": 2.3534,
+      "step": 700
+    },
+    {
+      "epoch": 0.0737506575486586,
+      "grad_norm": 0.7841008901596069,
+      "learning_rate": 0.0001990016694270199,
+      "loss": 1.9417,
+      "step": 701
+    },
+    {
+      "epoch": 0.07385586533403472,
+      "grad_norm": 0.7915363907814026,
+      "learning_rate": 0.0001989968604684294,
+      "loss": 2.337,
+      "step": 702
+    },
+    {
+      "epoch": 0.07396107311941083,
+      "grad_norm": 0.9618809819221497,
+      "learning_rate": 0.00019899204001365948,
+      "loss": 2.1485,
+      "step": 703
+    },
+    {
+      "epoch": 0.07406628090478695,
+      "grad_norm": 1.1301730871200562,
+      "learning_rate": 0.00019898720806326993,
+      "loss": 2.3046,
+      "step": 704
+    },
+    {
+      "epoch": 0.07417148869016307,
+      "grad_norm": 0.922090470790863,
+      "learning_rate": 0.00019898236461782186,
+      "loss": 1.9871,
+      "step": 705
+    },
+    {
+      "epoch": 0.0742766964755392,
+      "grad_norm": 0.9652605056762695,
+      "learning_rate": 0.0001989775096778777,
+      "loss": 2.1545,
+      "step": 706
+    },
+    {
+      "epoch": 0.0743819042609153,
+      "grad_norm": 0.8259909749031067,
+      "learning_rate": 0.00019897264324400128,
+      "loss": 1.9013,
+      "step": 707
+    },
+    {
+      "epoch": 0.07448711204629142,
+      "grad_norm": 0.7845616340637207,
+      "learning_rate": 0.00019896776531675773,
+      "loss": 2.0664,
+      "step": 708
+    },
+    {
+      "epoch": 0.07459231983166754,
+      "grad_norm": 0.9520933032035828,
+      "learning_rate": 0.0001989628758967135,
+      "loss": 1.898,
+      "step": 709
+    },
+    {
+      "epoch": 0.07469752761704367,
+      "grad_norm": 1.1913421154022217,
+      "learning_rate": 0.00019895797498443633,
+      "loss": 1.794,
+      "step": 710
+    },
+    {
+      "epoch": 0.07480273540241977,
+      "grad_norm": 0.8856778144836426,
+      "learning_rate": 0.00019895306258049542,
+      "loss": 2.2846,
+      "step": 711
+    },
+    {
+      "epoch": 0.0749079431877959,
+      "grad_norm": 1.1165918111801147,
+      "learning_rate": 0.00019894813868546115,
+      "loss": 1.8066,
+      "step": 712
+    },
+    {
+      "epoch": 0.07501315097317202,
+      "grad_norm": 1.1976250410079956,
+      "learning_rate": 0.0001989432032999054,
+      "loss": 1.9725,
+      "step": 713
+    },
+    {
+      "epoch": 0.07511835875854814,
+      "grad_norm": 1.1325089931488037,
+      "learning_rate": 0.00019893825642440128,
+      "loss": 2.4161,
+      "step": 714
+    },
+    {
+      "epoch": 0.07522356654392424,
+      "grad_norm": 1.7667649984359741,
+      "learning_rate": 0.0001989332980595232,
+      "loss": 1.7692,
+      "step": 715
+    },
+    {
+      "epoch": 0.07532877432930037,
+      "grad_norm": 0.7589969038963318,
+      "learning_rate": 0.00019892832820584704,
+      "loss": 1.9779,
+      "step": 716
+    },
+    {
+      "epoch": 0.07543398211467649,
+      "grad_norm": 1.1662501096725464,
+      "learning_rate": 0.00019892334686394985,
+      "loss": 2.1976,
+      "step": 717
+    },
+    {
+      "epoch": 0.07553918990005261,
+      "grad_norm": 0.9506317973136902,
+      "learning_rate": 0.00019891835403441013,
+      "loss": 2.2166,
+      "step": 718
+    },
+    {
+      "epoch": 0.07564439768542872,
+      "grad_norm": 0.8529815077781677,
+      "learning_rate": 0.00019891334971780772,
+      "loss": 2.1537,
+      "step": 719
+    },
+    {
+      "epoch": 0.07574960547080484,
+      "grad_norm": 0.7838415503501892,
+      "learning_rate": 0.0001989083339147237,
+      "loss": 2.0705,
+      "step": 720
+    },
+    {
+      "epoch": 0.07585481325618096,
+      "grad_norm": 0.8060374855995178,
+      "learning_rate": 0.00019890330662574056,
+      "loss": 2.0404,
+      "step": 721
+    },
+    {
+      "epoch": 0.07596002104155708,
+      "grad_norm": 1.2424758672714233,
+      "learning_rate": 0.0001988982678514421,
+      "loss": 2.1626,
+      "step": 722
+    },
+    {
+      "epoch": 0.07606522882693319,
+      "grad_norm": 0.752369225025177,
+      "learning_rate": 0.00019889321759241347,
+      "loss": 2.0623,
+      "step": 723
+    },
+    {
+      "epoch": 0.07617043661230931,
+      "grad_norm": 1.5461053848266602,
+      "learning_rate": 0.00019888815584924113,
+      "loss": 2.5155,
+      "step": 724
+    },
+    {
+      "epoch": 0.07627564439768543,
+      "grad_norm": 0.9477027654647827,
+      "learning_rate": 0.00019888308262251285,
+      "loss": 2.526,
+      "step": 725
+    },
+    {
+      "epoch": 0.07638085218306155,
+      "grad_norm": 1.75247061252594,
+      "learning_rate": 0.00019887799791281784,
+      "loss": 1.9596,
+      "step": 726
+    },
+    {
+      "epoch": 0.07648605996843766,
+      "grad_norm": 1.1626888513565063,
+      "learning_rate": 0.0001988729017207465,
+      "loss": 2.2181,
+      "step": 727
+    },
+    {
+      "epoch": 0.07659126775381378,
+      "grad_norm": 5.1132049560546875,
+      "learning_rate": 0.0001988677940468907,
+      "loss": 2.1096,
+      "step": 728
+    },
+    {
+      "epoch": 0.0766964755391899,
+      "grad_norm": 1.210626482963562,
+      "learning_rate": 0.0001988626748918435,
+      "loss": 1.8661,
+      "step": 729
+    },
+    {
+      "epoch": 0.07680168332456602,
+      "grad_norm": 1.0915313959121704,
+      "learning_rate": 0.00019885754425619945,
+      "loss": 1.9508,
+      "step": 730
+    },
+    {
+      "epoch": 0.07690689110994213,
+      "grad_norm": 0.9805365204811096,
+      "learning_rate": 0.0001988524021405543,
+      "loss": 1.8635,
+      "step": 731
+    },
+    {
+      "epoch": 0.07701209889531825,
+      "grad_norm": 1.2141715288162231,
+      "learning_rate": 0.0001988472485455052,
+      "loss": 2.0866,
+      "step": 732
+    },
+    {
+      "epoch": 0.07711730668069437,
+      "grad_norm": 2.095106601715088,
+      "learning_rate": 0.00019884208347165062,
+      "loss": 1.658,
+      "step": 733
+    },
+    {
+      "epoch": 0.0772225144660705,
+      "grad_norm": 1.0189520120620728,
+      "learning_rate": 0.00019883690691959035,
+      "loss": 2.0882,
+      "step": 734
+    },
+    {
+      "epoch": 0.0773277222514466,
+      "grad_norm": 1.0242335796356201,
+      "learning_rate": 0.00019883171888992557,
+      "loss": 1.9676,
+      "step": 735
+    },
+    {
+      "epoch": 0.07743293003682272,
+      "grad_norm": 1.7329810857772827,
+      "learning_rate": 0.00019882651938325872,
+      "loss": 2.0722,
+      "step": 736
+    },
+    {
+      "epoch": 0.07753813782219884,
+      "grad_norm": 0.8744789958000183,
+      "learning_rate": 0.00019882130840019358,
+      "loss": 2.3554,
+      "step": 737
+    },
+    {
+      "epoch": 0.07764334560757497,
+      "grad_norm": 1.0732308626174927,
+      "learning_rate": 0.0001988160859413353,
+      "loss": 2.057,
+      "step": 738
+    },
+    {
+      "epoch": 0.07774855339295107,
+      "grad_norm": 0.9206953048706055,
+      "learning_rate": 0.0001988108520072904,
+      "loss": 2.1232,
+      "step": 739
+    },
+    {
+      "epoch": 0.0778537611783272,
+      "grad_norm": 0.7946454286575317,
+      "learning_rate": 0.0001988056065986666,
+      "loss": 2.1398,
+      "step": 740
+    },
+    {
+      "epoch": 0.07795896896370332,
+      "grad_norm": 1.073172926902771,
+      "learning_rate": 0.00019880034971607308,
+      "loss": 2.0701,
+      "step": 741
+    },
+    {
+      "epoch": 0.07806417674907944,
+      "grad_norm": 0.7679953575134277,
+      "learning_rate": 0.00019879508136012026,
+      "loss": 2.3293,
+      "step": 742
+    },
+    {
+      "epoch": 0.07816938453445554,
+      "grad_norm": 1.1941078901290894,
+      "learning_rate": 0.00019878980153141998,
+      "loss": 2.2848,
+      "step": 743
+    },
+    {
+      "epoch": 0.07827459231983167,
+      "grad_norm": 0.7265371680259705,
+      "learning_rate": 0.00019878451023058537,
+      "loss": 2.1909,
+      "step": 744
+    },
+    {
+      "epoch": 0.07837980010520779,
+      "grad_norm": 1.4622846841812134,
+      "learning_rate": 0.00019877920745823085,
+      "loss": 2.1973,
+      "step": 745
+    },
+    {
+      "epoch": 0.07848500789058391,
+      "grad_norm": 1.132213830947876,
+      "learning_rate": 0.00019877389321497227,
+      "loss": 1.757,
+      "step": 746
+    },
+    {
+      "epoch": 0.07859021567596002,
+      "grad_norm": 1.0680114030838013,
+      "learning_rate": 0.00019876856750142673,
+      "loss": 1.8053,
+      "step": 747
+    },
+    {
+      "epoch": 0.07869542346133614,
+      "grad_norm": 1.1053400039672852,
+      "learning_rate": 0.00019876323031821266,
+      "loss": 2.0033,
+      "step": 748
+    },
+    {
+      "epoch": 0.07880063124671226,
+      "grad_norm": 1.0279300212860107,
+      "learning_rate": 0.0001987578816659499,
+      "loss": 1.9289,
+      "step": 749
+    },
+    {
+      "epoch": 0.07890583903208838,
+      "grad_norm": 0.8577696084976196,
+      "learning_rate": 0.00019875252154525952,
+      "loss": 2.3191,
+      "step": 750
+    },
+    {
+      "epoch": 0.07901104681746449,
+      "grad_norm": 2.2503817081451416,
+      "learning_rate": 0.000198747149956764,
+      "loss": 1.8925,
+      "step": 751
+    },
+    {
+      "epoch": 0.07911625460284061,
+      "grad_norm": 0.7883946299552917,
+      "learning_rate": 0.0001987417669010871,
+      "loss": 2.1394,
+      "step": 752
+    },
+    {
+      "epoch": 0.07922146238821673,
+      "grad_norm": 0.9540475010871887,
+      "learning_rate": 0.00019873637237885402,
+      "loss": 2.1549,
+      "step": 753
+    },
+    {
+      "epoch": 0.07932667017359285,
+      "grad_norm": 1.1005622148513794,
+      "learning_rate": 0.0001987309663906911,
+      "loss": 1.8121,
+      "step": 754
+    },
+    {
+      "epoch": 0.07943187795896896,
+      "grad_norm": 1.2828789949417114,
+      "learning_rate": 0.00019872554893722618,
+      "loss": 1.9085,
+      "step": 755
+    },
+    {
+      "epoch": 0.07953708574434508,
+      "grad_norm": 0.9298080205917358,
+      "learning_rate": 0.00019872012001908833,
+      "loss": 2.301,
+      "step": 756
+    },
+    {
+      "epoch": 0.0796422935297212,
+      "grad_norm": 1.2716705799102783,
+      "learning_rate": 0.00019871467963690807,
+      "loss": 2.0119,
+      "step": 757
+    },
+    {
+      "epoch": 0.07974750131509732,
+      "grad_norm": 1.2654720544815063,
+      "learning_rate": 0.0001987092277913171,
+      "loss": 1.7698,
+      "step": 758
+    },
+    {
+      "epoch": 0.07985270910047343,
+      "grad_norm": 1.111207127571106,
+      "learning_rate": 0.00019870376448294851,
+      "loss": 1.6179,
+      "step": 759
+    },
+    {
+      "epoch": 0.07995791688584955,
+      "grad_norm": 1.3444252014160156,
+      "learning_rate": 0.00019869828971243682,
+      "loss": 1.7896,
+      "step": 760
+    },
+    {
+      "epoch": 0.08006312467122567,
+      "grad_norm": 0.7708830237388611,
+      "learning_rate": 0.00019869280348041774,
+      "loss": 2.241,
+      "step": 761
+    },
+    {
+      "epoch": 0.08016833245660179,
+      "grad_norm": 1.6333880424499512,
+      "learning_rate": 0.0001986873057875284,
+      "loss": 1.9599,
+      "step": 762
+    },
+    {
+      "epoch": 0.0802735402419779,
+      "grad_norm": 0.9313520789146423,
+      "learning_rate": 0.00019868179663440718,
+      "loss": 2.3505,
+      "step": 763
+    },
+    {
+      "epoch": 0.08037874802735402,
+      "grad_norm": 1.2057533264160156,
+      "learning_rate": 0.00019867627602169387,
+      "loss": 2.1487,
+      "step": 764
+    },
+    {
+      "epoch": 0.08048395581273014,
+      "grad_norm": 1.0157853364944458,
+      "learning_rate": 0.00019867074395002958,
+      "loss": 2.0869,
+      "step": 765
+    },
+    {
+      "epoch": 0.08058916359810626,
+      "grad_norm": 0.8782801628112793,
+      "learning_rate": 0.00019866520042005669,
+      "loss": 2.1859,
+      "step": 766
+    },
+    {
+      "epoch": 0.08069437138348237,
+      "grad_norm": 0.8281430006027222,
+      "learning_rate": 0.00019865964543241897,
+      "loss": 2.1563,
+      "step": 767
+    },
+    {
+      "epoch": 0.08079957916885849,
+      "grad_norm": 0.9678866863250732,
+      "learning_rate": 0.00019865407898776152,
+      "loss": 2.1626,
+      "step": 768
+    },
+    {
+      "epoch": 0.08090478695423461,
+      "grad_norm": 0.9700407981872559,
+      "learning_rate": 0.00019864850108673073,
+      "loss": 2.2104,
+      "step": 769
+    },
+    {
+      "epoch": 0.08100999473961074,
+      "grad_norm": 1.1797422170639038,
+      "learning_rate": 0.00019864291172997435,
+      "loss": 1.8951,
+      "step": 770
+    },
+    {
+      "epoch": 0.08111520252498684,
+      "grad_norm": 1.1290271282196045,
+      "learning_rate": 0.00019863731091814146,
+      "loss": 2.2929,
+      "step": 771
+    },
+    {
+      "epoch": 0.08122041031036296,
+      "grad_norm": 1.6742466688156128,
+      "learning_rate": 0.00019863169865188244,
+      "loss": 1.3898,
+      "step": 772
+    },
+    {
+      "epoch": 0.08132561809573909,
+      "grad_norm": 1.3946906328201294,
+      "learning_rate": 0.00019862607493184906,
+      "loss": 1.6763,
+      "step": 773
+    },
+    {
+      "epoch": 0.0814308258811152,
+      "grad_norm": 1.0987610816955566,
+      "learning_rate": 0.00019862043975869438,
+      "loss": 1.9443,
+      "step": 774
+    },
+    {
+      "epoch": 0.08153603366649131,
+      "grad_norm": 1.0788226127624512,
+      "learning_rate": 0.00019861479313307273,
+      "loss": 1.8061,
+      "step": 775
+    },
+    {
+      "epoch": 0.08164124145186744,
+      "grad_norm": 0.9351698160171509,
+      "learning_rate": 0.0001986091350556399,
+      "loss": 2.159,
+      "step": 776
+    },
+    {
+      "epoch": 0.08174644923724356,
+      "grad_norm": 1.1127021312713623,
+      "learning_rate": 0.000198603465527053,
+      "loss": 2.0271,
+      "step": 777
+    },
+    {
+      "epoch": 0.08185165702261968,
+      "grad_norm": 0.7860623598098755,
+      "learning_rate": 0.00019859778454797027,
+      "loss": 1.8645,
+      "step": 778
+    },
+    {
+      "epoch": 0.08195686480799579,
+      "grad_norm": 1.112596869468689,
+      "learning_rate": 0.00019859209211905152,
+      "loss": 1.7839,
+      "step": 779
+    },
+    {
+      "epoch": 0.08206207259337191,
+      "grad_norm": 1.2952525615692139,
+      "learning_rate": 0.00019858638824095775,
+      "loss": 2.0274,
+      "step": 780
+    },
+    {
+      "epoch": 0.08216728037874803,
+      "grad_norm": 1.0121921300888062,
+      "learning_rate": 0.00019858067291435137,
+      "loss": 2.2605,
+      "step": 781
+    },
+    {
+      "epoch": 0.08227248816412415,
+      "grad_norm": 1.0447254180908203,
+      "learning_rate": 0.00019857494613989606,
+      "loss": 2.1212,
+      "step": 782
+    },
+    {
+      "epoch": 0.08237769594950026,
+      "grad_norm": 1.3232123851776123,
+      "learning_rate": 0.00019856920791825683,
+      "loss": 1.7452,
+      "step": 783
+    },
+    {
+      "epoch": 0.08248290373487638,
+      "grad_norm": 1.3245103359222412,
+      "learning_rate": 0.0001985634582501001,
+      "loss": 1.6841,
+      "step": 784
+    },
+    {
+      "epoch": 0.0825881115202525,
+      "grad_norm": 1.0271555185317993,
+      "learning_rate": 0.00019855769713609348,
+      "loss": 1.9158,
+      "step": 785
+    },
+    {
+      "epoch": 0.08269331930562862,
+      "grad_norm": 1.0767377614974976,
+      "learning_rate": 0.00019855192457690607,
+      "loss": 1.7579,
+      "step": 786
+    },
+    {
+      "epoch": 0.08279852709100473,
+      "grad_norm": 0.7473664283752441,
+      "learning_rate": 0.00019854614057320818,
+      "loss": 2.3149,
+      "step": 787
+    },
+    {
+      "epoch": 0.08290373487638085,
+      "grad_norm": 1.0371477603912354,
+      "learning_rate": 0.0001985403451256715,
+      "loss": 1.9446,
+      "step": 788
+    },
+    {
+      "epoch": 0.08300894266175697,
+      "grad_norm": 1.0556628704071045,
+      "learning_rate": 0.00019853453823496898,
+      "loss": 2.0536,
+      "step": 789
+    },
+    {
+      "epoch": 0.08311415044713309,
+      "grad_norm": 1.5194404125213623,
+      "learning_rate": 0.00019852871990177503,
+      "loss": 2.1898,
+      "step": 790
+    },
+    {
+      "epoch": 0.0832193582325092,
+      "grad_norm": 1.2382123470306396,
+      "learning_rate": 0.0001985228901267653,
+      "loss": 2.096,
+      "step": 791
+    },
+    {
+      "epoch": 0.08332456601788532,
+      "grad_norm": 0.805229902267456,
+      "learning_rate": 0.00019851704891061676,
+      "loss": 2.2309,
+      "step": 792
+    },
+    {
+      "epoch": 0.08342977380326144,
+      "grad_norm": 1.50439453125,
+      "learning_rate": 0.00019851119625400774,
+      "loss": 1.664,
+      "step": 793
+    },
+    {
+      "epoch": 0.08353498158863756,
+      "grad_norm": 1.2170758247375488,
+      "learning_rate": 0.0001985053321576179,
+      "loss": 2.2134,
+      "step": 794
+    },
+    {
+      "epoch": 0.08364018937401367,
+      "grad_norm": 0.898504376411438,
+      "learning_rate": 0.0001984994566221282,
+      "loss": 1.7387,
+      "step": 795
+    },
+    {
+      "epoch": 0.08374539715938979,
+      "grad_norm": 1.0071955919265747,
+      "learning_rate": 0.00019849356964822093,
+      "loss": 2.255,
+      "step": 796
+    },
+    {
+      "epoch": 0.08385060494476591,
+      "grad_norm": 1.0701338052749634,
+      "learning_rate": 0.00019848767123657976,
+      "loss": 2.1611,
+      "step": 797
+    },
+    {
+      "epoch": 0.08395581273014203,
+      "grad_norm": 0.9649009704589844,
+      "learning_rate": 0.00019848176138788964,
+      "loss": 2.2191,
+      "step": 798
+    },
+    {
+      "epoch": 0.08406102051551814,
+      "grad_norm": 2.6447019577026367,
+      "learning_rate": 0.00019847584010283686,
+      "loss": 1.9287,
+      "step": 799
+    },
+    {
+      "epoch": 0.08416622830089426,
+      "grad_norm": 0.8168442845344543,
+      "learning_rate": 0.00019846990738210907,
+      "loss": 1.7487,
+      "step": 800
+    },
+    {
+      "epoch": 0.08427143608627038,
+      "grad_norm": 1.1686725616455078,
+      "learning_rate": 0.00019846396322639514,
+      "loss": 1.7051,
+      "step": 801
+    },
+    {
+      "epoch": 0.0843766438716465,
+      "grad_norm": 0.8122353553771973,
+      "learning_rate": 0.00019845800763638544,
+      "loss": 1.7214,
+      "step": 802
+    },
+    {
+      "epoch": 0.08448185165702261,
+      "grad_norm": 1.4718796014785767,
+      "learning_rate": 0.0001984520406127715,
+      "loss": 1.8946,
+      "step": 803
+    },
+    {
+      "epoch": 0.08458705944239873,
+      "grad_norm": 1.2869932651519775,
+      "learning_rate": 0.0001984460621562463,
+      "loss": 1.6935,
+      "step": 804
+    },
+    {
+      "epoch": 0.08469226722777486,
+      "grad_norm": 0.8310649991035461,
+      "learning_rate": 0.00019844007226750408,
+      "loss": 1.7826,
+      "step": 805
+    },
+    {
+      "epoch": 0.08479747501315098,
+      "grad_norm": 1.2225552797317505,
+      "learning_rate": 0.0001984340709472404,
+      "loss": 1.9194,
+      "step": 806
+    },
+    {
+      "epoch": 0.08490268279852708,
+      "grad_norm": 0.819175660610199,
+      "learning_rate": 0.00019842805819615222,
+      "loss": 2.0277,
+      "step": 807
+    },
+    {
+      "epoch": 0.0850078905839032,
+      "grad_norm": 0.8296215534210205,
+      "learning_rate": 0.00019842203401493772,
+      "loss": 2.1235,
+      "step": 808
+    },
+    {
+      "epoch": 0.08511309836927933,
+      "grad_norm": 1.2839940786361694,
+      "learning_rate": 0.00019841599840429654,
+      "loss": 1.6553,
+      "step": 809
+    },
+    {
+      "epoch": 0.08521830615465545,
+      "grad_norm": 1.0420074462890625,
+      "learning_rate": 0.00019840995136492955,
+      "loss": 1.9328,
+      "step": 810
+    },
+    {
+      "epoch": 0.08532351394003156,
+      "grad_norm": 0.8936055898666382,
+      "learning_rate": 0.00019840389289753896,
+      "loss": 2.176,
+      "step": 811
+    },
+    {
+      "epoch": 0.08542872172540768,
+      "grad_norm": 0.811882495880127,
+      "learning_rate": 0.0001983978230028283,
+      "loss": 2.2401,
+      "step": 812
+    },
+    {
+      "epoch": 0.0855339295107838,
+      "grad_norm": 0.9509261846542358,
+      "learning_rate": 0.00019839174168150247,
+      "loss": 2.1676,
+      "step": 813
+    },
+    {
+      "epoch": 0.08563913729615992,
+      "grad_norm": 0.921415388584137,
+      "learning_rate": 0.0001983856489342677,
+      "loss": 2.3311,
+      "step": 814
+    },
+    {
+      "epoch": 0.08574434508153603,
+      "grad_norm": 0.9342227578163147,
+      "learning_rate": 0.00019837954476183148,
+      "loss": 2.2151,
+      "step": 815
+    },
+    {
+      "epoch": 0.08584955286691215,
+      "grad_norm": 0.9104680418968201,
+      "learning_rate": 0.00019837342916490268,
+      "loss": 2.3867,
+      "step": 816
+    },
+    {
+      "epoch": 0.08595476065228827,
+      "grad_norm": 1.001815915107727,
+      "learning_rate": 0.0001983673021441915,
+      "loss": 1.998,
+      "step": 817
+    },
+    {
+      "epoch": 0.08605996843766439,
+      "grad_norm": 1.704974889755249,
+      "learning_rate": 0.00019836116370040944,
+      "loss": 1.8599,
+      "step": 818
+    },
+    {
+      "epoch": 0.0861651762230405,
+      "grad_norm": 0.82003253698349,
+      "learning_rate": 0.0001983550138342693,
+      "loss": 2.3351,
+      "step": 819
+    },
+    {
+      "epoch": 0.08627038400841662,
+      "grad_norm": 0.7901598215103149,
+      "learning_rate": 0.00019834885254648533,
+      "loss": 2.0966,
+      "step": 820
+    },
+    {
+      "epoch": 0.08637559179379274,
+      "grad_norm": 1.0715858936309814,
+      "learning_rate": 0.00019834267983777292,
+      "loss": 1.8696,
+      "step": 821
+    },
+    {
+      "epoch": 0.08648079957916886,
+      "grad_norm": 1.0539333820343018,
+      "learning_rate": 0.000198336495708849,
+      "loss": 1.7904,
+      "step": 822
+    },
+    {
+      "epoch": 0.08658600736454497,
+      "grad_norm": 1.2801223993301392,
+      "learning_rate": 0.00019833030016043156,
+      "loss": 2.1986,
+      "step": 823
+    },
+    {
+      "epoch": 0.08669121514992109,
+      "grad_norm": 1.5818783044815063,
+      "learning_rate": 0.00019832409319324023,
+      "loss": 1.8651,
+      "step": 824
+    },
+    {
+      "epoch": 0.08679642293529721,
+      "grad_norm": 0.8712964057922363,
+      "learning_rate": 0.00019831787480799568,
+      "loss": 2.2397,
+      "step": 825
+    },
+    {
+      "epoch": 0.08690163072067333,
+      "grad_norm": 1.518312931060791,
+      "learning_rate": 0.00019831164500542012,
+      "loss": 2.1451,
+      "step": 826
+    },
+    {
+      "epoch": 0.08700683850604944,
+      "grad_norm": 1.9714057445526123,
+      "learning_rate": 0.00019830540378623694,
+      "loss": 2.4544,
+      "step": 827
+    },
+    {
+      "epoch": 0.08711204629142556,
+      "grad_norm": 1.1616429090499878,
+      "learning_rate": 0.00019829915115117093,
+      "loss": 1.8353,
+      "step": 828
+    },
+    {
+      "epoch": 0.08721725407680168,
+      "grad_norm": 12.633820533752441,
+      "learning_rate": 0.0001982928871009482,
+      "loss": 2.3892,
+      "step": 829
+    },
+    {
+      "epoch": 0.0873224618621778,
+      "grad_norm": 0.8267893195152283,
+      "learning_rate": 0.00019828661163629615,
+      "loss": 2.1383,
+      "step": 830
+    },
+    {
+      "epoch": 0.08742766964755391,
+      "grad_norm": 0.9138221144676208,
+      "learning_rate": 0.00019828032475794352,
+      "loss": 2.3029,
+      "step": 831
+    },
+    {
+      "epoch": 0.08753287743293003,
+      "grad_norm": 1.1330816745758057,
+      "learning_rate": 0.00019827402646662047,
+      "loss": 2.2558,
+      "step": 832
+    },
+    {
+      "epoch": 0.08763808521830616,
+      "grad_norm": 0.8642826676368713,
+      "learning_rate": 0.0001982677167630583,
+      "loss": 1.9996,
+      "step": 833
+    },
+    {
+      "epoch": 0.08774329300368228,
+      "grad_norm": 0.8680064678192139,
+      "learning_rate": 0.00019826139564798974,
+      "loss": 2.3248,
+      "step": 834
+    },
+    {
+      "epoch": 0.08784850078905838,
+      "grad_norm": 1.49010169506073,
+      "learning_rate": 0.0001982550631221489,
+      "loss": 1.9746,
+      "step": 835
+    },
+    {
+      "epoch": 0.0879537085744345,
+      "grad_norm": 2.9857089519500732,
+      "learning_rate": 0.00019824871918627115,
+      "loss": 2.3267,
+      "step": 836
+    },
+    {
+      "epoch": 0.08805891635981063,
+      "grad_norm": 1.274704933166504,
+      "learning_rate": 0.0001982423638410931,
+      "loss": 2.3621,
+      "step": 837
+    },
+    {
+      "epoch": 0.08816412414518675,
+      "grad_norm": 0.9977430105209351,
+      "learning_rate": 0.0001982359970873529,
+      "loss": 2.0158,
+      "step": 838
+    },
+    {
+      "epoch": 0.08826933193056286,
+      "grad_norm": 0.9175136685371399,
+      "learning_rate": 0.0001982296189257898,
+      "loss": 1.8731,
+      "step": 839
+    },
+    {
+      "epoch": 0.08837453971593898,
+      "grad_norm": 0.993022620677948,
+      "learning_rate": 0.00019822322935714458,
+      "loss": 1.9218,
+      "step": 840
+    },
+    {
+      "epoch": 0.0884797475013151,
+      "grad_norm": 1.6498595476150513,
+      "learning_rate": 0.00019821682838215915,
+      "loss": 2.0277,
+      "step": 841
+    },
+    {
+      "epoch": 0.08858495528669122,
+      "grad_norm": 0.8532339334487915,
+      "learning_rate": 0.00019821041600157682,
+      "loss": 1.7503,
+      "step": 842
+    },
+    {
+      "epoch": 0.08869016307206733,
+      "grad_norm": 0.8794986605644226,
+      "learning_rate": 0.0001982039922161423,
+      "loss": 2.2771,
+      "step": 843
+    },
+    {
+      "epoch": 0.08879537085744345,
+      "grad_norm": 1.4324920177459717,
+      "learning_rate": 0.00019819755702660155,
+      "loss": 2.0883,
+      "step": 844
+    },
+    {
+      "epoch": 0.08890057864281957,
+      "grad_norm": 1.0365421772003174,
+      "learning_rate": 0.00019819111043370186,
+      "loss": 2.0581,
+      "step": 845
+    },
+    {
+      "epoch": 0.08900578642819569,
+      "grad_norm": 0.9392815232276917,
+      "learning_rate": 0.00019818465243819184,
+      "loss": 2.3084,
+      "step": 846
+    },
+    {
+      "epoch": 0.0891109942135718,
+      "grad_norm": 0.9906817674636841,
+      "learning_rate": 0.00019817818304082146,
+      "loss": 2.3203,
+      "step": 847
+    },
+    {
+      "epoch": 0.08921620199894792,
+      "grad_norm": 1.1406745910644531,
+      "learning_rate": 0.000198171702242342,
+      "loss": 1.9426,
+      "step": 848
+    },
+    {
+      "epoch": 0.08932140978432404,
+      "grad_norm": 0.843298077583313,
+      "learning_rate": 0.00019816521004350596,
+      "loss": 2.2718,
+      "step": 849
+    },
+    {
+      "epoch": 0.08942661756970016,
+      "grad_norm": 0.8017824292182922,
+      "learning_rate": 0.00019815870644506738,
+      "loss": 1.9611,
+      "step": 850
+    },
+    {
+      "epoch": 0.08953182535507627,
+      "grad_norm": 1.6382418870925903,
+      "learning_rate": 0.00019815219144778143,
+      "loss": 1.5162,
+      "step": 851
+    },
+    {
+      "epoch": 0.08963703314045239,
+      "grad_norm": 0.6821427941322327,
+      "learning_rate": 0.00019814566505240472,
+      "loss": 1.8852,
+      "step": 852
+    },
+    {
+      "epoch": 0.08974224092582851,
+      "grad_norm": 0.967110276222229,
+      "learning_rate": 0.00019813912725969509,
+      "loss": 2.2542,
+      "step": 853
+    },
+    {
+      "epoch": 0.08984744871120463,
+      "grad_norm": 1.212904691696167,
+      "learning_rate": 0.00019813257807041178,
+      "loss": 2.1593,
+      "step": 854
+    },
+    {
+      "epoch": 0.08995265649658074,
+      "grad_norm": 0.7041783928871155,
+      "learning_rate": 0.00019812601748531533,
+      "loss": 2.0618,
+      "step": 855
+    },
+    {
+      "epoch": 0.09005786428195686,
+      "grad_norm": 1.1121630668640137,
+      "learning_rate": 0.00019811944550516758,
+      "loss": 1.94,
+      "step": 856
+    },
+    {
+      "epoch": 0.09016307206733298,
+      "grad_norm": 1.1690653562545776,
+      "learning_rate": 0.00019811286213073173,
+      "loss": 2.095,
+      "step": 857
+    },
+    {
+      "epoch": 0.0902682798527091,
+      "grad_norm": 1.0316269397735596,
+      "learning_rate": 0.00019810626736277228,
+      "loss": 1.8736,
+      "step": 858
+    },
+    {
+      "epoch": 0.09037348763808521,
+      "grad_norm": 0.9707139134407043,
+      "learning_rate": 0.00019809966120205505,
+      "loss": 1.9554,
+      "step": 859
+    },
+    {
+      "epoch": 0.09047869542346133,
+      "grad_norm": 1.1598957777023315,
+      "learning_rate": 0.0001980930436493472,
+      "loss": 2.1387,
+      "step": 860
+    },
+    {
+      "epoch": 0.09058390320883745,
+      "grad_norm": 0.881175696849823,
+      "learning_rate": 0.0001980864147054172,
+      "loss": 1.9518,
+      "step": 861
+    },
+    {
+      "epoch": 0.09068911099421358,
+      "grad_norm": 1.648833990097046,
+      "learning_rate": 0.0001980797743710349,
+      "loss": 1.8736,
+      "step": 862
+    },
+    {
+      "epoch": 0.09079431877958968,
+      "grad_norm": 1.2492873668670654,
+      "learning_rate": 0.0001980731226469713,
+      "loss": 2.2163,
+      "step": 863
+    },
+    {
+      "epoch": 0.0908995265649658,
+      "grad_norm": 0.6980217099189758,
+      "learning_rate": 0.00019806645953399893,
+      "loss": 1.8288,
+      "step": 864
+    },
+    {
+      "epoch": 0.09100473435034193,
+      "grad_norm": 1.0036383867263794,
+      "learning_rate": 0.00019805978503289158,
+      "loss": 2.3498,
+      "step": 865
+    },
+    {
+      "epoch": 0.09110994213571805,
+      "grad_norm": 0.788852870464325,
+      "learning_rate": 0.00019805309914442426,
+      "loss": 1.8863,
+      "step": 866
+    },
+    {
+      "epoch": 0.09121514992109415,
+      "grad_norm": 1.2762038707733154,
+      "learning_rate": 0.00019804640186937343,
+      "loss": 2.1017,
+      "step": 867
+    },
+    {
+      "epoch": 0.09132035770647028,
+      "grad_norm": 0.878587543964386,
+      "learning_rate": 0.0001980396932085168,
+      "loss": 2.1149,
+      "step": 868
+    },
+    {
+      "epoch": 0.0914255654918464,
+      "grad_norm": 1.13942551612854,
+      "learning_rate": 0.00019803297316263346,
+      "loss": 1.8322,
+      "step": 869
+    },
+    {
+      "epoch": 0.09153077327722252,
+      "grad_norm": 0.9839184284210205,
+      "learning_rate": 0.00019802624173250374,
+      "loss": 2.1218,
+      "step": 870
+    },
+    {
+      "epoch": 0.09163598106259863,
+      "grad_norm": 1.22666335105896,
+      "learning_rate": 0.00019801949891890938,
+      "loss": 1.9397,
+      "step": 871
+    },
+    {
+      "epoch": 0.09174118884797475,
+      "grad_norm": 1.162661075592041,
+      "learning_rate": 0.00019801274472263335,
+      "loss": 1.6768,
+      "step": 872
+    },
+    {
+      "epoch": 0.09184639663335087,
+      "grad_norm": 1.2919373512268066,
+      "learning_rate": 0.00019800597914446005,
+      "loss": 2.1166,
+      "step": 873
+    },
+    {
+      "epoch": 0.09195160441872699,
+      "grad_norm": 1.0327990055084229,
+      "learning_rate": 0.0001979992021851751,
+      "loss": 2.3549,
+      "step": 874
+    },
+    {
+      "epoch": 0.0920568122041031,
+      "grad_norm": 0.6984542012214661,
+      "learning_rate": 0.0001979924138455655,
+      "loss": 1.9739,
+      "step": 875
+    },
+    {
+      "epoch": 0.09216201998947922,
+      "grad_norm": 1.3677195310592651,
+      "learning_rate": 0.00019798561412641958,
+      "loss": 1.7295,
+      "step": 876
+    },
+    {
+      "epoch": 0.09226722777485534,
+      "grad_norm": 1.9641640186309814,
+      "learning_rate": 0.00019797880302852697,
+      "loss": 2.3405,
+      "step": 877
+    },
+    {
+      "epoch": 0.09237243556023146,
+      "grad_norm": 0.9281172752380371,
+      "learning_rate": 0.00019797198055267857,
+      "loss": 1.2817,
+      "step": 878
+    },
+    {
+      "epoch": 0.09247764334560757,
+      "grad_norm": 0.967189371585846,
+      "learning_rate": 0.0001979651466996667,
+      "loss": 2.2958,
+      "step": 879
+    },
+    {
+      "epoch": 0.09258285113098369,
+      "grad_norm": 1.9948618412017822,
+      "learning_rate": 0.0001979583014702849,
+      "loss": 1.6858,
+      "step": 880
+    },
+    {
+      "epoch": 0.09268805891635981,
+      "grad_norm": 0.7749442458152771,
+      "learning_rate": 0.00019795144486532814,
+      "loss": 2.4653,
+      "step": 881
+    },
+    {
+      "epoch": 0.09279326670173593,
+      "grad_norm": 1.0032241344451904,
+      "learning_rate": 0.00019794457688559265,
+      "loss": 2.2595,
+      "step": 882
+    },
+    {
+      "epoch": 0.09289847448711204,
+      "grad_norm": 0.8965802788734436,
+      "learning_rate": 0.00019793769753187595,
+      "loss": 1.9857,
+      "step": 883
+    },
+    {
+      "epoch": 0.09300368227248816,
+      "grad_norm": 0.8196418285369873,
+      "learning_rate": 0.00019793080680497696,
+      "loss": 2.0913,
+      "step": 884
+    },
+    {
+      "epoch": 0.09310889005786428,
+      "grad_norm": 0.9274260401725769,
+      "learning_rate": 0.00019792390470569583,
+      "loss": 2.0106,
+      "step": 885
+    },
+    {
+      "epoch": 0.0932140978432404,
+      "grad_norm": 0.8402416110038757,
+      "learning_rate": 0.00019791699123483412,
+      "loss": 2.5971,
+      "step": 886
+    },
+    {
+      "epoch": 0.09331930562861651,
+      "grad_norm": 1.1058905124664307,
+      "learning_rate": 0.00019791006639319463,
+      "loss": 2.134,
+      "step": 887
+    },
+    {
+      "epoch": 0.09342451341399263,
+      "grad_norm": 0.6589333415031433,
+      "learning_rate": 0.00019790313018158156,
+      "loss": 2.375,
+      "step": 888
+    },
+    {
+      "epoch": 0.09352972119936875,
+      "grad_norm": 1.1351195573806763,
+      "learning_rate": 0.00019789618260080034,
+      "loss": 1.8826,
+      "step": 889
+    },
+    {
+      "epoch": 0.09363492898474488,
+      "grad_norm": 1.2128400802612305,
+      "learning_rate": 0.00019788922365165785,
+      "loss": 2.023,
+      "step": 890
+    },
+    {
+      "epoch": 0.09374013677012098,
+      "grad_norm": 0.9158788919448853,
+      "learning_rate": 0.0001978822533349621,
+      "loss": 2.2709,
+      "step": 891
+    },
+    {
+      "epoch": 0.0938453445554971,
+      "grad_norm": 1.525220274925232,
+      "learning_rate": 0.00019787527165152265,
+      "loss": 2.0153,
+      "step": 892
+    },
+    {
+      "epoch": 0.09395055234087323,
+      "grad_norm": 0.9528924226760864,
+      "learning_rate": 0.00019786827860215014,
+      "loss": 2.1791,
+      "step": 893
+    },
+    {
+      "epoch": 0.09405576012624935,
+      "grad_norm": 1.2895303964614868,
+      "learning_rate": 0.00019786127418765673,
+      "loss": 1.9553,
+      "step": 894
+    },
+    {
+      "epoch": 0.09416096791162545,
+      "grad_norm": 0.9034793972969055,
+      "learning_rate": 0.0001978542584088558,
+      "loss": 2.3079,
+      "step": 895
+    },
+    {
+      "epoch": 0.09426617569700158,
+      "grad_norm": 1.507163643836975,
+      "learning_rate": 0.0001978472312665621,
+      "loss": 1.5934,
+      "step": 896
+    },
+    {
+      "epoch": 0.0943713834823777,
+      "grad_norm": 0.9403305649757385,
+      "learning_rate": 0.0001978401927615916,
+      "loss": 1.7338,
+      "step": 897
+    },
+    {
+      "epoch": 0.09447659126775382,
+      "grad_norm": 0.8427889347076416,
+      "learning_rate": 0.00019783314289476168,
+      "loss": 2.0062,
+      "step": 898
+    },
+    {
+      "epoch": 0.09458179905312993,
+      "grad_norm": 0.8502089977264404,
+      "learning_rate": 0.0001978260816668911,
+      "loss": 2.1841,
+      "step": 899
+    },
+    {
+      "epoch": 0.09468700683850605,
+      "grad_norm": 0.8778148889541626,
+      "learning_rate": 0.00019781900907879974,
+      "loss": 2.0756,
+      "step": 900
+    },
+    {
+      "epoch": 0.09479221462388217,
+      "grad_norm": 0.8359899520874023,
+      "learning_rate": 0.00019781192513130896,
+      "loss": 1.9334,
+      "step": 901
+    },
+    {
+      "epoch": 0.09489742240925829,
+      "grad_norm": 1.1402784585952759,
+      "learning_rate": 0.00019780482982524142,
+      "loss": 1.7855,
+      "step": 902
+    },
+    {
+      "epoch": 0.0950026301946344,
+      "grad_norm": 0.7871297597885132,
+      "learning_rate": 0.00019779772316142104,
+      "loss": 2.3578,
+      "step": 903
+    },
+    {
+      "epoch": 0.09510783798001052,
+      "grad_norm": 0.9127509593963623,
+      "learning_rate": 0.0001977906051406731,
+      "loss": 2.2987,
+      "step": 904
+    },
+    {
+      "epoch": 0.09521304576538664,
+      "grad_norm": 1.1920058727264404,
+      "learning_rate": 0.00019778347576382424,
+      "loss": 2.1563,
+      "step": 905
+    },
+    {
+      "epoch": 0.09531825355076276,
+      "grad_norm": 1.1233481168746948,
+      "learning_rate": 0.0001977763350317023,
+      "loss": 2.1715,
+      "step": 906
+    },
+    {
+      "epoch": 0.09542346133613887,
+      "grad_norm": 0.7940162420272827,
+      "learning_rate": 0.00019776918294513656,
+      "loss": 2.3458,
+      "step": 907
+    },
+    {
+      "epoch": 0.09552866912151499,
+      "grad_norm": 1.0063912868499756,
+      "learning_rate": 0.00019776201950495755,
+      "loss": 2.1605,
+      "step": 908
+    },
+    {
+      "epoch": 0.09563387690689111,
+      "grad_norm": 1.514668583869934,
+      "learning_rate": 0.00019775484471199715,
+      "loss": 1.5532,
+      "step": 909
+    },
+    {
+      "epoch": 0.09573908469226723,
+      "grad_norm": 0.8086494207382202,
+      "learning_rate": 0.0001977476585670885,
+      "loss": 1.8652,
+      "step": 910
+    },
+    {
+      "epoch": 0.09584429247764334,
+      "grad_norm": 0.925151526927948,
+      "learning_rate": 0.00019774046107106616,
+      "loss": 2.1997,
+      "step": 911
+    },
+    {
+      "epoch": 0.09594950026301946,
+      "grad_norm": 0.7572203278541565,
+      "learning_rate": 0.0001977332522247659,
+      "loss": 2.116,
+      "step": 912
+    },
+    {
+      "epoch": 0.09605470804839558,
+      "grad_norm": 1.877502202987671,
+      "learning_rate": 0.00019772603202902492,
+      "loss": 1.468,
+      "step": 913
+    },
+    {
+      "epoch": 0.0961599158337717,
+      "grad_norm": 1.2712554931640625,
+      "learning_rate": 0.00019771880048468163,
+      "loss": 2.0208,
+      "step": 914
+    },
+    {
+      "epoch": 0.09626512361914781,
+      "grad_norm": 0.7878434062004089,
+      "learning_rate": 0.00019771155759257584,
+      "loss": 2.0554,
+      "step": 915
+    },
+    {
+      "epoch": 0.09637033140452393,
+      "grad_norm": 0.7482603788375854,
+      "learning_rate": 0.0001977043033535486,
+      "loss": 1.954,
+      "step": 916
+    },
+    {
+      "epoch": 0.09647553918990005,
+      "grad_norm": 0.8648151755332947,
+      "learning_rate": 0.00019769703776844236,
+      "loss": 1.9668,
+      "step": 917
+    },
+    {
+      "epoch": 0.09658074697527617,
+      "grad_norm": 1.0494996309280396,
+      "learning_rate": 0.0001976897608381008,
+      "loss": 2.1018,
+      "step": 918
+    },
+    {
+      "epoch": 0.09668595476065228,
+      "grad_norm": 1.2748123407363892,
+      "learning_rate": 0.00019768247256336902,
+      "loss": 1.8108,
+      "step": 919
+    },
+    {
+      "epoch": 0.0967911625460284,
+      "grad_norm": 0.7436427474021912,
+      "learning_rate": 0.00019767517294509338,
+      "loss": 2.0819,
+      "step": 920
+    },
+    {
+      "epoch": 0.09689637033140452,
+      "grad_norm": 0.9573638439178467,
+      "learning_rate": 0.00019766786198412154,
+      "loss": 1.9145,
+      "step": 921
+    },
+    {
+      "epoch": 0.09700157811678065,
+      "grad_norm": 1.3364827632904053,
+      "learning_rate": 0.00019766053968130247,
+      "loss": 2.1283,
+      "step": 922
+    },
+    {
+      "epoch": 0.09710678590215675,
+      "grad_norm": 0.971378743648529,
+      "learning_rate": 0.00019765320603748655,
+      "loss": 2.0985,
+      "step": 923
+    },
+    {
+      "epoch": 0.09721199368753287,
+      "grad_norm": 0.8163911700248718,
+      "learning_rate": 0.00019764586105352534,
+      "loss": 2.3535,
+      "step": 924
+    },
+    {
+      "epoch": 0.097317201472909,
+      "grad_norm": 0.8818692564964294,
+      "learning_rate": 0.00019763850473027183,
+      "loss": 2.0812,
+      "step": 925
+    },
+    {
+      "epoch": 0.09742240925828512,
+      "grad_norm": 0.984257698059082,
+      "learning_rate": 0.00019763113706858031,
+      "loss": 1.9479,
+      "step": 926
+    },
+    {
+      "epoch": 0.09752761704366122,
+      "grad_norm": 0.9912437200546265,
+      "learning_rate": 0.00019762375806930632,
+      "loss": 1.6832,
+      "step": 927
+    },
+    {
+      "epoch": 0.09763282482903735,
+      "grad_norm": 0.8805968761444092,
+      "learning_rate": 0.0001976163677333068,
+      "loss": 2.3275,
+      "step": 928
+    },
+    {
+      "epoch": 0.09773803261441347,
+      "grad_norm": 1.2597976922988892,
+      "learning_rate": 0.00019760896606143988,
+      "loss": 1.8149,
+      "step": 929
+    },
+    {
+      "epoch": 0.09784324039978959,
+      "grad_norm": 0.709547758102417,
+      "learning_rate": 0.0001976015530545652,
+      "loss": 2.4011,
+      "step": 930
+    },
+    {
+      "epoch": 0.0979484481851657,
+      "grad_norm": 1.3702518939971924,
+      "learning_rate": 0.00019759412871354353,
+      "loss": 1.4587,
+      "step": 931
+    },
+    {
+      "epoch": 0.09805365597054182,
+      "grad_norm": 1.471561312675476,
+      "learning_rate": 0.00019758669303923706,
+      "loss": 1.9036,
+      "step": 932
+    },
+    {
+      "epoch": 0.09815886375591794,
+      "grad_norm": 0.7236111164093018,
+      "learning_rate": 0.0001975792460325093,
+      "loss": 2.8316,
+      "step": 933
+    },
+    {
+      "epoch": 0.09826407154129406,
+      "grad_norm": 0.9491783380508423,
+      "learning_rate": 0.000197571787694225,
+      "loss": 2.0152,
+      "step": 934
+    },
+    {
+      "epoch": 0.09836927932667017,
+      "grad_norm": 1.1442753076553345,
+      "learning_rate": 0.0001975643180252503,
+      "loss": 1.9371,
+      "step": 935
+    },
+    {
+      "epoch": 0.09847448711204629,
+      "grad_norm": 1.107806921005249,
+      "learning_rate": 0.00019755683702645262,
+      "loss": 2.0472,
+      "step": 936
+    },
+    {
+      "epoch": 0.09857969489742241,
+      "grad_norm": 1.2097374200820923,
+      "learning_rate": 0.0001975493446987007,
+      "loss": 2.0863,
+      "step": 937
+    },
+    {
+      "epoch": 0.09868490268279853,
+      "grad_norm": 0.8795415759086609,
+      "learning_rate": 0.0001975418410428646,
+      "loss": 2.0176,
+      "step": 938
+    },
+    {
+      "epoch": 0.09879011046817464,
+      "grad_norm": 0.8208872079849243,
+      "learning_rate": 0.0001975343260598157,
+      "loss": 2.0678,
+      "step": 939
+    },
+    {
+      "epoch": 0.09889531825355076,
+      "grad_norm": 1.147039532661438,
+      "learning_rate": 0.0001975267997504267,
+      "loss": 1.71,
+      "step": 940
+    },
+    {
+      "epoch": 0.09900052603892688,
+      "grad_norm": 1.0926604270935059,
+      "learning_rate": 0.00019751926211557157,
+      "loss": 2.2779,
+      "step": 941
+    },
+    {
+      "epoch": 0.099105733824303,
+      "grad_norm": 1.0695338249206543,
+      "learning_rate": 0.00019751171315612567,
+      "loss": 2.0222,
+      "step": 942
+    },
+    {
+      "epoch": 0.09921094160967911,
+      "grad_norm": 1.0999269485473633,
+      "learning_rate": 0.00019750415287296563,
+      "loss": 2.1906,
+      "step": 943
+    },
+    {
+      "epoch": 0.09931614939505523,
+      "grad_norm": 1.2324326038360596,
+      "learning_rate": 0.00019749658126696934,
+      "loss": 2.123,
+      "step": 944
+    },
+    {
+      "epoch": 0.09942135718043135,
+      "grad_norm": 0.9449137449264526,
+      "learning_rate": 0.00019748899833901614,
+      "loss": 1.8318,
+      "step": 945
+    },
+    {
+      "epoch": 0.09952656496580747,
+      "grad_norm": 1.1237130165100098,
+      "learning_rate": 0.0001974814040899866,
+      "loss": 1.9734,
+      "step": 946
+    },
+    {
+      "epoch": 0.09963177275118358,
+      "grad_norm": 0.7438077926635742,
+      "learning_rate": 0.00019747379852076263,
+      "loss": 2.406,
+      "step": 947
+    },
+    {
+      "epoch": 0.0997369805365597,
+      "grad_norm": 0.7513858675956726,
+      "learning_rate": 0.00019746618163222736,
+      "loss": 1.8843,
+      "step": 948
+    },
+    {
+      "epoch": 0.09984218832193582,
+      "grad_norm": 0.7331326007843018,
+      "learning_rate": 0.0001974585534252654,
+      "loss": 2.3763,
+      "step": 949
+    },
+    {
+      "epoch": 0.09994739610731194,
+      "grad_norm": 1.0843267440795898,
+      "learning_rate": 0.00019745091390076252,
+      "loss": 1.962,
+      "step": 950
+    },
+    {
+      "epoch": 0.10005260389268805,
+      "grad_norm": 1.1343190670013428,
+      "learning_rate": 0.00019744326305960595,
+      "loss": 2.0071,
+      "step": 951
+    },
+    {
+      "epoch": 0.10015781167806417,
+      "grad_norm": 1.4789917469024658,
+      "learning_rate": 0.0001974356009026841,
+      "loss": 1.715,
+      "step": 952
+    },
+    {
+      "epoch": 0.1002630194634403,
+      "grad_norm": 1.1908003091812134,
+      "learning_rate": 0.00019742792743088675,
+      "loss": 1.8924,
+      "step": 953
+    },
+    {
+      "epoch": 0.10036822724881642,
+      "grad_norm": 0.7355921864509583,
+      "learning_rate": 0.000197420242645105,
+      "loss": 2.3287,
+      "step": 954
+    },
+    {
+      "epoch": 0.10047343503419252,
+      "grad_norm": 0.7553874254226685,
+      "learning_rate": 0.0001974125465462313,
+      "loss": 1.7495,
+      "step": 955
+    },
+    {
+      "epoch": 0.10057864281956864,
+      "grad_norm": 1.1832842826843262,
+      "learning_rate": 0.00019740483913515932,
+      "loss": 2.0124,
+      "step": 956
+    },
+    {
+      "epoch": 0.10068385060494477,
+      "grad_norm": 1.039902925491333,
+      "learning_rate": 0.0001973971204127841,
+      "loss": 2.0525,
+      "step": 957
+    },
+    {
+      "epoch": 0.10078905839032089,
+      "grad_norm": 0.8891255855560303,
+      "learning_rate": 0.00019738939038000205,
+      "loss": 2.1702,
+      "step": 958
+    },
+    {
+      "epoch": 0.100894266175697,
+      "grad_norm": 1.692345142364502,
+      "learning_rate": 0.00019738164903771078,
+      "loss": 1.9213,
+      "step": 959
+    },
+    {
+      "epoch": 0.10099947396107312,
+      "grad_norm": 0.9816641807556152,
+      "learning_rate": 0.00019737389638680924,
+      "loss": 1.9241,
+      "step": 960
+    },
+    {
+      "epoch": 0.10110468174644924,
+      "grad_norm": 0.90645432472229,
+      "learning_rate": 0.0001973661324281978,
+      "loss": 1.6803,
+      "step": 961
+    },
+    {
+      "epoch": 0.10120988953182536,
+      "grad_norm": 1.1190398931503296,
+      "learning_rate": 0.00019735835716277802,
+      "loss": 2.0615,
+      "step": 962
+    },
+    {
+      "epoch": 0.10131509731720147,
+      "grad_norm": 0.9607227444648743,
+      "learning_rate": 0.0001973505705914528,
+      "loss": 1.9322,
+      "step": 963
+    },
+    {
+      "epoch": 0.10142030510257759,
+      "grad_norm": 0.7287774682044983,
+      "learning_rate": 0.00019734277271512638,
+      "loss": 2.3326,
+      "step": 964
+    },
+    {
+      "epoch": 0.10152551288795371,
+      "grad_norm": 1.8522443771362305,
+      "learning_rate": 0.00019733496353470433,
+      "loss": 2.1133,
+      "step": 965
+    },
+    {
+      "epoch": 0.10163072067332983,
+      "grad_norm": 0.899685800075531,
+      "learning_rate": 0.00019732714305109345,
+      "loss": 2.0211,
+      "step": 966
+    },
+    {
+      "epoch": 0.10173592845870594,
+      "grad_norm": 1.8534700870513916,
+      "learning_rate": 0.00019731931126520195,
+      "loss": 1.6331,
+      "step": 967
+    },
+    {
+      "epoch": 0.10184113624408206,
+      "grad_norm": 0.8898396492004395,
+      "learning_rate": 0.00019731146817793932,
+      "loss": 2.2339,
+      "step": 968
+    },
+    {
+      "epoch": 0.10194634402945818,
+      "grad_norm": 1.305379867553711,
+      "learning_rate": 0.00019730361379021632,
+      "loss": 1.9012,
+      "step": 969
+    },
+    {
+      "epoch": 0.1020515518148343,
+      "grad_norm": 1.1972163915634155,
+      "learning_rate": 0.00019729574810294507,
+      "loss": 1.6715,
+      "step": 970
+    },
+    {
+      "epoch": 0.10215675960021041,
+      "grad_norm": 1.1665350198745728,
+      "learning_rate": 0.00019728787111703895,
+      "loss": 2.3843,
+      "step": 971
+    },
+    {
+      "epoch": 0.10226196738558653,
+      "grad_norm": 0.8711062073707581,
+      "learning_rate": 0.00019727998283341274,
+      "loss": 2.2372,
+      "step": 972
+    },
+    {
+      "epoch": 0.10236717517096265,
+      "grad_norm": 0.9705300331115723,
+      "learning_rate": 0.00019727208325298246,
+      "loss": 2.2943,
+      "step": 973
+    },
+    {
+      "epoch": 0.10247238295633877,
+      "grad_norm": 1.7931549549102783,
+      "learning_rate": 0.00019726417237666546,
+      "loss": 2.1751,
+      "step": 974
+    },
+    {
+      "epoch": 0.10257759074171488,
+      "grad_norm": 1.630979061126709,
+      "learning_rate": 0.00019725625020538038,
+      "loss": 1.9252,
+      "step": 975
+    },
+    {
+      "epoch": 0.102682798527091,
+      "grad_norm": 1.412156105041504,
+      "learning_rate": 0.00019724831674004724,
+      "loss": 1.9631,
+      "step": 976
+    },
+    {
+      "epoch": 0.10278800631246712,
+      "grad_norm": 0.7268504500389099,
+      "learning_rate": 0.00019724037198158733,
+      "loss": 1.8943,
+      "step": 977
+    },
+    {
+      "epoch": 0.10289321409784324,
+      "grad_norm": 1.439155101776123,
+      "learning_rate": 0.00019723241593092318,
+      "loss": 1.9662,
+      "step": 978
+    },
+    {
+      "epoch": 0.10299842188321935,
+      "grad_norm": 1.4355570077896118,
+      "learning_rate": 0.00019722444858897878,
+      "loss": 2.2064,
+      "step": 979
+    },
+    {
+      "epoch": 0.10310362966859547,
+      "grad_norm": 0.9133340716362,
+      "learning_rate": 0.00019721646995667932,
+      "loss": 2.6105,
+      "step": 980
+    },
+    {
+      "epoch": 0.1032088374539716,
+      "grad_norm": 0.9703789949417114,
+      "learning_rate": 0.0001972084800349513,
+      "loss": 2.2393,
+      "step": 981
+    },
+    {
+      "epoch": 0.10331404523934772,
+      "grad_norm": 1.9817613363265991,
+      "learning_rate": 0.00019720047882472262,
+      "loss": 1.6778,
+      "step": 982
+    },
+    {
+      "epoch": 0.10341925302472382,
+      "grad_norm": 0.9752820134162903,
+      "learning_rate": 0.00019719246632692242,
+      "loss": 2.002,
+      "step": 983
+    },
+    {
+      "epoch": 0.10352446081009994,
+      "grad_norm": 1.4282121658325195,
+      "learning_rate": 0.00019718444254248114,
+      "loss": 1.8544,
+      "step": 984
+    },
+    {
+      "epoch": 0.10362966859547607,
+      "grad_norm": 0.9460940957069397,
+      "learning_rate": 0.00019717640747233056,
+      "loss": 2.1763,
+      "step": 985
+    },
+    {
+      "epoch": 0.10373487638085219,
+      "grad_norm": 1.4641571044921875,
+      "learning_rate": 0.00019716836111740378,
+      "loss": 2.2195,
+      "step": 986
+    },
+    {
+      "epoch": 0.1038400841662283,
+      "grad_norm": 0.9056519865989685,
+      "learning_rate": 0.00019716030347863517,
+      "loss": 2.0533,
+      "step": 987
+    },
+    {
+      "epoch": 0.10394529195160442,
+      "grad_norm": 1.7422082424163818,
+      "learning_rate": 0.00019715223455696047,
+      "loss": 2.1961,
+      "step": 988
+    },
+    {
+      "epoch": 0.10405049973698054,
+      "grad_norm": 0.7487703561782837,
+      "learning_rate": 0.0001971441543533167,
+      "loss": 2.3884,
+      "step": 989
+    },
+    {
+      "epoch": 0.10415570752235666,
+      "grad_norm": 1.123172402381897,
+      "learning_rate": 0.0001971360628686422,
+      "loss": 2.3124,
+      "step": 990
+    },
+    {
+      "epoch": 0.10426091530773277,
+      "grad_norm": 0.9539334177970886,
+      "learning_rate": 0.00019712796010387654,
+      "loss": 2.1406,
+      "step": 991
+    },
+    {
+      "epoch": 0.10436612309310889,
+      "grad_norm": 1.5190744400024414,
+      "learning_rate": 0.0001971198460599607,
+      "loss": 2.081,
+      "step": 992
+    },
+    {
+      "epoch": 0.10447133087848501,
+      "grad_norm": 0.7697191834449768,
+      "learning_rate": 0.00019711172073783696,
+      "loss": 2.2589,
+      "step": 993
+    },
+    {
+      "epoch": 0.10457653866386113,
+      "grad_norm": 1.2157846689224243,
+      "learning_rate": 0.0001971035841384489,
+      "loss": 1.9017,
+      "step": 994
+    },
+    {
+      "epoch": 0.10468174644923724,
+      "grad_norm": 0.8707440495491028,
+      "learning_rate": 0.00019709543626274131,
+      "loss": 2.2788,
+      "step": 995
+    },
+    {
+      "epoch": 0.10478695423461336,
+      "grad_norm": 1.0644099712371826,
+      "learning_rate": 0.00019708727711166047,
+      "loss": 1.8108,
+      "step": 996
+    },
+    {
+      "epoch": 0.10489216201998948,
+      "grad_norm": 1.3928571939468384,
+      "learning_rate": 0.00019707910668615382,
+      "loss": 1.811,
+      "step": 997
+    },
+    {
+      "epoch": 0.1049973698053656,
+      "grad_norm": 1.6720000505447388,
+      "learning_rate": 0.00019707092498717023,
+      "loss": 1.6834,
+      "step": 998
+    },
+    {
+      "epoch": 0.10510257759074171,
+      "grad_norm": 1.0016800165176392,
+      "learning_rate": 0.00019706273201565972,
+      "loss": 2.008,
+      "step": 999
+    },
+    {
+      "epoch": 0.10520778537611783,
+      "grad_norm": 0.7917373180389404,
+      "learning_rate": 0.00019705452777257377,
+      "loss": 2.1877,
+      "step": 1000
+    },
+    {
+      "epoch": 0.10531299316149395,
+      "grad_norm": 1.1289230585098267,
+      "learning_rate": 0.00019704631225886515,
+      "loss": 2.135,
+      "step": 1001
+    },
+    {
+      "epoch": 0.10541820094687007,
+      "grad_norm": 1.262541651725769,
+      "learning_rate": 0.00019703808547548782,
+      "loss": 2.1767,
+      "step": 1002
+    },
+    {
+      "epoch": 0.10552340873224618,
+      "grad_norm": 0.9078851938247681,
+      "learning_rate": 0.00019702984742339715,
+      "loss": 2.0435,
+      "step": 1003
+    },
+    {
+      "epoch": 0.1056286165176223,
+      "grad_norm": 1.3077322244644165,
+      "learning_rate": 0.00019702159810354978,
+      "loss": 1.8528,
+      "step": 1004
+    },
+    {
+      "epoch": 0.10573382430299842,
+      "grad_norm": 1.0650526285171509,
+      "learning_rate": 0.00019701333751690378,
+      "loss": 2.2322,
+      "step": 1005
+    },
+    {
+      "epoch": 0.10583903208837454,
+      "grad_norm": 1.7082046270370483,
+      "learning_rate": 0.0001970050656644183,
+      "loss": 2.0592,
+      "step": 1006
+    },
+    {
+      "epoch": 0.10594423987375065,
+      "grad_norm": 0.9092219471931458,
+      "learning_rate": 0.000196996782547054,
+      "loss": 1.929,
+      "step": 1007
+    },
+    {
+      "epoch": 0.10604944765912677,
+      "grad_norm": 1.0924969911575317,
+      "learning_rate": 0.00019698848816577274,
+      "loss": 2.0751,
+      "step": 1008
+    },
+    {
+      "epoch": 0.1061546554445029,
+      "grad_norm": 0.8354095220565796,
+      "learning_rate": 0.0001969801825215377,
+      "loss": 2.3727,
+      "step": 1009
+    },
+    {
+      "epoch": 0.10625986322987901,
+      "grad_norm": 0.8786736726760864,
+      "learning_rate": 0.00019697186561531345,
+      "loss": 2.0541,
+      "step": 1010
+    },
+    {
+      "epoch": 0.10636507101525512,
+      "grad_norm": 1.0986034870147705,
+      "learning_rate": 0.00019696353744806574,
+      "loss": 2.3069,
+      "step": 1011
+    },
+    {
+      "epoch": 0.10647027880063124,
+      "grad_norm": 0.7767875790596008,
+      "learning_rate": 0.00019695519802076175,
+      "loss": 2.2888,
+      "step": 1012
+    },
+    {
+      "epoch": 0.10657548658600736,
+      "grad_norm": 1.035365104675293,
+      "learning_rate": 0.00019694684733436986,
+      "loss": 1.9233,
+      "step": 1013
+    },
+    {
+      "epoch": 0.10668069437138349,
+      "grad_norm": 1.2048399448394775,
+      "learning_rate": 0.00019693848538985983,
+      "loss": 2.3406,
+      "step": 1014
+    },
+    {
+      "epoch": 0.1067859021567596,
+      "grad_norm": 0.9446375370025635,
+      "learning_rate": 0.0001969301121882027,
+      "loss": 1.9142,
+      "step": 1015
+    },
+    {
+      "epoch": 0.10689110994213571,
+      "grad_norm": 1.7339941263198853,
+      "learning_rate": 0.0001969217277303708,
+      "loss": 1.6881,
+      "step": 1016
+    },
+    {
+      "epoch": 0.10699631772751184,
+      "grad_norm": 1.0482121706008911,
+      "learning_rate": 0.00019691333201733786,
+      "loss": 2.0258,
+      "step": 1017
+    },
+    {
+      "epoch": 0.10710152551288796,
+      "grad_norm": 1.113286018371582,
+      "learning_rate": 0.00019690492505007877,
+      "loss": 1.7817,
+      "step": 1018
+    },
+    {
+      "epoch": 0.10720673329826406,
+      "grad_norm": 1.4576612710952759,
+      "learning_rate": 0.00019689650682956986,
+      "loss": 2.085,
+      "step": 1019
+    },
+    {
+      "epoch": 0.10731194108364019,
+      "grad_norm": 1.79568612575531,
+      "learning_rate": 0.00019688807735678866,
+      "loss": 2.0266,
+      "step": 1020
+    },
+    {
+      "epoch": 0.10741714886901631,
+      "grad_norm": 1.656327486038208,
+      "learning_rate": 0.00019687963663271409,
+      "loss": 2.1882,
+      "step": 1021
+    },
+    {
+      "epoch": 0.10752235665439243,
+      "grad_norm": 1.273909091949463,
+      "learning_rate": 0.00019687118465832636,
+      "loss": 1.9803,
+      "step": 1022
+    },
+    {
+      "epoch": 0.10762756443976854,
+      "grad_norm": 1.4731061458587646,
+      "learning_rate": 0.00019686272143460692,
+      "loss": 1.8951,
+      "step": 1023
+    },
+    {
+      "epoch": 0.10773277222514466,
+      "grad_norm": 0.7379525899887085,
+      "learning_rate": 0.00019685424696253858,
+      "loss": 2.0547,
+      "step": 1024
+    },
+    {
+      "epoch": 0.10783798001052078,
+      "grad_norm": 0.7799714803695679,
+      "learning_rate": 0.0001968457612431055,
+      "loss": 1.9213,
+      "step": 1025
+    },
+    {
+      "epoch": 0.1079431877958969,
+      "grad_norm": 2.3484795093536377,
+      "learning_rate": 0.00019683726427729306,
+      "loss": 2.5578,
+      "step": 1026
+    },
+    {
+      "epoch": 0.10804839558127301,
+      "grad_norm": 1.0442508459091187,
+      "learning_rate": 0.000196828756066088,
+      "loss": 2.1757,
+      "step": 1027
+    },
+    {
+      "epoch": 0.10815360336664913,
+      "grad_norm": 1.0097154378890991,
+      "learning_rate": 0.00019682023661047836,
+      "loss": 1.8894,
+      "step": 1028
+    },
+    {
+      "epoch": 0.10825881115202525,
+      "grad_norm": 1.412798523902893,
+      "learning_rate": 0.00019681170591145345,
+      "loss": 2.2574,
+      "step": 1029
+    },
+    {
+      "epoch": 0.10836401893740137,
+      "grad_norm": 0.9488776922225952,
+      "learning_rate": 0.00019680316397000395,
+      "loss": 1.6457,
+      "step": 1030
+    },
+    {
+      "epoch": 0.10846922672277748,
+      "grad_norm": 0.8774062395095825,
+      "learning_rate": 0.00019679461078712178,
+      "loss": 2.0935,
+      "step": 1031
+    },
+    {
+      "epoch": 0.1085744345081536,
+      "grad_norm": 1.0559821128845215,
+      "learning_rate": 0.00019678604636380018,
+      "loss": 1.9539,
+      "step": 1032
+    },
+    {
+      "epoch": 0.10867964229352972,
+      "grad_norm": 1.1021692752838135,
+      "learning_rate": 0.0001967774707010337,
+      "loss": 2.1199,
+      "step": 1033
+    },
+    {
+      "epoch": 0.10878485007890584,
+      "grad_norm": 1.441303014755249,
+      "learning_rate": 0.0001967688837998183,
+      "loss": 1.7238,
+      "step": 1034
+    },
+    {
+      "epoch": 0.10889005786428195,
+      "grad_norm": 1.2004631757736206,
+      "learning_rate": 0.00019676028566115102,
+      "loss": 2.3156,
+      "step": 1035
+    },
+    {
+      "epoch": 0.10899526564965807,
+      "grad_norm": 1.1571768522262573,
+      "learning_rate": 0.0001967516762860304,
+      "loss": 2.2756,
+      "step": 1036
+    },
+    {
+      "epoch": 0.10910047343503419,
+      "grad_norm": 0.7429097890853882,
+      "learning_rate": 0.0001967430556754562,
+      "loss": 1.4574,
+      "step": 1037
+    },
+    {
+      "epoch": 0.10920568122041031,
+      "grad_norm": 1.2119436264038086,
+      "learning_rate": 0.00019673442383042952,
+      "loss": 2.3933,
+      "step": 1038
+    },
+    {
+      "epoch": 0.10931088900578642,
+      "grad_norm": 1.4365086555480957,
+      "learning_rate": 0.00019672578075195272,
+      "loss": 1.9312,
+      "step": 1039
+    },
+    {
+      "epoch": 0.10941609679116254,
+      "grad_norm": 1.0024199485778809,
+      "learning_rate": 0.00019671712644102956,
+      "loss": 1.9121,
+      "step": 1040
+    },
+    {
+      "epoch": 0.10952130457653866,
+      "grad_norm": 0.6678622961044312,
+      "learning_rate": 0.00019670846089866496,
+      "loss": 1.9272,
+      "step": 1041
+    },
+    {
+      "epoch": 0.10962651236191479,
+      "grad_norm": 1.1651445627212524,
+      "learning_rate": 0.00019669978412586528,
+      "loss": 2.1199,
+      "step": 1042
+    },
+    {
+      "epoch": 0.10973172014729089,
+      "grad_norm": 1.0801807641983032,
+      "learning_rate": 0.00019669109612363803,
+      "loss": 1.9533,
+      "step": 1043
+    },
+    {
+      "epoch": 0.10983692793266701,
+      "grad_norm": 1.6487115621566772,
+      "learning_rate": 0.00019668239689299224,
+      "loss": 1.4197,
+      "step": 1044
+    },
+    {
+      "epoch": 0.10994213571804314,
+      "grad_norm": 1.0398926734924316,
+      "learning_rate": 0.00019667368643493804,
+      "loss": 1.8756,
+      "step": 1045
+    },
+    {
+      "epoch": 0.11004734350341926,
+      "grad_norm": 1.2380412817001343,
+      "learning_rate": 0.00019666496475048698,
+      "loss": 1.731,
+      "step": 1046
+    },
+    {
+      "epoch": 0.11015255128879536,
+      "grad_norm": 1.0695997476577759,
+      "learning_rate": 0.00019665623184065187,
+      "loss": 1.816,
+      "step": 1047
+    },
+    {
+      "epoch": 0.11025775907417149,
+      "grad_norm": 1.0123080015182495,
+      "learning_rate": 0.00019664748770644686,
+      "loss": 2.0089,
+      "step": 1048
+    },
+    {
+      "epoch": 0.1103629668595476,
+      "grad_norm": 0.980810821056366,
+      "learning_rate": 0.00019663873234888733,
+      "loss": 2.0396,
+      "step": 1049
+    },
+    {
+      "epoch": 0.11046817464492373,
+      "grad_norm": 1.2700474262237549,
+      "learning_rate": 0.00019662996576899004,
+      "loss": 2.3745,
+      "step": 1050
+    },
+    {
+      "epoch": 0.11057338243029984,
+      "grad_norm": 1.1388194561004639,
+      "learning_rate": 0.00019662118796777303,
+      "loss": 1.8479,
+      "step": 1051
+    },
+    {
+      "epoch": 0.11067859021567596,
+      "grad_norm": 1.463036298751831,
+      "learning_rate": 0.0001966123989462556,
+      "loss": 1.9751,
+      "step": 1052
+    },
+    {
+      "epoch": 0.11078379800105208,
+      "grad_norm": 1.2596359252929688,
+      "learning_rate": 0.00019660359870545845,
+      "loss": 2.0322,
+      "step": 1053
+    },
+    {
+      "epoch": 0.1108890057864282,
+      "grad_norm": 0.8281905055046082,
+      "learning_rate": 0.00019659478724640348,
+      "loss": 1.9301,
+      "step": 1054
+    },
+    {
+      "epoch": 0.1109942135718043,
+      "grad_norm": 1.469774603843689,
+      "learning_rate": 0.00019658596457011393,
+      "loss": 1.5288,
+      "step": 1055
+    },
+    {
+      "epoch": 0.11109942135718043,
+      "grad_norm": 0.8355114459991455,
+      "learning_rate": 0.0001965771306776144,
+      "loss": 2.1748,
+      "step": 1056
+    },
+    {
+      "epoch": 0.11120462914255655,
+      "grad_norm": 0.8642022013664246,
+      "learning_rate": 0.00019656828556993068,
+      "loss": 1.6075,
+      "step": 1057
+    },
+    {
+      "epoch": 0.11130983692793267,
+      "grad_norm": 0.6303163170814514,
+      "learning_rate": 0.00019655942924808994,
+      "loss": 2.3366,
+      "step": 1058
+    },
+    {
+      "epoch": 0.11141504471330879,
+      "grad_norm": 0.9480810761451721,
+      "learning_rate": 0.00019655056171312069,
+      "loss": 1.8519,
+      "step": 1059
+    },
+    {
+      "epoch": 0.1115202524986849,
+      "grad_norm": 0.8360817432403564,
+      "learning_rate": 0.0001965416829660526,
+      "loss": 2.1247,
+      "step": 1060
+    },
+    {
+      "epoch": 0.11162546028406102,
+      "grad_norm": 0.9040228724479675,
+      "learning_rate": 0.0001965327930079168,
+      "loss": 2.2492,
+      "step": 1061
+    },
+    {
+      "epoch": 0.11173066806943714,
+      "grad_norm": 0.8252432942390442,
+      "learning_rate": 0.00019652389183974557,
+      "loss": 1.7289,
+      "step": 1062
+    },
+    {
+      "epoch": 0.11183587585481326,
+      "grad_norm": 0.783162534236908,
+      "learning_rate": 0.00019651497946257266,
+      "loss": 2.3029,
+      "step": 1063
+    },
+    {
+      "epoch": 0.11194108364018937,
+      "grad_norm": 0.8001965284347534,
+      "learning_rate": 0.00019650605587743302,
+      "loss": 1.6801,
+      "step": 1064
+    },
+    {
+      "epoch": 0.11204629142556549,
+      "grad_norm": 1.3523527383804321,
+      "learning_rate": 0.00019649712108536286,
+      "loss": 2.1644,
+      "step": 1065
+    },
+    {
+      "epoch": 0.11215149921094161,
+      "grad_norm": 1.2581042051315308,
+      "learning_rate": 0.00019648817508739983,
+      "loss": 2.1609,
+      "step": 1066
+    },
+    {
+      "epoch": 0.11225670699631773,
+      "grad_norm": 1.0684157609939575,
+      "learning_rate": 0.00019647921788458272,
+      "loss": 2.4841,
+      "step": 1067
+    },
+    {
+      "epoch": 0.11236191478169384,
+      "grad_norm": 0.9492899775505066,
+      "learning_rate": 0.00019647024947795175,
+      "loss": 1.8177,
+      "step": 1068
+    },
+    {
+      "epoch": 0.11246712256706996,
+      "grad_norm": 1.1650155782699585,
+      "learning_rate": 0.00019646126986854837,
+      "loss": 1.6481,
+      "step": 1069
+    },
+    {
+      "epoch": 0.11257233035244608,
+      "grad_norm": 1.2191792726516724,
+      "learning_rate": 0.00019645227905741534,
+      "loss": 1.7948,
+      "step": 1070
+    },
+    {
+      "epoch": 0.1126775381378222,
+      "grad_norm": 1.074036717414856,
+      "learning_rate": 0.0001964432770455968,
+      "loss": 1.764,
+      "step": 1071
+    },
+    {
+      "epoch": 0.11278274592319831,
+      "grad_norm": 1.0380913019180298,
+      "learning_rate": 0.00019643426383413805,
+      "loss": 1.526,
+      "step": 1072
+    },
+    {
+      "epoch": 0.11288795370857443,
+      "grad_norm": 1.4645812511444092,
+      "learning_rate": 0.0001964252394240858,
+      "loss": 2.016,
+      "step": 1073
+    },
+    {
+      "epoch": 0.11299316149395056,
+      "grad_norm": 0.8805195093154907,
+      "learning_rate": 0.000196416203816488,
+      "loss": 2.0084,
+      "step": 1074
+    },
+    {
+      "epoch": 0.11309836927932668,
+      "grad_norm": 1.593756079673767,
+      "learning_rate": 0.00019640715701239395,
+      "loss": 2.1586,
+      "step": 1075
+    },
+    {
+      "epoch": 0.11320357706470278,
+      "grad_norm": 0.9434992671012878,
+      "learning_rate": 0.00019639809901285423,
+      "loss": 1.8971,
+      "step": 1076
+    },
+    {
+      "epoch": 0.1133087848500789,
+      "grad_norm": 1.0768349170684814,
+      "learning_rate": 0.00019638902981892068,
+      "loss": 2.289,
+      "step": 1077
+    },
+    {
+      "epoch": 0.11341399263545503,
+      "grad_norm": 1.1029759645462036,
+      "learning_rate": 0.0001963799494316465,
+      "loss": 2.2014,
+      "step": 1078
+    },
+    {
+      "epoch": 0.11351920042083115,
+      "grad_norm": 1.1551514863967896,
+      "learning_rate": 0.0001963708578520862,
+      "loss": 2.3707,
+      "step": 1079
+    },
+    {
+      "epoch": 0.11362440820620726,
+      "grad_norm": 1.2115519046783447,
+      "learning_rate": 0.00019636175508129552,
+      "loss": 1.4232,
+      "step": 1080
+    },
+    {
+      "epoch": 0.11372961599158338,
+      "grad_norm": 1.1642786264419556,
+      "learning_rate": 0.0001963526411203315,
+      "loss": 2.1658,
+      "step": 1081
+    },
+    {
+      "epoch": 0.1138348237769595,
+      "grad_norm": 1.1630796194076538,
+      "learning_rate": 0.00019634351597025255,
+      "loss": 2.0171,
+      "step": 1082
+    },
+    {
+      "epoch": 0.11394003156233562,
+      "grad_norm": 1.058016061782837,
+      "learning_rate": 0.00019633437963211832,
+      "loss": 1.5272,
+      "step": 1083
+    },
+    {
+      "epoch": 0.11404523934771173,
+      "grad_norm": 1.1244498491287231,
+      "learning_rate": 0.00019632523210698987,
+      "loss": 2.4017,
+      "step": 1084
+    },
+    {
+      "epoch": 0.11415044713308785,
+      "grad_norm": 1.0353667736053467,
+      "learning_rate": 0.00019631607339592937,
+      "loss": 1.9151,
+      "step": 1085
+    },
+    {
+      "epoch": 0.11425565491846397,
+      "grad_norm": 0.735931396484375,
+      "learning_rate": 0.00019630690350000042,
+      "loss": 2.3843,
+      "step": 1086
+    },
+    {
+      "epoch": 0.11436086270384009,
+      "grad_norm": 0.9920666217803955,
+      "learning_rate": 0.00019629772242026793,
+      "loss": 2.1867,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1144660704892162,
+      "grad_norm": 0.9778599739074707,
+      "learning_rate": 0.000196288530157798,
+      "loss": 2.3535,
+      "step": 1088
+    },
+    {
+      "epoch": 0.11457127827459232,
+      "grad_norm": 1.1692768335342407,
+      "learning_rate": 0.00019627932671365813,
+      "loss": 1.9833,
+      "step": 1089
+    },
+    {
+      "epoch": 0.11467648605996844,
+      "grad_norm": 1.3122508525848389,
+      "learning_rate": 0.00019627011208891713,
+      "loss": 1.8988,
+      "step": 1090
+    },
+    {
+      "epoch": 0.11478169384534456,
+      "grad_norm": 1.044777512550354,
+      "learning_rate": 0.00019626088628464498,
+      "loss": 1.92,
+      "step": 1091
+    },
+    {
+      "epoch": 0.11488690163072067,
+      "grad_norm": 0.8848743438720703,
+      "learning_rate": 0.00019625164930191311,
+      "loss": 2.4497,
+      "step": 1092
+    },
+    {
+      "epoch": 0.11499210941609679,
+      "grad_norm": 0.9451887607574463,
+      "learning_rate": 0.00019624240114179416,
+      "loss": 1.7285,
+      "step": 1093
+    },
+    {
+      "epoch": 0.11509731720147291,
+      "grad_norm": 0.9410362243652344,
+      "learning_rate": 0.00019623314180536205,
+      "loss": 1.9463,
+      "step": 1094
+    },
+    {
+      "epoch": 0.11520252498684903,
+      "grad_norm": 1.023637294769287,
+      "learning_rate": 0.00019622387129369212,
+      "loss": 1.8648,
+      "step": 1095
+    },
+    {
+      "epoch": 0.11530773277222514,
+      "grad_norm": 1.2272717952728271,
+      "learning_rate": 0.00019621458960786083,
+      "loss": 2.1191,
+      "step": 1096
+    },
+    {
+      "epoch": 0.11541294055760126,
+      "grad_norm": 1.150048851966858,
+      "learning_rate": 0.0001962052967489461,
+      "loss": 1.7046,
+      "step": 1097
+    },
+    {
+      "epoch": 0.11551814834297738,
+      "grad_norm": 0.7519487738609314,
+      "learning_rate": 0.00019619599271802706,
+      "loss": 2.2551,
+      "step": 1098
+    },
+    {
+      "epoch": 0.1156233561283535,
+      "grad_norm": 1.0038561820983887,
+      "learning_rate": 0.00019618667751618416,
+      "loss": 2.2381,
+      "step": 1099
+    },
+    {
+      "epoch": 0.11572856391372961,
+      "grad_norm": 0.9891253113746643,
+      "learning_rate": 0.0001961773511444991,
+      "loss": 2.2862,
+      "step": 1100
+    },
+    {
+      "epoch": 0.11583377169910573,
+      "grad_norm": 0.9823799729347229,
+      "learning_rate": 0.00019616801360405499,
+      "loss": 1.7907,
+      "step": 1101
+    },
+    {
+      "epoch": 0.11593897948448185,
+      "grad_norm": 2.5130529403686523,
+      "learning_rate": 0.0001961586648959361,
+      "loss": 2.1288,
+      "step": 1102
+    },
+    {
+      "epoch": 0.11604418726985798,
+      "grad_norm": 0.940342366695404,
+      "learning_rate": 0.00019614930502122812,
+      "loss": 2.1486,
+      "step": 1103
+    },
+    {
+      "epoch": 0.11614939505523408,
+      "grad_norm": 1.1634947061538696,
+      "learning_rate": 0.00019613993398101795,
+      "loss": 1.3548,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1162546028406102,
+      "grad_norm": 1.2739289999008179,
+      "learning_rate": 0.00019613055177639384,
+      "loss": 2.4006,
+      "step": 1105
+    },
+    {
+      "epoch": 0.11635981062598633,
+      "grad_norm": 0.7770984768867493,
+      "learning_rate": 0.0001961211584084453,
+      "loss": 2.2417,
+      "step": 1106
+    },
+    {
+      "epoch": 0.11646501841136245,
+      "grad_norm": 0.9923128485679626,
+      "learning_rate": 0.00019611175387826315,
+      "loss": 1.9447,
+      "step": 1107
+    },
+    {
+      "epoch": 0.11657022619673856,
+      "grad_norm": 1.3460139036178589,
+      "learning_rate": 0.00019610233818693953,
+      "loss": 1.8778,
+      "step": 1108
+    },
+    {
+      "epoch": 0.11667543398211468,
+      "grad_norm": 0.7878180146217346,
+      "learning_rate": 0.0001960929113355678,
+      "loss": 1.8391,
+      "step": 1109
+    },
+    {
+      "epoch": 0.1167806417674908,
+      "grad_norm": 1.1408841609954834,
+      "learning_rate": 0.00019608347332524272,
+      "loss": 1.8246,
+      "step": 1110
+    },
+    {
+      "epoch": 0.11688584955286692,
+      "grad_norm": 1.3990023136138916,
+      "learning_rate": 0.00019607402415706027,
+      "loss": 1.8562,
+      "step": 1111
+    },
+    {
+      "epoch": 0.11699105733824303,
+      "grad_norm": 1.045021414756775,
+      "learning_rate": 0.00019606456383211777,
+      "loss": 1.9724,
+      "step": 1112
+    },
+    {
+      "epoch": 0.11709626512361915,
+      "grad_norm": 0.6789584755897522,
+      "learning_rate": 0.0001960550923515138,
+      "loss": 2.149,
+      "step": 1113
+    },
+    {
+      "epoch": 0.11720147290899527,
+      "grad_norm": 0.8170668482780457,
+      "learning_rate": 0.00019604560971634826,
+      "loss": 2.0608,
+      "step": 1114
+    },
+    {
+      "epoch": 0.11730668069437139,
+      "grad_norm": 0.6909334659576416,
+      "learning_rate": 0.00019603611592772233,
+      "loss": 2.1037,
+      "step": 1115
+    },
+    {
+      "epoch": 0.1174118884797475,
+      "grad_norm": 0.8929466605186462,
+      "learning_rate": 0.0001960266109867385,
+      "loss": 1.7609,
+      "step": 1116
+    },
+    {
+      "epoch": 0.11751709626512362,
+      "grad_norm": 1.2429050207138062,
+      "learning_rate": 0.00019601709489450056,
+      "loss": 1.6607,
+      "step": 1117
+    },
+    {
+      "epoch": 0.11762230405049974,
+      "grad_norm": 0.7718959450721741,
+      "learning_rate": 0.00019600756765211354,
+      "loss": 2.273,
+      "step": 1118
+    },
+    {
+      "epoch": 0.11772751183587586,
+      "grad_norm": 1.4449671506881714,
+      "learning_rate": 0.00019599802926068384,
+      "loss": 1.9564,
+      "step": 1119
+    },
+    {
+      "epoch": 0.11783271962125197,
+      "grad_norm": 0.955116868019104,
+      "learning_rate": 0.00019598847972131914,
+      "loss": 1.8775,
+      "step": 1120
+    },
+    {
+      "epoch": 0.11793792740662809,
+      "grad_norm": 1.5084177255630493,
+      "learning_rate": 0.00019597891903512835,
+      "loss": 2.0589,
+      "step": 1121
+    },
+    {
+      "epoch": 0.11804313519200421,
+      "grad_norm": 1.039664626121521,
+      "learning_rate": 0.00019596934720322176,
+      "loss": 2.331,
+      "step": 1122
+    },
+    {
+      "epoch": 0.11814834297738033,
+      "grad_norm": 1.2235255241394043,
+      "learning_rate": 0.00019595976422671086,
+      "loss": 1.7996,
+      "step": 1123
+    },
+    {
+      "epoch": 0.11825355076275644,
+      "grad_norm": 0.9834601283073425,
+      "learning_rate": 0.00019595017010670858,
+      "loss": 2.4876,
+      "step": 1124
+    },
+    {
+      "epoch": 0.11835875854813256,
+      "grad_norm": 1.0267283916473389,
+      "learning_rate": 0.00019594056484432897,
+      "loss": 2.1381,
+      "step": 1125
+    },
+    {
+      "epoch": 0.11846396633350868,
+      "grad_norm": 1.187152624130249,
+      "learning_rate": 0.00019593094844068748,
+      "loss": 2.1336,
+      "step": 1126
+    },
+    {
+      "epoch": 0.1185691741188848,
+      "grad_norm": 0.8260311484336853,
+      "learning_rate": 0.00019592132089690085,
+      "loss": 2.0608,
+      "step": 1127
+    },
+    {
+      "epoch": 0.11867438190426091,
+      "grad_norm": 0.8362654447555542,
+      "learning_rate": 0.0001959116822140871,
+      "loss": 2.2765,
+      "step": 1128
+    },
+    {
+      "epoch": 0.11877958968963703,
+      "grad_norm": 0.9070550799369812,
+      "learning_rate": 0.00019590203239336552,
+      "loss": 2.1471,
+      "step": 1129
+    },
+    {
+      "epoch": 0.11888479747501315,
+      "grad_norm": 0.8633410334587097,
+      "learning_rate": 0.0001958923714358567,
+      "loss": 2.2632,
+      "step": 1130
+    },
+    {
+      "epoch": 0.11899000526038928,
+      "grad_norm": 1.0233980417251587,
+      "learning_rate": 0.00019588269934268257,
+      "loss": 2.0631,
+      "step": 1131
+    },
+    {
+      "epoch": 0.11909521304576538,
+      "grad_norm": 1.4428153038024902,
+      "learning_rate": 0.00019587301611496632,
+      "loss": 1.646,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1192004208311415,
+      "grad_norm": 1.2302439212799072,
+      "learning_rate": 0.00019586332175383238,
+      "loss": 2.0555,
+      "step": 1133
+    },
+    {
+      "epoch": 0.11930562861651763,
+      "grad_norm": 0.8173778057098389,
+      "learning_rate": 0.00019585361626040654,
+      "loss": 1.8569,
+      "step": 1134
+    },
+    {
+      "epoch": 0.11941083640189375,
+      "grad_norm": 1.5377962589263916,
+      "learning_rate": 0.00019584389963581592,
+      "loss": 1.9263,
+      "step": 1135
+    },
+    {
+      "epoch": 0.11951604418726985,
+      "grad_norm": 0.9011675119400024,
+      "learning_rate": 0.00019583417188118882,
+      "loss": 1.7328,
+      "step": 1136
+    },
+    {
+      "epoch": 0.11962125197264598,
+      "grad_norm": 1.0108987092971802,
+      "learning_rate": 0.0001958244329976549,
+      "loss": 1.8854,
+      "step": 1137
+    },
+    {
+      "epoch": 0.1197264597580221,
+      "grad_norm": 1.1729274988174438,
+      "learning_rate": 0.00019581468298634515,
+      "loss": 2.1649,
+      "step": 1138
+    },
+    {
+      "epoch": 0.11983166754339822,
+      "grad_norm": 1.0889396667480469,
+      "learning_rate": 0.00019580492184839175,
+      "loss": 2.0792,
+      "step": 1139
+    },
+    {
+      "epoch": 0.11993687532877433,
+      "grad_norm": 1.3217434883117676,
+      "learning_rate": 0.00019579514958492826,
+      "loss": 2.1042,
+      "step": 1140
+    },
+    {
+      "epoch": 0.12004208311415045,
+      "grad_norm": 1.0219428539276123,
+      "learning_rate": 0.00019578536619708952,
+      "loss": 2.3675,
+      "step": 1141
+    },
+    {
+      "epoch": 0.12014729089952657,
+      "grad_norm": 1.771694302558899,
+      "learning_rate": 0.0001957755716860116,
+      "loss": 2.1605,
+      "step": 1142
+    },
+    {
+      "epoch": 0.12025249868490269,
+      "grad_norm": 1.005615472793579,
+      "learning_rate": 0.0001957657660528319,
+      "loss": 1.9396,
+      "step": 1143
+    },
+    {
+      "epoch": 0.1203577064702788,
+      "grad_norm": 0.8795095086097717,
+      "learning_rate": 0.00019575594929868918,
+      "loss": 2.2603,
+      "step": 1144
+    },
+    {
+      "epoch": 0.12046291425565492,
+      "grad_norm": 1.1998109817504883,
+      "learning_rate": 0.00019574612142472334,
+      "loss": 2.5464,
+      "step": 1145
+    },
+    {
+      "epoch": 0.12056812204103104,
+      "grad_norm": 0.9663223624229431,
+      "learning_rate": 0.00019573628243207573,
+      "loss": 1.8903,
+      "step": 1146
+    },
+    {
+      "epoch": 0.12067332982640716,
+      "grad_norm": 0.9603292346000671,
+      "learning_rate": 0.0001957264323218889,
+      "loss": 2.1182,
+      "step": 1147
+    },
+    {
+      "epoch": 0.12077853761178327,
+      "grad_norm": 0.8244251608848572,
+      "learning_rate": 0.00019571657109530667,
+      "loss": 2.0749,
+      "step": 1148
+    },
+    {
+      "epoch": 0.12088374539715939,
+      "grad_norm": 1.2357412576675415,
+      "learning_rate": 0.00019570669875347427,
+      "loss": 1.8142,
+      "step": 1149
+    },
+    {
+      "epoch": 0.12098895318253551,
+      "grad_norm": 1.6487301588058472,
+      "learning_rate": 0.00019569681529753806,
+      "loss": 2.057,
+      "step": 1150
+    },
+    {
+      "epoch": 0.12109416096791163,
+      "grad_norm": 1.3535126447677612,
+      "learning_rate": 0.00019568692072864581,
+      "loss": 1.8645,
+      "step": 1151
+    },
+    {
+      "epoch": 0.12119936875328774,
+      "grad_norm": 0.983331024646759,
+      "learning_rate": 0.0001956770150479466,
+      "loss": 2.2773,
+      "step": 1152
+    },
+    {
+      "epoch": 0.12130457653866386,
+      "grad_norm": 0.625877320766449,
+      "learning_rate": 0.00019566709825659064,
+      "loss": 2.2146,
+      "step": 1153
+    },
+    {
+      "epoch": 0.12140978432403998,
+      "grad_norm": 1.0601180791854858,
+      "learning_rate": 0.0001956571703557296,
+      "loss": 2.2351,
+      "step": 1154
+    },
+    {
+      "epoch": 0.1215149921094161,
+      "grad_norm": 1.0137640237808228,
+      "learning_rate": 0.00019564723134651634,
+      "loss": 2.5221,
+      "step": 1155
+    },
+    {
+      "epoch": 0.12162019989479221,
+      "grad_norm": 1.044480323791504,
+      "learning_rate": 0.0001956372812301051,
+      "loss": 2.14,
+      "step": 1156
+    },
+    {
+      "epoch": 0.12172540768016833,
+      "grad_norm": 1.44232976436615,
+      "learning_rate": 0.00019562732000765127,
+      "loss": 1.9134,
+      "step": 1157
+    },
+    {
+      "epoch": 0.12183061546554445,
+      "grad_norm": 1.2370481491088867,
+      "learning_rate": 0.0001956173476803117,
+      "loss": 1.727,
+      "step": 1158
+    },
+    {
+      "epoch": 0.12193582325092057,
+      "grad_norm": 1.006666898727417,
+      "learning_rate": 0.00019560736424924439,
+      "loss": 2.221,
+      "step": 1159
+    },
+    {
+      "epoch": 0.12204103103629668,
+      "grad_norm": 0.8660513162612915,
+      "learning_rate": 0.0001955973697156087,
+      "loss": 2.1285,
+      "step": 1160
+    },
+    {
+      "epoch": 0.1221462388216728,
+      "grad_norm": 0.7105367183685303,
+      "learning_rate": 0.00019558736408056525,
+      "loss": 2.0834,
+      "step": 1161
+    },
+    {
+      "epoch": 0.12225144660704892,
+      "grad_norm": 1.236364483833313,
+      "learning_rate": 0.000195577347345276,
+      "loss": 1.8932,
+      "step": 1162
+    },
+    {
+      "epoch": 0.12235665439242505,
+      "grad_norm": 1.5001050233840942,
+      "learning_rate": 0.0001955673195109041,
+      "loss": 2.263,
+      "step": 1163
+    },
+    {
+      "epoch": 0.12246186217780115,
+      "grad_norm": 0.8300707936286926,
+      "learning_rate": 0.0001955572805786141,
+      "loss": 2.3241,
+      "step": 1164
+    },
+    {
+      "epoch": 0.12256706996317727,
+      "grad_norm": 1.1796188354492188,
+      "learning_rate": 0.00019554723054957175,
+      "loss": 1.7304,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1226722777485534,
+      "grad_norm": 1.1886162757873535,
+      "learning_rate": 0.00019553716942494415,
+      "loss": 1.8389,
+      "step": 1166
+    },
+    {
+      "epoch": 0.12277748553392952,
+      "grad_norm": 1.0573186874389648,
+      "learning_rate": 0.00019552709720589966,
+      "loss": 2.0146,
+      "step": 1167
+    },
+    {
+      "epoch": 0.12288269331930562,
+      "grad_norm": 0.8099238276481628,
+      "learning_rate": 0.00019551701389360795,
+      "loss": 1.593,
+      "step": 1168
+    },
+    {
+      "epoch": 0.12298790110468175,
+      "grad_norm": 1.0094003677368164,
+      "learning_rate": 0.00019550691948923992,
+      "loss": 2.4218,
+      "step": 1169
+    },
+    {
+      "epoch": 0.12309310889005787,
+      "grad_norm": 0.8765414953231812,
+      "learning_rate": 0.00019549681399396785,
+      "loss": 2.2399,
+      "step": 1170
+    },
+    {
+      "epoch": 0.12319831667543399,
+      "grad_norm": 1.4000588655471802,
+      "learning_rate": 0.00019548669740896525,
+      "loss": 1.7225,
+      "step": 1171
+    },
+    {
+      "epoch": 0.1233035244608101,
+      "grad_norm": 1.909455418586731,
+      "learning_rate": 0.0001954765697354069,
+      "loss": 2.0569,
+      "step": 1172
+    },
+    {
+      "epoch": 0.12340873224618622,
+      "grad_norm": 1.0092931985855103,
+      "learning_rate": 0.00019546643097446888,
+      "loss": 1.8983,
+      "step": 1173
+    },
+    {
+      "epoch": 0.12351394003156234,
+      "grad_norm": 1.216364860534668,
+      "learning_rate": 0.0001954562811273286,
+      "loss": 2.0644,
+      "step": 1174
+    },
+    {
+      "epoch": 0.12361914781693846,
+      "grad_norm": 1.245184302330017,
+      "learning_rate": 0.00019544612019516472,
+      "loss": 1.4739,
+      "step": 1175
+    },
+    {
+      "epoch": 0.12372435560231457,
+      "grad_norm": 1.5006418228149414,
+      "learning_rate": 0.00019543594817915722,
+      "loss": 2.0464,
+      "step": 1176
+    },
+    {
+      "epoch": 0.12382956338769069,
+      "grad_norm": 1.2883630990982056,
+      "learning_rate": 0.00019542576508048732,
+      "loss": 1.9773,
+      "step": 1177
+    },
+    {
+      "epoch": 0.12393477117306681,
+      "grad_norm": 1.4748666286468506,
+      "learning_rate": 0.00019541557090033753,
+      "loss": 2.0239,
+      "step": 1178
+    },
+    {
+      "epoch": 0.12403997895844293,
+      "grad_norm": 1.251774549484253,
+      "learning_rate": 0.0001954053656398917,
+      "loss": 1.8751,
+      "step": 1179
+    },
+    {
+      "epoch": 0.12414518674381904,
+      "grad_norm": 1.2670092582702637,
+      "learning_rate": 0.00019539514930033493,
+      "loss": 2.093,
+      "step": 1180
+    },
+    {
+      "epoch": 0.12425039452919516,
+      "grad_norm": 0.845373272895813,
+      "learning_rate": 0.00019538492188285358,
+      "loss": 2.2009,
+      "step": 1181
+    },
+    {
+      "epoch": 0.12435560231457128,
+      "grad_norm": 1.3684707880020142,
+      "learning_rate": 0.00019537468338863537,
+      "loss": 2.1124,
+      "step": 1182
+    },
+    {
+      "epoch": 0.1244608100999474,
+      "grad_norm": 1.0904892683029175,
+      "learning_rate": 0.0001953644338188692,
+      "loss": 2.2674,
+      "step": 1183
+    },
+    {
+      "epoch": 0.12456601788532351,
+      "grad_norm": 1.0843944549560547,
+      "learning_rate": 0.0001953541731747454,
+      "loss": 2.1427,
+      "step": 1184
+    },
+    {
+      "epoch": 0.12467122567069963,
+      "grad_norm": 1.4140195846557617,
+      "learning_rate": 0.00019534390145745545,
+      "loss": 1.9423,
+      "step": 1185
+    },
+    {
+      "epoch": 0.12477643345607575,
+      "grad_norm": 1.2071352005004883,
+      "learning_rate": 0.00019533361866819218,
+      "loss": 1.7771,
+      "step": 1186
+    },
+    {
+      "epoch": 0.12488164124145187,
+      "grad_norm": 0.9805304408073425,
+      "learning_rate": 0.0001953233248081497,
+      "loss": 2.1534,
+      "step": 1187
+    },
+    {
+      "epoch": 0.12498684902682798,
+      "grad_norm": 0.8106799721717834,
+      "learning_rate": 0.0001953130198785234,
+      "loss": 1.8775,
+      "step": 1188
+    },
+    {
+      "epoch": 0.12509205681220412,
+      "grad_norm": 1.2186012268066406,
+      "learning_rate": 0.00019530270388050998,
+      "loss": 2.0908,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1251972645975802,
+      "grad_norm": 1.270003318786621,
+      "learning_rate": 0.00019529237681530735,
+      "loss": 1.9182,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12530247238295633,
+      "grad_norm": 1.1836543083190918,
+      "learning_rate": 0.00019528203868411482,
+      "loss": 2.2695,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12540768016833245,
+      "grad_norm": 1.4582277536392212,
+      "learning_rate": 0.00019527168948813288,
+      "loss": 1.7637,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12551288795370857,
+      "grad_norm": 1.24919593334198,
+      "learning_rate": 0.00019526132922856334,
+      "loss": 1.8924,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1256180957390847,
+      "grad_norm": 1.0060888528823853,
+      "learning_rate": 0.00019525095790660937,
+      "loss": 2.2829,
+      "step": 1194
+    },
+    {
+      "epoch": 0.12572330352446082,
+      "grad_norm": 1.7646416425704956,
+      "learning_rate": 0.00019524057552347527,
+      "loss": 2.0088,
+      "step": 1195
+    },
+    {
+      "epoch": 0.12582851130983694,
+      "grad_norm": 1.0595672130584717,
+      "learning_rate": 0.00019523018208036677,
+      "loss": 1.9792,
+      "step": 1196
+    },
+    {
+      "epoch": 0.12593371909521306,
+      "grad_norm": 0.9665113091468811,
+      "learning_rate": 0.00019521977757849083,
+      "loss": 1.8552,
+      "step": 1197
+    },
+    {
+      "epoch": 0.12603892688058915,
+      "grad_norm": 0.996906042098999,
+      "learning_rate": 0.00019520936201905566,
+      "loss": 1.9122,
+      "step": 1198
+    },
+    {
+      "epoch": 0.12614413466596527,
+      "grad_norm": 1.0974137783050537,
+      "learning_rate": 0.0001951989354032708,
+      "loss": 2.3548,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1262493424513414,
+      "grad_norm": 1.2760165929794312,
+      "learning_rate": 0.00019518849773234704,
+      "loss": 1.9927,
+      "step": 1200
+    },
+    {
+      "epoch": 0.12635455023671752,
+      "grad_norm": 0.8448911309242249,
+      "learning_rate": 0.0001951780490074965,
+      "loss": 2.1343,
+      "step": 1201
+    },
+    {
+      "epoch": 0.12645975802209364,
+      "grad_norm": 0.8463537096977234,
+      "learning_rate": 0.00019516758922993256,
+      "loss": 2.4417,
+      "step": 1202
+    },
+    {
+      "epoch": 0.12656496580746976,
+      "grad_norm": 0.971531331539154,
+      "learning_rate": 0.0001951571184008698,
+      "loss": 2.0901,
+      "step": 1203
+    },
+    {
+      "epoch": 0.12667017359284588,
+      "grad_norm": 1.1150871515274048,
+      "learning_rate": 0.00019514663652152428,
+      "loss": 2.3631,
+      "step": 1204
+    },
+    {
+      "epoch": 0.126775381378222,
+      "grad_norm": 1.44049870967865,
+      "learning_rate": 0.00019513614359311315,
+      "loss": 1.9413,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1268805891635981,
+      "grad_norm": 1.1765494346618652,
+      "learning_rate": 0.00019512563961685494,
+      "loss": 2.2888,
+      "step": 1206
+    },
+    {
+      "epoch": 0.12698579694897422,
+      "grad_norm": 1.1426244974136353,
+      "learning_rate": 0.00019511512459396944,
+      "loss": 1.7372,
+      "step": 1207
+    },
+    {
+      "epoch": 0.12709100473435034,
+      "grad_norm": 1.2937824726104736,
+      "learning_rate": 0.00019510459852567773,
+      "loss": 1.7583,
+      "step": 1208
+    },
+    {
+      "epoch": 0.12719621251972646,
+      "grad_norm": 1.0427653789520264,
+      "learning_rate": 0.0001950940614132022,
+      "loss": 2.2694,
+      "step": 1209
+    },
+    {
+      "epoch": 0.12730142030510258,
+      "grad_norm": 1.0768921375274658,
+      "learning_rate": 0.00019508351325776642,
+      "loss": 2.0166,
+      "step": 1210
+    },
+    {
+      "epoch": 0.1274066280904787,
+      "grad_norm": 1.1876052618026733,
+      "learning_rate": 0.00019507295406059533,
+      "loss": 2.2435,
+      "step": 1211
+    },
+    {
+      "epoch": 0.12751183587585482,
+      "grad_norm": 1.4047654867172241,
+      "learning_rate": 0.0001950623838229152,
+      "loss": 1.9735,
+      "step": 1212
+    },
+    {
+      "epoch": 0.12761704366123094,
+      "grad_norm": 1.5884112119674683,
+      "learning_rate": 0.00019505180254595343,
+      "loss": 1.8768,
+      "step": 1213
+    },
+    {
+      "epoch": 0.12772225144660704,
+      "grad_norm": 0.8576595187187195,
+      "learning_rate": 0.00019504121023093888,
+      "loss": 2.186,
+      "step": 1214
+    },
+    {
+      "epoch": 0.12782745923198316,
+      "grad_norm": 1.0435853004455566,
+      "learning_rate": 0.00019503060687910148,
+      "loss": 1.963,
+      "step": 1215
+    },
+    {
+      "epoch": 0.12793266701735928,
+      "grad_norm": 1.279901385307312,
+      "learning_rate": 0.0001950199924916727,
+      "loss": 1.7715,
+      "step": 1216
+    },
+    {
+      "epoch": 0.1280378748027354,
+      "grad_norm": 0.7529373168945312,
+      "learning_rate": 0.00019500936706988502,
+      "loss": 2.0821,
+      "step": 1217
+    },
+    {
+      "epoch": 0.12814308258811152,
+      "grad_norm": 1.64811372756958,
+      "learning_rate": 0.00019499873061497246,
+      "loss": 2.1265,
+      "step": 1218
+    },
+    {
+      "epoch": 0.12824829037348764,
+      "grad_norm": 1.1517812013626099,
+      "learning_rate": 0.00019498808312817006,
+      "loss": 1.9678,
+      "step": 1219
+    },
+    {
+      "epoch": 0.12835349815886377,
+      "grad_norm": 1.0397827625274658,
+      "learning_rate": 0.00019497742461071441,
+      "loss": 1.9882,
+      "step": 1220
+    },
+    {
+      "epoch": 0.1284587059442399,
+      "grad_norm": 1.1834861040115356,
+      "learning_rate": 0.0001949667550638432,
+      "loss": 2.175,
+      "step": 1221
+    },
+    {
+      "epoch": 0.12856391372961598,
+      "grad_norm": 1.0176396369934082,
+      "learning_rate": 0.00019495607448879546,
+      "loss": 1.9798,
+      "step": 1222
+    },
+    {
+      "epoch": 0.1286691215149921,
+      "grad_norm": 1.2823526859283447,
+      "learning_rate": 0.00019494538288681145,
+      "loss": 1.8454,
+      "step": 1223
+    },
+    {
+      "epoch": 0.12877432930036822,
+      "grad_norm": 1.0017826557159424,
+      "learning_rate": 0.00019493468025913276,
+      "loss": 2.1296,
+      "step": 1224
+    },
+    {
+      "epoch": 0.12887953708574434,
+      "grad_norm": 0.8220755457878113,
+      "learning_rate": 0.00019492396660700226,
+      "loss": 1.9987,
+      "step": 1225
+    },
+    {
+      "epoch": 0.12898474487112047,
+      "grad_norm": 0.9506868124008179,
+      "learning_rate": 0.00019491324193166408,
+      "loss": 2.3586,
+      "step": 1226
+    },
+    {
+      "epoch": 0.1290899526564966,
+      "grad_norm": 0.796829342842102,
+      "learning_rate": 0.00019490250623436367,
+      "loss": 1.4214,
+      "step": 1227
+    },
+    {
+      "epoch": 0.1291951604418727,
+      "grad_norm": 1.117042899131775,
+      "learning_rate": 0.00019489175951634775,
+      "loss": 1.7607,
+      "step": 1228
+    },
+    {
+      "epoch": 0.12930036822724883,
+      "grad_norm": 1.1482354402542114,
+      "learning_rate": 0.00019488100177886427,
+      "loss": 1.782,
+      "step": 1229
+    },
+    {
+      "epoch": 0.12940557601262492,
+      "grad_norm": 0.8992315530776978,
+      "learning_rate": 0.00019487023302316243,
+      "loss": 2.3632,
+      "step": 1230
+    },
+    {
+      "epoch": 0.12951078379800104,
+      "grad_norm": 0.9906787872314453,
+      "learning_rate": 0.00019485945325049288,
+      "loss": 1.8036,
+      "step": 1231
+    },
+    {
+      "epoch": 0.12961599158337717,
+      "grad_norm": 1.0407614707946777,
+      "learning_rate": 0.00019484866246210738,
+      "loss": 1.9722,
+      "step": 1232
+    },
+    {
+      "epoch": 0.1297211993687533,
+      "grad_norm": 1.3917356729507446,
+      "learning_rate": 0.00019483786065925904,
+      "loss": 1.6965,
+      "step": 1233
+    },
+    {
+      "epoch": 0.1298264071541294,
+      "grad_norm": 0.9043229818344116,
+      "learning_rate": 0.0001948270478432022,
+      "loss": 1.9409,
+      "step": 1234
+    },
+    {
+      "epoch": 0.12993161493950553,
+      "grad_norm": 1.2875796556472778,
+      "learning_rate": 0.0001948162240151926,
+      "loss": 2.0062,
+      "step": 1235
+    },
+    {
+      "epoch": 0.13003682272488165,
+      "grad_norm": 1.3274993896484375,
+      "learning_rate": 0.00019480538917648711,
+      "loss": 1.8964,
+      "step": 1236
+    },
+    {
+      "epoch": 0.13014203051025777,
+      "grad_norm": 0.974236786365509,
+      "learning_rate": 0.00019479454332834396,
+      "loss": 1.4836,
+      "step": 1237
+    },
+    {
+      "epoch": 0.13024723829563387,
+      "grad_norm": 0.9269886016845703,
+      "learning_rate": 0.00019478368647202264,
+      "loss": 2.19,
+      "step": 1238
+    },
+    {
+      "epoch": 0.13035244608101,
+      "grad_norm": 0.7717685699462891,
+      "learning_rate": 0.0001947728186087839,
+      "loss": 2.4556,
+      "step": 1239
+    },
+    {
+      "epoch": 0.1304576538663861,
+      "grad_norm": 1.7023308277130127,
+      "learning_rate": 0.00019476193973988988,
+      "loss": 2.433,
+      "step": 1240
+    },
+    {
+      "epoch": 0.13056286165176223,
+      "grad_norm": 1.3712750673294067,
+      "learning_rate": 0.0001947510498666038,
+      "loss": 2.1469,
+      "step": 1241
+    },
+    {
+      "epoch": 0.13066806943713835,
+      "grad_norm": 1.1237108707427979,
+      "learning_rate": 0.0001947401489901903,
+      "loss": 2.3985,
+      "step": 1242
+    },
+    {
+      "epoch": 0.13077327722251447,
+      "grad_norm": 1.0519534349441528,
+      "learning_rate": 0.0001947292371119153,
+      "loss": 2.3266,
+      "step": 1243
+    },
+    {
+      "epoch": 0.1308784850078906,
+      "grad_norm": 0.9474523663520813,
+      "learning_rate": 0.0001947183142330459,
+      "loss": 2.0388,
+      "step": 1244
+    },
+    {
+      "epoch": 0.13098369279326671,
+      "grad_norm": 1.21075439453125,
+      "learning_rate": 0.00019470738035485058,
+      "loss": 1.7828,
+      "step": 1245
+    },
+    {
+      "epoch": 0.1310889005786428,
+      "grad_norm": 0.9895322322845459,
+      "learning_rate": 0.00019469643547859904,
+      "loss": 2.0604,
+      "step": 1246
+    },
+    {
+      "epoch": 0.13119410836401893,
+      "grad_norm": 1.5811057090759277,
+      "learning_rate": 0.0001946854796055623,
+      "loss": 1.7441,
+      "step": 1247
+    },
+    {
+      "epoch": 0.13129931614939505,
+      "grad_norm": 1.3504817485809326,
+      "learning_rate": 0.00019467451273701256,
+      "loss": 2.4175,
+      "step": 1248
+    },
+    {
+      "epoch": 0.13140452393477117,
+      "grad_norm": 1.2100443840026855,
+      "learning_rate": 0.00019466353487422345,
+      "loss": 2.0371,
+      "step": 1249
+    },
+    {
+      "epoch": 0.1315097317201473,
+      "grad_norm": 0.8898875713348389,
+      "learning_rate": 0.00019465254601846974,
+      "loss": 2.3904,
+      "step": 1250
+    },
+    {
+      "epoch": 0.13161493950552342,
+      "grad_norm": 1.0012081861495972,
+      "learning_rate": 0.00019464154617102755,
+      "loss": 1.8023,
+      "step": 1251
+    },
+    {
+      "epoch": 0.13172014729089954,
+      "grad_norm": 1.0802556276321411,
+      "learning_rate": 0.00019463053533317425,
+      "loss": 2.0063,
+      "step": 1252
+    },
+    {
+      "epoch": 0.13182535507627566,
+      "grad_norm": 1.0085639953613281,
+      "learning_rate": 0.00019461951350618849,
+      "loss": 1.7158,
+      "step": 1253
+    },
+    {
+      "epoch": 0.13193056286165175,
+      "grad_norm": 0.7718974351882935,
+      "learning_rate": 0.00019460848069135017,
+      "loss": 2.2337,
+      "step": 1254
+    },
+    {
+      "epoch": 0.13203577064702787,
+      "grad_norm": 1.247361421585083,
+      "learning_rate": 0.0001945974368899406,
+      "loss": 2.0403,
+      "step": 1255
+    },
+    {
+      "epoch": 0.132140978432404,
+      "grad_norm": 0.9886590838432312,
+      "learning_rate": 0.00019458638210324212,
+      "loss": 2.1477,
+      "step": 1256
+    },
+    {
+      "epoch": 0.13224618621778012,
+      "grad_norm": 0.9976163506507874,
+      "learning_rate": 0.00019457531633253856,
+      "loss": 2.0871,
+      "step": 1257
+    },
+    {
+      "epoch": 0.13235139400315624,
+      "grad_norm": 0.9172899127006531,
+      "learning_rate": 0.00019456423957911497,
+      "loss": 2.0631,
+      "step": 1258
+    },
+    {
+      "epoch": 0.13245660178853236,
+      "grad_norm": 1.053161382675171,
+      "learning_rate": 0.0001945531518442576,
+      "loss": 2.0908,
+      "step": 1259
+    },
+    {
+      "epoch": 0.13256180957390848,
+      "grad_norm": 1.0828722715377808,
+      "learning_rate": 0.00019454205312925408,
+      "loss": 1.9512,
+      "step": 1260
+    },
+    {
+      "epoch": 0.1326670173592846,
+      "grad_norm": 0.9323759078979492,
+      "learning_rate": 0.00019453094343539325,
+      "loss": 2.3588,
+      "step": 1261
+    },
+    {
+      "epoch": 0.1327722251446607,
+      "grad_norm": 1.2428725957870483,
+      "learning_rate": 0.00019451982276396526,
+      "loss": 1.9449,
+      "step": 1262
+    },
+    {
+      "epoch": 0.13287743293003682,
+      "grad_norm": 1.8878873586654663,
+      "learning_rate": 0.00019450869111626147,
+      "loss": 1.4895,
+      "step": 1263
+    },
+    {
+      "epoch": 0.13298264071541294,
+      "grad_norm": 1.343825101852417,
+      "learning_rate": 0.0001944975484935746,
+      "loss": 1.713,
+      "step": 1264
+    },
+    {
+      "epoch": 0.13308784850078906,
+      "grad_norm": 0.9605780243873596,
+      "learning_rate": 0.0001944863948971986,
+      "loss": 2.0469,
+      "step": 1265
+    },
+    {
+      "epoch": 0.13319305628616518,
+      "grad_norm": 1.6616069078445435,
+      "learning_rate": 0.0001944752303284287,
+      "loss": 2.0865,
+      "step": 1266
+    },
+    {
+      "epoch": 0.1332982640715413,
+      "grad_norm": 1.17947518825531,
+      "learning_rate": 0.0001944640547885614,
+      "loss": 1.984,
+      "step": 1267
+    },
+    {
+      "epoch": 0.13340347185691742,
+      "grad_norm": 1.6048225164413452,
+      "learning_rate": 0.00019445286827889446,
+      "loss": 1.7243,
+      "step": 1268
+    },
+    {
+      "epoch": 0.13350867964229354,
+      "grad_norm": 0.9817343950271606,
+      "learning_rate": 0.00019444167080072698,
+      "loss": 1.7584,
+      "step": 1269
+    },
+    {
+      "epoch": 0.13361388742766964,
+      "grad_norm": 1.3782764673233032,
+      "learning_rate": 0.00019443046235535923,
+      "loss": 1.9988,
+      "step": 1270
+    },
+    {
+      "epoch": 0.13371909521304576,
+      "grad_norm": 1.0830388069152832,
+      "learning_rate": 0.00019441924294409289,
+      "loss": 2.0023,
+      "step": 1271
+    },
+    {
+      "epoch": 0.13382430299842188,
+      "grad_norm": 1.0523784160614014,
+      "learning_rate": 0.00019440801256823074,
+      "loss": 2.1114,
+      "step": 1272
+    },
+    {
+      "epoch": 0.133929510783798,
+      "grad_norm": 0.7290390729904175,
+      "learning_rate": 0.00019439677122907697,
+      "loss": 2.0477,
+      "step": 1273
+    },
+    {
+      "epoch": 0.13403471856917412,
+      "grad_norm": 1.1659435033798218,
+      "learning_rate": 0.00019438551892793701,
+      "loss": 2.0237,
+      "step": 1274
+    },
+    {
+      "epoch": 0.13413992635455024,
+      "grad_norm": 0.9051696062088013,
+      "learning_rate": 0.00019437425566611754,
+      "loss": 1.4642,
+      "step": 1275
+    },
+    {
+      "epoch": 0.13424513413992636,
+      "grad_norm": 0.9961209297180176,
+      "learning_rate": 0.0001943629814449265,
+      "loss": 2.1073,
+      "step": 1276
+    },
+    {
+      "epoch": 0.13435034192530249,
+      "grad_norm": 1.1717655658721924,
+      "learning_rate": 0.0001943516962656732,
+      "loss": 1.878,
+      "step": 1277
+    },
+    {
+      "epoch": 0.13445554971067858,
+      "grad_norm": 1.0735725164413452,
+      "learning_rate": 0.00019434040012966807,
+      "loss": 1.8939,
+      "step": 1278
+    },
+    {
+      "epoch": 0.1345607574960547,
+      "grad_norm": 1.1567238569259644,
+      "learning_rate": 0.00019432909303822296,
+      "loss": 2.2329,
+      "step": 1279
+    },
+    {
+      "epoch": 0.13466596528143082,
+      "grad_norm": 2.026787757873535,
+      "learning_rate": 0.00019431777499265087,
+      "loss": 1.5744,
+      "step": 1280
+    },
+    {
+      "epoch": 0.13477117306680694,
+      "grad_norm": 1.0909669399261475,
+      "learning_rate": 0.00019430644599426614,
+      "loss": 2.273,
+      "step": 1281
+    },
+    {
+      "epoch": 0.13487638085218306,
+      "grad_norm": 1.2396634817123413,
+      "learning_rate": 0.0001942951060443844,
+      "loss": 1.8932,
+      "step": 1282
+    },
+    {
+      "epoch": 0.13498158863755919,
+      "grad_norm": 1.163996696472168,
+      "learning_rate": 0.00019428375514432254,
+      "loss": 2.2442,
+      "step": 1283
+    },
+    {
+      "epoch": 0.1350867964229353,
+      "grad_norm": 0.9469679594039917,
+      "learning_rate": 0.0001942723932953986,
+      "loss": 2.4131,
+      "step": 1284
+    },
+    {
+      "epoch": 0.13519200420831143,
+      "grad_norm": 0.993421196937561,
+      "learning_rate": 0.00019426102049893208,
+      "loss": 2.3047,
+      "step": 1285
+    },
+    {
+      "epoch": 0.13529721199368752,
+      "grad_norm": 1.0937494039535522,
+      "learning_rate": 0.00019424963675624364,
+      "loss": 1.9778,
+      "step": 1286
+    },
+    {
+      "epoch": 0.13540241977906364,
+      "grad_norm": 1.0145161151885986,
+      "learning_rate": 0.00019423824206865527,
+      "loss": 2.3466,
+      "step": 1287
+    },
+    {
+      "epoch": 0.13550762756443976,
+      "grad_norm": 1.2493735551834106,
+      "learning_rate": 0.00019422683643749013,
+      "loss": 2.1595,
+      "step": 1288
+    },
+    {
+      "epoch": 0.13561283534981589,
+      "grad_norm": 0.7461625933647156,
+      "learning_rate": 0.00019421541986407276,
+      "loss": 1.9241,
+      "step": 1289
+    },
+    {
+      "epoch": 0.135718043135192,
+      "grad_norm": 1.2610676288604736,
+      "learning_rate": 0.00019420399234972894,
+      "loss": 1.9008,
+      "step": 1290
+    },
+    {
+      "epoch": 0.13582325092056813,
+      "grad_norm": 1.5543406009674072,
+      "learning_rate": 0.0001941925538957857,
+      "loss": 1.8748,
+      "step": 1291
+    },
+    {
+      "epoch": 0.13592845870594425,
+      "grad_norm": 0.6136517524719238,
+      "learning_rate": 0.00019418110450357135,
+      "loss": 1.6854,
+      "step": 1292
+    },
+    {
+      "epoch": 0.13603366649132037,
+      "grad_norm": 1.5480318069458008,
+      "learning_rate": 0.00019416964417441542,
+      "loss": 2.0514,
+      "step": 1293
+    },
+    {
+      "epoch": 0.13613887427669646,
+      "grad_norm": 0.9149646759033203,
+      "learning_rate": 0.00019415817290964883,
+      "loss": 2.1665,
+      "step": 1294
+    },
+    {
+      "epoch": 0.13624408206207259,
+      "grad_norm": 0.9369128346443176,
+      "learning_rate": 0.0001941466907106037,
+      "loss": 1.9612,
+      "step": 1295
+    },
+    {
+      "epoch": 0.1363492898474487,
+      "grad_norm": 0.8914728164672852,
+      "learning_rate": 0.0001941351975786134,
+      "loss": 1.5696,
+      "step": 1296
+    },
+    {
+      "epoch": 0.13645449763282483,
+      "grad_norm": 1.2152458429336548,
+      "learning_rate": 0.00019412369351501255,
+      "loss": 2.3433,
+      "step": 1297
+    },
+    {
+      "epoch": 0.13655970541820095,
+      "grad_norm": 0.6514436602592468,
+      "learning_rate": 0.0001941121785211371,
+      "loss": 2.3054,
+      "step": 1298
+    },
+    {
+      "epoch": 0.13666491320357707,
+      "grad_norm": 1.1356117725372314,
+      "learning_rate": 0.0001941006525983243,
+      "loss": 2.2101,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1367701209889532,
+      "grad_norm": 1.195465326309204,
+      "learning_rate": 0.00019408911574791255,
+      "loss": 1.6433,
+      "step": 1300
+    },
+    {
+      "epoch": 0.1368753287743293,
+      "grad_norm": 1.0879439115524292,
+      "learning_rate": 0.00019407756797124164,
+      "loss": 2.0753,
+      "step": 1301
+    },
+    {
+      "epoch": 0.1369805365597054,
+      "grad_norm": 1.4740591049194336,
+      "learning_rate": 0.00019406600926965255,
+      "loss": 2.1329,
+      "step": 1302
+    },
+    {
+      "epoch": 0.13708574434508153,
+      "grad_norm": 1.2994344234466553,
+      "learning_rate": 0.00019405443964448757,
+      "loss": 1.9924,
+      "step": 1303
+    },
+    {
+      "epoch": 0.13719095213045765,
+      "grad_norm": 1.0827962160110474,
+      "learning_rate": 0.0001940428590970902,
+      "loss": 2.1453,
+      "step": 1304
+    },
+    {
+      "epoch": 0.13729615991583377,
+      "grad_norm": 0.9276465177536011,
+      "learning_rate": 0.0001940312676288053,
+      "loss": 2.3533,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1374013677012099,
+      "grad_norm": 1.7329941987991333,
+      "learning_rate": 0.00019401966524097892,
+      "loss": 1.963,
+      "step": 1306
+    },
+    {
+      "epoch": 0.137506575486586,
+      "grad_norm": 1.348925232887268,
+      "learning_rate": 0.00019400805193495839,
+      "loss": 1.9953,
+      "step": 1307
+    },
+    {
+      "epoch": 0.13761178327196213,
+      "grad_norm": 1.512202262878418,
+      "learning_rate": 0.00019399642771209238,
+      "loss": 1.7091,
+      "step": 1308
+    },
+    {
+      "epoch": 0.13771699105733826,
+      "grad_norm": 1.0821994543075562,
+      "learning_rate": 0.00019398479257373073,
+      "loss": 2.0883,
+      "step": 1309
+    },
+    {
+      "epoch": 0.13782219884271435,
+      "grad_norm": 0.9924389719963074,
+      "learning_rate": 0.00019397314652122463,
+      "loss": 2.0062,
+      "step": 1310
+    },
+    {
+      "epoch": 0.13792740662809047,
+      "grad_norm": 1.1338258981704712,
+      "learning_rate": 0.00019396148955592643,
+      "loss": 2.1751,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1380326144134666,
+      "grad_norm": 1.3012595176696777,
+      "learning_rate": 0.00019394982167918987,
+      "loss": 1.8652,
+      "step": 1312
+    },
+    {
+      "epoch": 0.1381378221988427,
+      "grad_norm": 1.109009861946106,
+      "learning_rate": 0.0001939381428923699,
+      "loss": 2.3266,
+      "step": 1313
+    },
+    {
+      "epoch": 0.13824302998421883,
+      "grad_norm": 0.9063061475753784,
+      "learning_rate": 0.00019392645319682273,
+      "loss": 1.8569,
+      "step": 1314
+    },
+    {
+      "epoch": 0.13834823776959496,
+      "grad_norm": 0.9491289854049683,
+      "learning_rate": 0.00019391475259390584,
+      "loss": 2.031,
+      "step": 1315
+    },
+    {
+      "epoch": 0.13845344555497108,
+      "grad_norm": 1.2872332334518433,
+      "learning_rate": 0.00019390304108497794,
+      "loss": 2.0533,
+      "step": 1316
+    },
+    {
+      "epoch": 0.1385586533403472,
+      "grad_norm": 1.0027546882629395,
+      "learning_rate": 0.00019389131867139913,
+      "loss": 2.1953,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1386638611257233,
+      "grad_norm": 0.9697312712669373,
+      "learning_rate": 0.00019387958535453068,
+      "loss": 1.8958,
+      "step": 1318
+    },
+    {
+      "epoch": 0.1387690689110994,
+      "grad_norm": 0.9787667393684387,
+      "learning_rate": 0.00019386784113573508,
+      "loss": 2.074,
+      "step": 1319
+    },
+    {
+      "epoch": 0.13887427669647553,
+      "grad_norm": 1.1985487937927246,
+      "learning_rate": 0.00019385608601637624,
+      "loss": 2.2367,
+      "step": 1320
+    },
+    {
+      "epoch": 0.13897948448185166,
+      "grad_norm": 0.9592642188072205,
+      "learning_rate": 0.00019384431999781916,
+      "loss": 1.7508,
+      "step": 1321
+    },
+    {
+      "epoch": 0.13908469226722778,
+      "grad_norm": 2.198991060256958,
+      "learning_rate": 0.0001938325430814302,
+      "loss": 1.7532,
+      "step": 1322
+    },
+    {
+      "epoch": 0.1391899000526039,
+      "grad_norm": 1.6652568578720093,
+      "learning_rate": 0.00019382075526857705,
+      "loss": 1.5727,
+      "step": 1323
+    },
+    {
+      "epoch": 0.13929510783798002,
+      "grad_norm": 0.9190332889556885,
+      "learning_rate": 0.00019380895656062846,
+      "loss": 2.206,
+      "step": 1324
+    },
+    {
+      "epoch": 0.13940031562335614,
+      "grad_norm": 1.020340085029602,
+      "learning_rate": 0.00019379714695895472,
+      "loss": 2.0502,
+      "step": 1325
+    },
+    {
+      "epoch": 0.13950552340873223,
+      "grad_norm": 1.3590896129608154,
+      "learning_rate": 0.00019378532646492714,
+      "loss": 2.3099,
+      "step": 1326
+    },
+    {
+      "epoch": 0.13961073119410836,
+      "grad_norm": 0.7917930483818054,
+      "learning_rate": 0.00019377349507991842,
+      "loss": 2.2146,
+      "step": 1327
+    },
+    {
+      "epoch": 0.13971593897948448,
+      "grad_norm": 1.0664583444595337,
+      "learning_rate": 0.00019376165280530252,
+      "loss": 2.5567,
+      "step": 1328
+    },
+    {
+      "epoch": 0.1398211467648606,
+      "grad_norm": 0.7282881736755371,
+      "learning_rate": 0.00019374979964245463,
+      "loss": 1.9319,
+      "step": 1329
+    },
+    {
+      "epoch": 0.13992635455023672,
+      "grad_norm": 0.9543462991714478,
+      "learning_rate": 0.0001937379355927512,
+      "loss": 1.9968,
+      "step": 1330
+    },
+    {
+      "epoch": 0.14003156233561284,
+      "grad_norm": 0.8911160826683044,
+      "learning_rate": 0.00019372606065757003,
+      "loss": 1.9732,
+      "step": 1331
+    },
+    {
+      "epoch": 0.14013677012098896,
+      "grad_norm": 0.8454022407531738,
+      "learning_rate": 0.00019371417483829003,
+      "loss": 1.6059,
+      "step": 1332
+    },
+    {
+      "epoch": 0.14024197790636508,
+      "grad_norm": 0.9383593797683716,
+      "learning_rate": 0.00019370227813629147,
+      "loss": 2.2838,
+      "step": 1333
+    },
+    {
+      "epoch": 0.14034718569174118,
+      "grad_norm": 1.4767099618911743,
+      "learning_rate": 0.00019369037055295594,
+      "loss": 1.5007,
+      "step": 1334
+    },
+    {
+      "epoch": 0.1404523934771173,
+      "grad_norm": 1.4801223278045654,
+      "learning_rate": 0.00019367845208966618,
+      "loss": 2.0174,
+      "step": 1335
+    },
+    {
+      "epoch": 0.14055760126249342,
+      "grad_norm": 0.995274543762207,
+      "learning_rate": 0.00019366652274780628,
+      "loss": 2.0762,
+      "step": 1336
+    },
+    {
+      "epoch": 0.14066280904786954,
+      "grad_norm": 0.9528397917747498,
+      "learning_rate": 0.0001936545825287615,
+      "loss": 2.177,
+      "step": 1337
+    },
+    {
+      "epoch": 0.14076801683324566,
+      "grad_norm": 1.0369575023651123,
+      "learning_rate": 0.00019364263143391847,
+      "loss": 1.8945,
+      "step": 1338
+    },
+    {
+      "epoch": 0.14087322461862178,
+      "grad_norm": 0.8390545845031738,
+      "learning_rate": 0.00019363066946466502,
+      "loss": 1.9435,
+      "step": 1339
+    },
+    {
+      "epoch": 0.1409784324039979,
+      "grad_norm": 1.2658418416976929,
+      "learning_rate": 0.0001936186966223902,
+      "loss": 2.3394,
+      "step": 1340
+    },
+    {
+      "epoch": 0.14108364018937403,
+      "grad_norm": 0.9371610283851624,
+      "learning_rate": 0.00019360671290848447,
+      "loss": 2.1424,
+      "step": 1341
+    },
+    {
+      "epoch": 0.14118884797475012,
+      "grad_norm": 0.9509402513504028,
+      "learning_rate": 0.00019359471832433936,
+      "loss": 1.9365,
+      "step": 1342
+    },
+    {
+      "epoch": 0.14129405576012624,
+      "grad_norm": 1.1386808156967163,
+      "learning_rate": 0.00019358271287134784,
+      "loss": 1.7742,
+      "step": 1343
+    },
+    {
+      "epoch": 0.14139926354550236,
+      "grad_norm": 0.946628987789154,
+      "learning_rate": 0.00019357069655090404,
+      "loss": 2.1615,
+      "step": 1344
+    },
+    {
+      "epoch": 0.14150447133087848,
+      "grad_norm": 0.8672672510147095,
+      "learning_rate": 0.00019355866936440337,
+      "loss": 1.9622,
+      "step": 1345
+    },
+    {
+      "epoch": 0.1416096791162546,
+      "grad_norm": 1.2565221786499023,
+      "learning_rate": 0.0001935466313132425,
+      "loss": 1.6286,
+      "step": 1346
+    },
+    {
+      "epoch": 0.14171488690163073,
+      "grad_norm": 1.0216825008392334,
+      "learning_rate": 0.00019353458239881936,
+      "loss": 1.6907,
+      "step": 1347
+    },
+    {
+      "epoch": 0.14182009468700685,
+      "grad_norm": 1.3723381757736206,
+      "learning_rate": 0.00019352252262253318,
+      "loss": 1.64,
+      "step": 1348
+    },
+    {
+      "epoch": 0.14192530247238297,
+      "grad_norm": 1.0351169109344482,
+      "learning_rate": 0.00019351045198578445,
+      "loss": 2.2069,
+      "step": 1349
+    },
+    {
+      "epoch": 0.14203051025775906,
+      "grad_norm": 1.1500434875488281,
+      "learning_rate": 0.00019349837048997478,
+      "loss": 1.8967,
+      "step": 1350
+    },
+    {
+      "epoch": 0.14213571804313518,
+      "grad_norm": 1.0211580991744995,
+      "learning_rate": 0.00019348627813650727,
+      "loss": 1.6682,
+      "step": 1351
+    },
+    {
+      "epoch": 0.1422409258285113,
+      "grad_norm": 0.8861009478569031,
+      "learning_rate": 0.00019347417492678615,
+      "loss": 2.383,
+      "step": 1352
+    },
+    {
+      "epoch": 0.14234613361388743,
+      "grad_norm": 1.089370608329773,
+      "learning_rate": 0.00019346206086221686,
+      "loss": 2.0606,
+      "step": 1353
+    },
+    {
+      "epoch": 0.14245134139926355,
+      "grad_norm": 0.9320117235183716,
+      "learning_rate": 0.00019344993594420622,
+      "loss": 2.165,
+      "step": 1354
+    },
+    {
+      "epoch": 0.14255654918463967,
+      "grad_norm": 0.9916388988494873,
+      "learning_rate": 0.00019343780017416223,
+      "loss": 2.1013,
+      "step": 1355
+    },
+    {
+      "epoch": 0.1426617569700158,
+      "grad_norm": 0.894123375415802,
+      "learning_rate": 0.00019342565355349417,
+      "loss": 2.3444,
+      "step": 1356
+    },
+    {
+      "epoch": 0.1427669647553919,
+      "grad_norm": 0.835313081741333,
+      "learning_rate": 0.00019341349608361267,
+      "loss": 2.1591,
+      "step": 1357
+    },
+    {
+      "epoch": 0.142872172540768,
+      "grad_norm": 0.9938138723373413,
+      "learning_rate": 0.00019340132776592942,
+      "loss": 2.1381,
+      "step": 1358
+    },
+    {
+      "epoch": 0.14297738032614413,
+      "grad_norm": 1.408219337463379,
+      "learning_rate": 0.00019338914860185752,
+      "loss": 1.9765,
+      "step": 1359
+    },
+    {
+      "epoch": 0.14308258811152025,
+      "grad_norm": 1.1644525527954102,
+      "learning_rate": 0.00019337695859281137,
+      "loss": 1.788,
+      "step": 1360
+    },
+    {
+      "epoch": 0.14318779589689637,
+      "grad_norm": 0.6040117144584656,
+      "learning_rate": 0.00019336475774020648,
+      "loss": 2.1476,
+      "step": 1361
+    },
+    {
+      "epoch": 0.1432930036822725,
+      "grad_norm": 0.9746436476707458,
+      "learning_rate": 0.0001933525460454597,
+      "loss": 1.6193,
+      "step": 1362
+    },
+    {
+      "epoch": 0.1433982114676486,
+      "grad_norm": 1.6310737133026123,
+      "learning_rate": 0.00019334032350998919,
+      "loss": 2.009,
+      "step": 1363
+    },
+    {
+      "epoch": 0.14350341925302473,
+      "grad_norm": 0.9230230450630188,
+      "learning_rate": 0.00019332809013521428,
+      "loss": 2.1239,
+      "step": 1364
+    },
+    {
+      "epoch": 0.14360862703840085,
+      "grad_norm": 0.9308440685272217,
+      "learning_rate": 0.00019331584592255553,
+      "loss": 1.8617,
+      "step": 1365
+    },
+    {
+      "epoch": 0.14371383482377695,
+      "grad_norm": 1.1313167810440063,
+      "learning_rate": 0.0001933035908734349,
+      "loss": 1.4825,
+      "step": 1366
+    },
+    {
+      "epoch": 0.14381904260915307,
+      "grad_norm": 1.0187263488769531,
+      "learning_rate": 0.0001932913249892755,
+      "loss": 2.1542,
+      "step": 1367
+    },
+    {
+      "epoch": 0.1439242503945292,
+      "grad_norm": 1.1843492984771729,
+      "learning_rate": 0.00019327904827150176,
+      "loss": 1.8278,
+      "step": 1368
+    },
+    {
+      "epoch": 0.1440294581799053,
+      "grad_norm": 1.6638554334640503,
+      "learning_rate": 0.00019326676072153927,
+      "loss": 1.9298,
+      "step": 1369
+    },
+    {
+      "epoch": 0.14413466596528143,
+      "grad_norm": 0.9473015666007996,
+      "learning_rate": 0.00019325446234081498,
+      "loss": 2.2619,
+      "step": 1370
+    },
+    {
+      "epoch": 0.14423987375065755,
+      "grad_norm": 0.9866594672203064,
+      "learning_rate": 0.00019324215313075706,
+      "loss": 1.7889,
+      "step": 1371
+    },
+    {
+      "epoch": 0.14434508153603368,
+      "grad_norm": 1.3924474716186523,
+      "learning_rate": 0.00019322983309279495,
+      "loss": 1.8161,
+      "step": 1372
+    },
+    {
+      "epoch": 0.1444502893214098,
+      "grad_norm": 1.1654231548309326,
+      "learning_rate": 0.00019321750222835933,
+      "loss": 1.8105,
+      "step": 1373
+    },
+    {
+      "epoch": 0.1445554971067859,
+      "grad_norm": 0.7712641358375549,
+      "learning_rate": 0.0001932051605388821,
+      "loss": 1.7791,
+      "step": 1374
+    },
+    {
+      "epoch": 0.144660704892162,
+      "grad_norm": 0.7534984350204468,
+      "learning_rate": 0.00019319280802579654,
+      "loss": 1.9788,
+      "step": 1375
+    },
+    {
+      "epoch": 0.14476591267753813,
+      "grad_norm": 0.9810335040092468,
+      "learning_rate": 0.00019318044469053702,
+      "loss": 2.0382,
+      "step": 1376
+    },
+    {
+      "epoch": 0.14487112046291425,
+      "grad_norm": 1.1605186462402344,
+      "learning_rate": 0.0001931680705345393,
+      "loss": 2.2038,
+      "step": 1377
+    },
+    {
+      "epoch": 0.14497632824829038,
+      "grad_norm": 1.2045458555221558,
+      "learning_rate": 0.00019315568555924035,
+      "loss": 2.3353,
+      "step": 1378
+    },
+    {
+      "epoch": 0.1450815360336665,
+      "grad_norm": 1.1022682189941406,
+      "learning_rate": 0.0001931432897660784,
+      "loss": 1.8617,
+      "step": 1379
+    },
+    {
+      "epoch": 0.14518674381904262,
+      "grad_norm": 1.5427874326705933,
+      "learning_rate": 0.0001931308831564929,
+      "loss": 2.1449,
+      "step": 1380
+    },
+    {
+      "epoch": 0.14529195160441874,
+      "grad_norm": 1.1985043287277222,
+      "learning_rate": 0.00019311846573192461,
+      "loss": 2.0909,
+      "step": 1381
+    },
+    {
+      "epoch": 0.14539715938979483,
+      "grad_norm": 1.052314043045044,
+      "learning_rate": 0.00019310603749381558,
+      "loss": 1.9035,
+      "step": 1382
+    },
+    {
+      "epoch": 0.14550236717517095,
+      "grad_norm": 1.17412269115448,
+      "learning_rate": 0.00019309359844360893,
+      "loss": 1.6764,
+      "step": 1383
+    },
+    {
+      "epoch": 0.14560757496054708,
+      "grad_norm": 1.25741446018219,
+      "learning_rate": 0.00019308114858274932,
+      "loss": 2.1154,
+      "step": 1384
+    },
+    {
+      "epoch": 0.1457127827459232,
+      "grad_norm": 1.7540018558502197,
+      "learning_rate": 0.0001930686879126824,
+      "loss": 1.1782,
+      "step": 1385
+    },
+    {
+      "epoch": 0.14581799053129932,
+      "grad_norm": 1.423569917678833,
+      "learning_rate": 0.00019305621643485522,
+      "loss": 2.217,
+      "step": 1386
+    },
+    {
+      "epoch": 0.14592319831667544,
+      "grad_norm": 1.670623540878296,
+      "learning_rate": 0.00019304373415071605,
+      "loss": 2.1301,
+      "step": 1387
+    },
+    {
+      "epoch": 0.14602840610205156,
+      "grad_norm": 1.0802117586135864,
+      "learning_rate": 0.00019303124106171443,
+      "loss": 2.0065,
+      "step": 1388
+    },
+    {
+      "epoch": 0.14613361388742768,
+      "grad_norm": 0.7987265586853027,
+      "learning_rate": 0.00019301873716930107,
+      "loss": 2.3087,
+      "step": 1389
+    },
+    {
+      "epoch": 0.14623882167280378,
+      "grad_norm": 1.3220068216323853,
+      "learning_rate": 0.00019300622247492814,
+      "loss": 2.1964,
+      "step": 1390
+    },
+    {
+      "epoch": 0.1463440294581799,
+      "grad_norm": 1.0792020559310913,
+      "learning_rate": 0.00019299369698004884,
+      "loss": 2.0078,
+      "step": 1391
+    },
+    {
+      "epoch": 0.14644923724355602,
+      "grad_norm": 2.057413101196289,
+      "learning_rate": 0.0001929811606861177,
+      "loss": 1.783,
+      "step": 1392
+    },
+    {
+      "epoch": 0.14655444502893214,
+      "grad_norm": 1.4398462772369385,
+      "learning_rate": 0.0001929686135945906,
+      "loss": 1.3909,
+      "step": 1393
+    },
+    {
+      "epoch": 0.14665965281430826,
+      "grad_norm": 1.0750278234481812,
+      "learning_rate": 0.0001929560557069245,
+      "loss": 1.8938,
+      "step": 1394
+    },
+    {
+      "epoch": 0.14676486059968438,
+      "grad_norm": 0.8204072713851929,
+      "learning_rate": 0.00019294348702457773,
+      "loss": 2.4199,
+      "step": 1395
+    },
+    {
+      "epoch": 0.1468700683850605,
+      "grad_norm": 0.9489021897315979,
+      "learning_rate": 0.0001929309075490099,
+      "loss": 2.0696,
+      "step": 1396
+    },
+    {
+      "epoch": 0.14697527617043663,
+      "grad_norm": 0.8120673298835754,
+      "learning_rate": 0.00019291831728168182,
+      "loss": 2.0151,
+      "step": 1397
+    },
+    {
+      "epoch": 0.14708048395581272,
+      "grad_norm": 1.4392826557159424,
+      "learning_rate": 0.00019290571622405548,
+      "loss": 1.8832,
+      "step": 1398
+    },
+    {
+      "epoch": 0.14718569174118884,
+      "grad_norm": 1.1055705547332764,
+      "learning_rate": 0.00019289310437759427,
+      "loss": 1.829,
+      "step": 1399
+    },
+    {
+      "epoch": 0.14729089952656496,
+      "grad_norm": 0.5811572074890137,
+      "learning_rate": 0.00019288048174376273,
+      "loss": 2.2987,
+      "step": 1400
+    },
+    {
+      "epoch": 0.14739610731194108,
+      "grad_norm": 1.0770584344863892,
+      "learning_rate": 0.0001928678483240267,
+      "loss": 1.9941,
+      "step": 1401
+    },
+    {
+      "epoch": 0.1475013150973172,
+      "grad_norm": 1.227790117263794,
+      "learning_rate": 0.00019285520411985326,
+      "loss": 2.0154,
+      "step": 1402
+    },
+    {
+      "epoch": 0.14760652288269333,
+      "grad_norm": 0.6818186640739441,
+      "learning_rate": 0.0001928425491327107,
+      "loss": 2.2323,
+      "step": 1403
+    },
+    {
+      "epoch": 0.14771173066806945,
+      "grad_norm": 1.290170431137085,
+      "learning_rate": 0.00019282988336406865,
+      "loss": 2.1058,
+      "step": 1404
+    },
+    {
+      "epoch": 0.14781693845344557,
+      "grad_norm": 0.9615659713745117,
+      "learning_rate": 0.0001928172068153979,
+      "loss": 2.0933,
+      "step": 1405
+    },
+    {
+      "epoch": 0.14792214623882166,
+      "grad_norm": 0.9037755131721497,
+      "learning_rate": 0.00019280451948817059,
+      "loss": 2.0112,
+      "step": 1406
+    },
+    {
+      "epoch": 0.14802735402419778,
+      "grad_norm": 1.2614983320236206,
+      "learning_rate": 0.00019279182138386003,
+      "loss": 1.8352,
+      "step": 1407
+    },
+    {
+      "epoch": 0.1481325618095739,
+      "grad_norm": 1.1680060625076294,
+      "learning_rate": 0.0001927791125039408,
+      "loss": 2.0655,
+      "step": 1408
+    },
+    {
+      "epoch": 0.14823776959495003,
+      "grad_norm": 1.1708638668060303,
+      "learning_rate": 0.00019276639284988875,
+      "loss": 1.9568,
+      "step": 1409
+    },
+    {
+      "epoch": 0.14834297738032615,
+      "grad_norm": 1.2957854270935059,
+      "learning_rate": 0.00019275366242318097,
+      "loss": 1.9057,
+      "step": 1410
+    },
+    {
+      "epoch": 0.14844818516570227,
+      "grad_norm": 1.3663761615753174,
+      "learning_rate": 0.00019274092122529584,
+      "loss": 1.9964,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1485533929510784,
+      "grad_norm": 1.0850753784179688,
+      "learning_rate": 0.00019272816925771288,
+      "loss": 1.8218,
+      "step": 1412
+    },
+    {
+      "epoch": 0.1486586007364545,
+      "grad_norm": 0.7503268718719482,
+      "learning_rate": 0.00019271540652191296,
+      "loss": 2.2849,
+      "step": 1413
+    },
+    {
+      "epoch": 0.1487638085218306,
+      "grad_norm": 1.1095421314239502,
+      "learning_rate": 0.0001927026330193782,
+      "loss": 2.0907,
+      "step": 1414
+    },
+    {
+      "epoch": 0.14886901630720673,
+      "grad_norm": 1.947831630706787,
+      "learning_rate": 0.00019268984875159191,
+      "loss": 2.3621,
+      "step": 1415
+    },
+    {
+      "epoch": 0.14897422409258285,
+      "grad_norm": 1.045089602470398,
+      "learning_rate": 0.00019267705372003876,
+      "loss": 1.8236,
+      "step": 1416
+    },
+    {
+      "epoch": 0.14907943187795897,
+      "grad_norm": 1.137406826019287,
+      "learning_rate": 0.0001926642479262045,
+      "loss": 2.1149,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1491846396633351,
+      "grad_norm": 1.469742774963379,
+      "learning_rate": 0.00019265143137157627,
+      "loss": 1.8987,
+      "step": 1418
+    },
+    {
+      "epoch": 0.1492898474487112,
+      "grad_norm": 1.202427864074707,
+      "learning_rate": 0.00019263860405764241,
+      "loss": 1.8768,
+      "step": 1419
+    },
+    {
+      "epoch": 0.14939505523408733,
+      "grad_norm": 0.9847029447555542,
+      "learning_rate": 0.0001926257659858925,
+      "loss": 1.9719,
+      "step": 1420
+    },
+    {
+      "epoch": 0.14950026301946345,
+      "grad_norm": 1.1331363916397095,
+      "learning_rate": 0.00019261291715781743,
+      "loss": 2.2542,
+      "step": 1421
+    },
+    {
+      "epoch": 0.14960547080483955,
+      "grad_norm": 1.19657564163208,
+      "learning_rate": 0.00019260005757490922,
+      "loss": 2.2169,
+      "step": 1422
+    },
+    {
+      "epoch": 0.14971067859021567,
+      "grad_norm": 1.2917791604995728,
+      "learning_rate": 0.00019258718723866127,
+      "loss": 2.0383,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1498158863755918,
+      "grad_norm": 1.163900375366211,
+      "learning_rate": 0.00019257430615056816,
+      "loss": 1.9061,
+      "step": 1424
+    },
+    {
+      "epoch": 0.1499210941609679,
+      "grad_norm": 1.1655508279800415,
+      "learning_rate": 0.00019256141431212568,
+      "loss": 2.2014,
+      "step": 1425
+    },
+    {
+      "epoch": 0.15002630194634403,
+      "grad_norm": 0.8750263452529907,
+      "learning_rate": 0.00019254851172483098,
+      "loss": 2.0095,
+      "step": 1426
+    },
+    {
+      "epoch": 0.15013150973172015,
+      "grad_norm": 1.1399056911468506,
+      "learning_rate": 0.00019253559839018235,
+      "loss": 1.8426,
+      "step": 1427
+    },
+    {
+      "epoch": 0.15023671751709627,
+      "grad_norm": 1.3897678852081299,
+      "learning_rate": 0.00019252267430967942,
+      "loss": 1.3708,
+      "step": 1428
+    },
+    {
+      "epoch": 0.1503419253024724,
+      "grad_norm": 0.97458416223526,
+      "learning_rate": 0.00019250973948482298,
+      "loss": 1.6044,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1504471330878485,
+      "grad_norm": 1.1727815866470337,
+      "learning_rate": 0.0001924967939171151,
+      "loss": 2.0115,
+      "step": 1430
+    },
+    {
+      "epoch": 0.1505523408732246,
+      "grad_norm": 1.3556091785430908,
+      "learning_rate": 0.00019248383760805916,
+      "loss": 1.7871,
+      "step": 1431
+    },
+    {
+      "epoch": 0.15065754865860073,
+      "grad_norm": 0.9026738405227661,
+      "learning_rate": 0.00019247087055915968,
+      "loss": 1.9852,
+      "step": 1432
+    },
+    {
+      "epoch": 0.15076275644397685,
+      "grad_norm": 1.1884734630584717,
+      "learning_rate": 0.0001924578927719225,
+      "loss": 2.3177,
+      "step": 1433
+    },
+    {
+      "epoch": 0.15086796422935297,
+      "grad_norm": 1.2027161121368408,
+      "learning_rate": 0.00019244490424785468,
+      "loss": 1.6248,
+      "step": 1434
+    },
+    {
+      "epoch": 0.1509731720147291,
+      "grad_norm": 1.37662935256958,
+      "learning_rate": 0.00019243190498846458,
+      "loss": 2.0508,
+      "step": 1435
+    },
+    {
+      "epoch": 0.15107837980010522,
+      "grad_norm": 1.3299387693405151,
+      "learning_rate": 0.00019241889499526169,
+      "loss": 1.862,
+      "step": 1436
+    },
+    {
+      "epoch": 0.15118358758548134,
+      "grad_norm": 1.6885327100753784,
+      "learning_rate": 0.00019240587426975686,
+      "loss": 2.6639,
+      "step": 1437
+    },
+    {
+      "epoch": 0.15128879537085743,
+      "grad_norm": 1.0421466827392578,
+      "learning_rate": 0.00019239284281346214,
+      "loss": 1.8263,
+      "step": 1438
+    },
+    {
+      "epoch": 0.15139400315623355,
+      "grad_norm": 1.2554149627685547,
+      "learning_rate": 0.00019237980062789082,
+      "loss": 2.4153,
+      "step": 1439
+    },
+    {
+      "epoch": 0.15149921094160967,
+      "grad_norm": 0.9526044130325317,
+      "learning_rate": 0.00019236674771455747,
+      "loss": 2.236,
+      "step": 1440
+    },
+    {
+      "epoch": 0.1516044187269858,
+      "grad_norm": 0.9509533643722534,
+      "learning_rate": 0.00019235368407497788,
+      "loss": 1.966,
+      "step": 1441
+    },
+    {
+      "epoch": 0.15170962651236192,
+      "grad_norm": 1.2150377035140991,
+      "learning_rate": 0.00019234060971066902,
+      "loss": 1.6399,
+      "step": 1442
+    },
+    {
+      "epoch": 0.15181483429773804,
+      "grad_norm": 1.228989601135254,
+      "learning_rate": 0.00019232752462314923,
+      "loss": 2.1176,
+      "step": 1443
+    },
+    {
+      "epoch": 0.15192004208311416,
+      "grad_norm": 1.2654527425765991,
+      "learning_rate": 0.0001923144288139381,
+      "loss": 1.5625,
+      "step": 1444
+    },
+    {
+      "epoch": 0.15202524986849028,
+      "grad_norm": 1.086130976676941,
+      "learning_rate": 0.00019230132228455628,
+      "loss": 1.8363,
+      "step": 1445
+    },
+    {
+      "epoch": 0.15213045765386637,
+      "grad_norm": 1.0267728567123413,
+      "learning_rate": 0.00019228820503652586,
+      "loss": 1.7803,
+      "step": 1446
+    },
+    {
+      "epoch": 0.1522356654392425,
+      "grad_norm": 1.3312456607818604,
+      "learning_rate": 0.00019227507707137006,
+      "loss": 1.7026,
+      "step": 1447
+    },
+    {
+      "epoch": 0.15234087322461862,
+      "grad_norm": 1.434946060180664,
+      "learning_rate": 0.00019226193839061347,
+      "loss": 1.7181,
+      "step": 1448
+    },
+    {
+      "epoch": 0.15244608100999474,
+      "grad_norm": 2.0896921157836914,
+      "learning_rate": 0.00019224878899578175,
+      "loss": 2.235,
+      "step": 1449
+    },
+    {
+      "epoch": 0.15255128879537086,
+      "grad_norm": 0.7048740386962891,
+      "learning_rate": 0.00019223562888840193,
+      "loss": 2.3047,
+      "step": 1450
+    },
+    {
+      "epoch": 0.15265649658074698,
+      "grad_norm": 1.5920569896697998,
+      "learning_rate": 0.00019222245807000223,
+      "loss": 2.2711,
+      "step": 1451
+    },
+    {
+      "epoch": 0.1527617043661231,
+      "grad_norm": 0.8454103469848633,
+      "learning_rate": 0.00019220927654211217,
+      "loss": 1.8763,
+      "step": 1452
+    },
+    {
+      "epoch": 0.15286691215149922,
+      "grad_norm": 0.9767968654632568,
+      "learning_rate": 0.0001921960843062625,
+      "loss": 2.1151,
+      "step": 1453
+    },
+    {
+      "epoch": 0.15297211993687532,
+      "grad_norm": 1.5461039543151855,
+      "learning_rate": 0.00019218288136398513,
+      "loss": 1.5243,
+      "step": 1454
+    },
+    {
+      "epoch": 0.15307732772225144,
+      "grad_norm": 1.1929031610488892,
+      "learning_rate": 0.0001921696677168133,
+      "loss": 1.914,
+      "step": 1455
+    },
+    {
+      "epoch": 0.15318253550762756,
+      "grad_norm": 1.3064931631088257,
+      "learning_rate": 0.00019215644336628148,
+      "loss": 1.9663,
+      "step": 1456
+    },
+    {
+      "epoch": 0.15328774329300368,
+      "grad_norm": 1.21585214138031,
+      "learning_rate": 0.0001921432083139253,
+      "loss": 2.0572,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1533929510783798,
+      "grad_norm": 1.0481001138687134,
+      "learning_rate": 0.00019212996256128182,
+      "loss": 2.2602,
+      "step": 1458
+    },
+    {
+      "epoch": 0.15349815886375592,
+      "grad_norm": 1.2363125085830688,
+      "learning_rate": 0.00019211670610988913,
+      "loss": 1.9804,
+      "step": 1459
+    },
+    {
+      "epoch": 0.15360336664913204,
+      "grad_norm": 0.9035821557044983,
+      "learning_rate": 0.0001921034389612867,
+      "loss": 2.1159,
+      "step": 1460
+    },
+    {
+      "epoch": 0.15370857443450817,
+      "grad_norm": 1.1142661571502686,
+      "learning_rate": 0.00019209016111701522,
+      "loss": 2.3541,
+      "step": 1461
+    },
+    {
+      "epoch": 0.15381378221988426,
+      "grad_norm": 0.9691978096961975,
+      "learning_rate": 0.00019207687257861655,
+      "loss": 2.0898,
+      "step": 1462
+    },
+    {
+      "epoch": 0.15391899000526038,
+      "grad_norm": 1.0944209098815918,
+      "learning_rate": 0.00019206357334763388,
+      "loss": 1.7832,
+      "step": 1463
+    },
+    {
+      "epoch": 0.1540241977906365,
+      "grad_norm": 1.0283271074295044,
+      "learning_rate": 0.00019205026342561157,
+      "loss": 2.3259,
+      "step": 1464
+    },
+    {
+      "epoch": 0.15412940557601262,
+      "grad_norm": 0.9159666299819946,
+      "learning_rate": 0.0001920369428140953,
+      "loss": 2.1943,
+      "step": 1465
+    },
+    {
+      "epoch": 0.15423461336138874,
+      "grad_norm": 1.10454261302948,
+      "learning_rate": 0.00019202361151463194,
+      "loss": 1.7206,
+      "step": 1466
+    },
+    {
+      "epoch": 0.15433982114676487,
+      "grad_norm": 1.2136410474777222,
+      "learning_rate": 0.00019201026952876958,
+      "loss": 2.1148,
+      "step": 1467
+    },
+    {
+      "epoch": 0.154445028932141,
+      "grad_norm": 1.4272185564041138,
+      "learning_rate": 0.00019199691685805763,
+      "loss": 2.401,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1545502367175171,
+      "grad_norm": 1.2169204950332642,
+      "learning_rate": 0.00019198355350404667,
+      "loss": 2.039,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1546554445028932,
+      "grad_norm": 0.9736878275871277,
+      "learning_rate": 0.0001919701794682885,
+      "loss": 2.3821,
+      "step": 1470
+    },
+    {
+      "epoch": 0.15476065228826932,
+      "grad_norm": 1.0728427171707153,
+      "learning_rate": 0.00019195679475233625,
+      "loss": 2.079,
+      "step": 1471
+    },
+    {
+      "epoch": 0.15486586007364544,
+      "grad_norm": 0.9223616719245911,
+      "learning_rate": 0.00019194339935774422,
+      "loss": 1.9857,
+      "step": 1472
+    },
+    {
+      "epoch": 0.15497106785902157,
+      "grad_norm": 0.7518747448921204,
+      "learning_rate": 0.00019192999328606803,
+      "loss": 2.2304,
+      "step": 1473
+    },
+    {
+      "epoch": 0.1550762756443977,
+      "grad_norm": 1.1213191747665405,
+      "learning_rate": 0.0001919165765388644,
+      "loss": 2.1738,
+      "step": 1474
+    },
+    {
+      "epoch": 0.1551814834297738,
+      "grad_norm": 1.642525553703308,
+      "learning_rate": 0.00019190314911769142,
+      "loss": 1.8865,
+      "step": 1475
+    },
+    {
+      "epoch": 0.15528669121514993,
+      "grad_norm": 0.9955556392669678,
+      "learning_rate": 0.00019188971102410837,
+      "loss": 2.1058,
+      "step": 1476
+    },
+    {
+      "epoch": 0.15539189900052605,
+      "grad_norm": 0.9576829075813293,
+      "learning_rate": 0.00019187626225967576,
+      "loss": 1.9758,
+      "step": 1477
+    },
+    {
+      "epoch": 0.15549710678590214,
+      "grad_norm": 1.4288785457611084,
+      "learning_rate": 0.00019186280282595535,
+      "loss": 1.7091,
+      "step": 1478
+    },
+    {
+      "epoch": 0.15560231457127827,
+      "grad_norm": 1.115845799446106,
+      "learning_rate": 0.00019184933272451015,
+      "loss": 2.1405,
+      "step": 1479
+    },
+    {
+      "epoch": 0.1557075223566544,
+      "grad_norm": 1.4371943473815918,
+      "learning_rate": 0.0001918358519569044,
+      "loss": 2.0125,
+      "step": 1480
+    },
+    {
+      "epoch": 0.1558127301420305,
+      "grad_norm": 1.4342418909072876,
+      "learning_rate": 0.00019182236052470354,
+      "loss": 2.0421,
+      "step": 1481
+    },
+    {
+      "epoch": 0.15591793792740663,
+      "grad_norm": 1.1712485551834106,
+      "learning_rate": 0.00019180885842947436,
+      "loss": 2.1814,
+      "step": 1482
+    },
+    {
+      "epoch": 0.15602314571278275,
+      "grad_norm": 0.816021740436554,
+      "learning_rate": 0.00019179534567278475,
+      "loss": 2.2745,
+      "step": 1483
+    },
+    {
+      "epoch": 0.15612835349815887,
+      "grad_norm": 1.6961162090301514,
+      "learning_rate": 0.0001917818222562039,
+      "loss": 1.8956,
+      "step": 1484
+    },
+    {
+      "epoch": 0.156233561283535,
+      "grad_norm": 1.0247284173965454,
+      "learning_rate": 0.0001917682881813023,
+      "loss": 1.9351,
+      "step": 1485
+    },
+    {
+      "epoch": 0.1563387690689111,
+      "grad_norm": 0.9849500060081482,
+      "learning_rate": 0.00019175474344965157,
+      "loss": 1.6681,
+      "step": 1486
+    },
+    {
+      "epoch": 0.1564439768542872,
+      "grad_norm": 1.3697773218154907,
+      "learning_rate": 0.00019174118806282458,
+      "loss": 1.9333,
+      "step": 1487
+    },
+    {
+      "epoch": 0.15654918463966333,
+      "grad_norm": 1.421255111694336,
+      "learning_rate": 0.00019172762202239558,
+      "loss": 1.9458,
+      "step": 1488
+    },
+    {
+      "epoch": 0.15665439242503945,
+      "grad_norm": 1.1675089597702026,
+      "learning_rate": 0.00019171404532993986,
+      "loss": 2.0317,
+      "step": 1489
+    },
+    {
+      "epoch": 0.15675960021041557,
+      "grad_norm": 0.8886706233024597,
+      "learning_rate": 0.00019170045798703406,
+      "loss": 2.0854,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1568648079957917,
+      "grad_norm": 1.510884404182434,
+      "learning_rate": 0.00019168685999525607,
+      "loss": 2.1653,
+      "step": 1491
+    },
+    {
+      "epoch": 0.15697001578116782,
+      "grad_norm": 1.1406757831573486,
+      "learning_rate": 0.00019167325135618487,
+      "loss": 2.0393,
+      "step": 1492
+    },
+    {
+      "epoch": 0.15707522356654394,
+      "grad_norm": 0.9392661452293396,
+      "learning_rate": 0.00019165963207140095,
+      "loss": 2.1535,
+      "step": 1493
+    },
+    {
+      "epoch": 0.15718043135192003,
+      "grad_norm": 0.9734488725662231,
+      "learning_rate": 0.00019164600214248575,
+      "loss": 1.9538,
+      "step": 1494
+    },
+    {
+      "epoch": 0.15728563913729615,
+      "grad_norm": 1.0362365245819092,
+      "learning_rate": 0.0001916323615710221,
+      "loss": 1.6926,
+      "step": 1495
+    },
+    {
+      "epoch": 0.15739084692267227,
+      "grad_norm": 1.099932074546814,
+      "learning_rate": 0.00019161871035859403,
+      "loss": 1.9106,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1574960547080484,
+      "grad_norm": 1.4597340822219849,
+      "learning_rate": 0.0001916050485067868,
+      "loss": 2.1146,
+      "step": 1497
+    },
+    {
+      "epoch": 0.15760126249342452,
+      "grad_norm": 1.3075284957885742,
+      "learning_rate": 0.00019159137601718697,
+      "loss": 2.6296,
+      "step": 1498
+    },
+    {
+      "epoch": 0.15770647027880064,
+      "grad_norm": 1.0450359582901,
+      "learning_rate": 0.00019157769289138225,
+      "loss": 1.7802,
+      "step": 1499
+    },
+    {
+      "epoch": 0.15781167806417676,
+      "grad_norm": 1.9684633016586304,
+      "learning_rate": 0.0001915639991309616,
+      "loss": 2.1394,
+      "step": 1500
+    },
+    {
+      "epoch": 0.15791688584955288,
+      "grad_norm": 1.4822932481765747,
+      "learning_rate": 0.00019155029473751526,
+      "loss": 1.9256,
+      "step": 1501
+    },
+    {
+      "epoch": 0.15802209363492897,
+      "grad_norm": 0.9977061748504639,
+      "learning_rate": 0.00019153657971263463,
+      "loss": 1.7428,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1581273014203051,
+      "grad_norm": 1.2301387786865234,
+      "learning_rate": 0.00019152285405791243,
+      "loss": 1.8108,
+      "step": 1503
+    },
+    {
+      "epoch": 0.15823250920568122,
+      "grad_norm": 0.9384319186210632,
+      "learning_rate": 0.00019150911777494258,
+      "loss": 2.55,
+      "step": 1504
+    },
+    {
+      "epoch": 0.15833771699105734,
+      "grad_norm": 0.9354838728904724,
+      "learning_rate": 0.00019149537086532022,
+      "loss": 2.1337,
+      "step": 1505
+    },
+    {
+      "epoch": 0.15844292477643346,
+      "grad_norm": 1.6690521240234375,
+      "learning_rate": 0.0001914816133306417,
+      "loss": 2.1056,
+      "step": 1506
+    },
+    {
+      "epoch": 0.15854813256180958,
+      "grad_norm": 1.3110581636428833,
+      "learning_rate": 0.0001914678451725047,
+      "loss": 2.1933,
+      "step": 1507
+    },
+    {
+      "epoch": 0.1586533403471857,
+      "grad_norm": 1.4883490800857544,
+      "learning_rate": 0.000191454066392508,
+      "loss": 1.751,
+      "step": 1508
+    },
+    {
+      "epoch": 0.15875854813256182,
+      "grad_norm": 1.056282639503479,
+      "learning_rate": 0.00019144027699225172,
+      "loss": 1.4755,
+      "step": 1509
+    },
+    {
+      "epoch": 0.15886375591793792,
+      "grad_norm": 1.4437083005905151,
+      "learning_rate": 0.00019142647697333723,
+      "loss": 2.0506,
+      "step": 1510
+    },
+    {
+      "epoch": 0.15896896370331404,
+      "grad_norm": 1.2706918716430664,
+      "learning_rate": 0.00019141266633736697,
+      "loss": 2.2959,
+      "step": 1511
+    },
+    {
+      "epoch": 0.15907417148869016,
+      "grad_norm": 0.7875407934188843,
+      "learning_rate": 0.00019139884508594484,
+      "loss": 1.9302,
+      "step": 1512
+    },
+    {
+      "epoch": 0.15917937927406628,
+      "grad_norm": 1.6068692207336426,
+      "learning_rate": 0.00019138501322067577,
+      "loss": 2.0472,
+      "step": 1513
+    },
+    {
+      "epoch": 0.1592845870594424,
+      "grad_norm": 1.250916600227356,
+      "learning_rate": 0.00019137117074316602,
+      "loss": 2.3526,
+      "step": 1514
+    },
+    {
+      "epoch": 0.15938979484481852,
+      "grad_norm": 0.9924522638320923,
+      "learning_rate": 0.00019135731765502313,
+      "loss": 1.804,
+      "step": 1515
+    },
+    {
+      "epoch": 0.15949500263019464,
+      "grad_norm": 0.9752918481826782,
+      "learning_rate": 0.00019134345395785572,
+      "loss": 1.527,
+      "step": 1516
+    },
+    {
+      "epoch": 0.15960021041557076,
+      "grad_norm": 0.8332353830337524,
+      "learning_rate": 0.00019132957965327382,
+      "loss": 2.2386,
+      "step": 1517
+    },
+    {
+      "epoch": 0.15970541820094686,
+      "grad_norm": 1.0230636596679688,
+      "learning_rate": 0.0001913156947428886,
+      "loss": 1.935,
+      "step": 1518
+    },
+    {
+      "epoch": 0.15981062598632298,
+      "grad_norm": 1.3288902044296265,
+      "learning_rate": 0.00019130179922831241,
+      "loss": 2.178,
+      "step": 1519
+    },
+    {
+      "epoch": 0.1599158337716991,
+      "grad_norm": 1.1439303159713745,
+      "learning_rate": 0.00019128789311115892,
+      "loss": 2.2228,
+      "step": 1520
+    },
+    {
+      "epoch": 0.16002104155707522,
+      "grad_norm": 1.0876315832138062,
+      "learning_rate": 0.00019127397639304305,
+      "loss": 2.0773,
+      "step": 1521
+    },
+    {
+      "epoch": 0.16012624934245134,
+      "grad_norm": 1.0651696920394897,
+      "learning_rate": 0.00019126004907558085,
+      "loss": 2.0508,
+      "step": 1522
+    },
+    {
+      "epoch": 0.16023145712782746,
+      "grad_norm": 1.309206485748291,
+      "learning_rate": 0.00019124611116038963,
+      "loss": 2.4417,
+      "step": 1523
+    },
+    {
+      "epoch": 0.16033666491320359,
+      "grad_norm": 1.2913589477539062,
+      "learning_rate": 0.00019123216264908802,
+      "loss": 2.1339,
+      "step": 1524
+    },
+    {
+      "epoch": 0.1604418726985797,
+      "grad_norm": 1.7865016460418701,
+      "learning_rate": 0.00019121820354329577,
+      "loss": 2.0545,
+      "step": 1525
+    },
+    {
+      "epoch": 0.1605470804839558,
+      "grad_norm": 1.2007263898849487,
+      "learning_rate": 0.00019120423384463392,
+      "loss": 1.865,
+      "step": 1526
+    },
+    {
+      "epoch": 0.16065228826933192,
+      "grad_norm": 1.5868077278137207,
+      "learning_rate": 0.0001911902535547247,
+      "loss": 2.2881,
+      "step": 1527
+    },
+    {
+      "epoch": 0.16075749605470804,
+      "grad_norm": 0.9651376605033875,
+      "learning_rate": 0.00019117626267519162,
+      "loss": 1.9428,
+      "step": 1528
+    },
+    {
+      "epoch": 0.16086270384008416,
+      "grad_norm": 1.0492637157440186,
+      "learning_rate": 0.0001911622612076594,
+      "loss": 1.8641,
+      "step": 1529
+    },
+    {
+      "epoch": 0.16096791162546029,
+      "grad_norm": 1.4848203659057617,
+      "learning_rate": 0.000191148249153754,
+      "loss": 2.256,
+      "step": 1530
+    },
+    {
+      "epoch": 0.1610731194108364,
+      "grad_norm": 1.3082940578460693,
+      "learning_rate": 0.00019113422651510255,
+      "loss": 2.2692,
+      "step": 1531
+    },
+    {
+      "epoch": 0.16117832719621253,
+      "grad_norm": 0.7588707804679871,
+      "learning_rate": 0.00019112019329333346,
+      "loss": 2.3249,
+      "step": 1532
+    },
+    {
+      "epoch": 0.16128353498158865,
+      "grad_norm": 0.7588993906974792,
+      "learning_rate": 0.0001911061494900764,
+      "loss": 2.1544,
+      "step": 1533
+    },
+    {
+      "epoch": 0.16138874276696474,
+      "grad_norm": 1.9041491746902466,
+      "learning_rate": 0.00019109209510696217,
+      "loss": 1.9289,
+      "step": 1534
+    },
+    {
+      "epoch": 0.16149395055234086,
+      "grad_norm": 1.3282427787780762,
+      "learning_rate": 0.00019107803014562294,
+      "loss": 1.6547,
+      "step": 1535
+    },
+    {
+      "epoch": 0.16159915833771699,
+      "grad_norm": 1.084847092628479,
+      "learning_rate": 0.00019106395460769196,
+      "loss": 1.9799,
+      "step": 1536
+    },
+    {
+      "epoch": 0.1617043661230931,
+      "grad_norm": 1.2308772802352905,
+      "learning_rate": 0.0001910498684948038,
+      "loss": 1.7734,
+      "step": 1537
+    },
+    {
+      "epoch": 0.16180957390846923,
+      "grad_norm": 1.1046706438064575,
+      "learning_rate": 0.0001910357718085942,
+      "loss": 1.8926,
+      "step": 1538
+    },
+    {
+      "epoch": 0.16191478169384535,
+      "grad_norm": 0.9326602816581726,
+      "learning_rate": 0.00019102166455070024,
+      "loss": 1.679,
+      "step": 1539
+    },
+    {
+      "epoch": 0.16201998947922147,
+      "grad_norm": 2.735374927520752,
+      "learning_rate": 0.0001910075467227601,
+      "loss": 2.2685,
+      "step": 1540
+    },
+    {
+      "epoch": 0.1621251972645976,
+      "grad_norm": 0.7285892367362976,
+      "learning_rate": 0.00019099341832641323,
+      "loss": 2.1306,
+      "step": 1541
+    },
+    {
+      "epoch": 0.16223040504997369,
+      "grad_norm": 1.3543635606765747,
+      "learning_rate": 0.0001909792793633003,
+      "loss": 1.7381,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1623356128353498,
+      "grad_norm": 1.4372515678405762,
+      "learning_rate": 0.00019096512983506327,
+      "loss": 1.3866,
+      "step": 1543
+    },
+    {
+      "epoch": 0.16244082062072593,
+      "grad_norm": 1.202860951423645,
+      "learning_rate": 0.00019095096974334523,
+      "loss": 1.9439,
+      "step": 1544
+    },
+    {
+      "epoch": 0.16254602840610205,
+      "grad_norm": 1.5187021493911743,
+      "learning_rate": 0.0001909367990897906,
+      "loss": 1.6743,
+      "step": 1545
+    },
+    {
+      "epoch": 0.16265123619147817,
+      "grad_norm": 0.9257557392120361,
+      "learning_rate": 0.00019092261787604492,
+      "loss": 2.1086,
+      "step": 1546
+    },
+    {
+      "epoch": 0.1627564439768543,
+      "grad_norm": 1.202397108078003,
+      "learning_rate": 0.00019090842610375503,
+      "loss": 2.1001,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1628616517622304,
+      "grad_norm": 1.3095595836639404,
+      "learning_rate": 0.000190894223774569,
+      "loss": 2.525,
+      "step": 1548
+    },
+    {
+      "epoch": 0.16296685954760654,
+      "grad_norm": 1.2810163497924805,
+      "learning_rate": 0.00019088001089013603,
+      "loss": 2.0124,
+      "step": 1549
+    },
+    {
+      "epoch": 0.16307206733298263,
+      "grad_norm": 2.0133657455444336,
+      "learning_rate": 0.00019086578745210666,
+      "loss": 1.9361,
+      "step": 1550
+    },
+    {
+      "epoch": 0.16317727511835875,
+      "grad_norm": 1.1870371103286743,
+      "learning_rate": 0.00019085155346213264,
+      "loss": 1.7296,
+      "step": 1551
+    },
+    {
+      "epoch": 0.16328248290373487,
+      "grad_norm": 1.0325980186462402,
+      "learning_rate": 0.00019083730892186686,
+      "loss": 2.4486,
+      "step": 1552
+    },
+    {
+      "epoch": 0.163387690689111,
+      "grad_norm": 1.313650131225586,
+      "learning_rate": 0.00019082305383296352,
+      "loss": 2.4732,
+      "step": 1553
+    },
+    {
+      "epoch": 0.16349289847448711,
+      "grad_norm": 0.8540735840797424,
+      "learning_rate": 0.00019080878819707802,
+      "loss": 1.8071,
+      "step": 1554
+    },
+    {
+      "epoch": 0.16359810625986324,
+      "grad_norm": 1.3443700075149536,
+      "learning_rate": 0.00019079451201586695,
+      "loss": 2.3741,
+      "step": 1555
+    },
+    {
+      "epoch": 0.16370331404523936,
+      "grad_norm": 1.2170543670654297,
+      "learning_rate": 0.0001907802252909882,
+      "loss": 2.0448,
+      "step": 1556
+    },
+    {
+      "epoch": 0.16380852183061548,
+      "grad_norm": 0.8038926720619202,
+      "learning_rate": 0.0001907659280241008,
+      "loss": 2.1864,
+      "step": 1557
+    },
+    {
+      "epoch": 0.16391372961599157,
+      "grad_norm": 1.2641410827636719,
+      "learning_rate": 0.00019075162021686505,
+      "loss": 1.7906,
+      "step": 1558
+    },
+    {
+      "epoch": 0.1640189374013677,
+      "grad_norm": 0.9859477877616882,
+      "learning_rate": 0.0001907373018709425,
+      "loss": 2.3066,
+      "step": 1559
+    },
+    {
+      "epoch": 0.16412414518674381,
+      "grad_norm": 1.0549753904342651,
+      "learning_rate": 0.00019072297298799589,
+      "loss": 2.0271,
+      "step": 1560
+    },
+    {
+      "epoch": 0.16422935297211994,
+      "grad_norm": 1.015430212020874,
+      "learning_rate": 0.0001907086335696892,
+      "loss": 2.1745,
+      "step": 1561
+    },
+    {
+      "epoch": 0.16433456075749606,
+      "grad_norm": 0.9478970766067505,
+      "learning_rate": 0.00019069428361768754,
+      "loss": 2.1743,
+      "step": 1562
+    },
+    {
+      "epoch": 0.16443976854287218,
+      "grad_norm": 0.8643477559089661,
+      "learning_rate": 0.00019067992313365735,
+      "loss": 2.4163,
+      "step": 1563
+    },
+    {
+      "epoch": 0.1645449763282483,
+      "grad_norm": 1.1437067985534668,
+      "learning_rate": 0.00019066555211926634,
+      "loss": 2.0313,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16465018411362442,
+      "grad_norm": 1.8715989589691162,
+      "learning_rate": 0.00019065117057618332,
+      "loss": 2.3086,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16475539189900051,
+      "grad_norm": 1.1657463312149048,
+      "learning_rate": 0.00019063677850607834,
+      "loss": 2.4182,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16486059968437664,
+      "grad_norm": 1.0688217878341675,
+      "learning_rate": 0.00019062237591062272,
+      "loss": 2.3177,
+      "step": 1567
+    },
+    {
+      "epoch": 0.16496580746975276,
+      "grad_norm": 1.5269672870635986,
+      "learning_rate": 0.000190607962791489,
+      "loss": 1.4833,
+      "step": 1568
+    },
+    {
+      "epoch": 0.16507101525512888,
+      "grad_norm": 0.9268459677696228,
+      "learning_rate": 0.00019059353915035096,
+      "loss": 2.1505,
+      "step": 1569
+    },
+    {
+      "epoch": 0.165176223040505,
+      "grad_norm": 1.0593661069869995,
+      "learning_rate": 0.00019057910498888352,
+      "loss": 2.1309,
+      "step": 1570
+    },
+    {
+      "epoch": 0.16528143082588112,
+      "grad_norm": 1.552679181098938,
+      "learning_rate": 0.00019056466030876288,
+      "loss": 1.8734,
+      "step": 1571
+    },
+    {
+      "epoch": 0.16538663861125724,
+      "grad_norm": 1.9197827577590942,
+      "learning_rate": 0.00019055020511166647,
+      "loss": 1.7033,
+      "step": 1572
+    },
+    {
+      "epoch": 0.16549184639663336,
+      "grad_norm": 1.1575535535812378,
+      "learning_rate": 0.0001905357393992729,
+      "loss": 2.3521,
+      "step": 1573
+    },
+    {
+      "epoch": 0.16559705418200946,
+      "grad_norm": 1.0774550437927246,
+      "learning_rate": 0.00019052126317326207,
+      "loss": 2.1065,
+      "step": 1574
+    },
+    {
+      "epoch": 0.16570226196738558,
+      "grad_norm": 1.0978038311004639,
+      "learning_rate": 0.000190506776435315,
+      "loss": 1.7549,
+      "step": 1575
+    },
+    {
+      "epoch": 0.1658074697527617,
+      "grad_norm": 0.8305821418762207,
+      "learning_rate": 0.00019049227918711402,
+      "loss": 2.5738,
+      "step": 1576
+    },
+    {
+      "epoch": 0.16591267753813782,
+      "grad_norm": 1.5406334400177002,
+      "learning_rate": 0.00019047777143034266,
+      "loss": 1.5855,
+      "step": 1577
+    },
+    {
+      "epoch": 0.16601788532351394,
+      "grad_norm": 1.2726386785507202,
+      "learning_rate": 0.00019046325316668562,
+      "loss": 1.9179,
+      "step": 1578
+    },
+    {
+      "epoch": 0.16612309310889006,
+      "grad_norm": 0.9605535864830017,
+      "learning_rate": 0.0001904487243978289,
+      "loss": 2.2189,
+      "step": 1579
+    },
+    {
+      "epoch": 0.16622830089426618,
+      "grad_norm": 1.0368754863739014,
+      "learning_rate": 0.00019043418512545963,
+      "loss": 2.1578,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1663335086796423,
+      "grad_norm": 0.9062163829803467,
+      "learning_rate": 0.00019041963535126625,
+      "loss": 2.2097,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1664387164650184,
+      "grad_norm": 1.6533317565917969,
+      "learning_rate": 0.00019040507507693836,
+      "loss": 2.1455,
+      "step": 1582
+    },
+    {
+      "epoch": 0.16654392425039452,
+      "grad_norm": 0.9693316221237183,
+      "learning_rate": 0.0001903905043041668,
+      "loss": 2.1971,
+      "step": 1583
+    },
+    {
+      "epoch": 0.16664913203577064,
+      "grad_norm": 0.8169092535972595,
+      "learning_rate": 0.00019037592303464362,
+      "loss": 2.4642,
+      "step": 1584
+    },
+    {
+      "epoch": 0.16675433982114676,
+      "grad_norm": 1.0367541313171387,
+      "learning_rate": 0.0001903613312700621,
+      "loss": 2.2354,
+      "step": 1585
+    },
+    {
+      "epoch": 0.16685954760652288,
+      "grad_norm": 1.3058645725250244,
+      "learning_rate": 0.00019034672901211672,
+      "loss": 1.8324,
+      "step": 1586
+    },
+    {
+      "epoch": 0.166964755391899,
+      "grad_norm": 1.1962265968322754,
+      "learning_rate": 0.0001903321162625032,
+      "loss": 2.0343,
+      "step": 1587
+    },
+    {
+      "epoch": 0.16706996317727513,
+      "grad_norm": 1.3322851657867432,
+      "learning_rate": 0.0001903174930229185,
+      "loss": 2.0793,
+      "step": 1588
+    },
+    {
+      "epoch": 0.16717517096265125,
+      "grad_norm": 0.9945839643478394,
+      "learning_rate": 0.00019030285929506075,
+      "loss": 1.8876,
+      "step": 1589
+    },
+    {
+      "epoch": 0.16728037874802734,
+      "grad_norm": 1.0389971733093262,
+      "learning_rate": 0.0001902882150806293,
+      "loss": 2.1234,
+      "step": 1590
+    },
+    {
+      "epoch": 0.16738558653340346,
+      "grad_norm": 1.1944520473480225,
+      "learning_rate": 0.00019027356038132473,
+      "loss": 2.3191,
+      "step": 1591
+    },
+    {
+      "epoch": 0.16749079431877958,
+      "grad_norm": 1.4879711866378784,
+      "learning_rate": 0.00019025889519884887,
+      "loss": 1.7512,
+      "step": 1592
+    },
+    {
+      "epoch": 0.1675960021041557,
+      "grad_norm": 0.7883063554763794,
+      "learning_rate": 0.00019024421953490472,
+      "loss": 2.2303,
+      "step": 1593
+    },
+    {
+      "epoch": 0.16770120988953183,
+      "grad_norm": 1.214263677597046,
+      "learning_rate": 0.00019022953339119654,
+      "loss": 1.8914,
+      "step": 1594
+    },
+    {
+      "epoch": 0.16780641767490795,
+      "grad_norm": 1.0355852842330933,
+      "learning_rate": 0.00019021483676942973,
+      "loss": 2.1548,
+      "step": 1595
+    },
+    {
+      "epoch": 0.16791162546028407,
+      "grad_norm": 1.011046290397644,
+      "learning_rate": 0.00019020012967131106,
+      "loss": 1.8183,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1680168332456602,
+      "grad_norm": 1.062743067741394,
+      "learning_rate": 0.0001901854120985483,
+      "loss": 1.9263,
+      "step": 1597
+    },
+    {
+      "epoch": 0.16812204103103628,
+      "grad_norm": 0.8611815571784973,
+      "learning_rate": 0.00019017068405285058,
+      "loss": 2.2722,
+      "step": 1598
+    },
+    {
+      "epoch": 0.1682272488164124,
+      "grad_norm": 1.0582350492477417,
+      "learning_rate": 0.0001901559455359283,
+      "loss": 2.1025,
+      "step": 1599
+    },
+    {
+      "epoch": 0.16833245660178853,
+      "grad_norm": 0.7138919234275818,
+      "learning_rate": 0.00019014119654949294,
+      "loss": 1.8953,
+      "step": 1600
+    },
+    {
+      "epoch": 0.16843766438716465,
+      "grad_norm": 1.2566789388656616,
+      "learning_rate": 0.00019012643709525722,
+      "loss": 2.494,
+      "step": 1601
+    },
+    {
+      "epoch": 0.16854287217254077,
+      "grad_norm": 2.009523391723633,
+      "learning_rate": 0.00019011166717493517,
+      "loss": 1.0732,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1686480799579169,
+      "grad_norm": 1.5301228761672974,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 1.6838,
+      "step": 1603
+    },
+    {
+      "epoch": 0.168753287743293,
+      "grad_norm": 1.3879108428955078,
+      "learning_rate": 0.0001900820959428939,
+      "loss": 2.1531,
+      "step": 1604
+    },
+    {
+      "epoch": 0.16885849552866913,
+      "grad_norm": 1.2042590379714966,
+      "learning_rate": 0.0001900672946346087,
+      "loss": 2.2596,
+      "step": 1605
+    },
+    {
+      "epoch": 0.16896370331404523,
+      "grad_norm": 1.7592346668243408,
+      "learning_rate": 0.00019005248286710518,
+      "loss": 2.145,
+      "step": 1606
+    },
+    {
+      "epoch": 0.16906891109942135,
+      "grad_norm": 0.7835527658462524,
+      "learning_rate": 0.00019003766064210336,
+      "loss": 2.4961,
+      "step": 1607
+    },
+    {
+      "epoch": 0.16917411888479747,
+      "grad_norm": 1.1461858749389648,
+      "learning_rate": 0.00019002282796132448,
+      "loss": 1.8269,
+      "step": 1608
+    },
+    {
+      "epoch": 0.1692793266701736,
+      "grad_norm": 1.6687688827514648,
+      "learning_rate": 0.000190007984826491,
+      "loss": 1.8541,
+      "step": 1609
+    },
+    {
+      "epoch": 0.1693845344555497,
+      "grad_norm": 0.8094751238822937,
+      "learning_rate": 0.0001899931312393267,
+      "loss": 2.0899,
+      "step": 1610
+    },
+    {
+      "epoch": 0.16948974224092583,
+      "grad_norm": 1.5134220123291016,
+      "learning_rate": 0.00018997826720155636,
+      "loss": 2.0814,
+      "step": 1611
+    },
+    {
+      "epoch": 0.16959495002630195,
+      "grad_norm": 1.034388542175293,
+      "learning_rate": 0.00018996339271490616,
+      "loss": 1.7283,
+      "step": 1612
+    },
+    {
+      "epoch": 0.16970015781167808,
+      "grad_norm": 0.7677897214889526,
+      "learning_rate": 0.0001899485077811034,
+      "loss": 2.0904,
+      "step": 1613
+    },
+    {
+      "epoch": 0.16980536559705417,
+      "grad_norm": 1.4724828004837036,
+      "learning_rate": 0.00018993361240187665,
+      "loss": 1.9291,
+      "step": 1614
+    },
+    {
+      "epoch": 0.1699105733824303,
+      "grad_norm": 1.181376338005066,
+      "learning_rate": 0.00018991870657895558,
+      "loss": 1.8633,
+      "step": 1615
+    },
+    {
+      "epoch": 0.1700157811678064,
+      "grad_norm": 1.2235981225967407,
+      "learning_rate": 0.00018990379031407124,
+      "loss": 1.8902,
+      "step": 1616
+    },
+    {
+      "epoch": 0.17012098895318253,
+      "grad_norm": 1.796578288078308,
+      "learning_rate": 0.0001898888636089558,
+      "loss": 1.7133,
+      "step": 1617
+    },
+    {
+      "epoch": 0.17022619673855865,
+      "grad_norm": 0.8562111854553223,
+      "learning_rate": 0.00018987392646534258,
+      "loss": 1.918,
+      "step": 1618
+    },
+    {
+      "epoch": 0.17033140452393478,
+      "grad_norm": 1.1729540824890137,
+      "learning_rate": 0.00018985897888496627,
+      "loss": 2.0491,
+      "step": 1619
+    },
+    {
+      "epoch": 0.1704366123093109,
+      "grad_norm": 1.2273286581039429,
+      "learning_rate": 0.0001898440208695626,
+      "loss": 1.9927,
+      "step": 1620
+    },
+    {
+      "epoch": 0.17054182009468702,
+      "grad_norm": 1.4244452714920044,
+      "learning_rate": 0.00018982905242086867,
+      "loss": 1.7184,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1706470278800631,
+      "grad_norm": 1.2891523838043213,
+      "learning_rate": 0.00018981407354062268,
+      "loss": 2.1326,
+      "step": 1622
+    },
+    {
+      "epoch": 0.17075223566543923,
+      "grad_norm": 1.7193819284439087,
+      "learning_rate": 0.00018979908423056408,
+      "loss": 2.0726,
+      "step": 1623
+    },
+    {
+      "epoch": 0.17085744345081536,
+      "grad_norm": 1.0913747549057007,
+      "learning_rate": 0.00018978408449243353,
+      "loss": 2.1798,
+      "step": 1624
+    },
+    {
+      "epoch": 0.17096265123619148,
+      "grad_norm": 1.2384060621261597,
+      "learning_rate": 0.00018976907432797287,
+      "loss": 2.2418,
+      "step": 1625
+    },
+    {
+      "epoch": 0.1710678590215676,
+      "grad_norm": 1.3474977016448975,
+      "learning_rate": 0.00018975405373892524,
+      "loss": 2.2404,
+      "step": 1626
+    },
+    {
+      "epoch": 0.17117306680694372,
+      "grad_norm": 1.227752685546875,
+      "learning_rate": 0.0001897390227270349,
+      "loss": 2.2127,
+      "step": 1627
+    },
+    {
+      "epoch": 0.17127827459231984,
+      "grad_norm": 1.1790754795074463,
+      "learning_rate": 0.00018972398129404736,
+      "loss": 2.3705,
+      "step": 1628
+    },
+    {
+      "epoch": 0.17138348237769596,
+      "grad_norm": 1.323519229888916,
+      "learning_rate": 0.00018970892944170933,
+      "loss": 1.9858,
+      "step": 1629
+    },
+    {
+      "epoch": 0.17148869016307206,
+      "grad_norm": 0.9499973058700562,
+      "learning_rate": 0.0001896938671717687,
+      "loss": 1.9775,
+      "step": 1630
+    },
+    {
+      "epoch": 0.17159389794844818,
+      "grad_norm": 0.8820815086364746,
+      "learning_rate": 0.00018967879448597463,
+      "loss": 2.2519,
+      "step": 1631
+    },
+    {
+      "epoch": 0.1716991057338243,
+      "grad_norm": 1.1254478693008423,
+      "learning_rate": 0.00018966371138607748,
+      "loss": 2.1188,
+      "step": 1632
+    },
+    {
+      "epoch": 0.17180431351920042,
+      "grad_norm": 0.9858778715133667,
+      "learning_rate": 0.00018964861787382876,
+      "loss": 2.0559,
+      "step": 1633
+    },
+    {
+      "epoch": 0.17190952130457654,
+      "grad_norm": 0.9913214445114136,
+      "learning_rate": 0.0001896335139509813,
+      "loss": 2.1155,
+      "step": 1634
+    },
+    {
+      "epoch": 0.17201472908995266,
+      "grad_norm": 0.8039067983627319,
+      "learning_rate": 0.00018961839961928898,
+      "loss": 1.7209,
+      "step": 1635
+    },
+    {
+      "epoch": 0.17211993687532878,
+      "grad_norm": 1.1919358968734741,
+      "learning_rate": 0.00018960327488050705,
+      "loss": 1.7773,
+      "step": 1636
+    },
+    {
+      "epoch": 0.1722251446607049,
+      "grad_norm": 1.3590924739837646,
+      "learning_rate": 0.00018958813973639184,
+      "loss": 1.134,
+      "step": 1637
+    },
+    {
+      "epoch": 0.172330352446081,
+      "grad_norm": 0.8363476991653442,
+      "learning_rate": 0.00018957299418870095,
+      "loss": 2.165,
+      "step": 1638
+    },
+    {
+      "epoch": 0.17243556023145712,
+      "grad_norm": 0.8952641487121582,
+      "learning_rate": 0.00018955783823919325,
+      "loss": 2.4521,
+      "step": 1639
+    },
+    {
+      "epoch": 0.17254076801683324,
+      "grad_norm": 1.4345828294754028,
+      "learning_rate": 0.0001895426718896287,
+      "loss": 1.983,
+      "step": 1640
+    },
+    {
+      "epoch": 0.17264597580220936,
+      "grad_norm": 1.2068872451782227,
+      "learning_rate": 0.00018952749514176848,
+      "loss": 2.0875,
+      "step": 1641
+    },
+    {
+      "epoch": 0.17275118358758548,
+      "grad_norm": 1.1335731744766235,
+      "learning_rate": 0.00018951230799737508,
+      "loss": 2.2086,
+      "step": 1642
+    },
+    {
+      "epoch": 0.1728563913729616,
+      "grad_norm": 1.1118179559707642,
+      "learning_rate": 0.0001894971104582121,
+      "loss": 2.2505,
+      "step": 1643
+    },
+    {
+      "epoch": 0.17296159915833773,
+      "grad_norm": 1.528136968612671,
+      "learning_rate": 0.0001894819025260444,
+      "loss": 1.9484,
+      "step": 1644
+    },
+    {
+      "epoch": 0.17306680694371385,
+      "grad_norm": 1.4697853326797485,
+      "learning_rate": 0.00018946668420263802,
+      "loss": 2.1765,
+      "step": 1645
+    },
+    {
+      "epoch": 0.17317201472908994,
+      "grad_norm": 1.0284318923950195,
+      "learning_rate": 0.0001894514554897602,
+      "loss": 2.173,
+      "step": 1646
+    },
+    {
+      "epoch": 0.17327722251446606,
+      "grad_norm": 0.9980498552322388,
+      "learning_rate": 0.0001894362163891794,
+      "loss": 2.1735,
+      "step": 1647
+    },
+    {
+      "epoch": 0.17338243029984218,
+      "grad_norm": 1.4780910015106201,
+      "learning_rate": 0.00018942096690266534,
+      "loss": 1.8087,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1734876380852183,
+      "grad_norm": 0.8312754034996033,
+      "learning_rate": 0.00018940570703198878,
+      "loss": 2.2845,
+      "step": 1649
+    },
+    {
+      "epoch": 0.17359284587059443,
+      "grad_norm": 0.8467996120452881,
+      "learning_rate": 0.00018939043677892192,
+      "loss": 2.1111,
+      "step": 1650
+    },
+    {
+      "epoch": 0.17369805365597055,
+      "grad_norm": 1.1280200481414795,
+      "learning_rate": 0.00018937515614523797,
+      "loss": 1.7534,
+      "step": 1651
+    },
+    {
+      "epoch": 0.17380326144134667,
+      "grad_norm": 1.277297019958496,
+      "learning_rate": 0.00018935986513271146,
+      "loss": 2.1373,
+      "step": 1652
+    },
+    {
+      "epoch": 0.1739084692267228,
+      "grad_norm": 0.9894921779632568,
+      "learning_rate": 0.00018934456374311806,
+      "loss": 1.9866,
+      "step": 1653
+    },
+    {
+      "epoch": 0.17401367701209888,
+      "grad_norm": 1.4973275661468506,
+      "learning_rate": 0.00018932925197823468,
+      "loss": 1.9207,
+      "step": 1654
+    },
+    {
+      "epoch": 0.174118884797475,
+      "grad_norm": 0.9163431525230408,
+      "learning_rate": 0.0001893139298398394,
+      "loss": 2.214,
+      "step": 1655
+    },
+    {
+      "epoch": 0.17422409258285113,
+      "grad_norm": 1.1179120540618896,
+      "learning_rate": 0.00018929859732971162,
+      "loss": 2.3422,
+      "step": 1656
+    },
+    {
+      "epoch": 0.17432930036822725,
+      "grad_norm": 1.4268755912780762,
+      "learning_rate": 0.00018928325444963172,
+      "loss": 1.8902,
+      "step": 1657
+    },
+    {
+      "epoch": 0.17443450815360337,
+      "grad_norm": 0.9193028211593628,
+      "learning_rate": 0.0001892679012013815,
+      "loss": 2.2039,
+      "step": 1658
+    },
+    {
+      "epoch": 0.1745397159389795,
+      "grad_norm": 0.7367107272148132,
+      "learning_rate": 0.00018925253758674386,
+      "loss": 2.1194,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1746449237243556,
+      "grad_norm": 1.8018954992294312,
+      "learning_rate": 0.00018923716360750293,
+      "loss": 2.2415,
+      "step": 1660
+    },
+    {
+      "epoch": 0.17475013150973173,
+      "grad_norm": 1.6338098049163818,
+      "learning_rate": 0.00018922177926544405,
+      "loss": 1.9905,
+      "step": 1661
+    },
+    {
+      "epoch": 0.17485533929510783,
+      "grad_norm": 1.4636485576629639,
+      "learning_rate": 0.00018920638456235375,
+      "loss": 2.3444,
+      "step": 1662
+    },
+    {
+      "epoch": 0.17496054708048395,
+      "grad_norm": 0.698484480381012,
+      "learning_rate": 0.00018919097950001977,
+      "loss": 1.9483,
+      "step": 1663
+    },
+    {
+      "epoch": 0.17506575486586007,
+      "grad_norm": 0.85833340883255,
+      "learning_rate": 0.00018917556408023102,
+      "loss": 2.0118,
+      "step": 1664
+    },
+    {
+      "epoch": 0.1751709626512362,
+      "grad_norm": 1.2028058767318726,
+      "learning_rate": 0.00018916013830477766,
+      "loss": 2.0637,
+      "step": 1665
+    },
+    {
+      "epoch": 0.1752761704366123,
+      "grad_norm": 0.8081074357032776,
+      "learning_rate": 0.00018914470217545103,
+      "loss": 1.7637,
+      "step": 1666
+    },
+    {
+      "epoch": 0.17538137822198843,
+      "grad_norm": 0.9753775000572205,
+      "learning_rate": 0.0001891292556940437,
+      "loss": 2.4205,
+      "step": 1667
+    },
+    {
+      "epoch": 0.17548658600736455,
+      "grad_norm": 0.8399510383605957,
+      "learning_rate": 0.00018911379886234938,
+      "loss": 1.6638,
+      "step": 1668
+    },
+    {
+      "epoch": 0.17559179379274067,
+      "grad_norm": 1.3433152437210083,
+      "learning_rate": 0.00018909833168216306,
+      "loss": 2.049,
+      "step": 1669
+    },
+    {
+      "epoch": 0.17569700157811677,
+      "grad_norm": 0.8703423142433167,
+      "learning_rate": 0.00018908285415528088,
+      "loss": 1.973,
+      "step": 1670
+    },
+    {
+      "epoch": 0.1758022093634929,
+      "grad_norm": 1.1213691234588623,
+      "learning_rate": 0.0001890673662835002,
+      "loss": 2.2099,
+      "step": 1671
+    },
+    {
+      "epoch": 0.175907417148869,
+      "grad_norm": 1.0827136039733887,
+      "learning_rate": 0.00018905186806861957,
+      "loss": 2.2668,
+      "step": 1672
+    },
+    {
+      "epoch": 0.17601262493424513,
+      "grad_norm": 1.356737494468689,
+      "learning_rate": 0.0001890363595124387,
+      "loss": 2.206,
+      "step": 1673
+    },
+    {
+      "epoch": 0.17611783271962125,
+      "grad_norm": 1.1863815784454346,
+      "learning_rate": 0.00018902084061675863,
+      "loss": 1.8536,
+      "step": 1674
+    },
+    {
+      "epoch": 0.17622304050499737,
+      "grad_norm": 0.8727356791496277,
+      "learning_rate": 0.00018900531138338144,
+      "loss": 2.0044,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1763282482903735,
+      "grad_norm": 1.1327565908432007,
+      "learning_rate": 0.00018898977181411054,
+      "loss": 1.8945,
+      "step": 1676
+    },
+    {
+      "epoch": 0.17643345607574962,
+      "grad_norm": 0.8317870497703552,
+      "learning_rate": 0.0001889742219107505,
+      "loss": 2.3596,
+      "step": 1677
+    },
+    {
+      "epoch": 0.1765386638611257,
+      "grad_norm": 1.6865041255950928,
+      "learning_rate": 0.00018895866167510704,
+      "loss": 2.0753,
+      "step": 1678
+    },
+    {
+      "epoch": 0.17664387164650183,
+      "grad_norm": 1.1657840013504028,
+      "learning_rate": 0.00018894309110898712,
+      "loss": 2.2913,
+      "step": 1679
+    },
+    {
+      "epoch": 0.17674907943187795,
+      "grad_norm": 1.2765939235687256,
+      "learning_rate": 0.0001889275102141989,
+      "loss": 1.9399,
+      "step": 1680
+    },
+    {
+      "epoch": 0.17685428721725407,
+      "grad_norm": 1.0933256149291992,
+      "learning_rate": 0.0001889119189925518,
+      "loss": 2.0644,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1769594950026302,
+      "grad_norm": 1.196715235710144,
+      "learning_rate": 0.0001888963174458563,
+      "loss": 2.2077,
+      "step": 1682
+    },
+    {
+      "epoch": 0.17706470278800632,
+      "grad_norm": 0.8405833840370178,
+      "learning_rate": 0.00018888070557592418,
+      "loss": 2.35,
+      "step": 1683
+    },
+    {
+      "epoch": 0.17716991057338244,
+      "grad_norm": 0.6497626304626465,
+      "learning_rate": 0.0001888650833845684,
+      "loss": 2.0005,
+      "step": 1684
+    },
+    {
+      "epoch": 0.17727511835875856,
+      "grad_norm": 1.1812434196472168,
+      "learning_rate": 0.00018884945087360312,
+      "loss": 1.9776,
+      "step": 1685
+    },
+    {
+      "epoch": 0.17738032614413465,
+      "grad_norm": 1.2201387882232666,
+      "learning_rate": 0.00018883380804484367,
+      "loss": 1.9608,
+      "step": 1686
+    },
+    {
+      "epoch": 0.17748553392951077,
+      "grad_norm": 1.2153204679489136,
+      "learning_rate": 0.00018881815490010662,
+      "loss": 2.1224,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1775907417148869,
+      "grad_norm": 0.8220624327659607,
+      "learning_rate": 0.00018880249144120973,
+      "loss": 2.0308,
+      "step": 1688
+    },
+    {
+      "epoch": 0.17769594950026302,
+      "grad_norm": 1.218924880027771,
+      "learning_rate": 0.0001887868176699719,
+      "loss": 2.0924,
+      "step": 1689
+    },
+    {
+      "epoch": 0.17780115728563914,
+      "grad_norm": 1.4680471420288086,
+      "learning_rate": 0.0001887711335882133,
+      "loss": 1.4879,
+      "step": 1690
+    },
+    {
+      "epoch": 0.17790636507101526,
+      "grad_norm": 0.9292452335357666,
+      "learning_rate": 0.00018875543919775534,
+      "loss": 2.3925,
+      "step": 1691
+    },
+    {
+      "epoch": 0.17801157285639138,
+      "grad_norm": 0.935463011264801,
+      "learning_rate": 0.00018873973450042044,
+      "loss": 1.7671,
+      "step": 1692
+    },
+    {
+      "epoch": 0.1781167806417675,
+      "grad_norm": 1.0145299434661865,
+      "learning_rate": 0.00018872401949803237,
+      "loss": 1.9564,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1782219884271436,
+      "grad_norm": 1.4179667234420776,
+      "learning_rate": 0.00018870829419241608,
+      "loss": 1.9507,
+      "step": 1694
+    },
+    {
+      "epoch": 0.17832719621251972,
+      "grad_norm": 1.5500625371932983,
+      "learning_rate": 0.0001886925585853977,
+      "loss": 1.6376,
+      "step": 1695
+    },
+    {
+      "epoch": 0.17843240399789584,
+      "grad_norm": 1.2018243074417114,
+      "learning_rate": 0.0001886768126788046,
+      "loss": 1.9299,
+      "step": 1696
+    },
+    {
+      "epoch": 0.17853761178327196,
+      "grad_norm": 1.6733994483947754,
+      "learning_rate": 0.0001886610564744652,
+      "loss": 2.1827,
+      "step": 1697
+    },
+    {
+      "epoch": 0.17864281956864808,
+      "grad_norm": 1.0895193815231323,
+      "learning_rate": 0.00018864528997420928,
+      "loss": 2.2253,
+      "step": 1698
+    },
+    {
+      "epoch": 0.1787480273540242,
+      "grad_norm": 1.5276850461959839,
+      "learning_rate": 0.0001886295131798677,
+      "loss": 2.1097,
+      "step": 1699
+    },
+    {
+      "epoch": 0.17885323513940032,
+      "grad_norm": 1.0547844171524048,
+      "learning_rate": 0.00018861372609327263,
+      "loss": 1.8359,
+      "step": 1700
+    },
+    {
+      "epoch": 0.17895844292477645,
+      "grad_norm": 1.3544683456420898,
+      "learning_rate": 0.00018859792871625736,
+      "loss": 1.8435,
+      "step": 1701
+    },
+    {
+      "epoch": 0.17906365071015254,
+      "grad_norm": 1.3501803874969482,
+      "learning_rate": 0.0001885821210506564,
+      "loss": 1.7588,
+      "step": 1702
+    },
+    {
+      "epoch": 0.17916885849552866,
+      "grad_norm": 1.4099911451339722,
+      "learning_rate": 0.00018856630309830536,
+      "loss": 2.0634,
+      "step": 1703
+    },
+    {
+      "epoch": 0.17927406628090478,
+      "grad_norm": 1.3355512619018555,
+      "learning_rate": 0.0001885504748610412,
+      "loss": 1.6948,
+      "step": 1704
+    },
+    {
+      "epoch": 0.1793792740662809,
+      "grad_norm": 1.5482978820800781,
+      "learning_rate": 0.000188534636340702,
+      "loss": 1.7464,
+      "step": 1705
+    },
+    {
+      "epoch": 0.17948448185165702,
+      "grad_norm": 1.2930607795715332,
+      "learning_rate": 0.000188518787539127,
+      "loss": 2.3852,
+      "step": 1706
+    },
+    {
+      "epoch": 0.17958968963703315,
+      "grad_norm": 1.4282091856002808,
+      "learning_rate": 0.00018850292845815672,
+      "loss": 1.8695,
+      "step": 1707
+    },
+    {
+      "epoch": 0.17969489742240927,
+      "grad_norm": 1.1130648851394653,
+      "learning_rate": 0.00018848705909963275,
+      "loss": 2.0868,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1798001052077854,
+      "grad_norm": 1.1909211874008179,
+      "learning_rate": 0.000188471179465398,
+      "loss": 1.8749,
+      "step": 1709
+    },
+    {
+      "epoch": 0.17990531299316148,
+      "grad_norm": 0.7360309958457947,
+      "learning_rate": 0.00018845528955729654,
+      "loss": 1.9737,
+      "step": 1710
+    },
+    {
+      "epoch": 0.1800105207785376,
+      "grad_norm": 1.4310061931610107,
+      "learning_rate": 0.00018843938937717356,
+      "loss": 2.1225,
+      "step": 1711
+    },
+    {
+      "epoch": 0.18011572856391372,
+      "grad_norm": 1.2631195783615112,
+      "learning_rate": 0.00018842347892687552,
+      "loss": 1.7833,
+      "step": 1712
+    },
+    {
+      "epoch": 0.18022093634928985,
+      "grad_norm": 1.025079607963562,
+      "learning_rate": 0.00018840755820825002,
+      "loss": 2.2132,
+      "step": 1713
+    },
+    {
+      "epoch": 0.18032614413466597,
+      "grad_norm": 0.9161180257797241,
+      "learning_rate": 0.0001883916272231459,
+      "loss": 2.1401,
+      "step": 1714
+    },
+    {
+      "epoch": 0.1804313519200421,
+      "grad_norm": 0.9029863476753235,
+      "learning_rate": 0.0001883756859734132,
+      "loss": 1.8179,
+      "step": 1715
+    },
+    {
+      "epoch": 0.1805365597054182,
+      "grad_norm": 1.060342788696289,
+      "learning_rate": 0.00018835973446090312,
+      "loss": 1.8488,
+      "step": 1716
+    },
+    {
+      "epoch": 0.18064176749079433,
+      "grad_norm": 1.106149435043335,
+      "learning_rate": 0.000188343772687468,
+      "loss": 2.1516,
+      "step": 1717
+    },
+    {
+      "epoch": 0.18074697527617042,
+      "grad_norm": 1.5603445768356323,
+      "learning_rate": 0.0001883278006549615,
+      "loss": 1.7992,
+      "step": 1718
+    },
+    {
+      "epoch": 0.18085218306154655,
+      "grad_norm": 1.4736990928649902,
+      "learning_rate": 0.00018831181836523832,
+      "loss": 1.5836,
+      "step": 1719
+    },
+    {
+      "epoch": 0.18095739084692267,
+      "grad_norm": 1.0077097415924072,
+      "learning_rate": 0.00018829582582015453,
+      "loss": 1.9346,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1810625986322988,
+      "grad_norm": 1.0348749160766602,
+      "learning_rate": 0.0001882798230215672,
+      "loss": 1.9961,
+      "step": 1721
+    },
+    {
+      "epoch": 0.1811678064176749,
+      "grad_norm": 1.658117413520813,
+      "learning_rate": 0.00018826380997133475,
+      "loss": 1.9501,
+      "step": 1722
+    },
+    {
+      "epoch": 0.18127301420305103,
+      "grad_norm": 0.9015898704528809,
+      "learning_rate": 0.00018824778667131669,
+      "loss": 2.412,
+      "step": 1723
+    },
+    {
+      "epoch": 0.18137822198842715,
+      "grad_norm": 1.5265978574752808,
+      "learning_rate": 0.00018823175312337374,
+      "loss": 1.9471,
+      "step": 1724
+    },
+    {
+      "epoch": 0.18148342977380327,
+      "grad_norm": 1.099857211112976,
+      "learning_rate": 0.00018821570932936785,
+      "loss": 2.1251,
+      "step": 1725
+    },
+    {
+      "epoch": 0.18158863755917937,
+      "grad_norm": 0.8147949576377869,
+      "learning_rate": 0.0001881996552911621,
+      "loss": 1.7347,
+      "step": 1726
+    },
+    {
+      "epoch": 0.1816938453445555,
+      "grad_norm": 1.1035587787628174,
+      "learning_rate": 0.00018818359101062087,
+      "loss": 2.0451,
+      "step": 1727
+    },
+    {
+      "epoch": 0.1817990531299316,
+      "grad_norm": 1.2525732517242432,
+      "learning_rate": 0.00018816751648960956,
+      "loss": 2.0103,
+      "step": 1728
+    },
+    {
+      "epoch": 0.18190426091530773,
+      "grad_norm": 1.0820599794387817,
+      "learning_rate": 0.0001881514317299949,
+      "loss": 2.0033,
+      "step": 1729
+    },
+    {
+      "epoch": 0.18200946870068385,
+      "grad_norm": 1.1944822072982788,
+      "learning_rate": 0.00018813533673364474,
+      "loss": 1.5899,
+      "step": 1730
+    },
+    {
+      "epoch": 0.18211467648605997,
+      "grad_norm": 0.9098942875862122,
+      "learning_rate": 0.00018811923150242814,
+      "loss": 1.6848,
+      "step": 1731
+    },
+    {
+      "epoch": 0.1822198842714361,
+      "grad_norm": 1.7374067306518555,
+      "learning_rate": 0.00018810311603821534,
+      "loss": 1.9592,
+      "step": 1732
+    },
+    {
+      "epoch": 0.18232509205681222,
+      "grad_norm": 1.2925838232040405,
+      "learning_rate": 0.00018808699034287784,
+      "loss": 2.1075,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1824302998421883,
+      "grad_norm": 1.4516898393630981,
+      "learning_rate": 0.0001880708544182882,
+      "loss": 1.8392,
+      "step": 1734
+    },
+    {
+      "epoch": 0.18253550762756443,
+      "grad_norm": 1.116715908050537,
+      "learning_rate": 0.00018805470826632024,
+      "loss": 2.0185,
+      "step": 1735
+    },
+    {
+      "epoch": 0.18264071541294055,
+      "grad_norm": 0.9944667816162109,
+      "learning_rate": 0.00018803855188884896,
+      "loss": 2.3831,
+      "step": 1736
+    },
+    {
+      "epoch": 0.18274592319831667,
+      "grad_norm": 0.8294582962989807,
+      "learning_rate": 0.00018802238528775055,
+      "loss": 2.1609,
+      "step": 1737
+    },
+    {
+      "epoch": 0.1828511309836928,
+      "grad_norm": 1.0243922472000122,
+      "learning_rate": 0.0001880062084649024,
+      "loss": 2.3259,
+      "step": 1738
+    },
+    {
+      "epoch": 0.18295633876906892,
+      "grad_norm": 1.0900909900665283,
+      "learning_rate": 0.00018799002142218306,
+      "loss": 2.0693,
+      "step": 1739
+    },
+    {
+      "epoch": 0.18306154655444504,
+      "grad_norm": 1.310567855834961,
+      "learning_rate": 0.00018797382416147227,
+      "loss": 2.1476,
+      "step": 1740
+    },
+    {
+      "epoch": 0.18316675433982116,
+      "grad_norm": 0.9449117183685303,
+      "learning_rate": 0.00018795761668465098,
+      "loss": 2.0165,
+      "step": 1741
+    },
+    {
+      "epoch": 0.18327196212519725,
+      "grad_norm": 1.4861035346984863,
+      "learning_rate": 0.0001879413989936013,
+      "loss": 1.583,
+      "step": 1742
+    },
+    {
+      "epoch": 0.18337716991057337,
+      "grad_norm": 1.134390950202942,
+      "learning_rate": 0.00018792517109020654,
+      "loss": 1.9263,
+      "step": 1743
+    },
+    {
+      "epoch": 0.1834823776959495,
+      "grad_norm": 1.9425288438796997,
+      "learning_rate": 0.00018790893297635118,
+      "loss": 2.2563,
+      "step": 1744
+    },
+    {
+      "epoch": 0.18358758548132562,
+      "grad_norm": 1.343895673751831,
+      "learning_rate": 0.0001878926846539209,
+      "loss": 1.8078,
+      "step": 1745
+    },
+    {
+      "epoch": 0.18369279326670174,
+      "grad_norm": 1.0244340896606445,
+      "learning_rate": 0.00018787642612480261,
+      "loss": 2.0321,
+      "step": 1746
+    },
+    {
+      "epoch": 0.18379800105207786,
+      "grad_norm": 1.5598095655441284,
+      "learning_rate": 0.0001878601573908843,
+      "loss": 2.149,
+      "step": 1747
+    },
+    {
+      "epoch": 0.18390320883745398,
+      "grad_norm": 1.1912977695465088,
+      "learning_rate": 0.00018784387845405525,
+      "loss": 2.3885,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1840084166228301,
+      "grad_norm": 1.210438847541809,
+      "learning_rate": 0.00018782758931620584,
+      "loss": 1.8181,
+      "step": 1749
+    },
+    {
+      "epoch": 0.1841136244082062,
+      "grad_norm": 1.1713554859161377,
+      "learning_rate": 0.0001878112899792277,
+      "loss": 1.9201,
+      "step": 1750
+    },
+    {
+      "epoch": 0.18421883219358232,
+      "grad_norm": 1.1458523273468018,
+      "learning_rate": 0.0001877949804450136,
+      "loss": 1.8281,
+      "step": 1751
+    },
+    {
+      "epoch": 0.18432403997895844,
+      "grad_norm": 0.9206444621086121,
+      "learning_rate": 0.00018777866071545751,
+      "loss": 2.0485,
+      "step": 1752
+    },
+    {
+      "epoch": 0.18442924776433456,
+      "grad_norm": 1.196629285812378,
+      "learning_rate": 0.0001877623307924546,
+      "loss": 2.2331,
+      "step": 1753
+    },
+    {
+      "epoch": 0.18453445554971068,
+      "grad_norm": 1.2622110843658447,
+      "learning_rate": 0.00018774599067790127,
+      "loss": 1.6533,
+      "step": 1754
+    },
+    {
+      "epoch": 0.1846396633350868,
+      "grad_norm": 1.3255887031555176,
+      "learning_rate": 0.0001877296403736949,
+      "loss": 1.9875,
+      "step": 1755
+    },
+    {
+      "epoch": 0.18474487112046292,
+      "grad_norm": 0.9285155534744263,
+      "learning_rate": 0.00018771327988173435,
+      "loss": 2.0829,
+      "step": 1756
+    },
+    {
+      "epoch": 0.18485007890583904,
+      "grad_norm": 1.3429813385009766,
+      "learning_rate": 0.0001876969092039194,
+      "loss": 2.3519,
+      "step": 1757
+    },
+    {
+      "epoch": 0.18495528669121514,
+      "grad_norm": 0.8573184013366699,
+      "learning_rate": 0.0001876805283421512,
+      "loss": 2.2899,
+      "step": 1758
+    },
+    {
+      "epoch": 0.18506049447659126,
+      "grad_norm": 1.3685758113861084,
+      "learning_rate": 0.00018766413729833192,
+      "loss": 2.1059,
+      "step": 1759
+    },
+    {
+      "epoch": 0.18516570226196738,
+      "grad_norm": 1.2804449796676636,
+      "learning_rate": 0.0001876477360743651,
+      "loss": 2.1972,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1852709100473435,
+      "grad_norm": 1.6442826986312866,
+      "learning_rate": 0.00018763132467215527,
+      "loss": 1.6334,
+      "step": 1761
+    },
+    {
+      "epoch": 0.18537611783271962,
+      "grad_norm": 1.3029580116271973,
+      "learning_rate": 0.00018761490309360826,
+      "loss": 1.8276,
+      "step": 1762
+    },
+    {
+      "epoch": 0.18548132561809574,
+      "grad_norm": 1.166380524635315,
+      "learning_rate": 0.00018759847134063108,
+      "loss": 1.6849,
+      "step": 1763
+    },
+    {
+      "epoch": 0.18558653340347187,
+      "grad_norm": 1.6139475107192993,
+      "learning_rate": 0.0001875820294151319,
+      "loss": 1.8161,
+      "step": 1764
+    },
+    {
+      "epoch": 0.185691741188848,
+      "grad_norm": 1.228589653968811,
+      "learning_rate": 0.00018756557731902,
+      "loss": 1.7924,
+      "step": 1765
+    },
+    {
+      "epoch": 0.18579694897422408,
+      "grad_norm": 1.2991068363189697,
+      "learning_rate": 0.00018754911505420598,
+      "loss": 1.9833,
+      "step": 1766
+    },
+    {
+      "epoch": 0.1859021567596002,
+      "grad_norm": 1.323332667350769,
+      "learning_rate": 0.00018753264262260153,
+      "loss": 2.072,
+      "step": 1767
+    },
+    {
+      "epoch": 0.18600736454497632,
+      "grad_norm": 1.4269095659255981,
+      "learning_rate": 0.0001875161600261195,
+      "loss": 2.2458,
+      "step": 1768
+    },
+    {
+      "epoch": 0.18611257233035244,
+      "grad_norm": 1.6868364810943604,
+      "learning_rate": 0.000187499667266674,
+      "loss": 1.4101,
+      "step": 1769
+    },
+    {
+      "epoch": 0.18621778011572857,
+      "grad_norm": 1.9576934576034546,
+      "learning_rate": 0.0001874831643461803,
+      "loss": 2.193,
+      "step": 1770
+    },
+    {
+      "epoch": 0.1863229879011047,
+      "grad_norm": 1.4027034044265747,
+      "learning_rate": 0.00018746665126655477,
+      "loss": 2.0634,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1864281956864808,
+      "grad_norm": 1.0985709428787231,
+      "learning_rate": 0.00018745012802971503,
+      "loss": 2.1046,
+      "step": 1772
+    },
+    {
+      "epoch": 0.18653340347185693,
+      "grad_norm": 0.9427407383918762,
+      "learning_rate": 0.00018743359463757996,
+      "loss": 2.001,
+      "step": 1773
+    },
+    {
+      "epoch": 0.18663861125723302,
+      "grad_norm": 1.0140053033828735,
+      "learning_rate": 0.0001874170510920694,
+      "loss": 2.0172,
+      "step": 1774
+    },
+    {
+      "epoch": 0.18674381904260914,
+      "grad_norm": 1.3929756879806519,
+      "learning_rate": 0.00018740049739510454,
+      "loss": 1.9716,
+      "step": 1775
+    },
+    {
+      "epoch": 0.18684902682798527,
+      "grad_norm": 0.9349243640899658,
+      "learning_rate": 0.00018738393354860775,
+      "loss": 2.0728,
+      "step": 1776
+    },
+    {
+      "epoch": 0.1869542346133614,
+      "grad_norm": 0.9166633486747742,
+      "learning_rate": 0.00018736735955450251,
+      "loss": 1.9776,
+      "step": 1777
+    },
+    {
+      "epoch": 0.1870594423987375,
+      "grad_norm": 1.3984131813049316,
+      "learning_rate": 0.0001873507754147135,
+      "loss": 2.0968,
+      "step": 1778
+    },
+    {
+      "epoch": 0.18716465018411363,
+      "grad_norm": 1.1432921886444092,
+      "learning_rate": 0.0001873341811311666,
+      "loss": 1.687,
+      "step": 1779
+    },
+    {
+      "epoch": 0.18726985796948975,
+      "grad_norm": 1.310202956199646,
+      "learning_rate": 0.00018731757670578878,
+      "loss": 1.7605,
+      "step": 1780
+    },
+    {
+      "epoch": 0.18737506575486587,
+      "grad_norm": 1.1294410228729248,
+      "learning_rate": 0.00018730096214050832,
+      "loss": 1.805,
+      "step": 1781
+    },
+    {
+      "epoch": 0.18748027354024197,
+      "grad_norm": 0.7705867886543274,
+      "learning_rate": 0.0001872843374372546,
+      "loss": 2.1706,
+      "step": 1782
+    },
+    {
+      "epoch": 0.1875854813256181,
+      "grad_norm": 1.1728864908218384,
+      "learning_rate": 0.00018726770259795821,
+      "loss": 2.1194,
+      "step": 1783
+    },
+    {
+      "epoch": 0.1876906891109942,
+      "grad_norm": 1.0230637788772583,
+      "learning_rate": 0.0001872510576245509,
+      "loss": 2.2653,
+      "step": 1784
+    },
+    {
+      "epoch": 0.18779589689637033,
+      "grad_norm": 1.1115864515304565,
+      "learning_rate": 0.00018723440251896552,
+      "loss": 2.1956,
+      "step": 1785
+    },
+    {
+      "epoch": 0.18790110468174645,
+      "grad_norm": 1.0122426748275757,
+      "learning_rate": 0.00018721773728313628,
+      "loss": 1.6465,
+      "step": 1786
+    },
+    {
+      "epoch": 0.18800631246712257,
+      "grad_norm": 0.9406682848930359,
+      "learning_rate": 0.0001872010619189984,
+      "loss": 1.7737,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1881115202524987,
+      "grad_norm": 1.0271477699279785,
+      "learning_rate": 0.00018718437642848833,
+      "loss": 2.3859,
+      "step": 1788
+    },
+    {
+      "epoch": 0.18821672803787481,
+      "grad_norm": 1.2405545711517334,
+      "learning_rate": 0.00018716768081354374,
+      "loss": 1.8296,
+      "step": 1789
+    },
+    {
+      "epoch": 0.1883219358232509,
+      "grad_norm": 1.5526014566421509,
+      "learning_rate": 0.0001871509750761034,
+      "loss": 2.0319,
+      "step": 1790
+    },
+    {
+      "epoch": 0.18842714360862703,
+      "grad_norm": 1.2660319805145264,
+      "learning_rate": 0.00018713425921810733,
+      "loss": 2.2965,
+      "step": 1791
+    },
+    {
+      "epoch": 0.18853235139400315,
+      "grad_norm": 0.9081699848175049,
+      "learning_rate": 0.00018711753324149663,
+      "loss": 2.2645,
+      "step": 1792
+    },
+    {
+      "epoch": 0.18863755917937927,
+      "grad_norm": 0.7778252363204956,
+      "learning_rate": 0.00018710079714821367,
+      "loss": 2.1503,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1887427669647554,
+      "grad_norm": 0.903505802154541,
+      "learning_rate": 0.00018708405094020197,
+      "loss": 2.1276,
+      "step": 1794
+    },
+    {
+      "epoch": 0.18884797475013151,
+      "grad_norm": 0.9733725190162659,
+      "learning_rate": 0.00018706729461940617,
+      "loss": 2.1536,
+      "step": 1795
+    },
+    {
+      "epoch": 0.18895318253550764,
+      "grad_norm": 1.1625339984893799,
+      "learning_rate": 0.00018705052818777219,
+      "loss": 1.7845,
+      "step": 1796
+    },
+    {
+      "epoch": 0.18905839032088376,
+      "grad_norm": 0.9139503836631775,
+      "learning_rate": 0.000187033751647247,
+      "loss": 2.1074,
+      "step": 1797
+    },
+    {
+      "epoch": 0.18916359810625985,
+      "grad_norm": 1.0095267295837402,
+      "learning_rate": 0.00018701696499977884,
+      "loss": 1.7807,
+      "step": 1798
+    },
+    {
+      "epoch": 0.18926880589163597,
+      "grad_norm": 0.9188694357872009,
+      "learning_rate": 0.00018700016824731706,
+      "loss": 1.9302,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1893740136770121,
+      "grad_norm": 0.9856252074241638,
+      "learning_rate": 0.0001869833613918122,
+      "loss": 2.1511,
+      "step": 1800
+    },
+    {
+      "epoch": 0.18947922146238821,
+      "grad_norm": 1.3189128637313843,
+      "learning_rate": 0.00018696654443521607,
+      "loss": 1.8588,
+      "step": 1801
+    },
+    {
+      "epoch": 0.18958442924776434,
+      "grad_norm": 1.3261507749557495,
+      "learning_rate": 0.00018694971737948145,
+      "loss": 1.9224,
+      "step": 1802
+    },
+    {
+      "epoch": 0.18968963703314046,
+      "grad_norm": 1.4043636322021484,
+      "learning_rate": 0.00018693288022656252,
+      "loss": 1.8017,
+      "step": 1803
+    },
+    {
+      "epoch": 0.18979484481851658,
+      "grad_norm": 1.6043760776519775,
+      "learning_rate": 0.00018691603297841446,
+      "loss": 1.9915,
+      "step": 1804
+    },
+    {
+      "epoch": 0.1899000526038927,
+      "grad_norm": 0.9396890997886658,
+      "learning_rate": 0.0001868991756369937,
+      "loss": 1.8151,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1900052603892688,
+      "grad_norm": 1.0986272096633911,
+      "learning_rate": 0.0001868823082042578,
+      "loss": 2.1091,
+      "step": 1806
+    },
+    {
+      "epoch": 0.19011046817464491,
+      "grad_norm": 0.7267233729362488,
+      "learning_rate": 0.00018686543068216556,
+      "loss": 2.2471,
+      "step": 1807
+    },
+    {
+      "epoch": 0.19021567596002104,
+      "grad_norm": 0.7171627283096313,
+      "learning_rate": 0.0001868485430726769,
+      "loss": 1.785,
+      "step": 1808
+    },
+    {
+      "epoch": 0.19032088374539716,
+      "grad_norm": 1.197615146636963,
+      "learning_rate": 0.0001868316453777529,
+      "loss": 2.0766,
+      "step": 1809
+    },
+    {
+      "epoch": 0.19042609153077328,
+      "grad_norm": 1.0680128335952759,
+      "learning_rate": 0.00018681473759935585,
+      "loss": 2.1709,
+      "step": 1810
+    },
+    {
+      "epoch": 0.1905312993161494,
+      "grad_norm": 1.3420791625976562,
+      "learning_rate": 0.00018679781973944922,
+      "loss": 2.1537,
+      "step": 1811
+    },
+    {
+      "epoch": 0.19063650710152552,
+      "grad_norm": 0.9118467569351196,
+      "learning_rate": 0.00018678089179999762,
+      "loss": 1.8085,
+      "step": 1812
+    },
+    {
+      "epoch": 0.19074171488690164,
+      "grad_norm": 0.9484670162200928,
+      "learning_rate": 0.00018676395378296678,
+      "loss": 2.1784,
+      "step": 1813
+    },
+    {
+      "epoch": 0.19084692267227774,
+      "grad_norm": 0.7800343632698059,
+      "learning_rate": 0.0001867470056903237,
+      "loss": 2.0045,
+      "step": 1814
+    },
+    {
+      "epoch": 0.19095213045765386,
+      "grad_norm": 1.0197774171829224,
+      "learning_rate": 0.00018673004752403651,
+      "loss": 2.1594,
+      "step": 1815
+    },
+    {
+      "epoch": 0.19105733824302998,
+      "grad_norm": 1.9036952257156372,
+      "learning_rate": 0.0001867130792860745,
+      "loss": 2.3911,
+      "step": 1816
+    },
+    {
+      "epoch": 0.1911625460284061,
+      "grad_norm": 1.0456516742706299,
+      "learning_rate": 0.00018669610097840812,
+      "loss": 2.0251,
+      "step": 1817
+    },
+    {
+      "epoch": 0.19126775381378222,
+      "grad_norm": 1.23896324634552,
+      "learning_rate": 0.00018667911260300904,
+      "loss": 2.091,
+      "step": 1818
+    },
+    {
+      "epoch": 0.19137296159915834,
+      "grad_norm": 1.0979390144348145,
+      "learning_rate": 0.00018666211416184999,
+      "loss": 1.8643,
+      "step": 1819
+    },
+    {
+      "epoch": 0.19147816938453446,
+      "grad_norm": 1.4400193691253662,
+      "learning_rate": 0.00018664510565690506,
+      "loss": 1.9773,
+      "step": 1820
+    },
+    {
+      "epoch": 0.19158337716991058,
+      "grad_norm": 1.1655758619308472,
+      "learning_rate": 0.0001866280870901493,
+      "loss": 2.0926,
+      "step": 1821
+    },
+    {
+      "epoch": 0.19168858495528668,
+      "grad_norm": 1.0224357843399048,
+      "learning_rate": 0.00018661105846355902,
+      "loss": 1.7359,
+      "step": 1822
+    },
+    {
+      "epoch": 0.1917937927406628,
+      "grad_norm": 1.4255973100662231,
+      "learning_rate": 0.00018659401977911175,
+      "loss": 2.5401,
+      "step": 1823
+    },
+    {
+      "epoch": 0.19189900052603892,
+      "grad_norm": 1.0111758708953857,
+      "learning_rate": 0.0001865769710387861,
+      "loss": 2.319,
+      "step": 1824
+    },
+    {
+      "epoch": 0.19200420831141504,
+      "grad_norm": 1.0198891162872314,
+      "learning_rate": 0.00018655991224456191,
+      "loss": 1.7324,
+      "step": 1825
+    },
+    {
+      "epoch": 0.19210941609679116,
+      "grad_norm": 1.0950661897659302,
+      "learning_rate": 0.00018654284339842013,
+      "loss": 1.8189,
+      "step": 1826
+    },
+    {
+      "epoch": 0.19221462388216728,
+      "grad_norm": 1.0327266454696655,
+      "learning_rate": 0.0001865257645023429,
+      "loss": 1.9624,
+      "step": 1827
+    },
+    {
+      "epoch": 0.1923198316675434,
+      "grad_norm": 1.1767691373825073,
+      "learning_rate": 0.0001865086755583136,
+      "loss": 2.2126,
+      "step": 1828
+    },
+    {
+      "epoch": 0.19242503945291953,
+      "grad_norm": 0.9736087918281555,
+      "learning_rate": 0.0001864915765683167,
+      "loss": 2.2341,
+      "step": 1829
+    },
+    {
+      "epoch": 0.19253024723829562,
+      "grad_norm": 1.7082476615905762,
+      "learning_rate": 0.00018647446753433777,
+      "loss": 1.7682,
+      "step": 1830
+    },
+    {
+      "epoch": 0.19263545502367174,
+      "grad_norm": 1.0709311962127686,
+      "learning_rate": 0.00018645734845836368,
+      "loss": 1.7948,
+      "step": 1831
+    },
+    {
+      "epoch": 0.19274066280904786,
+      "grad_norm": 1.171373724937439,
+      "learning_rate": 0.00018644021934238243,
+      "loss": 2.1489,
+      "step": 1832
+    },
+    {
+      "epoch": 0.19284587059442398,
+      "grad_norm": 1.0802509784698486,
+      "learning_rate": 0.00018642308018838316,
+      "loss": 2.4032,
+      "step": 1833
+    },
+    {
+      "epoch": 0.1929510783798001,
+      "grad_norm": 1.047210693359375,
+      "learning_rate": 0.00018640593099835618,
+      "loss": 2.2197,
+      "step": 1834
+    },
+    {
+      "epoch": 0.19305628616517623,
+      "grad_norm": 0.7914498448371887,
+      "learning_rate": 0.00018638877177429292,
+      "loss": 1.5993,
+      "step": 1835
+    },
+    {
+      "epoch": 0.19316149395055235,
+      "grad_norm": 0.8980052471160889,
+      "learning_rate": 0.0001863716025181861,
+      "loss": 2.2428,
+      "step": 1836
+    },
+    {
+      "epoch": 0.19326670173592847,
+      "grad_norm": 1.1372369527816772,
+      "learning_rate": 0.00018635442323202946,
+      "loss": 2.2588,
+      "step": 1837
+    },
+    {
+      "epoch": 0.19337190952130456,
+      "grad_norm": 0.995132327079773,
+      "learning_rate": 0.00018633723391781802,
+      "loss": 2.2008,
+      "step": 1838
+    },
+    {
+      "epoch": 0.19347711730668068,
+      "grad_norm": 1.5439562797546387,
+      "learning_rate": 0.00018632003457754793,
+      "loss": 2.0667,
+      "step": 1839
+    },
+    {
+      "epoch": 0.1935823250920568,
+      "grad_norm": 1.945412039756775,
+      "learning_rate": 0.00018630282521321645,
+      "loss": 2.2433,
+      "step": 1840
+    },
+    {
+      "epoch": 0.19368753287743293,
+      "grad_norm": 0.7718631029129028,
+      "learning_rate": 0.00018628560582682207,
+      "loss": 2.3626,
+      "step": 1841
+    },
+    {
+      "epoch": 0.19379274066280905,
+      "grad_norm": 1.3353549242019653,
+      "learning_rate": 0.0001862683764203644,
+      "loss": 1.9693,
+      "step": 1842
+    },
+    {
+      "epoch": 0.19389794844818517,
+      "grad_norm": 1.200984239578247,
+      "learning_rate": 0.00018625113699584426,
+      "loss": 2.0508,
+      "step": 1843
+    },
+    {
+      "epoch": 0.1940031562335613,
+      "grad_norm": 1.5393805503845215,
+      "learning_rate": 0.00018623388755526364,
+      "loss": 1.7895,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1941083640189374,
+      "grad_norm": 1.4993560314178467,
+      "learning_rate": 0.00018621662810062558,
+      "loss": 1.9832,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1942135718043135,
+      "grad_norm": 1.0732812881469727,
+      "learning_rate": 0.00018619935863393444,
+      "loss": 1.6443,
+      "step": 1846
+    },
+    {
+      "epoch": 0.19431877958968963,
+      "grad_norm": 1.4361530542373657,
+      "learning_rate": 0.0001861820791571956,
+      "loss": 2.2628,
+      "step": 1847
+    },
+    {
+      "epoch": 0.19442398737506575,
+      "grad_norm": 0.9760112166404724,
+      "learning_rate": 0.00018616478967241568,
+      "loss": 2.043,
+      "step": 1848
+    },
+    {
+      "epoch": 0.19452919516044187,
+      "grad_norm": 1.2888370752334595,
+      "learning_rate": 0.00018614749018160248,
+      "loss": 2.0368,
+      "step": 1849
+    },
+    {
+      "epoch": 0.194634402945818,
+      "grad_norm": 1.2557207345962524,
+      "learning_rate": 0.00018613018068676493,
+      "loss": 2.0163,
+      "step": 1850
+    },
+    {
+      "epoch": 0.1947396107311941,
+      "grad_norm": 1.2219709157943726,
+      "learning_rate": 0.00018611286118991313,
+      "loss": 1.4421,
+      "step": 1851
+    },
+    {
+      "epoch": 0.19484481851657023,
+      "grad_norm": 0.9364914894104004,
+      "learning_rate": 0.0001860955316930583,
+      "loss": 1.4738,
+      "step": 1852
+    },
+    {
+      "epoch": 0.19495002630194636,
+      "grad_norm": 1.3384712934494019,
+      "learning_rate": 0.0001860781921982129,
+      "loss": 2.2564,
+      "step": 1853
+    },
+    {
+      "epoch": 0.19505523408732245,
+      "grad_norm": 2.036452054977417,
+      "learning_rate": 0.00018606084270739049,
+      "loss": 2.0251,
+      "step": 1854
+    },
+    {
+      "epoch": 0.19516044187269857,
+      "grad_norm": 1.2831714153289795,
+      "learning_rate": 0.00018604348322260578,
+      "loss": 1.8063,
+      "step": 1855
+    },
+    {
+      "epoch": 0.1952656496580747,
+      "grad_norm": 0.8491331934928894,
+      "learning_rate": 0.0001860261137458747,
+      "loss": 1.6988,
+      "step": 1856
+    },
+    {
+      "epoch": 0.1953708574434508,
+      "grad_norm": 1.3487755060195923,
+      "learning_rate": 0.00018600873427921435,
+      "loss": 1.9603,
+      "step": 1857
+    },
+    {
+      "epoch": 0.19547606522882693,
+      "grad_norm": 1.7233006954193115,
+      "learning_rate": 0.00018599134482464287,
+      "loss": 1.8172,
+      "step": 1858
+    },
+    {
+      "epoch": 0.19558127301420306,
+      "grad_norm": 1.1606876850128174,
+      "learning_rate": 0.0001859739453841797,
+      "loss": 1.8372,
+      "step": 1859
+    },
+    {
+      "epoch": 0.19568648079957918,
+      "grad_norm": 1.3234946727752686,
+      "learning_rate": 0.00018595653595984536,
+      "loss": 2.0716,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1957916885849553,
+      "grad_norm": 1.0269123315811157,
+      "learning_rate": 0.0001859391165536615,
+      "loss": 1.8958,
+      "step": 1861
+    },
+    {
+      "epoch": 0.1958968963703314,
+      "grad_norm": 1.1604347229003906,
+      "learning_rate": 0.0001859216871676511,
+      "loss": 2.1997,
+      "step": 1862
+    },
+    {
+      "epoch": 0.1960021041557075,
+      "grad_norm": 0.8779052495956421,
+      "learning_rate": 0.00018590424780383805,
+      "loss": 1.8132,
+      "step": 1863
+    },
+    {
+      "epoch": 0.19610731194108363,
+      "grad_norm": 1.0170305967330933,
+      "learning_rate": 0.0001858867984642476,
+      "loss": 2.1858,
+      "step": 1864
+    },
+    {
+      "epoch": 0.19621251972645976,
+      "grad_norm": 0.9716585874557495,
+      "learning_rate": 0.00018586933915090605,
+      "loss": 2.1944,
+      "step": 1865
+    },
+    {
+      "epoch": 0.19631772751183588,
+      "grad_norm": 1.326249361038208,
+      "learning_rate": 0.0001858518698658409,
+      "loss": 1.8595,
+      "step": 1866
+    },
+    {
+      "epoch": 0.196422935297212,
+      "grad_norm": 0.9899535179138184,
+      "learning_rate": 0.00018583439061108084,
+      "loss": 2.2828,
+      "step": 1867
+    },
+    {
+      "epoch": 0.19652814308258812,
+      "grad_norm": 0.9519728422164917,
+      "learning_rate": 0.0001858169013886556,
+      "loss": 2.2894,
+      "step": 1868
+    },
+    {
+      "epoch": 0.19663335086796424,
+      "grad_norm": 1.11469304561615,
+      "learning_rate": 0.0001857994022005962,
+      "loss": 2.0972,
+      "step": 1869
+    },
+    {
+      "epoch": 0.19673855865334033,
+      "grad_norm": 0.7735962271690369,
+      "learning_rate": 0.00018578189304893479,
+      "loss": 1.8865,
+      "step": 1870
+    },
+    {
+      "epoch": 0.19684376643871646,
+      "grad_norm": 1.504098653793335,
+      "learning_rate": 0.00018576437393570458,
+      "loss": 2.2802,
+      "step": 1871
+    },
+    {
+      "epoch": 0.19694897422409258,
+      "grad_norm": 0.8838121891021729,
+      "learning_rate": 0.00018574684486294006,
+      "loss": 2.0275,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1970541820094687,
+      "grad_norm": 0.868427574634552,
+      "learning_rate": 0.0001857293058326768,
+      "loss": 2.2436,
+      "step": 1873
+    },
+    {
+      "epoch": 0.19715938979484482,
+      "grad_norm": 1.472787857055664,
+      "learning_rate": 0.00018571175684695154,
+      "loss": 2.3729,
+      "step": 1874
+    },
+    {
+      "epoch": 0.19726459758022094,
+      "grad_norm": 0.9417343139648438,
+      "learning_rate": 0.00018569419790780218,
+      "loss": 2.0141,
+      "step": 1875
+    },
+    {
+      "epoch": 0.19736980536559706,
+      "grad_norm": 1.3230361938476562,
+      "learning_rate": 0.00018567662901726784,
+      "loss": 2.1847,
+      "step": 1876
+    },
+    {
+      "epoch": 0.19747501315097318,
+      "grad_norm": 0.8992961049079895,
+      "learning_rate": 0.00018565905017738868,
+      "loss": 2.3444,
+      "step": 1877
+    },
+    {
+      "epoch": 0.19758022093634928,
+      "grad_norm": 1.455939769744873,
+      "learning_rate": 0.00018564146139020608,
+      "loss": 2.0235,
+      "step": 1878
+    },
+    {
+      "epoch": 0.1976854287217254,
+      "grad_norm": 1.1618740558624268,
+      "learning_rate": 0.00018562386265776263,
+      "loss": 2.627,
+      "step": 1879
+    },
+    {
+      "epoch": 0.19779063650710152,
+      "grad_norm": 0.8262205719947815,
+      "learning_rate": 0.00018560625398210192,
+      "loss": 2.0536,
+      "step": 1880
+    },
+    {
+      "epoch": 0.19789584429247764,
+      "grad_norm": 0.9469594955444336,
+      "learning_rate": 0.00018558863536526885,
+      "loss": 2.3183,
+      "step": 1881
+    },
+    {
+      "epoch": 0.19800105207785376,
+      "grad_norm": 1.1301559209823608,
+      "learning_rate": 0.00018557100680930937,
+      "loss": 2.3325,
+      "step": 1882
+    },
+    {
+      "epoch": 0.19810625986322988,
+      "grad_norm": 0.8819060921669006,
+      "learning_rate": 0.00018555336831627063,
+      "loss": 1.6358,
+      "step": 1883
+    },
+    {
+      "epoch": 0.198211467648606,
+      "grad_norm": 1.2020803689956665,
+      "learning_rate": 0.000185535719888201,
+      "loss": 2.0548,
+      "step": 1884
+    },
+    {
+      "epoch": 0.19831667543398213,
+      "grad_norm": 1.0018656253814697,
+      "learning_rate": 0.00018551806152714985,
+      "loss": 2.1452,
+      "step": 1885
+    },
+    {
+      "epoch": 0.19842188321935822,
+      "grad_norm": 1.1717722415924072,
+      "learning_rate": 0.00018550039323516783,
+      "loss": 2.0638,
+      "step": 1886
+    },
+    {
+      "epoch": 0.19852709100473434,
+      "grad_norm": 1.4248872995376587,
+      "learning_rate": 0.00018548271501430668,
+      "loss": 1.6008,
+      "step": 1887
+    },
+    {
+      "epoch": 0.19863229879011046,
+      "grad_norm": 1.5258474349975586,
+      "learning_rate": 0.00018546502686661934,
+      "loss": 1.6953,
+      "step": 1888
+    },
+    {
+      "epoch": 0.19873750657548658,
+      "grad_norm": 1.2465327978134155,
+      "learning_rate": 0.00018544732879415986,
+      "loss": 2.2588,
+      "step": 1889
+    },
+    {
+      "epoch": 0.1988427143608627,
+      "grad_norm": 1.0645866394042969,
+      "learning_rate": 0.00018542962079898346,
+      "loss": 2.2034,
+      "step": 1890
+    },
+    {
+      "epoch": 0.19894792214623883,
+      "grad_norm": 1.7710703611373901,
+      "learning_rate": 0.00018541190288314647,
+      "loss": 2.1372,
+      "step": 1891
+    },
+    {
+      "epoch": 0.19905312993161495,
+      "grad_norm": 1.4344483613967896,
+      "learning_rate": 0.00018539417504870648,
+      "loss": 1.7988,
+      "step": 1892
+    },
+    {
+      "epoch": 0.19915833771699107,
+      "grad_norm": 1.4004452228546143,
+      "learning_rate": 0.00018537643729772216,
+      "loss": 1.9769,
+      "step": 1893
+    },
+    {
+      "epoch": 0.19926354550236716,
+      "grad_norm": 0.8972628116607666,
+      "learning_rate": 0.00018535868963225326,
+      "loss": 2.2142,
+      "step": 1894
+    },
+    {
+      "epoch": 0.19936875328774328,
+      "grad_norm": 0.7937827706336975,
+      "learning_rate": 0.00018534093205436087,
+      "loss": 2.0791,
+      "step": 1895
+    },
+    {
+      "epoch": 0.1994739610731194,
+      "grad_norm": 1.6337438821792603,
+      "learning_rate": 0.00018532316456610704,
+      "loss": 1.486,
+      "step": 1896
+    },
+    {
+      "epoch": 0.19957916885849553,
+      "grad_norm": 2.0083436965942383,
+      "learning_rate": 0.00018530538716955504,
+      "loss": 1.8192,
+      "step": 1897
+    },
+    {
+      "epoch": 0.19968437664387165,
+      "grad_norm": 1.693180799484253,
+      "learning_rate": 0.0001852875998667694,
+      "loss": 1.4399,
+      "step": 1898
+    },
+    {
+      "epoch": 0.19978958442924777,
+      "grad_norm": 0.8119178414344788,
+      "learning_rate": 0.0001852698026598156,
+      "loss": 2.0997,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1998947922146239,
+      "grad_norm": 1.1830841302871704,
+      "learning_rate": 0.0001852519955507604,
+      "loss": 1.7776,
+      "step": 1900
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7960675358772278,
+      "learning_rate": 0.00018523417854167168,
+      "loss": 2.2009,
+      "step": 1901
+    },
+    {
+      "epoch": 0.2001052077853761,
+      "grad_norm": 0.9493720531463623,
+      "learning_rate": 0.00018521635163461846,
+      "loss": 1.9135,
+      "step": 1902
+    },
+    {
+      "epoch": 0.20021041557075223,
+      "grad_norm": 1.2081838846206665,
+      "learning_rate": 0.00018519851483167097,
+      "loss": 1.9611,
+      "step": 1903
+    },
+    {
+      "epoch": 0.20031562335612835,
+      "grad_norm": 1.2762458324432373,
+      "learning_rate": 0.00018518066813490047,
+      "loss": 2.1839,
+      "step": 1904
+    },
+    {
+      "epoch": 0.20042083114150447,
+      "grad_norm": 1.249342679977417,
+      "learning_rate": 0.0001851628115463795,
+      "loss": 1.6887,
+      "step": 1905
+    },
+    {
+      "epoch": 0.2005260389268806,
+      "grad_norm": 1.3049440383911133,
+      "learning_rate": 0.00018514494506818166,
+      "loss": 2.3789,
+      "step": 1906
+    },
+    {
+      "epoch": 0.2006312467122567,
+      "grad_norm": 1.0053954124450684,
+      "learning_rate": 0.0001851270687023817,
+      "loss": 2.2105,
+      "step": 1907
+    },
+    {
+      "epoch": 0.20073645449763283,
+      "grad_norm": 1.1582366228103638,
+      "learning_rate": 0.0001851091824510556,
+      "loss": 1.9553,
+      "step": 1908
+    },
+    {
+      "epoch": 0.20084166228300895,
+      "grad_norm": 1.2567706108093262,
+      "learning_rate": 0.00018509128631628036,
+      "loss": 2.3391,
+      "step": 1909
+    },
+    {
+      "epoch": 0.20094687006838505,
+      "grad_norm": 0.823357105255127,
+      "learning_rate": 0.00018507338030013427,
+      "loss": 2.1023,
+      "step": 1910
+    },
+    {
+      "epoch": 0.20105207785376117,
+      "grad_norm": 0.7399005889892578,
+      "learning_rate": 0.0001850554644046967,
+      "loss": 1.947,
+      "step": 1911
+    },
+    {
+      "epoch": 0.2011572856391373,
+      "grad_norm": 1.4578897953033447,
+      "learning_rate": 0.00018503753863204807,
+      "loss": 1.2573,
+      "step": 1912
+    },
+    {
+      "epoch": 0.2012624934245134,
+      "grad_norm": 0.9766344428062439,
+      "learning_rate": 0.00018501960298427013,
+      "loss": 1.9681,
+      "step": 1913
+    },
+    {
+      "epoch": 0.20136770120988953,
+      "grad_norm": 1.1950548887252808,
+      "learning_rate": 0.00018500165746344562,
+      "loss": 2.1528,
+      "step": 1914
+    },
+    {
+      "epoch": 0.20147290899526565,
+      "grad_norm": 1.4169249534606934,
+      "learning_rate": 0.0001849837020716586,
+      "loss": 2.0959,
+      "step": 1915
+    },
+    {
+      "epoch": 0.20157811678064178,
+      "grad_norm": 1.3415583372116089,
+      "learning_rate": 0.00018496573681099401,
+      "loss": 1.7363,
+      "step": 1916
+    },
+    {
+      "epoch": 0.2016833245660179,
+      "grad_norm": 1.3929237127304077,
+      "learning_rate": 0.00018494776168353827,
+      "loss": 1.8159,
+      "step": 1917
+    },
+    {
+      "epoch": 0.201788532351394,
+      "grad_norm": 1.296832799911499,
+      "learning_rate": 0.00018492977669137868,
+      "loss": 2.2187,
+      "step": 1918
+    },
+    {
+      "epoch": 0.2018937401367701,
+      "grad_norm": 0.8477041125297546,
+      "learning_rate": 0.00018491178183660376,
+      "loss": 2.0901,
+      "step": 1919
+    },
+    {
+      "epoch": 0.20199894792214623,
+      "grad_norm": 1.353830099105835,
+      "learning_rate": 0.00018489377712130326,
+      "loss": 2.3666,
+      "step": 1920
+    },
+    {
+      "epoch": 0.20210415570752235,
+      "grad_norm": 1.109371304512024,
+      "learning_rate": 0.00018487576254756793,
+      "loss": 2.1329,
+      "step": 1921
+    },
+    {
+      "epoch": 0.20220936349289848,
+      "grad_norm": 1.1594985723495483,
+      "learning_rate": 0.0001848577381174898,
+      "loss": 2.059,
+      "step": 1922
+    },
+    {
+      "epoch": 0.2023145712782746,
+      "grad_norm": 1.2802788019180298,
+      "learning_rate": 0.00018483970383316198,
+      "loss": 2.0002,
+      "step": 1923
+    },
+    {
+      "epoch": 0.20241977906365072,
+      "grad_norm": 1.8601975440979004,
+      "learning_rate": 0.00018482165969667874,
+      "loss": 2.3373,
+      "step": 1924
+    },
+    {
+      "epoch": 0.20252498684902684,
+      "grad_norm": 1.242659330368042,
+      "learning_rate": 0.00018480360571013544,
+      "loss": 1.755,
+      "step": 1925
+    },
+    {
+      "epoch": 0.20263019463440293,
+      "grad_norm": 1.3230918645858765,
+      "learning_rate": 0.00018478554187562868,
+      "loss": 2.2089,
+      "step": 1926
+    },
+    {
+      "epoch": 0.20273540241977905,
+      "grad_norm": 1.1816704273223877,
+      "learning_rate": 0.00018476746819525613,
+      "loss": 1.6534,
+      "step": 1927
+    },
+    {
+      "epoch": 0.20284061020515518,
+      "grad_norm": 1.1363966464996338,
+      "learning_rate": 0.00018474938467111663,
+      "loss": 2.1807,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2029458179905313,
+      "grad_norm": 1.0613765716552734,
+      "learning_rate": 0.00018473129130531016,
+      "loss": 1.87,
+      "step": 1929
+    },
+    {
+      "epoch": 0.20305102577590742,
+      "grad_norm": 0.7839847207069397,
+      "learning_rate": 0.00018471318809993784,
+      "loss": 1.6391,
+      "step": 1930
+    },
+    {
+      "epoch": 0.20315623356128354,
+      "grad_norm": 1.5794448852539062,
+      "learning_rate": 0.00018469507505710194,
+      "loss": 2.2534,
+      "step": 1931
+    },
+    {
+      "epoch": 0.20326144134665966,
+      "grad_norm": 1.1429455280303955,
+      "learning_rate": 0.0001846769521789059,
+      "loss": 1.6465,
+      "step": 1932
+    },
+    {
+      "epoch": 0.20336664913203578,
+      "grad_norm": 1.0982143878936768,
+      "learning_rate": 0.0001846588194674542,
+      "loss": 1.8907,
+      "step": 1933
+    },
+    {
+      "epoch": 0.20347185691741188,
+      "grad_norm": 1.2898775339126587,
+      "learning_rate": 0.00018464067692485254,
+      "loss": 1.6681,
+      "step": 1934
+    },
+    {
+      "epoch": 0.203577064702788,
+      "grad_norm": 1.1818253993988037,
+      "learning_rate": 0.00018462252455320785,
+      "loss": 2.2638,
+      "step": 1935
+    },
+    {
+      "epoch": 0.20368227248816412,
+      "grad_norm": 1.7338184118270874,
+      "learning_rate": 0.000184604362354628,
+      "loss": 2.2689,
+      "step": 1936
+    },
+    {
+      "epoch": 0.20378748027354024,
+      "grad_norm": 1.2104718685150146,
+      "learning_rate": 0.00018458619033122218,
+      "loss": 1.9735,
+      "step": 1937
+    },
+    {
+      "epoch": 0.20389268805891636,
+      "grad_norm": 0.9602882266044617,
+      "learning_rate": 0.00018456800848510056,
+      "loss": 2.2646,
+      "step": 1938
+    },
+    {
+      "epoch": 0.20399789584429248,
+      "grad_norm": 0.9840019941329956,
+      "learning_rate": 0.00018454981681837463,
+      "loss": 1.8798,
+      "step": 1939
+    },
+    {
+      "epoch": 0.2041031036296686,
+      "grad_norm": 0.9403349161148071,
+      "learning_rate": 0.0001845316153331569,
+      "loss": 2.063,
+      "step": 1940
+    },
+    {
+      "epoch": 0.20420831141504472,
+      "grad_norm": 1.7658482789993286,
+      "learning_rate": 0.000184513404031561,
+      "loss": 2.1487,
+      "step": 1941
+    },
+    {
+      "epoch": 0.20431351920042082,
+      "grad_norm": 1.6682535409927368,
+      "learning_rate": 0.00018449518291570183,
+      "loss": 2.1085,
+      "step": 1942
+    },
+    {
+      "epoch": 0.20441872698579694,
+      "grad_norm": 1.355338215827942,
+      "learning_rate": 0.00018447695198769526,
+      "loss": 2.2586,
+      "step": 1943
+    },
+    {
+      "epoch": 0.20452393477117306,
+      "grad_norm": 1.2086683511734009,
+      "learning_rate": 0.00018445871124965843,
+      "loss": 2.0692,
+      "step": 1944
+    },
+    {
+      "epoch": 0.20462914255654918,
+      "grad_norm": 1.9813165664672852,
+      "learning_rate": 0.00018444046070370963,
+      "loss": 1.7036,
+      "step": 1945
+    },
+    {
+      "epoch": 0.2047343503419253,
+      "grad_norm": 1.0692365169525146,
+      "learning_rate": 0.00018442220035196812,
+      "loss": 2.2982,
+      "step": 1946
+    },
+    {
+      "epoch": 0.20483955812730142,
+      "grad_norm": 1.4883077144622803,
+      "learning_rate": 0.00018440393019655452,
+      "loss": 2.304,
+      "step": 1947
+    },
+    {
+      "epoch": 0.20494476591267755,
+      "grad_norm": 2.0655646324157715,
+      "learning_rate": 0.00018438565023959043,
+      "loss": 2.0391,
+      "step": 1948
+    },
+    {
+      "epoch": 0.20504997369805367,
+      "grad_norm": 1.2688896656036377,
+      "learning_rate": 0.00018436736048319866,
+      "loss": 1.6661,
+      "step": 1949
+    },
+    {
+      "epoch": 0.20515518148342976,
+      "grad_norm": 1.0065981149673462,
+      "learning_rate": 0.00018434906092950313,
+      "loss": 2.1238,
+      "step": 1950
+    },
+    {
+      "epoch": 0.20526038926880588,
+      "grad_norm": 1.2444024085998535,
+      "learning_rate": 0.00018433075158062891,
+      "loss": 2.0663,
+      "step": 1951
+    },
+    {
+      "epoch": 0.205365597054182,
+      "grad_norm": 1.5449178218841553,
+      "learning_rate": 0.00018431243243870223,
+      "loss": 2.1242,
+      "step": 1952
+    },
+    {
+      "epoch": 0.20547080483955812,
+      "grad_norm": 0.7938444018363953,
+      "learning_rate": 0.00018429410350585034,
+      "loss": 2.2292,
+      "step": 1953
+    },
+    {
+      "epoch": 0.20557601262493425,
+      "grad_norm": 1.2025595903396606,
+      "learning_rate": 0.00018427576478420186,
+      "loss": 2.057,
+      "step": 1954
+    },
+    {
+      "epoch": 0.20568122041031037,
+      "grad_norm": 1.0300257205963135,
+      "learning_rate": 0.00018425741627588627,
+      "loss": 1.9549,
+      "step": 1955
+    },
+    {
+      "epoch": 0.2057864281956865,
+      "grad_norm": 1.1756985187530518,
+      "learning_rate": 0.0001842390579830344,
+      "loss": 1.7581,
+      "step": 1956
+    },
+    {
+      "epoch": 0.2058916359810626,
+      "grad_norm": 1.201172947883606,
+      "learning_rate": 0.00018422068990777812,
+      "loss": 1.9362,
+      "step": 1957
+    },
+    {
+      "epoch": 0.2059968437664387,
+      "grad_norm": 1.3182251453399658,
+      "learning_rate": 0.00018420231205225048,
+      "loss": 2.057,
+      "step": 1958
+    },
+    {
+      "epoch": 0.20610205155181482,
+      "grad_norm": 1.0256531238555908,
+      "learning_rate": 0.00018418392441858555,
+      "loss": 1.3062,
+      "step": 1959
+    },
+    {
+      "epoch": 0.20620725933719095,
+      "grad_norm": 1.0511935949325562,
+      "learning_rate": 0.00018416552700891873,
+      "loss": 2.1073,
+      "step": 1960
+    },
+    {
+      "epoch": 0.20631246712256707,
+      "grad_norm": 1.2152209281921387,
+      "learning_rate": 0.0001841471198253864,
+      "loss": 2.1933,
+      "step": 1961
+    },
+    {
+      "epoch": 0.2064176749079432,
+      "grad_norm": 1.1236765384674072,
+      "learning_rate": 0.00018412870287012612,
+      "loss": 1.9926,
+      "step": 1962
+    },
+    {
+      "epoch": 0.2065228826933193,
+      "grad_norm": 1.0651775598526,
+      "learning_rate": 0.00018411027614527665,
+      "loss": 2.0337,
+      "step": 1963
+    },
+    {
+      "epoch": 0.20662809047869543,
+      "grad_norm": 1.3447325229644775,
+      "learning_rate": 0.00018409183965297776,
+      "loss": 2.0808,
+      "step": 1964
+    },
+    {
+      "epoch": 0.20673329826407155,
+      "grad_norm": 1.0653175115585327,
+      "learning_rate": 0.0001840733933953704,
+      "loss": 2.1391,
+      "step": 1965
+    },
+    {
+      "epoch": 0.20683850604944765,
+      "grad_norm": 0.9823555946350098,
+      "learning_rate": 0.0001840549373745968,
+      "loss": 2.2805,
+      "step": 1966
+    },
+    {
+      "epoch": 0.20694371383482377,
+      "grad_norm": 1.1034152507781982,
+      "learning_rate": 0.00018403647159280002,
+      "loss": 2.1311,
+      "step": 1967
+    },
+    {
+      "epoch": 0.2070489216201999,
+      "grad_norm": 1.0444360971450806,
+      "learning_rate": 0.00018401799605212457,
+      "loss": 2.3337,
+      "step": 1968
+    },
+    {
+      "epoch": 0.207154129405576,
+      "grad_norm": 0.9108467102050781,
+      "learning_rate": 0.00018399951075471588,
+      "loss": 1.8849,
+      "step": 1969
+    },
+    {
+      "epoch": 0.20725933719095213,
+      "grad_norm": 1.4065741300582886,
+      "learning_rate": 0.0001839810157027206,
+      "loss": 1.8689,
+      "step": 1970
+    },
+    {
+      "epoch": 0.20736454497632825,
+      "grad_norm": 1.062463641166687,
+      "learning_rate": 0.00018396251089828654,
+      "loss": 1.9281,
+      "step": 1971
+    },
+    {
+      "epoch": 0.20746975276170437,
+      "grad_norm": 0.9371905326843262,
+      "learning_rate": 0.00018394399634356256,
+      "loss": 2.3794,
+      "step": 1972
+    },
+    {
+      "epoch": 0.2075749605470805,
+      "grad_norm": 1.0700284242630005,
+      "learning_rate": 0.0001839254720406987,
+      "loss": 2.0316,
+      "step": 1973
+    },
+    {
+      "epoch": 0.2076801683324566,
+      "grad_norm": 1.718322992324829,
+      "learning_rate": 0.00018390693799184613,
+      "loss": 1.9874,
+      "step": 1974
+    },
+    {
+      "epoch": 0.2077853761178327,
+      "grad_norm": 1.9158811569213867,
+      "learning_rate": 0.00018388839419915715,
+      "loss": 1.595,
+      "step": 1975
+    },
+    {
+      "epoch": 0.20789058390320883,
+      "grad_norm": 0.9104257225990295,
+      "learning_rate": 0.00018386984066478518,
+      "loss": 1.8649,
+      "step": 1976
+    },
+    {
+      "epoch": 0.20799579168858495,
+      "grad_norm": 1.0303380489349365,
+      "learning_rate": 0.00018385127739088482,
+      "loss": 1.9061,
+      "step": 1977
+    },
+    {
+      "epoch": 0.20810099947396107,
+      "grad_norm": 0.9922066926956177,
+      "learning_rate": 0.0001838327043796117,
+      "loss": 1.966,
+      "step": 1978
+    },
+    {
+      "epoch": 0.2082062072593372,
+      "grad_norm": 1.0381168127059937,
+      "learning_rate": 0.0001838141216331227,
+      "loss": 2.3589,
+      "step": 1979
+    },
+    {
+      "epoch": 0.20831141504471332,
+      "grad_norm": 1.5640273094177246,
+      "learning_rate": 0.00018379552915357575,
+      "loss": 1.9377,
+      "step": 1980
+    },
+    {
+      "epoch": 0.20841662283008944,
+      "grad_norm": 1.299121618270874,
+      "learning_rate": 0.00018377692694312994,
+      "loss": 2.0013,
+      "step": 1981
+    },
+    {
+      "epoch": 0.20852183061546553,
+      "grad_norm": 1.0334281921386719,
+      "learning_rate": 0.0001837583150039454,
+      "loss": 1.9686,
+      "step": 1982
+    },
+    {
+      "epoch": 0.20862703840084165,
+      "grad_norm": 1.1700999736785889,
+      "learning_rate": 0.00018373969333818364,
+      "loss": 1.6355,
+      "step": 1983
+    },
+    {
+      "epoch": 0.20873224618621777,
+      "grad_norm": 1.4064104557037354,
+      "learning_rate": 0.00018372106194800703,
+      "loss": 2.3571,
+      "step": 1984
+    },
+    {
+      "epoch": 0.2088374539715939,
+      "grad_norm": 1.18000328540802,
+      "learning_rate": 0.00018370242083557914,
+      "loss": 2.0538,
+      "step": 1985
+    },
+    {
+      "epoch": 0.20894266175697002,
+      "grad_norm": 0.7155065536499023,
+      "learning_rate": 0.00018368377000306475,
+      "loss": 2.2469,
+      "step": 1986
+    },
+    {
+      "epoch": 0.20904786954234614,
+      "grad_norm": 0.7191174626350403,
+      "learning_rate": 0.00018366510945262972,
+      "loss": 2.0709,
+      "step": 1987
+    },
+    {
+      "epoch": 0.20915307732772226,
+      "grad_norm": 1.1349201202392578,
+      "learning_rate": 0.00018364643918644108,
+      "loss": 1.3495,
+      "step": 1988
+    },
+    {
+      "epoch": 0.20925828511309838,
+      "grad_norm": 0.8776866793632507,
+      "learning_rate": 0.00018362775920666684,
+      "loss": 2.28,
+      "step": 1989
+    },
+    {
+      "epoch": 0.20936349289847447,
+      "grad_norm": 1.0915156602859497,
+      "learning_rate": 0.00018360906951547633,
+      "loss": 2.3553,
+      "step": 1990
+    },
+    {
+      "epoch": 0.2094687006838506,
+      "grad_norm": 1.0722743272781372,
+      "learning_rate": 0.00018359037011503988,
+      "loss": 2.0794,
+      "step": 1991
+    },
+    {
+      "epoch": 0.20957390846922672,
+      "grad_norm": 1.764479398727417,
+      "learning_rate": 0.000183571661007529,
+      "loss": 2.4184,
+      "step": 1992
+    },
+    {
+      "epoch": 0.20967911625460284,
+      "grad_norm": 0.9234581589698792,
+      "learning_rate": 0.00018355294219511633,
+      "loss": 2.3174,
+      "step": 1993
+    },
+    {
+      "epoch": 0.20978432403997896,
+      "grad_norm": 1.334053635597229,
+      "learning_rate": 0.00018353421367997563,
+      "loss": 1.8774,
+      "step": 1994
+    },
+    {
+      "epoch": 0.20988953182535508,
+      "grad_norm": 0.8430811166763306,
+      "learning_rate": 0.00018351547546428175,
+      "loss": 2.094,
+      "step": 1995
+    },
+    {
+      "epoch": 0.2099947396107312,
+      "grad_norm": 1.3585329055786133,
+      "learning_rate": 0.00018349672755021073,
+      "loss": 1.9657,
+      "step": 1996
+    },
+    {
+      "epoch": 0.21009994739610732,
+      "grad_norm": 1.2807389497756958,
+      "learning_rate": 0.00018347796993993968,
+      "loss": 1.1998,
+      "step": 1997
+    },
+    {
+      "epoch": 0.21020515518148342,
+      "grad_norm": 1.249908685684204,
+      "learning_rate": 0.00018345920263564683,
+      "loss": 2.0678,
+      "step": 1998
+    },
+    {
+      "epoch": 0.21031036296685954,
+      "grad_norm": 1.2682554721832275,
+      "learning_rate": 0.00018344042563951167,
+      "loss": 2.2967,
+      "step": 1999
+    },
+    {
+      "epoch": 0.21041557075223566,
+      "grad_norm": 1.0497610569000244,
+      "learning_rate": 0.0001834216389537146,
+      "loss": 2.2477,
+      "step": 2000
+    },
+    {
+      "epoch": 0.21052077853761178,
+      "grad_norm": 1.0887739658355713,
+      "learning_rate": 0.00018340284258043732,
+      "loss": 1.9668,
+      "step": 2001
+    },
+    {
+      "epoch": 0.2106259863229879,
+      "grad_norm": 1.3399494886398315,
+      "learning_rate": 0.00018338403652186255,
+      "loss": 1.4247,
+      "step": 2002
+    },
+    {
+      "epoch": 0.21073119410836402,
+      "grad_norm": 1.225252389907837,
+      "learning_rate": 0.0001833652207801742,
+      "loss": 1.7742,
+      "step": 2003
+    },
+    {
+      "epoch": 0.21083640189374014,
+      "grad_norm": 0.9872320294380188,
+      "learning_rate": 0.0001833463953575573,
+      "loss": 2.131,
+      "step": 2004
+    },
+    {
+      "epoch": 0.21094160967911627,
+      "grad_norm": 1.161528468132019,
+      "learning_rate": 0.00018332756025619796,
+      "loss": 2.1038,
+      "step": 2005
+    },
+    {
+      "epoch": 0.21104681746449236,
+      "grad_norm": 0.8207192420959473,
+      "learning_rate": 0.00018330871547828342,
+      "loss": 1.9778,
+      "step": 2006
+    },
+    {
+      "epoch": 0.21115202524986848,
+      "grad_norm": 1.2393897771835327,
+      "learning_rate": 0.00018328986102600207,
+      "loss": 1.8538,
+      "step": 2007
+    },
+    {
+      "epoch": 0.2112572330352446,
+      "grad_norm": 1.052492380142212,
+      "learning_rate": 0.00018327099690154344,
+      "loss": 2.0008,
+      "step": 2008
+    },
+    {
+      "epoch": 0.21136244082062072,
+      "grad_norm": 1.1104390621185303,
+      "learning_rate": 0.00018325212310709815,
+      "loss": 2.2228,
+      "step": 2009
+    },
+    {
+      "epoch": 0.21146764860599684,
+      "grad_norm": 1.1710113286972046,
+      "learning_rate": 0.00018323323964485795,
+      "loss": 2.3255,
+      "step": 2010
+    },
+    {
+      "epoch": 0.21157285639137297,
+      "grad_norm": 1.5669982433319092,
+      "learning_rate": 0.00018321434651701567,
+      "loss": 2.0485,
+      "step": 2011
+    },
+    {
+      "epoch": 0.2116780641767491,
+      "grad_norm": 1.486722707748413,
+      "learning_rate": 0.00018319544372576537,
+      "loss": 1.396,
+      "step": 2012
+    },
+    {
+      "epoch": 0.2117832719621252,
+      "grad_norm": 0.9690135717391968,
+      "learning_rate": 0.00018317653127330216,
+      "loss": 1.9004,
+      "step": 2013
+    },
+    {
+      "epoch": 0.2118884797475013,
+      "grad_norm": 1.1157475709915161,
+      "learning_rate": 0.00018315760916182228,
+      "loss": 1.9181,
+      "step": 2014
+    },
+    {
+      "epoch": 0.21199368753287742,
+      "grad_norm": 1.8870946168899536,
+      "learning_rate": 0.00018313867739352304,
+      "loss": 1.8504,
+      "step": 2015
+    },
+    {
+      "epoch": 0.21209889531825354,
+      "grad_norm": 1.5294593572616577,
+      "learning_rate": 0.000183119735970603,
+      "loss": 2.0109,
+      "step": 2016
+    },
+    {
+      "epoch": 0.21220410310362967,
+      "grad_norm": 1.3320099115371704,
+      "learning_rate": 0.00018310078489526172,
+      "loss": 2.149,
+      "step": 2017
+    },
+    {
+      "epoch": 0.2123093108890058,
+      "grad_norm": 1.6566550731658936,
+      "learning_rate": 0.0001830818241696999,
+      "loss": 1.8748,
+      "step": 2018
+    },
+    {
+      "epoch": 0.2124145186743819,
+      "grad_norm": 1.4480758905410767,
+      "learning_rate": 0.00018306285379611947,
+      "loss": 1.9223,
+      "step": 2019
+    },
+    {
+      "epoch": 0.21251972645975803,
+      "grad_norm": 1.1992216110229492,
+      "learning_rate": 0.00018304387377672331,
+      "loss": 2.334,
+      "step": 2020
+    },
+    {
+      "epoch": 0.21262493424513415,
+      "grad_norm": 1.6246322393417358,
+      "learning_rate": 0.00018302488411371556,
+      "loss": 1.818,
+      "step": 2021
+    },
+    {
+      "epoch": 0.21273014203051024,
+      "grad_norm": 1.3721061944961548,
+      "learning_rate": 0.00018300588480930143,
+      "loss": 2.0362,
+      "step": 2022
+    },
+    {
+      "epoch": 0.21283534981588637,
+      "grad_norm": 1.2342585325241089,
+      "learning_rate": 0.00018298687586568721,
+      "loss": 1.7764,
+      "step": 2023
+    },
+    {
+      "epoch": 0.2129405576012625,
+      "grad_norm": 1.8841065168380737,
+      "learning_rate": 0.00018296785728508038,
+      "loss": 1.5991,
+      "step": 2024
+    },
+    {
+      "epoch": 0.2130457653866386,
+      "grad_norm": 1.1739152669906616,
+      "learning_rate": 0.00018294882906968947,
+      "loss": 2.1558,
+      "step": 2025
+    },
+    {
+      "epoch": 0.21315097317201473,
+      "grad_norm": 0.8841888904571533,
+      "learning_rate": 0.00018292979122172418,
+      "loss": 1.7489,
+      "step": 2026
+    },
+    {
+      "epoch": 0.21325618095739085,
+      "grad_norm": 1.0950676202774048,
+      "learning_rate": 0.00018291074374339534,
+      "loss": 2.2481,
+      "step": 2027
+    },
+    {
+      "epoch": 0.21336138874276697,
+      "grad_norm": 1.693406581878662,
+      "learning_rate": 0.00018289168663691486,
+      "loss": 2.3203,
+      "step": 2028
+    },
+    {
+      "epoch": 0.2134665965281431,
+      "grad_norm": 0.8693836331367493,
+      "learning_rate": 0.0001828726199044957,
+      "loss": 2.1194,
+      "step": 2029
+    },
+    {
+      "epoch": 0.2135718043135192,
+      "grad_norm": 0.9883258938789368,
+      "learning_rate": 0.00018285354354835215,
+      "loss": 2.0936,
+      "step": 2030
+    },
+    {
+      "epoch": 0.2136770120988953,
+      "grad_norm": 1.0930025577545166,
+      "learning_rate": 0.00018283445757069942,
+      "loss": 1.9489,
+      "step": 2031
+    },
+    {
+      "epoch": 0.21378221988427143,
+      "grad_norm": 1.8851250410079956,
+      "learning_rate": 0.00018281536197375386,
+      "loss": 1.5551,
+      "step": 2032
+    },
+    {
+      "epoch": 0.21388742766964755,
+      "grad_norm": 0.933934211730957,
+      "learning_rate": 0.00018279625675973304,
+      "loss": 1.7964,
+      "step": 2033
+    },
+    {
+      "epoch": 0.21399263545502367,
+      "grad_norm": 0.9329856038093567,
+      "learning_rate": 0.00018277714193085554,
+      "loss": 1.8394,
+      "step": 2034
+    },
+    {
+      "epoch": 0.2140978432403998,
+      "grad_norm": 2.0386810302734375,
+      "learning_rate": 0.00018275801748934115,
+      "loss": 2.0189,
+      "step": 2035
+    },
+    {
+      "epoch": 0.21420305102577591,
+      "grad_norm": 1.8662831783294678,
+      "learning_rate": 0.0001827388834374107,
+      "loss": 1.7599,
+      "step": 2036
+    },
+    {
+      "epoch": 0.21430825881115204,
+      "grad_norm": 1.1685835123062134,
+      "learning_rate": 0.0001827197397772862,
+      "loss": 1.8818,
+      "step": 2037
+    },
+    {
+      "epoch": 0.21441346659652813,
+      "grad_norm": 0.8799912333488464,
+      "learning_rate": 0.00018270058651119063,
+      "loss": 2.1813,
+      "step": 2038
+    },
+    {
+      "epoch": 0.21451867438190425,
+      "grad_norm": 0.8218079805374146,
+      "learning_rate": 0.00018268142364134834,
+      "loss": 2.3432,
+      "step": 2039
+    },
+    {
+      "epoch": 0.21462388216728037,
+      "grad_norm": 1.1055721044540405,
+      "learning_rate": 0.00018266225116998457,
+      "loss": 1.585,
+      "step": 2040
+    },
+    {
+      "epoch": 0.2147290899526565,
+      "grad_norm": 1.1287375688552856,
+      "learning_rate": 0.00018264306909932575,
+      "loss": 2.0216,
+      "step": 2041
+    },
+    {
+      "epoch": 0.21483429773803261,
+      "grad_norm": 1.085099220275879,
+      "learning_rate": 0.0001826238774315995,
+      "loss": 1.9095,
+      "step": 2042
+    },
+    {
+      "epoch": 0.21493950552340874,
+      "grad_norm": 1.0446321964263916,
+      "learning_rate": 0.0001826046761690344,
+      "loss": 2.0777,
+      "step": 2043
+    },
+    {
+      "epoch": 0.21504471330878486,
+      "grad_norm": 1.3021361827850342,
+      "learning_rate": 0.0001825854653138603,
+      "loss": 2.0682,
+      "step": 2044
+    },
+    {
+      "epoch": 0.21514992109416098,
+      "grad_norm": 0.9500402808189392,
+      "learning_rate": 0.00018256624486830803,
+      "loss": 2.3274,
+      "step": 2045
+    },
+    {
+      "epoch": 0.21525512887953707,
+      "grad_norm": 1.077355980873108,
+      "learning_rate": 0.00018254701483460964,
+      "loss": 1.9925,
+      "step": 2046
+    },
+    {
+      "epoch": 0.2153603366649132,
+      "grad_norm": 1.6127686500549316,
+      "learning_rate": 0.00018252777521499821,
+      "loss": 1.9386,
+      "step": 2047
+    },
+    {
+      "epoch": 0.21546554445028931,
+      "grad_norm": 0.7945027351379395,
+      "learning_rate": 0.00018250852601170805,
+      "loss": 2.1476,
+      "step": 2048
+    },
+    {
+      "epoch": 0.21557075223566544,
+      "grad_norm": 1.248733401298523,
+      "learning_rate": 0.00018248926722697444,
+      "loss": 2.1239,
+      "step": 2049
+    },
+    {
+      "epoch": 0.21567596002104156,
+      "grad_norm": 1.2039722204208374,
+      "learning_rate": 0.00018246999886303383,
+      "loss": 2.132,
+      "step": 2050
+    },
+    {
+      "epoch": 0.21578116780641768,
+      "grad_norm": 1.0198581218719482,
+      "learning_rate": 0.00018245072092212388,
+      "loss": 1.758,
+      "step": 2051
+    },
+    {
+      "epoch": 0.2158863755917938,
+      "grad_norm": 1.2120723724365234,
+      "learning_rate": 0.00018243143340648316,
+      "loss": 2.1045,
+      "step": 2052
+    },
+    {
+      "epoch": 0.21599158337716992,
+      "grad_norm": 0.8687598705291748,
+      "learning_rate": 0.00018241213631835153,
+      "loss": 1.568,
+      "step": 2053
+    },
+    {
+      "epoch": 0.21609679116254601,
+      "grad_norm": 1.0547373294830322,
+      "learning_rate": 0.0001823928296599699,
+      "loss": 2.3057,
+      "step": 2054
+    },
+    {
+      "epoch": 0.21620199894792214,
+      "grad_norm": 0.9490458369255066,
+      "learning_rate": 0.00018237351343358026,
+      "loss": 2.2151,
+      "step": 2055
+    },
+    {
+      "epoch": 0.21630720673329826,
+      "grad_norm": 1.1601332426071167,
+      "learning_rate": 0.00018235418764142575,
+      "loss": 2.0495,
+      "step": 2056
+    },
+    {
+      "epoch": 0.21641241451867438,
+      "grad_norm": 1.8616313934326172,
+      "learning_rate": 0.00018233485228575063,
+      "loss": 1.877,
+      "step": 2057
+    },
+    {
+      "epoch": 0.2165176223040505,
+      "grad_norm": 1.3617507219314575,
+      "learning_rate": 0.00018231550736880024,
+      "loss": 2.1049,
+      "step": 2058
+    },
+    {
+      "epoch": 0.21662283008942662,
+      "grad_norm": 1.3656481504440308,
+      "learning_rate": 0.00018229615289282102,
+      "loss": 2.0189,
+      "step": 2059
+    },
+    {
+      "epoch": 0.21672803787480274,
+      "grad_norm": 1.2620902061462402,
+      "learning_rate": 0.0001822767888600606,
+      "loss": 2.1029,
+      "step": 2060
+    },
+    {
+      "epoch": 0.21683324566017886,
+      "grad_norm": 0.9363958239555359,
+      "learning_rate": 0.00018225741527276755,
+      "loss": 1.9721,
+      "step": 2061
+    },
+    {
+      "epoch": 0.21693845344555496,
+      "grad_norm": 0.7752951979637146,
+      "learning_rate": 0.0001822380321331918,
+      "loss": 1.9241,
+      "step": 2062
+    },
+    {
+      "epoch": 0.21704366123093108,
+      "grad_norm": 1.2256505489349365,
+      "learning_rate": 0.00018221863944358412,
+      "loss": 2.0667,
+      "step": 2063
+    },
+    {
+      "epoch": 0.2171488690163072,
+      "grad_norm": 1.6273034811019897,
+      "learning_rate": 0.00018219923720619663,
+      "loss": 1.9696,
+      "step": 2064
+    },
+    {
+      "epoch": 0.21725407680168332,
+      "grad_norm": 1.1030242443084717,
+      "learning_rate": 0.00018217982542328238,
+      "loss": 2.0116,
+      "step": 2065
+    },
+    {
+      "epoch": 0.21735928458705944,
+      "grad_norm": 0.9453311562538147,
+      "learning_rate": 0.00018216040409709563,
+      "loss": 1.7222,
+      "step": 2066
+    },
+    {
+      "epoch": 0.21746449237243556,
+      "grad_norm": 1.2243266105651855,
+      "learning_rate": 0.00018214097322989168,
+      "loss": 2.2173,
+      "step": 2067
+    },
+    {
+      "epoch": 0.21756970015781169,
+      "grad_norm": 2.4034857749938965,
+      "learning_rate": 0.000182121532823927,
+      "loss": 1.7776,
+      "step": 2068
+    },
+    {
+      "epoch": 0.2176749079431878,
+      "grad_norm": 1.4039760828018188,
+      "learning_rate": 0.00018210208288145914,
+      "loss": 2.3769,
+      "step": 2069
+    },
+    {
+      "epoch": 0.2177801157285639,
+      "grad_norm": 1.2816754579544067,
+      "learning_rate": 0.00018208262340474677,
+      "loss": 1.9379,
+      "step": 2070
+    },
+    {
+      "epoch": 0.21788532351394002,
+      "grad_norm": 1.6516449451446533,
+      "learning_rate": 0.0001820631543960496,
+      "loss": 1.7285,
+      "step": 2071
+    },
+    {
+      "epoch": 0.21799053129931614,
+      "grad_norm": 1.015907645225525,
+      "learning_rate": 0.00018204367585762855,
+      "loss": 1.749,
+      "step": 2072
+    },
+    {
+      "epoch": 0.21809573908469226,
+      "grad_norm": 1.1783592700958252,
+      "learning_rate": 0.00018202418779174556,
+      "loss": 2.1641,
+      "step": 2073
+    },
+    {
+      "epoch": 0.21820094687006839,
+      "grad_norm": 0.8992342352867126,
+      "learning_rate": 0.00018200469020066378,
+      "loss": 1.8566,
+      "step": 2074
+    },
+    {
+      "epoch": 0.2183061546554445,
+      "grad_norm": 1.1242965459823608,
+      "learning_rate": 0.00018198518308664734,
+      "loss": 1.4988,
+      "step": 2075
+    },
+    {
+      "epoch": 0.21841136244082063,
+      "grad_norm": 1.6040006875991821,
+      "learning_rate": 0.00018196566645196157,
+      "loss": 2.4172,
+      "step": 2076
+    },
+    {
+      "epoch": 0.21851657022619675,
+      "grad_norm": 1.6206622123718262,
+      "learning_rate": 0.00018194614029887286,
+      "loss": 1.8929,
+      "step": 2077
+    },
+    {
+      "epoch": 0.21862177801157284,
+      "grad_norm": 1.8726282119750977,
+      "learning_rate": 0.0001819266046296487,
+      "loss": 1.8977,
+      "step": 2078
+    },
+    {
+      "epoch": 0.21872698579694896,
+      "grad_norm": 0.9685164093971252,
+      "learning_rate": 0.00018190705944655776,
+      "loss": 2.2144,
+      "step": 2079
+    },
+    {
+      "epoch": 0.21883219358232509,
+      "grad_norm": 1.0125783681869507,
+      "learning_rate": 0.00018188750475186968,
+      "loss": 1.9572,
+      "step": 2080
+    },
+    {
+      "epoch": 0.2189374013677012,
+      "grad_norm": 1.1588927507400513,
+      "learning_rate": 0.00018186794054785534,
+      "loss": 1.7875,
+      "step": 2081
+    },
+    {
+      "epoch": 0.21904260915307733,
+      "grad_norm": 0.96497642993927,
+      "learning_rate": 0.00018184836683678667,
+      "loss": 2.2885,
+      "step": 2082
+    },
+    {
+      "epoch": 0.21914781693845345,
+      "grad_norm": 1.0914496183395386,
+      "learning_rate": 0.00018182878362093665,
+      "loss": 1.9495,
+      "step": 2083
+    },
+    {
+      "epoch": 0.21925302472382957,
+      "grad_norm": 1.0270262956619263,
+      "learning_rate": 0.00018180919090257945,
+      "loss": 1.7052,
+      "step": 2084
+    },
+    {
+      "epoch": 0.2193582325092057,
+      "grad_norm": 1.0222289562225342,
+      "learning_rate": 0.00018178958868399033,
+      "loss": 2.1131,
+      "step": 2085
+    },
+    {
+      "epoch": 0.21946344029458179,
+      "grad_norm": 2.0089921951293945,
+      "learning_rate": 0.00018176997696744556,
+      "loss": 2.0778,
+      "step": 2086
+    },
+    {
+      "epoch": 0.2195686480799579,
+      "grad_norm": 1.4937543869018555,
+      "learning_rate": 0.00018175035575522264,
+      "loss": 1.8517,
+      "step": 2087
+    },
+    {
+      "epoch": 0.21967385586533403,
+      "grad_norm": 1.0364367961883545,
+      "learning_rate": 0.00018173072504960012,
+      "loss": 1.9049,
+      "step": 2088
+    },
+    {
+      "epoch": 0.21977906365071015,
+      "grad_norm": 1.225338339805603,
+      "learning_rate": 0.00018171108485285763,
+      "loss": 2.0639,
+      "step": 2089
+    },
+    {
+      "epoch": 0.21988427143608627,
+      "grad_norm": 1.0059758424758911,
+      "learning_rate": 0.0001816914351672759,
+      "loss": 1.9738,
+      "step": 2090
+    },
+    {
+      "epoch": 0.2199894792214624,
+      "grad_norm": 1.3843426704406738,
+      "learning_rate": 0.00018167177599513683,
+      "loss": 2.1056,
+      "step": 2091
+    },
+    {
+      "epoch": 0.2200946870068385,
+      "grad_norm": 1.4146124124526978,
+      "learning_rate": 0.00018165210733872336,
+      "loss": 2.014,
+      "step": 2092
+    },
+    {
+      "epoch": 0.22019989479221463,
+      "grad_norm": 0.9480724334716797,
+      "learning_rate": 0.00018163242920031953,
+      "loss": 2.1525,
+      "step": 2093
+    },
+    {
+      "epoch": 0.22030510257759073,
+      "grad_norm": 1.4912493228912354,
+      "learning_rate": 0.00018161274158221048,
+      "loss": 2.697,
+      "step": 2094
+    },
+    {
+      "epoch": 0.22041031036296685,
+      "grad_norm": 1.2136460542678833,
+      "learning_rate": 0.00018159304448668253,
+      "loss": 1.5769,
+      "step": 2095
+    },
+    {
+      "epoch": 0.22051551814834297,
+      "grad_norm": 0.991580069065094,
+      "learning_rate": 0.00018157333791602297,
+      "loss": 2.2276,
+      "step": 2096
+    },
+    {
+      "epoch": 0.2206207259337191,
+      "grad_norm": 0.8912844061851501,
+      "learning_rate": 0.00018155362187252032,
+      "loss": 2.0501,
+      "step": 2097
+    },
+    {
+      "epoch": 0.2207259337190952,
+      "grad_norm": 1.1324025392532349,
+      "learning_rate": 0.00018153389635846412,
+      "loss": 2.2222,
+      "step": 2098
+    },
+    {
+      "epoch": 0.22083114150447133,
+      "grad_norm": 0.7306355237960815,
+      "learning_rate": 0.000181514161376145,
+      "loss": 2.0178,
+      "step": 2099
+    },
+    {
+      "epoch": 0.22093634928984746,
+      "grad_norm": 1.2878000736236572,
+      "learning_rate": 0.00018149441692785474,
+      "loss": 2.1164,
+      "step": 2100
+    },
+    {
+      "epoch": 0.22104155707522358,
+      "grad_norm": 1.2069727182388306,
+      "learning_rate": 0.00018147466301588622,
+      "loss": 1.9003,
+      "step": 2101
+    },
+    {
+      "epoch": 0.22114676486059967,
+      "grad_norm": 0.8253376483917236,
+      "learning_rate": 0.00018145489964253332,
+      "loss": 1.7612,
+      "step": 2102
+    },
+    {
+      "epoch": 0.2212519726459758,
+      "grad_norm": 1.2956628799438477,
+      "learning_rate": 0.0001814351268100912,
+      "loss": 2.1136,
+      "step": 2103
+    },
+    {
+      "epoch": 0.2213571804313519,
+      "grad_norm": 0.9152336120605469,
+      "learning_rate": 0.00018141534452085595,
+      "loss": 1.8014,
+      "step": 2104
+    },
+    {
+      "epoch": 0.22146238821672803,
+      "grad_norm": 1.4479109048843384,
+      "learning_rate": 0.00018139555277712482,
+      "loss": 1.951,
+      "step": 2105
+    },
+    {
+      "epoch": 0.22156759600210416,
+      "grad_norm": 1.5760619640350342,
+      "learning_rate": 0.0001813757515811962,
+      "loss": 1.9071,
+      "step": 2106
+    },
+    {
+      "epoch": 0.22167280378748028,
+      "grad_norm": 1.1229628324508667,
+      "learning_rate": 0.0001813559409353695,
+      "loss": 1.8727,
+      "step": 2107
+    },
+    {
+      "epoch": 0.2217780115728564,
+      "grad_norm": 0.8988067507743835,
+      "learning_rate": 0.0001813361208419453,
+      "loss": 2.2383,
+      "step": 2108
+    },
+    {
+      "epoch": 0.22188321935823252,
+      "grad_norm": 1.3481158018112183,
+      "learning_rate": 0.0001813162913032252,
+      "loss": 2.0623,
+      "step": 2109
+    },
+    {
+      "epoch": 0.2219884271436086,
+      "grad_norm": 1.297059416770935,
+      "learning_rate": 0.00018129645232151193,
+      "loss": 2.4296,
+      "step": 2110
+    },
+    {
+      "epoch": 0.22209363492898473,
+      "grad_norm": 1.9435932636260986,
+      "learning_rate": 0.0001812766038991094,
+      "loss": 2.1651,
+      "step": 2111
+    },
+    {
+      "epoch": 0.22219884271436086,
+      "grad_norm": 1.3611021041870117,
+      "learning_rate": 0.00018125674603832248,
+      "loss": 1.9258,
+      "step": 2112
+    },
+    {
+      "epoch": 0.22230405049973698,
+      "grad_norm": 1.4265265464782715,
+      "learning_rate": 0.00018123687874145721,
+      "loss": 1.6352,
+      "step": 2113
+    },
+    {
+      "epoch": 0.2224092582851131,
+      "grad_norm": 0.9899458289146423,
+      "learning_rate": 0.00018121700201082072,
+      "loss": 2.3061,
+      "step": 2114
+    },
+    {
+      "epoch": 0.22251446607048922,
+      "grad_norm": 0.9266303777694702,
+      "learning_rate": 0.00018119711584872123,
+      "loss": 2.1004,
+      "step": 2115
+    },
+    {
+      "epoch": 0.22261967385586534,
+      "grad_norm": 1.2452328205108643,
+      "learning_rate": 0.00018117722025746806,
+      "loss": 2.2262,
+      "step": 2116
+    },
+    {
+      "epoch": 0.22272488164124146,
+      "grad_norm": 1.2869268655776978,
+      "learning_rate": 0.0001811573152393716,
+      "loss": 2.3989,
+      "step": 2117
+    },
+    {
+      "epoch": 0.22283008942661758,
+      "grad_norm": 0.9200719594955444,
+      "learning_rate": 0.00018113740079674337,
+      "loss": 1.9065,
+      "step": 2118
+    },
+    {
+      "epoch": 0.22293529721199368,
+      "grad_norm": 0.869301974773407,
+      "learning_rate": 0.00018111747693189595,
+      "loss": 1.7604,
+      "step": 2119
+    },
+    {
+      "epoch": 0.2230405049973698,
+      "grad_norm": 1.119777798652649,
+      "learning_rate": 0.00018109754364714305,
+      "loss": 2.128,
+      "step": 2120
+    },
+    {
+      "epoch": 0.22314571278274592,
+      "grad_norm": 1.4669591188430786,
+      "learning_rate": 0.00018107760094479948,
+      "loss": 2.174,
+      "step": 2121
+    },
+    {
+      "epoch": 0.22325092056812204,
+      "grad_norm": 0.9276453256607056,
+      "learning_rate": 0.0001810576488271811,
+      "loss": 2.1165,
+      "step": 2122
+    },
+    {
+      "epoch": 0.22335612835349816,
+      "grad_norm": 0.9469357132911682,
+      "learning_rate": 0.00018103768729660485,
+      "loss": 2.2046,
+      "step": 2123
+    },
+    {
+      "epoch": 0.22346133613887428,
+      "grad_norm": 1.525495171546936,
+      "learning_rate": 0.00018101771635538883,
+      "loss": 1.4605,
+      "step": 2124
+    },
+    {
+      "epoch": 0.2235665439242504,
+      "grad_norm": 1.022232174873352,
+      "learning_rate": 0.00018099773600585223,
+      "loss": 2.0554,
+      "step": 2125
+    },
+    {
+      "epoch": 0.22367175170962653,
+      "grad_norm": 1.015611171722412,
+      "learning_rate": 0.00018097774625031523,
+      "loss": 2.064,
+      "step": 2126
+    },
+    {
+      "epoch": 0.22377695949500262,
+      "grad_norm": 1.5566353797912598,
+      "learning_rate": 0.0001809577470910992,
+      "loss": 1.9989,
+      "step": 2127
+    },
+    {
+      "epoch": 0.22388216728037874,
+      "grad_norm": 1.2023249864578247,
+      "learning_rate": 0.0001809377385305266,
+      "loss": 1.9456,
+      "step": 2128
+    },
+    {
+      "epoch": 0.22398737506575486,
+      "grad_norm": 1.1629719734191895,
+      "learning_rate": 0.00018091772057092097,
+      "loss": 2.3635,
+      "step": 2129
+    },
+    {
+      "epoch": 0.22409258285113098,
+      "grad_norm": 1.1971515417099,
+      "learning_rate": 0.00018089769321460688,
+      "loss": 2.101,
+      "step": 2130
+    },
+    {
+      "epoch": 0.2241977906365071,
+      "grad_norm": 0.9968559741973877,
+      "learning_rate": 0.00018087765646391008,
+      "loss": 1.94,
+      "step": 2131
+    },
+    {
+      "epoch": 0.22430299842188323,
+      "grad_norm": 1.0076571702957153,
+      "learning_rate": 0.00018085761032115736,
+      "loss": 2.1199,
+      "step": 2132
+    },
+    {
+      "epoch": 0.22440820620725935,
+      "grad_norm": 1.0761382579803467,
+      "learning_rate": 0.00018083755478867658,
+      "loss": 1.8256,
+      "step": 2133
+    },
+    {
+      "epoch": 0.22451341399263547,
+      "grad_norm": 1.2399243116378784,
+      "learning_rate": 0.00018081748986879679,
+      "loss": 1.6247,
+      "step": 2134
+    },
+    {
+      "epoch": 0.22461862177801156,
+      "grad_norm": 0.6975569725036621,
+      "learning_rate": 0.00018079741556384802,
+      "loss": 2.0154,
+      "step": 2135
+    },
+    {
+      "epoch": 0.22472382956338768,
+      "grad_norm": 0.8129298686981201,
+      "learning_rate": 0.00018077733187616142,
+      "loss": 1.9586,
+      "step": 2136
+    },
+    {
+      "epoch": 0.2248290373487638,
+      "grad_norm": 1.2736451625823975,
+      "learning_rate": 0.0001807572388080693,
+      "loss": 2.3598,
+      "step": 2137
+    },
+    {
+      "epoch": 0.22493424513413993,
+      "grad_norm": 1.5190147161483765,
+      "learning_rate": 0.00018073713636190494,
+      "loss": 2.1019,
+      "step": 2138
+    },
+    {
+      "epoch": 0.22503945291951605,
+      "grad_norm": 1.2464752197265625,
+      "learning_rate": 0.0001807170245400028,
+      "loss": 2.0247,
+      "step": 2139
+    },
+    {
+      "epoch": 0.22514466070489217,
+      "grad_norm": 1.5845669507980347,
+      "learning_rate": 0.0001806969033446984,
+      "loss": 1.6235,
+      "step": 2140
+    },
+    {
+      "epoch": 0.2252498684902683,
+      "grad_norm": 1.4465845823287964,
+      "learning_rate": 0.00018067677277832834,
+      "loss": 1.8367,
+      "step": 2141
+    },
+    {
+      "epoch": 0.2253550762756444,
+      "grad_norm": 1.5385208129882812,
+      "learning_rate": 0.00018065663284323028,
+      "loss": 1.9726,
+      "step": 2142
+    },
+    {
+      "epoch": 0.2254602840610205,
+      "grad_norm": 1.2862590551376343,
+      "learning_rate": 0.0001806364835417431,
+      "loss": 2.216,
+      "step": 2143
+    },
+    {
+      "epoch": 0.22556549184639663,
+      "grad_norm": 1.4813519716262817,
+      "learning_rate": 0.0001806163248762066,
+      "loss": 2.0377,
+      "step": 2144
+    },
+    {
+      "epoch": 0.22567069963177275,
+      "grad_norm": 3.828671455383301,
+      "learning_rate": 0.00018059615684896176,
+      "loss": 2.1483,
+      "step": 2145
+    },
+    {
+      "epoch": 0.22577590741714887,
+      "grad_norm": 1.3726669549942017,
+      "learning_rate": 0.00018057597946235062,
+      "loss": 1.8516,
+      "step": 2146
+    },
+    {
+      "epoch": 0.225881115202525,
+      "grad_norm": 1.2943693399429321,
+      "learning_rate": 0.0001805557927187163,
+      "loss": 1.9399,
+      "step": 2147
+    },
+    {
+      "epoch": 0.2259863229879011,
+      "grad_norm": 1.1112357378005981,
+      "learning_rate": 0.00018053559662040302,
+      "loss": 2.064,
+      "step": 2148
+    },
+    {
+      "epoch": 0.22609153077327723,
+      "grad_norm": 1.11170494556427,
+      "learning_rate": 0.00018051539116975613,
+      "loss": 1.9364,
+      "step": 2149
+    },
+    {
+      "epoch": 0.22619673855865335,
+      "grad_norm": 0.9152960181236267,
+      "learning_rate": 0.000180495176369122,
+      "loss": 2.0502,
+      "step": 2150
+    },
+    {
+      "epoch": 0.22630194634402945,
+      "grad_norm": 1.5197174549102783,
+      "learning_rate": 0.00018047495222084812,
+      "loss": 2.0064,
+      "step": 2151
+    },
+    {
+      "epoch": 0.22640715412940557,
+      "grad_norm": 1.3327728509902954,
+      "learning_rate": 0.000180454718727283,
+      "loss": 1.2452,
+      "step": 2152
+    },
+    {
+      "epoch": 0.2265123619147817,
+      "grad_norm": 1.113082766532898,
+      "learning_rate": 0.00018043447589077634,
+      "loss": 1.7021,
+      "step": 2153
+    },
+    {
+      "epoch": 0.2266175697001578,
+      "grad_norm": 1.596432089805603,
+      "learning_rate": 0.00018041422371367885,
+      "loss": 1.9555,
+      "step": 2154
+    },
+    {
+      "epoch": 0.22672277748553393,
+      "grad_norm": 1.9314936399459839,
+      "learning_rate": 0.00018039396219834237,
+      "loss": 1.774,
+      "step": 2155
+    },
+    {
+      "epoch": 0.22682798527091005,
+      "grad_norm": 1.492937445640564,
+      "learning_rate": 0.00018037369134711977,
+      "loss": 2.2366,
+      "step": 2156
+    },
+    {
+      "epoch": 0.22693319305628618,
+      "grad_norm": 1.4675370454788208,
+      "learning_rate": 0.00018035341116236507,
+      "loss": 1.9737,
+      "step": 2157
+    },
+    {
+      "epoch": 0.2270384008416623,
+      "grad_norm": 1.595471978187561,
+      "learning_rate": 0.00018033312164643332,
+      "loss": 2.141,
+      "step": 2158
+    },
+    {
+      "epoch": 0.2271436086270384,
+      "grad_norm": 1.6596059799194336,
+      "learning_rate": 0.0001803128228016807,
+      "loss": 1.5989,
+      "step": 2159
+    },
+    {
+      "epoch": 0.2272488164124145,
+      "grad_norm": 1.2517389059066772,
+      "learning_rate": 0.00018029251463046444,
+      "loss": 1.805,
+      "step": 2160
+    },
+    {
+      "epoch": 0.22735402419779063,
+      "grad_norm": 1.7200475931167603,
+      "learning_rate": 0.00018027219713514283,
+      "loss": 0.9951,
+      "step": 2161
+    },
+    {
+      "epoch": 0.22745923198316675,
+      "grad_norm": 1.616018295288086,
+      "learning_rate": 0.00018025187031807532,
+      "loss": 1.9761,
+      "step": 2162
+    },
+    {
+      "epoch": 0.22756443976854288,
+      "grad_norm": 1.3265496492385864,
+      "learning_rate": 0.00018023153418162235,
+      "loss": 1.8146,
+      "step": 2163
+    },
+    {
+      "epoch": 0.227669647553919,
+      "grad_norm": 1.3496745824813843,
+      "learning_rate": 0.0001802111887281455,
+      "loss": 1.7593,
+      "step": 2164
+    },
+    {
+      "epoch": 0.22777485533929512,
+      "grad_norm": 1.171179175376892,
+      "learning_rate": 0.0001801908339600075,
+      "loss": 1.7885,
+      "step": 2165
+    },
+    {
+      "epoch": 0.22788006312467124,
+      "grad_norm": 1.5108046531677246,
+      "learning_rate": 0.00018017046987957197,
+      "loss": 1.9684,
+      "step": 2166
+    },
+    {
+      "epoch": 0.22798527091004733,
+      "grad_norm": 1.351288914680481,
+      "learning_rate": 0.00018015009648920374,
+      "loss": 1.671,
+      "step": 2167
+    },
+    {
+      "epoch": 0.22809047869542345,
+      "grad_norm": 1.2217696905136108,
+      "learning_rate": 0.00018012971379126875,
+      "loss": 1.9302,
+      "step": 2168
+    },
+    {
+      "epoch": 0.22819568648079958,
+      "grad_norm": 1.644564151763916,
+      "learning_rate": 0.00018010932178813397,
+      "loss": 1.7754,
+      "step": 2169
+    },
+    {
+      "epoch": 0.2283008942661757,
+      "grad_norm": 1.167407751083374,
+      "learning_rate": 0.00018008892048216744,
+      "loss": 1.9996,
+      "step": 2170
+    },
+    {
+      "epoch": 0.22840610205155182,
+      "grad_norm": 1.467685341835022,
+      "learning_rate": 0.00018006850987573834,
+      "loss": 1.9082,
+      "step": 2171
+    },
+    {
+      "epoch": 0.22851130983692794,
+      "grad_norm": 1.2934365272521973,
+      "learning_rate": 0.0001800480899712168,
+      "loss": 2.4512,
+      "step": 2172
+    },
+    {
+      "epoch": 0.22861651762230406,
+      "grad_norm": 1.2714877128601074,
+      "learning_rate": 0.00018002766077097415,
+      "loss": 2.1337,
+      "step": 2173
+    },
+    {
+      "epoch": 0.22872172540768018,
+      "grad_norm": 1.2429519891738892,
+      "learning_rate": 0.0001800072222773828,
+      "loss": 1.7869,
+      "step": 2174
+    },
+    {
+      "epoch": 0.22882693319305628,
+      "grad_norm": 1.2201809883117676,
+      "learning_rate": 0.00017998677449281621,
+      "loss": 1.6531,
+      "step": 2175
+    },
+    {
+      "epoch": 0.2289321409784324,
+      "grad_norm": 1.0419690608978271,
+      "learning_rate": 0.00017996631741964888,
+      "loss": 2.3172,
+      "step": 2176
+    },
+    {
+      "epoch": 0.22903734876380852,
+      "grad_norm": 1.217642903327942,
+      "learning_rate": 0.0001799458510602564,
+      "loss": 1.8395,
+      "step": 2177
+    },
+    {
+      "epoch": 0.22914255654918464,
+      "grad_norm": 1.229682445526123,
+      "learning_rate": 0.0001799253754170155,
+      "loss": 2.1528,
+      "step": 2178
+    },
+    {
+      "epoch": 0.22924776433456076,
+      "grad_norm": 0.8546053767204285,
+      "learning_rate": 0.00017990489049230396,
+      "loss": 2.0306,
+      "step": 2179
+    },
+    {
+      "epoch": 0.22935297211993688,
+      "grad_norm": 1.6027251482009888,
+      "learning_rate": 0.0001798843962885006,
+      "loss": 2.03,
+      "step": 2180
+    },
+    {
+      "epoch": 0.229458179905313,
+      "grad_norm": 1.3929436206817627,
+      "learning_rate": 0.00017986389280798533,
+      "loss": 2.0689,
+      "step": 2181
+    },
+    {
+      "epoch": 0.22956338769068912,
+      "grad_norm": 1.221983790397644,
+      "learning_rate": 0.00017984338005313922,
+      "loss": 2.0643,
+      "step": 2182
+    },
+    {
+      "epoch": 0.22966859547606522,
+      "grad_norm": 1.0649733543395996,
+      "learning_rate": 0.00017982285802634426,
+      "loss": 1.6149,
+      "step": 2183
+    },
+    {
+      "epoch": 0.22977380326144134,
+      "grad_norm": 1.2897831201553345,
+      "learning_rate": 0.00017980232672998368,
+      "loss": 2.0364,
+      "step": 2184
+    },
+    {
+      "epoch": 0.22987901104681746,
+      "grad_norm": 0.9529754519462585,
+      "learning_rate": 0.00017978178616644166,
+      "loss": 1.6717,
+      "step": 2185
+    },
+    {
+      "epoch": 0.22998421883219358,
+      "grad_norm": 1.1082221269607544,
+      "learning_rate": 0.00017976123633810354,
+      "loss": 2.2986,
+      "step": 2186
+    },
+    {
+      "epoch": 0.2300894266175697,
+      "grad_norm": 1.5675081014633179,
+      "learning_rate": 0.00017974067724735567,
+      "loss": 1.7103,
+      "step": 2187
+    },
+    {
+      "epoch": 0.23019463440294582,
+      "grad_norm": 1.7995283603668213,
+      "learning_rate": 0.00017972010889658554,
+      "loss": 2.0048,
+      "step": 2188
+    },
+    {
+      "epoch": 0.23029984218832195,
+      "grad_norm": 1.0870211124420166,
+      "learning_rate": 0.00017969953128818168,
+      "loss": 2.0764,
+      "step": 2189
+    },
+    {
+      "epoch": 0.23040504997369807,
+      "grad_norm": 1.2951061725616455,
+      "learning_rate": 0.0001796789444245337,
+      "loss": 2.4356,
+      "step": 2190
+    },
+    {
+      "epoch": 0.23051025775907416,
+      "grad_norm": 1.5553895235061646,
+      "learning_rate": 0.00017965834830803228,
+      "loss": 2.0726,
+      "step": 2191
+    },
+    {
+      "epoch": 0.23061546554445028,
+      "grad_norm": 1.519358515739441,
+      "learning_rate": 0.0001796377429410692,
+      "loss": 2.0531,
+      "step": 2192
+    },
+    {
+      "epoch": 0.2307206733298264,
+      "grad_norm": 1.6568381786346436,
+      "learning_rate": 0.00017961712832603724,
+      "loss": 1.7599,
+      "step": 2193
+    },
+    {
+      "epoch": 0.23082588111520252,
+      "grad_norm": 0.9388352036476135,
+      "learning_rate": 0.00017959650446533037,
+      "loss": 2.2012,
+      "step": 2194
+    },
+    {
+      "epoch": 0.23093108890057865,
+      "grad_norm": 1.2496514320373535,
+      "learning_rate": 0.0001795758713613435,
+      "loss": 1.9036,
+      "step": 2195
+    },
+    {
+      "epoch": 0.23103629668595477,
+      "grad_norm": 1.169584035873413,
+      "learning_rate": 0.00017955522901647275,
+      "loss": 2.3784,
+      "step": 2196
+    },
+    {
+      "epoch": 0.2311415044713309,
+      "grad_norm": 0.9535863399505615,
+      "learning_rate": 0.00017953457743311523,
+      "loss": 1.9531,
+      "step": 2197
+    },
+    {
+      "epoch": 0.231246712256707,
+      "grad_norm": 0.8409315347671509,
+      "learning_rate": 0.00017951391661366912,
+      "loss": 1.9737,
+      "step": 2198
+    },
+    {
+      "epoch": 0.2313519200420831,
+      "grad_norm": 1.8287293910980225,
+      "learning_rate": 0.00017949324656053373,
+      "loss": 1.6721,
+      "step": 2199
+    },
+    {
+      "epoch": 0.23145712782745922,
+      "grad_norm": 1.6631537675857544,
+      "learning_rate": 0.00017947256727610935,
+      "loss": 1.6647,
+      "step": 2200
+    },
+    {
+      "epoch": 0.23156233561283535,
+      "grad_norm": 2.273388147354126,
+      "learning_rate": 0.00017945187876279746,
+      "loss": 1.5766,
+      "step": 2201
+    },
+    {
+      "epoch": 0.23166754339821147,
+      "grad_norm": 1.3821104764938354,
+      "learning_rate": 0.0001794311810230005,
+      "loss": 2.6111,
+      "step": 2202
+    },
+    {
+      "epoch": 0.2317727511835876,
+      "grad_norm": 0.8510130643844604,
+      "learning_rate": 0.00017941047405912203,
+      "loss": 2.1294,
+      "step": 2203
+    },
+    {
+      "epoch": 0.2318779589689637,
+      "grad_norm": 1.8961708545684814,
+      "learning_rate": 0.00017938975787356673,
+      "loss": 2.1342,
+      "step": 2204
+    },
+    {
+      "epoch": 0.23198316675433983,
+      "grad_norm": 1.116129994392395,
+      "learning_rate": 0.00017936903246874026,
+      "loss": 1.8121,
+      "step": 2205
+    },
+    {
+      "epoch": 0.23208837453971595,
+      "grad_norm": 1.1960958242416382,
+      "learning_rate": 0.0001793482978470494,
+      "loss": 1.9618,
+      "step": 2206
+    },
+    {
+      "epoch": 0.23219358232509205,
+      "grad_norm": 1.3476707935333252,
+      "learning_rate": 0.00017932755401090203,
+      "loss": 2.1292,
+      "step": 2207
+    },
+    {
+      "epoch": 0.23229879011046817,
+      "grad_norm": 1.285470962524414,
+      "learning_rate": 0.00017930680096270697,
+      "loss": 2.2751,
+      "step": 2208
+    },
+    {
+      "epoch": 0.2324039978958443,
+      "grad_norm": 1.4089267253875732,
+      "learning_rate": 0.00017928603870487434,
+      "loss": 1.8112,
+      "step": 2209
+    },
+    {
+      "epoch": 0.2325092056812204,
+      "grad_norm": 0.8862564563751221,
+      "learning_rate": 0.00017926526723981506,
+      "loss": 2.086,
+      "step": 2210
+    },
+    {
+      "epoch": 0.23261441346659653,
+      "grad_norm": 1.1878046989440918,
+      "learning_rate": 0.00017924448656994133,
+      "loss": 1.8758,
+      "step": 2211
+    },
+    {
+      "epoch": 0.23271962125197265,
+      "grad_norm": 1.1390782594680786,
+      "learning_rate": 0.00017922369669766633,
+      "loss": 1.7815,
+      "step": 2212
+    },
+    {
+      "epoch": 0.23282482903734877,
+      "grad_norm": 1.1724015474319458,
+      "learning_rate": 0.0001792028976254043,
+      "loss": 1.8632,
+      "step": 2213
+    },
+    {
+      "epoch": 0.2329300368227249,
+      "grad_norm": 1.2103734016418457,
+      "learning_rate": 0.00017918208935557058,
+      "loss": 1.739,
+      "step": 2214
+    },
+    {
+      "epoch": 0.233035244608101,
+      "grad_norm": 1.0550634860992432,
+      "learning_rate": 0.00017916127189058158,
+      "loss": 2.301,
+      "step": 2215
+    },
+    {
+      "epoch": 0.2331404523934771,
+      "grad_norm": 1.6345337629318237,
+      "learning_rate": 0.0001791404452328547,
+      "loss": 1.9167,
+      "step": 2216
+    },
+    {
+      "epoch": 0.23324566017885323,
+      "grad_norm": 0.9604158997535706,
+      "learning_rate": 0.00017911960938480858,
+      "loss": 1.8264,
+      "step": 2217
+    },
+    {
+      "epoch": 0.23335086796422935,
+      "grad_norm": 1.6715139150619507,
+      "learning_rate": 0.00017909876434886273,
+      "loss": 1.826,
+      "step": 2218
+    },
+    {
+      "epoch": 0.23345607574960547,
+      "grad_norm": 1.1755411624908447,
+      "learning_rate": 0.00017907791012743783,
+      "loss": 1.5755,
+      "step": 2219
+    },
+    {
+      "epoch": 0.2335612835349816,
+      "grad_norm": 1.8296947479248047,
+      "learning_rate": 0.00017905704672295563,
+      "loss": 1.5263,
+      "step": 2220
+    },
+    {
+      "epoch": 0.23366649132035772,
+      "grad_norm": 1.162333607673645,
+      "learning_rate": 0.00017903617413783893,
+      "loss": 2.2662,
+      "step": 2221
+    },
+    {
+      "epoch": 0.23377169910573384,
+      "grad_norm": 1.608921766281128,
+      "learning_rate": 0.0001790152923745116,
+      "loss": 2.2386,
+      "step": 2222
+    },
+    {
+      "epoch": 0.23387690689110993,
+      "grad_norm": 0.9484199285507202,
+      "learning_rate": 0.00017899440143539854,
+      "loss": 1.5288,
+      "step": 2223
+    },
+    {
+      "epoch": 0.23398211467648605,
+      "grad_norm": 0.7746570110321045,
+      "learning_rate": 0.00017897350132292577,
+      "loss": 1.7404,
+      "step": 2224
+    },
+    {
+      "epoch": 0.23408732246186217,
+      "grad_norm": 1.0333335399627686,
+      "learning_rate": 0.00017895259203952032,
+      "loss": 2.0346,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2341925302472383,
+      "grad_norm": 1.308410882949829,
+      "learning_rate": 0.00017893167358761037,
+      "loss": 2.0028,
+      "step": 2226
+    },
+    {
+      "epoch": 0.23429773803261442,
+      "grad_norm": 1.7827560901641846,
+      "learning_rate": 0.00017891074596962508,
+      "loss": 1.665,
+      "step": 2227
+    },
+    {
+      "epoch": 0.23440294581799054,
+      "grad_norm": 2.48459792137146,
+      "learning_rate": 0.0001788898091879947,
+      "loss": 1.9477,
+      "step": 2228
+    },
+    {
+      "epoch": 0.23450815360336666,
+      "grad_norm": 1.5465545654296875,
+      "learning_rate": 0.00017886886324515054,
+      "loss": 2.1428,
+      "step": 2229
+    },
+    {
+      "epoch": 0.23461336138874278,
+      "grad_norm": 1.2667497396469116,
+      "learning_rate": 0.00017884790814352502,
+      "loss": 1.6537,
+      "step": 2230
+    },
+    {
+      "epoch": 0.23471856917411887,
+      "grad_norm": 1.3898303508758545,
+      "learning_rate": 0.00017882694388555157,
+      "loss": 2.175,
+      "step": 2231
+    },
+    {
+      "epoch": 0.234823776959495,
+      "grad_norm": 1.1822909116744995,
+      "learning_rate": 0.0001788059704736647,
+      "loss": 1.6444,
+      "step": 2232
+    },
+    {
+      "epoch": 0.23492898474487112,
+      "grad_norm": 1.4326550960540771,
+      "learning_rate": 0.00017878498791029998,
+      "loss": 2.4223,
+      "step": 2233
+    },
+    {
+      "epoch": 0.23503419253024724,
+      "grad_norm": 1.0603430271148682,
+      "learning_rate": 0.00017876399619789406,
+      "loss": 1.8874,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23513940031562336,
+      "grad_norm": 0.9951735138893127,
+      "learning_rate": 0.0001787429953388846,
+      "loss": 1.9595,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23524460810099948,
+      "grad_norm": 0.889345645904541,
+      "learning_rate": 0.0001787219853357104,
+      "loss": 1.5804,
+      "step": 2236
+    },
+    {
+      "epoch": 0.2353498158863756,
+      "grad_norm": 1.0618045330047607,
+      "learning_rate": 0.00017870096619081123,
+      "loss": 1.8707,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23545502367175172,
+      "grad_norm": 0.9510637521743774,
+      "learning_rate": 0.00017867993790662804,
+      "loss": 2.1408,
+      "step": 2238
+    },
+    {
+      "epoch": 0.23556023145712782,
+      "grad_norm": 0.9027166366577148,
+      "learning_rate": 0.00017865890048560277,
+      "loss": 2.3268,
+      "step": 2239
+    },
+    {
+      "epoch": 0.23566543924250394,
+      "grad_norm": 2.181145191192627,
+      "learning_rate": 0.00017863785393017838,
+      "loss": 2.1089,
+      "step": 2240
+    },
+    {
+      "epoch": 0.23577064702788006,
+      "grad_norm": 0.9953977465629578,
+      "learning_rate": 0.00017861679824279897,
+      "loss": 2.0335,
+      "step": 2241
+    },
+    {
+      "epoch": 0.23587585481325618,
+      "grad_norm": 1.1575350761413574,
+      "learning_rate": 0.00017859573342590964,
+      "loss": 1.7528,
+      "step": 2242
+    },
+    {
+      "epoch": 0.2359810625986323,
+      "grad_norm": 1.475678563117981,
+      "learning_rate": 0.00017857465948195662,
+      "loss": 2.1524,
+      "step": 2243
+    },
+    {
+      "epoch": 0.23608627038400842,
+      "grad_norm": 1.2343095541000366,
+      "learning_rate": 0.00017855357641338712,
+      "loss": 2.0534,
+      "step": 2244
+    },
+    {
+      "epoch": 0.23619147816938454,
+      "grad_norm": 1.7605794668197632,
+      "learning_rate": 0.0001785324842226495,
+      "loss": 2.1463,
+      "step": 2245
+    },
+    {
+      "epoch": 0.23629668595476067,
+      "grad_norm": 1.5780607461929321,
+      "learning_rate": 0.00017851138291219301,
+      "loss": 1.9925,
+      "step": 2246
+    },
+    {
+      "epoch": 0.23640189374013676,
+      "grad_norm": 1.1927117109298706,
+      "learning_rate": 0.00017849027248446824,
+      "loss": 1.7036,
+      "step": 2247
+    },
+    {
+      "epoch": 0.23650710152551288,
+      "grad_norm": 0.9393693208694458,
+      "learning_rate": 0.00017846915294192654,
+      "loss": 1.765,
+      "step": 2248
+    },
+    {
+      "epoch": 0.236612309310889,
+      "grad_norm": 2.274392604827881,
+      "learning_rate": 0.00017844802428702052,
+      "loss": 1.7931,
+      "step": 2249
+    },
+    {
+      "epoch": 0.23671751709626512,
+      "grad_norm": 1.2075339555740356,
+      "learning_rate": 0.00017842688652220377,
+      "loss": 2.3247,
+      "step": 2250
+    },
+    {
+      "epoch": 0.23682272488164124,
+      "grad_norm": 1.0093438625335693,
+      "learning_rate": 0.00017840573964993093,
+      "loss": 2.0814,
+      "step": 2251
+    },
+    {
+      "epoch": 0.23692793266701737,
+      "grad_norm": 1.147470474243164,
+      "learning_rate": 0.00017838458367265772,
+      "loss": 1.8764,
+      "step": 2252
+    },
+    {
+      "epoch": 0.2370331404523935,
+      "grad_norm": 1.6974701881408691,
+      "learning_rate": 0.00017836341859284093,
+      "loss": 1.4435,
+      "step": 2253
+    },
+    {
+      "epoch": 0.2371383482377696,
+      "grad_norm": 1.7535107135772705,
+      "learning_rate": 0.00017834224441293836,
+      "loss": 1.4787,
+      "step": 2254
+    },
+    {
+      "epoch": 0.2372435560231457,
+      "grad_norm": 1.3474122285842896,
+      "learning_rate": 0.00017832106113540897,
+      "loss": 1.3433,
+      "step": 2255
+    },
+    {
+      "epoch": 0.23734876380852182,
+      "grad_norm": 1.8467276096343994,
+      "learning_rate": 0.0001782998687627126,
+      "loss": 2.1831,
+      "step": 2256
+    },
+    {
+      "epoch": 0.23745397159389794,
+      "grad_norm": 1.08902907371521,
+      "learning_rate": 0.00017827866729731035,
+      "loss": 2.0054,
+      "step": 2257
+    },
+    {
+      "epoch": 0.23755917937927407,
+      "grad_norm": 0.9274120926856995,
+      "learning_rate": 0.0001782574567416642,
+      "loss": 2.1129,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2376643871646502,
+      "grad_norm": 0.9804845452308655,
+      "learning_rate": 0.00017823623709823733,
+      "loss": 1.9947,
+      "step": 2259
+    },
+    {
+      "epoch": 0.2377695949500263,
+      "grad_norm": 2.3834564685821533,
+      "learning_rate": 0.00017821500836949386,
+      "loss": 1.8823,
+      "step": 2260
+    },
+    {
+      "epoch": 0.23787480273540243,
+      "grad_norm": 1.2635396718978882,
+      "learning_rate": 0.000178193770557899,
+      "loss": 2.2756,
+      "step": 2261
+    },
+    {
+      "epoch": 0.23798001052077855,
+      "grad_norm": 0.8009362816810608,
+      "learning_rate": 0.00017817252366591907,
+      "loss": 2.1824,
+      "step": 2262
+    },
+    {
+      "epoch": 0.23808521830615464,
+      "grad_norm": 1.8110861778259277,
+      "learning_rate": 0.0001781512676960214,
+      "loss": 1.3582,
+      "step": 2263
+    },
+    {
+      "epoch": 0.23819042609153077,
+      "grad_norm": 1.50228750705719,
+      "learning_rate": 0.00017813000265067433,
+      "loss": 1.5663,
+      "step": 2264
+    },
+    {
+      "epoch": 0.2382956338769069,
+      "grad_norm": 1.1287262439727783,
+      "learning_rate": 0.00017810872853234733,
+      "loss": 1.9131,
+      "step": 2265
+    },
+    {
+      "epoch": 0.238400841662283,
+      "grad_norm": 1.1263632774353027,
+      "learning_rate": 0.0001780874453435109,
+      "loss": 1.8182,
+      "step": 2266
+    },
+    {
+      "epoch": 0.23850604944765913,
+      "grad_norm": 1.2594873905181885,
+      "learning_rate": 0.0001780661530866366,
+      "loss": 1.8225,
+      "step": 2267
+    },
+    {
+      "epoch": 0.23861125723303525,
+      "grad_norm": 1.1252297163009644,
+      "learning_rate": 0.00017804485176419697,
+      "loss": 2.0544,
+      "step": 2268
+    },
+    {
+      "epoch": 0.23871646501841137,
+      "grad_norm": 1.2109135389328003,
+      "learning_rate": 0.00017802354137866572,
+      "loss": 2.2275,
+      "step": 2269
+    },
+    {
+      "epoch": 0.2388216728037875,
+      "grad_norm": 1.0338249206542969,
+      "learning_rate": 0.00017800222193251752,
+      "loss": 2.1604,
+      "step": 2270
+    },
+    {
+      "epoch": 0.2389268805891636,
+      "grad_norm": 1.1437997817993164,
+      "learning_rate": 0.00017798089342822816,
+      "loss": 2.406,
+      "step": 2271
+    },
+    {
+      "epoch": 0.2390320883745397,
+      "grad_norm": 1.0176033973693848,
+      "learning_rate": 0.00017795955586827442,
+      "loss": 2.3426,
+      "step": 2272
+    },
+    {
+      "epoch": 0.23913729615991583,
+      "grad_norm": 1.5589518547058105,
+      "learning_rate": 0.00017793820925513418,
+      "loss": 2.1739,
+      "step": 2273
+    },
+    {
+      "epoch": 0.23924250394529195,
+      "grad_norm": 0.8495704531669617,
+      "learning_rate": 0.00017791685359128633,
+      "loss": 2.1634,
+      "step": 2274
+    },
+    {
+      "epoch": 0.23934771173066807,
+      "grad_norm": 1.0862325429916382,
+      "learning_rate": 0.00017789548887921087,
+      "loss": 2.2329,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2394529195160442,
+      "grad_norm": 1.8624638319015503,
+      "learning_rate": 0.00017787411512138875,
+      "loss": 1.7344,
+      "step": 2276
+    },
+    {
+      "epoch": 0.23955812730142031,
+      "grad_norm": 1.0375189781188965,
+      "learning_rate": 0.0001778527323203021,
+      "loss": 1.7976,
+      "step": 2277
+    },
+    {
+      "epoch": 0.23966333508679644,
+      "grad_norm": 0.92099928855896,
+      "learning_rate": 0.000177831340478434,
+      "loss": 2.173,
+      "step": 2278
+    },
+    {
+      "epoch": 0.23976854287217253,
+      "grad_norm": 2.462179183959961,
+      "learning_rate": 0.00017780993959826865,
+      "loss": 1.6677,
+      "step": 2279
+    },
+    {
+      "epoch": 0.23987375065754865,
+      "grad_norm": 1.415610432624817,
+      "learning_rate": 0.00017778852968229123,
+      "loss": 2.1977,
+      "step": 2280
+    },
+    {
+      "epoch": 0.23997895844292477,
+      "grad_norm": 1.6684101819992065,
+      "learning_rate": 0.000177767110732988,
+      "loss": 1.7925,
+      "step": 2281
+    },
+    {
+      "epoch": 0.2400841662283009,
+      "grad_norm": 0.9390462040901184,
+      "learning_rate": 0.00017774568275284627,
+      "loss": 1.7441,
+      "step": 2282
+    },
+    {
+      "epoch": 0.24018937401367702,
+      "grad_norm": 1.377502679824829,
+      "learning_rate": 0.00017772424574435443,
+      "loss": 1.847,
+      "step": 2283
+    },
+    {
+      "epoch": 0.24029458179905314,
+      "grad_norm": 1.1089184284210205,
+      "learning_rate": 0.00017770279971000185,
+      "loss": 2.3168,
+      "step": 2284
+    },
+    {
+      "epoch": 0.24039978958442926,
+      "grad_norm": 1.1749531030654907,
+      "learning_rate": 0.00017768134465227903,
+      "loss": 2.303,
+      "step": 2285
+    },
+    {
+      "epoch": 0.24050499736980538,
+      "grad_norm": 1.2190381288528442,
+      "learning_rate": 0.00017765988057367747,
+      "loss": 1.7057,
+      "step": 2286
+    },
+    {
+      "epoch": 0.24061020515518147,
+      "grad_norm": 0.8646054267883301,
+      "learning_rate": 0.00017763840747668966,
+      "loss": 1.7069,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2407154129405576,
+      "grad_norm": 0.8684479594230652,
+      "learning_rate": 0.00017761692536380928,
+      "loss": 1.9386,
+      "step": 2288
+    },
+    {
+      "epoch": 0.24082062072593372,
+      "grad_norm": 1.0403144359588623,
+      "learning_rate": 0.00017759543423753093,
+      "loss": 1.9768,
+      "step": 2289
+    },
+    {
+      "epoch": 0.24092582851130984,
+      "grad_norm": 1.298659086227417,
+      "learning_rate": 0.00017757393410035033,
+      "loss": 1.9757,
+      "step": 2290
+    },
+    {
+      "epoch": 0.24103103629668596,
+      "grad_norm": 1.4886418581008911,
+      "learning_rate": 0.00017755242495476418,
+      "loss": 1.6223,
+      "step": 2291
+    },
+    {
+      "epoch": 0.24113624408206208,
+      "grad_norm": 1.2547330856323242,
+      "learning_rate": 0.00017753090680327032,
+      "loss": 2.0024,
+      "step": 2292
+    },
+    {
+      "epoch": 0.2412414518674382,
+      "grad_norm": 1.3677291870117188,
+      "learning_rate": 0.00017750937964836755,
+      "loss": 1.797,
+      "step": 2293
+    },
+    {
+      "epoch": 0.24134665965281432,
+      "grad_norm": 1.4770119190216064,
+      "learning_rate": 0.00017748784349255577,
+      "loss": 2.2053,
+      "step": 2294
+    },
+    {
+      "epoch": 0.24145186743819042,
+      "grad_norm": 0.817295491695404,
+      "learning_rate": 0.00017746629833833585,
+      "loss": 1.9076,
+      "step": 2295
+    },
+    {
+      "epoch": 0.24155707522356654,
+      "grad_norm": 1.4393826723098755,
+      "learning_rate": 0.00017744474418820985,
+      "loss": 1.6945,
+      "step": 2296
+    },
+    {
+      "epoch": 0.24166228300894266,
+      "grad_norm": 1.3872381448745728,
+      "learning_rate": 0.00017742318104468067,
+      "loss": 1.8003,
+      "step": 2297
+    },
+    {
+      "epoch": 0.24176749079431878,
+      "grad_norm": 1.1846232414245605,
+      "learning_rate": 0.00017740160891025245,
+      "loss": 1.9204,
+      "step": 2298
+    },
+    {
+      "epoch": 0.2418726985796949,
+      "grad_norm": 0.8881604075431824,
+      "learning_rate": 0.00017738002778743027,
+      "loss": 2.1186,
+      "step": 2299
+    },
+    {
+      "epoch": 0.24197790636507102,
+      "grad_norm": 1.211893081665039,
+      "learning_rate": 0.00017735843767872024,
+      "loss": 2.0164,
+      "step": 2300
+    },
+    {
+      "epoch": 0.24208311415044714,
+      "grad_norm": 1.456984281539917,
+      "learning_rate": 0.0001773368385866296,
+      "loss": 1.7958,
+      "step": 2301
+    },
+    {
+      "epoch": 0.24218832193582326,
+      "grad_norm": 1.3485262393951416,
+      "learning_rate": 0.00017731523051366658,
+      "loss": 1.935,
+      "step": 2302
+    },
+    {
+      "epoch": 0.24229352972119936,
+      "grad_norm": 1.3391015529632568,
+      "learning_rate": 0.0001772936134623404,
+      "loss": 1.9303,
+      "step": 2303
+    },
+    {
+      "epoch": 0.24239873750657548,
+      "grad_norm": 1.4360216856002808,
+      "learning_rate": 0.00017727198743516142,
+      "loss": 1.9268,
+      "step": 2304
+    },
+    {
+      "epoch": 0.2425039452919516,
+      "grad_norm": 1.5892516374588013,
+      "learning_rate": 0.00017725035243464099,
+      "loss": 1.6656,
+      "step": 2305
+    },
+    {
+      "epoch": 0.24260915307732772,
+      "grad_norm": 1.308066964149475,
+      "learning_rate": 0.0001772287084632915,
+      "loss": 1.7519,
+      "step": 2306
+    },
+    {
+      "epoch": 0.24271436086270384,
+      "grad_norm": 1.079184651374817,
+      "learning_rate": 0.00017720705552362644,
+      "loss": 2.3011,
+      "step": 2307
+    },
+    {
+      "epoch": 0.24281956864807996,
+      "grad_norm": 1.0954896211624146,
+      "learning_rate": 0.00017718539361816023,
+      "loss": 2.275,
+      "step": 2308
+    },
+    {
+      "epoch": 0.24292477643345609,
+      "grad_norm": 1.0315781831741333,
+      "learning_rate": 0.00017716372274940843,
+      "loss": 1.8872,
+      "step": 2309
+    },
+    {
+      "epoch": 0.2430299842188322,
+      "grad_norm": 0.9103153944015503,
+      "learning_rate": 0.00017714204291988762,
+      "loss": 2.0967,
+      "step": 2310
+    },
+    {
+      "epoch": 0.2431351920042083,
+      "grad_norm": 1.182989239692688,
+      "learning_rate": 0.00017712035413211535,
+      "loss": 1.6962,
+      "step": 2311
+    },
+    {
+      "epoch": 0.24324039978958442,
+      "grad_norm": 0.8709691166877747,
+      "learning_rate": 0.00017709865638861034,
+      "loss": 1.6754,
+      "step": 2312
+    },
+    {
+      "epoch": 0.24334560757496054,
+      "grad_norm": 1.3362677097320557,
+      "learning_rate": 0.0001770769496918922,
+      "loss": 1.6432,
+      "step": 2313
+    },
+    {
+      "epoch": 0.24345081536033666,
+      "grad_norm": 1.2893030643463135,
+      "learning_rate": 0.00017705523404448176,
+      "loss": 2.0481,
+      "step": 2314
+    },
+    {
+      "epoch": 0.24355602314571279,
+      "grad_norm": 0.913367748260498,
+      "learning_rate": 0.00017703350944890068,
+      "loss": 2.1723,
+      "step": 2315
+    },
+    {
+      "epoch": 0.2436612309310889,
+      "grad_norm": 0.9429225325584412,
+      "learning_rate": 0.00017701177590767183,
+      "loss": 2.47,
+      "step": 2316
+    },
+    {
+      "epoch": 0.24376643871646503,
+      "grad_norm": 0.9314096570014954,
+      "learning_rate": 0.00017699003342331904,
+      "loss": 2.2355,
+      "step": 2317
+    },
+    {
+      "epoch": 0.24387164650184115,
+      "grad_norm": 1.0147193670272827,
+      "learning_rate": 0.0001769682819983672,
+      "loss": 2.2912,
+      "step": 2318
+    },
+    {
+      "epoch": 0.24397685428721724,
+      "grad_norm": 0.9757153987884521,
+      "learning_rate": 0.00017694652163534222,
+      "loss": 1.817,
+      "step": 2319
+    },
+    {
+      "epoch": 0.24408206207259336,
+      "grad_norm": 1.4220627546310425,
+      "learning_rate": 0.00017692475233677105,
+      "loss": 2.0166,
+      "step": 2320
+    },
+    {
+      "epoch": 0.24418726985796949,
+      "grad_norm": 1.103240728378296,
+      "learning_rate": 0.0001769029741051817,
+      "loss": 1.9228,
+      "step": 2321
+    },
+    {
+      "epoch": 0.2442924776433456,
+      "grad_norm": 1.1101772785186768,
+      "learning_rate": 0.0001768811869431032,
+      "loss": 2.1224,
+      "step": 2322
+    },
+    {
+      "epoch": 0.24439768542872173,
+      "grad_norm": 1.5481605529785156,
+      "learning_rate": 0.00017685939085306562,
+      "loss": 2.0526,
+      "step": 2323
+    },
+    {
+      "epoch": 0.24450289321409785,
+      "grad_norm": 1.3062520027160645,
+      "learning_rate": 0.00017683758583760008,
+      "loss": 1.8925,
+      "step": 2324
+    },
+    {
+      "epoch": 0.24460810099947397,
+      "grad_norm": 1.5073399543762207,
+      "learning_rate": 0.00017681577189923873,
+      "loss": 2.0869,
+      "step": 2325
+    },
+    {
+      "epoch": 0.2447133087848501,
+      "grad_norm": 0.8444550037384033,
+      "learning_rate": 0.00017679394904051473,
+      "loss": 2.1741,
+      "step": 2326
+    },
+    {
+      "epoch": 0.24481851657022619,
+      "grad_norm": 0.9408750534057617,
+      "learning_rate": 0.0001767721172639623,
+      "loss": 1.9257,
+      "step": 2327
+    },
+    {
+      "epoch": 0.2449237243556023,
+      "grad_norm": 1.572348713874817,
+      "learning_rate": 0.0001767502765721167,
+      "loss": 2.2418,
+      "step": 2328
+    },
+    {
+      "epoch": 0.24502893214097843,
+      "grad_norm": 1.0508718490600586,
+      "learning_rate": 0.0001767284269675142,
+      "loss": 2.1225,
+      "step": 2329
+    },
+    {
+      "epoch": 0.24513413992635455,
+      "grad_norm": 1.1722843647003174,
+      "learning_rate": 0.00017670656845269214,
+      "loss": 2.103,
+      "step": 2330
+    },
+    {
+      "epoch": 0.24523934771173067,
+      "grad_norm": 1.4209791421890259,
+      "learning_rate": 0.00017668470103018887,
+      "loss": 1.5394,
+      "step": 2331
+    },
+    {
+      "epoch": 0.2453445554971068,
+      "grad_norm": 1.2421069145202637,
+      "learning_rate": 0.00017666282470254381,
+      "loss": 2.0012,
+      "step": 2332
+    },
+    {
+      "epoch": 0.2454497632824829,
+      "grad_norm": 0.9292833209037781,
+      "learning_rate": 0.00017664093947229736,
+      "loss": 2.2945,
+      "step": 2333
+    },
+    {
+      "epoch": 0.24555497106785903,
+      "grad_norm": 1.0548170804977417,
+      "learning_rate": 0.00017661904534199097,
+      "loss": 2.0605,
+      "step": 2334
+    },
+    {
+      "epoch": 0.24566017885323513,
+      "grad_norm": 0.8535258769989014,
+      "learning_rate": 0.00017659714231416714,
+      "loss": 2.3665,
+      "step": 2335
+    },
+    {
+      "epoch": 0.24576538663861125,
+      "grad_norm": 1.2200136184692383,
+      "learning_rate": 0.00017657523039136942,
+      "loss": 2.0179,
+      "step": 2336
+    },
+    {
+      "epoch": 0.24587059442398737,
+      "grad_norm": 1.0717980861663818,
+      "learning_rate": 0.00017655330957614234,
+      "loss": 2.1129,
+      "step": 2337
+    },
+    {
+      "epoch": 0.2459758022093635,
+      "grad_norm": 0.9226016402244568,
+      "learning_rate": 0.0001765313798710315,
+      "loss": 2.2021,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2460810099947396,
+      "grad_norm": 1.2788020372390747,
+      "learning_rate": 0.00017650944127858354,
+      "loss": 1.7879,
+      "step": 2339
+    },
+    {
+      "epoch": 0.24618621778011573,
+      "grad_norm": 1.342400074005127,
+      "learning_rate": 0.00017648749380134608,
+      "loss": 1.8949,
+      "step": 2340
+    },
+    {
+      "epoch": 0.24629142556549186,
+      "grad_norm": 0.979184627532959,
+      "learning_rate": 0.00017646553744186784,
+      "loss": 1.773,
+      "step": 2341
+    },
+    {
+      "epoch": 0.24639663335086798,
+      "grad_norm": 1.4060213565826416,
+      "learning_rate": 0.00017644357220269856,
+      "loss": 2.2494,
+      "step": 2342
+    },
+    {
+      "epoch": 0.24650184113624407,
+      "grad_norm": 1.8778382539749146,
+      "learning_rate": 0.0001764215980863889,
+      "loss": 2.1433,
+      "step": 2343
+    },
+    {
+      "epoch": 0.2466070489216202,
+      "grad_norm": 1.214531421661377,
+      "learning_rate": 0.00017639961509549078,
+      "loss": 1.936,
+      "step": 2344
+    },
+    {
+      "epoch": 0.2467122567069963,
+      "grad_norm": 1.4594982862472534,
+      "learning_rate": 0.0001763776232325569,
+      "loss": 1.9461,
+      "step": 2345
+    },
+    {
+      "epoch": 0.24681746449237243,
+      "grad_norm": 1.5665651559829712,
+      "learning_rate": 0.00017635562250014112,
+      "loss": 2.5673,
+      "step": 2346
+    },
+    {
+      "epoch": 0.24692267227774856,
+      "grad_norm": 2.044776439666748,
+      "learning_rate": 0.00017633361290079837,
+      "loss": 2.6654,
+      "step": 2347
+    },
+    {
+      "epoch": 0.24702788006312468,
+      "grad_norm": 1.7862943410873413,
+      "learning_rate": 0.0001763115944370845,
+      "loss": 2.3661,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2471330878485008,
+      "grad_norm": 1.3187406063079834,
+      "learning_rate": 0.00017628956711155644,
+      "loss": 1.5839,
+      "step": 2349
+    },
+    {
+      "epoch": 0.24723829563387692,
+      "grad_norm": 1.1463435888290405,
+      "learning_rate": 0.0001762675309267722,
+      "loss": 2.1677,
+      "step": 2350
+    },
+    {
+      "epoch": 0.247343503419253,
+      "grad_norm": 1.8317863941192627,
+      "learning_rate": 0.00017624548588529072,
+      "loss": 2.1135,
+      "step": 2351
+    },
+    {
+      "epoch": 0.24744871120462913,
+      "grad_norm": 1.2014631032943726,
+      "learning_rate": 0.00017622343198967202,
+      "loss": 1.9363,
+      "step": 2352
+    },
+    {
+      "epoch": 0.24755391899000526,
+      "grad_norm": 1.3410941362380981,
+      "learning_rate": 0.00017620136924247719,
+      "loss": 2.149,
+      "step": 2353
+    },
+    {
+      "epoch": 0.24765912677538138,
+      "grad_norm": 1.2935651540756226,
+      "learning_rate": 0.00017617929764626825,
+      "loss": 2.3697,
+      "step": 2354
+    },
+    {
+      "epoch": 0.2477643345607575,
+      "grad_norm": 1.6687452793121338,
+      "learning_rate": 0.00017615721720360834,
+      "loss": 1.979,
+      "step": 2355
+    },
+    {
+      "epoch": 0.24786954234613362,
+      "grad_norm": 1.249712347984314,
+      "learning_rate": 0.00017613512791706155,
+      "loss": 1.6578,
+      "step": 2356
+    },
+    {
+      "epoch": 0.24797475013150974,
+      "grad_norm": 1.1926326751708984,
+      "learning_rate": 0.0001761130297891931,
+      "loss": 2.3114,
+      "step": 2357
+    },
+    {
+      "epoch": 0.24807995791688586,
+      "grad_norm": 1.159462332725525,
+      "learning_rate": 0.00017609092282256912,
+      "loss": 1.8345,
+      "step": 2358
+    },
+    {
+      "epoch": 0.24818516570226196,
+      "grad_norm": 1.2362000942230225,
+      "learning_rate": 0.00017606880701975683,
+      "loss": 2.6279,
+      "step": 2359
+    },
+    {
+      "epoch": 0.24829037348763808,
+      "grad_norm": 1.116899013519287,
+      "learning_rate": 0.00017604668238332448,
+      "loss": 2.3413,
+      "step": 2360
+    },
+    {
+      "epoch": 0.2483955812730142,
+      "grad_norm": 1.3379223346710205,
+      "learning_rate": 0.0001760245489158413,
+      "loss": 1.7864,
+      "step": 2361
+    },
+    {
+      "epoch": 0.24850078905839032,
+      "grad_norm": 1.2308326959609985,
+      "learning_rate": 0.00017600240661987763,
+      "loss": 2.0573,
+      "step": 2362
+    },
+    {
+      "epoch": 0.24860599684376644,
+      "grad_norm": 1.1983873844146729,
+      "learning_rate": 0.00017598025549800473,
+      "loss": 1.7869,
+      "step": 2363
+    },
+    {
+      "epoch": 0.24871120462914256,
+      "grad_norm": 1.1547279357910156,
+      "learning_rate": 0.00017595809555279494,
+      "loss": 1.9424,
+      "step": 2364
+    },
+    {
+      "epoch": 0.24881641241451868,
+      "grad_norm": 1.0015462636947632,
+      "learning_rate": 0.00017593592678682166,
+      "loss": 2.0275,
+      "step": 2365
+    },
+    {
+      "epoch": 0.2489216201998948,
+      "grad_norm": 0.9650968313217163,
+      "learning_rate": 0.00017591374920265923,
+      "loss": 1.815,
+      "step": 2366
+    },
+    {
+      "epoch": 0.2490268279852709,
+      "grad_norm": 1.2822868824005127,
+      "learning_rate": 0.00017589156280288311,
+      "loss": 2.1693,
+      "step": 2367
+    },
+    {
+      "epoch": 0.24913203577064702,
+      "grad_norm": 1.4724416732788086,
+      "learning_rate": 0.00017586936759006968,
+      "loss": 2.15,
+      "step": 2368
+    },
+    {
+      "epoch": 0.24923724355602314,
+      "grad_norm": 1.5108137130737305,
+      "learning_rate": 0.00017584716356679647,
+      "loss": 1.6316,
+      "step": 2369
+    },
+    {
+      "epoch": 0.24934245134139926,
+      "grad_norm": 1.047283411026001,
+      "learning_rate": 0.0001758249507356419,
+      "loss": 1.9975,
+      "step": 2370
+    },
+    {
+      "epoch": 0.24944765912677538,
+      "grad_norm": 1.0887430906295776,
+      "learning_rate": 0.00017580272909918545,
+      "loss": 1.6365,
+      "step": 2371
+    },
+    {
+      "epoch": 0.2495528669121515,
+      "grad_norm": 0.8896718621253967,
+      "learning_rate": 0.0001757804986600077,
+      "loss": 1.9713,
+      "step": 2372
+    },
+    {
+      "epoch": 0.24965807469752763,
+      "grad_norm": 0.9527761340141296,
+      "learning_rate": 0.00017575825942069018,
+      "loss": 1.6953,
+      "step": 2373
+    },
+    {
+      "epoch": 0.24976328248290375,
+      "grad_norm": 1.2873328924179077,
+      "learning_rate": 0.00017573601138381548,
+      "loss": 2.5118,
+      "step": 2374
+    },
+    {
+      "epoch": 0.24986849026827984,
+      "grad_norm": 0.9107836484909058,
+      "learning_rate": 0.00017571375455196714,
+      "loss": 1.8383,
+      "step": 2375
+    },
+    {
+      "epoch": 0.24997369805365596,
+      "grad_norm": 1.5500174760818481,
+      "learning_rate": 0.00017569148892772983,
+      "loss": 1.7582,
+      "step": 2376
+    },
+    {
+      "epoch": 0.2500789058390321,
+      "grad_norm": 1.8706609010696411,
+      "learning_rate": 0.0001756692145136891,
+      "loss": 1.4764,
+      "step": 2377
+    },
+    {
+      "epoch": 0.25018411362440823,
+      "grad_norm": 1.1527718305587769,
+      "learning_rate": 0.00017564693131243172,
+      "loss": 2.3996,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2502893214097843,
+      "grad_norm": 1.112915277481079,
+      "learning_rate": 0.0001756246393265453,
+      "loss": 2.4006,
+      "step": 2379
+    },
+    {
+      "epoch": 0.2503945291951604,
+      "grad_norm": 1.4158530235290527,
+      "learning_rate": 0.00017560233855861855,
+      "loss": 2.4866,
+      "step": 2380
+    },
+    {
+      "epoch": 0.25049973698053657,
+      "grad_norm": 1.486147403717041,
+      "learning_rate": 0.00017558002901124113,
+      "loss": 1.7218,
+      "step": 2381
+    },
+    {
+      "epoch": 0.25060494476591266,
+      "grad_norm": 1.7639609575271606,
+      "learning_rate": 0.00017555771068700386,
+      "loss": 1.7781,
+      "step": 2382
+    },
+    {
+      "epoch": 0.2507101525512888,
+      "grad_norm": 1.4476866722106934,
+      "learning_rate": 0.00017553538358849844,
+      "loss": 1.8637,
+      "step": 2383
+    },
+    {
+      "epoch": 0.2508153603366649,
+      "grad_norm": 1.6569194793701172,
+      "learning_rate": 0.00017551304771831766,
+      "loss": 1.8196,
+      "step": 2384
+    },
+    {
+      "epoch": 0.25092056812204105,
+      "grad_norm": 1.3412871360778809,
+      "learning_rate": 0.0001754907030790553,
+      "loss": 1.9298,
+      "step": 2385
+    },
+    {
+      "epoch": 0.25102577590741715,
+      "grad_norm": 1.0702909231185913,
+      "learning_rate": 0.00017546834967330617,
+      "loss": 2.0842,
+      "step": 2386
+    },
+    {
+      "epoch": 0.25113098369279324,
+      "grad_norm": 0.887639045715332,
+      "learning_rate": 0.00017544598750366614,
+      "loss": 2.1446,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2512361914781694,
+      "grad_norm": 0.9364414811134338,
+      "learning_rate": 0.000175423616572732,
+      "loss": 2.218,
+      "step": 2388
+    },
+    {
+      "epoch": 0.2513413992635455,
+      "grad_norm": 1.431705355644226,
+      "learning_rate": 0.00017540123688310162,
+      "loss": 1.7036,
+      "step": 2389
+    },
+    {
+      "epoch": 0.25144660704892163,
+      "grad_norm": 1.2119871377944946,
+      "learning_rate": 0.00017537884843737392,
+      "loss": 2.1576,
+      "step": 2390
+    },
+    {
+      "epoch": 0.2515518148342977,
+      "grad_norm": 1.0442527532577515,
+      "learning_rate": 0.00017535645123814873,
+      "loss": 2.1174,
+      "step": 2391
+    },
+    {
+      "epoch": 0.2516570226196739,
+      "grad_norm": 1.7288084030151367,
+      "learning_rate": 0.000175334045288027,
+      "loss": 1.7428,
+      "step": 2392
+    },
+    {
+      "epoch": 0.25176223040504997,
+      "grad_norm": 0.9551234245300293,
+      "learning_rate": 0.00017531163058961066,
+      "loss": 1.8753,
+      "step": 2393
+    },
+    {
+      "epoch": 0.2518674381904261,
+      "grad_norm": 1.1774919033050537,
+      "learning_rate": 0.0001752892071455027,
+      "loss": 2.1542,
+      "step": 2394
+    },
+    {
+      "epoch": 0.2519726459758022,
+      "grad_norm": 1.0245354175567627,
+      "learning_rate": 0.000175266774958307,
+      "loss": 1.8211,
+      "step": 2395
+    },
+    {
+      "epoch": 0.2520778537611783,
+      "grad_norm": 1.1799890995025635,
+      "learning_rate": 0.0001752443340306286,
+      "loss": 1.9991,
+      "step": 2396
+    },
+    {
+      "epoch": 0.25218306154655445,
+      "grad_norm": 2.0197417736053467,
+      "learning_rate": 0.00017522188436507342,
+      "loss": 1.7232,
+      "step": 2397
+    },
+    {
+      "epoch": 0.25228826933193055,
+      "grad_norm": 1.235841155052185,
+      "learning_rate": 0.0001751994259642485,
+      "loss": 2.0587,
+      "step": 2398
+    },
+    {
+      "epoch": 0.2523934771173067,
+      "grad_norm": 1.3500566482543945,
+      "learning_rate": 0.00017517695883076192,
+      "loss": 1.611,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2524986849026828,
+      "grad_norm": 1.6199384927749634,
+      "learning_rate": 0.00017515448296722262,
+      "loss": 1.7509,
+      "step": 2400
+    },
+    {
+      "epoch": 0.25260389268805894,
+      "grad_norm": 1.55356764793396,
+      "learning_rate": 0.00017513199837624073,
+      "loss": 1.9007,
+      "step": 2401
+    },
+    {
+      "epoch": 0.25270910047343503,
+      "grad_norm": 1.423980474472046,
+      "learning_rate": 0.00017510950506042727,
+      "loss": 1.6977,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2528143082588111,
+      "grad_norm": 1.0989725589752197,
+      "learning_rate": 0.00017508700302239428,
+      "loss": 1.8371,
+      "step": 2403
+    },
+    {
+      "epoch": 0.2529195160441873,
+      "grad_norm": 1.6548444032669067,
+      "learning_rate": 0.00017506449226475492,
+      "loss": 1.9493,
+      "step": 2404
+    },
+    {
+      "epoch": 0.25302472382956337,
+      "grad_norm": 1.6886351108551025,
+      "learning_rate": 0.00017504197279012321,
+      "loss": 2.4116,
+      "step": 2405
+    },
+    {
+      "epoch": 0.2531299316149395,
+      "grad_norm": 0.9853400588035583,
+      "learning_rate": 0.00017501944460111436,
+      "loss": 2.1326,
+      "step": 2406
+    },
+    {
+      "epoch": 0.2532351394003156,
+      "grad_norm": 0.8209208250045776,
+      "learning_rate": 0.00017499690770034443,
+      "loss": 2.1675,
+      "step": 2407
+    },
+    {
+      "epoch": 0.25334034718569176,
+      "grad_norm": 1.097050666809082,
+      "learning_rate": 0.00017497436209043055,
+      "loss": 2.0179,
+      "step": 2408
+    },
+    {
+      "epoch": 0.25344555497106785,
+      "grad_norm": 1.007564663887024,
+      "learning_rate": 0.00017495180777399088,
+      "loss": 2.0509,
+      "step": 2409
+    },
+    {
+      "epoch": 0.253550762756444,
+      "grad_norm": 2.713827133178711,
+      "learning_rate": 0.00017492924475364462,
+      "loss": 1.6709,
+      "step": 2410
+    },
+    {
+      "epoch": 0.2536559705418201,
+      "grad_norm": 1.4967379570007324,
+      "learning_rate": 0.00017490667303201186,
+      "loss": 1.4405,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2537611783271962,
+      "grad_norm": 1.0364857912063599,
+      "learning_rate": 0.00017488409261171386,
+      "loss": 2.3166,
+      "step": 2412
+    },
+    {
+      "epoch": 0.25386638611257234,
+      "grad_norm": 1.656065583229065,
+      "learning_rate": 0.00017486150349537276,
+      "loss": 1.8276,
+      "step": 2413
+    },
+    {
+      "epoch": 0.25397159389794843,
+      "grad_norm": 1.0365979671478271,
+      "learning_rate": 0.00017483890568561173,
+      "loss": 2.1835,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2540768016833246,
+      "grad_norm": 0.7417734861373901,
+      "learning_rate": 0.0001748162991850551,
+      "loss": 2.2011,
+      "step": 2415
+    },
+    {
+      "epoch": 0.2541820094687007,
+      "grad_norm": 1.164849042892456,
+      "learning_rate": 0.00017479368399632797,
+      "loss": 2.0921,
+      "step": 2416
+    },
+    {
+      "epoch": 0.2542872172540768,
+      "grad_norm": 1.243613362312317,
+      "learning_rate": 0.0001747710601220566,
+      "loss": 1.3508,
+      "step": 2417
+    },
+    {
+      "epoch": 0.2543924250394529,
+      "grad_norm": 1.5702705383300781,
+      "learning_rate": 0.0001747484275648682,
+      "loss": 2.0768,
+      "step": 2418
+    },
+    {
+      "epoch": 0.254497632824829,
+      "grad_norm": 1.5284066200256348,
+      "learning_rate": 0.0001747257863273911,
+      "loss": 2.144,
+      "step": 2419
+    },
+    {
+      "epoch": 0.25460284061020516,
+      "grad_norm": 1.222301721572876,
+      "learning_rate": 0.0001747031364122545,
+      "loss": 1.833,
+      "step": 2420
+    },
+    {
+      "epoch": 0.25470804839558125,
+      "grad_norm": 1.341837763786316,
+      "learning_rate": 0.00017468047782208865,
+      "loss": 1.9542,
+      "step": 2421
+    },
+    {
+      "epoch": 0.2548132561809574,
+      "grad_norm": 1.10076904296875,
+      "learning_rate": 0.00017465781055952482,
+      "loss": 1.6218,
+      "step": 2422
+    },
+    {
+      "epoch": 0.2549184639663335,
+      "grad_norm": 1.2603121995925903,
+      "learning_rate": 0.0001746351346271953,
+      "loss": 2.0813,
+      "step": 2423
+    },
+    {
+      "epoch": 0.25502367175170965,
+      "grad_norm": 1.3991905450820923,
+      "learning_rate": 0.00017461245002773336,
+      "loss": 2.2184,
+      "step": 2424
+    },
+    {
+      "epoch": 0.25512887953708574,
+      "grad_norm": 0.874963641166687,
+      "learning_rate": 0.00017458975676377326,
+      "loss": 2.2639,
+      "step": 2425
+    },
+    {
+      "epoch": 0.2552340873224619,
+      "grad_norm": 1.275423288345337,
+      "learning_rate": 0.00017456705483795038,
+      "loss": 1.8021,
+      "step": 2426
+    },
+    {
+      "epoch": 0.255339295107838,
+      "grad_norm": 1.325136423110962,
+      "learning_rate": 0.00017454434425290093,
+      "loss": 2.2287,
+      "step": 2427
+    },
+    {
+      "epoch": 0.2554445028932141,
+      "grad_norm": 1.783543348312378,
+      "learning_rate": 0.00017452162501126227,
+      "loss": 1.6924,
+      "step": 2428
+    },
+    {
+      "epoch": 0.2555497106785902,
+      "grad_norm": 1.0781643390655518,
+      "learning_rate": 0.0001744988971156727,
+      "loss": 1.8557,
+      "step": 2429
+    },
+    {
+      "epoch": 0.2556549184639663,
+      "grad_norm": 1.3726019859313965,
+      "learning_rate": 0.00017447616056877148,
+      "loss": 1.7486,
+      "step": 2430
+    },
+    {
+      "epoch": 0.25576012624934247,
+      "grad_norm": 1.8802924156188965,
+      "learning_rate": 0.000174453415373199,
+      "loss": 1.8817,
+      "step": 2431
+    },
+    {
+      "epoch": 0.25586533403471856,
+      "grad_norm": 1.2403502464294434,
+      "learning_rate": 0.00017443066153159656,
+      "loss": 2.513,
+      "step": 2432
+    },
+    {
+      "epoch": 0.2559705418200947,
+      "grad_norm": 0.8978909850120544,
+      "learning_rate": 0.00017440789904660652,
+      "loss": 2.163,
+      "step": 2433
+    },
+    {
+      "epoch": 0.2560757496054708,
+      "grad_norm": 1.4527589082717896,
+      "learning_rate": 0.00017438512792087218,
+      "loss": 2.2242,
+      "step": 2434
+    },
+    {
+      "epoch": 0.2561809573908469,
+      "grad_norm": 1.3555766344070435,
+      "learning_rate": 0.00017436234815703788,
+      "loss": 2.0565,
+      "step": 2435
+    },
+    {
+      "epoch": 0.25628616517622305,
+      "grad_norm": 1.2842960357666016,
+      "learning_rate": 0.0001743395597577489,
+      "loss": 2.1045,
+      "step": 2436
+    },
+    {
+      "epoch": 0.25639137296159914,
+      "grad_norm": 0.9545568823814392,
+      "learning_rate": 0.0001743167627256517,
+      "loss": 1.7778,
+      "step": 2437
+    },
+    {
+      "epoch": 0.2564965807469753,
+      "grad_norm": 1.2293113470077515,
+      "learning_rate": 0.00017429395706339355,
+      "loss": 2.1365,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2566017885323514,
+      "grad_norm": 1.2413405179977417,
+      "learning_rate": 0.0001742711427736228,
+      "loss": 2.3555,
+      "step": 2439
+    },
+    {
+      "epoch": 0.25670699631772753,
+      "grad_norm": 0.9717143774032593,
+      "learning_rate": 0.00017424831985898883,
+      "loss": 1.5027,
+      "step": 2440
+    },
+    {
+      "epoch": 0.2568122041031036,
+      "grad_norm": 1.797363042831421,
+      "learning_rate": 0.0001742254883221419,
+      "loss": 1.6288,
+      "step": 2441
+    },
+    {
+      "epoch": 0.2569174118884798,
+      "grad_norm": 1.5854530334472656,
+      "learning_rate": 0.0001742026481657335,
+      "loss": 2.0474,
+      "step": 2442
+    },
+    {
+      "epoch": 0.25702261967385587,
+      "grad_norm": 1.1225696802139282,
+      "learning_rate": 0.0001741797993924159,
+      "loss": 2.2607,
+      "step": 2443
+    },
+    {
+      "epoch": 0.25712782745923196,
+      "grad_norm": 1.4167970418930054,
+      "learning_rate": 0.00017415694200484247,
+      "loss": 1.8007,
+      "step": 2444
+    },
+    {
+      "epoch": 0.2572330352446081,
+      "grad_norm": 1.275691032409668,
+      "learning_rate": 0.00017413407600566755,
+      "loss": 1.6061,
+      "step": 2445
+    },
+    {
+      "epoch": 0.2573382430299842,
+      "grad_norm": 1.4618197679519653,
+      "learning_rate": 0.00017411120139754652,
+      "loss": 2.1667,
+      "step": 2446
+    },
+    {
+      "epoch": 0.25744345081536035,
+      "grad_norm": 1.4069091081619263,
+      "learning_rate": 0.00017408831818313566,
+      "loss": 1.4578,
+      "step": 2447
+    },
+    {
+      "epoch": 0.25754865860073645,
+      "grad_norm": 1.2108229398727417,
+      "learning_rate": 0.0001740654263650924,
+      "loss": 2.1124,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2576538663861126,
+      "grad_norm": 1.4401681423187256,
+      "learning_rate": 0.0001740425259460751,
+      "loss": 1.7578,
+      "step": 2449
+    },
+    {
+      "epoch": 0.2577590741714887,
+      "grad_norm": 1.814821481704712,
+      "learning_rate": 0.00017401961692874304,
+      "loss": 2.2632,
+      "step": 2450
+    },
+    {
+      "epoch": 0.2578642819568648,
+      "grad_norm": 1.439861536026001,
+      "learning_rate": 0.00017399669931575663,
+      "loss": 2.0677,
+      "step": 2451
+    },
+    {
+      "epoch": 0.25796948974224093,
+      "grad_norm": 1.9982361793518066,
+      "learning_rate": 0.0001739737731097772,
+      "loss": 2.1271,
+      "step": 2452
+    },
+    {
+      "epoch": 0.258074697527617,
+      "grad_norm": 1.41429603099823,
+      "learning_rate": 0.00017395083831346707,
+      "loss": 1.7051,
+      "step": 2453
+    },
+    {
+      "epoch": 0.2581799053129932,
+      "grad_norm": 1.6925151348114014,
+      "learning_rate": 0.0001739278949294896,
+      "loss": 2.1179,
+      "step": 2454
+    },
+    {
+      "epoch": 0.25828511309836927,
+      "grad_norm": 1.9045064449310303,
+      "learning_rate": 0.0001739049429605091,
+      "loss": 1.7847,
+      "step": 2455
+    },
+    {
+      "epoch": 0.2583903208837454,
+      "grad_norm": 1.0189872980117798,
+      "learning_rate": 0.00017388198240919102,
+      "loss": 2.0259,
+      "step": 2456
+    },
+    {
+      "epoch": 0.2584955286691215,
+      "grad_norm": 1.117735505104065,
+      "learning_rate": 0.00017385901327820157,
+      "loss": 2.3478,
+      "step": 2457
+    },
+    {
+      "epoch": 0.25860073645449766,
+      "grad_norm": 1.060653567314148,
+      "learning_rate": 0.0001738360355702081,
+      "loss": 2.1214,
+      "step": 2458
+    },
+    {
+      "epoch": 0.25870594423987375,
+      "grad_norm": 1.3654861450195312,
+      "learning_rate": 0.00017381304928787897,
+      "loss": 1.7643,
+      "step": 2459
+    },
+    {
+      "epoch": 0.25881115202524985,
+      "grad_norm": 0.8313367366790771,
+      "learning_rate": 0.00017379005443388348,
+      "loss": 1.7377,
+      "step": 2460
+    },
+    {
+      "epoch": 0.258916359810626,
+      "grad_norm": 0.9707303643226624,
+      "learning_rate": 0.00017376705101089198,
+      "loss": 2.323,
+      "step": 2461
+    },
+    {
+      "epoch": 0.2590215675960021,
+      "grad_norm": 1.2245289087295532,
+      "learning_rate": 0.0001737440390215757,
+      "loss": 1.6078,
+      "step": 2462
+    },
+    {
+      "epoch": 0.25912677538137824,
+      "grad_norm": 1.0803886651992798,
+      "learning_rate": 0.00017372101846860707,
+      "loss": 1.9575,
+      "step": 2463
+    },
+    {
+      "epoch": 0.25923198316675433,
+      "grad_norm": 0.9316054582595825,
+      "learning_rate": 0.00017369798935465926,
+      "loss": 1.5473,
+      "step": 2464
+    },
+    {
+      "epoch": 0.2593371909521305,
+      "grad_norm": 1.0661040544509888,
+      "learning_rate": 0.00017367495168240667,
+      "loss": 2.1103,
+      "step": 2465
+    },
+    {
+      "epoch": 0.2594423987375066,
+      "grad_norm": 0.8651044368743896,
+      "learning_rate": 0.00017365190545452452,
+      "loss": 2.0468,
+      "step": 2466
+    },
+    {
+      "epoch": 0.25954760652288267,
+      "grad_norm": 1.3294917345046997,
+      "learning_rate": 0.00017362885067368915,
+      "loss": 1.7681,
+      "step": 2467
+    },
+    {
+      "epoch": 0.2596528143082588,
+      "grad_norm": 1.514784574508667,
+      "learning_rate": 0.0001736057873425778,
+      "loss": 2.0482,
+      "step": 2468
+    },
+    {
+      "epoch": 0.2597580220936349,
+      "grad_norm": 1.4999936819076538,
+      "learning_rate": 0.00017358271546386874,
+      "loss": 1.8877,
+      "step": 2469
+    },
+    {
+      "epoch": 0.25986322987901106,
+      "grad_norm": 0.9134921431541443,
+      "learning_rate": 0.00017355963504024123,
+      "loss": 2.1311,
+      "step": 2470
+    },
+    {
+      "epoch": 0.25996843766438715,
+      "grad_norm": 1.9929500818252563,
+      "learning_rate": 0.0001735365460743755,
+      "loss": 1.5913,
+      "step": 2471
+    },
+    {
+      "epoch": 0.2600736454497633,
+      "grad_norm": 1.1672104597091675,
+      "learning_rate": 0.00017351344856895287,
+      "loss": 1.9764,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2601788532351394,
+      "grad_norm": 1.5550178289413452,
+      "learning_rate": 0.0001734903425266555,
+      "loss": 2.003,
+      "step": 2473
+    },
+    {
+      "epoch": 0.26028406102051554,
+      "grad_norm": 0.8885960578918457,
+      "learning_rate": 0.00017346722795016665,
+      "loss": 1.7482,
+      "step": 2474
+    },
+    {
+      "epoch": 0.26038926880589164,
+      "grad_norm": 0.9775580763816833,
+      "learning_rate": 0.00017344410484217056,
+      "loss": 2.274,
+      "step": 2475
+    },
+    {
+      "epoch": 0.26049447659126773,
+      "grad_norm": 1.9509919881820679,
+      "learning_rate": 0.00017342097320535244,
+      "loss": 2.0511,
+      "step": 2476
+    },
+    {
+      "epoch": 0.2605996843766439,
+      "grad_norm": 1.423757553100586,
+      "learning_rate": 0.00017339783304239843,
+      "loss": 1.6038,
+      "step": 2477
+    },
+    {
+      "epoch": 0.26070489216202,
+      "grad_norm": 1.0820887088775635,
+      "learning_rate": 0.0001733746843559958,
+      "loss": 1.811,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2608100999473961,
+      "grad_norm": 1.143119215965271,
+      "learning_rate": 0.00017335152714883267,
+      "loss": 1.6442,
+      "step": 2479
+    },
+    {
+      "epoch": 0.2609153077327722,
+      "grad_norm": 0.8639788627624512,
+      "learning_rate": 0.00017332836142359823,
+      "loss": 2.1835,
+      "step": 2480
+    },
+    {
+      "epoch": 0.26102051551814837,
+      "grad_norm": 1.1601319313049316,
+      "learning_rate": 0.00017330518718298264,
+      "loss": 1.8785,
+      "step": 2481
+    },
+    {
+      "epoch": 0.26112572330352446,
+      "grad_norm": 1.2111576795578003,
+      "learning_rate": 0.00017328200442967706,
+      "loss": 2.1807,
+      "step": 2482
+    },
+    {
+      "epoch": 0.26123093108890055,
+      "grad_norm": 1.271608591079712,
+      "learning_rate": 0.00017325881316637362,
+      "loss": 2.0741,
+      "step": 2483
+    },
+    {
+      "epoch": 0.2613361388742767,
+      "grad_norm": 1.105197548866272,
+      "learning_rate": 0.00017323561339576543,
+      "loss": 1.9965,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2614413466596528,
+      "grad_norm": 1.9587740898132324,
+      "learning_rate": 0.00017321240512054663,
+      "loss": 1.8361,
+      "step": 2485
+    },
+    {
+      "epoch": 0.26154655444502894,
+      "grad_norm": 1.2962583303451538,
+      "learning_rate": 0.0001731891883434123,
+      "loss": 2.3446,
+      "step": 2486
+    },
+    {
+      "epoch": 0.26165176223040504,
+      "grad_norm": 1.0470434427261353,
+      "learning_rate": 0.00017316596306705853,
+      "loss": 2.0182,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2617569700157812,
+      "grad_norm": 2.4733150005340576,
+      "learning_rate": 0.0001731427292941824,
+      "loss": 2.066,
+      "step": 2488
+    },
+    {
+      "epoch": 0.2618621778011573,
+      "grad_norm": 1.580335259437561,
+      "learning_rate": 0.00017311948702748196,
+      "loss": 2.2136,
+      "step": 2489
+    },
+    {
+      "epoch": 0.26196738558653343,
+      "grad_norm": 1.2528669834136963,
+      "learning_rate": 0.0001730962362696563,
+      "loss": 2.0774,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2620725933719095,
+      "grad_norm": 1.2130522727966309,
+      "learning_rate": 0.0001730729770234054,
+      "loss": 1.7999,
+      "step": 2491
+    },
+    {
+      "epoch": 0.2621778011572856,
+      "grad_norm": 1.3857465982437134,
+      "learning_rate": 0.00017304970929143032,
+      "loss": 1.8678,
+      "step": 2492
+    },
+    {
+      "epoch": 0.26228300894266177,
+      "grad_norm": 1.4128077030181885,
+      "learning_rate": 0.00017302643307643304,
+      "loss": 2.1356,
+      "step": 2493
+    },
+    {
+      "epoch": 0.26238821672803786,
+      "grad_norm": 1.1475266218185425,
+      "learning_rate": 0.00017300314838111653,
+      "loss": 1.8912,
+      "step": 2494
+    },
+    {
+      "epoch": 0.262493424513414,
+      "grad_norm": 0.7525075674057007,
+      "learning_rate": 0.0001729798552081848,
+      "loss": 1.9335,
+      "step": 2495
+    },
+    {
+      "epoch": 0.2625986322987901,
+      "grad_norm": 1.9416919946670532,
+      "learning_rate": 0.00017295655356034284,
+      "loss": 2.0356,
+      "step": 2496
+    },
+    {
+      "epoch": 0.26270384008416625,
+      "grad_norm": 1.7642756700515747,
+      "learning_rate": 0.00017293324344029652,
+      "loss": 1.6941,
+      "step": 2497
+    },
+    {
+      "epoch": 0.26280904786954234,
+      "grad_norm": 1.2156851291656494,
+      "learning_rate": 0.00017290992485075282,
+      "loss": 1.9932,
+      "step": 2498
+    },
+    {
+      "epoch": 0.26291425565491844,
+      "grad_norm": 1.551274061203003,
+      "learning_rate": 0.00017288659779441962,
+      "loss": 2.309,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2630194634402946,
+      "grad_norm": 1.3318612575531006,
+      "learning_rate": 0.00017286326227400583,
+      "loss": 1.9814,
+      "step": 2500
+    },
+    {
+      "epoch": 0.2631246712256707,
+      "grad_norm": 0.949734628200531,
+      "learning_rate": 0.00017283991829222133,
+      "loss": 1.8459,
+      "step": 2501
+    },
+    {
+      "epoch": 0.26322987901104683,
+      "grad_norm": 1.3344646692276,
+      "learning_rate": 0.00017281656585177698,
+      "loss": 1.2661,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2633350867964229,
+      "grad_norm": 1.1102912425994873,
+      "learning_rate": 0.0001727932049553846,
+      "loss": 1.8715,
+      "step": 2503
+    },
+    {
+      "epoch": 0.2634402945817991,
+      "grad_norm": 1.1450645923614502,
+      "learning_rate": 0.00017276983560575703,
+      "loss": 1.9675,
+      "step": 2504
+    },
+    {
+      "epoch": 0.26354550236717517,
+      "grad_norm": 2.4335386753082275,
+      "learning_rate": 0.0001727464578056081,
+      "loss": 1.688,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2636507101525513,
+      "grad_norm": 1.290065884590149,
+      "learning_rate": 0.00017272307155765258,
+      "loss": 2.4197,
+      "step": 2506
+    },
+    {
+      "epoch": 0.2637559179379274,
+      "grad_norm": 1.097562313079834,
+      "learning_rate": 0.00017269967686460617,
+      "loss": 1.8908,
+      "step": 2507
+    },
+    {
+      "epoch": 0.2638611257233035,
+      "grad_norm": 1.137505292892456,
+      "learning_rate": 0.00017267627372918575,
+      "loss": 1.9866,
+      "step": 2508
+    },
+    {
+      "epoch": 0.26396633350867965,
+      "grad_norm": 1.3950042724609375,
+      "learning_rate": 0.00017265286215410893,
+      "loss": 2.0389,
+      "step": 2509
+    },
+    {
+      "epoch": 0.26407154129405574,
+      "grad_norm": 1.7362642288208008,
+      "learning_rate": 0.00017262944214209452,
+      "loss": 1.6798,
+      "step": 2510
+    },
+    {
+      "epoch": 0.2641767490794319,
+      "grad_norm": 1.443983793258667,
+      "learning_rate": 0.0001726060136958621,
+      "loss": 2.0108,
+      "step": 2511
+    },
+    {
+      "epoch": 0.264281956864808,
+      "grad_norm": 1.4797279834747314,
+      "learning_rate": 0.00017258257681813244,
+      "loss": 2.1318,
+      "step": 2512
+    },
+    {
+      "epoch": 0.26438716465018414,
+      "grad_norm": 3.060091257095337,
+      "learning_rate": 0.00017255913151162714,
+      "loss": 2.1173,
+      "step": 2513
+    },
+    {
+      "epoch": 0.26449237243556023,
+      "grad_norm": 1.2308971881866455,
+      "learning_rate": 0.00017253567777906882,
+      "loss": 1.9013,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2645975802209363,
+      "grad_norm": 1.0173864364624023,
+      "learning_rate": 0.00017251221562318108,
+      "loss": 2.111,
+      "step": 2515
+    },
+    {
+      "epoch": 0.2647027880063125,
+      "grad_norm": 1.6821645498275757,
+      "learning_rate": 0.0001724887450466885,
+      "loss": 2.4451,
+      "step": 2516
+    },
+    {
+      "epoch": 0.26480799579168857,
+      "grad_norm": 1.388883352279663,
+      "learning_rate": 0.0001724652660523167,
+      "loss": 2.4442,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2649132035770647,
+      "grad_norm": 1.0989224910736084,
+      "learning_rate": 0.00017244177864279215,
+      "loss": 1.6966,
+      "step": 2518
+    },
+    {
+      "epoch": 0.2650184113624408,
+      "grad_norm": 1.3560047149658203,
+      "learning_rate": 0.0001724182828208424,
+      "loss": 1.9355,
+      "step": 2519
+    },
+    {
+      "epoch": 0.26512361914781696,
+      "grad_norm": 0.8736411333084106,
+      "learning_rate": 0.00017239477858919594,
+      "loss": 1.9303,
+      "step": 2520
+    },
+    {
+      "epoch": 0.26522882693319305,
+      "grad_norm": 0.9820002317428589,
+      "learning_rate": 0.00017237126595058224,
+      "loss": 2.123,
+      "step": 2521
+    },
+    {
+      "epoch": 0.2653340347185692,
+      "grad_norm": 1.382918357849121,
+      "learning_rate": 0.0001723477449077317,
+      "loss": 2.0202,
+      "step": 2522
+    },
+    {
+      "epoch": 0.2654392425039453,
+      "grad_norm": 1.1429475545883179,
+      "learning_rate": 0.00017232421546337583,
+      "loss": 1.7081,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2655444502893214,
+      "grad_norm": 1.5039023160934448,
+      "learning_rate": 0.00017230067762024693,
+      "loss": 1.82,
+      "step": 2524
+    },
+    {
+      "epoch": 0.26564965807469754,
+      "grad_norm": 1.7104127407073975,
+      "learning_rate": 0.00017227713138107844,
+      "loss": 2.1606,
+      "step": 2525
+    },
+    {
+      "epoch": 0.26575486586007363,
+      "grad_norm": 1.3519093990325928,
+      "learning_rate": 0.0001722535767486047,
+      "loss": 1.8595,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2658600736454498,
+      "grad_norm": 1.0122334957122803,
+      "learning_rate": 0.000172230013725561,
+      "loss": 2.1709,
+      "step": 2527
+    },
+    {
+      "epoch": 0.2659652814308259,
+      "grad_norm": 1.7543777227401733,
+      "learning_rate": 0.00017220644231468366,
+      "loss": 2.0135,
+      "step": 2528
+    },
+    {
+      "epoch": 0.266070489216202,
+      "grad_norm": 0.9146006107330322,
+      "learning_rate": 0.00017218286251870994,
+      "loss": 1.9128,
+      "step": 2529
+    },
+    {
+      "epoch": 0.2661756970015781,
+      "grad_norm": 1.8088133335113525,
+      "learning_rate": 0.0001721592743403781,
+      "loss": 1.7378,
+      "step": 2530
+    },
+    {
+      "epoch": 0.2662809047869542,
+      "grad_norm": 0.819983720779419,
+      "learning_rate": 0.00017213567778242731,
+      "loss": 2.0829,
+      "step": 2531
+    },
+    {
+      "epoch": 0.26638611257233036,
+      "grad_norm": 0.8384549021720886,
+      "learning_rate": 0.00017211207284759784,
+      "loss": 1.9466,
+      "step": 2532
+    },
+    {
+      "epoch": 0.26649132035770645,
+      "grad_norm": 2.0535356998443604,
+      "learning_rate": 0.00017208845953863076,
+      "loss": 2.1554,
+      "step": 2533
+    },
+    {
+      "epoch": 0.2665965281430826,
+      "grad_norm": 0.8103688955307007,
+      "learning_rate": 0.00017206483785826832,
+      "loss": 2.2735,
+      "step": 2534
+    },
+    {
+      "epoch": 0.2667017359284587,
+      "grad_norm": 1.2302720546722412,
+      "learning_rate": 0.00017204120780925353,
+      "loss": 1.7633,
+      "step": 2535
+    },
+    {
+      "epoch": 0.26680694371383484,
+      "grad_norm": 1.417521595954895,
+      "learning_rate": 0.0001720175693943305,
+      "loss": 1.8596,
+      "step": 2536
+    },
+    {
+      "epoch": 0.26691215149921094,
+      "grad_norm": 1.3307616710662842,
+      "learning_rate": 0.00017199392261624429,
+      "loss": 1.6087,
+      "step": 2537
+    },
+    {
+      "epoch": 0.2670173592845871,
+      "grad_norm": 1.0965129137039185,
+      "learning_rate": 0.0001719702674777409,
+      "loss": 2.2497,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2671225670699632,
+      "grad_norm": 1.9108279943466187,
+      "learning_rate": 0.00017194660398156737,
+      "loss": 1.6981,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2672277748553393,
+      "grad_norm": 1.4486432075500488,
+      "learning_rate": 0.0001719229321304716,
+      "loss": 1.6847,
+      "step": 2540
+    },
+    {
+      "epoch": 0.2673329826407154,
+      "grad_norm": 1.0305055379867554,
+      "learning_rate": 0.00017189925192720258,
+      "loss": 2.2104,
+      "step": 2541
+    },
+    {
+      "epoch": 0.2674381904260915,
+      "grad_norm": 1.1246416568756104,
+      "learning_rate": 0.0001718755633745102,
+      "loss": 1.5966,
+      "step": 2542
+    },
+    {
+      "epoch": 0.26754339821146766,
+      "grad_norm": 1.09429931640625,
+      "learning_rate": 0.00017185186647514531,
+      "loss": 2.6052,
+      "step": 2543
+    },
+    {
+      "epoch": 0.26764860599684376,
+      "grad_norm": 1.0281459093093872,
+      "learning_rate": 0.0001718281612318598,
+      "loss": 1.998,
+      "step": 2544
+    },
+    {
+      "epoch": 0.2677538137822199,
+      "grad_norm": 1.2437368631362915,
+      "learning_rate": 0.0001718044476474064,
+      "loss": 2.2582,
+      "step": 2545
+    },
+    {
+      "epoch": 0.267859021567596,
+      "grad_norm": 1.0594416856765747,
+      "learning_rate": 0.00017178072572453896,
+      "loss": 1.4396,
+      "step": 2546
+    },
+    {
+      "epoch": 0.2679642293529721,
+      "grad_norm": 1.3116780519485474,
+      "learning_rate": 0.00017175699546601223,
+      "loss": 2.0078,
+      "step": 2547
+    },
+    {
+      "epoch": 0.26806943713834824,
+      "grad_norm": 1.8310174942016602,
+      "learning_rate": 0.00017173325687458188,
+      "loss": 1.6103,
+      "step": 2548
+    },
+    {
+      "epoch": 0.26817464492372434,
+      "grad_norm": 1.762455940246582,
+      "learning_rate": 0.00017170950995300466,
+      "loss": 1.9907,
+      "step": 2549
+    },
+    {
+      "epoch": 0.2682798527091005,
+      "grad_norm": 1.2596118450164795,
+      "learning_rate": 0.00017168575470403815,
+      "loss": 1.6975,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2683850604944766,
+      "grad_norm": 1.0478299856185913,
+      "learning_rate": 0.000171661991130441,
+      "loss": 1.7204,
+      "step": 2551
+    },
+    {
+      "epoch": 0.26849026827985273,
+      "grad_norm": 2.9845941066741943,
+      "learning_rate": 0.0001716382192349728,
+      "loss": 1.4462,
+      "step": 2552
+    },
+    {
+      "epoch": 0.2685954760652288,
+      "grad_norm": 1.6229699850082397,
+      "learning_rate": 0.00017161443902039412,
+      "loss": 1.9217,
+      "step": 2553
+    },
+    {
+      "epoch": 0.26870068385060497,
+      "grad_norm": 1.0252301692962646,
+      "learning_rate": 0.00017159065048946644,
+      "loss": 1.9798,
+      "step": 2554
+    },
+    {
+      "epoch": 0.26880589163598106,
+      "grad_norm": 1.1335787773132324,
+      "learning_rate": 0.00017156685364495226,
+      "loss": 2.2337,
+      "step": 2555
+    },
+    {
+      "epoch": 0.26891109942135716,
+      "grad_norm": 1.192686676979065,
+      "learning_rate": 0.00017154304848961504,
+      "loss": 2.0498,
+      "step": 2556
+    },
+    {
+      "epoch": 0.2690163072067333,
+      "grad_norm": 1.2302987575531006,
+      "learning_rate": 0.00017151923502621918,
+      "loss": 2.0112,
+      "step": 2557
+    },
+    {
+      "epoch": 0.2691215149921094,
+      "grad_norm": 1.11729097366333,
+      "learning_rate": 0.00017149541325753008,
+      "loss": 1.5503,
+      "step": 2558
+    },
+    {
+      "epoch": 0.26922672277748555,
+      "grad_norm": 1.276808261871338,
+      "learning_rate": 0.00017147158318631402,
+      "loss": 1.7576,
+      "step": 2559
+    },
+    {
+      "epoch": 0.26933193056286164,
+      "grad_norm": 1.1111189126968384,
+      "learning_rate": 0.0001714477448153384,
+      "loss": 1.7231,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2694371383482378,
+      "grad_norm": 1.3912397623062134,
+      "learning_rate": 0.00017142389814737142,
+      "loss": 1.9727,
+      "step": 2561
+    },
+    {
+      "epoch": 0.2695423461336139,
+      "grad_norm": 1.2298855781555176,
+      "learning_rate": 0.00017140004318518236,
+      "loss": 1.8264,
+      "step": 2562
+    },
+    {
+      "epoch": 0.26964755391899,
+      "grad_norm": 0.9467676877975464,
+      "learning_rate": 0.0001713761799315414,
+      "loss": 2.2141,
+      "step": 2563
+    },
+    {
+      "epoch": 0.26975276170436613,
+      "grad_norm": 1.3311164379119873,
+      "learning_rate": 0.00017135230838921967,
+      "loss": 1.7519,
+      "step": 2564
+    },
+    {
+      "epoch": 0.2698579694897422,
+      "grad_norm": 1.4334806203842163,
+      "learning_rate": 0.00017132842856098937,
+      "loss": 2.1874,
+      "step": 2565
+    },
+    {
+      "epoch": 0.26996317727511837,
+      "grad_norm": 1.0612467527389526,
+      "learning_rate": 0.0001713045404496235,
+      "loss": 2.026,
+      "step": 2566
+    },
+    {
+      "epoch": 0.27006838506049446,
+      "grad_norm": 2.4904496669769287,
+      "learning_rate": 0.00017128064405789618,
+      "loss": 1.9635,
+      "step": 2567
+    },
+    {
+      "epoch": 0.2701735928458706,
+      "grad_norm": 1.3276349306106567,
+      "learning_rate": 0.00017125673938858237,
+      "loss": 1.9939,
+      "step": 2568
+    },
+    {
+      "epoch": 0.2702788006312467,
+      "grad_norm": 1.099576711654663,
+      "learning_rate": 0.0001712328264444581,
+      "loss": 2.0464,
+      "step": 2569
+    },
+    {
+      "epoch": 0.27038400841662286,
+      "grad_norm": 1.4065220355987549,
+      "learning_rate": 0.00017120890522830017,
+      "loss": 2.1369,
+      "step": 2570
+    },
+    {
+      "epoch": 0.27048921620199895,
+      "grad_norm": 1.4357450008392334,
+      "learning_rate": 0.00017118497574288664,
+      "loss": 1.6478,
+      "step": 2571
+    },
+    {
+      "epoch": 0.27059442398737504,
+      "grad_norm": 1.6182183027267456,
+      "learning_rate": 0.00017116103799099625,
+      "loss": 1.3441,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2706996317727512,
+      "grad_norm": 1.831085443496704,
+      "learning_rate": 0.00017113709197540887,
+      "loss": 1.8925,
+      "step": 2573
+    },
+    {
+      "epoch": 0.2708048395581273,
+      "grad_norm": 1.2130355834960938,
+      "learning_rate": 0.0001711131376989052,
+      "loss": 1.8865,
+      "step": 2574
+    },
+    {
+      "epoch": 0.27091004734350344,
+      "grad_norm": 0.9673578143119812,
+      "learning_rate": 0.00017108917516426704,
+      "loss": 1.6841,
+      "step": 2575
+    },
+    {
+      "epoch": 0.27101525512887953,
+      "grad_norm": 1.0448040962219238,
+      "learning_rate": 0.00017106520437427708,
+      "loss": 2.3308,
+      "step": 2576
+    },
+    {
+      "epoch": 0.2711204629142557,
+      "grad_norm": 0.8160276412963867,
+      "learning_rate": 0.00017104122533171895,
+      "loss": 2.3082,
+      "step": 2577
+    },
+    {
+      "epoch": 0.27122567069963177,
+      "grad_norm": 1.134219765663147,
+      "learning_rate": 0.00017101723803937722,
+      "loss": 2.3576,
+      "step": 2578
+    },
+    {
+      "epoch": 0.27133087848500786,
+      "grad_norm": 1.3649998903274536,
+      "learning_rate": 0.00017099324250003753,
+      "loss": 2.0624,
+      "step": 2579
+    },
+    {
+      "epoch": 0.271436086270384,
+      "grad_norm": 1.4927576780319214,
+      "learning_rate": 0.00017096923871648634,
+      "loss": 1.884,
+      "step": 2580
+    },
+    {
+      "epoch": 0.2715412940557601,
+      "grad_norm": 1.5744551420211792,
+      "learning_rate": 0.00017094522669151117,
+      "loss": 1.2819,
+      "step": 2581
+    },
+    {
+      "epoch": 0.27164650184113626,
+      "grad_norm": 1.319411277770996,
+      "learning_rate": 0.00017092120642790042,
+      "loss": 1.7153,
+      "step": 2582
+    },
+    {
+      "epoch": 0.27175170962651235,
+      "grad_norm": 1.3168647289276123,
+      "learning_rate": 0.00017089717792844353,
+      "loss": 1.0528,
+      "step": 2583
+    },
+    {
+      "epoch": 0.2718569174118885,
+      "grad_norm": 1.3358632326126099,
+      "learning_rate": 0.00017087314119593078,
+      "loss": 2.0915,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2719621251972646,
+      "grad_norm": 1.1743086576461792,
+      "learning_rate": 0.00017084909623315357,
+      "loss": 1.875,
+      "step": 2585
+    },
+    {
+      "epoch": 0.27206733298264074,
+      "grad_norm": 1.6610081195831299,
+      "learning_rate": 0.00017082504304290408,
+      "loss": 2.0099,
+      "step": 2586
+    },
+    {
+      "epoch": 0.27217254076801684,
+      "grad_norm": 1.2851457595825195,
+      "learning_rate": 0.0001708009816279756,
+      "loss": 1.8666,
+      "step": 2587
+    },
+    {
+      "epoch": 0.27227774855339293,
+      "grad_norm": 0.9746414422988892,
+      "learning_rate": 0.00017077691199116223,
+      "loss": 2.0146,
+      "step": 2588
+    },
+    {
+      "epoch": 0.2723829563387691,
+      "grad_norm": 0.9980962872505188,
+      "learning_rate": 0.00017075283413525916,
+      "loss": 2.1908,
+      "step": 2589
+    },
+    {
+      "epoch": 0.27248816412414517,
+      "grad_norm": 1.0672366619110107,
+      "learning_rate": 0.00017072874806306246,
+      "loss": 2.4187,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2725933719095213,
+      "grad_norm": 1.1528050899505615,
+      "learning_rate": 0.00017070465377736914,
+      "loss": 1.5965,
+      "step": 2591
+    },
+    {
+      "epoch": 0.2726985796948974,
+      "grad_norm": 1.3461138010025024,
+      "learning_rate": 0.00017068055128097718,
+      "loss": 2.3055,
+      "step": 2592
+    },
+    {
+      "epoch": 0.27280378748027356,
+      "grad_norm": 0.876352071762085,
+      "learning_rate": 0.00017065644057668555,
+      "loss": 1.8871,
+      "step": 2593
+    },
+    {
+      "epoch": 0.27290899526564966,
+      "grad_norm": 1.1798096895217896,
+      "learning_rate": 0.00017063232166729413,
+      "loss": 1.7881,
+      "step": 2594
+    },
+    {
+      "epoch": 0.27301420305102575,
+      "grad_norm": 0.9815201163291931,
+      "learning_rate": 0.00017060819455560382,
+      "loss": 1.9022,
+      "step": 2595
+    },
+    {
+      "epoch": 0.2731194108364019,
+      "grad_norm": 1.1576168537139893,
+      "learning_rate": 0.00017058405924441636,
+      "loss": 1.8402,
+      "step": 2596
+    },
+    {
+      "epoch": 0.273224618621778,
+      "grad_norm": 1.1924322843551636,
+      "learning_rate": 0.00017055991573653454,
+      "loss": 1.7346,
+      "step": 2597
+    },
+    {
+      "epoch": 0.27332982640715414,
+      "grad_norm": 1.4050421714782715,
+      "learning_rate": 0.00017053576403476206,
+      "loss": 1.8869,
+      "step": 2598
+    },
+    {
+      "epoch": 0.27343503419253024,
+      "grad_norm": 1.221558928489685,
+      "learning_rate": 0.00017051160414190353,
+      "loss": 2.056,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2735402419779064,
+      "grad_norm": 1.141769528388977,
+      "learning_rate": 0.00017048743606076463,
+      "loss": 2.0981,
+      "step": 2600
+    },
+    {
+      "epoch": 0.2736454497632825,
+      "grad_norm": 1.7314319610595703,
+      "learning_rate": 0.0001704632597941519,
+      "loss": 1.6496,
+      "step": 2601
+    },
+    {
+      "epoch": 0.2737506575486586,
+      "grad_norm": 1.8792369365692139,
+      "learning_rate": 0.0001704390753448728,
+      "loss": 2.0167,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2738558653340347,
+      "grad_norm": 1.1898305416107178,
+      "learning_rate": 0.00017041488271573587,
+      "loss": 2.1775,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2739610731194108,
+      "grad_norm": 1.3105398416519165,
+      "learning_rate": 0.00017039068190955047,
+      "loss": 2.2253,
+      "step": 2604
+    },
+    {
+      "epoch": 0.27406628090478696,
+      "grad_norm": 1.4331053495407104,
+      "learning_rate": 0.00017036647292912696,
+      "loss": 1.6136,
+      "step": 2605
+    },
+    {
+      "epoch": 0.27417148869016306,
+      "grad_norm": 1.059443712234497,
+      "learning_rate": 0.00017034225577727667,
+      "loss": 1.6369,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2742766964755392,
+      "grad_norm": 1.0023260116577148,
+      "learning_rate": 0.00017031803045681188,
+      "loss": 2.3698,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2743819042609153,
+      "grad_norm": 2.0491862297058105,
+      "learning_rate": 0.00017029379697054573,
+      "loss": 1.6626,
+      "step": 2608
+    },
+    {
+      "epoch": 0.27448711204629145,
+      "grad_norm": 1.3866617679595947,
+      "learning_rate": 0.0001702695553212924,
+      "loss": 1.8494,
+      "step": 2609
+    },
+    {
+      "epoch": 0.27459231983166754,
+      "grad_norm": 1.2196578979492188,
+      "learning_rate": 0.00017024530551186702,
+      "loss": 1.9425,
+      "step": 2610
+    },
+    {
+      "epoch": 0.27469752761704364,
+      "grad_norm": 0.9433722496032715,
+      "learning_rate": 0.00017022104754508562,
+      "loss": 2.2177,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2748027354024198,
+      "grad_norm": 1.2304083108901978,
+      "learning_rate": 0.0001701967814237652,
+      "loss": 1.6013,
+      "step": 2612
+    },
+    {
+      "epoch": 0.2749079431877959,
+      "grad_norm": 1.4996156692504883,
+      "learning_rate": 0.0001701725071507237,
+      "loss": 1.9624,
+      "step": 2613
+    },
+    {
+      "epoch": 0.275013150973172,
+      "grad_norm": 2.1100172996520996,
+      "learning_rate": 0.00017014822472878,
+      "loss": 1.4203,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2751183587585481,
+      "grad_norm": 1.0386134386062622,
+      "learning_rate": 0.00017012393416075398,
+      "loss": 2.0622,
+      "step": 2615
+    },
+    {
+      "epoch": 0.27522356654392427,
+      "grad_norm": 2.1451127529144287,
+      "learning_rate": 0.0001700996354494664,
+      "loss": 1.6151,
+      "step": 2616
+    },
+    {
+      "epoch": 0.27532877432930036,
+      "grad_norm": 1.6166155338287354,
+      "learning_rate": 0.000170075328597739,
+      "loss": 1.8475,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2754339821146765,
+      "grad_norm": 1.1394798755645752,
+      "learning_rate": 0.00017005101360839442,
+      "loss": 1.9892,
+      "step": 2618
+    },
+    {
+      "epoch": 0.2755391899000526,
+      "grad_norm": 1.3837336301803589,
+      "learning_rate": 0.00017002669048425632,
+      "loss": 2.2187,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2756443976854287,
+      "grad_norm": 1.8999110460281372,
+      "learning_rate": 0.00017000235922814922,
+      "loss": 2.209,
+      "step": 2620
+    },
+    {
+      "epoch": 0.27574960547080485,
+      "grad_norm": 1.2286471128463745,
+      "learning_rate": 0.00016997801984289866,
+      "loss": 1.8712,
+      "step": 2621
+    },
+    {
+      "epoch": 0.27585481325618094,
+      "grad_norm": 1.0605058670043945,
+      "learning_rate": 0.00016995367233133113,
+      "loss": 2.3471,
+      "step": 2622
+    },
+    {
+      "epoch": 0.2759600210415571,
+      "grad_norm": 1.838855266571045,
+      "learning_rate": 0.00016992931669627392,
+      "loss": 1.7718,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2760652288269332,
+      "grad_norm": 1.4109746217727661,
+      "learning_rate": 0.00016990495294055548,
+      "loss": 1.8921,
+      "step": 2624
+    },
+    {
+      "epoch": 0.27617043661230933,
+      "grad_norm": 1.4461547136306763,
+      "learning_rate": 0.00016988058106700505,
+      "loss": 2.1139,
+      "step": 2625
+    },
+    {
+      "epoch": 0.2762756443976854,
+      "grad_norm": 1.5509105920791626,
+      "learning_rate": 0.00016985620107845282,
+      "loss": 1.7556,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2763808521830615,
+      "grad_norm": 0.8982523679733276,
+      "learning_rate": 0.00016983181297773,
+      "loss": 1.9906,
+      "step": 2627
+    },
+    {
+      "epoch": 0.27648605996843767,
+      "grad_norm": 1.2106822729110718,
+      "learning_rate": 0.0001698074167676687,
+      "loss": 1.9958,
+      "step": 2628
+    },
+    {
+      "epoch": 0.27659126775381376,
+      "grad_norm": 1.0443248748779297,
+      "learning_rate": 0.00016978301245110195,
+      "loss": 1.5866,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2766964755391899,
+      "grad_norm": 0.8765720129013062,
+      "learning_rate": 0.00016975860003086378,
+      "loss": 2.1551,
+      "step": 2630
+    },
+    {
+      "epoch": 0.276801683324566,
+      "grad_norm": 1.5277172327041626,
+      "learning_rate": 0.00016973417950978906,
+      "loss": 1.5894,
+      "step": 2631
+    },
+    {
+      "epoch": 0.27690689110994215,
+      "grad_norm": 1.0646064281463623,
+      "learning_rate": 0.00016970975089071371,
+      "loss": 2.0302,
+      "step": 2632
+    },
+    {
+      "epoch": 0.27701209889531825,
+      "grad_norm": 0.9117512702941895,
+      "learning_rate": 0.00016968531417647456,
+      "loss": 2.1135,
+      "step": 2633
+    },
+    {
+      "epoch": 0.2771173066806944,
+      "grad_norm": 1.429713487625122,
+      "learning_rate": 0.0001696608693699093,
+      "loss": 1.8757,
+      "step": 2634
+    },
+    {
+      "epoch": 0.2772225144660705,
+      "grad_norm": 1.1439270973205566,
+      "learning_rate": 0.00016963641647385673,
+      "loss": 1.9884,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2773277222514466,
+      "grad_norm": 0.8106045126914978,
+      "learning_rate": 0.00016961195549115637,
+      "loss": 2.0197,
+      "step": 2636
+    },
+    {
+      "epoch": 0.27743293003682273,
+      "grad_norm": 1.2242646217346191,
+      "learning_rate": 0.00016958748642464887,
+      "loss": 2.0144,
+      "step": 2637
+    },
+    {
+      "epoch": 0.2775381378221988,
+      "grad_norm": 1.4185434579849243,
+      "learning_rate": 0.00016956300927717575,
+      "loss": 2.1412,
+      "step": 2638
+    },
+    {
+      "epoch": 0.277643345607575,
+      "grad_norm": 1.560724139213562,
+      "learning_rate": 0.0001695385240515794,
+      "loss": 2.1001,
+      "step": 2639
+    },
+    {
+      "epoch": 0.27774855339295107,
+      "grad_norm": 1.4623364210128784,
+      "learning_rate": 0.00016951403075070323,
+      "loss": 1.9271,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2778537611783272,
+      "grad_norm": 0.8837911486625671,
+      "learning_rate": 0.00016948952937739155,
+      "loss": 1.621,
+      "step": 2641
+    },
+    {
+      "epoch": 0.2779589689637033,
+      "grad_norm": 1.3567240238189697,
+      "learning_rate": 0.00016946501993448968,
+      "loss": 1.9028,
+      "step": 2642
+    },
+    {
+      "epoch": 0.2780641767490794,
+      "grad_norm": 1.2848210334777832,
+      "learning_rate": 0.00016944050242484378,
+      "loss": 1.6725,
+      "step": 2643
+    },
+    {
+      "epoch": 0.27816938453445555,
+      "grad_norm": 1.0903668403625488,
+      "learning_rate": 0.00016941597685130098,
+      "loss": 1.9121,
+      "step": 2644
+    },
+    {
+      "epoch": 0.27827459231983165,
+      "grad_norm": 1.519652247428894,
+      "learning_rate": 0.0001693914432167094,
+      "loss": 2.1452,
+      "step": 2645
+    },
+    {
+      "epoch": 0.2783798001052078,
+      "grad_norm": 1.244167447090149,
+      "learning_rate": 0.000169366901523918,
+      "loss": 2.3742,
+      "step": 2646
+    },
+    {
+      "epoch": 0.2784850078905839,
+      "grad_norm": 1.5084514617919922,
+      "learning_rate": 0.00016934235177577673,
+      "loss": 1.4103,
+      "step": 2647
+    },
+    {
+      "epoch": 0.27859021567596004,
+      "grad_norm": 1.1031715869903564,
+      "learning_rate": 0.00016931779397513652,
+      "loss": 1.1998,
+      "step": 2648
+    },
+    {
+      "epoch": 0.27869542346133613,
+      "grad_norm": 1.4771610498428345,
+      "learning_rate": 0.0001692932281248491,
+      "loss": 1.7789,
+      "step": 2649
+    },
+    {
+      "epoch": 0.2788006312467123,
+      "grad_norm": 1.3364812135696411,
+      "learning_rate": 0.00016926865422776737,
+      "loss": 1.576,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2789058390320884,
+      "grad_norm": 1.3395233154296875,
+      "learning_rate": 0.00016924407228674485,
+      "loss": 2.1849,
+      "step": 2651
+    },
+    {
+      "epoch": 0.27901104681746447,
+      "grad_norm": 1.4787706136703491,
+      "learning_rate": 0.00016921948230463625,
+      "loss": 1.7253,
+      "step": 2652
+    },
+    {
+      "epoch": 0.2791162546028406,
+      "grad_norm": 1.3256549835205078,
+      "learning_rate": 0.0001691948842842971,
+      "loss": 1.7977,
+      "step": 2653
+    },
+    {
+      "epoch": 0.2792214623882167,
+      "grad_norm": 1.2367287874221802,
+      "learning_rate": 0.0001691702782285839,
+      "loss": 2.3646,
+      "step": 2654
+    },
+    {
+      "epoch": 0.27932667017359286,
+      "grad_norm": 1.422200083732605,
+      "learning_rate": 0.00016914566414035403,
+      "loss": 1.379,
+      "step": 2655
+    },
+    {
+      "epoch": 0.27943187795896895,
+      "grad_norm": 1.134897232055664,
+      "learning_rate": 0.0001691210420224659,
+      "loss": 2.3472,
+      "step": 2656
+    },
+    {
+      "epoch": 0.2795370857443451,
+      "grad_norm": 1.5080149173736572,
+      "learning_rate": 0.00016909641187777877,
+      "loss": 1.8669,
+      "step": 2657
+    },
+    {
+      "epoch": 0.2796422935297212,
+      "grad_norm": 0.9561126828193665,
+      "learning_rate": 0.00016907177370915287,
+      "loss": 2.1019,
+      "step": 2658
+    },
+    {
+      "epoch": 0.2797475013150973,
+      "grad_norm": 1.668290138244629,
+      "learning_rate": 0.00016904712751944931,
+      "loss": 2.1505,
+      "step": 2659
+    },
+    {
+      "epoch": 0.27985270910047344,
+      "grad_norm": 1.1527786254882812,
+      "learning_rate": 0.0001690224733115302,
+      "loss": 2.1848,
+      "step": 2660
+    },
+    {
+      "epoch": 0.27995791688584953,
+      "grad_norm": 1.4785900115966797,
+      "learning_rate": 0.0001689978110882586,
+      "loss": 2.2729,
+      "step": 2661
+    },
+    {
+      "epoch": 0.2800631246712257,
+      "grad_norm": 1.4332002401351929,
+      "learning_rate": 0.00016897314085249834,
+      "loss": 2.0227,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2801683324566018,
+      "grad_norm": 1.4204810857772827,
+      "learning_rate": 0.00016894846260711438,
+      "loss": 2.3247,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2802735402419779,
+      "grad_norm": 1.2798484563827515,
+      "learning_rate": 0.00016892377635497252,
+      "loss": 1.9749,
+      "step": 2664
+    },
+    {
+      "epoch": 0.280378748027354,
+      "grad_norm": 1.1675047874450684,
+      "learning_rate": 0.00016889908209893943,
+      "loss": 2.3451,
+      "step": 2665
+    },
+    {
+      "epoch": 0.28048395581273017,
+      "grad_norm": 0.8004907369613647,
+      "learning_rate": 0.00016887437984188286,
+      "loss": 2.2131,
+      "step": 2666
+    },
+    {
+      "epoch": 0.28058916359810626,
+      "grad_norm": 1.09357750415802,
+      "learning_rate": 0.00016884966958667132,
+      "loss": 2.1521,
+      "step": 2667
+    },
+    {
+      "epoch": 0.28069437138348236,
+      "grad_norm": 0.926487147808075,
+      "learning_rate": 0.00016882495133617437,
+      "loss": 2.0104,
+      "step": 2668
+    },
+    {
+      "epoch": 0.2807995791688585,
+      "grad_norm": 1.332603096961975,
+      "learning_rate": 0.0001688002250932625,
+      "loss": 1.4487,
+      "step": 2669
+    },
+    {
+      "epoch": 0.2809047869542346,
+      "grad_norm": 0.9102864265441895,
+      "learning_rate": 0.000168775490860807,
+      "loss": 2.0708,
+      "step": 2670
+    },
+    {
+      "epoch": 0.28100999473961075,
+      "grad_norm": 1.8600473403930664,
+      "learning_rate": 0.0001687507486416802,
+      "loss": 1.901,
+      "step": 2671
+    },
+    {
+      "epoch": 0.28111520252498684,
+      "grad_norm": 1.7245584726333618,
+      "learning_rate": 0.00016872599843875544,
+      "loss": 1.7357,
+      "step": 2672
+    },
+    {
+      "epoch": 0.281220410310363,
+      "grad_norm": 0.940122663974762,
+      "learning_rate": 0.00016870124025490673,
+      "loss": 1.7142,
+      "step": 2673
+    },
+    {
+      "epoch": 0.2813256180957391,
+      "grad_norm": 1.2285487651824951,
+      "learning_rate": 0.0001686764740930092,
+      "loss": 2.1235,
+      "step": 2674
+    },
+    {
+      "epoch": 0.2814308258811152,
+      "grad_norm": 1.5378491878509521,
+      "learning_rate": 0.0001686516999559389,
+      "loss": 1.9053,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2815360336664913,
+      "grad_norm": 1.1821976900100708,
+      "learning_rate": 0.00016862691784657273,
+      "loss": 2.2413,
+      "step": 2676
+    },
+    {
+      "epoch": 0.2816412414518674,
+      "grad_norm": 0.931596040725708,
+      "learning_rate": 0.0001686021277677886,
+      "loss": 1.9538,
+      "step": 2677
+    },
+    {
+      "epoch": 0.28174644923724357,
+      "grad_norm": 2.422799825668335,
+      "learning_rate": 0.00016857732972246528,
+      "loss": 2.2234,
+      "step": 2678
+    },
+    {
+      "epoch": 0.28185165702261966,
+      "grad_norm": 1.1872382164001465,
+      "learning_rate": 0.00016855252371348245,
+      "loss": 2.0544,
+      "step": 2679
+    },
+    {
+      "epoch": 0.2819568648079958,
+      "grad_norm": 0.8804065585136414,
+      "learning_rate": 0.0001685277097437208,
+      "loss": 1.9305,
+      "step": 2680
+    },
+    {
+      "epoch": 0.2820620725933719,
+      "grad_norm": 0.8961731195449829,
+      "learning_rate": 0.00016850288781606186,
+      "loss": 1.8034,
+      "step": 2681
+    },
+    {
+      "epoch": 0.28216728037874805,
+      "grad_norm": 1.362924575805664,
+      "learning_rate": 0.00016847805793338818,
+      "loss": 2.5519,
+      "step": 2682
+    },
+    {
+      "epoch": 0.28227248816412415,
+      "grad_norm": 1.105214238166809,
+      "learning_rate": 0.00016845322009858307,
+      "loss": 2.212,
+      "step": 2683
+    },
+    {
+      "epoch": 0.28237769594950024,
+      "grad_norm": 1.1936612129211426,
+      "learning_rate": 0.00016842837431453093,
+      "loss": 2.4178,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2824829037348764,
+      "grad_norm": 1.873302936553955,
+      "learning_rate": 0.000168403520584117,
+      "loss": 1.7252,
+      "step": 2685
+    },
+    {
+      "epoch": 0.2825881115202525,
+      "grad_norm": 1.440673589706421,
+      "learning_rate": 0.0001683786589102275,
+      "loss": 2.1752,
+      "step": 2686
+    },
+    {
+      "epoch": 0.28269331930562863,
+      "grad_norm": 1.075239658355713,
+      "learning_rate": 0.0001683537892957495,
+      "loss": 1.8977,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2827985270910047,
+      "grad_norm": 1.338554859161377,
+      "learning_rate": 0.00016832891174357103,
+      "loss": 1.5485,
+      "step": 2688
+    },
+    {
+      "epoch": 0.2829037348763809,
+      "grad_norm": 1.2033013105392456,
+      "learning_rate": 0.00016830402625658104,
+      "loss": 2.182,
+      "step": 2689
+    },
+    {
+      "epoch": 0.28300894266175697,
+      "grad_norm": 1.282525658607483,
+      "learning_rate": 0.00016827913283766938,
+      "loss": 1.8707,
+      "step": 2690
+    },
+    {
+      "epoch": 0.28311415044713306,
+      "grad_norm": 1.3480403423309326,
+      "learning_rate": 0.0001682542314897269,
+      "loss": 1.5163,
+      "step": 2691
+    },
+    {
+      "epoch": 0.2832193582325092,
+      "grad_norm": 1.0361692905426025,
+      "learning_rate": 0.00016822932221564524,
+      "loss": 2.2251,
+      "step": 2692
+    },
+    {
+      "epoch": 0.2833245660178853,
+      "grad_norm": 1.486932635307312,
+      "learning_rate": 0.0001682044050183171,
+      "loss": 1.9494,
+      "step": 2693
+    },
+    {
+      "epoch": 0.28342977380326145,
+      "grad_norm": 1.127858281135559,
+      "learning_rate": 0.00016817947990063598,
+      "loss": 2.0038,
+      "step": 2694
+    },
+    {
+      "epoch": 0.28353498158863755,
+      "grad_norm": 1.3391180038452148,
+      "learning_rate": 0.00016815454686549636,
+      "loss": 1.6913,
+      "step": 2695
+    },
+    {
+      "epoch": 0.2836401893740137,
+      "grad_norm": 1.6128203868865967,
+      "learning_rate": 0.00016812960591579366,
+      "loss": 1.7916,
+      "step": 2696
+    },
+    {
+      "epoch": 0.2837453971593898,
+      "grad_norm": 1.50484037399292,
+      "learning_rate": 0.00016810465705442416,
+      "loss": 1.7511,
+      "step": 2697
+    },
+    {
+      "epoch": 0.28385060494476594,
+      "grad_norm": 1.4077752828598022,
+      "learning_rate": 0.00016807970028428508,
+      "loss": 2.2318,
+      "step": 2698
+    },
+    {
+      "epoch": 0.28395581273014203,
+      "grad_norm": 1.3436038494110107,
+      "learning_rate": 0.0001680547356082746,
+      "loss": 2.0339,
+      "step": 2699
+    },
+    {
+      "epoch": 0.2840610205155181,
+      "grad_norm": 1.6516331434249878,
+      "learning_rate": 0.00016802976302929178,
+      "loss": 1.5578,
+      "step": 2700
+    },
+    {
+      "epoch": 0.2841662283008943,
+      "grad_norm": 2.065028429031372,
+      "learning_rate": 0.0001680047825502366,
+      "loss": 1.8351,
+      "step": 2701
+    },
+    {
+      "epoch": 0.28427143608627037,
+      "grad_norm": 1.5128923654556274,
+      "learning_rate": 0.00016797979417400996,
+      "loss": 1.9321,
+      "step": 2702
+    },
+    {
+      "epoch": 0.2843766438716465,
+      "grad_norm": 2.0304954051971436,
+      "learning_rate": 0.00016795479790351366,
+      "loss": 1.3774,
+      "step": 2703
+    },
+    {
+      "epoch": 0.2844818516570226,
+      "grad_norm": 1.0998564958572388,
+      "learning_rate": 0.00016792979374165046,
+      "loss": 2.1468,
+      "step": 2704
+    },
+    {
+      "epoch": 0.28458705944239876,
+      "grad_norm": 2.0375499725341797,
+      "learning_rate": 0.00016790478169132397,
+      "loss": 2.3572,
+      "step": 2705
+    },
+    {
+      "epoch": 0.28469226722777485,
+      "grad_norm": 1.1634581089019775,
+      "learning_rate": 0.00016787976175543882,
+      "loss": 2.1935,
+      "step": 2706
+    },
+    {
+      "epoch": 0.28479747501315095,
+      "grad_norm": 1.2639390230178833,
+      "learning_rate": 0.00016785473393690045,
+      "loss": 1.9523,
+      "step": 2707
+    },
+    {
+      "epoch": 0.2849026827985271,
+      "grad_norm": 1.566070795059204,
+      "learning_rate": 0.00016782969823861526,
+      "loss": 2.3277,
+      "step": 2708
+    },
+    {
+      "epoch": 0.2850078905839032,
+      "grad_norm": 1.1275237798690796,
+      "learning_rate": 0.0001678046546634906,
+      "loss": 1.8413,
+      "step": 2709
+    },
+    {
+      "epoch": 0.28511309836927934,
+      "grad_norm": 1.370336890220642,
+      "learning_rate": 0.00016777960321443463,
+      "loss": 2.0708,
+      "step": 2710
+    },
+    {
+      "epoch": 0.28521830615465543,
+      "grad_norm": 1.3864731788635254,
+      "learning_rate": 0.00016775454389435655,
+      "loss": 1.8423,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2853235139400316,
+      "grad_norm": 1.9456918239593506,
+      "learning_rate": 0.0001677294767061664,
+      "loss": 2.1698,
+      "step": 2712
+    },
+    {
+      "epoch": 0.2854287217254077,
+      "grad_norm": 1.2052172422409058,
+      "learning_rate": 0.00016770440165277516,
+      "loss": 2.0182,
+      "step": 2713
+    },
+    {
+      "epoch": 0.2855339295107838,
+      "grad_norm": 1.085540533065796,
+      "learning_rate": 0.0001676793187370947,
+      "loss": 2.1485,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2856391372961599,
+      "grad_norm": 1.0382028818130493,
+      "learning_rate": 0.0001676542279620378,
+      "loss": 1.9553,
+      "step": 2715
+    },
+    {
+      "epoch": 0.285744345081536,
+      "grad_norm": 1.1799119710922241,
+      "learning_rate": 0.0001676291293305182,
+      "loss": 1.8491,
+      "step": 2716
+    },
+    {
+      "epoch": 0.28584955286691216,
+      "grad_norm": 1.027492880821228,
+      "learning_rate": 0.00016760402284545048,
+      "loss": 2.1497,
+      "step": 2717
+    },
+    {
+      "epoch": 0.28595476065228825,
+      "grad_norm": 1.3454300165176392,
+      "learning_rate": 0.00016757890850975025,
+      "loss": 2.153,
+      "step": 2718
+    },
+    {
+      "epoch": 0.2860599684376644,
+      "grad_norm": 1.4455000162124634,
+      "learning_rate": 0.00016755378632633388,
+      "loss": 1.9732,
+      "step": 2719
+    },
+    {
+      "epoch": 0.2861651762230405,
+      "grad_norm": 1.3513553142547607,
+      "learning_rate": 0.00016752865629811873,
+      "loss": 1.5569,
+      "step": 2720
+    },
+    {
+      "epoch": 0.28627038400841665,
+      "grad_norm": 0.8183251619338989,
+      "learning_rate": 0.00016750351842802314,
+      "loss": 1.4716,
+      "step": 2721
+    },
+    {
+      "epoch": 0.28637559179379274,
+      "grad_norm": 2.2286715507507324,
+      "learning_rate": 0.00016747837271896622,
+      "loss": 1.947,
+      "step": 2722
+    },
+    {
+      "epoch": 0.28648079957916883,
+      "grad_norm": 0.9977225065231323,
+      "learning_rate": 0.00016745321917386804,
+      "loss": 2.348,
+      "step": 2723
+    },
+    {
+      "epoch": 0.286586007364545,
+      "grad_norm": 0.8259665369987488,
+      "learning_rate": 0.00016742805779564968,
+      "loss": 2.4167,
+      "step": 2724
+    },
+    {
+      "epoch": 0.2866912151499211,
+      "grad_norm": 0.9765976667404175,
+      "learning_rate": 0.00016740288858723302,
+      "loss": 1.8912,
+      "step": 2725
+    },
+    {
+      "epoch": 0.2867964229352972,
+      "grad_norm": 1.2133702039718628,
+      "learning_rate": 0.0001673777115515408,
+      "loss": 2.5914,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2869016307206733,
+      "grad_norm": 1.3183568716049194,
+      "learning_rate": 0.00016735252669149685,
+      "loss": 1.7136,
+      "step": 2727
+    },
+    {
+      "epoch": 0.28700683850604947,
+      "grad_norm": 1.4817814826965332,
+      "learning_rate": 0.00016732733401002574,
+      "loss": 1.8334,
+      "step": 2728
+    },
+    {
+      "epoch": 0.28711204629142556,
+      "grad_norm": 1.6449335813522339,
+      "learning_rate": 0.00016730213351005303,
+      "loss": 2.3261,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2872172540768017,
+      "grad_norm": 1.0374553203582764,
+      "learning_rate": 0.0001672769251945052,
+      "loss": 2.4651,
+      "step": 2730
+    },
+    {
+      "epoch": 0.2873224618621778,
+      "grad_norm": 1.436585783958435,
+      "learning_rate": 0.0001672517090663096,
+      "loss": 1.9923,
+      "step": 2731
+    },
+    {
+      "epoch": 0.2874276696475539,
+      "grad_norm": 1.1871074438095093,
+      "learning_rate": 0.00016722648512839446,
+      "loss": 2.0888,
+      "step": 2732
+    },
+    {
+      "epoch": 0.28753287743293005,
+      "grad_norm": 1.0876706838607788,
+      "learning_rate": 0.00016720125338368894,
+      "loss": 1.427,
+      "step": 2733
+    },
+    {
+      "epoch": 0.28763808521830614,
+      "grad_norm": 0.8011285662651062,
+      "learning_rate": 0.00016717601383512318,
+      "loss": 1.8155,
+      "step": 2734
+    },
+    {
+      "epoch": 0.2877432930036823,
+      "grad_norm": 1.4743174314498901,
+      "learning_rate": 0.00016715076648562814,
+      "loss": 2.1159,
+      "step": 2735
+    },
+    {
+      "epoch": 0.2878485007890584,
+      "grad_norm": 2.063098907470703,
+      "learning_rate": 0.00016712551133813572,
+      "loss": 2.3086,
+      "step": 2736
+    },
+    {
+      "epoch": 0.28795370857443453,
+      "grad_norm": 1.3487850427627563,
+      "learning_rate": 0.0001671002483955787,
+      "loss": 2.1455,
+      "step": 2737
+    },
+    {
+      "epoch": 0.2880589163598106,
+      "grad_norm": 1.5087181329727173,
+      "learning_rate": 0.00016707497766089082,
+      "loss": 1.8298,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2881641241451867,
+      "grad_norm": 1.2308396100997925,
+      "learning_rate": 0.00016704969913700662,
+      "loss": 2.4025,
+      "step": 2739
+    },
+    {
+      "epoch": 0.28826933193056287,
+      "grad_norm": 1.2315316200256348,
+      "learning_rate": 0.00016702441282686166,
+      "loss": 1.8982,
+      "step": 2740
+    },
+    {
+      "epoch": 0.28837453971593896,
+      "grad_norm": 2.0835769176483154,
+      "learning_rate": 0.00016699911873339232,
+      "loss": 1.9045,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2884797475013151,
+      "grad_norm": 0.9750843048095703,
+      "learning_rate": 0.00016697381685953596,
+      "loss": 1.8595,
+      "step": 2742
+    },
+    {
+      "epoch": 0.2885849552866912,
+      "grad_norm": 1.373468279838562,
+      "learning_rate": 0.0001669485072082308,
+      "loss": 2.2024,
+      "step": 2743
+    },
+    {
+      "epoch": 0.28869016307206735,
+      "grad_norm": 1.492432951927185,
+      "learning_rate": 0.00016692318978241594,
+      "loss": 2.0474,
+      "step": 2744
+    },
+    {
+      "epoch": 0.28879537085744345,
+      "grad_norm": 2.0323355197906494,
+      "learning_rate": 0.00016689786458503141,
+      "loss": 2.0644,
+      "step": 2745
+    },
+    {
+      "epoch": 0.2889005786428196,
+      "grad_norm": 1.1842961311340332,
+      "learning_rate": 0.0001668725316190182,
+      "loss": 2.1992,
+      "step": 2746
+    },
+    {
+      "epoch": 0.2890057864281957,
+      "grad_norm": 1.259732961654663,
+      "learning_rate": 0.00016684719088731807,
+      "loss": 1.7769,
+      "step": 2747
+    },
+    {
+      "epoch": 0.2891109942135718,
+      "grad_norm": 0.8478556275367737,
+      "learning_rate": 0.0001668218423928738,
+      "loss": 1.7708,
+      "step": 2748
+    },
+    {
+      "epoch": 0.28921620199894793,
+      "grad_norm": 0.826978325843811,
+      "learning_rate": 0.000166796486138629,
+      "loss": 1.8427,
+      "step": 2749
+    },
+    {
+      "epoch": 0.289321409784324,
+      "grad_norm": 0.9389429688453674,
+      "learning_rate": 0.00016677112212752824,
+      "loss": 1.7696,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2894266175697002,
+      "grad_norm": 1.0777794122695923,
+      "learning_rate": 0.0001667457503625169,
+      "loss": 1.8971,
+      "step": 2751
+    },
+    {
+      "epoch": 0.28953182535507627,
+      "grad_norm": 1.1226226091384888,
+      "learning_rate": 0.00016672037084654139,
+      "loss": 1.8703,
+      "step": 2752
+    },
+    {
+      "epoch": 0.2896370331404524,
+      "grad_norm": 1.2667369842529297,
+      "learning_rate": 0.0001666949835825489,
+      "loss": 1.8995,
+      "step": 2753
+    },
+    {
+      "epoch": 0.2897422409258285,
+      "grad_norm": 1.5252774953842163,
+      "learning_rate": 0.0001666695885734876,
+      "loss": 2.2547,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2898474487112046,
+      "grad_norm": 1.3169254064559937,
+      "learning_rate": 0.0001666441858223065,
+      "loss": 1.7265,
+      "step": 2755
+    },
+    {
+      "epoch": 0.28995265649658075,
+      "grad_norm": 1.177160382270813,
+      "learning_rate": 0.00016661877533195556,
+      "loss": 1.8653,
+      "step": 2756
+    },
+    {
+      "epoch": 0.29005786428195685,
+      "grad_norm": 1.028063416481018,
+      "learning_rate": 0.00016659335710538564,
+      "loss": 2.0327,
+      "step": 2757
+    },
+    {
+      "epoch": 0.290163072067333,
+      "grad_norm": 1.1455798149108887,
+      "learning_rate": 0.00016656793114554842,
+      "loss": 2.0376,
+      "step": 2758
+    },
+    {
+      "epoch": 0.2902682798527091,
+      "grad_norm": 1.1332205533981323,
+      "learning_rate": 0.00016654249745539656,
+      "loss": 1.9765,
+      "step": 2759
+    },
+    {
+      "epoch": 0.29037348763808524,
+      "grad_norm": 2.2703967094421387,
+      "learning_rate": 0.00016651705603788362,
+      "loss": 2.2095,
+      "step": 2760
+    },
+    {
+      "epoch": 0.29047869542346133,
+      "grad_norm": 1.228402853012085,
+      "learning_rate": 0.00016649160689596396,
+      "loss": 2.0485,
+      "step": 2761
+    },
+    {
+      "epoch": 0.2905839032088375,
+      "grad_norm": 0.901681125164032,
+      "learning_rate": 0.00016646615003259295,
+      "loss": 1.8973,
+      "step": 2762
+    },
+    {
+      "epoch": 0.2906891109942136,
+      "grad_norm": 1.9370111227035522,
+      "learning_rate": 0.00016644068545072682,
+      "loss": 2.3251,
+      "step": 2763
+    },
+    {
+      "epoch": 0.29079431877958967,
+      "grad_norm": 1.6907079219818115,
+      "learning_rate": 0.00016641521315332265,
+      "loss": 1.8123,
+      "step": 2764
+    },
+    {
+      "epoch": 0.2908995265649658,
+      "grad_norm": 0.9530484080314636,
+      "learning_rate": 0.0001663897331433385,
+      "loss": 1.4562,
+      "step": 2765
+    },
+    {
+      "epoch": 0.2910047343503419,
+      "grad_norm": 0.8451758623123169,
+      "learning_rate": 0.00016636424542373324,
+      "loss": 1.7969,
+      "step": 2766
+    },
+    {
+      "epoch": 0.29110994213571806,
+      "grad_norm": 0.9732938408851624,
+      "learning_rate": 0.00016633874999746667,
+      "loss": 1.7659,
+      "step": 2767
+    },
+    {
+      "epoch": 0.29121514992109415,
+      "grad_norm": 1.020836591720581,
+      "learning_rate": 0.00016631324686749958,
+      "loss": 2.0531,
+      "step": 2768
+    },
+    {
+      "epoch": 0.2913203577064703,
+      "grad_norm": 1.1106858253479004,
+      "learning_rate": 0.0001662877360367934,
+      "loss": 1.9912,
+      "step": 2769
+    },
+    {
+      "epoch": 0.2914255654918464,
+      "grad_norm": 0.9008655548095703,
+      "learning_rate": 0.0001662622175083108,
+      "loss": 1.8145,
+      "step": 2770
+    },
+    {
+      "epoch": 0.2915307732772225,
+      "grad_norm": 0.9763224124908447,
+      "learning_rate": 0.00016623669128501504,
+      "loss": 2.3397,
+      "step": 2771
+    },
+    {
+      "epoch": 0.29163598106259864,
+      "grad_norm": 1.1344367265701294,
+      "learning_rate": 0.00016621115736987046,
+      "loss": 1.7173,
+      "step": 2772
+    },
+    {
+      "epoch": 0.29174118884797473,
+      "grad_norm": 1.7770837545394897,
+      "learning_rate": 0.00016618561576584216,
+      "loss": 1.9058,
+      "step": 2773
+    },
+    {
+      "epoch": 0.2918463966333509,
+      "grad_norm": 0.8264276385307312,
+      "learning_rate": 0.00016616006647589626,
+      "loss": 2.0614,
+      "step": 2774
+    },
+    {
+      "epoch": 0.291951604418727,
+      "grad_norm": 1.0680514574050903,
+      "learning_rate": 0.0001661345095029997,
+      "loss": 1.7038,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2920568122041031,
+      "grad_norm": 1.3225088119506836,
+      "learning_rate": 0.00016610894485012033,
+      "loss": 1.9587,
+      "step": 2776
+    },
+    {
+      "epoch": 0.2921620199894792,
+      "grad_norm": 2.3858277797698975,
+      "learning_rate": 0.0001660833725202269,
+      "loss": 2.4027,
+      "step": 2777
+    },
+    {
+      "epoch": 0.29226722777485536,
+      "grad_norm": 1.472761869430542,
+      "learning_rate": 0.00016605779251628903,
+      "loss": 1.394,
+      "step": 2778
+    },
+    {
+      "epoch": 0.29237243556023146,
+      "grad_norm": 1.8934521675109863,
+      "learning_rate": 0.00016603220484127723,
+      "loss": 1.8053,
+      "step": 2779
+    },
+    {
+      "epoch": 0.29247764334560755,
+      "grad_norm": 1.2446554899215698,
+      "learning_rate": 0.00016600660949816291,
+      "loss": 2.0428,
+      "step": 2780
+    },
+    {
+      "epoch": 0.2925828511309837,
+      "grad_norm": 1.7331035137176514,
+      "learning_rate": 0.00016598100648991838,
+      "loss": 1.9776,
+      "step": 2781
+    },
+    {
+      "epoch": 0.2926880589163598,
+      "grad_norm": 1.300978422164917,
+      "learning_rate": 0.00016595539581951686,
+      "loss": 1.6838,
+      "step": 2782
+    },
+    {
+      "epoch": 0.29279326670173594,
+      "grad_norm": 1.042041540145874,
+      "learning_rate": 0.0001659297774899324,
+      "loss": 1.783,
+      "step": 2783
+    },
+    {
+      "epoch": 0.29289847448711204,
+      "grad_norm": 1.229884147644043,
+      "learning_rate": 0.00016590415150413997,
+      "loss": 1.9093,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2930036822724882,
+      "grad_norm": 1.859123706817627,
+      "learning_rate": 0.00016587851786511543,
+      "loss": 2.0027,
+      "step": 2785
+    },
+    {
+      "epoch": 0.2931088900578643,
+      "grad_norm": 1.4690340757369995,
+      "learning_rate": 0.00016585287657583557,
+      "loss": 1.4002,
+      "step": 2786
+    },
+    {
+      "epoch": 0.2932140978432404,
+      "grad_norm": 1.5459439754486084,
+      "learning_rate": 0.00016582722763927802,
+      "loss": 1.4447,
+      "step": 2787
+    },
+    {
+      "epoch": 0.2933193056286165,
+      "grad_norm": 1.0940881967544556,
+      "learning_rate": 0.00016580157105842123,
+      "loss": 1.9658,
+      "step": 2788
+    },
+    {
+      "epoch": 0.2934245134139926,
+      "grad_norm": 1.102705478668213,
+      "learning_rate": 0.00016577590683624472,
+      "loss": 2.0337,
+      "step": 2789
+    },
+    {
+      "epoch": 0.29352972119936876,
+      "grad_norm": 1.6445786952972412,
+      "learning_rate": 0.00016575023497572872,
+      "loss": 1.674,
+      "step": 2790
+    },
+    {
+      "epoch": 0.29363492898474486,
+      "grad_norm": 1.018031120300293,
+      "learning_rate": 0.00016572455547985446,
+      "loss": 2.1392,
+      "step": 2791
+    },
+    {
+      "epoch": 0.293740136770121,
+      "grad_norm": 1.7538018226623535,
+      "learning_rate": 0.00016569886835160399,
+      "loss": 1.8097,
+      "step": 2792
+    },
+    {
+      "epoch": 0.2938453445554971,
+      "grad_norm": 1.397821307182312,
+      "learning_rate": 0.00016567317359396028,
+      "loss": 2.0166,
+      "step": 2793
+    },
+    {
+      "epoch": 0.29395055234087325,
+      "grad_norm": 1.5573742389678955,
+      "learning_rate": 0.0001656474712099072,
+      "loss": 1.9029,
+      "step": 2794
+    },
+    {
+      "epoch": 0.29405576012624934,
+      "grad_norm": 0.9295175671577454,
+      "learning_rate": 0.0001656217612024294,
+      "loss": 2.1548,
+      "step": 2795
+    },
+    {
+      "epoch": 0.29416096791162544,
+      "grad_norm": 0.8610682487487793,
+      "learning_rate": 0.00016559604357451263,
+      "loss": 2.2194,
+      "step": 2796
+    },
+    {
+      "epoch": 0.2942661756970016,
+      "grad_norm": 1.385067343711853,
+      "learning_rate": 0.00016557031832914327,
+      "loss": 1.9511,
+      "step": 2797
+    },
+    {
+      "epoch": 0.2943713834823777,
+      "grad_norm": 0.9952557682991028,
+      "learning_rate": 0.00016554458546930878,
+      "loss": 2.6092,
+      "step": 2798
+    },
+    {
+      "epoch": 0.29447659126775383,
+      "grad_norm": 1.1171859502792358,
+      "learning_rate": 0.0001655188449979974,
+      "loss": 1.2882,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2945817990531299,
+      "grad_norm": 1.6443709135055542,
+      "learning_rate": 0.00016549309691819833,
+      "loss": 1.6552,
+      "step": 2800
+    },
+    {
+      "epoch": 0.29468700683850607,
+      "grad_norm": 0.9413725137710571,
+      "learning_rate": 0.00016546734123290156,
+      "loss": 2.4733,
+      "step": 2801
+    },
+    {
+      "epoch": 0.29479221462388217,
+      "grad_norm": 1.1774075031280518,
+      "learning_rate": 0.000165441577945098,
+      "loss": 1.8616,
+      "step": 2802
+    },
+    {
+      "epoch": 0.29489742240925826,
+      "grad_norm": 0.9130268692970276,
+      "learning_rate": 0.00016541580705777955,
+      "loss": 2.1065,
+      "step": 2803
+    },
+    {
+      "epoch": 0.2950026301946344,
+      "grad_norm": 1.0994932651519775,
+      "learning_rate": 0.0001653900285739388,
+      "loss": 2.1053,
+      "step": 2804
+    },
+    {
+      "epoch": 0.2951078379800105,
+      "grad_norm": 0.8119322657585144,
+      "learning_rate": 0.00016536424249656933,
+      "loss": 2.3161,
+      "step": 2805
+    },
+    {
+      "epoch": 0.29521304576538665,
+      "grad_norm": 1.0407122373580933,
+      "learning_rate": 0.00016533844882866568,
+      "loss": 2.0664,
+      "step": 2806
+    },
+    {
+      "epoch": 0.29531825355076274,
+      "grad_norm": 1.3417972326278687,
+      "learning_rate": 0.00016531264757322308,
+      "loss": 1.924,
+      "step": 2807
+    },
+    {
+      "epoch": 0.2954234613361389,
+      "grad_norm": 1.2302452325820923,
+      "learning_rate": 0.0001652868387332378,
+      "loss": 1.8267,
+      "step": 2808
+    },
+    {
+      "epoch": 0.295528669121515,
+      "grad_norm": 1.2221964597702026,
+      "learning_rate": 0.00016526102231170691,
+      "loss": 2.0054,
+      "step": 2809
+    },
+    {
+      "epoch": 0.29563387690689114,
+      "grad_norm": 1.5506309270858765,
+      "learning_rate": 0.0001652351983116284,
+      "loss": 1.9843,
+      "step": 2810
+    },
+    {
+      "epoch": 0.29573908469226723,
+      "grad_norm": 1.3232911825180054,
+      "learning_rate": 0.00016520936673600117,
+      "loss": 2.2001,
+      "step": 2811
+    },
+    {
+      "epoch": 0.2958442924776433,
+      "grad_norm": 1.1889708042144775,
+      "learning_rate": 0.00016518352758782486,
+      "loss": 1.3271,
+      "step": 2812
+    },
+    {
+      "epoch": 0.29594950026301947,
+      "grad_norm": 1.4930849075317383,
+      "learning_rate": 0.00016515768087010013,
+      "loss": 2.121,
+      "step": 2813
+    },
+    {
+      "epoch": 0.29605470804839557,
+      "grad_norm": 1.3592177629470825,
+      "learning_rate": 0.0001651318265858285,
+      "loss": 2.0658,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2961599158337717,
+      "grad_norm": 1.2253750562667847,
+      "learning_rate": 0.00016510596473801232,
+      "loss": 1.8405,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2962651236191478,
+      "grad_norm": 1.1184381246566772,
+      "learning_rate": 0.00016508009532965485,
+      "loss": 1.7584,
+      "step": 2816
+    },
+    {
+      "epoch": 0.29637033140452396,
+      "grad_norm": 1.332506537437439,
+      "learning_rate": 0.0001650542183637602,
+      "loss": 1.8449,
+      "step": 2817
+    },
+    {
+      "epoch": 0.29647553918990005,
+      "grad_norm": 1.4818626642227173,
+      "learning_rate": 0.00016502833384333338,
+      "loss": 2.0897,
+      "step": 2818
+    },
+    {
+      "epoch": 0.29658074697527614,
+      "grad_norm": 0.7966058254241943,
+      "learning_rate": 0.0001650024417713803,
+      "loss": 1.9058,
+      "step": 2819
+    },
+    {
+      "epoch": 0.2966859547606523,
+      "grad_norm": 1.4730877876281738,
+      "learning_rate": 0.00016497654215090772,
+      "loss": 1.8389,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2967911625460284,
+      "grad_norm": 1.3068904876708984,
+      "learning_rate": 0.00016495063498492326,
+      "loss": 2.4661,
+      "step": 2821
+    },
+    {
+      "epoch": 0.29689637033140454,
+      "grad_norm": 1.5194685459136963,
+      "learning_rate": 0.00016492472027643541,
+      "loss": 1.685,
+      "step": 2822
+    },
+    {
+      "epoch": 0.29700157811678063,
+      "grad_norm": 0.9527503848075867,
+      "learning_rate": 0.00016489879802845361,
+      "loss": 2.2541,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2971067859021568,
+      "grad_norm": 1.2973871231079102,
+      "learning_rate": 0.0001648728682439881,
+      "loss": 2.3992,
+      "step": 2824
+    },
+    {
+      "epoch": 0.29721199368753287,
+      "grad_norm": 1.154152750968933,
+      "learning_rate": 0.00016484693092605002,
+      "loss": 1.7631,
+      "step": 2825
+    },
+    {
+      "epoch": 0.297317201472909,
+      "grad_norm": 1.4371882677078247,
+      "learning_rate": 0.00016482098607765137,
+      "loss": 1.8837,
+      "step": 2826
+    },
+    {
+      "epoch": 0.2974224092582851,
+      "grad_norm": 1.1324549913406372,
+      "learning_rate": 0.00016479503370180507,
+      "loss": 2.2255,
+      "step": 2827
+    },
+    {
+      "epoch": 0.2975276170436612,
+      "grad_norm": 0.9485172629356384,
+      "learning_rate": 0.0001647690738015249,
+      "loss": 2.1955,
+      "step": 2828
+    },
+    {
+      "epoch": 0.29763282482903736,
+      "grad_norm": 1.0813753604888916,
+      "learning_rate": 0.0001647431063798254,
+      "loss": 1.8017,
+      "step": 2829
+    },
+    {
+      "epoch": 0.29773803261441345,
+      "grad_norm": 1.0293214321136475,
+      "learning_rate": 0.0001647171314397222,
+      "loss": 1.9523,
+      "step": 2830
+    },
+    {
+      "epoch": 0.2978432403997896,
+      "grad_norm": 1.2349849939346313,
+      "learning_rate": 0.00016469114898423165,
+      "loss": 1.8325,
+      "step": 2831
+    },
+    {
+      "epoch": 0.2979484481851657,
+      "grad_norm": 1.0925050973892212,
+      "learning_rate": 0.00016466515901637096,
+      "loss": 2.1776,
+      "step": 2832
+    },
+    {
+      "epoch": 0.29805365597054184,
+      "grad_norm": 1.1909685134887695,
+      "learning_rate": 0.0001646391615391583,
+      "loss": 1.9048,
+      "step": 2833
+    },
+    {
+      "epoch": 0.29815886375591794,
+      "grad_norm": 1.5420279502868652,
+      "learning_rate": 0.00016461315655561263,
+      "loss": 1.8361,
+      "step": 2834
+    },
+    {
+      "epoch": 0.29826407154129403,
+      "grad_norm": 1.3840932846069336,
+      "learning_rate": 0.00016458714406875392,
+      "loss": 1.9673,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2983692793266702,
+      "grad_norm": 1.4501628875732422,
+      "learning_rate": 0.0001645611240816028,
+      "loss": 1.3108,
+      "step": 2836
+    },
+    {
+      "epoch": 0.29847448711204627,
+      "grad_norm": 1.304292917251587,
+      "learning_rate": 0.00016453509659718093,
+      "loss": 1.4608,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2985796948974224,
+      "grad_norm": 1.1828423738479614,
+      "learning_rate": 0.00016450906161851082,
+      "loss": 1.8168,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2986849026827985,
+      "grad_norm": 1.2528148889541626,
+      "learning_rate": 0.00016448301914861584,
+      "loss": 1.5209,
+      "step": 2839
+    },
+    {
+      "epoch": 0.29879011046817466,
+      "grad_norm": 0.9708895683288574,
+      "learning_rate": 0.00016445696919052013,
+      "loss": 2.2244,
+      "step": 2840
+    },
+    {
+      "epoch": 0.29889531825355076,
+      "grad_norm": 1.3018192052841187,
+      "learning_rate": 0.00016443091174724885,
+      "loss": 1.6512,
+      "step": 2841
+    },
+    {
+      "epoch": 0.2990005260389269,
+      "grad_norm": 1.1458721160888672,
+      "learning_rate": 0.00016440484682182799,
+      "loss": 2.0022,
+      "step": 2842
+    },
+    {
+      "epoch": 0.299105733824303,
+      "grad_norm": 2.0912718772888184,
+      "learning_rate": 0.00016437877441728433,
+      "loss": 1.9703,
+      "step": 2843
+    },
+    {
+      "epoch": 0.2992109416096791,
+      "grad_norm": 1.1203174591064453,
+      "learning_rate": 0.00016435269453664558,
+      "loss": 1.9248,
+      "step": 2844
+    },
+    {
+      "epoch": 0.29931614939505524,
+      "grad_norm": 1.6757680177688599,
+      "learning_rate": 0.00016432660718294033,
+      "loss": 1.8856,
+      "step": 2845
+    },
+    {
+      "epoch": 0.29942135718043134,
+      "grad_norm": 1.9626332521438599,
+      "learning_rate": 0.00016430051235919802,
+      "loss": 1.3917,
+      "step": 2846
+    },
+    {
+      "epoch": 0.2995265649658075,
+      "grad_norm": 1.4652724266052246,
+      "learning_rate": 0.00016427441006844893,
+      "loss": 2.1621,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2996317727511836,
+      "grad_norm": 1.40691339969635,
+      "learning_rate": 0.00016424830031372425,
+      "loss": 1.8587,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2997369805365597,
+      "grad_norm": 1.054113745689392,
+      "learning_rate": 0.000164222183098056,
+      "loss": 1.9425,
+      "step": 2849
+    },
+    {
+      "epoch": 0.2998421883219358,
+      "grad_norm": 1.6056944131851196,
+      "learning_rate": 0.00016419605842447714,
+      "loss": 1.5958,
+      "step": 2850
+    },
+    {
+      "epoch": 0.2999473961073119,
+      "grad_norm": 1.3629722595214844,
+      "learning_rate": 0.00016416992629602142,
+      "loss": 1.524,
+      "step": 2851
+    },
+    {
+      "epoch": 0.30005260389268806,
+      "grad_norm": 1.3340497016906738,
+      "learning_rate": 0.00016414378671572344,
+      "loss": 1.2919,
+      "step": 2852
+    },
+    {
+      "epoch": 0.30015781167806416,
+      "grad_norm": 1.4080469608306885,
+      "learning_rate": 0.00016411763968661873,
+      "loss": 2.1183,
+      "step": 2853
+    },
+    {
+      "epoch": 0.3002630194634403,
+      "grad_norm": 1.1511781215667725,
+      "learning_rate": 0.00016409148521174367,
+      "loss": 1.4441,
+      "step": 2854
+    },
+    {
+      "epoch": 0.3003682272488164,
+      "grad_norm": 1.404482126235962,
+      "learning_rate": 0.00016406532329413546,
+      "loss": 1.9554,
+      "step": 2855
+    },
+    {
+      "epoch": 0.30047343503419255,
+      "grad_norm": 1.4892315864562988,
+      "learning_rate": 0.00016403915393683221,
+      "loss": 1.9983,
+      "step": 2856
+    },
+    {
+      "epoch": 0.30057864281956864,
+      "grad_norm": 1.159440040588379,
+      "learning_rate": 0.00016401297714287294,
+      "loss": 2.1476,
+      "step": 2857
+    },
+    {
+      "epoch": 0.3006838506049448,
+      "grad_norm": 1.5921545028686523,
+      "learning_rate": 0.00016398679291529738,
+      "loss": 2.1005,
+      "step": 2858
+    },
+    {
+      "epoch": 0.3007890583903209,
+      "grad_norm": 0.9857234358787537,
+      "learning_rate": 0.00016396060125714628,
+      "loss": 1.659,
+      "step": 2859
+    },
+    {
+      "epoch": 0.300894266175697,
+      "grad_norm": 1.3065168857574463,
+      "learning_rate": 0.00016393440217146114,
+      "loss": 2.1434,
+      "step": 2860
+    },
+    {
+      "epoch": 0.3009994739610731,
+      "grad_norm": 1.689706563949585,
+      "learning_rate": 0.00016390819566128445,
+      "loss": 1.8146,
+      "step": 2861
+    },
+    {
+      "epoch": 0.3011046817464492,
+      "grad_norm": 2.270273208618164,
+      "learning_rate": 0.00016388198172965942,
+      "loss": 2.1251,
+      "step": 2862
+    },
+    {
+      "epoch": 0.30120988953182537,
+      "grad_norm": 1.3508756160736084,
+      "learning_rate": 0.00016385576037963021,
+      "loss": 1.9374,
+      "step": 2863
+    },
+    {
+      "epoch": 0.30131509731720146,
+      "grad_norm": 1.7013899087905884,
+      "learning_rate": 0.00016382953161424185,
+      "loss": 1.9598,
+      "step": 2864
+    },
+    {
+      "epoch": 0.3014203051025776,
+      "grad_norm": 1.457298755645752,
+      "learning_rate": 0.00016380329543654013,
+      "loss": 2.2662,
+      "step": 2865
+    },
+    {
+      "epoch": 0.3015255128879537,
+      "grad_norm": 1.280228853225708,
+      "learning_rate": 0.00016377705184957185,
+      "loss": 1.9162,
+      "step": 2866
+    },
+    {
+      "epoch": 0.3016307206733298,
+      "grad_norm": 1.6916749477386475,
+      "learning_rate": 0.00016375080085638451,
+      "loss": 2.1827,
+      "step": 2867
+    },
+    {
+      "epoch": 0.30173592845870595,
+      "grad_norm": 1.3950283527374268,
+      "learning_rate": 0.00016372454246002663,
+      "loss": 1.6382,
+      "step": 2868
+    },
+    {
+      "epoch": 0.30184113624408204,
+      "grad_norm": 1.4823285341262817,
+      "learning_rate": 0.00016369827666354745,
+      "loss": 1.5104,
+      "step": 2869
+    },
+    {
+      "epoch": 0.3019463440294582,
+      "grad_norm": 0.9238750338554382,
+      "learning_rate": 0.00016367200346999714,
+      "loss": 2.1355,
+      "step": 2870
+    },
+    {
+      "epoch": 0.3020515518148343,
+      "grad_norm": 1.290266990661621,
+      "learning_rate": 0.00016364572288242677,
+      "loss": 1.9956,
+      "step": 2871
+    },
+    {
+      "epoch": 0.30215675960021043,
+      "grad_norm": 0.9198346138000488,
+      "learning_rate": 0.00016361943490388815,
+      "loss": 2.0828,
+      "step": 2872
+    },
+    {
+      "epoch": 0.3022619673855865,
+      "grad_norm": 1.6661968231201172,
+      "learning_rate": 0.00016359313953743406,
+      "loss": 1.7579,
+      "step": 2873
+    },
+    {
+      "epoch": 0.3023671751709627,
+      "grad_norm": 1.031909465789795,
+      "learning_rate": 0.00016356683678611807,
+      "loss": 2.0152,
+      "step": 2874
+    },
+    {
+      "epoch": 0.30247238295633877,
+      "grad_norm": 1.662083387374878,
+      "learning_rate": 0.00016354052665299468,
+      "loss": 2.1388,
+      "step": 2875
+    },
+    {
+      "epoch": 0.30257759074171486,
+      "grad_norm": 1.226048231124878,
+      "learning_rate": 0.00016351420914111916,
+      "loss": 2.4793,
+      "step": 2876
+    },
+    {
+      "epoch": 0.302682798527091,
+      "grad_norm": 1.6051714420318604,
+      "learning_rate": 0.00016348788425354766,
+      "loss": 1.7133,
+      "step": 2877
+    },
+    {
+      "epoch": 0.3027880063124671,
+      "grad_norm": 0.972946047782898,
+      "learning_rate": 0.00016346155199333721,
+      "loss": 1.8606,
+      "step": 2878
+    },
+    {
+      "epoch": 0.30289321409784326,
+      "grad_norm": 1.1665260791778564,
+      "learning_rate": 0.00016343521236354574,
+      "loss": 2.1413,
+      "step": 2879
+    },
+    {
+      "epoch": 0.30299842188321935,
+      "grad_norm": 1.250375509262085,
+      "learning_rate": 0.00016340886536723192,
+      "loss": 1.6386,
+      "step": 2880
+    },
+    {
+      "epoch": 0.3031036296685955,
+      "grad_norm": 1.2102625370025635,
+      "learning_rate": 0.00016338251100745537,
+      "loss": 1.9187,
+      "step": 2881
+    },
+    {
+      "epoch": 0.3032088374539716,
+      "grad_norm": 2.1711556911468506,
+      "learning_rate": 0.00016335614928727652,
+      "loss": 1.7001,
+      "step": 2882
+    },
+    {
+      "epoch": 0.3033140452393477,
+      "grad_norm": 2.243464231491089,
+      "learning_rate": 0.0001633297802097567,
+      "loss": 1.7878,
+      "step": 2883
+    },
+    {
+      "epoch": 0.30341925302472383,
+      "grad_norm": 1.4908397197723389,
+      "learning_rate": 0.00016330340377795804,
+      "loss": 1.846,
+      "step": 2884
+    },
+    {
+      "epoch": 0.3035244608100999,
+      "grad_norm": 1.6034961938858032,
+      "learning_rate": 0.00016327701999494353,
+      "loss": 1.6955,
+      "step": 2885
+    },
+    {
+      "epoch": 0.3036296685954761,
+      "grad_norm": 1.2690043449401855,
+      "learning_rate": 0.0001632506288637771,
+      "loss": 2.009,
+      "step": 2886
+    },
+    {
+      "epoch": 0.30373487638085217,
+      "grad_norm": 1.138432502746582,
+      "learning_rate": 0.00016322423038752336,
+      "loss": 1.8181,
+      "step": 2887
+    },
+    {
+      "epoch": 0.3038400841662283,
+      "grad_norm": 0.7713850140571594,
+      "learning_rate": 0.000163197824569248,
+      "loss": 2.1781,
+      "step": 2888
+    },
+    {
+      "epoch": 0.3039452919516044,
+      "grad_norm": 1.5833183526992798,
+      "learning_rate": 0.00016317141141201731,
+      "loss": 2.0445,
+      "step": 2889
+    },
+    {
+      "epoch": 0.30405049973698056,
+      "grad_norm": 1.196471095085144,
+      "learning_rate": 0.0001631449909188987,
+      "loss": 1.6292,
+      "step": 2890
+    },
+    {
+      "epoch": 0.30415570752235666,
+      "grad_norm": 1.7439603805541992,
+      "learning_rate": 0.0001631185630929602,
+      "loss": 2.1841,
+      "step": 2891
+    },
+    {
+      "epoch": 0.30426091530773275,
+      "grad_norm": 1.4162222146987915,
+      "learning_rate": 0.00016309212793727077,
+      "loss": 1.9838,
+      "step": 2892
+    },
+    {
+      "epoch": 0.3043661230931089,
+      "grad_norm": 1.5100573301315308,
+      "learning_rate": 0.00016306568545490033,
+      "loss": 2.1413,
+      "step": 2893
+    },
+    {
+      "epoch": 0.304471330878485,
+      "grad_norm": 1.3554414510726929,
+      "learning_rate": 0.00016303923564891948,
+      "loss": 2.3482,
+      "step": 2894
+    },
+    {
+      "epoch": 0.30457653866386114,
+      "grad_norm": 1.4444999694824219,
+      "learning_rate": 0.0001630127785223998,
+      "loss": 2.3344,
+      "step": 2895
+    },
+    {
+      "epoch": 0.30468174644923723,
+      "grad_norm": 1.267393708229065,
+      "learning_rate": 0.00016298631407841361,
+      "loss": 1.4887,
+      "step": 2896
+    },
+    {
+      "epoch": 0.3047869542346134,
+      "grad_norm": 1.479801893234253,
+      "learning_rate": 0.00016295984232003426,
+      "loss": 2.2112,
+      "step": 2897
+    },
+    {
+      "epoch": 0.3048921620199895,
+      "grad_norm": 1.5580319166183472,
+      "learning_rate": 0.0001629333632503357,
+      "loss": 1.7964,
+      "step": 2898
+    },
+    {
+      "epoch": 0.30499736980536557,
+      "grad_norm": 1.1315181255340576,
+      "learning_rate": 0.00016290687687239283,
+      "loss": 1.6472,
+      "step": 2899
+    },
+    {
+      "epoch": 0.3051025775907417,
+      "grad_norm": 1.8991228342056274,
+      "learning_rate": 0.00016288038318928156,
+      "loss": 1.6517,
+      "step": 2900
+    },
+    {
+      "epoch": 0.3052077853761178,
+      "grad_norm": 1.7622857093811035,
+      "learning_rate": 0.00016285388220407847,
+      "loss": 1.8343,
+      "step": 2901
+    },
+    {
+      "epoch": 0.30531299316149396,
+      "grad_norm": 1.0036766529083252,
+      "learning_rate": 0.00016282737391986097,
+      "loss": 2.2478,
+      "step": 2902
+    },
+    {
+      "epoch": 0.30541820094687006,
+      "grad_norm": 1.0161508321762085,
+      "learning_rate": 0.00016280085833970744,
+      "loss": 2.0217,
+      "step": 2903
+    },
+    {
+      "epoch": 0.3055234087322462,
+      "grad_norm": 1.6816000938415527,
+      "learning_rate": 0.00016277433546669703,
+      "loss": 1.6079,
+      "step": 2904
+    },
+    {
+      "epoch": 0.3056286165176223,
+      "grad_norm": 1.2823631763458252,
+      "learning_rate": 0.00016274780530390977,
+      "loss": 1.6762,
+      "step": 2905
+    },
+    {
+      "epoch": 0.30573382430299845,
+      "grad_norm": 1.2397310733795166,
+      "learning_rate": 0.00016272126785442644,
+      "loss": 1.9937,
+      "step": 2906
+    },
+    {
+      "epoch": 0.30583903208837454,
+      "grad_norm": 1.5026395320892334,
+      "learning_rate": 0.0001626947231213289,
+      "loss": 1.6613,
+      "step": 2907
+    },
+    {
+      "epoch": 0.30594423987375063,
+      "grad_norm": 1.2583715915679932,
+      "learning_rate": 0.00016266817110769955,
+      "loss": 2.0048,
+      "step": 2908
+    },
+    {
+      "epoch": 0.3060494476591268,
+      "grad_norm": 0.9579928517341614,
+      "learning_rate": 0.00016264161181662188,
+      "loss": 2.0814,
+      "step": 2909
+    },
+    {
+      "epoch": 0.3061546554445029,
+      "grad_norm": 1.003151297569275,
+      "learning_rate": 0.0001626150452511801,
+      "loss": 2.2672,
+      "step": 2910
+    },
+    {
+      "epoch": 0.306259863229879,
+      "grad_norm": 1.4802247285842896,
+      "learning_rate": 0.00016258847141445928,
+      "loss": 2.5276,
+      "step": 2911
+    },
+    {
+      "epoch": 0.3063650710152551,
+      "grad_norm": 1.5277745723724365,
+      "learning_rate": 0.00016256189030954538,
+      "loss": 2.1005,
+      "step": 2912
+    },
+    {
+      "epoch": 0.30647027880063127,
+      "grad_norm": 1.1211442947387695,
+      "learning_rate": 0.00016253530193952517,
+      "loss": 1.8205,
+      "step": 2913
+    },
+    {
+      "epoch": 0.30657548658600736,
+      "grad_norm": 1.083178997039795,
+      "learning_rate": 0.0001625087063074863,
+      "loss": 2.128,
+      "step": 2914
+    },
+    {
+      "epoch": 0.30668069437138346,
+      "grad_norm": 2.118847370147705,
+      "learning_rate": 0.00016248210341651716,
+      "loss": 1.9704,
+      "step": 2915
+    },
+    {
+      "epoch": 0.3067859021567596,
+      "grad_norm": 1.371398687362671,
+      "learning_rate": 0.00016245549326970713,
+      "loss": 1.5593,
+      "step": 2916
+    },
+    {
+      "epoch": 0.3068911099421357,
+      "grad_norm": 1.180167317390442,
+      "learning_rate": 0.0001624288758701463,
+      "loss": 1.7496,
+      "step": 2917
+    },
+    {
+      "epoch": 0.30699631772751185,
+      "grad_norm": 1.3497270345687866,
+      "learning_rate": 0.00016240225122092573,
+      "loss": 2.0314,
+      "step": 2918
+    },
+    {
+      "epoch": 0.30710152551288794,
+      "grad_norm": 1.6975984573364258,
+      "learning_rate": 0.00016237561932513718,
+      "loss": 1.77,
+      "step": 2919
+    },
+    {
+      "epoch": 0.3072067332982641,
+      "grad_norm": 1.6393413543701172,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 1.93,
+      "step": 2920
+    },
+    {
+      "epoch": 0.3073119410836402,
+      "grad_norm": 1.7652052640914917,
+      "learning_rate": 0.00016232233380622779,
+      "loss": 1.8103,
+      "step": 2921
+    },
+    {
+      "epoch": 0.30741714886901633,
+      "grad_norm": 1.489980936050415,
+      "learning_rate": 0.00016229568018929483,
+      "loss": 1.8286,
+      "step": 2922
+    },
+    {
+      "epoch": 0.3075223566543924,
+      "grad_norm": 1.4050182104110718,
+      "learning_rate": 0.00016226901933816962,
+      "loss": 1.5679,
+      "step": 2923
+    },
+    {
+      "epoch": 0.3076275644397685,
+      "grad_norm": 1.0518805980682373,
+      "learning_rate": 0.0001622423512559483,
+      "loss": 1.6626,
+      "step": 2924
+    },
+    {
+      "epoch": 0.30773277222514467,
+      "grad_norm": 1.4546114206314087,
+      "learning_rate": 0.00016221567594572762,
+      "loss": 1.9034,
+      "step": 2925
+    },
+    {
+      "epoch": 0.30783798001052076,
+      "grad_norm": 0.9748585820198059,
+      "learning_rate": 0.00016218899341060542,
+      "loss": 2.2178,
+      "step": 2926
+    },
+    {
+      "epoch": 0.3079431877958969,
+      "grad_norm": 1.3530354499816895,
+      "learning_rate": 0.00016216230365368017,
+      "loss": 2.2276,
+      "step": 2927
+    },
+    {
+      "epoch": 0.308048395581273,
+      "grad_norm": 0.9455937743186951,
+      "learning_rate": 0.00016213560667805127,
+      "loss": 1.9019,
+      "step": 2928
+    },
+    {
+      "epoch": 0.30815360336664915,
+      "grad_norm": 1.0966365337371826,
+      "learning_rate": 0.00016210890248681906,
+      "loss": 1.5131,
+      "step": 2929
+    },
+    {
+      "epoch": 0.30825881115202525,
+      "grad_norm": 0.9960981607437134,
+      "learning_rate": 0.00016208219108308444,
+      "loss": 1.9477,
+      "step": 2930
+    },
+    {
+      "epoch": 0.30836401893740134,
+      "grad_norm": 1.0510518550872803,
+      "learning_rate": 0.00016205547246994945,
+      "loss": 1.7093,
+      "step": 2931
+    },
+    {
+      "epoch": 0.3084692267227775,
+      "grad_norm": 1.185834527015686,
+      "learning_rate": 0.00016202874665051674,
+      "loss": 2.012,
+      "step": 2932
+    },
+    {
+      "epoch": 0.3085744345081536,
+      "grad_norm": 0.9046087861061096,
+      "learning_rate": 0.00016200201362788995,
+      "loss": 1.9285,
+      "step": 2933
+    },
+    {
+      "epoch": 0.30867964229352973,
+      "grad_norm": 1.357476830482483,
+      "learning_rate": 0.00016197527340517352,
+      "loss": 1.8586,
+      "step": 2934
+    },
+    {
+      "epoch": 0.3087848500789058,
+      "grad_norm": 0.9783453941345215,
+      "learning_rate": 0.00016194852598547263,
+      "loss": 2.1739,
+      "step": 2935
+    },
+    {
+      "epoch": 0.308890057864282,
+      "grad_norm": 1.2122621536254883,
+      "learning_rate": 0.00016192177137189345,
+      "loss": 1.7506,
+      "step": 2936
+    },
+    {
+      "epoch": 0.30899526564965807,
+      "grad_norm": 1.8045505285263062,
+      "learning_rate": 0.00016189500956754284,
+      "loss": 1.656,
+      "step": 2937
+    },
+    {
+      "epoch": 0.3091004734350342,
+      "grad_norm": 1.205743670463562,
+      "learning_rate": 0.00016186824057552856,
+      "loss": 1.9285,
+      "step": 2938
+    },
+    {
+      "epoch": 0.3092056812204103,
+      "grad_norm": 1.1010884046554565,
+      "learning_rate": 0.00016184146439895928,
+      "loss": 2.1431,
+      "step": 2939
+    },
+    {
+      "epoch": 0.3093108890057864,
+      "grad_norm": 0.7133980393409729,
+      "learning_rate": 0.00016181468104094435,
+      "loss": 2.1475,
+      "step": 2940
+    },
+    {
+      "epoch": 0.30941609679116255,
+      "grad_norm": 1.0357547998428345,
+      "learning_rate": 0.00016178789050459407,
+      "loss": 1.8142,
+      "step": 2941
+    },
+    {
+      "epoch": 0.30952130457653865,
+      "grad_norm": 1.1060856580734253,
+      "learning_rate": 0.0001617610927930195,
+      "loss": 1.4338,
+      "step": 2942
+    },
+    {
+      "epoch": 0.3096265123619148,
+      "grad_norm": 2.330672264099121,
+      "learning_rate": 0.00016173428790933265,
+      "loss": 1.5461,
+      "step": 2943
+    },
+    {
+      "epoch": 0.3097317201472909,
+      "grad_norm": 1.4529541730880737,
+      "learning_rate": 0.0001617074758566462,
+      "loss": 1.8743,
+      "step": 2944
+    },
+    {
+      "epoch": 0.30983692793266704,
+      "grad_norm": 0.984163224697113,
+      "learning_rate": 0.00016168065663807376,
+      "loss": 2.2996,
+      "step": 2945
+    },
+    {
+      "epoch": 0.30994213571804313,
+      "grad_norm": 1.3428571224212646,
+      "learning_rate": 0.00016165383025672981,
+      "loss": 2.0059,
+      "step": 2946
+    },
+    {
+      "epoch": 0.3100473435034192,
+      "grad_norm": 0.8479394316673279,
+      "learning_rate": 0.00016162699671572956,
+      "loss": 2.0439,
+      "step": 2947
+    },
+    {
+      "epoch": 0.3101525512887954,
+      "grad_norm": 0.9885584712028503,
+      "learning_rate": 0.0001616001560181891,
+      "loss": 2.2262,
+      "step": 2948
+    },
+    {
+      "epoch": 0.31025775907417147,
+      "grad_norm": 1.3757953643798828,
+      "learning_rate": 0.0001615733081672254,
+      "loss": 1.9734,
+      "step": 2949
+    },
+    {
+      "epoch": 0.3103629668595476,
+      "grad_norm": 1.0954573154449463,
+      "learning_rate": 0.00016154645316595616,
+      "loss": 1.7084,
+      "step": 2950
+    },
+    {
+      "epoch": 0.3104681746449237,
+      "grad_norm": 1.236785888671875,
+      "learning_rate": 0.00016151959101749996,
+      "loss": 1.4454,
+      "step": 2951
+    },
+    {
+      "epoch": 0.31057338243029986,
+      "grad_norm": 1.077486515045166,
+      "learning_rate": 0.00016149272172497626,
+      "loss": 1.6408,
+      "step": 2952
+    },
+    {
+      "epoch": 0.31067859021567595,
+      "grad_norm": 0.9998572468757629,
+      "learning_rate": 0.00016146584529150526,
+      "loss": 2.4455,
+      "step": 2953
+    },
+    {
+      "epoch": 0.3107837980010521,
+      "grad_norm": 1.189342737197876,
+      "learning_rate": 0.00016143896172020808,
+      "loss": 1.8845,
+      "step": 2954
+    },
+    {
+      "epoch": 0.3108890057864282,
+      "grad_norm": 1.0967530012130737,
+      "learning_rate": 0.00016141207101420655,
+      "loss": 2.104,
+      "step": 2955
+    },
+    {
+      "epoch": 0.3109942135718043,
+      "grad_norm": 1.4149389266967773,
+      "learning_rate": 0.00016138517317662346,
+      "loss": 1.6627,
+      "step": 2956
+    },
+    {
+      "epoch": 0.31109942135718044,
+      "grad_norm": 0.9122399687767029,
+      "learning_rate": 0.00016135826821058233,
+      "loss": 1.7756,
+      "step": 2957
+    },
+    {
+      "epoch": 0.31120462914255653,
+      "grad_norm": 1.4171504974365234,
+      "learning_rate": 0.00016133135611920757,
+      "loss": 1.9604,
+      "step": 2958
+    },
+    {
+      "epoch": 0.3113098369279327,
+      "grad_norm": 1.4796384572982788,
+      "learning_rate": 0.0001613044369056244,
+      "loss": 1.6938,
+      "step": 2959
+    },
+    {
+      "epoch": 0.3114150447133088,
+      "grad_norm": 1.3851723670959473,
+      "learning_rate": 0.0001612775105729588,
+      "loss": 1.7427,
+      "step": 2960
+    },
+    {
+      "epoch": 0.3115202524986849,
+      "grad_norm": 1.3645011186599731,
+      "learning_rate": 0.00016125057712433773,
+      "loss": 1.7735,
+      "step": 2961
+    },
+    {
+      "epoch": 0.311625460284061,
+      "grad_norm": 1.73271906375885,
+      "learning_rate": 0.00016122363656288882,
+      "loss": 2.1165,
+      "step": 2962
+    },
+    {
+      "epoch": 0.3117306680694371,
+      "grad_norm": 0.9199891686439514,
+      "learning_rate": 0.0001611966888917406,
+      "loss": 1.6816,
+      "step": 2963
+    },
+    {
+      "epoch": 0.31183587585481326,
+      "grad_norm": 1.2702879905700684,
+      "learning_rate": 0.00016116973411402238,
+      "loss": 2.1454,
+      "step": 2964
+    },
+    {
+      "epoch": 0.31194108364018935,
+      "grad_norm": 1.4072431325912476,
+      "learning_rate": 0.0001611427722328644,
+      "loss": 1.9006,
+      "step": 2965
+    },
+    {
+      "epoch": 0.3120462914255655,
+      "grad_norm": 0.7870166301727295,
+      "learning_rate": 0.0001611158032513976,
+      "loss": 1.8407,
+      "step": 2966
+    },
+    {
+      "epoch": 0.3121514992109416,
+      "grad_norm": 1.3497145175933838,
+      "learning_rate": 0.00016108882717275384,
+      "loss": 1.6683,
+      "step": 2967
+    },
+    {
+      "epoch": 0.31225670699631775,
+      "grad_norm": 1.090868592262268,
+      "learning_rate": 0.00016106184400006569,
+      "loss": 1.9826,
+      "step": 2968
+    },
+    {
+      "epoch": 0.31236191478169384,
+      "grad_norm": 1.025602102279663,
+      "learning_rate": 0.00016103485373646672,
+      "loss": 2.011,
+      "step": 2969
+    },
+    {
+      "epoch": 0.31246712256707,
+      "grad_norm": 1.3139764070510864,
+      "learning_rate": 0.00016100785638509114,
+      "loss": 2.148,
+      "step": 2970
+    },
+    {
+      "epoch": 0.3125723303524461,
+      "grad_norm": 1.5432015657424927,
+      "learning_rate": 0.00016098085194907413,
+      "loss": 1.8339,
+      "step": 2971
+    },
+    {
+      "epoch": 0.3126775381378222,
+      "grad_norm": 1.3246941566467285,
+      "learning_rate": 0.00016095384043155156,
+      "loss": 1.5852,
+      "step": 2972
+    },
+    {
+      "epoch": 0.3127827459231983,
+      "grad_norm": 1.229443907737732,
+      "learning_rate": 0.00016092682183566025,
+      "loss": 1.4615,
+      "step": 2973
+    },
+    {
+      "epoch": 0.3128879537085744,
+      "grad_norm": 1.3812941312789917,
+      "learning_rate": 0.0001608997961645377,
+      "loss": 2.0711,
+      "step": 2974
+    },
+    {
+      "epoch": 0.31299316149395057,
+      "grad_norm": 1.3320645093917847,
+      "learning_rate": 0.0001608727634213224,
+      "loss": 2.0261,
+      "step": 2975
+    },
+    {
+      "epoch": 0.31309836927932666,
+      "grad_norm": 1.8210208415985107,
+      "learning_rate": 0.00016084572360915348,
+      "loss": 1.314,
+      "step": 2976
+    },
+    {
+      "epoch": 0.3132035770647028,
+      "grad_norm": 0.9022918939590454,
+      "learning_rate": 0.00016081867673117106,
+      "loss": 1.8564,
+      "step": 2977
+    },
+    {
+      "epoch": 0.3133087848500789,
+      "grad_norm": 1.5104812383651733,
+      "learning_rate": 0.00016079162279051602,
+      "loss": 1.8883,
+      "step": 2978
+    },
+    {
+      "epoch": 0.313413992635455,
+      "grad_norm": 1.4188772439956665,
+      "learning_rate": 0.00016076456179032998,
+      "loss": 2.1038,
+      "step": 2979
+    },
+    {
+      "epoch": 0.31351920042083115,
+      "grad_norm": 1.3051005601882935,
+      "learning_rate": 0.00016073749373375545,
+      "loss": 2.2416,
+      "step": 2980
+    },
+    {
+      "epoch": 0.31362440820620724,
+      "grad_norm": 1.0661115646362305,
+      "learning_rate": 0.00016071041862393578,
+      "loss": 1.8527,
+      "step": 2981
+    },
+    {
+      "epoch": 0.3137296159915834,
+      "grad_norm": 1.1394139528274536,
+      "learning_rate": 0.00016068333646401516,
+      "loss": 1.468,
+      "step": 2982
+    },
+    {
+      "epoch": 0.3138348237769595,
+      "grad_norm": 1.1041537523269653,
+      "learning_rate": 0.00016065624725713847,
+      "loss": 1.4634,
+      "step": 2983
+    },
+    {
+      "epoch": 0.31394003156233563,
+      "grad_norm": 1.4715754985809326,
+      "learning_rate": 0.00016062915100645153,
+      "loss": 1.8815,
+      "step": 2984
+    },
+    {
+      "epoch": 0.3140452393477117,
+      "grad_norm": 1.0146689414978027,
+      "learning_rate": 0.0001606020477151009,
+      "loss": 1.9733,
+      "step": 2985
+    },
+    {
+      "epoch": 0.3141504471330879,
+      "grad_norm": 1.4208067655563354,
+      "learning_rate": 0.00016057493738623406,
+      "loss": 1.6583,
+      "step": 2986
+    },
+    {
+      "epoch": 0.31425565491846397,
+      "grad_norm": 0.9717585444450378,
+      "learning_rate": 0.0001605478200229992,
+      "loss": 1.8575,
+      "step": 2987
+    },
+    {
+      "epoch": 0.31436086270384006,
+      "grad_norm": 2.195236921310425,
+      "learning_rate": 0.0001605206956285454,
+      "loss": 1.4584,
+      "step": 2988
+    },
+    {
+      "epoch": 0.3144660704892162,
+      "grad_norm": 1.941888451576233,
+      "learning_rate": 0.00016049356420602247,
+      "loss": 1.9882,
+      "step": 2989
+    },
+    {
+      "epoch": 0.3145712782745923,
+      "grad_norm": 1.0348442792892456,
+      "learning_rate": 0.00016046642575858115,
+      "loss": 2.0475,
+      "step": 2990
+    },
+    {
+      "epoch": 0.31467648605996845,
+      "grad_norm": 1.9281662702560425,
+      "learning_rate": 0.00016043928028937292,
+      "loss": 2.1587,
+      "step": 2991
+    },
+    {
+      "epoch": 0.31478169384534455,
+      "grad_norm": 1.6395374536514282,
+      "learning_rate": 0.00016041212780155007,
+      "loss": 1.9435,
+      "step": 2992
+    },
+    {
+      "epoch": 0.3148869016307207,
+      "grad_norm": 1.1363707780838013,
+      "learning_rate": 0.0001603849682982658,
+      "loss": 1.6754,
+      "step": 2993
+    },
+    {
+      "epoch": 0.3149921094160968,
+      "grad_norm": 1.3249703645706177,
+      "learning_rate": 0.00016035780178267394,
+      "loss": 2.2371,
+      "step": 2994
+    },
+    {
+      "epoch": 0.3150973172014729,
+      "grad_norm": 1.0637754201889038,
+      "learning_rate": 0.00016033062825792935,
+      "loss": 2.1387,
+      "step": 2995
+    },
+    {
+      "epoch": 0.31520252498684903,
+      "grad_norm": 0.9320810437202454,
+      "learning_rate": 0.00016030344772718756,
+      "loss": 2.0091,
+      "step": 2996
+    },
+    {
+      "epoch": 0.3153077327722251,
+      "grad_norm": 1.375025987625122,
+      "learning_rate": 0.00016027626019360496,
+      "loss": 2.2975,
+      "step": 2997
+    },
+    {
+      "epoch": 0.3154129405576013,
+      "grad_norm": 1.532082200050354,
+      "learning_rate": 0.00016024906566033874,
+      "loss": 2.0983,
+      "step": 2998
+    },
+    {
+      "epoch": 0.31551814834297737,
+      "grad_norm": 0.982284426689148,
+      "learning_rate": 0.00016022186413054693,
+      "loss": 1.4862,
+      "step": 2999
+    },
+    {
+      "epoch": 0.3156233561283535,
+      "grad_norm": 1.5432162284851074,
+      "learning_rate": 0.00016019465560738834,
+      "loss": 2.0209,
+      "step": 3000
+    },
+    {
+      "epoch": 0.3157285639137296,
+      "grad_norm": 1.4942352771759033,
+      "learning_rate": 0.0001601674400940226,
+      "loss": 2.1858,
+      "step": 3001
+    },
+    {
+      "epoch": 0.31583377169910576,
+      "grad_norm": 1.9148730039596558,
+      "learning_rate": 0.0001601402175936102,
+      "loss": 1.2212,
+      "step": 3002
+    },
+    {
+      "epoch": 0.31593897948448185,
+      "grad_norm": 1.283167839050293,
+      "learning_rate": 0.00016011298810931232,
+      "loss": 1.617,
+      "step": 3003
+    },
+    {
+      "epoch": 0.31604418726985795,
+      "grad_norm": 1.2399694919586182,
+      "learning_rate": 0.00016008575164429113,
+      "loss": 1.8931,
+      "step": 3004
+    },
+    {
+      "epoch": 0.3161493950552341,
+      "grad_norm": 1.6097055673599243,
+      "learning_rate": 0.00016005850820170943,
+      "loss": 1.6942,
+      "step": 3005
+    },
+    {
+      "epoch": 0.3162546028406102,
+      "grad_norm": 0.8781225085258484,
+      "learning_rate": 0.00016003125778473097,
+      "loss": 1.8621,
+      "step": 3006
+    },
+    {
+      "epoch": 0.31635981062598634,
+      "grad_norm": 1.7202116250991821,
+      "learning_rate": 0.0001600040003965202,
+      "loss": 2.01,
+      "step": 3007
+    },
+    {
+      "epoch": 0.31646501841136243,
+      "grad_norm": 1.6193592548370361,
+      "learning_rate": 0.00015997673604024244,
+      "loss": 1.5022,
+      "step": 3008
+    },
+    {
+      "epoch": 0.3165702261967386,
+      "grad_norm": 1.6493405103683472,
+      "learning_rate": 0.00015994946471906382,
+      "loss": 1.9556,
+      "step": 3009
+    },
+    {
+      "epoch": 0.3166754339821147,
+      "grad_norm": 1.438834309577942,
+      "learning_rate": 0.0001599221864361513,
+      "loss": 2.3363,
+      "step": 3010
+    },
+    {
+      "epoch": 0.31678064176749077,
+      "grad_norm": 1.1828105449676514,
+      "learning_rate": 0.00015989490119467257,
+      "loss": 2.0282,
+      "step": 3011
+    },
+    {
+      "epoch": 0.3168858495528669,
+      "grad_norm": 1.8960435390472412,
+      "learning_rate": 0.00015986760899779618,
+      "loss": 1.9431,
+      "step": 3012
+    },
+    {
+      "epoch": 0.316991057338243,
+      "grad_norm": 1.4587379693984985,
+      "learning_rate": 0.0001598403098486915,
+      "loss": 1.9177,
+      "step": 3013
+    },
+    {
+      "epoch": 0.31709626512361916,
+      "grad_norm": 2.3861501216888428,
+      "learning_rate": 0.00015981300375052872,
+      "loss": 1.7569,
+      "step": 3014
+    },
+    {
+      "epoch": 0.31720147290899525,
+      "grad_norm": 1.6264148950576782,
+      "learning_rate": 0.00015978569070647876,
+      "loss": 1.835,
+      "step": 3015
+    },
+    {
+      "epoch": 0.3173066806943714,
+      "grad_norm": 1.7277299165725708,
+      "learning_rate": 0.0001597583707197134,
+      "loss": 2.2641,
+      "step": 3016
+    },
+    {
+      "epoch": 0.3174118884797475,
+      "grad_norm": 1.0360989570617676,
+      "learning_rate": 0.00015973104379340524,
+      "loss": 1.9667,
+      "step": 3017
+    },
+    {
+      "epoch": 0.31751709626512364,
+      "grad_norm": 0.7866742014884949,
+      "learning_rate": 0.00015970370993072762,
+      "loss": 1.7332,
+      "step": 3018
+    },
+    {
+      "epoch": 0.31762230405049974,
+      "grad_norm": 1.069047212600708,
+      "learning_rate": 0.0001596763691348548,
+      "loss": 2.3551,
+      "step": 3019
+    },
+    {
+      "epoch": 0.31772751183587583,
+      "grad_norm": 1.3254644870758057,
+      "learning_rate": 0.00015964902140896175,
+      "loss": 2.3056,
+      "step": 3020
+    },
+    {
+      "epoch": 0.317832719621252,
+      "grad_norm": 0.88875812292099,
+      "learning_rate": 0.00015962166675622424,
+      "loss": 2.1039,
+      "step": 3021
+    },
+    {
+      "epoch": 0.3179379274066281,
+      "grad_norm": 1.263662338256836,
+      "learning_rate": 0.0001595943051798189,
+      "loss": 2.3112,
+      "step": 3022
+    },
+    {
+      "epoch": 0.3180431351920042,
+      "grad_norm": 0.8133673667907715,
+      "learning_rate": 0.00015956693668292313,
+      "loss": 2.2302,
+      "step": 3023
+    },
+    {
+      "epoch": 0.3181483429773803,
+      "grad_norm": 3.3079943656921387,
+      "learning_rate": 0.00015953956126871517,
+      "loss": 2.2323,
+      "step": 3024
+    },
+    {
+      "epoch": 0.31825355076275647,
+      "grad_norm": 1.0143603086471558,
+      "learning_rate": 0.00015951217894037402,
+      "loss": 2.016,
+      "step": 3025
+    },
+    {
+      "epoch": 0.31835875854813256,
+      "grad_norm": 1.2601126432418823,
+      "learning_rate": 0.0001594847897010795,
+      "loss": 2.0657,
+      "step": 3026
+    },
+    {
+      "epoch": 0.31846396633350865,
+      "grad_norm": 0.7915671467781067,
+      "learning_rate": 0.00015945739355401222,
+      "loss": 1.8548,
+      "step": 3027
+    },
+    {
+      "epoch": 0.3185691741188848,
+      "grad_norm": 1.0004899501800537,
+      "learning_rate": 0.0001594299905023536,
+      "loss": 2.4322,
+      "step": 3028
+    },
+    {
+      "epoch": 0.3186743819042609,
+      "grad_norm": 0.8855611681938171,
+      "learning_rate": 0.0001594025805492859,
+      "loss": 2.1558,
+      "step": 3029
+    },
+    {
+      "epoch": 0.31877958968963704,
+      "grad_norm": 1.2269375324249268,
+      "learning_rate": 0.00015937516369799216,
+      "loss": 2.2797,
+      "step": 3030
+    },
+    {
+      "epoch": 0.31888479747501314,
+      "grad_norm": 1.0635052919387817,
+      "learning_rate": 0.00015934773995165613,
+      "loss": 2.2713,
+      "step": 3031
+    },
+    {
+      "epoch": 0.3189900052603893,
+      "grad_norm": 1.1982238292694092,
+      "learning_rate": 0.0001593203093134625,
+      "loss": 1.9763,
+      "step": 3032
+    },
+    {
+      "epoch": 0.3190952130457654,
+      "grad_norm": 1.1229660511016846,
+      "learning_rate": 0.0001592928717865967,
+      "loss": 2.0421,
+      "step": 3033
+    },
+    {
+      "epoch": 0.31920042083114153,
+      "grad_norm": 0.9558283686637878,
+      "learning_rate": 0.00015926542737424492,
+      "loss": 2.0596,
+      "step": 3034
+    },
+    {
+      "epoch": 0.3193056286165176,
+      "grad_norm": 1.1449071168899536,
+      "learning_rate": 0.00015923797607959422,
+      "loss": 2.0033,
+      "step": 3035
+    },
+    {
+      "epoch": 0.3194108364018937,
+      "grad_norm": 2.168914794921875,
+      "learning_rate": 0.00015921051790583247,
+      "loss": 1.6018,
+      "step": 3036
+    },
+    {
+      "epoch": 0.31951604418726987,
+      "grad_norm": 0.9856159090995789,
+      "learning_rate": 0.00015918305285614822,
+      "loss": 1.8371,
+      "step": 3037
+    },
+    {
+      "epoch": 0.31962125197264596,
+      "grad_norm": 1.085970163345337,
+      "learning_rate": 0.0001591555809337309,
+      "loss": 2.1444,
+      "step": 3038
+    },
+    {
+      "epoch": 0.3197264597580221,
+      "grad_norm": 1.2033791542053223,
+      "learning_rate": 0.0001591281021417708,
+      "loss": 1.6321,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3198316675433982,
+      "grad_norm": 1.5555825233459473,
+      "learning_rate": 0.0001591006164834589,
+      "loss": 1.6976,
+      "step": 3040
+    },
+    {
+      "epoch": 0.31993687532877435,
+      "grad_norm": 1.6176249980926514,
+      "learning_rate": 0.00015907312396198697,
+      "loss": 2.028,
+      "step": 3041
+    },
+    {
+      "epoch": 0.32004208311415044,
+      "grad_norm": 1.1690542697906494,
+      "learning_rate": 0.00015904562458054773,
+      "loss": 1.9713,
+      "step": 3042
+    },
+    {
+      "epoch": 0.32014729089952654,
+      "grad_norm": 1.2883795499801636,
+      "learning_rate": 0.00015901811834233452,
+      "loss": 2.1284,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3202524986849027,
+      "grad_norm": 1.3291345834732056,
+      "learning_rate": 0.00015899060525054157,
+      "loss": 1.5719,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3203577064702788,
+      "grad_norm": 1.5232844352722168,
+      "learning_rate": 0.0001589630853083639,
+      "loss": 1.8812,
+      "step": 3045
+    },
+    {
+      "epoch": 0.32046291425565493,
+      "grad_norm": 0.951919436454773,
+      "learning_rate": 0.0001589355585189973,
+      "loss": 2.0791,
+      "step": 3046
+    },
+    {
+      "epoch": 0.320568122041031,
+      "grad_norm": 1.3650517463684082,
+      "learning_rate": 0.0001589080248856383,
+      "loss": 1.834,
+      "step": 3047
+    },
+    {
+      "epoch": 0.32067332982640717,
+      "grad_norm": 1.4583851099014282,
+      "learning_rate": 0.00015888048441148442,
+      "loss": 2.0812,
+      "step": 3048
+    },
+    {
+      "epoch": 0.32077853761178327,
+      "grad_norm": 1.3375087976455688,
+      "learning_rate": 0.00015885293709973374,
+      "loss": 1.7964,
+      "step": 3049
+    },
+    {
+      "epoch": 0.3208837453971594,
+      "grad_norm": 1.1333143711090088,
+      "learning_rate": 0.0001588253829535853,
+      "loss": 2.2698,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3209889531825355,
+      "grad_norm": 0.9990382790565491,
+      "learning_rate": 0.0001587978219762388,
+      "loss": 1.4186,
+      "step": 3051
+    },
+    {
+      "epoch": 0.3210941609679116,
+      "grad_norm": 1.230421543121338,
+      "learning_rate": 0.0001587702541708949,
+      "loss": 1.8636,
+      "step": 3052
+    },
+    {
+      "epoch": 0.32119936875328775,
+      "grad_norm": 1.4861817359924316,
+      "learning_rate": 0.00015874267954075485,
+      "loss": 2.0363,
+      "step": 3053
+    },
+    {
+      "epoch": 0.32130457653866384,
+      "grad_norm": 1.4042917490005493,
+      "learning_rate": 0.0001587150980890209,
+      "loss": 1.6804,
+      "step": 3054
+    },
+    {
+      "epoch": 0.32140978432404,
+      "grad_norm": 1.5827716588974,
+      "learning_rate": 0.00015868750981889594,
+      "loss": 1.7993,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3215149921094161,
+      "grad_norm": 1.274047613143921,
+      "learning_rate": 0.00015865991473358373,
+      "loss": 1.8368,
+      "step": 3056
+    },
+    {
+      "epoch": 0.32162019989479224,
+      "grad_norm": 1.2829877138137817,
+      "learning_rate": 0.00015863231283628877,
+      "loss": 1.9838,
+      "step": 3057
+    },
+    {
+      "epoch": 0.32172540768016833,
+      "grad_norm": 0.8687334060668945,
+      "learning_rate": 0.00015860470413021642,
+      "loss": 1.5084,
+      "step": 3058
+    },
+    {
+      "epoch": 0.3218306154655444,
+      "grad_norm": 1.9248266220092773,
+      "learning_rate": 0.00015857708861857274,
+      "loss": 1.6113,
+      "step": 3059
+    },
+    {
+      "epoch": 0.32193582325092057,
+      "grad_norm": 1.0948201417922974,
+      "learning_rate": 0.00015854946630456467,
+      "loss": 2.1865,
+      "step": 3060
+    },
+    {
+      "epoch": 0.32204103103629667,
+      "grad_norm": 1.5295681953430176,
+      "learning_rate": 0.00015852183719139985,
+      "loss": 2.0163,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3221462388216728,
+      "grad_norm": 2.258373737335205,
+      "learning_rate": 0.00015849420128228678,
+      "loss": 2.2256,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3222514466070489,
+      "grad_norm": 1.5048192739486694,
+      "learning_rate": 0.00015846655858043477,
+      "loss": 1.7284,
+      "step": 3063
+    },
+    {
+      "epoch": 0.32235665439242506,
+      "grad_norm": 1.728192925453186,
+      "learning_rate": 0.0001584389090890538,
+      "loss": 2.0966,
+      "step": 3064
+    },
+    {
+      "epoch": 0.32246186217780115,
+      "grad_norm": 1.150688886642456,
+      "learning_rate": 0.00015841125281135473,
+      "loss": 1.9364,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3225670699631773,
+      "grad_norm": 1.404353141784668,
+      "learning_rate": 0.0001583835897505493,
+      "loss": 1.6709,
+      "step": 3066
+    },
+    {
+      "epoch": 0.3226722777485534,
+      "grad_norm": 1.3692048788070679,
+      "learning_rate": 0.00015835591990984974,
+      "loss": 1.8023,
+      "step": 3067
+    },
+    {
+      "epoch": 0.3227774855339295,
+      "grad_norm": 1.486565351486206,
+      "learning_rate": 0.00015832824329246946,
+      "loss": 1.8788,
+      "step": 3068
+    },
+    {
+      "epoch": 0.32288269331930564,
+      "grad_norm": 0.9628434777259827,
+      "learning_rate": 0.0001583005599016223,
+      "loss": 2.3548,
+      "step": 3069
+    },
+    {
+      "epoch": 0.32298790110468173,
+      "grad_norm": 0.9888852834701538,
+      "learning_rate": 0.0001582728697405231,
+      "loss": 2.1615,
+      "step": 3070
+    },
+    {
+      "epoch": 0.3230931088900579,
+      "grad_norm": 1.8097474575042725,
+      "learning_rate": 0.00015824517281238745,
+      "loss": 1.4427,
+      "step": 3071
+    },
+    {
+      "epoch": 0.32319831667543397,
+      "grad_norm": 1.5706349611282349,
+      "learning_rate": 0.00015821746912043165,
+      "loss": 1.9168,
+      "step": 3072
+    },
+    {
+      "epoch": 0.3233035244608101,
+      "grad_norm": 3.1145801544189453,
+      "learning_rate": 0.0001581897586678729,
+      "loss": 1.4371,
+      "step": 3073
+    },
+    {
+      "epoch": 0.3234087322461862,
+      "grad_norm": 1.5123764276504517,
+      "learning_rate": 0.00015816204145792904,
+      "loss": 1.9067,
+      "step": 3074
+    },
+    {
+      "epoch": 0.3235139400315623,
+      "grad_norm": 0.8363044261932373,
+      "learning_rate": 0.00015813431749381887,
+      "loss": 2.0816,
+      "step": 3075
+    },
+    {
+      "epoch": 0.32361914781693846,
+      "grad_norm": 1.4123753309249878,
+      "learning_rate": 0.00015810658677876184,
+      "loss": 1.9094,
+      "step": 3076
+    },
+    {
+      "epoch": 0.32372435560231455,
+      "grad_norm": 1.268897533416748,
+      "learning_rate": 0.0001580788493159782,
+      "loss": 1.6229,
+      "step": 3077
+    },
+    {
+      "epoch": 0.3238295633876907,
+      "grad_norm": 1.2455761432647705,
+      "learning_rate": 0.00015805110510868907,
+      "loss": 1.9262,
+      "step": 3078
+    },
+    {
+      "epoch": 0.3239347711730668,
+      "grad_norm": 1.2342606782913208,
+      "learning_rate": 0.00015802335416011625,
+      "loss": 1.8135,
+      "step": 3079
+    },
+    {
+      "epoch": 0.32403997895844294,
+      "grad_norm": 1.7797157764434814,
+      "learning_rate": 0.00015799559647348236,
+      "loss": 1.5395,
+      "step": 3080
+    },
+    {
+      "epoch": 0.32414518674381904,
+      "grad_norm": 1.31904935836792,
+      "learning_rate": 0.00015796783205201086,
+      "loss": 1.5797,
+      "step": 3081
+    },
+    {
+      "epoch": 0.3242503945291952,
+      "grad_norm": 1.5645530223846436,
+      "learning_rate": 0.00015794006089892587,
+      "loss": 1.6131,
+      "step": 3082
+    },
+    {
+      "epoch": 0.3243556023145713,
+      "grad_norm": 1.483992099761963,
+      "learning_rate": 0.00015791228301745245,
+      "loss": 2.2207,
+      "step": 3083
+    },
+    {
+      "epoch": 0.32446081009994737,
+      "grad_norm": 1.0606383085250854,
+      "learning_rate": 0.00015788449841081626,
+      "loss": 1.5808,
+      "step": 3084
+    },
+    {
+      "epoch": 0.3245660178853235,
+      "grad_norm": 1.3656562566757202,
+      "learning_rate": 0.00015785670708224389,
+      "loss": 2.1471,
+      "step": 3085
+    },
+    {
+      "epoch": 0.3246712256706996,
+      "grad_norm": 1.2391986846923828,
+      "learning_rate": 0.00015782890903496264,
+      "loss": 2.1729,
+      "step": 3086
+    },
+    {
+      "epoch": 0.32477643345607576,
+      "grad_norm": 1.3538752794265747,
+      "learning_rate": 0.0001578011042722006,
+      "loss": 2.518,
+      "step": 3087
+    },
+    {
+      "epoch": 0.32488164124145186,
+      "grad_norm": 1.5246939659118652,
+      "learning_rate": 0.0001577732927971867,
+      "loss": 2.2928,
+      "step": 3088
+    },
+    {
+      "epoch": 0.324986849026828,
+      "grad_norm": 1.872592806816101,
+      "learning_rate": 0.0001577454746131505,
+      "loss": 2.0892,
+      "step": 3089
+    },
+    {
+      "epoch": 0.3250920568122041,
+      "grad_norm": 1.573846697807312,
+      "learning_rate": 0.00015771764972332254,
+      "loss": 2.2432,
+      "step": 3090
+    },
+    {
+      "epoch": 0.3251972645975802,
+      "grad_norm": 0.8727262020111084,
+      "learning_rate": 0.00015768981813093393,
+      "loss": 1.7948,
+      "step": 3091
+    },
+    {
+      "epoch": 0.32530247238295634,
+      "grad_norm": 1.4603215456008911,
+      "learning_rate": 0.00015766197983921673,
+      "loss": 2.5792,
+      "step": 3092
+    },
+    {
+      "epoch": 0.32540768016833244,
+      "grad_norm": 1.198878526687622,
+      "learning_rate": 0.00015763413485140365,
+      "loss": 2.1937,
+      "step": 3093
+    },
+    {
+      "epoch": 0.3255128879537086,
+      "grad_norm": 1.553514003753662,
+      "learning_rate": 0.00015760628317072834,
+      "loss": 2.3651,
+      "step": 3094
+    },
+    {
+      "epoch": 0.3256180957390847,
+      "grad_norm": 1.511500358581543,
+      "learning_rate": 0.00015757842480042502,
+      "loss": 2.1293,
+      "step": 3095
+    },
+    {
+      "epoch": 0.3257233035244608,
+      "grad_norm": 1.5197930335998535,
+      "learning_rate": 0.00015755055974372883,
+      "loss": 2.1377,
+      "step": 3096
+    },
+    {
+      "epoch": 0.3258285113098369,
+      "grad_norm": 1.308167815208435,
+      "learning_rate": 0.00015752268800387563,
+      "loss": 1.8966,
+      "step": 3097
+    },
+    {
+      "epoch": 0.32593371909521307,
+      "grad_norm": 0.7427685260772705,
+      "learning_rate": 0.0001574948095841021,
+      "loss": 2.4438,
+      "step": 3098
+    },
+    {
+      "epoch": 0.32603892688058916,
+      "grad_norm": 1.212828278541565,
+      "learning_rate": 0.00015746692448764568,
+      "loss": 1.8894,
+      "step": 3099
+    },
+    {
+      "epoch": 0.32614413466596526,
+      "grad_norm": 1.2902162075042725,
+      "learning_rate": 0.00015743903271774455,
+      "loss": 1.7276,
+      "step": 3100
+    },
+    {
+      "epoch": 0.3262493424513414,
+      "grad_norm": 1.072805404663086,
+      "learning_rate": 0.0001574111342776377,
+      "loss": 1.969,
+      "step": 3101
+    },
+    {
+      "epoch": 0.3263545502367175,
+      "grad_norm": 3.502418279647827,
+      "learning_rate": 0.00015738322917056486,
+      "loss": 2.6105,
+      "step": 3102
+    },
+    {
+      "epoch": 0.32645975802209365,
+      "grad_norm": 1.2981481552124023,
+      "learning_rate": 0.00015735531739976657,
+      "loss": 2.0216,
+      "step": 3103
+    },
+    {
+      "epoch": 0.32656496580746974,
+      "grad_norm": 0.8937673568725586,
+      "learning_rate": 0.00015732739896848414,
+      "loss": 1.7309,
+      "step": 3104
+    },
+    {
+      "epoch": 0.3266701735928459,
+      "grad_norm": 1.0238783359527588,
+      "learning_rate": 0.00015729947387995962,
+      "loss": 1.7886,
+      "step": 3105
+    },
+    {
+      "epoch": 0.326775381378222,
+      "grad_norm": 1.0271786451339722,
+      "learning_rate": 0.00015727154213743592,
+      "loss": 1.7114,
+      "step": 3106
+    },
+    {
+      "epoch": 0.3268805891635981,
+      "grad_norm": 1.1980253458023071,
+      "learning_rate": 0.0001572436037441566,
+      "loss": 2.0706,
+      "step": 3107
+    },
+    {
+      "epoch": 0.32698579694897423,
+      "grad_norm": 1.797404408454895,
+      "learning_rate": 0.0001572156587033661,
+      "loss": 2.1992,
+      "step": 3108
+    },
+    {
+      "epoch": 0.3270910047343503,
+      "grad_norm": 1.8935017585754395,
+      "learning_rate": 0.00015718770701830955,
+      "loss": 2.225,
+      "step": 3109
+    },
+    {
+      "epoch": 0.32719621251972647,
+      "grad_norm": 1.513788104057312,
+      "learning_rate": 0.0001571597486922329,
+      "loss": 2.0227,
+      "step": 3110
+    },
+    {
+      "epoch": 0.32730142030510256,
+      "grad_norm": 0.831935703754425,
+      "learning_rate": 0.00015713178372838286,
+      "loss": 1.5778,
+      "step": 3111
+    },
+    {
+      "epoch": 0.3274066280904787,
+      "grad_norm": 1.2550475597381592,
+      "learning_rate": 0.0001571038121300069,
+      "loss": 1.7091,
+      "step": 3112
+    },
+    {
+      "epoch": 0.3275118358758548,
+      "grad_norm": 1.2048956155776978,
+      "learning_rate": 0.00015707583390035327,
+      "loss": 2.2076,
+      "step": 3113
+    },
+    {
+      "epoch": 0.32761704366123096,
+      "grad_norm": 1.4993027448654175,
+      "learning_rate": 0.00015704784904267097,
+      "loss": 2.2555,
+      "step": 3114
+    },
+    {
+      "epoch": 0.32772225144660705,
+      "grad_norm": 1.5597890615463257,
+      "learning_rate": 0.00015701985756020985,
+      "loss": 1.8569,
+      "step": 3115
+    },
+    {
+      "epoch": 0.32782745923198314,
+      "grad_norm": 1.346541166305542,
+      "learning_rate": 0.00015699185945622043,
+      "loss": 2.031,
+      "step": 3116
+    },
+    {
+      "epoch": 0.3279326670173593,
+      "grad_norm": 1.1803152561187744,
+      "learning_rate": 0.00015696385473395403,
+      "loss": 1.9566,
+      "step": 3117
+    },
+    {
+      "epoch": 0.3280378748027354,
+      "grad_norm": 1.5781463384628296,
+      "learning_rate": 0.00015693584339666279,
+      "loss": 2.0671,
+      "step": 3118
+    },
+    {
+      "epoch": 0.32814308258811153,
+      "grad_norm": 1.163499355316162,
+      "learning_rate": 0.0001569078254475995,
+      "loss": 2.3857,
+      "step": 3119
+    },
+    {
+      "epoch": 0.32824829037348763,
+      "grad_norm": 1.9654390811920166,
+      "learning_rate": 0.00015687980089001787,
+      "loss": 2.0257,
+      "step": 3120
+    },
+    {
+      "epoch": 0.3283534981588638,
+      "grad_norm": 1.0074301958084106,
+      "learning_rate": 0.00015685176972717223,
+      "loss": 2.3659,
+      "step": 3121
+    },
+    {
+      "epoch": 0.32845870594423987,
+      "grad_norm": 1.2155638933181763,
+      "learning_rate": 0.00015682373196231782,
+      "loss": 2.0024,
+      "step": 3122
+    },
+    {
+      "epoch": 0.32856391372961596,
+      "grad_norm": 1.3712031841278076,
+      "learning_rate": 0.0001567956875987105,
+      "loss": 1.7298,
+      "step": 3123
+    },
+    {
+      "epoch": 0.3286691215149921,
+      "grad_norm": 1.682850956916809,
+      "learning_rate": 0.000156767636639607,
+      "loss": 2.3542,
+      "step": 3124
+    },
+    {
+      "epoch": 0.3287743293003682,
+      "grad_norm": 1.4702867269515991,
+      "learning_rate": 0.00015673957908826479,
+      "loss": 2.0367,
+      "step": 3125
+    },
+    {
+      "epoch": 0.32887953708574436,
+      "grad_norm": 1.9041625261306763,
+      "learning_rate": 0.00015671151494794211,
+      "loss": 2.015,
+      "step": 3126
+    },
+    {
+      "epoch": 0.32898474487112045,
+      "grad_norm": 1.1154800653457642,
+      "learning_rate": 0.00015668344422189794,
+      "loss": 1.9378,
+      "step": 3127
+    },
+    {
+      "epoch": 0.3290899526564966,
+      "grad_norm": 1.1576582193374634,
+      "learning_rate": 0.00015665536691339207,
+      "loss": 2.1949,
+      "step": 3128
+    },
+    {
+      "epoch": 0.3291951604418727,
+      "grad_norm": 0.9152946472167969,
+      "learning_rate": 0.00015662728302568498,
+      "loss": 1.8082,
+      "step": 3129
+    },
+    {
+      "epoch": 0.32930036822724884,
+      "grad_norm": 1.0784518718719482,
+      "learning_rate": 0.00015659919256203795,
+      "loss": 2.1528,
+      "step": 3130
+    },
+    {
+      "epoch": 0.32940557601262493,
+      "grad_norm": 1.0181763172149658,
+      "learning_rate": 0.00015657109552571312,
+      "loss": 2.1109,
+      "step": 3131
+    },
+    {
+      "epoch": 0.32951078379800103,
+      "grad_norm": 1.47239351272583,
+      "learning_rate": 0.00015654299191997324,
+      "loss": 2.1156,
+      "step": 3132
+    },
+    {
+      "epoch": 0.3296159915833772,
+      "grad_norm": 1.092498540878296,
+      "learning_rate": 0.0001565148817480819,
+      "loss": 1.7449,
+      "step": 3133
+    },
+    {
+      "epoch": 0.32972119936875327,
+      "grad_norm": 0.9805752635002136,
+      "learning_rate": 0.00015648676501330342,
+      "loss": 1.804,
+      "step": 3134
+    },
+    {
+      "epoch": 0.3298264071541294,
+      "grad_norm": 1.1172640323638916,
+      "learning_rate": 0.00015645864171890295,
+      "loss": 1.6379,
+      "step": 3135
+    },
+    {
+      "epoch": 0.3299316149395055,
+      "grad_norm": 1.313032627105713,
+      "learning_rate": 0.0001564305118681463,
+      "loss": 2.2116,
+      "step": 3136
+    },
+    {
+      "epoch": 0.33003682272488166,
+      "grad_norm": 1.0228915214538574,
+      "learning_rate": 0.00015640237546430018,
+      "loss": 2.1348,
+      "step": 3137
+    },
+    {
+      "epoch": 0.33014203051025776,
+      "grad_norm": 1.2504109144210815,
+      "learning_rate": 0.00015637423251063185,
+      "loss": 1.7011,
+      "step": 3138
+    },
+    {
+      "epoch": 0.33024723829563385,
+      "grad_norm": 1.2419320344924927,
+      "learning_rate": 0.00015634608301040958,
+      "loss": 1.5222,
+      "step": 3139
+    },
+    {
+      "epoch": 0.33035244608101,
+      "grad_norm": 1.6612623929977417,
+      "learning_rate": 0.00015631792696690225,
+      "loss": 1.9523,
+      "step": 3140
+    },
+    {
+      "epoch": 0.3304576538663861,
+      "grad_norm": 1.404956340789795,
+      "learning_rate": 0.00015628976438337948,
+      "loss": 1.6778,
+      "step": 3141
+    },
+    {
+      "epoch": 0.33056286165176224,
+      "grad_norm": 2.2217705249786377,
+      "learning_rate": 0.00015626159526311174,
+      "loss": 1.7111,
+      "step": 3142
+    },
+    {
+      "epoch": 0.33066806943713833,
+      "grad_norm": 1.682693362236023,
+      "learning_rate": 0.0001562334196093702,
+      "loss": 1.9422,
+      "step": 3143
+    },
+    {
+      "epoch": 0.3307732772225145,
+      "grad_norm": 2.735732078552246,
+      "learning_rate": 0.00015620523742542687,
+      "loss": 1.8769,
+      "step": 3144
+    },
+    {
+      "epoch": 0.3308784850078906,
+      "grad_norm": 1.4678375720977783,
+      "learning_rate": 0.00015617704871455433,
+      "loss": 2.3066,
+      "step": 3145
+    },
+    {
+      "epoch": 0.3309836927932667,
+      "grad_norm": 1.5375454425811768,
+      "learning_rate": 0.00015614885348002612,
+      "loss": 1.7219,
+      "step": 3146
+    },
+    {
+      "epoch": 0.3310889005786428,
+      "grad_norm": 1.09225332736969,
+      "learning_rate": 0.00015612065172511646,
+      "loss": 1.9515,
+      "step": 3147
+    },
+    {
+      "epoch": 0.3311941083640189,
+      "grad_norm": 1.3868629932403564,
+      "learning_rate": 0.0001560924434531003,
+      "loss": 2.0467,
+      "step": 3148
+    },
+    {
+      "epoch": 0.33129931614939506,
+      "grad_norm": 1.3525406122207642,
+      "learning_rate": 0.00015606422866725343,
+      "loss": 1.7597,
+      "step": 3149
+    },
+    {
+      "epoch": 0.33140452393477116,
+      "grad_norm": 1.2624315023422241,
+      "learning_rate": 0.00015603600737085227,
+      "loss": 1.9116,
+      "step": 3150
+    },
+    {
+      "epoch": 0.3315097317201473,
+      "grad_norm": 1.287229299545288,
+      "learning_rate": 0.00015600777956717408,
+      "loss": 1.8875,
+      "step": 3151
+    },
+    {
+      "epoch": 0.3316149395055234,
+      "grad_norm": 1.0664699077606201,
+      "learning_rate": 0.0001559795452594969,
+      "loss": 1.8313,
+      "step": 3152
+    },
+    {
+      "epoch": 0.33172014729089955,
+      "grad_norm": 1.5198394060134888,
+      "learning_rate": 0.00015595130445109946,
+      "loss": 1.6447,
+      "step": 3153
+    },
+    {
+      "epoch": 0.33182535507627564,
+      "grad_norm": 1.22080659866333,
+      "learning_rate": 0.0001559230571452613,
+      "loss": 2.5198,
+      "step": 3154
+    },
+    {
+      "epoch": 0.33193056286165173,
+      "grad_norm": 1.2474714517593384,
+      "learning_rate": 0.00015589480334526266,
+      "loss": 1.736,
+      "step": 3155
+    },
+    {
+      "epoch": 0.3320357706470279,
+      "grad_norm": 1.2309380769729614,
+      "learning_rate": 0.00015586654305438456,
+      "loss": 1.5404,
+      "step": 3156
+    },
+    {
+      "epoch": 0.332140978432404,
+      "grad_norm": 1.3968786001205444,
+      "learning_rate": 0.00015583827627590875,
+      "loss": 1.7834,
+      "step": 3157
+    },
+    {
+      "epoch": 0.3322461862177801,
+      "grad_norm": 1.3345340490341187,
+      "learning_rate": 0.00015581000301311782,
+      "loss": 1.7573,
+      "step": 3158
+    },
+    {
+      "epoch": 0.3323513940031562,
+      "grad_norm": 1.5625921487808228,
+      "learning_rate": 0.00015578172326929498,
+      "loss": 1.8578,
+      "step": 3159
+    },
+    {
+      "epoch": 0.33245660178853237,
+      "grad_norm": 0.9418458342552185,
+      "learning_rate": 0.0001557534370477243,
+      "loss": 2.0889,
+      "step": 3160
+    },
+    {
+      "epoch": 0.33256180957390846,
+      "grad_norm": 1.5350128412246704,
+      "learning_rate": 0.00015572514435169063,
+      "loss": 1.1002,
+      "step": 3161
+    },
+    {
+      "epoch": 0.3326670173592846,
+      "grad_norm": 1.865135908126831,
+      "learning_rate": 0.0001556968451844794,
+      "loss": 1.707,
+      "step": 3162
+    },
+    {
+      "epoch": 0.3327722251446607,
+      "grad_norm": 1.427790641784668,
+      "learning_rate": 0.00015566853954937694,
+      "loss": 1.8425,
+      "step": 3163
+    },
+    {
+      "epoch": 0.3328774329300368,
+      "grad_norm": 1.1270205974578857,
+      "learning_rate": 0.0001556402274496703,
+      "loss": 1.5166,
+      "step": 3164
+    },
+    {
+      "epoch": 0.33298264071541295,
+      "grad_norm": 1.270928978919983,
+      "learning_rate": 0.0001556119088886473,
+      "loss": 1.4399,
+      "step": 3165
+    },
+    {
+      "epoch": 0.33308784850078904,
+      "grad_norm": 1.3536688089370728,
+      "learning_rate": 0.0001555835838695964,
+      "loss": 1.5239,
+      "step": 3166
+    },
+    {
+      "epoch": 0.3331930562861652,
+      "grad_norm": 1.0648913383483887,
+      "learning_rate": 0.00015555525239580698,
+      "loss": 2.1152,
+      "step": 3167
+    },
+    {
+      "epoch": 0.3332982640715413,
+      "grad_norm": 1.2379913330078125,
+      "learning_rate": 0.00015552691447056903,
+      "loss": 2.2395,
+      "step": 3168
+    },
+    {
+      "epoch": 0.33340347185691743,
+      "grad_norm": 1.243102788925171,
+      "learning_rate": 0.0001554985700971733,
+      "loss": 1.8046,
+      "step": 3169
+    },
+    {
+      "epoch": 0.3335086796422935,
+      "grad_norm": 1.4841232299804688,
+      "learning_rate": 0.00015547021927891144,
+      "loss": 1.4817,
+      "step": 3170
+    },
+    {
+      "epoch": 0.3336138874276697,
+      "grad_norm": 1.3383890390396118,
+      "learning_rate": 0.00015544186201907562,
+      "loss": 1.7118,
+      "step": 3171
+    },
+    {
+      "epoch": 0.33371909521304577,
+      "grad_norm": 1.7296276092529297,
+      "learning_rate": 0.00015541349832095896,
+      "loss": 1.689,
+      "step": 3172
+    },
+    {
+      "epoch": 0.33382430299842186,
+      "grad_norm": 1.2010234594345093,
+      "learning_rate": 0.0001553851281878552,
+      "loss": 2.0383,
+      "step": 3173
+    },
+    {
+      "epoch": 0.333929510783798,
+      "grad_norm": 1.05833899974823,
+      "learning_rate": 0.00015535675162305887,
+      "loss": 1.6341,
+      "step": 3174
+    },
+    {
+      "epoch": 0.3340347185691741,
+      "grad_norm": 1.8304420709609985,
+      "learning_rate": 0.0001553283686298653,
+      "loss": 2.2299,
+      "step": 3175
+    },
+    {
+      "epoch": 0.33413992635455025,
+      "grad_norm": 1.1802849769592285,
+      "learning_rate": 0.00015529997921157044,
+      "loss": 1.7604,
+      "step": 3176
+    },
+    {
+      "epoch": 0.33424513413992635,
+      "grad_norm": 1.0604053735733032,
+      "learning_rate": 0.00015527158337147112,
+      "loss": 2.3878,
+      "step": 3177
+    },
+    {
+      "epoch": 0.3343503419253025,
+      "grad_norm": 1.17803955078125,
+      "learning_rate": 0.0001552431811128648,
+      "loss": 2.2497,
+      "step": 3178
+    },
+    {
+      "epoch": 0.3344555497106786,
+      "grad_norm": 1.691798210144043,
+      "learning_rate": 0.0001552147724390498,
+      "loss": 2.2167,
+      "step": 3179
+    },
+    {
+      "epoch": 0.3345607574960547,
+      "grad_norm": 1.0652263164520264,
+      "learning_rate": 0.00015518635735332507,
+      "loss": 1.471,
+      "step": 3180
+    },
+    {
+      "epoch": 0.33466596528143083,
+      "grad_norm": 1.2501590251922607,
+      "learning_rate": 0.00015515793585899038,
+      "loss": 1.8941,
+      "step": 3181
+    },
+    {
+      "epoch": 0.3347711730668069,
+      "grad_norm": 0.7410596013069153,
+      "learning_rate": 0.00015512950795934627,
+      "loss": 1.897,
+      "step": 3182
+    },
+    {
+      "epoch": 0.3348763808521831,
+      "grad_norm": 1.589612603187561,
+      "learning_rate": 0.0001551010736576939,
+      "loss": 2.0134,
+      "step": 3183
+    },
+    {
+      "epoch": 0.33498158863755917,
+      "grad_norm": 1.5273739099502563,
+      "learning_rate": 0.00015507263295733528,
+      "loss": 2.0521,
+      "step": 3184
+    },
+    {
+      "epoch": 0.3350867964229353,
+      "grad_norm": 1.4499183893203735,
+      "learning_rate": 0.00015504418586157316,
+      "loss": 2.2479,
+      "step": 3185
+    },
+    {
+      "epoch": 0.3351920042083114,
+      "grad_norm": 1.040443778038025,
+      "learning_rate": 0.000155015732373711,
+      "loss": 2.3157,
+      "step": 3186
+    },
+    {
+      "epoch": 0.33529721199368756,
+      "grad_norm": 1.6593208312988281,
+      "learning_rate": 0.000154987272497053,
+      "loss": 2.2596,
+      "step": 3187
+    },
+    {
+      "epoch": 0.33540241977906365,
+      "grad_norm": 2.389070510864258,
+      "learning_rate": 0.0001549588062349041,
+      "loss": 1.6402,
+      "step": 3188
+    },
+    {
+      "epoch": 0.33550762756443975,
+      "grad_norm": 1.5094181299209595,
+      "learning_rate": 0.00015493033359057003,
+      "loss": 1.8529,
+      "step": 3189
+    },
+    {
+      "epoch": 0.3356128353498159,
+      "grad_norm": 0.9902091026306152,
+      "learning_rate": 0.00015490185456735719,
+      "loss": 2.0056,
+      "step": 3190
+    },
+    {
+      "epoch": 0.335718043135192,
+      "grad_norm": 1.285557508468628,
+      "learning_rate": 0.00015487336916857278,
+      "loss": 2.1874,
+      "step": 3191
+    },
+    {
+      "epoch": 0.33582325092056814,
+      "grad_norm": 1.2860695123672485,
+      "learning_rate": 0.00015484487739752468,
+      "loss": 2.1247,
+      "step": 3192
+    },
+    {
+      "epoch": 0.33592845870594423,
+      "grad_norm": 1.0790120363235474,
+      "learning_rate": 0.00015481637925752155,
+      "loss": 2.0198,
+      "step": 3193
+    },
+    {
+      "epoch": 0.3360336664913204,
+      "grad_norm": 1.1655452251434326,
+      "learning_rate": 0.00015478787475187283,
+      "loss": 1.522,
+      "step": 3194
+    },
+    {
+      "epoch": 0.3361388742766965,
+      "grad_norm": 1.2023252248764038,
+      "learning_rate": 0.00015475936388388862,
+      "loss": 1.7807,
+      "step": 3195
+    },
+    {
+      "epoch": 0.33624408206207257,
+      "grad_norm": 1.0464221239089966,
+      "learning_rate": 0.00015473084665687984,
+      "loss": 1.6904,
+      "step": 3196
+    },
+    {
+      "epoch": 0.3363492898474487,
+      "grad_norm": 0.9979501366615295,
+      "learning_rate": 0.00015470232307415803,
+      "loss": 1.8852,
+      "step": 3197
+    },
+    {
+      "epoch": 0.3364544976328248,
+      "grad_norm": 1.2959654331207275,
+      "learning_rate": 0.00015467379313903557,
+      "loss": 2.162,
+      "step": 3198
+    },
+    {
+      "epoch": 0.33655970541820096,
+      "grad_norm": 1.7765225172042847,
+      "learning_rate": 0.00015464525685482557,
+      "loss": 1.5029,
+      "step": 3199
+    },
+    {
+      "epoch": 0.33666491320357705,
+      "grad_norm": 1.5161840915679932,
+      "learning_rate": 0.00015461671422484178,
+      "loss": 1.7194,
+      "step": 3200
+    },
+    {
+      "epoch": 0.3367701209889532,
+      "grad_norm": 1.1318210363388062,
+      "learning_rate": 0.00015458816525239886,
+      "loss": 1.8918,
+      "step": 3201
+    },
+    {
+      "epoch": 0.3368753287743293,
+      "grad_norm": 1.2241750955581665,
+      "learning_rate": 0.00015455960994081205,
+      "loss": 1.732,
+      "step": 3202
+    },
+    {
+      "epoch": 0.33698053655970545,
+      "grad_norm": 1.930464267730713,
+      "learning_rate": 0.0001545310482933974,
+      "loss": 1.8916,
+      "step": 3203
+    },
+    {
+      "epoch": 0.33708574434508154,
+      "grad_norm": 1.4464246034622192,
+      "learning_rate": 0.0001545024803134717,
+      "loss": 1.728,
+      "step": 3204
+    },
+    {
+      "epoch": 0.33719095213045763,
+      "grad_norm": 1.4596894979476929,
+      "learning_rate": 0.00015447390600435238,
+      "loss": 1.5601,
+      "step": 3205
+    },
+    {
+      "epoch": 0.3372961599158338,
+      "grad_norm": 1.1431268453598022,
+      "learning_rate": 0.00015444532536935777,
+      "loss": 2.2355,
+      "step": 3206
+    },
+    {
+      "epoch": 0.3374013677012099,
+      "grad_norm": 1.3519384860992432,
+      "learning_rate": 0.0001544167384118068,
+      "loss": 1.5459,
+      "step": 3207
+    },
+    {
+      "epoch": 0.337506575486586,
+      "grad_norm": 1.1788008213043213,
+      "learning_rate": 0.00015438814513501922,
+      "loss": 2.1492,
+      "step": 3208
+    },
+    {
+      "epoch": 0.3376117832719621,
+      "grad_norm": 1.527076244354248,
+      "learning_rate": 0.00015435954554231541,
+      "loss": 1.882,
+      "step": 3209
+    },
+    {
+      "epoch": 0.33771699105733827,
+      "grad_norm": 1.9754518270492554,
+      "learning_rate": 0.0001543309396370166,
+      "loss": 1.8531,
+      "step": 3210
+    },
+    {
+      "epoch": 0.33782219884271436,
+      "grad_norm": 1.0372203588485718,
+      "learning_rate": 0.00015430232742244467,
+      "loss": 2.1218,
+      "step": 3211
+    },
+    {
+      "epoch": 0.33792740662809045,
+      "grad_norm": 1.702439785003662,
+      "learning_rate": 0.00015427370890192224,
+      "loss": 1.7968,
+      "step": 3212
+    },
+    {
+      "epoch": 0.3380326144134666,
+      "grad_norm": 1.84778892993927,
+      "learning_rate": 0.0001542450840787727,
+      "loss": 1.9588,
+      "step": 3213
+    },
+    {
+      "epoch": 0.3381378221988427,
+      "grad_norm": 1.402566909790039,
+      "learning_rate": 0.00015421645295632023,
+      "loss": 2.245,
+      "step": 3214
+    },
+    {
+      "epoch": 0.33824302998421885,
+      "grad_norm": 1.550502061843872,
+      "learning_rate": 0.0001541878155378896,
+      "loss": 1.999,
+      "step": 3215
+    },
+    {
+      "epoch": 0.33834823776959494,
+      "grad_norm": 2.043599843978882,
+      "learning_rate": 0.00015415917182680638,
+      "loss": 1.7732,
+      "step": 3216
+    },
+    {
+      "epoch": 0.3384534455549711,
+      "grad_norm": 1.1463595628738403,
+      "learning_rate": 0.00015413052182639683,
+      "loss": 1.9726,
+      "step": 3217
+    },
+    {
+      "epoch": 0.3385586533403472,
+      "grad_norm": 1.1954436302185059,
+      "learning_rate": 0.0001541018655399881,
+      "loss": 2.0425,
+      "step": 3218
+    },
+    {
+      "epoch": 0.33866386112572333,
+      "grad_norm": 1.8237320184707642,
+      "learning_rate": 0.00015407320297090786,
+      "loss": 2.3845,
+      "step": 3219
+    },
+    {
+      "epoch": 0.3387690689110994,
+      "grad_norm": 1.442068338394165,
+      "learning_rate": 0.0001540445341224846,
+      "loss": 1.988,
+      "step": 3220
+    },
+    {
+      "epoch": 0.3388742766964755,
+      "grad_norm": 1.3725106716156006,
+      "learning_rate": 0.00015401585899804755,
+      "loss": 2.2294,
+      "step": 3221
+    },
+    {
+      "epoch": 0.33897948448185167,
+      "grad_norm": 1.1935747861862183,
+      "learning_rate": 0.00015398717760092666,
+      "loss": 2.0404,
+      "step": 3222
+    },
+    {
+      "epoch": 0.33908469226722776,
+      "grad_norm": 1.4062366485595703,
+      "learning_rate": 0.00015395848993445265,
+      "loss": 1.838,
+      "step": 3223
+    },
+    {
+      "epoch": 0.3391899000526039,
+      "grad_norm": 1.1401127576828003,
+      "learning_rate": 0.00015392979600195684,
+      "loss": 1.7744,
+      "step": 3224
+    },
+    {
+      "epoch": 0.33929510783798,
+      "grad_norm": 1.2197401523590088,
+      "learning_rate": 0.00015390109580677144,
+      "loss": 1.9057,
+      "step": 3225
+    },
+    {
+      "epoch": 0.33940031562335615,
+      "grad_norm": 1.2865253686904907,
+      "learning_rate": 0.00015387238935222927,
+      "loss": 1.8107,
+      "step": 3226
+    },
+    {
+      "epoch": 0.33950552340873225,
+      "grad_norm": 1.3906419277191162,
+      "learning_rate": 0.0001538436766416639,
+      "loss": 1.5141,
+      "step": 3227
+    },
+    {
+      "epoch": 0.33961073119410834,
+      "grad_norm": 1.4416093826293945,
+      "learning_rate": 0.00015381495767840967,
+      "loss": 1.9024,
+      "step": 3228
+    },
+    {
+      "epoch": 0.3397159389794845,
+      "grad_norm": 1.0017902851104736,
+      "learning_rate": 0.00015378623246580165,
+      "loss": 1.5883,
+      "step": 3229
+    },
+    {
+      "epoch": 0.3398211467648606,
+      "grad_norm": 1.0945497751235962,
+      "learning_rate": 0.00015375750100717555,
+      "loss": 2.3554,
+      "step": 3230
+    },
+    {
+      "epoch": 0.33992635455023673,
+      "grad_norm": 1.2821848392486572,
+      "learning_rate": 0.00015372876330586784,
+      "loss": 2.1862,
+      "step": 3231
+    },
+    {
+      "epoch": 0.3400315623356128,
+      "grad_norm": 0.8067452311515808,
+      "learning_rate": 0.00015370001936521583,
+      "loss": 2.2071,
+      "step": 3232
+    },
+    {
+      "epoch": 0.340136770120989,
+      "grad_norm": 1.01970374584198,
+      "learning_rate": 0.00015367126918855738,
+      "loss": 1.802,
+      "step": 3233
+    },
+    {
+      "epoch": 0.34024197790636507,
+      "grad_norm": 0.9139577746391296,
+      "learning_rate": 0.00015364251277923114,
+      "loss": 1.951,
+      "step": 3234
+    },
+    {
+      "epoch": 0.3403471856917412,
+      "grad_norm": 1.34104323387146,
+      "learning_rate": 0.00015361375014057656,
+      "loss": 1.7029,
+      "step": 3235
+    },
+    {
+      "epoch": 0.3404523934771173,
+      "grad_norm": 1.2936007976531982,
+      "learning_rate": 0.00015358498127593376,
+      "loss": 2.0298,
+      "step": 3236
+    },
+    {
+      "epoch": 0.3405576012624934,
+      "grad_norm": 1.3565287590026855,
+      "learning_rate": 0.00015355620618864348,
+      "loss": 1.7207,
+      "step": 3237
+    },
+    {
+      "epoch": 0.34066280904786955,
+      "grad_norm": 1.0391281843185425,
+      "learning_rate": 0.00015352742488204733,
+      "loss": 1.8902,
+      "step": 3238
+    },
+    {
+      "epoch": 0.34076801683324565,
+      "grad_norm": 1.0751245021820068,
+      "learning_rate": 0.0001534986373594876,
+      "loss": 1.6464,
+      "step": 3239
+    },
+    {
+      "epoch": 0.3408732246186218,
+      "grad_norm": 1.167717456817627,
+      "learning_rate": 0.0001534698436243073,
+      "loss": 1.8188,
+      "step": 3240
+    },
+    {
+      "epoch": 0.3409784324039979,
+      "grad_norm": 1.1624265909194946,
+      "learning_rate": 0.00015344104367985014,
+      "loss": 2.048,
+      "step": 3241
+    },
+    {
+      "epoch": 0.34108364018937404,
+      "grad_norm": 1.119253158569336,
+      "learning_rate": 0.00015341223752946052,
+      "loss": 2.1498,
+      "step": 3242
+    },
+    {
+      "epoch": 0.34118884797475013,
+      "grad_norm": 2.0956761837005615,
+      "learning_rate": 0.00015338342517648367,
+      "loss": 2.0089,
+      "step": 3243
+    },
+    {
+      "epoch": 0.3412940557601262,
+      "grad_norm": 2.5349323749542236,
+      "learning_rate": 0.0001533546066242654,
+      "loss": 1.4671,
+      "step": 3244
+    },
+    {
+      "epoch": 0.3413992635455024,
+      "grad_norm": 1.1138498783111572,
+      "learning_rate": 0.0001533257818761524,
+      "loss": 2.0825,
+      "step": 3245
+    },
+    {
+      "epoch": 0.34150447133087847,
+      "grad_norm": 1.0680429935455322,
+      "learning_rate": 0.00015329695093549192,
+      "loss": 1.474,
+      "step": 3246
+    },
+    {
+      "epoch": 0.3416096791162546,
+      "grad_norm": 1.8456655740737915,
+      "learning_rate": 0.00015326811380563204,
+      "loss": 2.3713,
+      "step": 3247
+    },
+    {
+      "epoch": 0.3417148869016307,
+      "grad_norm": 1.1764317750930786,
+      "learning_rate": 0.0001532392704899215,
+      "loss": 2.155,
+      "step": 3248
+    },
+    {
+      "epoch": 0.34182009468700686,
+      "grad_norm": 1.6202481985092163,
+      "learning_rate": 0.0001532104209917098,
+      "loss": 1.6939,
+      "step": 3249
+    },
+    {
+      "epoch": 0.34192530247238295,
+      "grad_norm": 0.7920756340026855,
+      "learning_rate": 0.00015318156531434713,
+      "loss": 1.8199,
+      "step": 3250
+    },
+    {
+      "epoch": 0.3420305102577591,
+      "grad_norm": 1.4184741973876953,
+      "learning_rate": 0.00015315270346118442,
+      "loss": 1.917,
+      "step": 3251
+    },
+    {
+      "epoch": 0.3421357180431352,
+      "grad_norm": 1.0654038190841675,
+      "learning_rate": 0.00015312383543557328,
+      "loss": 1.8263,
+      "step": 3252
+    },
+    {
+      "epoch": 0.3422409258285113,
+      "grad_norm": 0.845812976360321,
+      "learning_rate": 0.00015309496124086603,
+      "loss": 1.9213,
+      "step": 3253
+    },
+    {
+      "epoch": 0.34234613361388744,
+      "grad_norm": 1.5473440885543823,
+      "learning_rate": 0.0001530660808804158,
+      "loss": 1.8193,
+      "step": 3254
+    },
+    {
+      "epoch": 0.34245134139926353,
+      "grad_norm": 2.408548593521118,
+      "learning_rate": 0.00015303719435757633,
+      "loss": 1.9995,
+      "step": 3255
+    },
+    {
+      "epoch": 0.3425565491846397,
+      "grad_norm": 1.4737664461135864,
+      "learning_rate": 0.0001530083016757021,
+      "loss": 1.8523,
+      "step": 3256
+    },
+    {
+      "epoch": 0.3426617569700158,
+      "grad_norm": 1.6030080318450928,
+      "learning_rate": 0.0001529794028381484,
+      "loss": 1.8994,
+      "step": 3257
+    },
+    {
+      "epoch": 0.3427669647553919,
+      "grad_norm": 1.5346298217773438,
+      "learning_rate": 0.00015295049784827108,
+      "loss": 1.7178,
+      "step": 3258
+    },
+    {
+      "epoch": 0.342872172540768,
+      "grad_norm": 1.2144919633865356,
+      "learning_rate": 0.0001529215867094268,
+      "loss": 2.0721,
+      "step": 3259
+    },
+    {
+      "epoch": 0.3429773803261441,
+      "grad_norm": 1.6161445379257202,
+      "learning_rate": 0.00015289266942497293,
+      "loss": 2.2569,
+      "step": 3260
+    },
+    {
+      "epoch": 0.34308258811152026,
+      "grad_norm": 1.4736202955245972,
+      "learning_rate": 0.00015286374599826754,
+      "loss": 1.9167,
+      "step": 3261
+    },
+    {
+      "epoch": 0.34318779589689635,
+      "grad_norm": 1.4252387285232544,
+      "learning_rate": 0.0001528348164326694,
+      "loss": 2.426,
+      "step": 3262
+    },
+    {
+      "epoch": 0.3432930036822725,
+      "grad_norm": 0.9563824534416199,
+      "learning_rate": 0.000152805880731538,
+      "loss": 2.0023,
+      "step": 3263
+    },
+    {
+      "epoch": 0.3433982114676486,
+      "grad_norm": 1.182991862297058,
+      "learning_rate": 0.00015277693889823355,
+      "loss": 1.8019,
+      "step": 3264
+    },
+    {
+      "epoch": 0.34350341925302474,
+      "grad_norm": 0.8522358536720276,
+      "learning_rate": 0.000152747990936117,
+      "loss": 1.7264,
+      "step": 3265
+    },
+    {
+      "epoch": 0.34360862703840084,
+      "grad_norm": 1.2024897336959839,
+      "learning_rate": 0.0001527190368485499,
+      "loss": 1.8793,
+      "step": 3266
+    },
+    {
+      "epoch": 0.343713834823777,
+      "grad_norm": 1.762102484703064,
+      "learning_rate": 0.0001526900766388947,
+      "loss": 1.6688,
+      "step": 3267
+    },
+    {
+      "epoch": 0.3438190426091531,
+      "grad_norm": 1.8461966514587402,
+      "learning_rate": 0.00015266111031051442,
+      "loss": 1.385,
+      "step": 3268
+    },
+    {
+      "epoch": 0.3439242503945292,
+      "grad_norm": 1.0345498323440552,
+      "learning_rate": 0.00015263213786677278,
+      "loss": 1.694,
+      "step": 3269
+    },
+    {
+      "epoch": 0.3440294581799053,
+      "grad_norm": 1.057798981666565,
+      "learning_rate": 0.00015260315931103427,
+      "loss": 1.4546,
+      "step": 3270
+    },
+    {
+      "epoch": 0.3441346659652814,
+      "grad_norm": 1.6364519596099854,
+      "learning_rate": 0.00015257417464666412,
+      "loss": 1.8252,
+      "step": 3271
+    },
+    {
+      "epoch": 0.34423987375065757,
+      "grad_norm": 0.9418036341667175,
+      "learning_rate": 0.0001525451838770282,
+      "loss": 2.3333,
+      "step": 3272
+    },
+    {
+      "epoch": 0.34434508153603366,
+      "grad_norm": 1.4215340614318848,
+      "learning_rate": 0.00015251618700549307,
+      "loss": 1.7159,
+      "step": 3273
+    },
+    {
+      "epoch": 0.3444502893214098,
+      "grad_norm": 0.8506739139556885,
+      "learning_rate": 0.0001524871840354261,
+      "loss": 2.1433,
+      "step": 3274
+    },
+    {
+      "epoch": 0.3445554971067859,
+      "grad_norm": 1.7179830074310303,
+      "learning_rate": 0.00015245817497019524,
+      "loss": 2.0006,
+      "step": 3275
+    },
+    {
+      "epoch": 0.344660704892162,
+      "grad_norm": 1.3353736400604248,
+      "learning_rate": 0.0001524291598131693,
+      "loss": 2.0854,
+      "step": 3276
+    },
+    {
+      "epoch": 0.34476591267753814,
+      "grad_norm": 1.2626408338546753,
+      "learning_rate": 0.00015240013856771768,
+      "loss": 1.9261,
+      "step": 3277
+    },
+    {
+      "epoch": 0.34487112046291424,
+      "grad_norm": 1.2768040895462036,
+      "learning_rate": 0.00015237111123721052,
+      "loss": 2.1543,
+      "step": 3278
+    },
+    {
+      "epoch": 0.3449763282482904,
+      "grad_norm": 1.1452258825302124,
+      "learning_rate": 0.00015234207782501865,
+      "loss": 1.9403,
+      "step": 3279
+    },
+    {
+      "epoch": 0.3450815360336665,
+      "grad_norm": 1.7478617429733276,
+      "learning_rate": 0.0001523130383345136,
+      "loss": 1.8556,
+      "step": 3280
+    },
+    {
+      "epoch": 0.34518674381904263,
+      "grad_norm": 1.261497974395752,
+      "learning_rate": 0.00015228399276906774,
+      "loss": 2.3515,
+      "step": 3281
+    },
+    {
+      "epoch": 0.3452919516044187,
+      "grad_norm": 2.1745786666870117,
+      "learning_rate": 0.00015225494113205393,
+      "loss": 1.8923,
+      "step": 3282
+    },
+    {
+      "epoch": 0.34539715938979487,
+      "grad_norm": 1.2414960861206055,
+      "learning_rate": 0.0001522258834268459,
+      "loss": 1.8377,
+      "step": 3283
+    },
+    {
+      "epoch": 0.34550236717517097,
+      "grad_norm": 1.666604995727539,
+      "learning_rate": 0.00015219681965681798,
+      "loss": 2.1966,
+      "step": 3284
+    },
+    {
+      "epoch": 0.34560757496054706,
+      "grad_norm": 2.886913776397705,
+      "learning_rate": 0.0001521677498253453,
+      "loss": 1.9478,
+      "step": 3285
+    },
+    {
+      "epoch": 0.3457127827459232,
+      "grad_norm": 1.258230447769165,
+      "learning_rate": 0.00015213867393580358,
+      "loss": 2.0871,
+      "step": 3286
+    },
+    {
+      "epoch": 0.3458179905312993,
+      "grad_norm": 1.2245984077453613,
+      "learning_rate": 0.0001521095919915694,
+      "loss": 1.9927,
+      "step": 3287
+    },
+    {
+      "epoch": 0.34592319831667545,
+      "grad_norm": 1.6182457208633423,
+      "learning_rate": 0.00015208050399601985,
+      "loss": 1.6089,
+      "step": 3288
+    },
+    {
+      "epoch": 0.34602840610205154,
+      "grad_norm": 1.1847516298294067,
+      "learning_rate": 0.00015205140995253283,
+      "loss": 2.0237,
+      "step": 3289
+    },
+    {
+      "epoch": 0.3461336138874277,
+      "grad_norm": 1.0616167783737183,
+      "learning_rate": 0.00015202230986448704,
+      "loss": 1.7928,
+      "step": 3290
+    },
+    {
+      "epoch": 0.3462388216728038,
+      "grad_norm": 1.7937065362930298,
+      "learning_rate": 0.0001519932037352617,
+      "loss": 2.0342,
+      "step": 3291
+    },
+    {
+      "epoch": 0.3463440294581799,
+      "grad_norm": 1.3131473064422607,
+      "learning_rate": 0.0001519640915682368,
+      "loss": 1.8552,
+      "step": 3292
+    },
+    {
+      "epoch": 0.34644923724355603,
+      "grad_norm": 1.2970378398895264,
+      "learning_rate": 0.0001519349733667931,
+      "loss": 1.7052,
+      "step": 3293
+    },
+    {
+      "epoch": 0.3465544450289321,
+      "grad_norm": 1.1303088665008545,
+      "learning_rate": 0.00015190584913431194,
+      "loss": 1.8543,
+      "step": 3294
+    },
+    {
+      "epoch": 0.3466596528143083,
+      "grad_norm": 1.530392050743103,
+      "learning_rate": 0.00015187671887417542,
+      "loss": 2.0367,
+      "step": 3295
+    },
+    {
+      "epoch": 0.34676486059968437,
+      "grad_norm": 1.7276886701583862,
+      "learning_rate": 0.00015184758258976637,
+      "loss": 2.6418,
+      "step": 3296
+    },
+    {
+      "epoch": 0.3468700683850605,
+      "grad_norm": 1.323357343673706,
+      "learning_rate": 0.0001518184402844683,
+      "loss": 2.1965,
+      "step": 3297
+    },
+    {
+      "epoch": 0.3469752761704366,
+      "grad_norm": 4.387045383453369,
+      "learning_rate": 0.00015178929196166537,
+      "loss": 1.9068,
+      "step": 3298
+    },
+    {
+      "epoch": 0.34708048395581276,
+      "grad_norm": 1.7642402648925781,
+      "learning_rate": 0.00015176013762474252,
+      "loss": 1.7479,
+      "step": 3299
+    },
+    {
+      "epoch": 0.34718569174118885,
+      "grad_norm": 1.1075654029846191,
+      "learning_rate": 0.00015173097727708533,
+      "loss": 1.7473,
+      "step": 3300
+    },
+    {
+      "epoch": 0.34729089952656494,
+      "grad_norm": 1.3891255855560303,
+      "learning_rate": 0.0001517018109220801,
+      "loss": 1.863,
+      "step": 3301
+    },
+    {
+      "epoch": 0.3473961073119411,
+      "grad_norm": 0.9241618514060974,
+      "learning_rate": 0.0001516726385631138,
+      "loss": 2.2975,
+      "step": 3302
+    },
+    {
+      "epoch": 0.3475013150973172,
+      "grad_norm": 2.6608288288116455,
+      "learning_rate": 0.00015164346020357417,
+      "loss": 2.5467,
+      "step": 3303
+    },
+    {
+      "epoch": 0.34760652288269334,
+      "grad_norm": 1.2339940071105957,
+      "learning_rate": 0.00015161427584684954,
+      "loss": 1.4252,
+      "step": 3304
+    },
+    {
+      "epoch": 0.34771173066806943,
+      "grad_norm": 1.0386180877685547,
+      "learning_rate": 0.00015158508549632902,
+      "loss": 1.5439,
+      "step": 3305
+    },
+    {
+      "epoch": 0.3478169384534456,
+      "grad_norm": 1.0006986856460571,
+      "learning_rate": 0.0001515558891554024,
+      "loss": 2.3416,
+      "step": 3306
+    },
+    {
+      "epoch": 0.3479221462388217,
+      "grad_norm": 1.0212254524230957,
+      "learning_rate": 0.0001515266868274601,
+      "loss": 1.6541,
+      "step": 3307
+    },
+    {
+      "epoch": 0.34802735402419777,
+      "grad_norm": 1.2508071660995483,
+      "learning_rate": 0.0001514974785158934,
+      "loss": 1.5524,
+      "step": 3308
+    },
+    {
+      "epoch": 0.3481325618095739,
+      "grad_norm": 1.5572774410247803,
+      "learning_rate": 0.00015146826422409405,
+      "loss": 2.1225,
+      "step": 3309
+    },
+    {
+      "epoch": 0.34823776959495,
+      "grad_norm": 1.8511630296707153,
+      "learning_rate": 0.00015143904395545466,
+      "loss": 1.6325,
+      "step": 3310
+    },
+    {
+      "epoch": 0.34834297738032616,
+      "grad_norm": 1.0825985670089722,
+      "learning_rate": 0.00015140981771336848,
+      "loss": 1.9844,
+      "step": 3311
+    },
+    {
+      "epoch": 0.34844818516570225,
+      "grad_norm": 1.2613874673843384,
+      "learning_rate": 0.00015138058550122945,
+      "loss": 1.5684,
+      "step": 3312
+    },
+    {
+      "epoch": 0.3485533929510784,
+      "grad_norm": 1.5934251546859741,
+      "learning_rate": 0.00015135134732243227,
+      "loss": 2.6311,
+      "step": 3313
+    },
+    {
+      "epoch": 0.3486586007364545,
+      "grad_norm": 1.325727939605713,
+      "learning_rate": 0.00015132210318037214,
+      "loss": 1.2294,
+      "step": 3314
+    },
+    {
+      "epoch": 0.34876380852183064,
+      "grad_norm": 1.7962899208068848,
+      "learning_rate": 0.00015129285307844523,
+      "loss": 2.3393,
+      "step": 3315
+    },
+    {
+      "epoch": 0.34886901630720674,
+      "grad_norm": 1.5050058364868164,
+      "learning_rate": 0.00015126359702004818,
+      "loss": 1.8036,
+      "step": 3316
+    },
+    {
+      "epoch": 0.34897422409258283,
+      "grad_norm": 1.342183232307434,
+      "learning_rate": 0.0001512343350085784,
+      "loss": 1.9512,
+      "step": 3317
+    },
+    {
+      "epoch": 0.349079431877959,
+      "grad_norm": 1.6918140649795532,
+      "learning_rate": 0.00015120506704743402,
+      "loss": 2.1275,
+      "step": 3318
+    },
+    {
+      "epoch": 0.3491846396633351,
+      "grad_norm": 1.7969458103179932,
+      "learning_rate": 0.00015117579314001382,
+      "loss": 1.9085,
+      "step": 3319
+    },
+    {
+      "epoch": 0.3492898474487112,
+      "grad_norm": 1.1267280578613281,
+      "learning_rate": 0.00015114651328971727,
+      "loss": 1.8183,
+      "step": 3320
+    },
+    {
+      "epoch": 0.3493950552340873,
+      "grad_norm": 1.4564090967178345,
+      "learning_rate": 0.00015111722749994457,
+      "loss": 1.8921,
+      "step": 3321
+    },
+    {
+      "epoch": 0.34950026301946346,
+      "grad_norm": 1.4575985670089722,
+      "learning_rate": 0.00015108793577409656,
+      "loss": 1.7664,
+      "step": 3322
+    },
+    {
+      "epoch": 0.34960547080483956,
+      "grad_norm": 1.3683162927627563,
+      "learning_rate": 0.0001510586381155748,
+      "loss": 1.585,
+      "step": 3323
+    },
+    {
+      "epoch": 0.34971067859021565,
+      "grad_norm": 2.316188097000122,
+      "learning_rate": 0.0001510293345277815,
+      "loss": 1.8608,
+      "step": 3324
+    },
+    {
+      "epoch": 0.3498158863755918,
+      "grad_norm": 0.8995792269706726,
+      "learning_rate": 0.0001510000250141196,
+      "loss": 1.475,
+      "step": 3325
+    },
+    {
+      "epoch": 0.3499210941609679,
+      "grad_norm": 1.4614365100860596,
+      "learning_rate": 0.0001509707095779928,
+      "loss": 2.1418,
+      "step": 3326
+    },
+    {
+      "epoch": 0.35002630194634404,
+      "grad_norm": 2.0038840770721436,
+      "learning_rate": 0.00015094138822280533,
+      "loss": 2.1766,
+      "step": 3327
+    },
+    {
+      "epoch": 0.35013150973172014,
+      "grad_norm": 2.1047608852386475,
+      "learning_rate": 0.00015091206095196215,
+      "loss": 1.7852,
+      "step": 3328
+    },
+    {
+      "epoch": 0.3502367175170963,
+      "grad_norm": 1.098832607269287,
+      "learning_rate": 0.000150882727768869,
+      "loss": 1.973,
+      "step": 3329
+    },
+    {
+      "epoch": 0.3503419253024724,
+      "grad_norm": 1.2280389070510864,
+      "learning_rate": 0.00015085338867693225,
+      "loss": 1.7759,
+      "step": 3330
+    },
+    {
+      "epoch": 0.35044713308784853,
+      "grad_norm": 1.6479296684265137,
+      "learning_rate": 0.0001508240436795589,
+      "loss": 1.9789,
+      "step": 3331
+    },
+    {
+      "epoch": 0.3505523408732246,
+      "grad_norm": 1.37543785572052,
+      "learning_rate": 0.00015079469278015672,
+      "loss": 2.2909,
+      "step": 3332
+    },
+    {
+      "epoch": 0.3506575486586007,
+      "grad_norm": 0.9722198247909546,
+      "learning_rate": 0.00015076533598213415,
+      "loss": 1.6292,
+      "step": 3333
+    },
+    {
+      "epoch": 0.35076275644397686,
+      "grad_norm": 1.5459274053573608,
+      "learning_rate": 0.00015073597328890025,
+      "loss": 1.8489,
+      "step": 3334
+    },
+    {
+      "epoch": 0.35086796422935296,
+      "grad_norm": 1.3140374422073364,
+      "learning_rate": 0.00015070660470386485,
+      "loss": 2.0101,
+      "step": 3335
+    },
+    {
+      "epoch": 0.3509731720147291,
+      "grad_norm": 1.515724778175354,
+      "learning_rate": 0.00015067723023043844,
+      "loss": 1.9427,
+      "step": 3336
+    },
+    {
+      "epoch": 0.3510783798001052,
+      "grad_norm": 1.626247763633728,
+      "learning_rate": 0.00015064784987203216,
+      "loss": 1.574,
+      "step": 3337
+    },
+    {
+      "epoch": 0.35118358758548135,
+      "grad_norm": 1.1140034198760986,
+      "learning_rate": 0.00015061846363205784,
+      "loss": 2.173,
+      "step": 3338
+    },
+    {
+      "epoch": 0.35128879537085744,
+      "grad_norm": 1.1657525300979614,
+      "learning_rate": 0.000150589071513928,
+      "loss": 2.1967,
+      "step": 3339
+    },
+    {
+      "epoch": 0.35139400315623354,
+      "grad_norm": 0.87020343542099,
+      "learning_rate": 0.00015055967352105588,
+      "loss": 1.7997,
+      "step": 3340
+    },
+    {
+      "epoch": 0.3514992109416097,
+      "grad_norm": 1.1044055223464966,
+      "learning_rate": 0.0001505302696568554,
+      "loss": 2.051,
+      "step": 3341
+    },
+    {
+      "epoch": 0.3516044187269858,
+      "grad_norm": 2.5715765953063965,
+      "learning_rate": 0.00015050085992474106,
+      "loss": 2.1982,
+      "step": 3342
+    },
+    {
+      "epoch": 0.35170962651236193,
+      "grad_norm": 1.7368378639221191,
+      "learning_rate": 0.00015047144432812814,
+      "loss": 1.2449,
+      "step": 3343
+    },
+    {
+      "epoch": 0.351814834297738,
+      "grad_norm": 1.0947346687316895,
+      "learning_rate": 0.00015044202287043263,
+      "loss": 1.4047,
+      "step": 3344
+    },
+    {
+      "epoch": 0.35192004208311417,
+      "grad_norm": 1.4075251817703247,
+      "learning_rate": 0.00015041259555507108,
+      "loss": 1.8252,
+      "step": 3345
+    },
+    {
+      "epoch": 0.35202524986849026,
+      "grad_norm": 1.3330639600753784,
+      "learning_rate": 0.00015038316238546082,
+      "loss": 1.9027,
+      "step": 3346
+    },
+    {
+      "epoch": 0.3521304576538664,
+      "grad_norm": 1.0962144136428833,
+      "learning_rate": 0.00015035372336501984,
+      "loss": 1.3782,
+      "step": 3347
+    },
+    {
+      "epoch": 0.3522356654392425,
+      "grad_norm": 1.612261414527893,
+      "learning_rate": 0.00015032427849716675,
+      "loss": 2.1975,
+      "step": 3348
+    },
+    {
+      "epoch": 0.3523408732246186,
+      "grad_norm": 1.682795763015747,
+      "learning_rate": 0.0001502948277853209,
+      "loss": 1.8072,
+      "step": 3349
+    },
+    {
+      "epoch": 0.35244608100999475,
+      "grad_norm": 1.4576448202133179,
+      "learning_rate": 0.00015026537123290234,
+      "loss": 1.7883,
+      "step": 3350
+    },
+    {
+      "epoch": 0.35255128879537084,
+      "grad_norm": 1.385787844657898,
+      "learning_rate": 0.00015023590884333173,
+      "loss": 2.0371,
+      "step": 3351
+    },
+    {
+      "epoch": 0.352656496580747,
+      "grad_norm": 1.0585753917694092,
+      "learning_rate": 0.00015020644062003046,
+      "loss": 1.912,
+      "step": 3352
+    },
+    {
+      "epoch": 0.3527617043661231,
+      "grad_norm": 1.1872546672821045,
+      "learning_rate": 0.00015017696656642056,
+      "loss": 2.1347,
+      "step": 3353
+    },
+    {
+      "epoch": 0.35286691215149923,
+      "grad_norm": 1.5553888082504272,
+      "learning_rate": 0.00015014748668592477,
+      "loss": 1.8306,
+      "step": 3354
+    },
+    {
+      "epoch": 0.35297211993687533,
+      "grad_norm": 1.696855902671814,
+      "learning_rate": 0.00015011800098196646,
+      "loss": 1.5191,
+      "step": 3355
+    },
+    {
+      "epoch": 0.3530773277222514,
+      "grad_norm": 1.31337571144104,
+      "learning_rate": 0.00015008850945796975,
+      "loss": 2.1424,
+      "step": 3356
+    },
+    {
+      "epoch": 0.35318253550762757,
+      "grad_norm": 1.327221393585205,
+      "learning_rate": 0.00015005901211735938,
+      "loss": 1.8816,
+      "step": 3357
+    },
+    {
+      "epoch": 0.35328774329300366,
+      "grad_norm": 1.1838815212249756,
+      "learning_rate": 0.0001500295089635608,
+      "loss": 2.271,
+      "step": 3358
+    },
+    {
+      "epoch": 0.3533929510783798,
+      "grad_norm": 1.5247220993041992,
+      "learning_rate": 0.00015000000000000001,
+      "loss": 1.7908,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3534981588637559,
+      "grad_norm": 1.6162418127059937,
+      "learning_rate": 0.0001499704852301039,
+      "loss": 2.1344,
+      "step": 3360
+    },
+    {
+      "epoch": 0.35360336664913206,
+      "grad_norm": 1.533591628074646,
+      "learning_rate": 0.0001499409646572999,
+      "loss": 2.2785,
+      "step": 3361
+    },
+    {
+      "epoch": 0.35370857443450815,
+      "grad_norm": 1.3864167928695679,
+      "learning_rate": 0.00014991143828501613,
+      "loss": 2.2172,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3538137822198843,
+      "grad_norm": 1.580679178237915,
+      "learning_rate": 0.00014988190611668135,
+      "loss": 1.6399,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3539189900052604,
+      "grad_norm": 1.053759217262268,
+      "learning_rate": 0.00014985236815572513,
+      "loss": 1.3566,
+      "step": 3364
+    },
+    {
+      "epoch": 0.3540241977906365,
+      "grad_norm": 1.2287757396697998,
+      "learning_rate": 0.0001498228244055775,
+      "loss": 2.0034,
+      "step": 3365
+    },
+    {
+      "epoch": 0.35412940557601263,
+      "grad_norm": 1.3084381818771362,
+      "learning_rate": 0.00014979327486966938,
+      "loss": 1.9535,
+      "step": 3366
+    },
+    {
+      "epoch": 0.35423461336138873,
+      "grad_norm": 1.1982232332229614,
+      "learning_rate": 0.0001497637195514322,
+      "loss": 2.0076,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3543398211467649,
+      "grad_norm": 1.4315484762191772,
+      "learning_rate": 0.00014973415845429813,
+      "loss": 1.9953,
+      "step": 3368
+    },
+    {
+      "epoch": 0.35444502893214097,
+      "grad_norm": 1.0541669130325317,
+      "learning_rate": 0.0001497045915817,
+      "loss": 1.6665,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3545502367175171,
+      "grad_norm": 1.2444473505020142,
+      "learning_rate": 0.00014967501893707133,
+      "loss": 2.4482,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3546554445028932,
+      "grad_norm": 1.091001033782959,
+      "learning_rate": 0.00014964544052384628,
+      "loss": 2.1292,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3547606522882693,
+      "grad_norm": 1.2684603929519653,
+      "learning_rate": 0.0001496158563454597,
+      "loss": 1.6962,
+      "step": 3372
+    },
+    {
+      "epoch": 0.35486586007364546,
+      "grad_norm": 1.5170433521270752,
+      "learning_rate": 0.0001495862664053471,
+      "loss": 1.6657,
+      "step": 3373
+    },
+    {
+      "epoch": 0.35497106785902155,
+      "grad_norm": 1.2288258075714111,
+      "learning_rate": 0.0001495566707069447,
+      "loss": 1.9427,
+      "step": 3374
+    },
+    {
+      "epoch": 0.3550762756443977,
+      "grad_norm": 1.205127239227295,
+      "learning_rate": 0.0001495270692536893,
+      "loss": 1.958,
+      "step": 3375
+    },
+    {
+      "epoch": 0.3551814834297738,
+      "grad_norm": 1.3871060609817505,
+      "learning_rate": 0.0001494974620490184,
+      "loss": 1.9658,
+      "step": 3376
+    },
+    {
+      "epoch": 0.35528669121514994,
+      "grad_norm": 1.6164618730545044,
+      "learning_rate": 0.00014946784909637028,
+      "loss": 1.354,
+      "step": 3377
+    },
+    {
+      "epoch": 0.35539189900052603,
+      "grad_norm": 1.082649827003479,
+      "learning_rate": 0.00014943823039918373,
+      "loss": 1.5321,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3554971067859022,
+      "grad_norm": 1.9146784543991089,
+      "learning_rate": 0.00014940860596089828,
+      "loss": 1.5985,
+      "step": 3379
+    },
+    {
+      "epoch": 0.3556023145712783,
+      "grad_norm": 1.5512475967407227,
+      "learning_rate": 0.0001493789757849541,
+      "loss": 1.7136,
+      "step": 3380
+    },
+    {
+      "epoch": 0.35570752235665437,
+      "grad_norm": 1.2535208463668823,
+      "learning_rate": 0.00014934933987479206,
+      "loss": 1.5282,
+      "step": 3381
+    },
+    {
+      "epoch": 0.3558127301420305,
+      "grad_norm": 1.6112395524978638,
+      "learning_rate": 0.0001493196982338537,
+      "loss": 2.1735,
+      "step": 3382
+    },
+    {
+      "epoch": 0.3559179379274066,
+      "grad_norm": 1.6096539497375488,
+      "learning_rate": 0.00014929005086558117,
+      "loss": 2.2356,
+      "step": 3383
+    },
+    {
+      "epoch": 0.35602314571278276,
+      "grad_norm": 1.7732410430908203,
+      "learning_rate": 0.00014926039777341733,
+      "loss": 1.8119,
+      "step": 3384
+    },
+    {
+      "epoch": 0.35612835349815886,
+      "grad_norm": 2.6006152629852295,
+      "learning_rate": 0.00014923073896080575,
+      "loss": 1.6363,
+      "step": 3385
+    },
+    {
+      "epoch": 0.356233561283535,
+      "grad_norm": 1.6109579801559448,
+      "learning_rate": 0.00014920107443119052,
+      "loss": 1.8525,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3563387690689111,
+      "grad_norm": 1.49036705493927,
+      "learning_rate": 0.00014917140418801655,
+      "loss": 1.6045,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3564439768542872,
+      "grad_norm": 1.5587002038955688,
+      "learning_rate": 0.00014914172823472934,
+      "loss": 2.4678,
+      "step": 3388
+    },
+    {
+      "epoch": 0.35654918463966334,
+      "grad_norm": 1.2528430223464966,
+      "learning_rate": 0.00014911204657477506,
+      "loss": 2.4837,
+      "step": 3389
+    },
+    {
+      "epoch": 0.35665439242503943,
+      "grad_norm": 1.163336157798767,
+      "learning_rate": 0.00014908235921160055,
+      "loss": 2.1102,
+      "step": 3390
+    },
+    {
+      "epoch": 0.3567596002104156,
+      "grad_norm": 1.2272528409957886,
+      "learning_rate": 0.00014905266614865324,
+      "loss": 2.1122,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3568648079957917,
+      "grad_norm": 1.4555405378341675,
+      "learning_rate": 0.00014902296738938134,
+      "loss": 1.264,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3569700157811678,
+      "grad_norm": 2.172147274017334,
+      "learning_rate": 0.00014899326293723371,
+      "loss": 1.9308,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3570752235665439,
+      "grad_norm": 1.3015145063400269,
+      "learning_rate": 0.00014896355279565976,
+      "loss": 2.0421,
+      "step": 3394
+    },
+    {
+      "epoch": 0.35718043135192007,
+      "grad_norm": 1.9815869331359863,
+      "learning_rate": 0.00014893383696810964,
+      "loss": 1.5845,
+      "step": 3395
+    },
+    {
+      "epoch": 0.35728563913729616,
+      "grad_norm": 0.8126930594444275,
+      "learning_rate": 0.0001489041154580342,
+      "loss": 2.2649,
+      "step": 3396
+    },
+    {
+      "epoch": 0.35739084692267226,
+      "grad_norm": 1.3474737405776978,
+      "learning_rate": 0.00014887438826888483,
+      "loss": 1.7113,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3574960547080484,
+      "grad_norm": 1.247473120689392,
+      "learning_rate": 0.00014884465540411368,
+      "loss": 2.136,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3576012624934245,
+      "grad_norm": 1.0129485130310059,
+      "learning_rate": 0.00014881491686717362,
+      "loss": 2.1247,
+      "step": 3399
+    },
+    {
+      "epoch": 0.35770647027880065,
+      "grad_norm": 1.0661226511001587,
+      "learning_rate": 0.00014878517266151794,
+      "loss": 1.9584,
+      "step": 3400
+    },
+    {
+      "epoch": 0.35781167806417674,
+      "grad_norm": 1.4464398622512817,
+      "learning_rate": 0.00014875542279060085,
+      "loss": 1.9324,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3579168858495529,
+      "grad_norm": 1.864518165588379,
+      "learning_rate": 0.00014872566725787701,
+      "loss": 1.1379,
+      "step": 3402
+    },
+    {
+      "epoch": 0.358022093634929,
+      "grad_norm": 1.319411277770996,
+      "learning_rate": 0.00014869590606680192,
+      "loss": 1.796,
+      "step": 3403
+    },
+    {
+      "epoch": 0.3581273014203051,
+      "grad_norm": 0.8547179698944092,
+      "learning_rate": 0.0001486661392208316,
+      "loss": 1.9455,
+      "step": 3404
+    },
+    {
+      "epoch": 0.3582325092056812,
+      "grad_norm": 1.2236049175262451,
+      "learning_rate": 0.00014863636672342277,
+      "loss": 1.4572,
+      "step": 3405
+    },
+    {
+      "epoch": 0.3583377169910573,
+      "grad_norm": 1.5455988645553589,
+      "learning_rate": 0.00014860658857803285,
+      "loss": 1.9523,
+      "step": 3406
+    },
+    {
+      "epoch": 0.35844292477643347,
+      "grad_norm": 1.231605052947998,
+      "learning_rate": 0.00014857680478811984,
+      "loss": 2.3316,
+      "step": 3407
+    },
+    {
+      "epoch": 0.35854813256180956,
+      "grad_norm": 1.5083273649215698,
+      "learning_rate": 0.00014854701535714244,
+      "loss": 2.0371,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3586533403471857,
+      "grad_norm": 1.7702112197875977,
+      "learning_rate": 0.00014851722028856005,
+      "loss": 2.159,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3587585481325618,
+      "grad_norm": 1.20704984664917,
+      "learning_rate": 0.0001484874195858326,
+      "loss": 1.1983,
+      "step": 3410
+    },
+    {
+      "epoch": 0.35886375591793795,
+      "grad_norm": 1.5127192735671997,
+      "learning_rate": 0.00014845761325242077,
+      "loss": 1.638,
+      "step": 3411
+    },
+    {
+      "epoch": 0.35896896370331405,
+      "grad_norm": 1.1735661029815674,
+      "learning_rate": 0.0001484278012917859,
+      "loss": 1.6514,
+      "step": 3412
+    },
+    {
+      "epoch": 0.35907417148869014,
+      "grad_norm": 1.142444133758545,
+      "learning_rate": 0.00014839798370738994,
+      "loss": 1.6814,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3591793792740663,
+      "grad_norm": 2.166907787322998,
+      "learning_rate": 0.00014836816050269548,
+      "loss": 2.0189,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3592845870594424,
+      "grad_norm": 0.9603028893470764,
+      "learning_rate": 0.00014833833168116582,
+      "loss": 2.2928,
+      "step": 3415
+    },
+    {
+      "epoch": 0.35938979484481853,
+      "grad_norm": 1.3047958612442017,
+      "learning_rate": 0.00014830849724626488,
+      "loss": 2.1418,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3594950026301946,
+      "grad_norm": 2.171341896057129,
+      "learning_rate": 0.00014827865720145724,
+      "loss": 1.9426,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3596002104155708,
+      "grad_norm": 1.029051661491394,
+      "learning_rate": 0.0001482488115502081,
+      "loss": 2.0982,
+      "step": 3418
+    },
+    {
+      "epoch": 0.35970541820094687,
+      "grad_norm": 1.3975342512130737,
+      "learning_rate": 0.00014821896029598337,
+      "loss": 2.189,
+      "step": 3419
+    },
+    {
+      "epoch": 0.35981062598632296,
+      "grad_norm": 1.4755433797836304,
+      "learning_rate": 0.00014818910344224957,
+      "loss": 1.6457,
+      "step": 3420
+    },
+    {
+      "epoch": 0.3599158337716991,
+      "grad_norm": 1.5850578546524048,
+      "learning_rate": 0.00014815924099247384,
+      "loss": 2.2274,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3600210415570752,
+      "grad_norm": 1.3168679475784302,
+      "learning_rate": 0.00014812937295012406,
+      "loss": 2.4136,
+      "step": 3422
+    },
+    {
+      "epoch": 0.36012624934245135,
+      "grad_norm": 1.5056153535842896,
+      "learning_rate": 0.00014809949931866867,
+      "loss": 1.8313,
+      "step": 3423
+    },
+    {
+      "epoch": 0.36023145712782745,
+      "grad_norm": 3.0403895378112793,
+      "learning_rate": 0.00014806962010157683,
+      "loss": 2.1145,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3603366649132036,
+      "grad_norm": 0.8465742468833923,
+      "learning_rate": 0.00014803973530231828,
+      "loss": 2.0201,
+      "step": 3425
+    },
+    {
+      "epoch": 0.3604418726985797,
+      "grad_norm": 1.618240237236023,
+      "learning_rate": 0.00014800984492436346,
+      "loss": 2.0447,
+      "step": 3426
+    },
+    {
+      "epoch": 0.36054708048395584,
+      "grad_norm": 1.5372968912124634,
+      "learning_rate": 0.00014797994897118347,
+      "loss": 1.9388,
+      "step": 3427
+    },
+    {
+      "epoch": 0.36065228826933193,
+      "grad_norm": 1.471268892288208,
+      "learning_rate": 0.00014795004744625,
+      "loss": 2.3546,
+      "step": 3428
+    },
+    {
+      "epoch": 0.360757496054708,
+      "grad_norm": 1.4779199361801147,
+      "learning_rate": 0.00014792014035303535,
+      "loss": 1.9209,
+      "step": 3429
+    },
+    {
+      "epoch": 0.3608627038400842,
+      "grad_norm": 0.9379653930664062,
+      "learning_rate": 0.0001478902276950127,
+      "loss": 1.705,
+      "step": 3430
+    },
+    {
+      "epoch": 0.36096791162546027,
+      "grad_norm": 1.2077665328979492,
+      "learning_rate": 0.00014786030947565554,
+      "loss": 1.7868,
+      "step": 3431
+    },
+    {
+      "epoch": 0.3610731194108364,
+      "grad_norm": 1.1456810235977173,
+      "learning_rate": 0.00014783038569843822,
+      "loss": 2.3369,
+      "step": 3432
+    },
+    {
+      "epoch": 0.3611783271962125,
+      "grad_norm": 1.7450131177902222,
+      "learning_rate": 0.00014780045636683578,
+      "loss": 1.9921,
+      "step": 3433
+    },
+    {
+      "epoch": 0.36128353498158866,
+      "grad_norm": 1.1912232637405396,
+      "learning_rate": 0.00014777052148432372,
+      "loss": 1.9747,
+      "step": 3434
+    },
+    {
+      "epoch": 0.36138874276696475,
+      "grad_norm": 1.3234916925430298,
+      "learning_rate": 0.00014774058105437827,
+      "loss": 1.6403,
+      "step": 3435
+    },
+    {
+      "epoch": 0.36149395055234085,
+      "grad_norm": 1.4043394327163696,
+      "learning_rate": 0.00014771063508047636,
+      "loss": 1.8832,
+      "step": 3436
+    },
+    {
+      "epoch": 0.361599158337717,
+      "grad_norm": 1.1788146495819092,
+      "learning_rate": 0.00014768068356609554,
+      "loss": 1.997,
+      "step": 3437
+    },
+    {
+      "epoch": 0.3617043661230931,
+      "grad_norm": 1.0332565307617188,
+      "learning_rate": 0.00014765072651471393,
+      "loss": 1.7426,
+      "step": 3438
+    },
+    {
+      "epoch": 0.36180957390846924,
+      "grad_norm": 1.34288489818573,
+      "learning_rate": 0.00014762076392981033,
+      "loss": 2.0445,
+      "step": 3439
+    },
+    {
+      "epoch": 0.36191478169384533,
+      "grad_norm": 1.6269721984863281,
+      "learning_rate": 0.00014759079581486424,
+      "loss": 1.7493,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3620199894792215,
+      "grad_norm": 1.204362392425537,
+      "learning_rate": 0.00014756082217335577,
+      "loss": 2.179,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3621251972645976,
+      "grad_norm": 1.8871209621429443,
+      "learning_rate": 0.0001475308430087656,
+      "loss": 1.5849,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3622304050499737,
+      "grad_norm": 1.0620596408843994,
+      "learning_rate": 0.00014750085832457519,
+      "loss": 2.2185,
+      "step": 3443
+    },
+    {
+      "epoch": 0.3623356128353498,
+      "grad_norm": 1.24690842628479,
+      "learning_rate": 0.00014747086812426648,
+      "loss": 1.5221,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3624408206207259,
+      "grad_norm": 1.5834826231002808,
+      "learning_rate": 0.0001474408724113222,
+      "loss": 1.4759,
+      "step": 3445
+    },
+    {
+      "epoch": 0.36254602840610206,
+      "grad_norm": 1.1942899227142334,
+      "learning_rate": 0.0001474108711892256,
+      "loss": 2.2882,
+      "step": 3446
+    },
+    {
+      "epoch": 0.36265123619147815,
+      "grad_norm": 1.3137141466140747,
+      "learning_rate": 0.00014738086446146065,
+      "loss": 1.6804,
+      "step": 3447
+    },
+    {
+      "epoch": 0.3627564439768543,
+      "grad_norm": 1.0239195823669434,
+      "learning_rate": 0.00014735085223151198,
+      "loss": 2.2197,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3628616517622304,
+      "grad_norm": 1.18763267993927,
+      "learning_rate": 0.00014732083450286472,
+      "loss": 1.7646,
+      "step": 3449
+    },
+    {
+      "epoch": 0.36296685954760655,
+      "grad_norm": 1.4239435195922852,
+      "learning_rate": 0.00014729081127900476,
+      "loss": 1.8619,
+      "step": 3450
+    },
+    {
+      "epoch": 0.36307206733298264,
+      "grad_norm": 1.079707384109497,
+      "learning_rate": 0.00014726078256341863,
+      "loss": 2.0,
+      "step": 3451
+    },
+    {
+      "epoch": 0.36317727511835873,
+      "grad_norm": 1.7194312810897827,
+      "learning_rate": 0.00014723074835959346,
+      "loss": 1.6595,
+      "step": 3452
+    },
+    {
+      "epoch": 0.3632824829037349,
+      "grad_norm": 1.498183250427246,
+      "learning_rate": 0.000147200708671017,
+      "loss": 1.6809,
+      "step": 3453
+    },
+    {
+      "epoch": 0.363387690689111,
+      "grad_norm": 1.7695226669311523,
+      "learning_rate": 0.00014717066350117768,
+      "loss": 1.6466,
+      "step": 3454
+    },
+    {
+      "epoch": 0.3634928984744871,
+      "grad_norm": 1.1211014986038208,
+      "learning_rate": 0.00014714061285356453,
+      "loss": 1.772,
+      "step": 3455
+    },
+    {
+      "epoch": 0.3635981062598632,
+      "grad_norm": 1.286805510520935,
+      "learning_rate": 0.00014711055673166724,
+      "loss": 2.2176,
+      "step": 3456
+    },
+    {
+      "epoch": 0.36370331404523937,
+      "grad_norm": 2.3629977703094482,
+      "learning_rate": 0.0001470804951389761,
+      "loss": 1.4311,
+      "step": 3457
+    },
+    {
+      "epoch": 0.36380852183061546,
+      "grad_norm": 1.3376028537750244,
+      "learning_rate": 0.00014705042807898214,
+      "loss": 1.7872,
+      "step": 3458
+    },
+    {
+      "epoch": 0.3639137296159916,
+      "grad_norm": 2.238438129425049,
+      "learning_rate": 0.0001470203555551769,
+      "loss": 1.8577,
+      "step": 3459
+    },
+    {
+      "epoch": 0.3640189374013677,
+      "grad_norm": 1.5337350368499756,
+      "learning_rate": 0.00014699027757105254,
+      "loss": 1.7483,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3641241451867438,
+      "grad_norm": 2.3909807205200195,
+      "learning_rate": 0.00014696019413010204,
+      "loss": 2.0391,
+      "step": 3461
+    },
+    {
+      "epoch": 0.36422935297211995,
+      "grad_norm": 1.3781287670135498,
+      "learning_rate": 0.00014693010523581882,
+      "loss": 1.8577,
+      "step": 3462
+    },
+    {
+      "epoch": 0.36433456075749604,
+      "grad_norm": 1.3029046058654785,
+      "learning_rate": 0.00014690001089169702,
+      "loss": 1.7383,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3644397685428722,
+      "grad_norm": 1.002429485321045,
+      "learning_rate": 0.00014686991110123135,
+      "loss": 1.7997,
+      "step": 3464
+    },
+    {
+      "epoch": 0.3645449763282483,
+      "grad_norm": 1.2660531997680664,
+      "learning_rate": 0.0001468398058679173,
+      "loss": 1.7173,
+      "step": 3465
+    },
+    {
+      "epoch": 0.36465018411362443,
+      "grad_norm": 1.3863353729248047,
+      "learning_rate": 0.0001468096951952508,
+      "loss": 1.7758,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3647553918990005,
+      "grad_norm": 1.7203501462936401,
+      "learning_rate": 0.00014677957908672856,
+      "loss": 1.7279,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3648605996843766,
+      "grad_norm": 1.1552826166152954,
+      "learning_rate": 0.0001467494575458478,
+      "loss": 1.8039,
+      "step": 3468
+    },
+    {
+      "epoch": 0.36496580746975277,
+      "grad_norm": 0.9511840343475342,
+      "learning_rate": 0.00014671933057610654,
+      "loss": 1.8241,
+      "step": 3469
+    },
+    {
+      "epoch": 0.36507101525512886,
+      "grad_norm": 1.4740869998931885,
+      "learning_rate": 0.00014668919818100322,
+      "loss": 2.1223,
+      "step": 3470
+    },
+    {
+      "epoch": 0.365176223040505,
+      "grad_norm": 2.0013620853424072,
+      "learning_rate": 0.00014665906036403706,
+      "loss": 2.3534,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3652814308258811,
+      "grad_norm": 1.2783856391906738,
+      "learning_rate": 0.0001466289171287079,
+      "loss": 1.7754,
+      "step": 3472
+    },
+    {
+      "epoch": 0.36538663861125725,
+      "grad_norm": 1.5455641746520996,
+      "learning_rate": 0.00014659876847851607,
+      "loss": 2.1852,
+      "step": 3473
+    },
+    {
+      "epoch": 0.36549184639663335,
+      "grad_norm": 1.3689314126968384,
+      "learning_rate": 0.00014656861441696278,
+      "loss": 1.7998,
+      "step": 3474
+    },
+    {
+      "epoch": 0.3655970541820095,
+      "grad_norm": 1.5268868207931519,
+      "learning_rate": 0.00014653845494754962,
+      "loss": 1.4686,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3657022619673856,
+      "grad_norm": 1.6166832447052002,
+      "learning_rate": 0.00014650829007377894,
+      "loss": 1.9012,
+      "step": 3476
+    },
+    {
+      "epoch": 0.3658074697527617,
+      "grad_norm": 1.4161622524261475,
+      "learning_rate": 0.00014647811979915366,
+      "loss": 2.1802,
+      "step": 3477
+    },
+    {
+      "epoch": 0.36591267753813783,
+      "grad_norm": 1.2434589862823486,
+      "learning_rate": 0.00014644794412717736,
+      "loss": 1.9538,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3660178853235139,
+      "grad_norm": 1.0335787534713745,
+      "learning_rate": 0.00014641776306135431,
+      "loss": 1.8217,
+      "step": 3479
+    },
+    {
+      "epoch": 0.3661230931088901,
+      "grad_norm": 1.653498649597168,
+      "learning_rate": 0.00014638757660518923,
+      "loss": 1.8545,
+      "step": 3480
+    },
+    {
+      "epoch": 0.36622830089426617,
+      "grad_norm": 0.8305730223655701,
+      "learning_rate": 0.00014635738476218767,
+      "loss": 1.4702,
+      "step": 3481
+    },
+    {
+      "epoch": 0.3663335086796423,
+      "grad_norm": 1.3280541896820068,
+      "learning_rate": 0.00014632718753585566,
+      "loss": 1.7956,
+      "step": 3482
+    },
+    {
+      "epoch": 0.3664387164650184,
+      "grad_norm": 1.6454240083694458,
+      "learning_rate": 0.0001462969849296999,
+      "loss": 1.5739,
+      "step": 3483
+    },
+    {
+      "epoch": 0.3665439242503945,
+      "grad_norm": 1.0778430700302124,
+      "learning_rate": 0.00014626677694722773,
+      "loss": 2.2551,
+      "step": 3484
+    },
+    {
+      "epoch": 0.36664913203577065,
+      "grad_norm": 1.3582223653793335,
+      "learning_rate": 0.00014623656359194712,
+      "loss": 1.9954,
+      "step": 3485
+    },
+    {
+      "epoch": 0.36675433982114675,
+      "grad_norm": 1.7454923391342163,
+      "learning_rate": 0.00014620634486736667,
+      "loss": 2.0445,
+      "step": 3486
+    },
+    {
+      "epoch": 0.3668595476065229,
+      "grad_norm": 1.1220197677612305,
+      "learning_rate": 0.00014617612077699548,
+      "loss": 1.8598,
+      "step": 3487
+    },
+    {
+      "epoch": 0.366964755391899,
+      "grad_norm": 1.2056738138198853,
+      "learning_rate": 0.00014614589132434347,
+      "loss": 2.0579,
+      "step": 3488
+    },
+    {
+      "epoch": 0.36706996317727514,
+      "grad_norm": 1.1674832105636597,
+      "learning_rate": 0.00014611565651292106,
+      "loss": 1.7296,
+      "step": 3489
+    },
+    {
+      "epoch": 0.36717517096265123,
+      "grad_norm": 1.2299489974975586,
+      "learning_rate": 0.00014608541634623929,
+      "loss": 1.6829,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3672803787480274,
+      "grad_norm": 1.580704689025879,
+      "learning_rate": 0.00014605517082780988,
+      "loss": 2.2437,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3673855865334035,
+      "grad_norm": 1.4246400594711304,
+      "learning_rate": 0.00014602491996114516,
+      "loss": 1.8395,
+      "step": 3492
+    },
+    {
+      "epoch": 0.36749079431877957,
+      "grad_norm": 1.4935017824172974,
+      "learning_rate": 0.00014599466374975802,
+      "loss": 2.118,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3675960021041557,
+      "grad_norm": 1.4232902526855469,
+      "learning_rate": 0.00014596440219716205,
+      "loss": 1.8154,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3677012098895318,
+      "grad_norm": 1.707642912864685,
+      "learning_rate": 0.00014593413530687138,
+      "loss": 1.7055,
+      "step": 3495
+    },
+    {
+      "epoch": 0.36780641767490796,
+      "grad_norm": 1.4270106554031372,
+      "learning_rate": 0.0001459038630824009,
+      "loss": 2.0895,
+      "step": 3496
+    },
+    {
+      "epoch": 0.36791162546028405,
+      "grad_norm": 1.169714331626892,
+      "learning_rate": 0.00014587358552726592,
+      "loss": 1.8576,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3680168332456602,
+      "grad_norm": 1.1575000286102295,
+      "learning_rate": 0.0001458433026449825,
+      "loss": 1.8767,
+      "step": 3498
+    },
+    {
+      "epoch": 0.3681220410310363,
+      "grad_norm": 1.5656603574752808,
+      "learning_rate": 0.0001458130144390673,
+      "loss": 1.8864,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3682272488164124,
+      "grad_norm": 0.9591088891029358,
+      "learning_rate": 0.0001457827209130376,
+      "loss": 1.8874,
+      "step": 3500
+    },
+    {
+      "epoch": 0.36833245660178854,
+      "grad_norm": 1.8865973949432373,
+      "learning_rate": 0.00014575242207041128,
+      "loss": 1.846,
+      "step": 3501
+    },
+    {
+      "epoch": 0.36843766438716463,
+      "grad_norm": 1.3061480522155762,
+      "learning_rate": 0.00014572211791470685,
+      "loss": 2.1671,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3685428721725408,
+      "grad_norm": 1.3023380041122437,
+      "learning_rate": 0.00014569180844944344,
+      "loss": 1.8949,
+      "step": 3503
+    },
+    {
+      "epoch": 0.3686480799579169,
+      "grad_norm": 1.3070828914642334,
+      "learning_rate": 0.00014566149367814074,
+      "loss": 1.9491,
+      "step": 3504
+    },
+    {
+      "epoch": 0.368753287743293,
+      "grad_norm": 1.1962894201278687,
+      "learning_rate": 0.00014563117360431914,
+      "loss": 1.5836,
+      "step": 3505
+    },
+    {
+      "epoch": 0.3688584955286691,
+      "grad_norm": 1.190346121788025,
+      "learning_rate": 0.00014560084823149965,
+      "loss": 2.0028,
+      "step": 3506
+    },
+    {
+      "epoch": 0.36896370331404527,
+      "grad_norm": 1.0240740776062012,
+      "learning_rate": 0.00014557051756320378,
+      "loss": 1.9108,
+      "step": 3507
+    },
+    {
+      "epoch": 0.36906891109942136,
+      "grad_norm": 1.517669916152954,
+      "learning_rate": 0.0001455401816029538,
+      "loss": 1.7805,
+      "step": 3508
+    },
+    {
+      "epoch": 0.36917411888479745,
+      "grad_norm": 1.1098333597183228,
+      "learning_rate": 0.00014550984035427243,
+      "loss": 1.9193,
+      "step": 3509
+    },
+    {
+      "epoch": 0.3692793266701736,
+      "grad_norm": 1.9058518409729004,
+      "learning_rate": 0.00014547949382068322,
+      "loss": 2.471,
+      "step": 3510
+    },
+    {
+      "epoch": 0.3693845344555497,
+      "grad_norm": 0.9477731585502625,
+      "learning_rate": 0.0001454491420057101,
+      "loss": 1.2372,
+      "step": 3511
+    },
+    {
+      "epoch": 0.36948974224092584,
+      "grad_norm": 2.694037437438965,
+      "learning_rate": 0.00014541878491287783,
+      "loss": 2.3677,
+      "step": 3512
+    },
+    {
+      "epoch": 0.36959495002630194,
+      "grad_norm": 0.9963229298591614,
+      "learning_rate": 0.0001453884225457116,
+      "loss": 1.8101,
+      "step": 3513
+    },
+    {
+      "epoch": 0.3697001578116781,
+      "grad_norm": 1.3422759771347046,
+      "learning_rate": 0.00014535805490773732,
+      "loss": 1.9981,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3698053655970542,
+      "grad_norm": 0.9793868660926819,
+      "learning_rate": 0.0001453276820024815,
+      "loss": 2.0641,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3699105733824303,
+      "grad_norm": 1.3728710412979126,
+      "learning_rate": 0.0001452973038334712,
+      "loss": 2.2349,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3700157811678064,
+      "grad_norm": 1.3043874502182007,
+      "learning_rate": 0.0001452669204042342,
+      "loss": 2.091,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3701209889531825,
+      "grad_norm": 1.4674620628356934,
+      "learning_rate": 0.0001452365317182988,
+      "loss": 1.8651,
+      "step": 3518
+    },
+    {
+      "epoch": 0.37022619673855867,
+      "grad_norm": 1.3150261640548706,
+      "learning_rate": 0.00014520613777919392,
+      "loss": 1.8597,
+      "step": 3519
+    },
+    {
+      "epoch": 0.37033140452393476,
+      "grad_norm": 1.2791160345077515,
+      "learning_rate": 0.00014517573859044907,
+      "loss": 1.9082,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3704366123093109,
+      "grad_norm": 1.371948480606079,
+      "learning_rate": 0.00014514533415559453,
+      "loss": 1.6619,
+      "step": 3521
+    },
+    {
+      "epoch": 0.370541820094687,
+      "grad_norm": 1.731662392616272,
+      "learning_rate": 0.00014511492447816097,
+      "loss": 2.1011,
+      "step": 3522
+    },
+    {
+      "epoch": 0.37064702788006315,
+      "grad_norm": 1.2484428882598877,
+      "learning_rate": 0.0001450845095616798,
+      "loss": 1.4152,
+      "step": 3523
+    },
+    {
+      "epoch": 0.37075223566543924,
+      "grad_norm": 1.5884274244308472,
+      "learning_rate": 0.00014505408940968296,
+      "loss": 2.0706,
+      "step": 3524
+    },
+    {
+      "epoch": 0.37085744345081534,
+      "grad_norm": 1.000995397567749,
+      "learning_rate": 0.00014502366402570309,
+      "loss": 2.1573,
+      "step": 3525
+    },
+    {
+      "epoch": 0.3709626512361915,
+      "grad_norm": 0.8821168541908264,
+      "learning_rate": 0.00014499323341327338,
+      "loss": 1.902,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3710678590215676,
+      "grad_norm": 1.4428273439407349,
+      "learning_rate": 0.00014496279757592766,
+      "loss": 2.1331,
+      "step": 3527
+    },
+    {
+      "epoch": 0.37117306680694373,
+      "grad_norm": 1.5744878053665161,
+      "learning_rate": 0.00014493235651720027,
+      "loss": 1.7205,
+      "step": 3528
+    },
+    {
+      "epoch": 0.3712782745923198,
+      "grad_norm": 1.5465056896209717,
+      "learning_rate": 0.00014490191024062632,
+      "loss": 1.5546,
+      "step": 3529
+    },
+    {
+      "epoch": 0.371383482377696,
+      "grad_norm": 1.1197417974472046,
+      "learning_rate": 0.00014487145874974135,
+      "loss": 1.3374,
+      "step": 3530
+    },
+    {
+      "epoch": 0.37148869016307207,
+      "grad_norm": 0.9378413558006287,
+      "learning_rate": 0.00014484100204808167,
+      "loss": 2.3889,
+      "step": 3531
+    },
+    {
+      "epoch": 0.37159389794844816,
+      "grad_norm": 1.0312168598175049,
+      "learning_rate": 0.00014481054013918408,
+      "loss": 2.095,
+      "step": 3532
+    },
+    {
+      "epoch": 0.3716991057338243,
+      "grad_norm": 1.6550025939941406,
+      "learning_rate": 0.00014478007302658598,
+      "loss": 1.6941,
+      "step": 3533
+    },
+    {
+      "epoch": 0.3718043135192004,
+      "grad_norm": 1.5724120140075684,
+      "learning_rate": 0.0001447496007138255,
+      "loss": 2.3644,
+      "step": 3534
+    },
+    {
+      "epoch": 0.37190952130457655,
+      "grad_norm": 1.3543592691421509,
+      "learning_rate": 0.00014471912320444122,
+      "loss": 2.1931,
+      "step": 3535
+    },
+    {
+      "epoch": 0.37201472908995264,
+      "grad_norm": 1.7770755290985107,
+      "learning_rate": 0.00014468864050197242,
+      "loss": 1.5458,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3721199368753288,
+      "grad_norm": 1.3519712686538696,
+      "learning_rate": 0.00014465815260995894,
+      "loss": 2.119,
+      "step": 3537
+    },
+    {
+      "epoch": 0.3722251446607049,
+      "grad_norm": 2.1993541717529297,
+      "learning_rate": 0.0001446276595319413,
+      "loss": 1.3163,
+      "step": 3538
+    },
+    {
+      "epoch": 0.37233035244608104,
+      "grad_norm": 1.6572602987289429,
+      "learning_rate": 0.00014459716127146049,
+      "loss": 1.7625,
+      "step": 3539
+    },
+    {
+      "epoch": 0.37243556023145713,
+      "grad_norm": 1.320532202720642,
+      "learning_rate": 0.0001445666578320582,
+      "loss": 1.377,
+      "step": 3540
+    },
+    {
+      "epoch": 0.3725407680168332,
+      "grad_norm": 1.6189240217208862,
+      "learning_rate": 0.00014453614921727668,
+      "loss": 1.7391,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3726459758022094,
+      "grad_norm": 1.0008089542388916,
+      "learning_rate": 0.00014450563543065881,
+      "loss": 1.8331,
+      "step": 3542
+    },
+    {
+      "epoch": 0.37275118358758547,
+      "grad_norm": 1.0552862882614136,
+      "learning_rate": 0.00014447511647574805,
+      "loss": 2.0291,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3728563913729616,
+      "grad_norm": 1.1746324300765991,
+      "learning_rate": 0.00014444459235608847,
+      "loss": 2.281,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3729615991583377,
+      "grad_norm": 1.2236127853393555,
+      "learning_rate": 0.00014441406307522475,
+      "loss": 1.8409,
+      "step": 3545
+    },
+    {
+      "epoch": 0.37306680694371386,
+      "grad_norm": 1.948947548866272,
+      "learning_rate": 0.0001443835286367021,
+      "loss": 1.3529,
+      "step": 3546
+    },
+    {
+      "epoch": 0.37317201472908995,
+      "grad_norm": 1.3793987035751343,
+      "learning_rate": 0.00014435298904406642,
+      "loss": 1.852,
+      "step": 3547
+    },
+    {
+      "epoch": 0.37327722251446604,
+      "grad_norm": 0.9412884712219238,
+      "learning_rate": 0.00014432244430086423,
+      "loss": 2.2095,
+      "step": 3548
+    },
+    {
+      "epoch": 0.3733824302998422,
+      "grad_norm": 1.2927563190460205,
+      "learning_rate": 0.00014429189441064248,
+      "loss": 1.9944,
+      "step": 3549
+    },
+    {
+      "epoch": 0.3734876380852183,
+      "grad_norm": 1.234215497970581,
+      "learning_rate": 0.00014426133937694887,
+      "loss": 1.7411,
+      "step": 3550
+    },
+    {
+      "epoch": 0.37359284587059444,
+      "grad_norm": 1.1995365619659424,
+      "learning_rate": 0.00014423077920333173,
+      "loss": 2.0875,
+      "step": 3551
+    },
+    {
+      "epoch": 0.37369805365597053,
+      "grad_norm": 1.4042108058929443,
+      "learning_rate": 0.00014420021389333982,
+      "loss": 2.1245,
+      "step": 3552
+    },
+    {
+      "epoch": 0.3738032614413467,
+      "grad_norm": 1.579077959060669,
+      "learning_rate": 0.0001441696434505226,
+      "loss": 1.2663,
+      "step": 3553
+    },
+    {
+      "epoch": 0.3739084692267228,
+      "grad_norm": 1.5505826473236084,
+      "learning_rate": 0.00014413906787843014,
+      "loss": 1.8943,
+      "step": 3554
+    },
+    {
+      "epoch": 0.3740136770120989,
+      "grad_norm": 1.1203827857971191,
+      "learning_rate": 0.00014410848718061312,
+      "loss": 1.9703,
+      "step": 3555
+    },
+    {
+      "epoch": 0.374118884797475,
+      "grad_norm": 1.4529035091400146,
+      "learning_rate": 0.00014407790136062267,
+      "loss": 2.0986,
+      "step": 3556
+    },
+    {
+      "epoch": 0.3742240925828511,
+      "grad_norm": 2.0236105918884277,
+      "learning_rate": 0.0001440473104220107,
+      "loss": 1.363,
+      "step": 3557
+    },
+    {
+      "epoch": 0.37432930036822726,
+      "grad_norm": 1.0123807191848755,
+      "learning_rate": 0.0001440167143683296,
+      "loss": 1.7753,
+      "step": 3558
+    },
+    {
+      "epoch": 0.37443450815360335,
+      "grad_norm": 0.9799262285232544,
+      "learning_rate": 0.00014398611320313244,
+      "loss": 1.4174,
+      "step": 3559
+    },
+    {
+      "epoch": 0.3745397159389795,
+      "grad_norm": 1.738737940788269,
+      "learning_rate": 0.00014395550692997277,
+      "loss": 1.9,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3746449237243556,
+      "grad_norm": 1.3359789848327637,
+      "learning_rate": 0.00014392489555240486,
+      "loss": 1.7223,
+      "step": 3561
+    },
+    {
+      "epoch": 0.37475013150973174,
+      "grad_norm": 2.408348798751831,
+      "learning_rate": 0.00014389427907398342,
+      "loss": 1.6607,
+      "step": 3562
+    },
+    {
+      "epoch": 0.37485533929510784,
+      "grad_norm": 2.228571891784668,
+      "learning_rate": 0.0001438636574982639,
+      "loss": 1.2856,
+      "step": 3563
+    },
+    {
+      "epoch": 0.37496054708048393,
+      "grad_norm": 1.1137131452560425,
+      "learning_rate": 0.0001438330308288023,
+      "loss": 2.1641,
+      "step": 3564
+    },
+    {
+      "epoch": 0.3750657548658601,
+      "grad_norm": 1.3347491025924683,
+      "learning_rate": 0.00014380239906915514,
+      "loss": 2.0624,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3751709626512362,
+      "grad_norm": 1.4927902221679688,
+      "learning_rate": 0.00014377176222287965,
+      "loss": 1.6468,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3752761704366123,
+      "grad_norm": 1.1425906419754028,
+      "learning_rate": 0.0001437411202935335,
+      "loss": 2.0824,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3753813782219884,
+      "grad_norm": 1.4617263078689575,
+      "learning_rate": 0.00014371047328467511,
+      "loss": 2.1863,
+      "step": 3568
+    },
+    {
+      "epoch": 0.37548658600736456,
+      "grad_norm": 1.414243459701538,
+      "learning_rate": 0.00014367982119986342,
+      "loss": 1.8408,
+      "step": 3569
+    },
+    {
+      "epoch": 0.37559179379274066,
+      "grad_norm": 1.367104172706604,
+      "learning_rate": 0.00014364916404265788,
+      "loss": 1.6474,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3756970015781168,
+      "grad_norm": 2.073486089706421,
+      "learning_rate": 0.0001436185018166187,
+      "loss": 1.8869,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3758022093634929,
+      "grad_norm": 0.7273818850517273,
+      "learning_rate": 0.0001435878345253065,
+      "loss": 1.4154,
+      "step": 3572
+    },
+    {
+      "epoch": 0.375907417148869,
+      "grad_norm": 1.1615116596221924,
+      "learning_rate": 0.00014355716217228265,
+      "loss": 2.0725,
+      "step": 3573
+    },
+    {
+      "epoch": 0.37601262493424514,
+      "grad_norm": 1.9539501667022705,
+      "learning_rate": 0.00014352648476110896,
+      "loss": 1.8466,
+      "step": 3574
+    },
+    {
+      "epoch": 0.37611783271962124,
+      "grad_norm": 1.2255717515945435,
+      "learning_rate": 0.0001434958022953479,
+      "loss": 2.1784,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3762230405049974,
+      "grad_norm": 1.502297282218933,
+      "learning_rate": 0.00014346511477856259,
+      "loss": 1.8617,
+      "step": 3576
+    },
+    {
+      "epoch": 0.3763282482903735,
+      "grad_norm": 1.4873766899108887,
+      "learning_rate": 0.0001434344222143166,
+      "loss": 2.0098,
+      "step": 3577
+    },
+    {
+      "epoch": 0.37643345607574963,
+      "grad_norm": 1.7163630723953247,
+      "learning_rate": 0.0001434037246061742,
+      "loss": 2.5738,
+      "step": 3578
+    },
+    {
+      "epoch": 0.3765386638611257,
+      "grad_norm": 1.179367184638977,
+      "learning_rate": 0.0001433730219577002,
+      "loss": 1.5599,
+      "step": 3579
+    },
+    {
+      "epoch": 0.3766438716465018,
+      "grad_norm": 1.8585902452468872,
+      "learning_rate": 0.00014334231427245994,
+      "loss": 1.9643,
+      "step": 3580
+    },
+    {
+      "epoch": 0.37674907943187796,
+      "grad_norm": 1.168847918510437,
+      "learning_rate": 0.00014331160155401948,
+      "loss": 2.2399,
+      "step": 3581
+    },
+    {
+      "epoch": 0.37685428721725406,
+      "grad_norm": 1.2563929557800293,
+      "learning_rate": 0.00014328088380594534,
+      "loss": 2.0344,
+      "step": 3582
+    },
+    {
+      "epoch": 0.3769594950026302,
+      "grad_norm": 1.3818440437316895,
+      "learning_rate": 0.0001432501610318047,
+      "loss": 1.6415,
+      "step": 3583
+    },
+    {
+      "epoch": 0.3770647027880063,
+      "grad_norm": 1.4234402179718018,
+      "learning_rate": 0.00014321943323516526,
+      "loss": 1.7941,
+      "step": 3584
+    },
+    {
+      "epoch": 0.37716991057338245,
+      "grad_norm": 1.5692628622055054,
+      "learning_rate": 0.00014318870041959538,
+      "loss": 1.653,
+      "step": 3585
+    },
+    {
+      "epoch": 0.37727511835875854,
+      "grad_norm": 1.3026686906814575,
+      "learning_rate": 0.00014315796258866393,
+      "loss": 2.1109,
+      "step": 3586
+    },
+    {
+      "epoch": 0.3773803261441347,
+      "grad_norm": 1.237284541130066,
+      "learning_rate": 0.00014312721974594038,
+      "loss": 1.7561,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3774855339295108,
+      "grad_norm": 1.4320054054260254,
+      "learning_rate": 0.00014309647189499481,
+      "loss": 2.1673,
+      "step": 3588
+    },
+    {
+      "epoch": 0.3775907417148869,
+      "grad_norm": 1.15114164352417,
+      "learning_rate": 0.0001430657190393979,
+      "loss": 1.7654,
+      "step": 3589
+    },
+    {
+      "epoch": 0.37769594950026303,
+      "grad_norm": 1.3742176294326782,
+      "learning_rate": 0.00014303496118272084,
+      "loss": 1.8697,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3778011572856391,
+      "grad_norm": 1.233818531036377,
+      "learning_rate": 0.00014300419832853544,
+      "loss": 1.5884,
+      "step": 3591
+    },
+    {
+      "epoch": 0.37790636507101527,
+      "grad_norm": 1.7550405263900757,
+      "learning_rate": 0.0001429734304804141,
+      "loss": 1.99,
+      "step": 3592
+    },
+    {
+      "epoch": 0.37801157285639136,
+      "grad_norm": 1.5846307277679443,
+      "learning_rate": 0.0001429426576419298,
+      "loss": 1.879,
+      "step": 3593
+    },
+    {
+      "epoch": 0.3781167806417675,
+      "grad_norm": 1.696427822113037,
+      "learning_rate": 0.00014291187981665607,
+      "loss": 1.921,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3782219884271436,
+      "grad_norm": 1.756914496421814,
+      "learning_rate": 0.00014288109700816705,
+      "loss": 1.7842,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3783271962125197,
+      "grad_norm": 1.9030567407608032,
+      "learning_rate": 0.0001428503092200374,
+      "loss": 1.8851,
+      "step": 3596
+    },
+    {
+      "epoch": 0.37843240399789585,
+      "grad_norm": 0.8810452818870544,
+      "learning_rate": 0.0001428195164558425,
+      "loss": 1.7901,
+      "step": 3597
+    },
+    {
+      "epoch": 0.37853761178327194,
+      "grad_norm": 1.4262439012527466,
+      "learning_rate": 0.00014278871871915814,
+      "loss": 1.9453,
+      "step": 3598
+    },
+    {
+      "epoch": 0.3786428195686481,
+      "grad_norm": 1.6858327388763428,
+      "learning_rate": 0.00014275791601356074,
+      "loss": 1.8671,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3787480273540242,
+      "grad_norm": 0.9039170742034912,
+      "learning_rate": 0.0001427271083426274,
+      "loss": 2.199,
+      "step": 3600
+    },
+    {
+      "epoch": 0.37885323513940034,
+      "grad_norm": 1.2409204244613647,
+      "learning_rate": 0.00014269629570993564,
+      "loss": 1.6864,
+      "step": 3601
+    },
+    {
+      "epoch": 0.37895844292477643,
+      "grad_norm": 1.6588237285614014,
+      "learning_rate": 0.00014266547811906364,
+      "loss": 2.1573,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3790636507101526,
+      "grad_norm": 1.5845474004745483,
+      "learning_rate": 0.00014263465557359017,
+      "loss": 1.7634,
+      "step": 3603
+    },
+    {
+      "epoch": 0.37916885849552867,
+      "grad_norm": 1.337897539138794,
+      "learning_rate": 0.00014260382807709457,
+      "loss": 2.0049,
+      "step": 3604
+    },
+    {
+      "epoch": 0.37927406628090476,
+      "grad_norm": 1.3715331554412842,
+      "learning_rate": 0.00014257299563315667,
+      "loss": 1.8703,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3793792740662809,
+      "grad_norm": 0.8518646955490112,
+      "learning_rate": 0.00014254215824535698,
+      "loss": 2.0015,
+      "step": 3606
+    },
+    {
+      "epoch": 0.379484481851657,
+      "grad_norm": 0.998729407787323,
+      "learning_rate": 0.00014251131591727656,
+      "loss": 1.7552,
+      "step": 3607
+    },
+    {
+      "epoch": 0.37958968963703316,
+      "grad_norm": 1.5885865688323975,
+      "learning_rate": 0.00014248046865249697,
+      "loss": 1.7364,
+      "step": 3608
+    },
+    {
+      "epoch": 0.37969489742240925,
+      "grad_norm": 1.1192876100540161,
+      "learning_rate": 0.00014244961645460048,
+      "loss": 1.9686,
+      "step": 3609
+    },
+    {
+      "epoch": 0.3798001052077854,
+      "grad_norm": 1.3762092590332031,
+      "learning_rate": 0.0001424187593271698,
+      "loss": 1.668,
+      "step": 3610
+    },
+    {
+      "epoch": 0.3799053129931615,
+      "grad_norm": 1.3695660829544067,
+      "learning_rate": 0.0001423878972737883,
+      "loss": 1.939,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3800105207785376,
+      "grad_norm": 1.5928248167037964,
+      "learning_rate": 0.00014235703029803984,
+      "loss": 1.6483,
+      "step": 3612
+    },
+    {
+      "epoch": 0.38011572856391374,
+      "grad_norm": 1.027282953262329,
+      "learning_rate": 0.00014232615840350894,
+      "loss": 1.3274,
+      "step": 3613
+    },
+    {
+      "epoch": 0.38022093634928983,
+      "grad_norm": 1.2726476192474365,
+      "learning_rate": 0.00014229528159378065,
+      "loss": 2.1561,
+      "step": 3614
+    },
+    {
+      "epoch": 0.380326144134666,
+      "grad_norm": 1.3544089794158936,
+      "learning_rate": 0.00014226439987244057,
+      "loss": 1.5668,
+      "step": 3615
+    },
+    {
+      "epoch": 0.38043135192004207,
+      "grad_norm": 1.9423713684082031,
+      "learning_rate": 0.00014223351324307493,
+      "loss": 1.7549,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3805365597054182,
+      "grad_norm": 0.9007306694984436,
+      "learning_rate": 0.00014220262170927046,
+      "loss": 2.1485,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3806417674907943,
+      "grad_norm": 1.2513463497161865,
+      "learning_rate": 0.0001421717252746145,
+      "loss": 1.9919,
+      "step": 3618
+    },
+    {
+      "epoch": 0.38074697527617046,
+      "grad_norm": 1.4759798049926758,
+      "learning_rate": 0.00014214082394269493,
+      "loss": 2.2475,
+      "step": 3619
+    },
+    {
+      "epoch": 0.38085218306154656,
+      "grad_norm": 1.4133992195129395,
+      "learning_rate": 0.00014210991771710025,
+      "loss": 2.229,
+      "step": 3620
+    },
+    {
+      "epoch": 0.38095739084692265,
+      "grad_norm": 1.1135759353637695,
+      "learning_rate": 0.0001420790066014195,
+      "loss": 1.7897,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3810625986322988,
+      "grad_norm": 1.0141302347183228,
+      "learning_rate": 0.00014204809059924228,
+      "loss": 1.5876,
+      "step": 3622
+    },
+    {
+      "epoch": 0.3811678064176749,
+      "grad_norm": 1.2272731065750122,
+      "learning_rate": 0.00014201716971415875,
+      "loss": 1.9002,
+      "step": 3623
+    },
+    {
+      "epoch": 0.38127301420305104,
+      "grad_norm": 1.4292423725128174,
+      "learning_rate": 0.00014198624394975968,
+      "loss": 1.5974,
+      "step": 3624
+    },
+    {
+      "epoch": 0.38137822198842714,
+      "grad_norm": 1.052093744277954,
+      "learning_rate": 0.00014195531330963635,
+      "loss": 1.9424,
+      "step": 3625
+    },
+    {
+      "epoch": 0.3814834297738033,
+      "grad_norm": 1.377421498298645,
+      "learning_rate": 0.00014192437779738062,
+      "loss": 2.4788,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3815886375591794,
+      "grad_norm": 1.4875088930130005,
+      "learning_rate": 0.00014189343741658497,
+      "loss": 1.8751,
+      "step": 3627
+    },
+    {
+      "epoch": 0.38169384534455547,
+      "grad_norm": 1.2840195894241333,
+      "learning_rate": 0.0001418624921708424,
+      "loss": 1.8626,
+      "step": 3628
+    },
+    {
+      "epoch": 0.3817990531299316,
+      "grad_norm": 1.7244127988815308,
+      "learning_rate": 0.00014183154206374643,
+      "loss": 1.7683,
+      "step": 3629
+    },
+    {
+      "epoch": 0.3819042609153077,
+      "grad_norm": 1.5196006298065186,
+      "learning_rate": 0.0001418005870988912,
+      "loss": 2.0749,
+      "step": 3630
+    },
+    {
+      "epoch": 0.38200946870068386,
+      "grad_norm": 1.3021456003189087,
+      "learning_rate": 0.0001417696272798715,
+      "loss": 2.3025,
+      "step": 3631
+    },
+    {
+      "epoch": 0.38211467648605996,
+      "grad_norm": 0.9717100262641907,
+      "learning_rate": 0.0001417386626102825,
+      "loss": 2.2728,
+      "step": 3632
+    },
+    {
+      "epoch": 0.3822198842714361,
+      "grad_norm": 1.2531250715255737,
+      "learning_rate": 0.00014170769309372006,
+      "loss": 2.11,
+      "step": 3633
+    },
+    {
+      "epoch": 0.3823250920568122,
+      "grad_norm": 3.0939295291900635,
+      "learning_rate": 0.00014167671873378056,
+      "loss": 2.119,
+      "step": 3634
+    },
+    {
+      "epoch": 0.38243029984218835,
+      "grad_norm": 1.3304295539855957,
+      "learning_rate": 0.00014164573953406095,
+      "loss": 1.6821,
+      "step": 3635
+    },
+    {
+      "epoch": 0.38253550762756444,
+      "grad_norm": 1.434133768081665,
+      "learning_rate": 0.00014161475549815877,
+      "loss": 2.2729,
+      "step": 3636
+    },
+    {
+      "epoch": 0.38264071541294054,
+      "grad_norm": 1.4510753154754639,
+      "learning_rate": 0.00014158376662967202,
+      "loss": 2.2107,
+      "step": 3637
+    },
+    {
+      "epoch": 0.3827459231983167,
+      "grad_norm": 1.247135877609253,
+      "learning_rate": 0.0001415527729321994,
+      "loss": 1.846,
+      "step": 3638
+    },
+    {
+      "epoch": 0.3828511309836928,
+      "grad_norm": 0.9109544157981873,
+      "learning_rate": 0.00014152177440934012,
+      "loss": 1.9013,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3829563387690689,
+      "grad_norm": 0.9982622265815735,
+      "learning_rate": 0.00014149077106469387,
+      "loss": 2.0598,
+      "step": 3640
+    },
+    {
+      "epoch": 0.383061546554445,
+      "grad_norm": 1.1758394241333008,
+      "learning_rate": 0.00014145976290186102,
+      "loss": 2.3168,
+      "step": 3641
+    },
+    {
+      "epoch": 0.38316675433982117,
+      "grad_norm": 0.8653580546379089,
+      "learning_rate": 0.00014142874992444243,
+      "loss": 1.5041,
+      "step": 3642
+    },
+    {
+      "epoch": 0.38327196212519726,
+      "grad_norm": 1.015748143196106,
+      "learning_rate": 0.0001413977321360395,
+      "loss": 1.9884,
+      "step": 3643
+    },
+    {
+      "epoch": 0.38337716991057336,
+      "grad_norm": 1.737280249595642,
+      "learning_rate": 0.00014136670954025427,
+      "loss": 1.7972,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3834823776959495,
+      "grad_norm": 2.5558969974517822,
+      "learning_rate": 0.0001413356821406893,
+      "loss": 2.1549,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3835875854813256,
+      "grad_norm": 1.4405326843261719,
+      "learning_rate": 0.0001413046499409477,
+      "loss": 1.9342,
+      "step": 3646
+    },
+    {
+      "epoch": 0.38369279326670175,
+      "grad_norm": 1.235929012298584,
+      "learning_rate": 0.0001412736129446331,
+      "loss": 2.0326,
+      "step": 3647
+    },
+    {
+      "epoch": 0.38379800105207784,
+      "grad_norm": 1.8266171216964722,
+      "learning_rate": 0.0001412425711553497,
+      "loss": 1.9242,
+      "step": 3648
+    },
+    {
+      "epoch": 0.383903208837454,
+      "grad_norm": 1.1843401193618774,
+      "learning_rate": 0.00014121152457670234,
+      "loss": 2.4682,
+      "step": 3649
+    },
+    {
+      "epoch": 0.3840084166228301,
+      "grad_norm": 1.2029701471328735,
+      "learning_rate": 0.00014118047321229633,
+      "loss": 1.5255,
+      "step": 3650
+    },
+    {
+      "epoch": 0.38411362440820623,
+      "grad_norm": 1.264588713645935,
+      "learning_rate": 0.00014114941706573758,
+      "loss": 1.5459,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3842188321935823,
+      "grad_norm": 1.7081525325775146,
+      "learning_rate": 0.00014111835614063253,
+      "loss": 2.1296,
+      "step": 3652
+    },
+    {
+      "epoch": 0.3843240399789584,
+      "grad_norm": 0.9113522171974182,
+      "learning_rate": 0.0001410872904405882,
+      "loss": 1.5649,
+      "step": 3653
+    },
+    {
+      "epoch": 0.38442924776433457,
+      "grad_norm": 1.2940986156463623,
+      "learning_rate": 0.0001410562199692121,
+      "loss": 1.6233,
+      "step": 3654
+    },
+    {
+      "epoch": 0.38453445554971066,
+      "grad_norm": 1.201011300086975,
+      "learning_rate": 0.00014102514473011233,
+      "loss": 1.9928,
+      "step": 3655
+    },
+    {
+      "epoch": 0.3846396633350868,
+      "grad_norm": 1.310477614402771,
+      "learning_rate": 0.0001409940647268977,
+      "loss": 1.7539,
+      "step": 3656
+    },
+    {
+      "epoch": 0.3847448711204629,
+      "grad_norm": 0.9373623132705688,
+      "learning_rate": 0.00014096297996317724,
+      "loss": 1.9289,
+      "step": 3657
+    },
+    {
+      "epoch": 0.38485007890583905,
+      "grad_norm": 1.0480597019195557,
+      "learning_rate": 0.00014093189044256084,
+      "loss": 2.0449,
+      "step": 3658
+    },
+    {
+      "epoch": 0.38495528669121515,
+      "grad_norm": 1.2501201629638672,
+      "learning_rate": 0.00014090079616865882,
+      "loss": 1.7853,
+      "step": 3659
+    },
+    {
+      "epoch": 0.38506049447659124,
+      "grad_norm": 1.2125840187072754,
+      "learning_rate": 0.00014086969714508196,
+      "loss": 1.3741,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3851657022619674,
+      "grad_norm": 0.8791213631629944,
+      "learning_rate": 0.00014083859337544175,
+      "loss": 1.7084,
+      "step": 3661
+    },
+    {
+      "epoch": 0.3852709100473435,
+      "grad_norm": 1.8772777318954468,
+      "learning_rate": 0.00014080748486335022,
+      "loss": 1.6383,
+      "step": 3662
+    },
+    {
+      "epoch": 0.38537611783271963,
+      "grad_norm": 1.3970454931259155,
+      "learning_rate": 0.0001407763716124198,
+      "loss": 1.5493,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3854813256180957,
+      "grad_norm": 1.5599162578582764,
+      "learning_rate": 0.00014074525362626366,
+      "loss": 2.0028,
+      "step": 3664
+    },
+    {
+      "epoch": 0.3855865334034719,
+      "grad_norm": 1.0183049440383911,
+      "learning_rate": 0.00014071413090849534,
+      "loss": 2.151,
+      "step": 3665
+    },
+    {
+      "epoch": 0.38569174118884797,
+      "grad_norm": 1.4794666767120361,
+      "learning_rate": 0.0001406830034627291,
+      "loss": 1.8065,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3857969489742241,
+      "grad_norm": 1.77559232711792,
+      "learning_rate": 0.00014065187129257964,
+      "loss": 2.3033,
+      "step": 3667
+    },
+    {
+      "epoch": 0.3859021567596002,
+      "grad_norm": 2.2555582523345947,
+      "learning_rate": 0.00014062073440166222,
+      "loss": 1.48,
+      "step": 3668
+    },
+    {
+      "epoch": 0.3860073645449763,
+      "grad_norm": 0.9742986559867859,
+      "learning_rate": 0.00014058959279359266,
+      "loss": 1.6655,
+      "step": 3669
+    },
+    {
+      "epoch": 0.38611257233035245,
+      "grad_norm": 1.42844557762146,
+      "learning_rate": 0.00014055844647198738,
+      "loss": 2.1812,
+      "step": 3670
+    },
+    {
+      "epoch": 0.38621778011572855,
+      "grad_norm": 1.0865979194641113,
+      "learning_rate": 0.00014052729544046326,
+      "loss": 1.8439,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3863229879011047,
+      "grad_norm": 1.070984125137329,
+      "learning_rate": 0.00014049613970263774,
+      "loss": 1.7071,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3864281956864808,
+      "grad_norm": 1.4193851947784424,
+      "learning_rate": 0.0001404649792621289,
+      "loss": 2.1123,
+      "step": 3673
+    },
+    {
+      "epoch": 0.38653340347185694,
+      "grad_norm": 1.3338897228240967,
+      "learning_rate": 0.00014043381412255526,
+      "loss": 2.0661,
+      "step": 3674
+    },
+    {
+      "epoch": 0.38663861125723303,
+      "grad_norm": 1.063714861869812,
+      "learning_rate": 0.00014040264428753592,
+      "loss": 1.5893,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3867438190426091,
+      "grad_norm": 1.06549870967865,
+      "learning_rate": 0.00014037146976069055,
+      "loss": 1.7869,
+      "step": 3676
+    },
+    {
+      "epoch": 0.3868490268279853,
+      "grad_norm": 1.382040023803711,
+      "learning_rate": 0.00014034029054563933,
+      "loss": 1.5798,
+      "step": 3677
+    },
+    {
+      "epoch": 0.38695423461336137,
+      "grad_norm": 1.4620635509490967,
+      "learning_rate": 0.000140309106646003,
+      "loss": 1.7959,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3870594423987375,
+      "grad_norm": 1.2155455350875854,
+      "learning_rate": 0.0001402779180654029,
+      "loss": 1.5766,
+      "step": 3679
+    },
+    {
+      "epoch": 0.3871646501841136,
+      "grad_norm": 1.3501040935516357,
+      "learning_rate": 0.00014024672480746078,
+      "loss": 1.7963,
+      "step": 3680
+    },
+    {
+      "epoch": 0.38726985796948976,
+      "grad_norm": 1.1991915702819824,
+      "learning_rate": 0.00014021552687579902,
+      "loss": 1.9882,
+      "step": 3681
+    },
+    {
+      "epoch": 0.38737506575486585,
+      "grad_norm": 2.334041118621826,
+      "learning_rate": 0.00014018432427404055,
+      "loss": 1.3908,
+      "step": 3682
+    },
+    {
+      "epoch": 0.387480273540242,
+      "grad_norm": 1.5623788833618164,
+      "learning_rate": 0.0001401531170058088,
+      "loss": 1.2739,
+      "step": 3683
+    },
+    {
+      "epoch": 0.3875854813256181,
+      "grad_norm": 1.153784155845642,
+      "learning_rate": 0.00014012190507472783,
+      "loss": 1.6715,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3876906891109942,
+      "grad_norm": 0.9832100868225098,
+      "learning_rate": 0.00014009068848442214,
+      "loss": 2.1983,
+      "step": 3685
+    },
+    {
+      "epoch": 0.38779589689637034,
+      "grad_norm": 1.071271538734436,
+      "learning_rate": 0.0001400594672385168,
+      "loss": 2.1699,
+      "step": 3686
+    },
+    {
+      "epoch": 0.38790110468174643,
+      "grad_norm": 1.600494146347046,
+      "learning_rate": 0.00014002824134063747,
+      "loss": 1.9091,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3880063124671226,
+      "grad_norm": 1.5296038389205933,
+      "learning_rate": 0.00013999701079441028,
+      "loss": 2.0367,
+      "step": 3688
+    },
+    {
+      "epoch": 0.3881115202524987,
+      "grad_norm": 1.072365403175354,
+      "learning_rate": 0.0001399657756034619,
+      "loss": 2.0738,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3882167280378748,
+      "grad_norm": 1.1414501667022705,
+      "learning_rate": 0.00013993453577141964,
+      "loss": 1.5334,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3883219358232509,
+      "grad_norm": 1.220701813697815,
+      "learning_rate": 0.00013990329130191123,
+      "loss": 1.9406,
+      "step": 3691
+    },
+    {
+      "epoch": 0.388427143608627,
+      "grad_norm": 1.4676032066345215,
+      "learning_rate": 0.000139872042198565,
+      "loss": 1.6478,
+      "step": 3692
+    },
+    {
+      "epoch": 0.38853235139400316,
+      "grad_norm": 2.2082393169403076,
+      "learning_rate": 0.0001398407884650098,
+      "loss": 2.5011,
+      "step": 3693
+    },
+    {
+      "epoch": 0.38863755917937925,
+      "grad_norm": 1.4031996726989746,
+      "learning_rate": 0.000139809530104875,
+      "loss": 1.8676,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3887427669647554,
+      "grad_norm": 1.0090667009353638,
+      "learning_rate": 0.00013977826712179058,
+      "loss": 1.7266,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3888479747501315,
+      "grad_norm": 1.4006050825119019,
+      "learning_rate": 0.000139746999519387,
+      "loss": 1.8296,
+      "step": 3696
+    },
+    {
+      "epoch": 0.38895318253550765,
+      "grad_norm": 1.0253705978393555,
+      "learning_rate": 0.00013971572730129525,
+      "loss": 1.7789,
+      "step": 3697
+    },
+    {
+      "epoch": 0.38905839032088374,
+      "grad_norm": 1.1264349222183228,
+      "learning_rate": 0.00013968445047114685,
+      "loss": 2.0904,
+      "step": 3698
+    },
+    {
+      "epoch": 0.3891635981062599,
+      "grad_norm": 1.5611605644226074,
+      "learning_rate": 0.0001396531690325739,
+      "loss": 2.101,
+      "step": 3699
+    },
+    {
+      "epoch": 0.389268805891636,
+      "grad_norm": 1.7648766040802002,
+      "learning_rate": 0.00013962188298920902,
+      "loss": 2.1288,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3893740136770121,
+      "grad_norm": 2.0247011184692383,
+      "learning_rate": 0.00013959059234468536,
+      "loss": 1.7906,
+      "step": 3701
+    },
+    {
+      "epoch": 0.3894792214623882,
+      "grad_norm": 1.0505448579788208,
+      "learning_rate": 0.00013955929710263653,
+      "loss": 1.995,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3895844292477643,
+      "grad_norm": 1.6498749256134033,
+      "learning_rate": 0.00013952799726669682,
+      "loss": 1.8126,
+      "step": 3703
+    },
+    {
+      "epoch": 0.38968963703314047,
+      "grad_norm": 1.861400842666626,
+      "learning_rate": 0.00013949669284050092,
+      "loss": 2.3192,
+      "step": 3704
+    },
+    {
+      "epoch": 0.38979484481851656,
+      "grad_norm": 1.1413242816925049,
+      "learning_rate": 0.00013946538382768418,
+      "loss": 1.4746,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3899000526038927,
+      "grad_norm": 1.281286597251892,
+      "learning_rate": 0.00013943407023188234,
+      "loss": 2.3798,
+      "step": 3706
+    },
+    {
+      "epoch": 0.3900052603892688,
+      "grad_norm": 1.3880997896194458,
+      "learning_rate": 0.00013940275205673178,
+      "loss": 2.0053,
+      "step": 3707
+    },
+    {
+      "epoch": 0.3901104681746449,
+      "grad_norm": 1.1959868669509888,
+      "learning_rate": 0.0001393714293058694,
+      "loss": 2.0179,
+      "step": 3708
+    },
+    {
+      "epoch": 0.39021567596002105,
+      "grad_norm": 1.6288052797317505,
+      "learning_rate": 0.00013934010198293257,
+      "loss": 2.1802,
+      "step": 3709
+    },
+    {
+      "epoch": 0.39032088374539714,
+      "grad_norm": 1.229329228401184,
+      "learning_rate": 0.00013930877009155922,
+      "loss": 1.8899,
+      "step": 3710
+    },
+    {
+      "epoch": 0.3904260915307733,
+      "grad_norm": 1.2303109169006348,
+      "learning_rate": 0.00013927743363538787,
+      "loss": 1.6863,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3905312993161494,
+      "grad_norm": 1.653576374053955,
+      "learning_rate": 0.0001392460926180575,
+      "loss": 1.765,
+      "step": 3712
+    },
+    {
+      "epoch": 0.39063650710152553,
+      "grad_norm": 1.6338762044906616,
+      "learning_rate": 0.0001392147470432076,
+      "loss": 2.0503,
+      "step": 3713
+    },
+    {
+      "epoch": 0.3907417148869016,
+      "grad_norm": 1.4691399335861206,
+      "learning_rate": 0.00013918339691447825,
+      "loss": 2.1288,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3908469226722778,
+      "grad_norm": 1.0673540830612183,
+      "learning_rate": 0.0001391520422355101,
+      "loss": 1.8172,
+      "step": 3715
+    },
+    {
+      "epoch": 0.39095213045765387,
+      "grad_norm": 0.9647770524024963,
+      "learning_rate": 0.00013912068300994413,
+      "loss": 1.7418,
+      "step": 3716
+    },
+    {
+      "epoch": 0.39105733824302996,
+      "grad_norm": 1.7503970861434937,
+      "learning_rate": 0.0001390893192414221,
+      "loss": 1.78,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3911625460284061,
+      "grad_norm": 0.7777706980705261,
+      "learning_rate": 0.00013905795093358615,
+      "loss": 1.9819,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3912677538137822,
+      "grad_norm": 1.5159608125686646,
+      "learning_rate": 0.00013902657809007897,
+      "loss": 1.7175,
+      "step": 3719
+    },
+    {
+      "epoch": 0.39137296159915835,
+      "grad_norm": 1.442254662513733,
+      "learning_rate": 0.00013899520071454377,
+      "loss": 2.1266,
+      "step": 3720
+    },
+    {
+      "epoch": 0.39147816938453445,
+      "grad_norm": 1.2013967037200928,
+      "learning_rate": 0.00013896381881062437,
+      "loss": 1.9828,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3915833771699106,
+      "grad_norm": 2.616184949874878,
+      "learning_rate": 0.00013893243238196495,
+      "loss": 1.7142,
+      "step": 3722
+    },
+    {
+      "epoch": 0.3916885849552867,
+      "grad_norm": 1.076119065284729,
+      "learning_rate": 0.0001389010414322104,
+      "loss": 2.2307,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3917937927406628,
+      "grad_norm": 1.1124643087387085,
+      "learning_rate": 0.00013886964596500595,
+      "loss": 1.7956,
+      "step": 3724
+    },
+    {
+      "epoch": 0.39189900052603893,
+      "grad_norm": 1.8706750869750977,
+      "learning_rate": 0.00013883824598399756,
+      "loss": 2.3762,
+      "step": 3725
+    },
+    {
+      "epoch": 0.392004208311415,
+      "grad_norm": 1.2666409015655518,
+      "learning_rate": 0.00013880684149283152,
+      "loss": 1.5874,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3921094160967912,
+      "grad_norm": 1.3697994947433472,
+      "learning_rate": 0.00013877543249515476,
+      "loss": 2.4001,
+      "step": 3727
+    },
+    {
+      "epoch": 0.39221462388216727,
+      "grad_norm": 1.538824200630188,
+      "learning_rate": 0.00013874401899461474,
+      "loss": 1.8197,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3923198316675434,
+      "grad_norm": 1.4980809688568115,
+      "learning_rate": 0.00013871260099485936,
+      "loss": 1.434,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3924250394529195,
+      "grad_norm": 1.348250150680542,
+      "learning_rate": 0.0001386811784995371,
+      "loss": 1.6379,
+      "step": 3730
+    },
+    {
+      "epoch": 0.39253024723829566,
+      "grad_norm": 2.0274887084960938,
+      "learning_rate": 0.00013864975151229697,
+      "loss": 1.7713,
+      "step": 3731
+    },
+    {
+      "epoch": 0.39263545502367175,
+      "grad_norm": 1.4459730386734009,
+      "learning_rate": 0.00013861832003678846,
+      "loss": 1.68,
+      "step": 3732
+    },
+    {
+      "epoch": 0.39274066280904785,
+      "grad_norm": 1.5965417623519897,
+      "learning_rate": 0.00013858688407666163,
+      "loss": 1.962,
+      "step": 3733
+    },
+    {
+      "epoch": 0.392845870594424,
+      "grad_norm": 1.5508756637573242,
+      "learning_rate": 0.00013855544363556698,
+      "loss": 2.1783,
+      "step": 3734
+    },
+    {
+      "epoch": 0.3929510783798001,
+      "grad_norm": 1.3883273601531982,
+      "learning_rate": 0.00013852399871715562,
+      "loss": 2.3433,
+      "step": 3735
+    },
+    {
+      "epoch": 0.39305628616517624,
+      "grad_norm": 1.741829752922058,
+      "learning_rate": 0.00013849254932507917,
+      "loss": 2.3758,
+      "step": 3736
+    },
+    {
+      "epoch": 0.39316149395055233,
+      "grad_norm": 1.35794198513031,
+      "learning_rate": 0.00013846109546298971,
+      "loss": 2.0261,
+      "step": 3737
+    },
+    {
+      "epoch": 0.3932667017359285,
+      "grad_norm": 1.0668870210647583,
+      "learning_rate": 0.00013842963713453987,
+      "loss": 2.2452,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3933719095213046,
+      "grad_norm": 1.3212822675704956,
+      "learning_rate": 0.00013839817434338286,
+      "loss": 1.9696,
+      "step": 3739
+    },
+    {
+      "epoch": 0.39347711730668067,
+      "grad_norm": 0.7865563035011292,
+      "learning_rate": 0.00013836670709317225,
+      "loss": 1.6841,
+      "step": 3740
+    },
+    {
+      "epoch": 0.3935823250920568,
+      "grad_norm": 1.023016095161438,
+      "learning_rate": 0.0001383352353875623,
+      "loss": 2.1734,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3936875328774329,
+      "grad_norm": 0.966575562953949,
+      "learning_rate": 0.00013830375923020772,
+      "loss": 1.735,
+      "step": 3742
+    },
+    {
+      "epoch": 0.39379274066280906,
+      "grad_norm": 1.255327582359314,
+      "learning_rate": 0.00013827227862476372,
+      "loss": 2.0901,
+      "step": 3743
+    },
+    {
+      "epoch": 0.39389794844818515,
+      "grad_norm": 1.4241955280303955,
+      "learning_rate": 0.00013824079357488598,
+      "loss": 1.5569,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3940031562335613,
+      "grad_norm": 0.807941734790802,
+      "learning_rate": 0.00013820930408423086,
+      "loss": 1.9292,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3941083640189374,
+      "grad_norm": 1.034824013710022,
+      "learning_rate": 0.00013817781015645507,
+      "loss": 2.1646,
+      "step": 3746
+    },
+    {
+      "epoch": 0.39421357180431355,
+      "grad_norm": 1.143803358078003,
+      "learning_rate": 0.00013814631179521588,
+      "loss": 2.2347,
+      "step": 3747
+    },
+    {
+      "epoch": 0.39431877958968964,
+      "grad_norm": 1.7627426385879517,
+      "learning_rate": 0.0001381148090041711,
+      "loss": 1.8837,
+      "step": 3748
+    },
+    {
+      "epoch": 0.39442398737506573,
+      "grad_norm": 1.153597354888916,
+      "learning_rate": 0.0001380833017869791,
+      "loss": 1.989,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3945291951604419,
+      "grad_norm": 1.3962596654891968,
+      "learning_rate": 0.00013805179014729865,
+      "loss": 2.4745,
+      "step": 3750
+    },
+    {
+      "epoch": 0.394634402945818,
+      "grad_norm": 1.1619459390640259,
+      "learning_rate": 0.0001380202740887891,
+      "loss": 2.1719,
+      "step": 3751
+    },
+    {
+      "epoch": 0.3947396107311941,
+      "grad_norm": 1.4817026853561401,
+      "learning_rate": 0.00013798875361511033,
+      "loss": 2.6195,
+      "step": 3752
+    },
+    {
+      "epoch": 0.3948448185165702,
+      "grad_norm": 1.251517415046692,
+      "learning_rate": 0.00013795722872992272,
+      "loss": 1.9008,
+      "step": 3753
+    },
+    {
+      "epoch": 0.39495002630194637,
+      "grad_norm": 1.6343916654586792,
+      "learning_rate": 0.0001379256994368871,
+      "loss": 1.8265,
+      "step": 3754
+    },
+    {
+      "epoch": 0.39505523408732246,
+      "grad_norm": 1.4928406476974487,
+      "learning_rate": 0.0001378941657396649,
+      "loss": 2.0495,
+      "step": 3755
+    },
+    {
+      "epoch": 0.39516044187269855,
+      "grad_norm": 0.9346429109573364,
+      "learning_rate": 0.00013786262764191803,
+      "loss": 2.0691,
+      "step": 3756
+    },
+    {
+      "epoch": 0.3952656496580747,
+      "grad_norm": 1.5070823431015015,
+      "learning_rate": 0.00013783108514730884,
+      "loss": 1.6465,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3953708574434508,
+      "grad_norm": 1.225148320198059,
+      "learning_rate": 0.00013779953825950034,
+      "loss": 1.3778,
+      "step": 3758
+    },
+    {
+      "epoch": 0.39547606522882695,
+      "grad_norm": 1.4228816032409668,
+      "learning_rate": 0.00013776798698215593,
+      "loss": 1.559,
+      "step": 3759
+    },
+    {
+      "epoch": 0.39558127301420304,
+      "grad_norm": 1.1593899726867676,
+      "learning_rate": 0.00013773643131893956,
+      "loss": 2.1666,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3956864807995792,
+      "grad_norm": 1.2001755237579346,
+      "learning_rate": 0.00013770487127351568,
+      "loss": 1.5535,
+      "step": 3761
+    },
+    {
+      "epoch": 0.3957916885849553,
+      "grad_norm": 1.1403510570526123,
+      "learning_rate": 0.00013767330684954926,
+      "loss": 1.765,
+      "step": 3762
+    },
+    {
+      "epoch": 0.39589689637033143,
+      "grad_norm": 1.4081346988677979,
+      "learning_rate": 0.00013764173805070576,
+      "loss": 2.1388,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3960021041557075,
+      "grad_norm": 0.9673276543617249,
+      "learning_rate": 0.00013761016488065118,
+      "loss": 2.1821,
+      "step": 3764
+    },
+    {
+      "epoch": 0.3961073119410836,
+      "grad_norm": 1.567577838897705,
+      "learning_rate": 0.00013757858734305203,
+      "loss": 2.0253,
+      "step": 3765
+    },
+    {
+      "epoch": 0.39621251972645977,
+      "grad_norm": 1.2986990213394165,
+      "learning_rate": 0.00013754700544157524,
+      "loss": 2.1731,
+      "step": 3766
+    },
+    {
+      "epoch": 0.39631772751183586,
+      "grad_norm": 1.9788020849227905,
+      "learning_rate": 0.00013751541917988836,
+      "loss": 1.8981,
+      "step": 3767
+    },
+    {
+      "epoch": 0.396422935297212,
+      "grad_norm": 1.5377720594406128,
+      "learning_rate": 0.0001374838285616594,
+      "loss": 2.0393,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3965281430825881,
+      "grad_norm": 0.8659994006156921,
+      "learning_rate": 0.00013745223359055682,
+      "loss": 1.7419,
+      "step": 3769
+    },
+    {
+      "epoch": 0.39663335086796425,
+      "grad_norm": 1.1457507610321045,
+      "learning_rate": 0.0001374206342702497,
+      "loss": 2.1031,
+      "step": 3770
+    },
+    {
+      "epoch": 0.39673855865334035,
+      "grad_norm": 0.8939380645751953,
+      "learning_rate": 0.00013738903060440757,
+      "loss": 2.0608,
+      "step": 3771
+    },
+    {
+      "epoch": 0.39684376643871644,
+      "grad_norm": 1.0141644477844238,
+      "learning_rate": 0.0001373574225967004,
+      "loss": 1.9432,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3969489742240926,
+      "grad_norm": 1.7639262676239014,
+      "learning_rate": 0.0001373258102507988,
+      "loss": 1.7746,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3970541820094687,
+      "grad_norm": 1.8275938034057617,
+      "learning_rate": 0.00013729419357037372,
+      "loss": 1.9925,
+      "step": 3774
+    },
+    {
+      "epoch": 0.39715938979484483,
+      "grad_norm": 1.8225183486938477,
+      "learning_rate": 0.00013726257255909676,
+      "loss": 2.0511,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3972645975802209,
+      "grad_norm": 1.1750774383544922,
+      "learning_rate": 0.00013723094722063996,
+      "loss": 2.0788,
+      "step": 3776
+    },
+    {
+      "epoch": 0.3973698053655971,
+      "grad_norm": 1.4258540868759155,
+      "learning_rate": 0.00013719931755867587,
+      "loss": 2.0758,
+      "step": 3777
+    },
+    {
+      "epoch": 0.39747501315097317,
+      "grad_norm": 0.9708757400512695,
+      "learning_rate": 0.0001371676835768775,
+      "loss": 1.8594,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3975802209363493,
+      "grad_norm": 1.1266876459121704,
+      "learning_rate": 0.00013713604527891844,
+      "loss": 1.6579,
+      "step": 3779
+    },
+    {
+      "epoch": 0.3976854287217254,
+      "grad_norm": 2.256415605545044,
+      "learning_rate": 0.00013710440266847274,
+      "loss": 2.2314,
+      "step": 3780
+    },
+    {
+      "epoch": 0.3977906365071015,
+      "grad_norm": 1.4716403484344482,
+      "learning_rate": 0.0001370727557492149,
+      "loss": 1.7153,
+      "step": 3781
+    },
+    {
+      "epoch": 0.39789584429247765,
+      "grad_norm": 1.2496949434280396,
+      "learning_rate": 0.00013704110452482005,
+      "loss": 2.0778,
+      "step": 3782
+    },
+    {
+      "epoch": 0.39800105207785375,
+      "grad_norm": 0.9806689620018005,
+      "learning_rate": 0.0001370094489989637,
+      "loss": 2.1003,
+      "step": 3783
+    },
+    {
+      "epoch": 0.3981062598632299,
+      "grad_norm": 1.0199190378189087,
+      "learning_rate": 0.00013697778917532192,
+      "loss": 2.0489,
+      "step": 3784
+    },
+    {
+      "epoch": 0.398211467648606,
+      "grad_norm": 0.8735719919204712,
+      "learning_rate": 0.00013694612505757122,
+      "loss": 1.2784,
+      "step": 3785
+    },
+    {
+      "epoch": 0.39831667543398214,
+      "grad_norm": 1.2320698499679565,
+      "learning_rate": 0.00013691445664938866,
+      "loss": 2.0641,
+      "step": 3786
+    },
+    {
+      "epoch": 0.39842188321935823,
+      "grad_norm": 1.306445598602295,
+      "learning_rate": 0.00013688278395445185,
+      "loss": 1.8181,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3985270910047343,
+      "grad_norm": 1.5891810655593872,
+      "learning_rate": 0.00013685110697643878,
+      "loss": 1.7734,
+      "step": 3788
+    },
+    {
+      "epoch": 0.3986322987901105,
+      "grad_norm": 1.2579609155654907,
+      "learning_rate": 0.00013681942571902803,
+      "loss": 1.9562,
+      "step": 3789
+    },
+    {
+      "epoch": 0.39873750657548657,
+      "grad_norm": 1.3183737993240356,
+      "learning_rate": 0.00013678774018589855,
+      "loss": 1.824,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3988427143608627,
+      "grad_norm": 1.2794685363769531,
+      "learning_rate": 0.00013675605038072997,
+      "loss": 1.4231,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3989479221462388,
+      "grad_norm": 1.0361131429672241,
+      "learning_rate": 0.00013672435630720232,
+      "loss": 1.7537,
+      "step": 3792
+    },
+    {
+      "epoch": 0.39905312993161496,
+      "grad_norm": 1.2648097276687622,
+      "learning_rate": 0.00013669265796899607,
+      "loss": 2.534,
+      "step": 3793
+    },
+    {
+      "epoch": 0.39915833771699105,
+      "grad_norm": 1.1882283687591553,
+      "learning_rate": 0.00013666095536979232,
+      "loss": 1.7441,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3992635455023672,
+      "grad_norm": 1.5789291858673096,
+      "learning_rate": 0.00013662924851327247,
+      "loss": 1.7312,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3993687532877433,
+      "grad_norm": 1.9894529581069946,
+      "learning_rate": 0.00013659753740311866,
+      "loss": 1.8041,
+      "step": 3796
+    },
+    {
+      "epoch": 0.3994739610731194,
+      "grad_norm": 1.1728602647781372,
+      "learning_rate": 0.00013656582204301334,
+      "loss": 2.1131,
+      "step": 3797
+    },
+    {
+      "epoch": 0.39957916885849554,
+      "grad_norm": 2.1701881885528564,
+      "learning_rate": 0.00013653410243663952,
+      "loss": 2.1482,
+      "step": 3798
+    },
+    {
+      "epoch": 0.39968437664387163,
+      "grad_norm": 1.0504519939422607,
+      "learning_rate": 0.00013650237858768067,
+      "loss": 2.0582,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3997895844292478,
+      "grad_norm": 2.3325071334838867,
+      "learning_rate": 0.00013647065049982078,
+      "loss": 1.565,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3998947922146239,
+      "grad_norm": 1.386509895324707,
+      "learning_rate": 0.0001364389181767444,
+      "loss": 1.8805,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.0210331678390503,
+      "learning_rate": 0.0001364071816221364,
+      "loss": 2.0216,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4001052077853761,
+      "grad_norm": 1.7572848796844482,
+      "learning_rate": 0.00013637544083968227,
+      "loss": 2.1344,
+      "step": 3803
+    },
+    {
+      "epoch": 0.4002104155707522,
+      "grad_norm": 1.721763014793396,
+      "learning_rate": 0.00013634369583306798,
+      "loss": 1.8024,
+      "step": 3804
+    },
+    {
+      "epoch": 0.40031562335612836,
+      "grad_norm": 1.5794200897216797,
+      "learning_rate": 0.00013631194660598,
+      "loss": 1.9257,
+      "step": 3805
+    },
+    {
+      "epoch": 0.40042083114150445,
+      "grad_norm": 1.026024341583252,
+      "learning_rate": 0.00013628019316210522,
+      "loss": 2.4013,
+      "step": 3806
+    },
+    {
+      "epoch": 0.4005260389268806,
+      "grad_norm": 4.058107852935791,
+      "learning_rate": 0.0001362484355051311,
+      "loss": 1.554,
+      "step": 3807
+    },
+    {
+      "epoch": 0.4006312467122567,
+      "grad_norm": 1.9049569368362427,
+      "learning_rate": 0.00013621667363874552,
+      "loss": 2.2071,
+      "step": 3808
+    },
+    {
+      "epoch": 0.40073645449763284,
+      "grad_norm": 1.2772477865219116,
+      "learning_rate": 0.00013618490756663686,
+      "loss": 1.6058,
+      "step": 3809
+    },
+    {
+      "epoch": 0.40084166228300894,
+      "grad_norm": 1.0035552978515625,
+      "learning_rate": 0.00013615313729249405,
+      "loss": 1.8947,
+      "step": 3810
+    },
+    {
+      "epoch": 0.4009468700683851,
+      "grad_norm": 0.8478926420211792,
+      "learning_rate": 0.00013612136282000644,
+      "loss": 1.8861,
+      "step": 3811
+    },
+    {
+      "epoch": 0.4010520778537612,
+      "grad_norm": 1.2318350076675415,
+      "learning_rate": 0.00013608958415286396,
+      "loss": 1.8174,
+      "step": 3812
+    },
+    {
+      "epoch": 0.4011572856391373,
+      "grad_norm": 1.1952483654022217,
+      "learning_rate": 0.00013605780129475687,
+      "loss": 1.8322,
+      "step": 3813
+    },
+    {
+      "epoch": 0.4012624934245134,
+      "grad_norm": 1.1804511547088623,
+      "learning_rate": 0.00013602601424937604,
+      "loss": 1.7156,
+      "step": 3814
+    },
+    {
+      "epoch": 0.4013677012098895,
+      "grad_norm": 1.3360835313796997,
+      "learning_rate": 0.00013599422302041286,
+      "loss": 1.7492,
+      "step": 3815
+    },
+    {
+      "epoch": 0.40147290899526566,
+      "grad_norm": 1.1277227401733398,
+      "learning_rate": 0.00013596242761155903,
+      "loss": 1.4064,
+      "step": 3816
+    },
+    {
+      "epoch": 0.40157811678064176,
+      "grad_norm": 1.0985256433486938,
+      "learning_rate": 0.00013593062802650692,
+      "loss": 1.8555,
+      "step": 3817
+    },
+    {
+      "epoch": 0.4016833245660179,
+      "grad_norm": 1.0841742753982544,
+      "learning_rate": 0.0001358988242689493,
+      "loss": 1.4974,
+      "step": 3818
+    },
+    {
+      "epoch": 0.401788532351394,
+      "grad_norm": 1.3948936462402344,
+      "learning_rate": 0.0001358670163425795,
+      "loss": 1.8494,
+      "step": 3819
+    },
+    {
+      "epoch": 0.4018937401367701,
+      "grad_norm": 2.593919515609741,
+      "learning_rate": 0.0001358352042510911,
+      "loss": 1.9506,
+      "step": 3820
+    },
+    {
+      "epoch": 0.40199894792214624,
+      "grad_norm": 1.0042921304702759,
+      "learning_rate": 0.00013580338799817844,
+      "loss": 2.05,
+      "step": 3821
+    },
+    {
+      "epoch": 0.40210415570752234,
+      "grad_norm": 1.3841592073440552,
+      "learning_rate": 0.00013577156758753627,
+      "loss": 1.8187,
+      "step": 3822
+    },
+    {
+      "epoch": 0.4022093634928985,
+      "grad_norm": 2.3145501613616943,
+      "learning_rate": 0.00013573974302285972,
+      "loss": 2.1793,
+      "step": 3823
+    },
+    {
+      "epoch": 0.4023145712782746,
+      "grad_norm": 1.6691625118255615,
+      "learning_rate": 0.00013570791430784452,
+      "loss": 1.8489,
+      "step": 3824
+    },
+    {
+      "epoch": 0.40241977906365073,
+      "grad_norm": 1.7601234912872314,
+      "learning_rate": 0.0001356760814461868,
+      "loss": 1.055,
+      "step": 3825
+    },
+    {
+      "epoch": 0.4025249868490268,
+      "grad_norm": 1.5484458208084106,
+      "learning_rate": 0.00013564424444158324,
+      "loss": 1.8766,
+      "step": 3826
+    },
+    {
+      "epoch": 0.40263019463440297,
+      "grad_norm": 1.6130404472351074,
+      "learning_rate": 0.00013561240329773092,
+      "loss": 1.8005,
+      "step": 3827
+    },
+    {
+      "epoch": 0.40273540241977906,
+      "grad_norm": 1.7206354141235352,
+      "learning_rate": 0.00013558055801832748,
+      "loss": 1.5215,
+      "step": 3828
+    },
+    {
+      "epoch": 0.40284061020515516,
+      "grad_norm": 1.2879303693771362,
+      "learning_rate": 0.00013554870860707106,
+      "loss": 2.0524,
+      "step": 3829
+    },
+    {
+      "epoch": 0.4029458179905313,
+      "grad_norm": 0.978996217250824,
+      "learning_rate": 0.0001355168550676601,
+      "loss": 1.7266,
+      "step": 3830
+    },
+    {
+      "epoch": 0.4030510257759074,
+      "grad_norm": 1.726261019706726,
+      "learning_rate": 0.00013548499740379373,
+      "loss": 2.3547,
+      "step": 3831
+    },
+    {
+      "epoch": 0.40315623356128355,
+      "grad_norm": 1.3626611232757568,
+      "learning_rate": 0.00013545313561917144,
+      "loss": 1.5038,
+      "step": 3832
+    },
+    {
+      "epoch": 0.40326144134665964,
+      "grad_norm": 1.0368597507476807,
+      "learning_rate": 0.00013542126971749328,
+      "loss": 1.8802,
+      "step": 3833
+    },
+    {
+      "epoch": 0.4033666491320358,
+      "grad_norm": 1.394249439239502,
+      "learning_rate": 0.00013538939970245972,
+      "loss": 1.7687,
+      "step": 3834
+    },
+    {
+      "epoch": 0.4034718569174119,
+      "grad_norm": 1.244553565979004,
+      "learning_rate": 0.0001353575255777717,
+      "loss": 1.9525,
+      "step": 3835
+    },
+    {
+      "epoch": 0.403577064702788,
+      "grad_norm": 1.686198353767395,
+      "learning_rate": 0.00013532564734713068,
+      "loss": 1.6175,
+      "step": 3836
+    },
+    {
+      "epoch": 0.40368227248816413,
+      "grad_norm": 2.307166814804077,
+      "learning_rate": 0.00013529376501423852,
+      "loss": 2.3844,
+      "step": 3837
+    },
+    {
+      "epoch": 0.4037874802735402,
+      "grad_norm": 1.9087625741958618,
+      "learning_rate": 0.00013526187858279765,
+      "loss": 1.272,
+      "step": 3838
+    },
+    {
+      "epoch": 0.40389268805891637,
+      "grad_norm": 1.4016311168670654,
+      "learning_rate": 0.00013522998805651096,
+      "loss": 1.9822,
+      "step": 3839
+    },
+    {
+      "epoch": 0.40399789584429247,
+      "grad_norm": 1.2222285270690918,
+      "learning_rate": 0.00013519809343908178,
+      "loss": 2.1131,
+      "step": 3840
+    },
+    {
+      "epoch": 0.4041031036296686,
+      "grad_norm": 1.7945427894592285,
+      "learning_rate": 0.00013516619473421387,
+      "loss": 1.8994,
+      "step": 3841
+    },
+    {
+      "epoch": 0.4042083114150447,
+      "grad_norm": 1.665122628211975,
+      "learning_rate": 0.0001351342919456116,
+      "loss": 1.7643,
+      "step": 3842
+    },
+    {
+      "epoch": 0.40431351920042086,
+      "grad_norm": 1.1571388244628906,
+      "learning_rate": 0.00013510238507697967,
+      "loss": 2.1814,
+      "step": 3843
+    },
+    {
+      "epoch": 0.40441872698579695,
+      "grad_norm": 1.2247772216796875,
+      "learning_rate": 0.00013507047413202335,
+      "loss": 1.9968,
+      "step": 3844
+    },
+    {
+      "epoch": 0.40452393477117304,
+      "grad_norm": 1.482480525970459,
+      "learning_rate": 0.00013503855911444837,
+      "loss": 1.9141,
+      "step": 3845
+    },
+    {
+      "epoch": 0.4046291425565492,
+      "grad_norm": 1.5328335762023926,
+      "learning_rate": 0.00013500664002796093,
+      "loss": 1.8682,
+      "step": 3846
+    },
+    {
+      "epoch": 0.4047343503419253,
+      "grad_norm": 1.9800300598144531,
+      "learning_rate": 0.0001349747168762676,
+      "loss": 1.7248,
+      "step": 3847
+    },
+    {
+      "epoch": 0.40483955812730144,
+      "grad_norm": 1.3006877899169922,
+      "learning_rate": 0.0001349427896630756,
+      "loss": 1.5546,
+      "step": 3848
+    },
+    {
+      "epoch": 0.40494476591267753,
+      "grad_norm": 1.683077096939087,
+      "learning_rate": 0.0001349108583920925,
+      "loss": 2.1227,
+      "step": 3849
+    },
+    {
+      "epoch": 0.4050499736980537,
+      "grad_norm": 1.1028867959976196,
+      "learning_rate": 0.00013487892306702638,
+      "loss": 1.8122,
+      "step": 3850
+    },
+    {
+      "epoch": 0.40515518148342977,
+      "grad_norm": 1.3588521480560303,
+      "learning_rate": 0.00013484698369158578,
+      "loss": 1.7322,
+      "step": 3851
+    },
+    {
+      "epoch": 0.40526038926880587,
+      "grad_norm": 1.1769453287124634,
+      "learning_rate": 0.0001348150402694797,
+      "loss": 2.0872,
+      "step": 3852
+    },
+    {
+      "epoch": 0.405365597054182,
+      "grad_norm": 1.4604272842407227,
+      "learning_rate": 0.00013478309280441763,
+      "loss": 2.044,
+      "step": 3853
+    },
+    {
+      "epoch": 0.4054708048395581,
+      "grad_norm": 1.201572299003601,
+      "learning_rate": 0.00013475114130010954,
+      "loss": 2.048,
+      "step": 3854
+    },
+    {
+      "epoch": 0.40557601262493426,
+      "grad_norm": 2.324699878692627,
+      "learning_rate": 0.00013471918576026583,
+      "loss": 2.0551,
+      "step": 3855
+    },
+    {
+      "epoch": 0.40568122041031035,
+      "grad_norm": 1.0302995443344116,
+      "learning_rate": 0.00013468722618859743,
+      "loss": 1.7291,
+      "step": 3856
+    },
+    {
+      "epoch": 0.4057864281956865,
+      "grad_norm": 1.6787930727005005,
+      "learning_rate": 0.00013465526258881565,
+      "loss": 2.4362,
+      "step": 3857
+    },
+    {
+      "epoch": 0.4058916359810626,
+      "grad_norm": 1.2186408042907715,
+      "learning_rate": 0.00013462329496463236,
+      "loss": 1.8517,
+      "step": 3858
+    },
+    {
+      "epoch": 0.40599684376643874,
+      "grad_norm": 0.8594940900802612,
+      "learning_rate": 0.0001345913233197598,
+      "loss": 1.8623,
+      "step": 3859
+    },
+    {
+      "epoch": 0.40610205155181484,
+      "grad_norm": 1.7866066694259644,
+      "learning_rate": 0.00013455934765791084,
+      "loss": 1.7403,
+      "step": 3860
+    },
+    {
+      "epoch": 0.40620725933719093,
+      "grad_norm": 1.9410916566848755,
+      "learning_rate": 0.00013452736798279856,
+      "loss": 1.7438,
+      "step": 3861
+    },
+    {
+      "epoch": 0.4063124671225671,
+      "grad_norm": 0.9494317173957825,
+      "learning_rate": 0.0001344953842981368,
+      "loss": 1.8397,
+      "step": 3862
+    },
+    {
+      "epoch": 0.40641767490794317,
+      "grad_norm": 1.590301752090454,
+      "learning_rate": 0.0001344633966076396,
+      "loss": 2.0589,
+      "step": 3863
+    },
+    {
+      "epoch": 0.4065228826933193,
+      "grad_norm": 1.0571048259735107,
+      "learning_rate": 0.0001344314049150217,
+      "loss": 2.1632,
+      "step": 3864
+    },
+    {
+      "epoch": 0.4066280904786954,
+      "grad_norm": 1.2266767024993896,
+      "learning_rate": 0.00013439940922399806,
+      "loss": 2.1405,
+      "step": 3865
+    },
+    {
+      "epoch": 0.40673329826407156,
+      "grad_norm": 1.6102620363235474,
+      "learning_rate": 0.00013436740953828432,
+      "loss": 2.0555,
+      "step": 3866
+    },
+    {
+      "epoch": 0.40683850604944766,
+      "grad_norm": 1.640577793121338,
+      "learning_rate": 0.0001343354058615965,
+      "loss": 1.3221,
+      "step": 3867
+    },
+    {
+      "epoch": 0.40694371383482375,
+      "grad_norm": 1.5055091381072998,
+      "learning_rate": 0.00013430339819765105,
+      "loss": 1.9195,
+      "step": 3868
+    },
+    {
+      "epoch": 0.4070489216201999,
+      "grad_norm": 1.0952521562576294,
+      "learning_rate": 0.0001342713865501649,
+      "loss": 1.8598,
+      "step": 3869
+    },
+    {
+      "epoch": 0.407154129405576,
+      "grad_norm": 0.8317450881004333,
+      "learning_rate": 0.00013423937092285555,
+      "loss": 1.9036,
+      "step": 3870
+    },
+    {
+      "epoch": 0.40725933719095214,
+      "grad_norm": 1.3598711490631104,
+      "learning_rate": 0.00013420735131944073,
+      "loss": 1.9391,
+      "step": 3871
+    },
+    {
+      "epoch": 0.40736454497632824,
+      "grad_norm": 1.854884147644043,
+      "learning_rate": 0.0001341753277436389,
+      "loss": 1.5506,
+      "step": 3872
+    },
+    {
+      "epoch": 0.4074697527617044,
+      "grad_norm": 1.2758795022964478,
+      "learning_rate": 0.00013414330019916875,
+      "loss": 1.6345,
+      "step": 3873
+    },
+    {
+      "epoch": 0.4075749605470805,
+      "grad_norm": 1.4134248495101929,
+      "learning_rate": 0.0001341112686897496,
+      "loss": 1.3446,
+      "step": 3874
+    },
+    {
+      "epoch": 0.4076801683324566,
+      "grad_norm": 1.3641563653945923,
+      "learning_rate": 0.00013407923321910115,
+      "loss": 2.0881,
+      "step": 3875
+    },
+    {
+      "epoch": 0.4077853761178327,
+      "grad_norm": 1.287705659866333,
+      "learning_rate": 0.00013404719379094354,
+      "loss": 2.2842,
+      "step": 3876
+    },
+    {
+      "epoch": 0.4078905839032088,
+      "grad_norm": 1.201897382736206,
+      "learning_rate": 0.00013401515040899746,
+      "loss": 1.7941,
+      "step": 3877
+    },
+    {
+      "epoch": 0.40799579168858496,
+      "grad_norm": 1.092402696609497,
+      "learning_rate": 0.00013398310307698397,
+      "loss": 1.8657,
+      "step": 3878
+    },
+    {
+      "epoch": 0.40810099947396106,
+      "grad_norm": 1.349609136581421,
+      "learning_rate": 0.0001339510517986246,
+      "loss": 1.6198,
+      "step": 3879
+    },
+    {
+      "epoch": 0.4082062072593372,
+      "grad_norm": 1.9795316457748413,
+      "learning_rate": 0.0001339189965776414,
+      "loss": 1.7579,
+      "step": 3880
+    },
+    {
+      "epoch": 0.4083114150447133,
+      "grad_norm": 1.3774399757385254,
+      "learning_rate": 0.0001338869374177568,
+      "loss": 1.903,
+      "step": 3881
+    },
+    {
+      "epoch": 0.40841662283008945,
+      "grad_norm": 1.0936564207077026,
+      "learning_rate": 0.00013385487432269376,
+      "loss": 2.2156,
+      "step": 3882
+    },
+    {
+      "epoch": 0.40852183061546554,
+      "grad_norm": 1.4462928771972656,
+      "learning_rate": 0.00013382280729617568,
+      "loss": 2.0977,
+      "step": 3883
+    },
+    {
+      "epoch": 0.40862703840084164,
+      "grad_norm": 2.339925527572632,
+      "learning_rate": 0.00013379073634192632,
+      "loss": 1.6663,
+      "step": 3884
+    },
+    {
+      "epoch": 0.4087322461862178,
+      "grad_norm": 1.1641333103179932,
+      "learning_rate": 0.00013375866146367,
+      "loss": 2.5582,
+      "step": 3885
+    },
+    {
+      "epoch": 0.4088374539715939,
+      "grad_norm": 0.9723519086837769,
+      "learning_rate": 0.00013372658266513153,
+      "loss": 1.5445,
+      "step": 3886
+    },
+    {
+      "epoch": 0.40894266175697,
+      "grad_norm": 0.9060115218162537,
+      "learning_rate": 0.00013369449995003608,
+      "loss": 2.0697,
+      "step": 3887
+    },
+    {
+      "epoch": 0.4090478695423461,
+      "grad_norm": 1.7128041982650757,
+      "learning_rate": 0.00013366241332210928,
+      "loss": 1.4517,
+      "step": 3888
+    },
+    {
+      "epoch": 0.40915307732772227,
+      "grad_norm": 2.7365641593933105,
+      "learning_rate": 0.00013363032278507726,
+      "loss": 1.2883,
+      "step": 3889
+    },
+    {
+      "epoch": 0.40925828511309836,
+      "grad_norm": 1.3242186307907104,
+      "learning_rate": 0.00013359822834266662,
+      "loss": 2.3144,
+      "step": 3890
+    },
+    {
+      "epoch": 0.4093634928984745,
+      "grad_norm": 1.483139991760254,
+      "learning_rate": 0.00013356612999860436,
+      "loss": 1.9398,
+      "step": 3891
+    },
+    {
+      "epoch": 0.4094687006838506,
+      "grad_norm": 1.365196943283081,
+      "learning_rate": 0.00013353402775661795,
+      "loss": 1.5895,
+      "step": 3892
+    },
+    {
+      "epoch": 0.4095739084692267,
+      "grad_norm": 1.608028769493103,
+      "learning_rate": 0.0001335019216204353,
+      "loss": 1.9566,
+      "step": 3893
+    },
+    {
+      "epoch": 0.40967911625460285,
+      "grad_norm": 1.6372560262680054,
+      "learning_rate": 0.00013346981159378485,
+      "loss": 2.0881,
+      "step": 3894
+    },
+    {
+      "epoch": 0.40978432403997894,
+      "grad_norm": 1.121978521347046,
+      "learning_rate": 0.00013343769768039537,
+      "loss": 1.9097,
+      "step": 3895
+    },
+    {
+      "epoch": 0.4098895318253551,
+      "grad_norm": 1.4127432107925415,
+      "learning_rate": 0.00013340557988399617,
+      "loss": 2.1085,
+      "step": 3896
+    },
+    {
+      "epoch": 0.4099947396107312,
+      "grad_norm": 1.146700143814087,
+      "learning_rate": 0.00013337345820831696,
+      "loss": 1.546,
+      "step": 3897
+    },
+    {
+      "epoch": 0.41009994739610733,
+      "grad_norm": 1.1912237405776978,
+      "learning_rate": 0.000133341332657088,
+      "loss": 1.4891,
+      "step": 3898
+    },
+    {
+      "epoch": 0.4102051551814834,
+      "grad_norm": 1.4994091987609863,
+      "learning_rate": 0.0001333092032340398,
+      "loss": 2.1801,
+      "step": 3899
+    },
+    {
+      "epoch": 0.4103103629668595,
+      "grad_norm": 1.059524655342102,
+      "learning_rate": 0.00013327706994290355,
+      "loss": 1.9326,
+      "step": 3900
+    },
+    {
+      "epoch": 0.41041557075223567,
+      "grad_norm": 1.365159034729004,
+      "learning_rate": 0.00013324493278741073,
+      "loss": 1.833,
+      "step": 3901
+    },
+    {
+      "epoch": 0.41052077853761176,
+      "grad_norm": 2.979818344116211,
+      "learning_rate": 0.00013321279177129337,
+      "loss": 2.8629,
+      "step": 3902
+    },
+    {
+      "epoch": 0.4106259863229879,
+      "grad_norm": 2.018228530883789,
+      "learning_rate": 0.00013318064689828385,
+      "loss": 1.9213,
+      "step": 3903
+    },
+    {
+      "epoch": 0.410731194108364,
+      "grad_norm": 1.5214287042617798,
+      "learning_rate": 0.00013314849817211508,
+      "loss": 1.7545,
+      "step": 3904
+    },
+    {
+      "epoch": 0.41083640189374016,
+      "grad_norm": 1.0850343704223633,
+      "learning_rate": 0.00013311634559652036,
+      "loss": 1.5041,
+      "step": 3905
+    },
+    {
+      "epoch": 0.41094160967911625,
+      "grad_norm": 1.1685420274734497,
+      "learning_rate": 0.00013308418917523348,
+      "loss": 1.9563,
+      "step": 3906
+    },
+    {
+      "epoch": 0.4110468174644924,
+      "grad_norm": 1.6198590993881226,
+      "learning_rate": 0.00013305202891198862,
+      "loss": 1.7929,
+      "step": 3907
+    },
+    {
+      "epoch": 0.4111520252498685,
+      "grad_norm": 2.0444579124450684,
+      "learning_rate": 0.0001330198648105205,
+      "loss": 2.1797,
+      "step": 3908
+    },
+    {
+      "epoch": 0.4112572330352446,
+      "grad_norm": 1.8956327438354492,
+      "learning_rate": 0.00013298769687456426,
+      "loss": 2.1496,
+      "step": 3909
+    },
+    {
+      "epoch": 0.41136244082062073,
+      "grad_norm": 1.3414783477783203,
+      "learning_rate": 0.00013295552510785534,
+      "loss": 1.4914,
+      "step": 3910
+    },
+    {
+      "epoch": 0.4114676486059968,
+      "grad_norm": 1.368857502937317,
+      "learning_rate": 0.00013292334951412984,
+      "loss": 1.7118,
+      "step": 3911
+    },
+    {
+      "epoch": 0.411572856391373,
+      "grad_norm": 1.7612696886062622,
+      "learning_rate": 0.00013289117009712418,
+      "loss": 2.0468,
+      "step": 3912
+    },
+    {
+      "epoch": 0.41167806417674907,
+      "grad_norm": 1.1317222118377686,
+      "learning_rate": 0.00013285898686057524,
+      "loss": 1.9907,
+      "step": 3913
+    },
+    {
+      "epoch": 0.4117832719621252,
+      "grad_norm": 1.8835352659225464,
+      "learning_rate": 0.00013282679980822034,
+      "loss": 1.4493,
+      "step": 3914
+    },
+    {
+      "epoch": 0.4118884797475013,
+      "grad_norm": 1.1284127235412598,
+      "learning_rate": 0.00013279460894379729,
+      "loss": 1.5342,
+      "step": 3915
+    },
+    {
+      "epoch": 0.4119936875328774,
+      "grad_norm": 1.2283440828323364,
+      "learning_rate": 0.0001327624142710443,
+      "loss": 2.1995,
+      "step": 3916
+    },
+    {
+      "epoch": 0.41209889531825356,
+      "grad_norm": 1.2161046266555786,
+      "learning_rate": 0.00013273021579370003,
+      "loss": 2.1285,
+      "step": 3917
+    },
+    {
+      "epoch": 0.41220410310362965,
+      "grad_norm": 1.6340278387069702,
+      "learning_rate": 0.00013269801351550354,
+      "loss": 1.4938,
+      "step": 3918
+    },
+    {
+      "epoch": 0.4123093108890058,
+      "grad_norm": 1.4137933254241943,
+      "learning_rate": 0.00013266580744019445,
+      "loss": 1.6995,
+      "step": 3919
+    },
+    {
+      "epoch": 0.4124145186743819,
+      "grad_norm": 1.0340012311935425,
+      "learning_rate": 0.0001326335975715127,
+      "loss": 1.9991,
+      "step": 3920
+    },
+    {
+      "epoch": 0.41251972645975804,
+      "grad_norm": 1.2276008129119873,
+      "learning_rate": 0.00013260138391319872,
+      "loss": 1.3765,
+      "step": 3921
+    },
+    {
+      "epoch": 0.41262493424513413,
+      "grad_norm": 1.1833198070526123,
+      "learning_rate": 0.00013256916646899337,
+      "loss": 2.0268,
+      "step": 3922
+    },
+    {
+      "epoch": 0.4127301420305103,
+      "grad_norm": 2.14572811126709,
+      "learning_rate": 0.00013253694524263798,
+      "loss": 2.0048,
+      "step": 3923
+    },
+    {
+      "epoch": 0.4128353498158864,
+      "grad_norm": 1.3830515146255493,
+      "learning_rate": 0.00013250472023787425,
+      "loss": 2.113,
+      "step": 3924
+    },
+    {
+      "epoch": 0.41294055760126247,
+      "grad_norm": 1.290831208229065,
+      "learning_rate": 0.00013247249145844443,
+      "loss": 1.6246,
+      "step": 3925
+    },
+    {
+      "epoch": 0.4130457653866386,
+      "grad_norm": 1.4509975910186768,
+      "learning_rate": 0.00013244025890809112,
+      "loss": 1.8658,
+      "step": 3926
+    },
+    {
+      "epoch": 0.4131509731720147,
+      "grad_norm": 1.2265394926071167,
+      "learning_rate": 0.00013240802259055734,
+      "loss": 1.763,
+      "step": 3927
+    },
+    {
+      "epoch": 0.41325618095739086,
+      "grad_norm": 1.2114838361740112,
+      "learning_rate": 0.0001323757825095866,
+      "loss": 1.7796,
+      "step": 3928
+    },
+    {
+      "epoch": 0.41336138874276696,
+      "grad_norm": 1.1204650402069092,
+      "learning_rate": 0.00013234353866892285,
+      "loss": 1.5069,
+      "step": 3929
+    },
+    {
+      "epoch": 0.4134665965281431,
+      "grad_norm": 1.245553970336914,
+      "learning_rate": 0.00013231129107231052,
+      "loss": 1.9431,
+      "step": 3930
+    },
+    {
+      "epoch": 0.4135718043135192,
+      "grad_norm": 1.2581690549850464,
+      "learning_rate": 0.00013227903972349428,
+      "loss": 1.7897,
+      "step": 3931
+    },
+    {
+      "epoch": 0.4136770120988953,
+      "grad_norm": 1.647939920425415,
+      "learning_rate": 0.00013224678462621947,
+      "loss": 2.0146,
+      "step": 3932
+    },
+    {
+      "epoch": 0.41378221988427144,
+      "grad_norm": 1.1378495693206787,
+      "learning_rate": 0.00013221452578423176,
+      "loss": 1.7066,
+      "step": 3933
+    },
+    {
+      "epoch": 0.41388742766964753,
+      "grad_norm": 1.322014570236206,
+      "learning_rate": 0.00013218226320127724,
+      "loss": 1.4162,
+      "step": 3934
+    },
+    {
+      "epoch": 0.4139926354550237,
+      "grad_norm": 0.965604841709137,
+      "learning_rate": 0.00013214999688110249,
+      "loss": 2.0709,
+      "step": 3935
+    },
+    {
+      "epoch": 0.4140978432403998,
+      "grad_norm": 1.4124212265014648,
+      "learning_rate": 0.00013211772682745446,
+      "loss": 2.03,
+      "step": 3936
+    },
+    {
+      "epoch": 0.4142030510257759,
+      "grad_norm": 1.284655213356018,
+      "learning_rate": 0.00013208545304408057,
+      "loss": 1.9079,
+      "step": 3937
+    },
+    {
+      "epoch": 0.414308258811152,
+      "grad_norm": 1.4193320274353027,
+      "learning_rate": 0.00013205317553472868,
+      "loss": 1.6901,
+      "step": 3938
+    },
+    {
+      "epoch": 0.41441346659652817,
+      "grad_norm": 1.1792535781860352,
+      "learning_rate": 0.00013202089430314705,
+      "loss": 1.5996,
+      "step": 3939
+    },
+    {
+      "epoch": 0.41451867438190426,
+      "grad_norm": 1.242659568786621,
+      "learning_rate": 0.00013198860935308444,
+      "loss": 2.35,
+      "step": 3940
+    },
+    {
+      "epoch": 0.41462388216728036,
+      "grad_norm": 1.4708364009857178,
+      "learning_rate": 0.0001319563206882899,
+      "loss": 1.83,
+      "step": 3941
+    },
+    {
+      "epoch": 0.4147290899526565,
+      "grad_norm": 1.332457184791565,
+      "learning_rate": 0.00013192402831251312,
+      "loss": 2.0127,
+      "step": 3942
+    },
+    {
+      "epoch": 0.4148342977380326,
+      "grad_norm": 1.8387818336486816,
+      "learning_rate": 0.00013189173222950403,
+      "loss": 2.4401,
+      "step": 3943
+    },
+    {
+      "epoch": 0.41493950552340875,
+      "grad_norm": 1.1250078678131104,
+      "learning_rate": 0.0001318594324430131,
+      "loss": 2.0481,
+      "step": 3944
+    },
+    {
+      "epoch": 0.41504471330878484,
+      "grad_norm": 1.0061681270599365,
+      "learning_rate": 0.00013182712895679118,
+      "loss": 1.6706,
+      "step": 3945
+    },
+    {
+      "epoch": 0.415149921094161,
+      "grad_norm": 1.2926052808761597,
+      "learning_rate": 0.0001317948217745896,
+      "loss": 1.596,
+      "step": 3946
+    },
+    {
+      "epoch": 0.4152551288795371,
+      "grad_norm": 1.2874207496643066,
+      "learning_rate": 0.00013176251090016007,
+      "loss": 1.7986,
+      "step": 3947
+    },
+    {
+      "epoch": 0.4153603366649132,
+      "grad_norm": 1.1514983177185059,
+      "learning_rate": 0.00013173019633725474,
+      "loss": 2.1873,
+      "step": 3948
+    },
+    {
+      "epoch": 0.4154655444502893,
+      "grad_norm": 1.636364459991455,
+      "learning_rate": 0.00013169787808962617,
+      "loss": 1.9381,
+      "step": 3949
+    },
+    {
+      "epoch": 0.4155707522356654,
+      "grad_norm": 1.7293845415115356,
+      "learning_rate": 0.00013166555616102744,
+      "loss": 1.8849,
+      "step": 3950
+    },
+    {
+      "epoch": 0.41567596002104157,
+      "grad_norm": 1.3633328676223755,
+      "learning_rate": 0.0001316332305552119,
+      "loss": 1.2241,
+      "step": 3951
+    },
+    {
+      "epoch": 0.41578116780641766,
+      "grad_norm": 0.8640151023864746,
+      "learning_rate": 0.00013160090127593344,
+      "loss": 1.9409,
+      "step": 3952
+    },
+    {
+      "epoch": 0.4158863755917938,
+      "grad_norm": 1.580431580543518,
+      "learning_rate": 0.00013156856832694642,
+      "loss": 1.7308,
+      "step": 3953
+    },
+    {
+      "epoch": 0.4159915833771699,
+      "grad_norm": 0.9848331212997437,
+      "learning_rate": 0.0001315362317120055,
+      "loss": 1.8526,
+      "step": 3954
+    },
+    {
+      "epoch": 0.41609679116254605,
+      "grad_norm": 1.7985378503799438,
+      "learning_rate": 0.00013150389143486586,
+      "loss": 1.7818,
+      "step": 3955
+    },
+    {
+      "epoch": 0.41620199894792215,
+      "grad_norm": 1.3943655490875244,
+      "learning_rate": 0.000131471547499283,
+      "loss": 1.8059,
+      "step": 3956
+    },
+    {
+      "epoch": 0.41630720673329824,
+      "grad_norm": 1.2380815744400024,
+      "learning_rate": 0.00013143919990901302,
+      "loss": 1.652,
+      "step": 3957
+    },
+    {
+      "epoch": 0.4164124145186744,
+      "grad_norm": 1.415740966796875,
+      "learning_rate": 0.00013140684866781225,
+      "loss": 2.4166,
+      "step": 3958
+    },
+    {
+      "epoch": 0.4165176223040505,
+      "grad_norm": 1.2207032442092896,
+      "learning_rate": 0.00013137449377943755,
+      "loss": 1.7993,
+      "step": 3959
+    },
+    {
+      "epoch": 0.41662283008942663,
+      "grad_norm": 1.1489773988723755,
+      "learning_rate": 0.00013134213524764623,
+      "loss": 1.7678,
+      "step": 3960
+    },
+    {
+      "epoch": 0.4167280378748027,
+      "grad_norm": 1.381227970123291,
+      "learning_rate": 0.00013130977307619594,
+      "loss": 2.0441,
+      "step": 3961
+    },
+    {
+      "epoch": 0.4168332456601789,
+      "grad_norm": 1.6043063402175903,
+      "learning_rate": 0.0001312774072688448,
+      "loss": 1.9146,
+      "step": 3962
+    },
+    {
+      "epoch": 0.41693845344555497,
+      "grad_norm": 1.3174347877502441,
+      "learning_rate": 0.00013124503782935133,
+      "loss": 1.7681,
+      "step": 3963
+    },
+    {
+      "epoch": 0.41704366123093106,
+      "grad_norm": 1.854496955871582,
+      "learning_rate": 0.00013121266476147454,
+      "loss": 2.158,
+      "step": 3964
+    },
+    {
+      "epoch": 0.4171488690163072,
+      "grad_norm": 1.0530141592025757,
+      "learning_rate": 0.00013118028806897373,
+      "loss": 1.8426,
+      "step": 3965
+    },
+    {
+      "epoch": 0.4172540768016833,
+      "grad_norm": 1.5964198112487793,
+      "learning_rate": 0.00013114790775560877,
+      "loss": 1.3975,
+      "step": 3966
+    },
+    {
+      "epoch": 0.41735928458705945,
+      "grad_norm": 1.1978175640106201,
+      "learning_rate": 0.00013111552382513985,
+      "loss": 2.0528,
+      "step": 3967
+    },
+    {
+      "epoch": 0.41746449237243555,
+      "grad_norm": 1.6343151330947876,
+      "learning_rate": 0.0001310831362813276,
+      "loss": 1.9448,
+      "step": 3968
+    },
+    {
+      "epoch": 0.4175697001578117,
+      "grad_norm": 1.598139762878418,
+      "learning_rate": 0.0001310507451279331,
+      "loss": 1.6996,
+      "step": 3969
+    },
+    {
+      "epoch": 0.4176749079431878,
+      "grad_norm": 0.9592291712760925,
+      "learning_rate": 0.00013101835036871781,
+      "loss": 1.1519,
+      "step": 3970
+    },
+    {
+      "epoch": 0.41778011572856394,
+      "grad_norm": 2.241469621658325,
+      "learning_rate": 0.00013098595200744366,
+      "loss": 1.9488,
+      "step": 3971
+    },
+    {
+      "epoch": 0.41788532351394003,
+      "grad_norm": 2.4150736331939697,
+      "learning_rate": 0.0001309535500478729,
+      "loss": 2.2178,
+      "step": 3972
+    },
+    {
+      "epoch": 0.4179905312993161,
+      "grad_norm": 1.3556127548217773,
+      "learning_rate": 0.00013092114449376828,
+      "loss": 1.5269,
+      "step": 3973
+    },
+    {
+      "epoch": 0.4180957390846923,
+      "grad_norm": 1.4668810367584229,
+      "learning_rate": 0.00013088873534889304,
+      "loss": 2.0874,
+      "step": 3974
+    },
+    {
+      "epoch": 0.41820094687006837,
+      "grad_norm": 1.5009702444076538,
+      "learning_rate": 0.00013085632261701063,
+      "loss": 1.7267,
+      "step": 3975
+    },
+    {
+      "epoch": 0.4183061546554445,
+      "grad_norm": 1.6611928939819336,
+      "learning_rate": 0.0001308239063018851,
+      "loss": 1.5624,
+      "step": 3976
+    },
+    {
+      "epoch": 0.4184113624408206,
+      "grad_norm": 1.1202343702316284,
+      "learning_rate": 0.00013079148640728077,
+      "loss": 1.78,
+      "step": 3977
+    },
+    {
+      "epoch": 0.41851657022619676,
+      "grad_norm": 1.4417529106140137,
+      "learning_rate": 0.0001307590629369626,
+      "loss": 2.0658,
+      "step": 3978
+    },
+    {
+      "epoch": 0.41862177801157285,
+      "grad_norm": 1.2309139966964722,
+      "learning_rate": 0.0001307266358946957,
+      "loss": 1.5864,
+      "step": 3979
+    },
+    {
+      "epoch": 0.41872698579694895,
+      "grad_norm": 1.4191454648971558,
+      "learning_rate": 0.00013069420528424579,
+      "loss": 1.9842,
+      "step": 3980
+    },
+    {
+      "epoch": 0.4188321935823251,
+      "grad_norm": 1.4356913566589355,
+      "learning_rate": 0.00013066177110937884,
+      "loss": 1.7109,
+      "step": 3981
+    },
+    {
+      "epoch": 0.4189374013677012,
+      "grad_norm": 2.5019426345825195,
+      "learning_rate": 0.00013062933337386142,
+      "loss": 1.8636,
+      "step": 3982
+    },
+    {
+      "epoch": 0.41904260915307734,
+      "grad_norm": 1.5507819652557373,
+      "learning_rate": 0.00013059689208146035,
+      "loss": 1.5129,
+      "step": 3983
+    },
+    {
+      "epoch": 0.41914781693845343,
+      "grad_norm": 1.7946397066116333,
+      "learning_rate": 0.00013056444723594297,
+      "loss": 1.8957,
+      "step": 3984
+    },
+    {
+      "epoch": 0.4192530247238296,
+      "grad_norm": 1.1665949821472168,
+      "learning_rate": 0.000130531998841077,
+      "loss": 1.995,
+      "step": 3985
+    },
+    {
+      "epoch": 0.4193582325092057,
+      "grad_norm": 1.3788167238235474,
+      "learning_rate": 0.00013049954690063048,
+      "loss": 1.5433,
+      "step": 3986
+    },
+    {
+      "epoch": 0.4194634402945818,
+      "grad_norm": 1.3582801818847656,
+      "learning_rate": 0.00013046709141837205,
+      "loss": 1.6558,
+      "step": 3987
+    },
+    {
+      "epoch": 0.4195686480799579,
+      "grad_norm": 1.2572566270828247,
+      "learning_rate": 0.00013043463239807064,
+      "loss": 1.9563,
+      "step": 3988
+    },
+    {
+      "epoch": 0.419673855865334,
+      "grad_norm": 1.366934061050415,
+      "learning_rate": 0.00013040216984349555,
+      "loss": 1.987,
+      "step": 3989
+    },
+    {
+      "epoch": 0.41977906365071016,
+      "grad_norm": 1.2102018594741821,
+      "learning_rate": 0.0001303697037584166,
+      "loss": 2.09,
+      "step": 3990
+    },
+    {
+      "epoch": 0.41988427143608625,
+      "grad_norm": 1.2199288606643677,
+      "learning_rate": 0.000130337234146604,
+      "loss": 2.0518,
+      "step": 3991
+    },
+    {
+      "epoch": 0.4199894792214624,
+      "grad_norm": 1.3983739614486694,
+      "learning_rate": 0.00013030476101182824,
+      "loss": 1.9292,
+      "step": 3992
+    },
+    {
+      "epoch": 0.4200946870068385,
+      "grad_norm": 1.5765870809555054,
+      "learning_rate": 0.0001302722843578604,
+      "loss": 1.7341,
+      "step": 3993
+    },
+    {
+      "epoch": 0.42019989479221465,
+      "grad_norm": 1.6028943061828613,
+      "learning_rate": 0.00013023980418847185,
+      "loss": 2.1453,
+      "step": 3994
+    },
+    {
+      "epoch": 0.42030510257759074,
+      "grad_norm": 1.2651803493499756,
+      "learning_rate": 0.00013020732050743442,
+      "loss": 1.869,
+      "step": 3995
+    },
+    {
+      "epoch": 0.42041031036296683,
+      "grad_norm": 1.994928240776062,
+      "learning_rate": 0.00013017483331852035,
+      "loss": 2.1251,
+      "step": 3996
+    },
+    {
+      "epoch": 0.420515518148343,
+      "grad_norm": 1.914730191230774,
+      "learning_rate": 0.00013014234262550222,
+      "loss": 1.464,
+      "step": 3997
+    },
+    {
+      "epoch": 0.4206207259337191,
+      "grad_norm": 1.306941032409668,
+      "learning_rate": 0.00013010984843215312,
+      "loss": 1.7878,
+      "step": 3998
+    },
+    {
+      "epoch": 0.4207259337190952,
+      "grad_norm": 2.770078659057617,
+      "learning_rate": 0.00013007735074224645,
+      "loss": 1.7626,
+      "step": 3999
+    },
+    {
+      "epoch": 0.4208311415044713,
+      "grad_norm": 1.6361559629440308,
+      "learning_rate": 0.0001300448495595561,
+      "loss": 1.6572,
+      "step": 4000
+    },
+    {
+      "epoch": 0.42093634928984747,
+      "grad_norm": 1.2924569845199585,
+      "learning_rate": 0.0001300123448878563,
+      "loss": 1.9668,
+      "step": 4001
+    },
+    {
+      "epoch": 0.42104155707522356,
+      "grad_norm": 1.4006869792938232,
+      "learning_rate": 0.00012997983673092173,
+      "loss": 1.384,
+      "step": 4002
+    },
+    {
+      "epoch": 0.4211467648605997,
+      "grad_norm": 1.6290597915649414,
+      "learning_rate": 0.00012994732509252744,
+      "loss": 2.2109,
+      "step": 4003
+    },
+    {
+      "epoch": 0.4212519726459758,
+      "grad_norm": 1.2416553497314453,
+      "learning_rate": 0.00012991480997644886,
+      "loss": 2.3437,
+      "step": 4004
+    },
+    {
+      "epoch": 0.4213571804313519,
+      "grad_norm": 1.4514647722244263,
+      "learning_rate": 0.00012988229138646192,
+      "loss": 1.9429,
+      "step": 4005
+    },
+    {
+      "epoch": 0.42146238821672805,
+      "grad_norm": 1.3162492513656616,
+      "learning_rate": 0.00012984976932634292,
+      "loss": 1.8274,
+      "step": 4006
+    },
+    {
+      "epoch": 0.42156759600210414,
+      "grad_norm": 1.3386199474334717,
+      "learning_rate": 0.00012981724379986846,
+      "loss": 1.7178,
+      "step": 4007
+    },
+    {
+      "epoch": 0.4216728037874803,
+      "grad_norm": 1.1760342121124268,
+      "learning_rate": 0.00012978471481081566,
+      "loss": 1.4665,
+      "step": 4008
+    },
+    {
+      "epoch": 0.4217780115728564,
+      "grad_norm": 1.1634572744369507,
+      "learning_rate": 0.00012975218236296204,
+      "loss": 1.805,
+      "step": 4009
+    },
+    {
+      "epoch": 0.42188321935823253,
+      "grad_norm": 1.3325061798095703,
+      "learning_rate": 0.00012971964646008542,
+      "loss": 2.1894,
+      "step": 4010
+    },
+    {
+      "epoch": 0.4219884271436086,
+      "grad_norm": 1.4473451375961304,
+      "learning_rate": 0.00012968710710596417,
+      "loss": 2.0688,
+      "step": 4011
+    },
+    {
+      "epoch": 0.4220936349289847,
+      "grad_norm": 1.3676494359970093,
+      "learning_rate": 0.0001296545643043769,
+      "loss": 2.0859,
+      "step": 4012
+    },
+    {
+      "epoch": 0.42219884271436087,
+      "grad_norm": 1.168908953666687,
+      "learning_rate": 0.00012962201805910274,
+      "loss": 1.6922,
+      "step": 4013
+    },
+    {
+      "epoch": 0.42230405049973696,
+      "grad_norm": 1.1409196853637695,
+      "learning_rate": 0.00012958946837392113,
+      "loss": 1.9286,
+      "step": 4014
+    },
+    {
+      "epoch": 0.4224092582851131,
+      "grad_norm": 1.3931876420974731,
+      "learning_rate": 0.00012955691525261203,
+      "loss": 1.5437,
+      "step": 4015
+    },
+    {
+      "epoch": 0.4225144660704892,
+      "grad_norm": 1.6057072877883911,
+      "learning_rate": 0.00012952435869895569,
+      "loss": 2.0498,
+      "step": 4016
+    },
+    {
+      "epoch": 0.42261967385586535,
+      "grad_norm": 2.05791974067688,
+      "learning_rate": 0.00012949179871673278,
+      "loss": 1.7363,
+      "step": 4017
+    },
+    {
+      "epoch": 0.42272488164124145,
+      "grad_norm": 1.1081055402755737,
+      "learning_rate": 0.00012945923530972438,
+      "loss": 2.24,
+      "step": 4018
+    },
+    {
+      "epoch": 0.4228300894266176,
+      "grad_norm": 1.165257215499878,
+      "learning_rate": 0.00012942666848171202,
+      "loss": 2.1699,
+      "step": 4019
+    },
+    {
+      "epoch": 0.4229352972119937,
+      "grad_norm": 2.037813901901245,
+      "learning_rate": 0.00012939409823647753,
+      "loss": 1.3196,
+      "step": 4020
+    },
+    {
+      "epoch": 0.4230405049973698,
+      "grad_norm": 1.1846753358840942,
+      "learning_rate": 0.00012936152457780322,
+      "loss": 1.6817,
+      "step": 4021
+    },
+    {
+      "epoch": 0.42314571278274593,
+      "grad_norm": 0.9720628261566162,
+      "learning_rate": 0.00012932894750947177,
+      "loss": 1.6067,
+      "step": 4022
+    },
+    {
+      "epoch": 0.423250920568122,
+      "grad_norm": 1.4518640041351318,
+      "learning_rate": 0.00012929636703526618,
+      "loss": 1.9264,
+      "step": 4023
+    },
+    {
+      "epoch": 0.4233561283534982,
+      "grad_norm": 1.9571900367736816,
+      "learning_rate": 0.00012926378315896998,
+      "loss": 2.0874,
+      "step": 4024
+    },
+    {
+      "epoch": 0.42346133613887427,
+      "grad_norm": 1.8690723180770874,
+      "learning_rate": 0.00012923119588436702,
+      "loss": 1.3981,
+      "step": 4025
+    },
+    {
+      "epoch": 0.4235665439242504,
+      "grad_norm": 0.9987134337425232,
+      "learning_rate": 0.0001291986052152415,
+      "loss": 2.2083,
+      "step": 4026
+    },
+    {
+      "epoch": 0.4236717517096265,
+      "grad_norm": 1.6420297622680664,
+      "learning_rate": 0.0001291660111553781,
+      "loss": 2.006,
+      "step": 4027
+    },
+    {
+      "epoch": 0.4237769594950026,
+      "grad_norm": 1.2390166521072388,
+      "learning_rate": 0.0001291334137085619,
+      "loss": 1.4468,
+      "step": 4028
+    },
+    {
+      "epoch": 0.42388216728037875,
+      "grad_norm": 1.2712844610214233,
+      "learning_rate": 0.00012910081287857827,
+      "loss": 2.3897,
+      "step": 4029
+    },
+    {
+      "epoch": 0.42398737506575485,
+      "grad_norm": 1.1292376518249512,
+      "learning_rate": 0.0001290682086692131,
+      "loss": 2.0741,
+      "step": 4030
+    },
+    {
+      "epoch": 0.424092582851131,
+      "grad_norm": 1.479219913482666,
+      "learning_rate": 0.00012903560108425258,
+      "loss": 1.9733,
+      "step": 4031
+    },
+    {
+      "epoch": 0.4241977906365071,
+      "grad_norm": 0.8995711207389832,
+      "learning_rate": 0.00012900299012748328,
+      "loss": 1.9121,
+      "step": 4032
+    },
+    {
+      "epoch": 0.42430299842188324,
+      "grad_norm": 1.6135600805282593,
+      "learning_rate": 0.00012897037580269225,
+      "loss": 1.7079,
+      "step": 4033
+    },
+    {
+      "epoch": 0.42440820620725933,
+      "grad_norm": 1.3254718780517578,
+      "learning_rate": 0.0001289377581136669,
+      "loss": 1.5347,
+      "step": 4034
+    },
+    {
+      "epoch": 0.4245134139926355,
+      "grad_norm": 1.225527286529541,
+      "learning_rate": 0.00012890513706419497,
+      "loss": 0.9884,
+      "step": 4035
+    },
+    {
+      "epoch": 0.4246186217780116,
+      "grad_norm": 1.1481863260269165,
+      "learning_rate": 0.00012887251265806466,
+      "loss": 1.9885,
+      "step": 4036
+    },
+    {
+      "epoch": 0.42472382956338767,
+      "grad_norm": 1.3110417127609253,
+      "learning_rate": 0.00012883988489906454,
+      "loss": 1.7185,
+      "step": 4037
+    },
+    {
+      "epoch": 0.4248290373487638,
+      "grad_norm": 1.7494990825653076,
+      "learning_rate": 0.00012880725379098352,
+      "loss": 1.8137,
+      "step": 4038
+    },
+    {
+      "epoch": 0.4249342451341399,
+      "grad_norm": 1.0107097625732422,
+      "learning_rate": 0.00012877461933761102,
+      "loss": 1.7371,
+      "step": 4039
+    },
+    {
+      "epoch": 0.42503945291951606,
+      "grad_norm": 1.2290560007095337,
+      "learning_rate": 0.00012874198154273672,
+      "loss": 2.0912,
+      "step": 4040
+    },
+    {
+      "epoch": 0.42514466070489215,
+      "grad_norm": 1.1773860454559326,
+      "learning_rate": 0.00012870934041015071,
+      "loss": 1.9689,
+      "step": 4041
+    },
+    {
+      "epoch": 0.4252498684902683,
+      "grad_norm": 1.1857486963272095,
+      "learning_rate": 0.00012867669594364357,
+      "loss": 1.6265,
+      "step": 4042
+    },
+    {
+      "epoch": 0.4253550762756444,
+      "grad_norm": 1.5043977499008179,
+      "learning_rate": 0.00012864404814700618,
+      "loss": 1.9106,
+      "step": 4043
+    },
+    {
+      "epoch": 0.4254602840610205,
+      "grad_norm": 1.6733112335205078,
+      "learning_rate": 0.00012861139702402977,
+      "loss": 1.4501,
+      "step": 4044
+    },
+    {
+      "epoch": 0.42556549184639664,
+      "grad_norm": 1.2191860675811768,
+      "learning_rate": 0.00012857874257850605,
+      "loss": 1.5926,
+      "step": 4045
+    },
+    {
+      "epoch": 0.42567069963177273,
+      "grad_norm": 1.3937748670578003,
+      "learning_rate": 0.00012854608481422707,
+      "loss": 1.2054,
+      "step": 4046
+    },
+    {
+      "epoch": 0.4257759074171489,
+      "grad_norm": 1.3559739589691162,
+      "learning_rate": 0.00012851342373498525,
+      "loss": 1.5141,
+      "step": 4047
+    },
+    {
+      "epoch": 0.425881115202525,
+      "grad_norm": 1.3075133562088013,
+      "learning_rate": 0.0001284807593445734,
+      "loss": 1.3719,
+      "step": 4048
+    },
+    {
+      "epoch": 0.4259863229879011,
+      "grad_norm": 1.1074446439743042,
+      "learning_rate": 0.00012844809164678478,
+      "loss": 1.8366,
+      "step": 4049
+    },
+    {
+      "epoch": 0.4260915307732772,
+      "grad_norm": 2.5052926540374756,
+      "learning_rate": 0.00012841542064541292,
+      "loss": 1.7895,
+      "step": 4050
+    },
+    {
+      "epoch": 0.42619673855865337,
+      "grad_norm": 1.4112364053726196,
+      "learning_rate": 0.00012838274634425188,
+      "loss": 1.6224,
+      "step": 4051
+    },
+    {
+      "epoch": 0.42630194634402946,
+      "grad_norm": 1.2035661935806274,
+      "learning_rate": 0.00012835006874709594,
+      "loss": 1.7851,
+      "step": 4052
+    },
+    {
+      "epoch": 0.42640715412940555,
+      "grad_norm": 1.3116462230682373,
+      "learning_rate": 0.00012831738785773985,
+      "loss": 1.5662,
+      "step": 4053
+    },
+    {
+      "epoch": 0.4265123619147817,
+      "grad_norm": 1.2714296579360962,
+      "learning_rate": 0.00012828470367997884,
+      "loss": 1.3814,
+      "step": 4054
+    },
+    {
+      "epoch": 0.4266175697001578,
+      "grad_norm": 2.2298810482025146,
+      "learning_rate": 0.00012825201621760826,
+      "loss": 1.8722,
+      "step": 4055
+    },
+    {
+      "epoch": 0.42672277748553394,
+      "grad_norm": 1.7081495523452759,
+      "learning_rate": 0.00012821932547442408,
+      "loss": 2.1527,
+      "step": 4056
+    },
+    {
+      "epoch": 0.42682798527091004,
+      "grad_norm": 1.4093936681747437,
+      "learning_rate": 0.00012818663145422256,
+      "loss": 2.1676,
+      "step": 4057
+    },
+    {
+      "epoch": 0.4269331930562862,
+      "grad_norm": 1.707614541053772,
+      "learning_rate": 0.00012815393416080035,
+      "loss": 1.9155,
+      "step": 4058
+    },
+    {
+      "epoch": 0.4270384008416623,
+      "grad_norm": 1.1860383749008179,
+      "learning_rate": 0.00012812123359795446,
+      "loss": 1.788,
+      "step": 4059
+    },
+    {
+      "epoch": 0.4271436086270384,
+      "grad_norm": 1.6916757822036743,
+      "learning_rate": 0.00012808852976948232,
+      "loss": 1.3056,
+      "step": 4060
+    },
+    {
+      "epoch": 0.4272488164124145,
+      "grad_norm": 1.7355988025665283,
+      "learning_rate": 0.00012805582267918172,
+      "loss": 2.3532,
+      "step": 4061
+    },
+    {
+      "epoch": 0.4273540241977906,
+      "grad_norm": 0.883721113204956,
+      "learning_rate": 0.00012802311233085082,
+      "loss": 2.0607,
+      "step": 4062
+    },
+    {
+      "epoch": 0.42745923198316677,
+      "grad_norm": 1.3145248889923096,
+      "learning_rate": 0.00012799039872828812,
+      "loss": 2.2087,
+      "step": 4063
+    },
+    {
+      "epoch": 0.42756443976854286,
+      "grad_norm": 1.7425910234451294,
+      "learning_rate": 0.00012795768187529263,
+      "loss": 0.9703,
+      "step": 4064
+    },
+    {
+      "epoch": 0.427669647553919,
+      "grad_norm": 0.8028357028961182,
+      "learning_rate": 0.00012792496177566363,
+      "loss": 1.613,
+      "step": 4065
+    },
+    {
+      "epoch": 0.4277748553392951,
+      "grad_norm": 1.5005557537078857,
+      "learning_rate": 0.00012789223843320073,
+      "loss": 1.8242,
+      "step": 4066
+    },
+    {
+      "epoch": 0.42788006312467125,
+      "grad_norm": 1.2664399147033691,
+      "learning_rate": 0.00012785951185170403,
+      "loss": 2.2433,
+      "step": 4067
+    },
+    {
+      "epoch": 0.42798527091004734,
+      "grad_norm": 1.4799624681472778,
+      "learning_rate": 0.000127826782034974,
+      "loss": 2.0403,
+      "step": 4068
+    },
+    {
+      "epoch": 0.42809047869542344,
+      "grad_norm": 1.3635642528533936,
+      "learning_rate": 0.00012779404898681136,
+      "loss": 2.2521,
+      "step": 4069
+    },
+    {
+      "epoch": 0.4281956864807996,
+      "grad_norm": 1.3099942207336426,
+      "learning_rate": 0.00012776131271101732,
+      "loss": 2.0154,
+      "step": 4070
+    },
+    {
+      "epoch": 0.4283008942661757,
+      "grad_norm": 1.4328498840332031,
+      "learning_rate": 0.00012772857321139352,
+      "loss": 1.9576,
+      "step": 4071
+    },
+    {
+      "epoch": 0.42840610205155183,
+      "grad_norm": 1.504550576210022,
+      "learning_rate": 0.00012769583049174177,
+      "loss": 2.4232,
+      "step": 4072
+    },
+    {
+      "epoch": 0.4285113098369279,
+      "grad_norm": 1.3973898887634277,
+      "learning_rate": 0.0001276630845558644,
+      "loss": 1.5293,
+      "step": 4073
+    },
+    {
+      "epoch": 0.42861651762230407,
+      "grad_norm": 1.2882184982299805,
+      "learning_rate": 0.00012763033540756416,
+      "loss": 2.3094,
+      "step": 4074
+    },
+    {
+      "epoch": 0.42872172540768017,
+      "grad_norm": 1.2002280950546265,
+      "learning_rate": 0.00012759758305064405,
+      "loss": 1.8278,
+      "step": 4075
+    },
+    {
+      "epoch": 0.42882693319305626,
+      "grad_norm": 1.0488744974136353,
+      "learning_rate": 0.0001275648274889075,
+      "loss": 1.8504,
+      "step": 4076
+    },
+    {
+      "epoch": 0.4289321409784324,
+      "grad_norm": 1.2440074682235718,
+      "learning_rate": 0.00012753206872615825,
+      "loss": 1.8653,
+      "step": 4077
+    },
+    {
+      "epoch": 0.4290373487638085,
+      "grad_norm": 0.9582875967025757,
+      "learning_rate": 0.00012749930676620057,
+      "loss": 2.0311,
+      "step": 4078
+    },
+    {
+      "epoch": 0.42914255654918465,
+      "grad_norm": 1.1008398532867432,
+      "learning_rate": 0.00012746654161283896,
+      "loss": 1.6135,
+      "step": 4079
+    },
+    {
+      "epoch": 0.42924776433456074,
+      "grad_norm": 1.095392107963562,
+      "learning_rate": 0.00012743377326987826,
+      "loss": 1.7668,
+      "step": 4080
+    },
+    {
+      "epoch": 0.4293529721199369,
+      "grad_norm": 1.0410737991333008,
+      "learning_rate": 0.00012740100174112384,
+      "loss": 1.607,
+      "step": 4081
+    },
+    {
+      "epoch": 0.429458179905313,
+      "grad_norm": 0.979989767074585,
+      "learning_rate": 0.00012736822703038133,
+      "loss": 1.4283,
+      "step": 4082
+    },
+    {
+      "epoch": 0.42956338769068914,
+      "grad_norm": 1.1817355155944824,
+      "learning_rate": 0.00012733544914145673,
+      "loss": 1.7454,
+      "step": 4083
+    },
+    {
+      "epoch": 0.42966859547606523,
+      "grad_norm": 1.5766743421554565,
+      "learning_rate": 0.00012730266807815642,
+      "loss": 2.0112,
+      "step": 4084
+    },
+    {
+      "epoch": 0.4297738032614413,
+      "grad_norm": 1.0003412961959839,
+      "learning_rate": 0.0001272698838442872,
+      "loss": 1.9937,
+      "step": 4085
+    },
+    {
+      "epoch": 0.42987901104681747,
+      "grad_norm": 2.305222272872925,
+      "learning_rate": 0.00012723709644365614,
+      "loss": 2.0902,
+      "step": 4086
+    },
+    {
+      "epoch": 0.42998421883219357,
+      "grad_norm": 1.1228468418121338,
+      "learning_rate": 0.00012720430588007077,
+      "loss": 2.1054,
+      "step": 4087
+    },
+    {
+      "epoch": 0.4300894266175697,
+      "grad_norm": 1.6916412115097046,
+      "learning_rate": 0.00012717151215733892,
+      "loss": 1.3402,
+      "step": 4088
+    },
+    {
+      "epoch": 0.4301946344029458,
+      "grad_norm": 1.376791000366211,
+      "learning_rate": 0.0001271387152792689,
+      "loss": 2.2196,
+      "step": 4089
+    },
+    {
+      "epoch": 0.43029984218832196,
+      "grad_norm": 1.307098150253296,
+      "learning_rate": 0.00012710591524966918,
+      "loss": 1.9181,
+      "step": 4090
+    },
+    {
+      "epoch": 0.43040504997369805,
+      "grad_norm": 0.9531205892562866,
+      "learning_rate": 0.00012707311207234878,
+      "loss": 1.4014,
+      "step": 4091
+    },
+    {
+      "epoch": 0.43051025775907414,
+      "grad_norm": 1.9267292022705078,
+      "learning_rate": 0.00012704030575111705,
+      "loss": 1.4603,
+      "step": 4092
+    },
+    {
+      "epoch": 0.4306154655444503,
+      "grad_norm": 1.3264367580413818,
+      "learning_rate": 0.00012700749628978363,
+      "loss": 2.0973,
+      "step": 4093
+    },
+    {
+      "epoch": 0.4307206733298264,
+      "grad_norm": 1.430148720741272,
+      "learning_rate": 0.00012697468369215863,
+      "loss": 1.5874,
+      "step": 4094
+    },
+    {
+      "epoch": 0.43082588111520254,
+      "grad_norm": 1.3423150777816772,
+      "learning_rate": 0.00012694186796205243,
+      "loss": 1.6053,
+      "step": 4095
+    },
+    {
+      "epoch": 0.43093108890057863,
+      "grad_norm": 1.2820467948913574,
+      "learning_rate": 0.00012690904910327578,
+      "loss": 1.8371,
+      "step": 4096
+    },
+    {
+      "epoch": 0.4310362966859548,
+      "grad_norm": 1.965772271156311,
+      "learning_rate": 0.00012687622711963993,
+      "loss": 1.9467,
+      "step": 4097
+    },
+    {
+      "epoch": 0.43114150447133087,
+      "grad_norm": 1.4754441976547241,
+      "learning_rate": 0.0001268434020149563,
+      "loss": 1.8387,
+      "step": 4098
+    },
+    {
+      "epoch": 0.431246712256707,
+      "grad_norm": 1.0966287851333618,
+      "learning_rate": 0.00012681057379303678,
+      "loss": 2.0815,
+      "step": 4099
+    },
+    {
+      "epoch": 0.4313519200420831,
+      "grad_norm": 1.2483052015304565,
+      "learning_rate": 0.00012677774245769362,
+      "loss": 2.1525,
+      "step": 4100
+    },
+    {
+      "epoch": 0.4314571278274592,
+      "grad_norm": 1.1769499778747559,
+      "learning_rate": 0.00012674490801273938,
+      "loss": 2.0793,
+      "step": 4101
+    },
+    {
+      "epoch": 0.43156233561283536,
+      "grad_norm": 1.8558199405670166,
+      "learning_rate": 0.00012671207046198706,
+      "loss": 2.3923,
+      "step": 4102
+    },
+    {
+      "epoch": 0.43166754339821145,
+      "grad_norm": 1.0593931674957275,
+      "learning_rate": 0.00012667922980924998,
+      "loss": 2.0428,
+      "step": 4103
+    },
+    {
+      "epoch": 0.4317727511835876,
+      "grad_norm": 0.8765531182289124,
+      "learning_rate": 0.00012664638605834177,
+      "loss": 1.7572,
+      "step": 4104
+    },
+    {
+      "epoch": 0.4318779589689637,
+      "grad_norm": 0.9581341743469238,
+      "learning_rate": 0.00012661353921307648,
+      "loss": 2.2846,
+      "step": 4105
+    },
+    {
+      "epoch": 0.43198316675433984,
+      "grad_norm": 1.0098820924758911,
+      "learning_rate": 0.00012658068927726853,
+      "loss": 1.7325,
+      "step": 4106
+    },
+    {
+      "epoch": 0.43208837453971594,
+      "grad_norm": 1.224402666091919,
+      "learning_rate": 0.00012654783625473266,
+      "loss": 1.6651,
+      "step": 4107
+    },
+    {
+      "epoch": 0.43219358232509203,
+      "grad_norm": 0.9913700222969055,
+      "learning_rate": 0.00012651498014928402,
+      "loss": 2.4144,
+      "step": 4108
+    },
+    {
+      "epoch": 0.4322987901104682,
+      "grad_norm": 1.0635404586791992,
+      "learning_rate": 0.00012648212096473798,
+      "loss": 1.4415,
+      "step": 4109
+    },
+    {
+      "epoch": 0.43240399789584427,
+      "grad_norm": 0.8817548751831055,
+      "learning_rate": 0.00012644925870491052,
+      "loss": 1.8636,
+      "step": 4110
+    },
+    {
+      "epoch": 0.4325092056812204,
+      "grad_norm": 1.284338116645813,
+      "learning_rate": 0.0001264163933736177,
+      "loss": 1.6044,
+      "step": 4111
+    },
+    {
+      "epoch": 0.4326144134665965,
+      "grad_norm": 0.9684668779373169,
+      "learning_rate": 0.00012638352497467608,
+      "loss": 2.1041,
+      "step": 4112
+    },
+    {
+      "epoch": 0.43271962125197266,
+      "grad_norm": 1.4305078983306885,
+      "learning_rate": 0.00012635065351190261,
+      "loss": 1.9746,
+      "step": 4113
+    },
+    {
+      "epoch": 0.43282482903734876,
+      "grad_norm": 1.0623332262039185,
+      "learning_rate": 0.0001263177789891145,
+      "loss": 1.7623,
+      "step": 4114
+    },
+    {
+      "epoch": 0.4329300368227249,
+      "grad_norm": 1.431246042251587,
+      "learning_rate": 0.00012628490141012937,
+      "loss": 2.0701,
+      "step": 4115
+    },
+    {
+      "epoch": 0.433035244608101,
+      "grad_norm": 1.5637847185134888,
+      "learning_rate": 0.00012625202077876525,
+      "loss": 1.6522,
+      "step": 4116
+    },
+    {
+      "epoch": 0.4331404523934771,
+      "grad_norm": 1.6422128677368164,
+      "learning_rate": 0.00012621913709884037,
+      "loss": 1.8042,
+      "step": 4117
+    },
+    {
+      "epoch": 0.43324566017885324,
+      "grad_norm": 1.0828219652175903,
+      "learning_rate": 0.0001261862503741734,
+      "loss": 1.9911,
+      "step": 4118
+    },
+    {
+      "epoch": 0.43335086796422934,
+      "grad_norm": 1.0087640285491943,
+      "learning_rate": 0.00012615336060858344,
+      "loss": 2.199,
+      "step": 4119
+    },
+    {
+      "epoch": 0.4334560757496055,
+      "grad_norm": 1.5731931924819946,
+      "learning_rate": 0.00012612046780588986,
+      "loss": 1.9389,
+      "step": 4120
+    },
+    {
+      "epoch": 0.4335612835349816,
+      "grad_norm": 1.271106481552124,
+      "learning_rate": 0.00012608757196991234,
+      "loss": 1.8302,
+      "step": 4121
+    },
+    {
+      "epoch": 0.4336664913203577,
+      "grad_norm": 1.1410576105117798,
+      "learning_rate": 0.000126054673104471,
+      "loss": 1.6422,
+      "step": 4122
+    },
+    {
+      "epoch": 0.4337716991057338,
+      "grad_norm": 1.52702796459198,
+      "learning_rate": 0.00012602177121338626,
+      "loss": 2.0576,
+      "step": 4123
+    },
+    {
+      "epoch": 0.4338769068911099,
+      "grad_norm": 1.1503021717071533,
+      "learning_rate": 0.0001259888663004789,
+      "loss": 1.6653,
+      "step": 4124
+    },
+    {
+      "epoch": 0.43398211467648606,
+      "grad_norm": 1.6308045387268066,
+      "learning_rate": 0.00012595595836957006,
+      "loss": 1.885,
+      "step": 4125
+    },
+    {
+      "epoch": 0.43408732246186216,
+      "grad_norm": 1.5235882997512817,
+      "learning_rate": 0.0001259230474244813,
+      "loss": 1.9792,
+      "step": 4126
+    },
+    {
+      "epoch": 0.4341925302472383,
+      "grad_norm": 1.6996004581451416,
+      "learning_rate": 0.00012589013346903438,
+      "loss": 1.9094,
+      "step": 4127
+    },
+    {
+      "epoch": 0.4342977380326144,
+      "grad_norm": 1.2587865591049194,
+      "learning_rate": 0.0001258572165070515,
+      "loss": 1.8968,
+      "step": 4128
+    },
+    {
+      "epoch": 0.43440294581799055,
+      "grad_norm": 1.0128796100616455,
+      "learning_rate": 0.00012582429654235523,
+      "loss": 1.1083,
+      "step": 4129
+    },
+    {
+      "epoch": 0.43450815360336664,
+      "grad_norm": 1.4834532737731934,
+      "learning_rate": 0.00012579137357876844,
+      "loss": 2.2179,
+      "step": 4130
+    },
+    {
+      "epoch": 0.4346133613887428,
+      "grad_norm": 1.3126288652420044,
+      "learning_rate": 0.00012575844762011438,
+      "loss": 2.4677,
+      "step": 4131
+    },
+    {
+      "epoch": 0.4347185691741189,
+      "grad_norm": 0.8772363066673279,
+      "learning_rate": 0.0001257255186702166,
+      "loss": 1.4076,
+      "step": 4132
+    },
+    {
+      "epoch": 0.434823776959495,
+      "grad_norm": 1.3988395929336548,
+      "learning_rate": 0.00012569258673289903,
+      "loss": 1.8661,
+      "step": 4133
+    },
+    {
+      "epoch": 0.4349289847448711,
+      "grad_norm": 1.469456434249878,
+      "learning_rate": 0.000125659651811986,
+      "loss": 1.8328,
+      "step": 4134
+    },
+    {
+      "epoch": 0.4350341925302472,
+      "grad_norm": 1.3229504823684692,
+      "learning_rate": 0.00012562671391130208,
+      "loss": 1.7977,
+      "step": 4135
+    },
+    {
+      "epoch": 0.43513940031562337,
+      "grad_norm": 1.2461540699005127,
+      "learning_rate": 0.00012559377303467226,
+      "loss": 1.7205,
+      "step": 4136
+    },
+    {
+      "epoch": 0.43524460810099946,
+      "grad_norm": 1.4295867681503296,
+      "learning_rate": 0.00012556082918592187,
+      "loss": 2.3218,
+      "step": 4137
+    },
+    {
+      "epoch": 0.4353498158863756,
+      "grad_norm": 1.1670809984207153,
+      "learning_rate": 0.00012552788236887654,
+      "loss": 1.7761,
+      "step": 4138
+    },
+    {
+      "epoch": 0.4354550236717517,
+      "grad_norm": 2.9805972576141357,
+      "learning_rate": 0.0001254949325873623,
+      "loss": 1.9536,
+      "step": 4139
+    },
+    {
+      "epoch": 0.4355602314571278,
+      "grad_norm": 1.0447720289230347,
+      "learning_rate": 0.0001254619798452055,
+      "loss": 1.6888,
+      "step": 4140
+    },
+    {
+      "epoch": 0.43566543924250395,
+      "grad_norm": 1.6240369081497192,
+      "learning_rate": 0.00012542902414623282,
+      "loss": 2.3843,
+      "step": 4141
+    },
+    {
+      "epoch": 0.43577064702788004,
+      "grad_norm": 1.2419333457946777,
+      "learning_rate": 0.0001253960654942713,
+      "loss": 2.0027,
+      "step": 4142
+    },
+    {
+      "epoch": 0.4358758548132562,
+      "grad_norm": 1.9167548418045044,
+      "learning_rate": 0.00012536310389314832,
+      "loss": 1.8499,
+      "step": 4143
+    },
+    {
+      "epoch": 0.4359810625986323,
+      "grad_norm": 0.9558454155921936,
+      "learning_rate": 0.0001253301393466916,
+      "loss": 2.0347,
+      "step": 4144
+    },
+    {
+      "epoch": 0.43608627038400843,
+      "grad_norm": 1.5771024227142334,
+      "learning_rate": 0.0001252971718587292,
+      "loss": 2.0212,
+      "step": 4145
+    },
+    {
+      "epoch": 0.43619147816938453,
+      "grad_norm": 1.3093889951705933,
+      "learning_rate": 0.00012526420143308954,
+      "loss": 1.4546,
+      "step": 4146
+    },
+    {
+      "epoch": 0.4362966859547607,
+      "grad_norm": 1.1525970697402954,
+      "learning_rate": 0.00012523122807360138,
+      "loss": 1.8394,
+      "step": 4147
+    },
+    {
+      "epoch": 0.43640189374013677,
+      "grad_norm": 1.2308920621871948,
+      "learning_rate": 0.00012519825178409377,
+      "loss": 1.8985,
+      "step": 4148
+    },
+    {
+      "epoch": 0.43650710152551286,
+      "grad_norm": 1.5874372720718384,
+      "learning_rate": 0.00012516527256839616,
+      "loss": 1.814,
+      "step": 4149
+    },
+    {
+      "epoch": 0.436612309310889,
+      "grad_norm": 1.3634819984436035,
+      "learning_rate": 0.0001251322904303383,
+      "loss": 1.764,
+      "step": 4150
+    },
+    {
+      "epoch": 0.4367175170962651,
+      "grad_norm": 1.272567868232727,
+      "learning_rate": 0.00012509930537375036,
+      "loss": 2.2667,
+      "step": 4151
+    },
+    {
+      "epoch": 0.43682272488164126,
+      "grad_norm": 2.495673418045044,
+      "learning_rate": 0.0001250663174024627,
+      "loss": 1.88,
+      "step": 4152
+    },
+    {
+      "epoch": 0.43692793266701735,
+      "grad_norm": 1.3136742115020752,
+      "learning_rate": 0.00012503332652030613,
+      "loss": 1.7169,
+      "step": 4153
+    },
+    {
+      "epoch": 0.4370331404523935,
+      "grad_norm": 1.5158318281173706,
+      "learning_rate": 0.0001250003327311118,
+      "loss": 1.7009,
+      "step": 4154
+    },
+    {
+      "epoch": 0.4371383482377696,
+      "grad_norm": 1.6757663488388062,
+      "learning_rate": 0.00012496733603871115,
+      "loss": 2.4002,
+      "step": 4155
+    },
+    {
+      "epoch": 0.4372435560231457,
+      "grad_norm": 1.6522952318191528,
+      "learning_rate": 0.000124934336446936,
+      "loss": 1.8635,
+      "step": 4156
+    },
+    {
+      "epoch": 0.43734876380852183,
+      "grad_norm": 1.3090764284133911,
+      "learning_rate": 0.00012490133395961844,
+      "loss": 1.7477,
+      "step": 4157
+    },
+    {
+      "epoch": 0.43745397159389793,
+      "grad_norm": 1.1116405725479126,
+      "learning_rate": 0.000124868328580591,
+      "loss": 1.7589,
+      "step": 4158
+    },
+    {
+      "epoch": 0.4375591793792741,
+      "grad_norm": 1.0621029138565063,
+      "learning_rate": 0.0001248353203136864,
+      "loss": 1.7158,
+      "step": 4159
+    },
+    {
+      "epoch": 0.43766438716465017,
+      "grad_norm": 1.3473455905914307,
+      "learning_rate": 0.00012480230916273784,
+      "loss": 1.8552,
+      "step": 4160
+    },
+    {
+      "epoch": 0.4377695949500263,
+      "grad_norm": 1.4094268083572388,
+      "learning_rate": 0.00012476929513157881,
+      "loss": 1.7809,
+      "step": 4161
+    },
+    {
+      "epoch": 0.4378748027354024,
+      "grad_norm": 1.2711224555969238,
+      "learning_rate": 0.00012473627822404314,
+      "loss": 1.9564,
+      "step": 4162
+    },
+    {
+      "epoch": 0.43798001052077856,
+      "grad_norm": 1.670932650566101,
+      "learning_rate": 0.00012470325844396487,
+      "loss": 1.7578,
+      "step": 4163
+    },
+    {
+      "epoch": 0.43808521830615466,
+      "grad_norm": 1.6926169395446777,
+      "learning_rate": 0.00012467023579517856,
+      "loss": 1.9468,
+      "step": 4164
+    },
+    {
+      "epoch": 0.43819042609153075,
+      "grad_norm": 1.197359323501587,
+      "learning_rate": 0.000124637210281519,
+      "loss": 1.4122,
+      "step": 4165
+    },
+    {
+      "epoch": 0.4382956338769069,
+      "grad_norm": 1.7274643182754517,
+      "learning_rate": 0.00012460418190682134,
+      "loss": 2.1852,
+      "step": 4166
+    },
+    {
+      "epoch": 0.438400841662283,
+      "grad_norm": 1.5755963325500488,
+      "learning_rate": 0.00012457115067492108,
+      "loss": 2.3088,
+      "step": 4167
+    },
+    {
+      "epoch": 0.43850604944765914,
+      "grad_norm": 1.3150062561035156,
+      "learning_rate": 0.000124538116589654,
+      "loss": 1.9563,
+      "step": 4168
+    },
+    {
+      "epoch": 0.43861125723303523,
+      "grad_norm": 1.4447141885757446,
+      "learning_rate": 0.0001245050796548562,
+      "loss": 1.6322,
+      "step": 4169
+    },
+    {
+      "epoch": 0.4387164650184114,
+      "grad_norm": 1.5568017959594727,
+      "learning_rate": 0.0001244720398743642,
+      "loss": 1.272,
+      "step": 4170
+    },
+    {
+      "epoch": 0.4388216728037875,
+      "grad_norm": 1.4406249523162842,
+      "learning_rate": 0.00012443899725201482,
+      "loss": 1.4789,
+      "step": 4171
+    },
+    {
+      "epoch": 0.43892688058916357,
+      "grad_norm": 2.007683753967285,
+      "learning_rate": 0.0001244059517916452,
+      "loss": 2.0801,
+      "step": 4172
+    },
+    {
+      "epoch": 0.4390320883745397,
+      "grad_norm": 2.023632287979126,
+      "learning_rate": 0.00012437290349709271,
+      "loss": 2.739,
+      "step": 4173
+    },
+    {
+      "epoch": 0.4391372961599158,
+      "grad_norm": 1.7203633785247803,
+      "learning_rate": 0.0001243398523721952,
+      "loss": 2.2986,
+      "step": 4174
+    },
+    {
+      "epoch": 0.43924250394529196,
+      "grad_norm": 1.1265063285827637,
+      "learning_rate": 0.0001243067984207908,
+      "loss": 1.9861,
+      "step": 4175
+    },
+    {
+      "epoch": 0.43934771173066806,
+      "grad_norm": 1.3561371564865112,
+      "learning_rate": 0.00012427374164671794,
+      "loss": 2.1845,
+      "step": 4176
+    },
+    {
+      "epoch": 0.4394529195160442,
+      "grad_norm": 1.0997103452682495,
+      "learning_rate": 0.00012424068205381538,
+      "loss": 2.1557,
+      "step": 4177
+    },
+    {
+      "epoch": 0.4395581273014203,
+      "grad_norm": 1.7254860401153564,
+      "learning_rate": 0.00012420761964592223,
+      "loss": 1.6301,
+      "step": 4178
+    },
+    {
+      "epoch": 0.43966333508679645,
+      "grad_norm": 1.4888262748718262,
+      "learning_rate": 0.00012417455442687795,
+      "loss": 1.9758,
+      "step": 4179
+    },
+    {
+      "epoch": 0.43976854287217254,
+      "grad_norm": 1.2371491193771362,
+      "learning_rate": 0.00012414148640052227,
+      "loss": 2.0836,
+      "step": 4180
+    },
+    {
+      "epoch": 0.43987375065754863,
+      "grad_norm": 1.2289551496505737,
+      "learning_rate": 0.00012410841557069523,
+      "loss": 2.0425,
+      "step": 4181
+    },
+    {
+      "epoch": 0.4399789584429248,
+      "grad_norm": 1.2869768142700195,
+      "learning_rate": 0.0001240753419412373,
+      "loss": 1.9797,
+      "step": 4182
+    },
+    {
+      "epoch": 0.4400841662283009,
+      "grad_norm": 1.227531909942627,
+      "learning_rate": 0.00012404226551598923,
+      "loss": 2.0121,
+      "step": 4183
+    },
+    {
+      "epoch": 0.440189374013677,
+      "grad_norm": 1.2489982843399048,
+      "learning_rate": 0.000124009186298792,
+      "loss": 1.5442,
+      "step": 4184
+    },
+    {
+      "epoch": 0.4402945817990531,
+      "grad_norm": 1.3487237691879272,
+      "learning_rate": 0.000123976104293487,
+      "loss": 1.7614,
+      "step": 4185
+    },
+    {
+      "epoch": 0.44039978958442927,
+      "grad_norm": 1.567232370376587,
+      "learning_rate": 0.000123943019503916,
+      "loss": 1.7655,
+      "step": 4186
+    },
+    {
+      "epoch": 0.44050499736980536,
+      "grad_norm": 3.1679434776306152,
+      "learning_rate": 0.00012390993193392097,
+      "loss": 1.0603,
+      "step": 4187
+    },
+    {
+      "epoch": 0.44061020515518146,
+      "grad_norm": 1.1254740953445435,
+      "learning_rate": 0.00012387684158734425,
+      "loss": 1.781,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4407154129405576,
+      "grad_norm": 1.0286729335784912,
+      "learning_rate": 0.0001238437484680286,
+      "loss": 1.839,
+      "step": 4189
+    },
+    {
+      "epoch": 0.4408206207259337,
+      "grad_norm": 1.7165943384170532,
+      "learning_rate": 0.0001238106525798169,
+      "loss": 1.6264,
+      "step": 4190
+    },
+    {
+      "epoch": 0.44092582851130985,
+      "grad_norm": 1.339667797088623,
+      "learning_rate": 0.00012377755392655254,
+      "loss": 1.8619,
+      "step": 4191
+    },
+    {
+      "epoch": 0.44103103629668594,
+      "grad_norm": 1.1098575592041016,
+      "learning_rate": 0.00012374445251207914,
+      "loss": 1.8949,
+      "step": 4192
+    },
+    {
+      "epoch": 0.4411362440820621,
+      "grad_norm": 1.4389253854751587,
+      "learning_rate": 0.00012371134834024067,
+      "loss": 1.6791,
+      "step": 4193
+    },
+    {
+      "epoch": 0.4412414518674382,
+      "grad_norm": 1.6168919801712036,
+      "learning_rate": 0.00012367824141488142,
+      "loss": 1.251,
+      "step": 4194
+    },
+    {
+      "epoch": 0.44134665965281433,
+      "grad_norm": 1.0156736373901367,
+      "learning_rate": 0.00012364513173984592,
+      "loss": 2.281,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4414518674381904,
+      "grad_norm": 1.645992636680603,
+      "learning_rate": 0.00012361201931897916,
+      "loss": 1.8607,
+      "step": 4196
+    },
+    {
+      "epoch": 0.4415570752235665,
+      "grad_norm": 1.0895994901657104,
+      "learning_rate": 0.00012357890415612635,
+      "loss": 2.0752,
+      "step": 4197
+    },
+    {
+      "epoch": 0.44166228300894267,
+      "grad_norm": 1.4120984077453613,
+      "learning_rate": 0.00012354578625513302,
+      "loss": 1.6247,
+      "step": 4198
+    },
+    {
+      "epoch": 0.44176749079431876,
+      "grad_norm": 1.073986291885376,
+      "learning_rate": 0.00012351266561984507,
+      "loss": 1.4361,
+      "step": 4199
+    },
+    {
+      "epoch": 0.4418726985796949,
+      "grad_norm": 1.2587485313415527,
+      "learning_rate": 0.0001234795422541087,
+      "loss": 2.4348,
+      "step": 4200
+    },
+    {
+      "epoch": 0.441977906365071,
+      "grad_norm": 1.7424877882003784,
+      "learning_rate": 0.00012344641616177042,
+      "loss": 1.1056,
+      "step": 4201
+    },
+    {
+      "epoch": 0.44208311415044715,
+      "grad_norm": 1.4434423446655273,
+      "learning_rate": 0.00012341328734667698,
+      "loss": 2.2263,
+      "step": 4202
+    },
+    {
+      "epoch": 0.44218832193582325,
+      "grad_norm": 1.3455250263214111,
+      "learning_rate": 0.00012338015581267567,
+      "loss": 1.872,
+      "step": 4203
+    },
+    {
+      "epoch": 0.44229352972119934,
+      "grad_norm": 1.5791774988174438,
+      "learning_rate": 0.00012334702156361377,
+      "loss": 2.61,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4423987375065755,
+      "grad_norm": 1.0933024883270264,
+      "learning_rate": 0.0001233138846033392,
+      "loss": 2.0891,
+      "step": 4205
+    },
+    {
+      "epoch": 0.4425039452919516,
+      "grad_norm": 1.2376352548599243,
+      "learning_rate": 0.00012328074493569993,
+      "loss": 1.4209,
+      "step": 4206
+    },
+    {
+      "epoch": 0.44260915307732773,
+      "grad_norm": 1.549952507019043,
+      "learning_rate": 0.00012324760256454445,
+      "loss": 1.6895,
+      "step": 4207
+    },
+    {
+      "epoch": 0.4427143608627038,
+      "grad_norm": 1.5059970617294312,
+      "learning_rate": 0.0001232144574937214,
+      "loss": 1.8844,
+      "step": 4208
+    },
+    {
+      "epoch": 0.44281956864808,
+      "grad_norm": 1.6260056495666504,
+      "learning_rate": 0.00012318130972707985,
+      "loss": 2.5282,
+      "step": 4209
+    },
+    {
+      "epoch": 0.44292477643345607,
+      "grad_norm": 1.3028151988983154,
+      "learning_rate": 0.0001231481592684692,
+      "loss": 2.0735,
+      "step": 4210
+    },
+    {
+      "epoch": 0.4430299842188322,
+      "grad_norm": 0.9203398823738098,
+      "learning_rate": 0.00012311500612173897,
+      "loss": 2.0104,
+      "step": 4211
+    },
+    {
+      "epoch": 0.4431351920042083,
+      "grad_norm": 1.2784510850906372,
+      "learning_rate": 0.0001230818502907392,
+      "loss": 1.7065,
+      "step": 4212
+    },
+    {
+      "epoch": 0.4432403997895844,
+      "grad_norm": 1.1374523639678955,
+      "learning_rate": 0.0001230486917793202,
+      "loss": 1.569,
+      "step": 4213
+    },
+    {
+      "epoch": 0.44334560757496055,
+      "grad_norm": 1.1995996236801147,
+      "learning_rate": 0.00012301553059133248,
+      "loss": 1.6635,
+      "step": 4214
+    },
+    {
+      "epoch": 0.44345081536033665,
+      "grad_norm": 1.210629940032959,
+      "learning_rate": 0.000122982366730627,
+      "loss": 1.8382,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4435560231457128,
+      "grad_norm": 1.6628928184509277,
+      "learning_rate": 0.00012294920020105497,
+      "loss": 1.4067,
+      "step": 4216
+    },
+    {
+      "epoch": 0.4436612309310889,
+      "grad_norm": 1.001809000968933,
+      "learning_rate": 0.00012291603100646786,
+      "loss": 1.9476,
+      "step": 4217
+    },
+    {
+      "epoch": 0.44376643871646504,
+      "grad_norm": 2.0265283584594727,
+      "learning_rate": 0.00012288285915071752,
+      "loss": 1.9624,
+      "step": 4218
+    },
+    {
+      "epoch": 0.44387164650184113,
+      "grad_norm": 1.6050817966461182,
+      "learning_rate": 0.00012284968463765613,
+      "loss": 1.7624,
+      "step": 4219
+    },
+    {
+      "epoch": 0.4439768542872172,
+      "grad_norm": 1.2535775899887085,
+      "learning_rate": 0.00012281650747113612,
+      "loss": 2.0082,
+      "step": 4220
+    },
+    {
+      "epoch": 0.4440820620725934,
+      "grad_norm": 1.6964945793151855,
+      "learning_rate": 0.00012278332765501017,
+      "loss": 2.1988,
+      "step": 4221
+    },
+    {
+      "epoch": 0.44418726985796947,
+      "grad_norm": 1.662829041481018,
+      "learning_rate": 0.00012275014519313145,
+      "loss": 2.4153,
+      "step": 4222
+    },
+    {
+      "epoch": 0.4442924776433456,
+      "grad_norm": 1.6172958612442017,
+      "learning_rate": 0.00012271696008935324,
+      "loss": 1.6942,
+      "step": 4223
+    },
+    {
+      "epoch": 0.4443976854287217,
+      "grad_norm": 1.380727767944336,
+      "learning_rate": 0.0001226837723475293,
+      "loss": 1.9723,
+      "step": 4224
+    },
+    {
+      "epoch": 0.44450289321409786,
+      "grad_norm": 1.6496597528457642,
+      "learning_rate": 0.00012265058197151357,
+      "loss": 1.5928,
+      "step": 4225
+    },
+    {
+      "epoch": 0.44460810099947395,
+      "grad_norm": 1.51160728931427,
+      "learning_rate": 0.00012261738896516034,
+      "loss": 2.1494,
+      "step": 4226
+    },
+    {
+      "epoch": 0.4447133087848501,
+      "grad_norm": 1.5989267826080322,
+      "learning_rate": 0.0001225841933323242,
+      "loss": 1.6661,
+      "step": 4227
+    },
+    {
+      "epoch": 0.4448185165702262,
+      "grad_norm": 1.2797974348068237,
+      "learning_rate": 0.00012255099507686007,
+      "loss": 2.1762,
+      "step": 4228
+    },
+    {
+      "epoch": 0.4449237243556023,
+      "grad_norm": 0.9348226189613342,
+      "learning_rate": 0.00012251779420262312,
+      "loss": 1.6779,
+      "step": 4229
+    },
+    {
+      "epoch": 0.44502893214097844,
+      "grad_norm": 1.528698205947876,
+      "learning_rate": 0.0001224845907134689,
+      "loss": 1.4624,
+      "step": 4230
+    },
+    {
+      "epoch": 0.44513413992635453,
+      "grad_norm": 2.389446973800659,
+      "learning_rate": 0.00012245138461325318,
+      "loss": 1.4935,
+      "step": 4231
+    },
+    {
+      "epoch": 0.4452393477117307,
+      "grad_norm": 1.2253481149673462,
+      "learning_rate": 0.0001224181759058321,
+      "loss": 2.4449,
+      "step": 4232
+    },
+    {
+      "epoch": 0.4453445554971068,
+      "grad_norm": 1.3102329969406128,
+      "learning_rate": 0.00012238496459506207,
+      "loss": 1.9277,
+      "step": 4233
+    },
+    {
+      "epoch": 0.4454497632824829,
+      "grad_norm": 1.1610690355300903,
+      "learning_rate": 0.00012235175068479984,
+      "loss": 2.1319,
+      "step": 4234
+    },
+    {
+      "epoch": 0.445554971067859,
+      "grad_norm": 2.348677635192871,
+      "learning_rate": 0.00012231853417890237,
+      "loss": 2.2125,
+      "step": 4235
+    },
+    {
+      "epoch": 0.44566017885323517,
+      "grad_norm": 1.5758837461471558,
+      "learning_rate": 0.00012228531508122703,
+      "loss": 1.6423,
+      "step": 4236
+    },
+    {
+      "epoch": 0.44576538663861126,
+      "grad_norm": 1.0791726112365723,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 1.7604,
+      "step": 4237
+    },
+    {
+      "epoch": 0.44587059442398735,
+      "grad_norm": 1.3658674955368042,
+      "learning_rate": 0.00012221886912597353,
+      "loss": 1.8358,
+      "step": 4238
+    },
+    {
+      "epoch": 0.4459758022093635,
+      "grad_norm": 2.2098424434661865,
+      "learning_rate": 0.00012218564227611152,
+      "loss": 2.1306,
+      "step": 4239
+    },
+    {
+      "epoch": 0.4460810099947396,
+      "grad_norm": 1.6169191598892212,
+      "learning_rate": 0.0001221524128499039,
+      "loss": 2.0847,
+      "step": 4240
+    },
+    {
+      "epoch": 0.44618621778011575,
+      "grad_norm": 1.3867640495300293,
+      "learning_rate": 0.00012211918085120954,
+      "loss": 1.8926,
+      "step": 4241
+    },
+    {
+      "epoch": 0.44629142556549184,
+      "grad_norm": 1.7240827083587646,
+      "learning_rate": 0.00012208594628388753,
+      "loss": 2.1408,
+      "step": 4242
+    },
+    {
+      "epoch": 0.446396633350868,
+      "grad_norm": 1.132155179977417,
+      "learning_rate": 0.00012205270915179729,
+      "loss": 2.1424,
+      "step": 4243
+    },
+    {
+      "epoch": 0.4465018411362441,
+      "grad_norm": 1.2491706609725952,
+      "learning_rate": 0.00012201946945879856,
+      "loss": 2.039,
+      "step": 4244
+    },
+    {
+      "epoch": 0.4466070489216202,
+      "grad_norm": 1.3799735307693481,
+      "learning_rate": 0.00012198622720875139,
+      "loss": 1.7383,
+      "step": 4245
+    },
+    {
+      "epoch": 0.4467122567069963,
+      "grad_norm": 1.3152798414230347,
+      "learning_rate": 0.000121952982405516,
+      "loss": 2.1134,
+      "step": 4246
+    },
+    {
+      "epoch": 0.4468174644923724,
+      "grad_norm": 1.6262292861938477,
+      "learning_rate": 0.00012191973505295311,
+      "loss": 1.8854,
+      "step": 4247
+    },
+    {
+      "epoch": 0.44692267227774857,
+      "grad_norm": 1.3993144035339355,
+      "learning_rate": 0.00012188648515492355,
+      "loss": 1.8349,
+      "step": 4248
+    },
+    {
+      "epoch": 0.44702788006312466,
+      "grad_norm": 1.2754366397857666,
+      "learning_rate": 0.00012185323271528853,
+      "loss": 1.2813,
+      "step": 4249
+    },
+    {
+      "epoch": 0.4471330878485008,
+      "grad_norm": 1.437742829322815,
+      "learning_rate": 0.00012181997773790954,
+      "loss": 1.79,
+      "step": 4250
+    },
+    {
+      "epoch": 0.4472382956338769,
+      "grad_norm": 1.1852113008499146,
+      "learning_rate": 0.00012178672022664838,
+      "loss": 2.2712,
+      "step": 4251
+    },
+    {
+      "epoch": 0.44734350341925305,
+      "grad_norm": 1.1398671865463257,
+      "learning_rate": 0.00012175346018536717,
+      "loss": 2.3377,
+      "step": 4252
+    },
+    {
+      "epoch": 0.44744871120462915,
+      "grad_norm": 0.8755077123641968,
+      "learning_rate": 0.00012172019761792825,
+      "loss": 2.2223,
+      "step": 4253
+    },
+    {
+      "epoch": 0.44755391899000524,
+      "grad_norm": 0.9048967957496643,
+      "learning_rate": 0.00012168693252819433,
+      "loss": 1.8152,
+      "step": 4254
+    },
+    {
+      "epoch": 0.4476591267753814,
+      "grad_norm": 1.133655071258545,
+      "learning_rate": 0.00012165366492002832,
+      "loss": 1.8637,
+      "step": 4255
+    },
+    {
+      "epoch": 0.4477643345607575,
+      "grad_norm": 2.401837110519409,
+      "learning_rate": 0.00012162039479729351,
+      "loss": 2.07,
+      "step": 4256
+    },
+    {
+      "epoch": 0.44786954234613363,
+      "grad_norm": 1.3808231353759766,
+      "learning_rate": 0.00012158712216385344,
+      "loss": 1.9546,
+      "step": 4257
+    },
+    {
+      "epoch": 0.4479747501315097,
+      "grad_norm": 1.5541157722473145,
+      "learning_rate": 0.00012155384702357198,
+      "loss": 1.6161,
+      "step": 4258
+    },
+    {
+      "epoch": 0.4480799579168859,
+      "grad_norm": 1.5308620929718018,
+      "learning_rate": 0.00012152056938031324,
+      "loss": 2.0226,
+      "step": 4259
+    },
+    {
+      "epoch": 0.44818516570226197,
+      "grad_norm": 1.0472149848937988,
+      "learning_rate": 0.00012148728923794162,
+      "loss": 2.0012,
+      "step": 4260
+    },
+    {
+      "epoch": 0.44829037348763806,
+      "grad_norm": 1.0326931476593018,
+      "learning_rate": 0.00012145400660032187,
+      "loss": 1.6771,
+      "step": 4261
+    },
+    {
+      "epoch": 0.4483955812730142,
+      "grad_norm": 1.4375073909759521,
+      "learning_rate": 0.00012142072147131898,
+      "loss": 2.0649,
+      "step": 4262
+    },
+    {
+      "epoch": 0.4485007890583903,
+      "grad_norm": 0.880775511264801,
+      "learning_rate": 0.00012138743385479823,
+      "loss": 1.7757,
+      "step": 4263
+    },
+    {
+      "epoch": 0.44860599684376645,
+      "grad_norm": 1.205504298210144,
+      "learning_rate": 0.00012135414375462522,
+      "loss": 1.7152,
+      "step": 4264
+    },
+    {
+      "epoch": 0.44871120462914255,
+      "grad_norm": 0.8612405061721802,
+      "learning_rate": 0.00012132085117466582,
+      "loss": 1.7546,
+      "step": 4265
+    },
+    {
+      "epoch": 0.4488164124145187,
+      "grad_norm": 1.592641830444336,
+      "learning_rate": 0.00012128755611878617,
+      "loss": 1.8245,
+      "step": 4266
+    },
+    {
+      "epoch": 0.4489216201998948,
+      "grad_norm": 1.2067207098007202,
+      "learning_rate": 0.00012125425859085273,
+      "loss": 1.6449,
+      "step": 4267
+    },
+    {
+      "epoch": 0.44902682798527094,
+      "grad_norm": 1.0257954597473145,
+      "learning_rate": 0.00012122095859473223,
+      "loss": 1.13,
+      "step": 4268
+    },
+    {
+      "epoch": 0.44913203577064703,
+      "grad_norm": 2.4131839275360107,
+      "learning_rate": 0.00012118765613429173,
+      "loss": 1.3638,
+      "step": 4269
+    },
+    {
+      "epoch": 0.4492372435560231,
+      "grad_norm": 1.0637006759643555,
+      "learning_rate": 0.00012115435121339844,
+      "loss": 1.5358,
+      "step": 4270
+    },
+    {
+      "epoch": 0.4493424513413993,
+      "grad_norm": 1.3625959157943726,
+      "learning_rate": 0.00012112104383592,
+      "loss": 1.4193,
+      "step": 4271
+    },
+    {
+      "epoch": 0.44944765912677537,
+      "grad_norm": 1.6006038188934326,
+      "learning_rate": 0.00012108773400572431,
+      "loss": 1.8992,
+      "step": 4272
+    },
+    {
+      "epoch": 0.4495528669121515,
+      "grad_norm": 1.9256750345230103,
+      "learning_rate": 0.00012105442172667951,
+      "loss": 2.2142,
+      "step": 4273
+    },
+    {
+      "epoch": 0.4496580746975276,
+      "grad_norm": 1.241640567779541,
+      "learning_rate": 0.00012102110700265403,
+      "loss": 1.4468,
+      "step": 4274
+    },
+    {
+      "epoch": 0.44976328248290376,
+      "grad_norm": 1.4494163990020752,
+      "learning_rate": 0.00012098778983751662,
+      "loss": 1.8657,
+      "step": 4275
+    },
+    {
+      "epoch": 0.44986849026827985,
+      "grad_norm": 0.9596696496009827,
+      "learning_rate": 0.0001209544702351363,
+      "loss": 1.9698,
+      "step": 4276
+    },
+    {
+      "epoch": 0.44997369805365595,
+      "grad_norm": 0.9092001914978027,
+      "learning_rate": 0.00012092114819938233,
+      "loss": 1.3448,
+      "step": 4277
+    },
+    {
+      "epoch": 0.4500789058390321,
+      "grad_norm": 1.8261799812316895,
+      "learning_rate": 0.00012088782373412432,
+      "loss": 2.2352,
+      "step": 4278
+    },
+    {
+      "epoch": 0.4501841136244082,
+      "grad_norm": 1.3973294496536255,
+      "learning_rate": 0.00012085449684323216,
+      "loss": 1.4725,
+      "step": 4279
+    },
+    {
+      "epoch": 0.45028932140978434,
+      "grad_norm": 1.5682103633880615,
+      "learning_rate": 0.00012082116753057593,
+      "loss": 1.8422,
+      "step": 4280
+    },
+    {
+      "epoch": 0.45039452919516043,
+      "grad_norm": 1.5455161333084106,
+      "learning_rate": 0.00012078783580002607,
+      "loss": 1.604,
+      "step": 4281
+    },
+    {
+      "epoch": 0.4504997369805366,
+      "grad_norm": 1.6469101905822754,
+      "learning_rate": 0.00012075450165545328,
+      "loss": 2.1398,
+      "step": 4282
+    },
+    {
+      "epoch": 0.4506049447659127,
+      "grad_norm": 1.5728659629821777,
+      "learning_rate": 0.00012072116510072858,
+      "loss": 1.9734,
+      "step": 4283
+    },
+    {
+      "epoch": 0.4507101525512888,
+      "grad_norm": 1.4864883422851562,
+      "learning_rate": 0.00012068782613972318,
+      "loss": 1.7709,
+      "step": 4284
+    },
+    {
+      "epoch": 0.4508153603366649,
+      "grad_norm": 0.9983920454978943,
+      "learning_rate": 0.00012065448477630867,
+      "loss": 1.5441,
+      "step": 4285
+    },
+    {
+      "epoch": 0.450920568122041,
+      "grad_norm": 1.9383924007415771,
+      "learning_rate": 0.00012062114101435686,
+      "loss": 1.9526,
+      "step": 4286
+    },
+    {
+      "epoch": 0.45102577590741716,
+      "grad_norm": 1.959188461303711,
+      "learning_rate": 0.00012058779485773985,
+      "loss": 1.9704,
+      "step": 4287
+    },
+    {
+      "epoch": 0.45113098369279325,
+      "grad_norm": 1.8748353719711304,
+      "learning_rate": 0.00012055444631033,
+      "loss": 1.5454,
+      "step": 4288
+    },
+    {
+      "epoch": 0.4512361914781694,
+      "grad_norm": 1.4259321689605713,
+      "learning_rate": 0.000120521095376,
+      "loss": 1.6818,
+      "step": 4289
+    },
+    {
+      "epoch": 0.4513413992635455,
+      "grad_norm": 1.2353706359863281,
+      "learning_rate": 0.00012048774205862279,
+      "loss": 2.2079,
+      "step": 4290
+    },
+    {
+      "epoch": 0.45144660704892164,
+      "grad_norm": 1.5690447092056274,
+      "learning_rate": 0.00012045438636207151,
+      "loss": 1.7757,
+      "step": 4291
+    },
+    {
+      "epoch": 0.45155181483429774,
+      "grad_norm": 1.6075197458267212,
+      "learning_rate": 0.00012042102829021973,
+      "loss": 2.0568,
+      "step": 4292
+    },
+    {
+      "epoch": 0.45165702261967383,
+      "grad_norm": 1.3543256521224976,
+      "learning_rate": 0.00012038766784694117,
+      "loss": 1.6417,
+      "step": 4293
+    },
+    {
+      "epoch": 0.45176223040505,
+      "grad_norm": 1.3820946216583252,
+      "learning_rate": 0.00012035430503610988,
+      "loss": 2.0945,
+      "step": 4294
+    },
+    {
+      "epoch": 0.4518674381904261,
+      "grad_norm": 1.327938437461853,
+      "learning_rate": 0.00012032093986160015,
+      "loss": 1.7868,
+      "step": 4295
+    },
+    {
+      "epoch": 0.4519726459758022,
+      "grad_norm": 1.0907479524612427,
+      "learning_rate": 0.0001202875723272866,
+      "loss": 1.7495,
+      "step": 4296
+    },
+    {
+      "epoch": 0.4520778537611783,
+      "grad_norm": 1.2229701280593872,
+      "learning_rate": 0.0001202542024370441,
+      "loss": 2.1307,
+      "step": 4297
+    },
+    {
+      "epoch": 0.45218306154655447,
+      "grad_norm": 0.9999223947525024,
+      "learning_rate": 0.00012022083019474774,
+      "loss": 2.0101,
+      "step": 4298
+    },
+    {
+      "epoch": 0.45228826933193056,
+      "grad_norm": 1.5254671573638916,
+      "learning_rate": 0.00012018745560427298,
+      "loss": 1.6449,
+      "step": 4299
+    },
+    {
+      "epoch": 0.4523934771173067,
+      "grad_norm": 1.047942042350769,
+      "learning_rate": 0.00012015407866949548,
+      "loss": 1.5853,
+      "step": 4300
+    },
+    {
+      "epoch": 0.4524986849026828,
+      "grad_norm": 1.6619633436203003,
+      "learning_rate": 0.0001201206993942912,
+      "loss": 2.1214,
+      "step": 4301
+    },
+    {
+      "epoch": 0.4526038926880589,
+      "grad_norm": 1.228927731513977,
+      "learning_rate": 0.00012008731778253632,
+      "loss": 1.7196,
+      "step": 4302
+    },
+    {
+      "epoch": 0.45270910047343504,
+      "grad_norm": 1.4575189352035522,
+      "learning_rate": 0.00012005393383810737,
+      "loss": 2.0006,
+      "step": 4303
+    },
+    {
+      "epoch": 0.45281430825881114,
+      "grad_norm": 1.3251969814300537,
+      "learning_rate": 0.00012002054756488115,
+      "loss": 1.7704,
+      "step": 4304
+    },
+    {
+      "epoch": 0.4529195160441873,
+      "grad_norm": 1.6667139530181885,
+      "learning_rate": 0.00011998715896673465,
+      "loss": 1.2439,
+      "step": 4305
+    },
+    {
+      "epoch": 0.4530247238295634,
+      "grad_norm": 1.0954786539077759,
+      "learning_rate": 0.0001199537680475452,
+      "loss": 1.827,
+      "step": 4306
+    },
+    {
+      "epoch": 0.45312993161493953,
+      "grad_norm": 1.4709742069244385,
+      "learning_rate": 0.00011992037481119036,
+      "loss": 2.3623,
+      "step": 4307
+    },
+    {
+      "epoch": 0.4532351394003156,
+      "grad_norm": 1.467758059501648,
+      "learning_rate": 0.00011988697926154799,
+      "loss": 1.6894,
+      "step": 4308
+    },
+    {
+      "epoch": 0.4533403471856917,
+      "grad_norm": 1.011841058731079,
+      "learning_rate": 0.00011985358140249621,
+      "loss": 1.4649,
+      "step": 4309
+    },
+    {
+      "epoch": 0.45344555497106787,
+      "grad_norm": 1.7795685529708862,
+      "learning_rate": 0.0001198201812379134,
+      "loss": 1.4143,
+      "step": 4310
+    },
+    {
+      "epoch": 0.45355076275644396,
+      "grad_norm": 1.545654058456421,
+      "learning_rate": 0.00011978677877167822,
+      "loss": 2.1281,
+      "step": 4311
+    },
+    {
+      "epoch": 0.4536559705418201,
+      "grad_norm": 1.2907682657241821,
+      "learning_rate": 0.00011975337400766958,
+      "loss": 1.4049,
+      "step": 4312
+    },
+    {
+      "epoch": 0.4537611783271962,
+      "grad_norm": 1.0812395811080933,
+      "learning_rate": 0.00011971996694976663,
+      "loss": 2.2,
+      "step": 4313
+    },
+    {
+      "epoch": 0.45386638611257235,
+      "grad_norm": 1.4511226415634155,
+      "learning_rate": 0.00011968655760184891,
+      "loss": 2.0119,
+      "step": 4314
+    },
+    {
+      "epoch": 0.45397159389794844,
+      "grad_norm": 1.6760962009429932,
+      "learning_rate": 0.00011965314596779604,
+      "loss": 1.6423,
+      "step": 4315
+    },
+    {
+      "epoch": 0.4540768016833246,
+      "grad_norm": 1.0780059099197388,
+      "learning_rate": 0.00011961973205148804,
+      "loss": 1.5716,
+      "step": 4316
+    },
+    {
+      "epoch": 0.4541820094687007,
+      "grad_norm": 1.6850388050079346,
+      "learning_rate": 0.00011958631585680518,
+      "loss": 1.7895,
+      "step": 4317
+    },
+    {
+      "epoch": 0.4542872172540768,
+      "grad_norm": 1.061856746673584,
+      "learning_rate": 0.00011955289738762796,
+      "loss": 1.8926,
+      "step": 4318
+    },
+    {
+      "epoch": 0.45439242503945293,
+      "grad_norm": 1.7766098976135254,
+      "learning_rate": 0.00011951947664783713,
+      "loss": 1.3322,
+      "step": 4319
+    },
+    {
+      "epoch": 0.454497632824829,
+      "grad_norm": 1.695619821548462,
+      "learning_rate": 0.00011948605364131375,
+      "loss": 1.6927,
+      "step": 4320
+    },
+    {
+      "epoch": 0.45460284061020517,
+      "grad_norm": 1.510094404220581,
+      "learning_rate": 0.00011945262837193915,
+      "loss": 1.9738,
+      "step": 4321
+    },
+    {
+      "epoch": 0.45470804839558127,
+      "grad_norm": 1.4329527616500854,
+      "learning_rate": 0.0001194192008435949,
+      "loss": 1.2781,
+      "step": 4322
+    },
+    {
+      "epoch": 0.4548132561809574,
+      "grad_norm": 1.3772903680801392,
+      "learning_rate": 0.00011938577106016275,
+      "loss": 2.2099,
+      "step": 4323
+    },
+    {
+      "epoch": 0.4549184639663335,
+      "grad_norm": 1.3040528297424316,
+      "learning_rate": 0.00011935233902552485,
+      "loss": 1.9477,
+      "step": 4324
+    },
+    {
+      "epoch": 0.4550236717517096,
+      "grad_norm": 1.0545731782913208,
+      "learning_rate": 0.00011931890474356358,
+      "loss": 1.987,
+      "step": 4325
+    },
+    {
+      "epoch": 0.45512887953708575,
+      "grad_norm": 1.2383321523666382,
+      "learning_rate": 0.00011928546821816149,
+      "loss": 1.0446,
+      "step": 4326
+    },
+    {
+      "epoch": 0.45523408732246184,
+      "grad_norm": 0.9607561230659485,
+      "learning_rate": 0.00011925202945320146,
+      "loss": 1.446,
+      "step": 4327
+    },
+    {
+      "epoch": 0.455339295107838,
+      "grad_norm": 0.9800947308540344,
+      "learning_rate": 0.00011921858845256669,
+      "loss": 1.6101,
+      "step": 4328
+    },
+    {
+      "epoch": 0.4554445028932141,
+      "grad_norm": 1.430310845375061,
+      "learning_rate": 0.00011918514522014051,
+      "loss": 1.9691,
+      "step": 4329
+    },
+    {
+      "epoch": 0.45554971067859024,
+      "grad_norm": 1.495063304901123,
+      "learning_rate": 0.00011915169975980658,
+      "loss": 1.7227,
+      "step": 4330
+    },
+    {
+      "epoch": 0.45565491846396633,
+      "grad_norm": 1.0809447765350342,
+      "learning_rate": 0.00011911825207544885,
+      "loss": 2.2602,
+      "step": 4331
+    },
+    {
+      "epoch": 0.4557601262493425,
+      "grad_norm": 1.0745564699172974,
+      "learning_rate": 0.00011908480217095141,
+      "loss": 1.8813,
+      "step": 4332
+    },
+    {
+      "epoch": 0.4558653340347186,
+      "grad_norm": 1.3847942352294922,
+      "learning_rate": 0.00011905135005019881,
+      "loss": 2.1904,
+      "step": 4333
+    },
+    {
+      "epoch": 0.45597054182009467,
+      "grad_norm": 1.2953932285308838,
+      "learning_rate": 0.0001190178957170756,
+      "loss": 1.7487,
+      "step": 4334
+    },
+    {
+      "epoch": 0.4560757496054708,
+      "grad_norm": 1.1227591037750244,
+      "learning_rate": 0.00011898443917546682,
+      "loss": 1.7218,
+      "step": 4335
+    },
+    {
+      "epoch": 0.4561809573908469,
+      "grad_norm": 2.098883867263794,
+      "learning_rate": 0.00011895098042925763,
+      "loss": 2.1559,
+      "step": 4336
+    },
+    {
+      "epoch": 0.45628616517622306,
+      "grad_norm": 1.4620628356933594,
+      "learning_rate": 0.00011891751948233348,
+      "loss": 1.692,
+      "step": 4337
+    },
+    {
+      "epoch": 0.45639137296159915,
+      "grad_norm": 2.7015371322631836,
+      "learning_rate": 0.00011888405633858009,
+      "loss": 2.0226,
+      "step": 4338
+    },
+    {
+      "epoch": 0.4564965807469753,
+      "grad_norm": 1.333720326423645,
+      "learning_rate": 0.00011885059100188341,
+      "loss": 2.1821,
+      "step": 4339
+    },
+    {
+      "epoch": 0.4566017885323514,
+      "grad_norm": 1.3061823844909668,
+      "learning_rate": 0.00011881712347612965,
+      "loss": 1.7872,
+      "step": 4340
+    },
+    {
+      "epoch": 0.4567069963177275,
+      "grad_norm": 2.075654983520508,
+      "learning_rate": 0.00011878365376520535,
+      "loss": 1.7089,
+      "step": 4341
+    },
+    {
+      "epoch": 0.45681220410310364,
+      "grad_norm": 1.908001184463501,
+      "learning_rate": 0.00011875018187299719,
+      "loss": 1.7904,
+      "step": 4342
+    },
+    {
+      "epoch": 0.45691741188847973,
+      "grad_norm": 1.1886532306671143,
+      "learning_rate": 0.0001187167078033921,
+      "loss": 1.8649,
+      "step": 4343
+    },
+    {
+      "epoch": 0.4570226196738559,
+      "grad_norm": 1.1382215023040771,
+      "learning_rate": 0.00011868323156027742,
+      "loss": 1.6879,
+      "step": 4344
+    },
+    {
+      "epoch": 0.457127827459232,
+      "grad_norm": 1.0881881713867188,
+      "learning_rate": 0.00011864975314754058,
+      "loss": 1.816,
+      "step": 4345
+    },
+    {
+      "epoch": 0.4572330352446081,
+      "grad_norm": 1.6298319101333618,
+      "learning_rate": 0.00011861627256906929,
+      "loss": 1.1575,
+      "step": 4346
+    },
+    {
+      "epoch": 0.4573382430299842,
+      "grad_norm": 1.241731882095337,
+      "learning_rate": 0.00011858278982875157,
+      "loss": 1.7019,
+      "step": 4347
+    },
+    {
+      "epoch": 0.45744345081536036,
+      "grad_norm": 0.9705410599708557,
+      "learning_rate": 0.00011854930493047566,
+      "loss": 1.9562,
+      "step": 4348
+    },
+    {
+      "epoch": 0.45754865860073646,
+      "grad_norm": 1.3326573371887207,
+      "learning_rate": 0.00011851581787813006,
+      "loss": 2.3098,
+      "step": 4349
+    },
+    {
+      "epoch": 0.45765386638611255,
+      "grad_norm": 1.1833465099334717,
+      "learning_rate": 0.00011848232867560352,
+      "loss": 2.0325,
+      "step": 4350
+    },
+    {
+      "epoch": 0.4577590741714887,
+      "grad_norm": 1.501226544380188,
+      "learning_rate": 0.00011844883732678495,
+      "loss": 1.9941,
+      "step": 4351
+    },
+    {
+      "epoch": 0.4578642819568648,
+      "grad_norm": 1.8644059896469116,
+      "learning_rate": 0.00011841534383556372,
+      "loss": 1.6826,
+      "step": 4352
+    },
+    {
+      "epoch": 0.45796948974224094,
+      "grad_norm": 1.2248753309249878,
+      "learning_rate": 0.00011838184820582923,
+      "loss": 1.382,
+      "step": 4353
+    },
+    {
+      "epoch": 0.45807469752761704,
+      "grad_norm": 2.008018970489502,
+      "learning_rate": 0.00011834835044147121,
+      "loss": 1.5946,
+      "step": 4354
+    },
+    {
+      "epoch": 0.4581799053129932,
+      "grad_norm": 1.5521501302719116,
+      "learning_rate": 0.00011831485054637973,
+      "loss": 1.5836,
+      "step": 4355
+    },
+    {
+      "epoch": 0.4582851130983693,
+      "grad_norm": 1.707040548324585,
+      "learning_rate": 0.00011828134852444493,
+      "loss": 1.7551,
+      "step": 4356
+    },
+    {
+      "epoch": 0.4583903208837454,
+      "grad_norm": 1.7729437351226807,
+      "learning_rate": 0.00011824784437955732,
+      "loss": 2.3166,
+      "step": 4357
+    },
+    {
+      "epoch": 0.4584955286691215,
+      "grad_norm": 1.4290671348571777,
+      "learning_rate": 0.0001182143381156076,
+      "loss": 2.0124,
+      "step": 4358
+    },
+    {
+      "epoch": 0.4586007364544976,
+      "grad_norm": 2.462273359298706,
+      "learning_rate": 0.00011818082973648683,
+      "loss": 2.1217,
+      "step": 4359
+    },
+    {
+      "epoch": 0.45870594423987376,
+      "grad_norm": 1.7780591249465942,
+      "learning_rate": 0.00011814731924608616,
+      "loss": 1.4172,
+      "step": 4360
+    },
+    {
+      "epoch": 0.45881115202524986,
+      "grad_norm": 1.9604672193527222,
+      "learning_rate": 0.00011811380664829703,
+      "loss": 1.3744,
+      "step": 4361
+    },
+    {
+      "epoch": 0.458916359810626,
+      "grad_norm": 1.0941351652145386,
+      "learning_rate": 0.00011808029194701122,
+      "loss": 1.9376,
+      "step": 4362
+    },
+    {
+      "epoch": 0.4590215675960021,
+      "grad_norm": 1.587456226348877,
+      "learning_rate": 0.00011804677514612062,
+      "loss": 1.9933,
+      "step": 4363
+    },
+    {
+      "epoch": 0.45912677538137825,
+      "grad_norm": 1.38690984249115,
+      "learning_rate": 0.00011801325624951745,
+      "loss": 2.1653,
+      "step": 4364
+    },
+    {
+      "epoch": 0.45923198316675434,
+      "grad_norm": 1.6729885339736938,
+      "learning_rate": 0.00011797973526109416,
+      "loss": 1.7463,
+      "step": 4365
+    },
+    {
+      "epoch": 0.45933719095213044,
+      "grad_norm": 1.2510261535644531,
+      "learning_rate": 0.00011794621218474345,
+      "loss": 1.6988,
+      "step": 4366
+    },
+    {
+      "epoch": 0.4594423987375066,
+      "grad_norm": 1.0695215463638306,
+      "learning_rate": 0.00011791268702435816,
+      "loss": 2.1912,
+      "step": 4367
+    },
+    {
+      "epoch": 0.4595476065228827,
+      "grad_norm": 1.849915862083435,
+      "learning_rate": 0.00011787915978383151,
+      "loss": 1.7784,
+      "step": 4368
+    },
+    {
+      "epoch": 0.45965281430825883,
+      "grad_norm": 1.144481897354126,
+      "learning_rate": 0.00011784563046705695,
+      "loss": 1.928,
+      "step": 4369
+    },
+    {
+      "epoch": 0.4597580220936349,
+      "grad_norm": 1.9032231569290161,
+      "learning_rate": 0.00011781209907792805,
+      "loss": 1.3965,
+      "step": 4370
+    },
+    {
+      "epoch": 0.45986322987901107,
+      "grad_norm": 1.352990746498108,
+      "learning_rate": 0.00011777856562033876,
+      "loss": 1.987,
+      "step": 4371
+    },
+    {
+      "epoch": 0.45996843766438716,
+      "grad_norm": 1.394627571105957,
+      "learning_rate": 0.00011774503009818316,
+      "loss": 2.3461,
+      "step": 4372
+    },
+    {
+      "epoch": 0.46007364544976326,
+      "grad_norm": 0.8705309629440308,
+      "learning_rate": 0.00011771149251535569,
+      "loss": 2.7158,
+      "step": 4373
+    },
+    {
+      "epoch": 0.4601788532351394,
+      "grad_norm": 1.0310938358306885,
+      "learning_rate": 0.0001176779528757509,
+      "loss": 1.9135,
+      "step": 4374
+    },
+    {
+      "epoch": 0.4602840610205155,
+      "grad_norm": 1.4547808170318604,
+      "learning_rate": 0.00011764441118326364,
+      "loss": 1.8487,
+      "step": 4375
+    },
+    {
+      "epoch": 0.46038926880589165,
+      "grad_norm": 2.0553081035614014,
+      "learning_rate": 0.00011761086744178902,
+      "loss": 1.698,
+      "step": 4376
+    },
+    {
+      "epoch": 0.46049447659126774,
+      "grad_norm": 1.7115883827209473,
+      "learning_rate": 0.00011757732165522237,
+      "loss": 2.4097,
+      "step": 4377
+    },
+    {
+      "epoch": 0.4605996843766439,
+      "grad_norm": 1.6055662631988525,
+      "learning_rate": 0.00011754377382745922,
+      "loss": 1.6749,
+      "step": 4378
+    },
+    {
+      "epoch": 0.46070489216202,
+      "grad_norm": 1.6005985736846924,
+      "learning_rate": 0.00011751022396239539,
+      "loss": 2.0828,
+      "step": 4379
+    },
+    {
+      "epoch": 0.46081009994739613,
+      "grad_norm": 1.4807254076004028,
+      "learning_rate": 0.00011747667206392691,
+      "loss": 1.7367,
+      "step": 4380
+    },
+    {
+      "epoch": 0.46091530773277223,
+      "grad_norm": 1.8019040822982788,
+      "learning_rate": 0.00011744311813595006,
+      "loss": 2.1629,
+      "step": 4381
+    },
+    {
+      "epoch": 0.4610205155181483,
+      "grad_norm": 1.1676995754241943,
+      "learning_rate": 0.00011740956218236132,
+      "loss": 1.6825,
+      "step": 4382
+    },
+    {
+      "epoch": 0.46112572330352447,
+      "grad_norm": 1.9756308794021606,
+      "learning_rate": 0.00011737600420705748,
+      "loss": 2.0505,
+      "step": 4383
+    },
+    {
+      "epoch": 0.46123093108890056,
+      "grad_norm": 1.3423806428909302,
+      "learning_rate": 0.00011734244421393548,
+      "loss": 2.0603,
+      "step": 4384
+    },
+    {
+      "epoch": 0.4613361388742767,
+      "grad_norm": 1.3737515211105347,
+      "learning_rate": 0.00011730888220689251,
+      "loss": 1.5806,
+      "step": 4385
+    },
+    {
+      "epoch": 0.4614413466596528,
+      "grad_norm": 1.299750804901123,
+      "learning_rate": 0.0001172753181898261,
+      "loss": 1.4868,
+      "step": 4386
+    },
+    {
+      "epoch": 0.46154655444502896,
+      "grad_norm": 1.3489303588867188,
+      "learning_rate": 0.00011724175216663384,
+      "loss": 1.6957,
+      "step": 4387
+    },
+    {
+      "epoch": 0.46165176223040505,
+      "grad_norm": 1.3841158151626587,
+      "learning_rate": 0.00011720818414121368,
+      "loss": 1.5918,
+      "step": 4388
+    },
+    {
+      "epoch": 0.46175697001578114,
+      "grad_norm": 1.4252556562423706,
+      "learning_rate": 0.00011717461411746378,
+      "loss": 1.7677,
+      "step": 4389
+    },
+    {
+      "epoch": 0.4618621778011573,
+      "grad_norm": 1.7873703241348267,
+      "learning_rate": 0.0001171410420992825,
+      "loss": 1.6604,
+      "step": 4390
+    },
+    {
+      "epoch": 0.4619673855865334,
+      "grad_norm": 1.825914978981018,
+      "learning_rate": 0.00011710746809056841,
+      "loss": 1.7022,
+      "step": 4391
+    },
+    {
+      "epoch": 0.46207259337190953,
+      "grad_norm": 1.4188059568405151,
+      "learning_rate": 0.00011707389209522039,
+      "loss": 2.1184,
+      "step": 4392
+    },
+    {
+      "epoch": 0.46217780115728563,
+      "grad_norm": 1.4496338367462158,
+      "learning_rate": 0.0001170403141171375,
+      "loss": 2.0205,
+      "step": 4393
+    },
+    {
+      "epoch": 0.4622830089426618,
+      "grad_norm": 2.31512713432312,
+      "learning_rate": 0.00011700673416021908,
+      "loss": 1.822,
+      "step": 4394
+    },
+    {
+      "epoch": 0.46238821672803787,
+      "grad_norm": 1.1979174613952637,
+      "learning_rate": 0.00011697315222836458,
+      "loss": 2.3128,
+      "step": 4395
+    },
+    {
+      "epoch": 0.462493424513414,
+      "grad_norm": 1.3453646898269653,
+      "learning_rate": 0.00011693956832547384,
+      "loss": 2.0374,
+      "step": 4396
+    },
+    {
+      "epoch": 0.4625986322987901,
+      "grad_norm": 1.1278789043426514,
+      "learning_rate": 0.0001169059824554468,
+      "loss": 1.5538,
+      "step": 4397
+    },
+    {
+      "epoch": 0.4627038400841662,
+      "grad_norm": 1.6684142351150513,
+      "learning_rate": 0.00011687239462218369,
+      "loss": 2.0152,
+      "step": 4398
+    },
+    {
+      "epoch": 0.46280904786954236,
+      "grad_norm": 1.2076631784439087,
+      "learning_rate": 0.00011683880482958493,
+      "loss": 1.9577,
+      "step": 4399
+    },
+    {
+      "epoch": 0.46291425565491845,
+      "grad_norm": 1.3028810024261475,
+      "learning_rate": 0.00011680521308155124,
+      "loss": 1.413,
+      "step": 4400
+    },
+    {
+      "epoch": 0.4630194634402946,
+      "grad_norm": 1.2423094511032104,
+      "learning_rate": 0.00011677161938198348,
+      "loss": 1.7865,
+      "step": 4401
+    },
+    {
+      "epoch": 0.4631246712256707,
+      "grad_norm": 1.2606873512268066,
+      "learning_rate": 0.0001167380237347828,
+      "loss": 1.9457,
+      "step": 4402
+    },
+    {
+      "epoch": 0.46322987901104684,
+      "grad_norm": 1.7899268865585327,
+      "learning_rate": 0.00011670442614385053,
+      "loss": 2.2358,
+      "step": 4403
+    },
+    {
+      "epoch": 0.46333508679642293,
+      "grad_norm": 1.2076514959335327,
+      "learning_rate": 0.00011667082661308826,
+      "loss": 1.7803,
+      "step": 4404
+    },
+    {
+      "epoch": 0.46344029458179903,
+      "grad_norm": 1.6006786823272705,
+      "learning_rate": 0.00011663722514639778,
+      "loss": 1.696,
+      "step": 4405
+    },
+    {
+      "epoch": 0.4635455023671752,
+      "grad_norm": 1.0841476917266846,
+      "learning_rate": 0.00011660362174768114,
+      "loss": 1.8837,
+      "step": 4406
+    },
+    {
+      "epoch": 0.46365071015255127,
+      "grad_norm": 0.9759506583213806,
+      "learning_rate": 0.0001165700164208406,
+      "loss": 1.4346,
+      "step": 4407
+    },
+    {
+      "epoch": 0.4637559179379274,
+      "grad_norm": 1.5411173105239868,
+      "learning_rate": 0.00011653640916977861,
+      "loss": 1.925,
+      "step": 4408
+    },
+    {
+      "epoch": 0.4638611257233035,
+      "grad_norm": 1.1252624988555908,
+      "learning_rate": 0.00011650279999839787,
+      "loss": 1.8466,
+      "step": 4409
+    },
+    {
+      "epoch": 0.46396633350867966,
+      "grad_norm": 1.3989907503128052,
+      "learning_rate": 0.00011646918891060127,
+      "loss": 1.8891,
+      "step": 4410
+    },
+    {
+      "epoch": 0.46407154129405576,
+      "grad_norm": 1.7185852527618408,
+      "learning_rate": 0.00011643557591029206,
+      "loss": 1.7989,
+      "step": 4411
+    },
+    {
+      "epoch": 0.4641767490794319,
+      "grad_norm": 1.037192702293396,
+      "learning_rate": 0.0001164019610013735,
+      "loss": 1.3982,
+      "step": 4412
+    },
+    {
+      "epoch": 0.464281956864808,
+      "grad_norm": 1.2058227062225342,
+      "learning_rate": 0.00011636834418774922,
+      "loss": 1.605,
+      "step": 4413
+    },
+    {
+      "epoch": 0.4643871646501841,
+      "grad_norm": 1.8014076948165894,
+      "learning_rate": 0.00011633472547332305,
+      "loss": 1.3505,
+      "step": 4414
+    },
+    {
+      "epoch": 0.46449237243556024,
+      "grad_norm": 1.513741135597229,
+      "learning_rate": 0.00011630110486199899,
+      "loss": 1.6722,
+      "step": 4415
+    },
+    {
+      "epoch": 0.46459758022093633,
+      "grad_norm": 1.3467214107513428,
+      "learning_rate": 0.00011626748235768128,
+      "loss": 1.9053,
+      "step": 4416
+    },
+    {
+      "epoch": 0.4647027880063125,
+      "grad_norm": 1.5441854000091553,
+      "learning_rate": 0.00011623385796427442,
+      "loss": 1.7715,
+      "step": 4417
+    },
+    {
+      "epoch": 0.4648079957916886,
+      "grad_norm": 1.012117624282837,
+      "learning_rate": 0.00011620023168568311,
+      "loss": 1.9828,
+      "step": 4418
+    },
+    {
+      "epoch": 0.4649132035770647,
+      "grad_norm": 1.2613452672958374,
+      "learning_rate": 0.00011616660352581225,
+      "loss": 1.996,
+      "step": 4419
+    },
+    {
+      "epoch": 0.4650184113624408,
+      "grad_norm": 2.198415756225586,
+      "learning_rate": 0.00011613297348856693,
+      "loss": 2.229,
+      "step": 4420
+    },
+    {
+      "epoch": 0.4651236191478169,
+      "grad_norm": 1.369099736213684,
+      "learning_rate": 0.00011609934157785251,
+      "loss": 1.7323,
+      "step": 4421
+    },
+    {
+      "epoch": 0.46522882693319306,
+      "grad_norm": 1.4710428714752197,
+      "learning_rate": 0.00011606570779757461,
+      "loss": 1.7945,
+      "step": 4422
+    },
+    {
+      "epoch": 0.46533403471856916,
+      "grad_norm": 2.027179479598999,
+      "learning_rate": 0.00011603207215163894,
+      "loss": 1.0333,
+      "step": 4423
+    },
+    {
+      "epoch": 0.4654392425039453,
+      "grad_norm": 1.1320077180862427,
+      "learning_rate": 0.00011599843464395151,
+      "loss": 1.8657,
+      "step": 4424
+    },
+    {
+      "epoch": 0.4655444502893214,
+      "grad_norm": 1.3532216548919678,
+      "learning_rate": 0.00011596479527841859,
+      "loss": 2.1035,
+      "step": 4425
+    },
+    {
+      "epoch": 0.46564965807469755,
+      "grad_norm": 1.3773186206817627,
+      "learning_rate": 0.00011593115405894652,
+      "loss": 1.9023,
+      "step": 4426
+    },
+    {
+      "epoch": 0.46575486586007364,
+      "grad_norm": 1.1166491508483887,
+      "learning_rate": 0.00011589751098944202,
+      "loss": 1.5557,
+      "step": 4427
+    },
+    {
+      "epoch": 0.4658600736454498,
+      "grad_norm": 1.6609245538711548,
+      "learning_rate": 0.0001158638660738119,
+      "loss": 2.003,
+      "step": 4428
+    },
+    {
+      "epoch": 0.4659652814308259,
+      "grad_norm": 1.4807888269424438,
+      "learning_rate": 0.00011583021931596325,
+      "loss": 2.0677,
+      "step": 4429
+    },
+    {
+      "epoch": 0.466070489216202,
+      "grad_norm": 1.3660595417022705,
+      "learning_rate": 0.0001157965707198034,
+      "loss": 1.838,
+      "step": 4430
+    },
+    {
+      "epoch": 0.4661756970015781,
+      "grad_norm": 1.6003875732421875,
+      "learning_rate": 0.00011576292028923976,
+      "loss": 1.789,
+      "step": 4431
+    },
+    {
+      "epoch": 0.4662809047869542,
+      "grad_norm": 1.518566370010376,
+      "learning_rate": 0.00011572926802818011,
+      "loss": 2.4484,
+      "step": 4432
+    },
+    {
+      "epoch": 0.46638611257233037,
+      "grad_norm": 2.0211031436920166,
+      "learning_rate": 0.00011569561394053236,
+      "loss": 1.7331,
+      "step": 4433
+    },
+    {
+      "epoch": 0.46649132035770646,
+      "grad_norm": 1.0467982292175293,
+      "learning_rate": 0.00011566195803020464,
+      "loss": 1.5526,
+      "step": 4434
+    },
+    {
+      "epoch": 0.4665965281430826,
+      "grad_norm": 1.9077860116958618,
+      "learning_rate": 0.00011562830030110532,
+      "loss": 2.1115,
+      "step": 4435
+    },
+    {
+      "epoch": 0.4667017359284587,
+      "grad_norm": 1.1763733625411987,
+      "learning_rate": 0.00011559464075714292,
+      "loss": 1.5898,
+      "step": 4436
+    },
+    {
+      "epoch": 0.4668069437138348,
+      "grad_norm": 1.084130048751831,
+      "learning_rate": 0.00011556097940222628,
+      "loss": 2.0153,
+      "step": 4437
+    },
+    {
+      "epoch": 0.46691215149921095,
+      "grad_norm": 0.8493435978889465,
+      "learning_rate": 0.00011552731624026432,
+      "loss": 1.9469,
+      "step": 4438
+    },
+    {
+      "epoch": 0.46701735928458704,
+      "grad_norm": 2.182161331176758,
+      "learning_rate": 0.00011549365127516627,
+      "loss": 2.1621,
+      "step": 4439
+    },
+    {
+      "epoch": 0.4671225670699632,
+      "grad_norm": 1.1532615423202515,
+      "learning_rate": 0.0001154599845108415,
+      "loss": 2.1096,
+      "step": 4440
+    },
+    {
+      "epoch": 0.4672277748553393,
+      "grad_norm": 0.9221954345703125,
+      "learning_rate": 0.00011542631595119965,
+      "loss": 1.8867,
+      "step": 4441
+    },
+    {
+      "epoch": 0.46733298264071543,
+      "grad_norm": 1.4042004346847534,
+      "learning_rate": 0.00011539264560015052,
+      "loss": 1.7898,
+      "step": 4442
+    },
+    {
+      "epoch": 0.4674381904260915,
+      "grad_norm": 1.2298986911773682,
+      "learning_rate": 0.00011535897346160416,
+      "loss": 1.8017,
+      "step": 4443
+    },
+    {
+      "epoch": 0.4675433982114677,
+      "grad_norm": 0.9590116739273071,
+      "learning_rate": 0.00011532529953947075,
+      "loss": 1.823,
+      "step": 4444
+    },
+    {
+      "epoch": 0.46764860599684377,
+      "grad_norm": 1.2512177228927612,
+      "learning_rate": 0.00011529162383766079,
+      "loss": 1.6104,
+      "step": 4445
+    },
+    {
+      "epoch": 0.46775381378221986,
+      "grad_norm": 1.2496157884597778,
+      "learning_rate": 0.00011525794636008491,
+      "loss": 1.4588,
+      "step": 4446
+    },
+    {
+      "epoch": 0.467859021567596,
+      "grad_norm": 1.6144243478775024,
+      "learning_rate": 0.00011522426711065397,
+      "loss": 1.9968,
+      "step": 4447
+    },
+    {
+      "epoch": 0.4679642293529721,
+      "grad_norm": 1.1202630996704102,
+      "learning_rate": 0.000115190586093279,
+      "loss": 2.2607,
+      "step": 4448
+    },
+    {
+      "epoch": 0.46806943713834825,
+      "grad_norm": 1.414974570274353,
+      "learning_rate": 0.00011515690331187133,
+      "loss": 1.7647,
+      "step": 4449
+    },
+    {
+      "epoch": 0.46817464492372435,
+      "grad_norm": 1.3937184810638428,
+      "learning_rate": 0.00011512321877034234,
+      "loss": 1.5259,
+      "step": 4450
+    },
+    {
+      "epoch": 0.4682798527091005,
+      "grad_norm": 1.7348077297210693,
+      "learning_rate": 0.00011508953247260379,
+      "loss": 1.8668,
+      "step": 4451
+    },
+    {
+      "epoch": 0.4683850604944766,
+      "grad_norm": 1.4032397270202637,
+      "learning_rate": 0.00011505584442256752,
+      "loss": 1.1793,
+      "step": 4452
+    },
+    {
+      "epoch": 0.4684902682798527,
+      "grad_norm": 1.652096152305603,
+      "learning_rate": 0.00011502215462414561,
+      "loss": 1.5654,
+      "step": 4453
+    },
+    {
+      "epoch": 0.46859547606522883,
+      "grad_norm": 1.547971487045288,
+      "learning_rate": 0.00011498846308125033,
+      "loss": 1.9391,
+      "step": 4454
+    },
+    {
+      "epoch": 0.4687006838506049,
+      "grad_norm": 1.1444284915924072,
+      "learning_rate": 0.00011495476979779418,
+      "loss": 1.7784,
+      "step": 4455
+    },
+    {
+      "epoch": 0.4688058916359811,
+      "grad_norm": 1.4032714366912842,
+      "learning_rate": 0.00011492107477768992,
+      "loss": 1.7778,
+      "step": 4456
+    },
+    {
+      "epoch": 0.46891109942135717,
+      "grad_norm": 1.3546231985092163,
+      "learning_rate": 0.00011488737802485033,
+      "loss": 1.4267,
+      "step": 4457
+    },
+    {
+      "epoch": 0.4690163072067333,
+      "grad_norm": 1.4766918420791626,
+      "learning_rate": 0.00011485367954318856,
+      "loss": 1.6944,
+      "step": 4458
+    },
+    {
+      "epoch": 0.4691215149921094,
+      "grad_norm": 1.9072582721710205,
+      "learning_rate": 0.0001148199793366179,
+      "loss": 2.0898,
+      "step": 4459
+    },
+    {
+      "epoch": 0.46922672277748556,
+      "grad_norm": 1.6005427837371826,
+      "learning_rate": 0.00011478627740905183,
+      "loss": 1.5538,
+      "step": 4460
+    },
+    {
+      "epoch": 0.46933193056286165,
+      "grad_norm": 1.1679853200912476,
+      "learning_rate": 0.00011475257376440405,
+      "loss": 1.6107,
+      "step": 4461
+    },
+    {
+      "epoch": 0.46943713834823775,
+      "grad_norm": 1.3431743383407593,
+      "learning_rate": 0.0001147188684065885,
+      "loss": 1.99,
+      "step": 4462
+    },
+    {
+      "epoch": 0.4695423461336139,
+      "grad_norm": 1.37478768825531,
+      "learning_rate": 0.00011468516133951921,
+      "loss": 1.9783,
+      "step": 4463
+    },
+    {
+      "epoch": 0.46964755391899,
+      "grad_norm": 1.1095014810562134,
+      "learning_rate": 0.00011465145256711048,
+      "loss": 1.9507,
+      "step": 4464
+    },
+    {
+      "epoch": 0.46975276170436614,
+      "grad_norm": 1.8217538595199585,
+      "learning_rate": 0.0001146177420932768,
+      "loss": 2.4746,
+      "step": 4465
+    },
+    {
+      "epoch": 0.46985796948974223,
+      "grad_norm": 1.3353768587112427,
+      "learning_rate": 0.00011458402992193289,
+      "loss": 1.7805,
+      "step": 4466
+    },
+    {
+      "epoch": 0.4699631772751184,
+      "grad_norm": 1.624212384223938,
+      "learning_rate": 0.0001145503160569936,
+      "loss": 1.928,
+      "step": 4467
+    },
+    {
+      "epoch": 0.4700683850604945,
+      "grad_norm": 1.4001796245574951,
+      "learning_rate": 0.00011451660050237401,
+      "loss": 1.7054,
+      "step": 4468
+    },
+    {
+      "epoch": 0.47017359284587057,
+      "grad_norm": 1.3089070320129395,
+      "learning_rate": 0.00011448288326198939,
+      "loss": 2.2208,
+      "step": 4469
+    },
+    {
+      "epoch": 0.4702788006312467,
+      "grad_norm": 1.7245391607284546,
+      "learning_rate": 0.00011444916433975528,
+      "loss": 1.6884,
+      "step": 4470
+    },
+    {
+      "epoch": 0.4703840084166228,
+      "grad_norm": 1.450230360031128,
+      "learning_rate": 0.00011441544373958725,
+      "loss": 1.5062,
+      "step": 4471
+    },
+    {
+      "epoch": 0.47048921620199896,
+      "grad_norm": 1.8866242170333862,
+      "learning_rate": 0.00011438172146540123,
+      "loss": 1.6767,
+      "step": 4472
+    },
+    {
+      "epoch": 0.47059442398737505,
+      "grad_norm": 1.5962532758712769,
+      "learning_rate": 0.00011434799752111324,
+      "loss": 1.8297,
+      "step": 4473
+    },
+    {
+      "epoch": 0.4706996317727512,
+      "grad_norm": 1.6113159656524658,
+      "learning_rate": 0.00011431427191063957,
+      "loss": 1.84,
+      "step": 4474
+    },
+    {
+      "epoch": 0.4708048395581273,
+      "grad_norm": 1.374487042427063,
+      "learning_rate": 0.00011428054463789661,
+      "loss": 1.865,
+      "step": 4475
+    },
+    {
+      "epoch": 0.47091004734350345,
+      "grad_norm": 2.844097137451172,
+      "learning_rate": 0.00011424681570680105,
+      "loss": 1.3199,
+      "step": 4476
+    },
+    {
+      "epoch": 0.47101525512887954,
+      "grad_norm": 1.1103769540786743,
+      "learning_rate": 0.00011421308512126969,
+      "loss": 2.3692,
+      "step": 4477
+    },
+    {
+      "epoch": 0.47112046291425563,
+      "grad_norm": 1.4509694576263428,
+      "learning_rate": 0.00011417935288521955,
+      "loss": 2.0409,
+      "step": 4478
+    },
+    {
+      "epoch": 0.4712256706996318,
+      "grad_norm": 1.3323218822479248,
+      "learning_rate": 0.00011414561900256784,
+      "loss": 2.2169,
+      "step": 4479
+    },
+    {
+      "epoch": 0.4713308784850079,
+      "grad_norm": 1.0157517194747925,
+      "learning_rate": 0.00011411188347723198,
+      "loss": 2.1209,
+      "step": 4480
+    },
+    {
+      "epoch": 0.471436086270384,
+      "grad_norm": 1.2026599645614624,
+      "learning_rate": 0.00011407814631312957,
+      "loss": 1.934,
+      "step": 4481
+    },
+    {
+      "epoch": 0.4715412940557601,
+      "grad_norm": 1.1034396886825562,
+      "learning_rate": 0.00011404440751417838,
+      "loss": 1.7275,
+      "step": 4482
+    },
+    {
+      "epoch": 0.47164650184113627,
+      "grad_norm": 0.9263768196105957,
+      "learning_rate": 0.00011401066708429641,
+      "loss": 2.0278,
+      "step": 4483
+    },
+    {
+      "epoch": 0.47175170962651236,
+      "grad_norm": 1.4000307321548462,
+      "learning_rate": 0.0001139769250274018,
+      "loss": 2.0002,
+      "step": 4484
+    },
+    {
+      "epoch": 0.47185691741188845,
+      "grad_norm": 1.6272485256195068,
+      "learning_rate": 0.0001139431813474129,
+      "loss": 1.8365,
+      "step": 4485
+    },
+    {
+      "epoch": 0.4719621251972646,
+      "grad_norm": 1.3752802610397339,
+      "learning_rate": 0.00011390943604824826,
+      "loss": 2.1863,
+      "step": 4486
+    },
+    {
+      "epoch": 0.4720673329826407,
+      "grad_norm": 1.1118308305740356,
+      "learning_rate": 0.00011387568913382664,
+      "loss": 2.1644,
+      "step": 4487
+    },
+    {
+      "epoch": 0.47217254076801685,
+      "grad_norm": 1.0211619138717651,
+      "learning_rate": 0.00011384194060806692,
+      "loss": 1.9979,
+      "step": 4488
+    },
+    {
+      "epoch": 0.47227774855339294,
+      "grad_norm": 1.477390170097351,
+      "learning_rate": 0.0001138081904748882,
+      "loss": 1.9993,
+      "step": 4489
+    },
+    {
+      "epoch": 0.4723829563387691,
+      "grad_norm": 0.9481652975082397,
+      "learning_rate": 0.00011377443873820981,
+      "loss": 1.886,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4724881641241452,
+      "grad_norm": 1.2719368934631348,
+      "learning_rate": 0.00011374068540195122,
+      "loss": 1.6373,
+      "step": 4491
+    },
+    {
+      "epoch": 0.47259337190952133,
+      "grad_norm": 1.15910005569458,
+      "learning_rate": 0.00011370693047003205,
+      "loss": 2.0742,
+      "step": 4492
+    },
+    {
+      "epoch": 0.4726985796948974,
+      "grad_norm": 1.1939606666564941,
+      "learning_rate": 0.00011367317394637218,
+      "loss": 1.4249,
+      "step": 4493
+    },
+    {
+      "epoch": 0.4728037874802735,
+      "grad_norm": 1.500286340713501,
+      "learning_rate": 0.00011363941583489171,
+      "loss": 2.1489,
+      "step": 4494
+    },
+    {
+      "epoch": 0.47290899526564967,
+      "grad_norm": 1.2122026681900024,
+      "learning_rate": 0.00011360565613951073,
+      "loss": 2.0753,
+      "step": 4495
+    },
+    {
+      "epoch": 0.47301420305102576,
+      "grad_norm": 1.3212809562683105,
+      "learning_rate": 0.0001135718948641497,
+      "loss": 1.4172,
+      "step": 4496
+    },
+    {
+      "epoch": 0.4731194108364019,
+      "grad_norm": 1.0483636856079102,
+      "learning_rate": 0.00011353813201272921,
+      "loss": 1.5167,
+      "step": 4497
+    },
+    {
+      "epoch": 0.473224618621778,
+      "grad_norm": 0.8720149993896484,
+      "learning_rate": 0.00011350436758917007,
+      "loss": 1.6892,
+      "step": 4498
+    },
+    {
+      "epoch": 0.47332982640715415,
+      "grad_norm": 1.857450008392334,
+      "learning_rate": 0.00011347060159739315,
+      "loss": 2.0008,
+      "step": 4499
+    },
+    {
+      "epoch": 0.47343503419253025,
+      "grad_norm": 1.0890053510665894,
+      "learning_rate": 0.00011343683404131964,
+      "loss": 1.5298,
+      "step": 4500
+    },
+    {
+      "epoch": 0.47354024197790634,
+      "grad_norm": 1.181207299232483,
+      "learning_rate": 0.00011340306492487084,
+      "loss": 1.6269,
+      "step": 4501
+    },
+    {
+      "epoch": 0.4736454497632825,
+      "grad_norm": 1.423409104347229,
+      "learning_rate": 0.00011336929425196826,
+      "loss": 2.2304,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4737506575486586,
+      "grad_norm": 1.1393976211547852,
+      "learning_rate": 0.00011333552202653353,
+      "loss": 1.6851,
+      "step": 4503
+    },
+    {
+      "epoch": 0.47385586533403473,
+      "grad_norm": 1.761922001838684,
+      "learning_rate": 0.00011330174825248857,
+      "loss": 1.5886,
+      "step": 4504
+    },
+    {
+      "epoch": 0.4739610731194108,
+      "grad_norm": 2.4872066974639893,
+      "learning_rate": 0.0001132679729337554,
+      "loss": 1.8363,
+      "step": 4505
+    },
+    {
+      "epoch": 0.474066280904787,
+      "grad_norm": 2.0420877933502197,
+      "learning_rate": 0.00011323419607425618,
+      "loss": 1.7054,
+      "step": 4506
+    },
+    {
+      "epoch": 0.47417148869016307,
+      "grad_norm": 1.310332179069519,
+      "learning_rate": 0.00011320041767791336,
+      "loss": 2.296,
+      "step": 4507
+    },
+    {
+      "epoch": 0.4742766964755392,
+      "grad_norm": 1.570314884185791,
+      "learning_rate": 0.00011316663774864951,
+      "loss": 1.3577,
+      "step": 4508
+    },
+    {
+      "epoch": 0.4743819042609153,
+      "grad_norm": 1.4115073680877686,
+      "learning_rate": 0.00011313285629038737,
+      "loss": 1.744,
+      "step": 4509
+    },
+    {
+      "epoch": 0.4744871120462914,
+      "grad_norm": 1.366669774055481,
+      "learning_rate": 0.00011309907330704988,
+      "loss": 1.6847,
+      "step": 4510
+    },
+    {
+      "epoch": 0.47459231983166755,
+      "grad_norm": 1.2731913328170776,
+      "learning_rate": 0.00011306528880256016,
+      "loss": 2.1791,
+      "step": 4511
+    },
+    {
+      "epoch": 0.47469752761704365,
+      "grad_norm": 0.9359650015830994,
+      "learning_rate": 0.00011303150278084145,
+      "loss": 1.6904,
+      "step": 4512
+    },
+    {
+      "epoch": 0.4748027354024198,
+      "grad_norm": 1.9053103923797607,
+      "learning_rate": 0.00011299771524581722,
+      "loss": 1.1146,
+      "step": 4513
+    },
+    {
+      "epoch": 0.4749079431877959,
+      "grad_norm": 1.60822331905365,
+      "learning_rate": 0.00011296392620141114,
+      "loss": 1.47,
+      "step": 4514
+    },
+    {
+      "epoch": 0.47501315097317204,
+      "grad_norm": 1.3728163242340088,
+      "learning_rate": 0.00011293013565154702,
+      "loss": 2.0799,
+      "step": 4515
+    },
+    {
+      "epoch": 0.47511835875854813,
+      "grad_norm": 1.5891927480697632,
+      "learning_rate": 0.0001128963436001488,
+      "loss": 2.0376,
+      "step": 4516
+    },
+    {
+      "epoch": 0.4752235665439242,
+      "grad_norm": 1.4155621528625488,
+      "learning_rate": 0.00011286255005114065,
+      "loss": 1.6403,
+      "step": 4517
+    },
+    {
+      "epoch": 0.4753287743293004,
+      "grad_norm": 1.1898889541625977,
+      "learning_rate": 0.00011282875500844694,
+      "loss": 2.0779,
+      "step": 4518
+    },
+    {
+      "epoch": 0.47543398211467647,
+      "grad_norm": 1.205228328704834,
+      "learning_rate": 0.00011279495847599216,
+      "loss": 2.0405,
+      "step": 4519
+    },
+    {
+      "epoch": 0.4755391899000526,
+      "grad_norm": 1.6402561664581299,
+      "learning_rate": 0.00011276116045770096,
+      "loss": 1.8382,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4756443976854287,
+      "grad_norm": 1.9118852615356445,
+      "learning_rate": 0.00011272736095749823,
+      "loss": 2.0503,
+      "step": 4521
+    },
+    {
+      "epoch": 0.47574960547080486,
+      "grad_norm": 1.6586589813232422,
+      "learning_rate": 0.00011269355997930899,
+      "loss": 1.7323,
+      "step": 4522
+    },
+    {
+      "epoch": 0.47585481325618095,
+      "grad_norm": 1.5802953243255615,
+      "learning_rate": 0.00011265975752705842,
+      "loss": 2.3248,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4759600210415571,
+      "grad_norm": 1.256369709968567,
+      "learning_rate": 0.0001126259536046719,
+      "loss": 1.5487,
+      "step": 4524
+    },
+    {
+      "epoch": 0.4760652288269332,
+      "grad_norm": 1.2171026468276978,
+      "learning_rate": 0.00011259214821607496,
+      "loss": 1.3623,
+      "step": 4525
+    },
+    {
+      "epoch": 0.4761704366123093,
+      "grad_norm": 1.427194595336914,
+      "learning_rate": 0.00011255834136519334,
+      "loss": 1.1506,
+      "step": 4526
+    },
+    {
+      "epoch": 0.47627564439768544,
+      "grad_norm": 1.2046290636062622,
+      "learning_rate": 0.00011252453305595285,
+      "loss": 1.3518,
+      "step": 4527
+    },
+    {
+      "epoch": 0.47638085218306153,
+      "grad_norm": 1.498401403427124,
+      "learning_rate": 0.00011249072329227959,
+      "loss": 2.0285,
+      "step": 4528
+    },
+    {
+      "epoch": 0.4764860599684377,
+      "grad_norm": 1.4994467496871948,
+      "learning_rate": 0.00011245691207809978,
+      "loss": 1.9723,
+      "step": 4529
+    },
+    {
+      "epoch": 0.4765912677538138,
+      "grad_norm": 0.8767755627632141,
+      "learning_rate": 0.00011242309941733978,
+      "loss": 1.7982,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4766964755391899,
+      "grad_norm": 1.7396360635757446,
+      "learning_rate": 0.00011238928531392614,
+      "loss": 2.3044,
+      "step": 4531
+    },
+    {
+      "epoch": 0.476801683324566,
+      "grad_norm": 1.3658801317214966,
+      "learning_rate": 0.00011235546977178562,
+      "loss": 1.4995,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4769068911099421,
+      "grad_norm": 1.1424037218093872,
+      "learning_rate": 0.00011232165279484506,
+      "loss": 1.6863,
+      "step": 4533
+    },
+    {
+      "epoch": 0.47701209889531826,
+      "grad_norm": 1.8249849081039429,
+      "learning_rate": 0.00011228783438703154,
+      "loss": 1.6042,
+      "step": 4534
+    },
+    {
+      "epoch": 0.47711730668069435,
+      "grad_norm": 1.236474871635437,
+      "learning_rate": 0.0001122540145522723,
+      "loss": 2.4466,
+      "step": 4535
+    },
+    {
+      "epoch": 0.4772225144660705,
+      "grad_norm": 1.7508653402328491,
+      "learning_rate": 0.00011222019329449467,
+      "loss": 1.9367,
+      "step": 4536
+    },
+    {
+      "epoch": 0.4773277222514466,
+      "grad_norm": 1.29315185546875,
+      "learning_rate": 0.00011218637061762624,
+      "loss": 1.9788,
+      "step": 4537
+    },
+    {
+      "epoch": 0.47743293003682274,
+      "grad_norm": 1.5131714344024658,
+      "learning_rate": 0.00011215254652559472,
+      "loss": 2.1485,
+      "step": 4538
+    },
+    {
+      "epoch": 0.47753813782219884,
+      "grad_norm": 1.8211942911148071,
+      "learning_rate": 0.00011211872102232801,
+      "loss": 1.3127,
+      "step": 4539
+    },
+    {
+      "epoch": 0.477643345607575,
+      "grad_norm": 1.5046416521072388,
+      "learning_rate": 0.0001120848941117541,
+      "loss": 1.6588,
+      "step": 4540
+    },
+    {
+      "epoch": 0.4777485533929511,
+      "grad_norm": 2.2012786865234375,
+      "learning_rate": 0.00011205106579780125,
+      "loss": 2.1727,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4778537611783272,
+      "grad_norm": 2.173950433731079,
+      "learning_rate": 0.00011201723608439778,
+      "loss": 2.1504,
+      "step": 4542
+    },
+    {
+      "epoch": 0.4779589689637033,
+      "grad_norm": 1.459008812904358,
+      "learning_rate": 0.00011198340497547231,
+      "loss": 2.1339,
+      "step": 4543
+    },
+    {
+      "epoch": 0.4780641767490794,
+      "grad_norm": 1.1652424335479736,
+      "learning_rate": 0.00011194957247495344,
+      "loss": 1.6663,
+      "step": 4544
+    },
+    {
+      "epoch": 0.47816938453445557,
+      "grad_norm": 1.2521672248840332,
+      "learning_rate": 0.00011191573858677007,
+      "loss": 2.2594,
+      "step": 4545
+    },
+    {
+      "epoch": 0.47827459231983166,
+      "grad_norm": 1.4206370115280151,
+      "learning_rate": 0.00011188190331485125,
+      "loss": 2.0058,
+      "step": 4546
+    },
+    {
+      "epoch": 0.4783798001052078,
+      "grad_norm": 1.4350491762161255,
+      "learning_rate": 0.00011184806666312609,
+      "loss": 1.8621,
+      "step": 4547
+    },
+    {
+      "epoch": 0.4784850078905839,
+      "grad_norm": 0.9594536423683167,
+      "learning_rate": 0.00011181422863552398,
+      "loss": 2.0049,
+      "step": 4548
+    },
+    {
+      "epoch": 0.47859021567596,
+      "grad_norm": 1.3745486736297607,
+      "learning_rate": 0.0001117803892359744,
+      "loss": 2.1858,
+      "step": 4549
+    },
+    {
+      "epoch": 0.47869542346133614,
+      "grad_norm": 1.600081443786621,
+      "learning_rate": 0.00011174654846840701,
+      "loss": 1.9469,
+      "step": 4550
+    },
+    {
+      "epoch": 0.47880063124671224,
+      "grad_norm": 0.9910126328468323,
+      "learning_rate": 0.00011171270633675161,
+      "loss": 2.1031,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4789058390320884,
+      "grad_norm": 1.027575135231018,
+      "learning_rate": 0.00011167886284493821,
+      "loss": 1.5686,
+      "step": 4552
+    },
+    {
+      "epoch": 0.4790110468174645,
+      "grad_norm": 1.3736275434494019,
+      "learning_rate": 0.00011164501799689693,
+      "loss": 1.8597,
+      "step": 4553
+    },
+    {
+      "epoch": 0.47911625460284063,
+      "grad_norm": 1.4835340976715088,
+      "learning_rate": 0.00011161117179655804,
+      "loss": 2.0892,
+      "step": 4554
+    },
+    {
+      "epoch": 0.4792214623882167,
+      "grad_norm": 1.1146955490112305,
+      "learning_rate": 0.00011157732424785202,
+      "loss": 1.7941,
+      "step": 4555
+    },
+    {
+      "epoch": 0.4793266701735929,
+      "grad_norm": 1.7215523719787598,
+      "learning_rate": 0.00011154347535470947,
+      "loss": 1.6763,
+      "step": 4556
+    },
+    {
+      "epoch": 0.47943187795896897,
+      "grad_norm": 1.5364283323287964,
+      "learning_rate": 0.0001115096251210611,
+      "loss": 2.0692,
+      "step": 4557
+    },
+    {
+      "epoch": 0.47953708574434506,
+      "grad_norm": 1.7501965761184692,
+      "learning_rate": 0.00011147577355083789,
+      "loss": 2.2162,
+      "step": 4558
+    },
+    {
+      "epoch": 0.4796422935297212,
+      "grad_norm": 1.8486822843551636,
+      "learning_rate": 0.00011144192064797088,
+      "loss": 1.9551,
+      "step": 4559
+    },
+    {
+      "epoch": 0.4797475013150973,
+      "grad_norm": 2.2193543910980225,
+      "learning_rate": 0.0001114080664163913,
+      "loss": 1.7281,
+      "step": 4560
+    },
+    {
+      "epoch": 0.47985270910047345,
+      "grad_norm": 1.7399812936782837,
+      "learning_rate": 0.00011137421086003052,
+      "loss": 1.8969,
+      "step": 4561
+    },
+    {
+      "epoch": 0.47995791688584954,
+      "grad_norm": 1.6608930826187134,
+      "learning_rate": 0.0001113403539828201,
+      "loss": 1.8734,
+      "step": 4562
+    },
+    {
+      "epoch": 0.4800631246712257,
+      "grad_norm": 1.4710301160812378,
+      "learning_rate": 0.00011130649578869173,
+      "loss": 1.8415,
+      "step": 4563
+    },
+    {
+      "epoch": 0.4801683324566018,
+      "grad_norm": 1.3207552433013916,
+      "learning_rate": 0.00011127263628157722,
+      "loss": 1.8769,
+      "step": 4564
+    },
+    {
+      "epoch": 0.4802735402419779,
+      "grad_norm": 1.6613874435424805,
+      "learning_rate": 0.00011123877546540857,
+      "loss": 1.778,
+      "step": 4565
+    },
+    {
+      "epoch": 0.48037874802735403,
+      "grad_norm": 1.1604993343353271,
+      "learning_rate": 0.00011120491334411793,
+      "loss": 1.5905,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4804839558127301,
+      "grad_norm": 1.4732471704483032,
+      "learning_rate": 0.00011117104992163762,
+      "loss": 1.515,
+      "step": 4567
+    },
+    {
+      "epoch": 0.4805891635981063,
+      "grad_norm": 1.1420209407806396,
+      "learning_rate": 0.00011113718520190006,
+      "loss": 1.6139,
+      "step": 4568
+    },
+    {
+      "epoch": 0.48069437138348237,
+      "grad_norm": 1.2326006889343262,
+      "learning_rate": 0.00011110331918883787,
+      "loss": 1.5793,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4807995791688585,
+      "grad_norm": 1.0035511255264282,
+      "learning_rate": 0.00011106945188638378,
+      "loss": 2.121,
+      "step": 4570
+    },
+    {
+      "epoch": 0.4809047869542346,
+      "grad_norm": 0.9064120054244995,
+      "learning_rate": 0.0001110355832984707,
+      "loss": 1.6258,
+      "step": 4571
+    },
+    {
+      "epoch": 0.48100999473961076,
+      "grad_norm": 0.8680402040481567,
+      "learning_rate": 0.00011100171342903165,
+      "loss": 2.326,
+      "step": 4572
+    },
+    {
+      "epoch": 0.48111520252498685,
+      "grad_norm": 1.4351173639297485,
+      "learning_rate": 0.00011096784228199985,
+      "loss": 1.8333,
+      "step": 4573
+    },
+    {
+      "epoch": 0.48122041031036294,
+      "grad_norm": 1.4196946620941162,
+      "learning_rate": 0.00011093396986130866,
+      "loss": 1.5755,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4813256180957391,
+      "grad_norm": 1.547824501991272,
+      "learning_rate": 0.00011090009617089155,
+      "loss": 2.2177,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4814308258811152,
+      "grad_norm": 1.1497628688812256,
+      "learning_rate": 0.00011086622121468213,
+      "loss": 1.8107,
+      "step": 4576
+    },
+    {
+      "epoch": 0.48153603366649134,
+      "grad_norm": 0.9802242517471313,
+      "learning_rate": 0.00011083234499661426,
+      "loss": 1.9389,
+      "step": 4577
+    },
+    {
+      "epoch": 0.48164124145186743,
+      "grad_norm": 1.283870816230774,
+      "learning_rate": 0.00011079846752062182,
+      "loss": 2.4584,
+      "step": 4578
+    },
+    {
+      "epoch": 0.4817464492372436,
+      "grad_norm": 1.0405322313308716,
+      "learning_rate": 0.00011076458879063891,
+      "loss": 1.7302,
+      "step": 4579
+    },
+    {
+      "epoch": 0.4818516570226197,
+      "grad_norm": 1.3652944564819336,
+      "learning_rate": 0.00011073070881059977,
+      "loss": 1.8197,
+      "step": 4580
+    },
+    {
+      "epoch": 0.48195686480799577,
+      "grad_norm": 0.9112363457679749,
+      "learning_rate": 0.00011069682758443873,
+      "loss": 2.3815,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4820620725933719,
+      "grad_norm": 2.2471213340759277,
+      "learning_rate": 0.00011066294511609032,
+      "loss": 1.918,
+      "step": 4582
+    },
+    {
+      "epoch": 0.482167280378748,
+      "grad_norm": 1.5038715600967407,
+      "learning_rate": 0.00011062906140948922,
+      "loss": 1.571,
+      "step": 4583
+    },
+    {
+      "epoch": 0.48227248816412416,
+      "grad_norm": 1.7240707874298096,
+      "learning_rate": 0.00011059517646857023,
+      "loss": 1.7805,
+      "step": 4584
+    },
+    {
+      "epoch": 0.48237769594950025,
+      "grad_norm": 1.71571946144104,
+      "learning_rate": 0.00011056129029726825,
+      "loss": 2.1035,
+      "step": 4585
+    },
+    {
+      "epoch": 0.4824829037348764,
+      "grad_norm": 1.455073356628418,
+      "learning_rate": 0.00011052740289951842,
+      "loss": 1.5254,
+      "step": 4586
+    },
+    {
+      "epoch": 0.4825881115202525,
+      "grad_norm": 1.1514016389846802,
+      "learning_rate": 0.00011049351427925598,
+      "loss": 1.5474,
+      "step": 4587
+    },
+    {
+      "epoch": 0.48269331930562864,
+      "grad_norm": 1.1613560914993286,
+      "learning_rate": 0.00011045962444041624,
+      "loss": 2.2161,
+      "step": 4588
+    },
+    {
+      "epoch": 0.48279852709100474,
+      "grad_norm": 1.7991211414337158,
+      "learning_rate": 0.00011042573338693479,
+      "loss": 1.9893,
+      "step": 4589
+    },
+    {
+      "epoch": 0.48290373487638083,
+      "grad_norm": 1.4048147201538086,
+      "learning_rate": 0.00011039184112274725,
+      "loss": 1.771,
+      "step": 4590
+    },
+    {
+      "epoch": 0.483008942661757,
+      "grad_norm": 1.2304316759109497,
+      "learning_rate": 0.00011035794765178941,
+      "loss": 2.0752,
+      "step": 4591
+    },
+    {
+      "epoch": 0.4831141504471331,
+      "grad_norm": 1.0293245315551758,
+      "learning_rate": 0.00011032405297799722,
+      "loss": 2.0021,
+      "step": 4592
+    },
+    {
+      "epoch": 0.4832193582325092,
+      "grad_norm": 1.4348002672195435,
+      "learning_rate": 0.00011029015710530674,
+      "loss": 1.5027,
+      "step": 4593
+    },
+    {
+      "epoch": 0.4833245660178853,
+      "grad_norm": 1.2140568494796753,
+      "learning_rate": 0.0001102562600376542,
+      "loss": 1.864,
+      "step": 4594
+    },
+    {
+      "epoch": 0.48342977380326146,
+      "grad_norm": 1.5038398504257202,
+      "learning_rate": 0.000110222361778976,
+      "loss": 1.567,
+      "step": 4595
+    },
+    {
+      "epoch": 0.48353498158863756,
+      "grad_norm": 2.2957375049591064,
+      "learning_rate": 0.00011018846233320854,
+      "loss": 1.1497,
+      "step": 4596
+    },
+    {
+      "epoch": 0.48364018937401365,
+      "grad_norm": 1.7357734441757202,
+      "learning_rate": 0.0001101545617042885,
+      "loss": 2.1494,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4837453971593898,
+      "grad_norm": 1.3374382257461548,
+      "learning_rate": 0.0001101206598961527,
+      "loss": 2.0164,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4838506049447659,
+      "grad_norm": 1.0593690872192383,
+      "learning_rate": 0.00011008675691273793,
+      "loss": 1.7716,
+      "step": 4599
+    },
+    {
+      "epoch": 0.48395581273014204,
+      "grad_norm": 1.28423273563385,
+      "learning_rate": 0.00011005285275798132,
+      "loss": 1.9312,
+      "step": 4600
+    },
+    {
+      "epoch": 0.48406102051551814,
+      "grad_norm": 1.176078200340271,
+      "learning_rate": 0.00011001894743582004,
+      "loss": 2.0926,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4841662283008943,
+      "grad_norm": 1.3551251888275146,
+      "learning_rate": 0.00010998504095019137,
+      "loss": 1.9196,
+      "step": 4602
+    },
+    {
+      "epoch": 0.4842714360862704,
+      "grad_norm": 1.554956316947937,
+      "learning_rate": 0.00010995113330503278,
+      "loss": 1.5871,
+      "step": 4603
+    },
+    {
+      "epoch": 0.48437664387164653,
+      "grad_norm": 0.9341940879821777,
+      "learning_rate": 0.00010991722450428184,
+      "loss": 1.5597,
+      "step": 4604
+    },
+    {
+      "epoch": 0.4844818516570226,
+      "grad_norm": 1.026902437210083,
+      "learning_rate": 0.00010988331455187628,
+      "loss": 2.1465,
+      "step": 4605
+    },
+    {
+      "epoch": 0.4845870594423987,
+      "grad_norm": 1.683455228805542,
+      "learning_rate": 0.00010984940345175392,
+      "loss": 1.9747,
+      "step": 4606
+    },
+    {
+      "epoch": 0.48469226722777486,
+      "grad_norm": 1.6722251176834106,
+      "learning_rate": 0.0001098154912078528,
+      "loss": 1.8076,
+      "step": 4607
+    },
+    {
+      "epoch": 0.48479747501315096,
+      "grad_norm": 1.0394576787948608,
+      "learning_rate": 0.000109781577824111,
+      "loss": 1.9039,
+      "step": 4608
+    },
+    {
+      "epoch": 0.4849026827985271,
+      "grad_norm": 1.5216197967529297,
+      "learning_rate": 0.00010974766330446678,
+      "loss": 2.1068,
+      "step": 4609
+    },
+    {
+      "epoch": 0.4850078905839032,
+      "grad_norm": 2.0184504985809326,
+      "learning_rate": 0.00010971374765285851,
+      "loss": 1.9356,
+      "step": 4610
+    },
+    {
+      "epoch": 0.48511309836927935,
+      "grad_norm": 2.1446685791015625,
+      "learning_rate": 0.0001096798308732247,
+      "loss": 1.0265,
+      "step": 4611
+    },
+    {
+      "epoch": 0.48521830615465544,
+      "grad_norm": 1.6989480257034302,
+      "learning_rate": 0.00010964591296950406,
+      "loss": 1.8608,
+      "step": 4612
+    },
+    {
+      "epoch": 0.48532351394003154,
+      "grad_norm": 1.0963361263275146,
+      "learning_rate": 0.00010961199394563526,
+      "loss": 1.5642,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4854287217254077,
+      "grad_norm": 1.4704118967056274,
+      "learning_rate": 0.00010957807380555727,
+      "loss": 1.5999,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4855339295107838,
+      "grad_norm": 1.1268386840820312,
+      "learning_rate": 0.00010954415255320909,
+      "loss": 1.4149,
+      "step": 4615
+    },
+    {
+      "epoch": 0.48563913729615993,
+      "grad_norm": 1.2833198308944702,
+      "learning_rate": 0.00010951023019252993,
+      "loss": 1.3482,
+      "step": 4616
+    },
+    {
+      "epoch": 0.485744345081536,
+      "grad_norm": 1.0503816604614258,
+      "learning_rate": 0.00010947630672745906,
+      "loss": 1.5722,
+      "step": 4617
+    },
+    {
+      "epoch": 0.48584955286691217,
+      "grad_norm": 2.151472330093384,
+      "learning_rate": 0.00010944238216193586,
+      "loss": 2.0229,
+      "step": 4618
+    },
+    {
+      "epoch": 0.48595476065228826,
+      "grad_norm": 1.2306005954742432,
+      "learning_rate": 0.00010940845649989994,
+      "loss": 1.4716,
+      "step": 4619
+    },
+    {
+      "epoch": 0.4860599684376644,
+      "grad_norm": 1.4052913188934326,
+      "learning_rate": 0.00010937452974529093,
+      "loss": 1.6263,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4861651762230405,
+      "grad_norm": 1.0920840501785278,
+      "learning_rate": 0.00010934060190204865,
+      "loss": 2.1744,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4862703840084166,
+      "grad_norm": 2.252732276916504,
+      "learning_rate": 0.00010930667297411305,
+      "loss": 1.9662,
+      "step": 4622
+    },
+    {
+      "epoch": 0.48637559179379275,
+      "grad_norm": 1.5878801345825195,
+      "learning_rate": 0.00010927274296542416,
+      "loss": 1.7275,
+      "step": 4623
+    },
+    {
+      "epoch": 0.48648079957916884,
+      "grad_norm": 0.946306586265564,
+      "learning_rate": 0.00010923881187992215,
+      "loss": 1.5401,
+      "step": 4624
+    },
+    {
+      "epoch": 0.486586007364545,
+      "grad_norm": 1.2825038433074951,
+      "learning_rate": 0.00010920487972154734,
+      "loss": 2.1301,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4866912151499211,
+      "grad_norm": 2.0224785804748535,
+      "learning_rate": 0.00010917094649424018,
+      "loss": 1.6339,
+      "step": 4626
+    },
+    {
+      "epoch": 0.48679642293529723,
+      "grad_norm": 1.4419629573822021,
+      "learning_rate": 0.00010913701220194117,
+      "loss": 2.2219,
+      "step": 4627
+    },
+    {
+      "epoch": 0.48690163072067333,
+      "grad_norm": 1.4137790203094482,
+      "learning_rate": 0.00010910307684859102,
+      "loss": 2.2587,
+      "step": 4628
+    },
+    {
+      "epoch": 0.4870068385060494,
+      "grad_norm": 1.2425460815429688,
+      "learning_rate": 0.00010906914043813056,
+      "loss": 1.6606,
+      "step": 4629
+    },
+    {
+      "epoch": 0.48711204629142557,
+      "grad_norm": 1.4972299337387085,
+      "learning_rate": 0.00010903520297450067,
+      "loss": 1.3636,
+      "step": 4630
+    },
+    {
+      "epoch": 0.48721725407680166,
+      "grad_norm": 1.6092376708984375,
+      "learning_rate": 0.0001090012644616424,
+      "loss": 2.0573,
+      "step": 4631
+    },
+    {
+      "epoch": 0.4873224618621778,
+      "grad_norm": 1.6330995559692383,
+      "learning_rate": 0.00010896732490349697,
+      "loss": 1.7161,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4874276696475539,
+      "grad_norm": 2.3415868282318115,
+      "learning_rate": 0.00010893338430400562,
+      "loss": 1.524,
+      "step": 4633
+    },
+    {
+      "epoch": 0.48753287743293006,
+      "grad_norm": 1.2462005615234375,
+      "learning_rate": 0.00010889944266710972,
+      "loss": 1.5716,
+      "step": 4634
+    },
+    {
+      "epoch": 0.48763808521830615,
+      "grad_norm": 1.564387321472168,
+      "learning_rate": 0.00010886549999675088,
+      "loss": 1.6856,
+      "step": 4635
+    },
+    {
+      "epoch": 0.4877432930036823,
+      "grad_norm": 0.9912921786308289,
+      "learning_rate": 0.00010883155629687071,
+      "loss": 1.878,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4878485007890584,
+      "grad_norm": 1.348905086517334,
+      "learning_rate": 0.000108797611571411,
+      "loss": 2.4173,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4879537085744345,
+      "grad_norm": 1.316444754600525,
+      "learning_rate": 0.00010876366582431361,
+      "loss": 2.1035,
+      "step": 4638
+    },
+    {
+      "epoch": 0.48805891635981064,
+      "grad_norm": 1.0654300451278687,
+      "learning_rate": 0.00010872971905952057,
+      "loss": 2.1269,
+      "step": 4639
+    },
+    {
+      "epoch": 0.48816412414518673,
+      "grad_norm": 1.8568501472473145,
+      "learning_rate": 0.00010869577128097404,
+      "loss": 1.2797,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4882693319305629,
+      "grad_norm": 1.3833286762237549,
+      "learning_rate": 0.00010866182249261617,
+      "loss": 1.9619,
+      "step": 4641
+    },
+    {
+      "epoch": 0.48837453971593897,
+      "grad_norm": 0.8749449253082275,
+      "learning_rate": 0.00010862787269838939,
+      "loss": 1.6768,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4884797475013151,
+      "grad_norm": 1.499631643295288,
+      "learning_rate": 0.00010859392190223619,
+      "loss": 1.9632,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4885849552866912,
+      "grad_norm": 1.0441616773605347,
+      "learning_rate": 0.00010855997010809915,
+      "loss": 2.0146,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4886901630720673,
+      "grad_norm": 1.8089901208877563,
+      "learning_rate": 0.00010852601731992094,
+      "loss": 1.8159,
+      "step": 4645
+    },
+    {
+      "epoch": 0.48879537085744346,
+      "grad_norm": 1.6164319515228271,
+      "learning_rate": 0.00010849206354164439,
+      "loss": 1.3445,
+      "step": 4646
+    },
+    {
+      "epoch": 0.48890057864281955,
+      "grad_norm": 1.5210686922073364,
+      "learning_rate": 0.00010845810877721252,
+      "loss": 1.6565,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4890057864281957,
+      "grad_norm": 1.2863743305206299,
+      "learning_rate": 0.00010842415303056827,
+      "loss": 1.5714,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4891109942135718,
+      "grad_norm": 1.657315731048584,
+      "learning_rate": 0.0001083901963056549,
+      "loss": 1.7163,
+      "step": 4649
+    },
+    {
+      "epoch": 0.48921620199894794,
+      "grad_norm": 1.0181341171264648,
+      "learning_rate": 0.00010835623860641569,
+      "loss": 1.8581,
+      "step": 4650
+    },
+    {
+      "epoch": 0.48932140978432404,
+      "grad_norm": 0.8646203279495239,
+      "learning_rate": 0.00010832227993679396,
+      "loss": 1.8136,
+      "step": 4651
+    },
+    {
+      "epoch": 0.4894266175697002,
+      "grad_norm": 1.2895039319992065,
+      "learning_rate": 0.00010828832030073329,
+      "loss": 1.9504,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4895318253550763,
+      "grad_norm": 1.2667862176895142,
+      "learning_rate": 0.00010825435970217728,
+      "loss": 1.9774,
+      "step": 4653
+    },
+    {
+      "epoch": 0.48963703314045237,
+      "grad_norm": 1.3460382223129272,
+      "learning_rate": 0.00010822039814506964,
+      "loss": 2.0969,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4897422409258285,
+      "grad_norm": 1.2763968706130981,
+      "learning_rate": 0.00010818643563335424,
+      "loss": 2.2074,
+      "step": 4655
+    },
+    {
+      "epoch": 0.4898474487112046,
+      "grad_norm": 0.9694689512252808,
+      "learning_rate": 0.00010815247217097504,
+      "loss": 1.7114,
+      "step": 4656
+    },
+    {
+      "epoch": 0.48995265649658076,
+      "grad_norm": 2.234217882156372,
+      "learning_rate": 0.00010811850776187608,
+      "loss": 2.196,
+      "step": 4657
+    },
+    {
+      "epoch": 0.49005786428195686,
+      "grad_norm": 1.10568106174469,
+      "learning_rate": 0.00010808454241000155,
+      "loss": 1.66,
+      "step": 4658
+    },
+    {
+      "epoch": 0.490163072067333,
+      "grad_norm": 1.4655194282531738,
+      "learning_rate": 0.00010805057611929573,
+      "loss": 1.9986,
+      "step": 4659
+    },
+    {
+      "epoch": 0.4902682798527091,
+      "grad_norm": 1.9132064580917358,
+      "learning_rate": 0.00010801660889370301,
+      "loss": 1.7011,
+      "step": 4660
+    },
+    {
+      "epoch": 0.4903734876380852,
+      "grad_norm": 1.7220205068588257,
+      "learning_rate": 0.00010798264073716791,
+      "loss": 1.6763,
+      "step": 4661
+    },
+    {
+      "epoch": 0.49047869542346134,
+      "grad_norm": 0.8977935910224915,
+      "learning_rate": 0.000107948671653635,
+      "loss": 1.8618,
+      "step": 4662
+    },
+    {
+      "epoch": 0.49058390320883744,
+      "grad_norm": 1.6346113681793213,
+      "learning_rate": 0.00010791470164704904,
+      "loss": 1.6551,
+      "step": 4663
+    },
+    {
+      "epoch": 0.4906891109942136,
+      "grad_norm": 1.7571468353271484,
+      "learning_rate": 0.00010788073072135485,
+      "loss": 1.6048,
+      "step": 4664
+    },
+    {
+      "epoch": 0.4907943187795897,
+      "grad_norm": 1.8179447650909424,
+      "learning_rate": 0.00010784675888049735,
+      "loss": 2.1345,
+      "step": 4665
+    },
+    {
+      "epoch": 0.4908995265649658,
+      "grad_norm": 1.7753139734268188,
+      "learning_rate": 0.00010781278612842159,
+      "loss": 1.8888,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4910047343503419,
+      "grad_norm": 1.0005837678909302,
+      "learning_rate": 0.00010777881246907269,
+      "loss": 1.645,
+      "step": 4667
+    },
+    {
+      "epoch": 0.49110994213571807,
+      "grad_norm": 1.0366028547286987,
+      "learning_rate": 0.00010774483790639591,
+      "loss": 1.5939,
+      "step": 4668
+    },
+    {
+      "epoch": 0.49121514992109416,
+      "grad_norm": 0.8974822163581848,
+      "learning_rate": 0.00010771086244433662,
+      "loss": 1.9674,
+      "step": 4669
+    },
+    {
+      "epoch": 0.49132035770647026,
+      "grad_norm": 1.6691381931304932,
+      "learning_rate": 0.00010767688608684023,
+      "loss": 1.32,
+      "step": 4670
+    },
+    {
+      "epoch": 0.4914255654918464,
+      "grad_norm": 1.127601981163025,
+      "learning_rate": 0.00010764290883785237,
+      "loss": 1.8129,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4915307732772225,
+      "grad_norm": 1.6219996213912964,
+      "learning_rate": 0.00010760893070131868,
+      "loss": 1.8319,
+      "step": 4672
+    },
+    {
+      "epoch": 0.49163598106259865,
+      "grad_norm": 1.6142001152038574,
+      "learning_rate": 0.0001075749516811849,
+      "loss": 1.3877,
+      "step": 4673
+    },
+    {
+      "epoch": 0.49174118884797474,
+      "grad_norm": 1.2514292001724243,
+      "learning_rate": 0.00010754097178139695,
+      "loss": 1.4618,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4918463966333509,
+      "grad_norm": 2.3644227981567383,
+      "learning_rate": 0.00010750699100590076,
+      "loss": 1.4959,
+      "step": 4675
+    },
+    {
+      "epoch": 0.491951604418727,
+      "grad_norm": 0.9489824175834656,
+      "learning_rate": 0.00010747300935864243,
+      "loss": 1.775,
+      "step": 4676
+    },
+    {
+      "epoch": 0.4920568122041031,
+      "grad_norm": 1.1649144887924194,
+      "learning_rate": 0.00010743902684356815,
+      "loss": 1.8669,
+      "step": 4677
+    },
+    {
+      "epoch": 0.4921620199894792,
+      "grad_norm": 1.477669596672058,
+      "learning_rate": 0.00010740504346462417,
+      "loss": 1.9316,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4922672277748553,
+      "grad_norm": 1.8862009048461914,
+      "learning_rate": 0.00010737105922575685,
+      "loss": 2.0482,
+      "step": 4679
+    },
+    {
+      "epoch": 0.49237243556023147,
+      "grad_norm": 1.249657154083252,
+      "learning_rate": 0.00010733707413091269,
+      "loss": 1.5986,
+      "step": 4680
+    },
+    {
+      "epoch": 0.49247764334560756,
+      "grad_norm": 1.274321436882019,
+      "learning_rate": 0.00010730308818403832,
+      "loss": 2.0053,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4925828511309837,
+      "grad_norm": 1.5534741878509521,
+      "learning_rate": 0.00010726910138908032,
+      "loss": 1.7589,
+      "step": 4682
+    },
+    {
+      "epoch": 0.4926880589163598,
+      "grad_norm": 1.196625828742981,
+      "learning_rate": 0.00010723511374998554,
+      "loss": 1.6722,
+      "step": 4683
+    },
+    {
+      "epoch": 0.49279326670173595,
+      "grad_norm": 1.5351232290267944,
+      "learning_rate": 0.00010720112527070083,
+      "loss": 1.7944,
+      "step": 4684
+    },
+    {
+      "epoch": 0.49289847448711205,
+      "grad_norm": 1.5849000215530396,
+      "learning_rate": 0.00010716713595517313,
+      "loss": 1.7102,
+      "step": 4685
+    },
+    {
+      "epoch": 0.49300368227248814,
+      "grad_norm": 1.8203864097595215,
+      "learning_rate": 0.00010713314580734954,
+      "loss": 2.2117,
+      "step": 4686
+    },
+    {
+      "epoch": 0.4931088900578643,
+      "grad_norm": 1.308393955230713,
+      "learning_rate": 0.00010709915483117723,
+      "loss": 1.8919,
+      "step": 4687
+    },
+    {
+      "epoch": 0.4932140978432404,
+      "grad_norm": 1.2589519023895264,
+      "learning_rate": 0.00010706516303060345,
+      "loss": 1.3221,
+      "step": 4688
+    },
+    {
+      "epoch": 0.49331930562861653,
+      "grad_norm": 1.2415590286254883,
+      "learning_rate": 0.00010703117040957553,
+      "loss": 1.8339,
+      "step": 4689
+    },
+    {
+      "epoch": 0.4934245134139926,
+      "grad_norm": 1.551783800125122,
+      "learning_rate": 0.00010699717697204095,
+      "loss": 1.8219,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4935297211993688,
+      "grad_norm": 1.280253291130066,
+      "learning_rate": 0.00010696318272194726,
+      "loss": 1.7465,
+      "step": 4691
+    },
+    {
+      "epoch": 0.49363492898474487,
+      "grad_norm": 1.4209909439086914,
+      "learning_rate": 0.00010692918766324209,
+      "loss": 1.7862,
+      "step": 4692
+    },
+    {
+      "epoch": 0.49374013677012096,
+      "grad_norm": 1.7612025737762451,
+      "learning_rate": 0.00010689519179987316,
+      "loss": 1.8797,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4938453445554971,
+      "grad_norm": 1.5453522205352783,
+      "learning_rate": 0.00010686119513578831,
+      "loss": 2.0076,
+      "step": 4694
+    },
+    {
+      "epoch": 0.4939505523408732,
+      "grad_norm": 1.7672468423843384,
+      "learning_rate": 0.00010682719767493547,
+      "loss": 1.8416,
+      "step": 4695
+    },
+    {
+      "epoch": 0.49405576012624935,
+      "grad_norm": 1.4348251819610596,
+      "learning_rate": 0.00010679319942126264,
+      "loss": 1.872,
+      "step": 4696
+    },
+    {
+      "epoch": 0.49416096791162545,
+      "grad_norm": 1.9852848052978516,
+      "learning_rate": 0.00010675920037871794,
+      "loss": 1.921,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4942661756970016,
+      "grad_norm": 1.4216840267181396,
+      "learning_rate": 0.00010672520055124958,
+      "loss": 1.5686,
+      "step": 4698
+    },
+    {
+      "epoch": 0.4943713834823777,
+      "grad_norm": 1.5400251150131226,
+      "learning_rate": 0.00010669119994280581,
+      "loss": 2.021,
+      "step": 4699
+    },
+    {
+      "epoch": 0.49447659126775384,
+      "grad_norm": 1.3077402114868164,
+      "learning_rate": 0.00010665719855733501,
+      "loss": 2.1264,
+      "step": 4700
+    },
+    {
+      "epoch": 0.49458179905312993,
+      "grad_norm": 1.2699223756790161,
+      "learning_rate": 0.00010662319639878565,
+      "loss": 1.7802,
+      "step": 4701
+    },
+    {
+      "epoch": 0.494687006838506,
+      "grad_norm": 1.2943834066390991,
+      "learning_rate": 0.00010658919347110634,
+      "loss": 1.8573,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4947922146238822,
+      "grad_norm": 1.5511493682861328,
+      "learning_rate": 0.00010655518977824566,
+      "loss": 2.2023,
+      "step": 4703
+    },
+    {
+      "epoch": 0.49489742240925827,
+      "grad_norm": 1.4400460720062256,
+      "learning_rate": 0.00010652118532415236,
+      "loss": 1.9261,
+      "step": 4704
+    },
+    {
+      "epoch": 0.4950026301946344,
+      "grad_norm": 1.2507339715957642,
+      "learning_rate": 0.00010648718011277535,
+      "loss": 1.6863,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4951078379800105,
+      "grad_norm": 1.3244736194610596,
+      "learning_rate": 0.00010645317414806342,
+      "loss": 1.9085,
+      "step": 4706
+    },
+    {
+      "epoch": 0.49521304576538666,
+      "grad_norm": 1.1200934648513794,
+      "learning_rate": 0.00010641916743396563,
+      "loss": 1.5047,
+      "step": 4707
+    },
+    {
+      "epoch": 0.49531825355076275,
+      "grad_norm": 1.6609597206115723,
+      "learning_rate": 0.00010638515997443109,
+      "loss": 1.1343,
+      "step": 4708
+    },
+    {
+      "epoch": 0.49542346133613885,
+      "grad_norm": 1.0850027799606323,
+      "learning_rate": 0.00010635115177340893,
+      "loss": 1.6411,
+      "step": 4709
+    },
+    {
+      "epoch": 0.495528669121515,
+      "grad_norm": 1.537176489830017,
+      "learning_rate": 0.00010631714283484842,
+      "loss": 1.524,
+      "step": 4710
+    },
+    {
+      "epoch": 0.4956338769068911,
+      "grad_norm": 2.0602495670318604,
+      "learning_rate": 0.00010628313316269891,
+      "loss": 1.9078,
+      "step": 4711
+    },
+    {
+      "epoch": 0.49573908469226724,
+      "grad_norm": 1.759886622428894,
+      "learning_rate": 0.00010624912276090988,
+      "loss": 1.9751,
+      "step": 4712
+    },
+    {
+      "epoch": 0.49584429247764333,
+      "grad_norm": 1.4797873497009277,
+      "learning_rate": 0.00010621511163343077,
+      "loss": 2.0188,
+      "step": 4713
+    },
+    {
+      "epoch": 0.4959495002630195,
+      "grad_norm": 1.672033429145813,
+      "learning_rate": 0.00010618109978421119,
+      "loss": 1.5586,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4960547080483956,
+      "grad_norm": 1.9004497528076172,
+      "learning_rate": 0.00010614708721720085,
+      "loss": 1.4778,
+      "step": 4715
+    },
+    {
+      "epoch": 0.4961599158337717,
+      "grad_norm": 1.600061058998108,
+      "learning_rate": 0.00010611307393634955,
+      "loss": 1.4879,
+      "step": 4716
+    },
+    {
+      "epoch": 0.4962651236191478,
+      "grad_norm": 1.7574427127838135,
+      "learning_rate": 0.0001060790599456071,
+      "loss": 1.8051,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4963703314045239,
+      "grad_norm": 1.1964126825332642,
+      "learning_rate": 0.0001060450452489234,
+      "loss": 1.7907,
+      "step": 4718
+    },
+    {
+      "epoch": 0.49647553918990006,
+      "grad_norm": 1.7190804481506348,
+      "learning_rate": 0.00010601102985024853,
+      "loss": 2.033,
+      "step": 4719
+    },
+    {
+      "epoch": 0.49658074697527615,
+      "grad_norm": 1.3082791566848755,
+      "learning_rate": 0.00010597701375353257,
+      "loss": 1.7543,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4966859547606523,
+      "grad_norm": 1.5606935024261475,
+      "learning_rate": 0.00010594299696272565,
+      "loss": 1.6072,
+      "step": 4721
+    },
+    {
+      "epoch": 0.4967911625460284,
+      "grad_norm": 1.4001544713974,
+      "learning_rate": 0.00010590897948177806,
+      "loss": 2.1012,
+      "step": 4722
+    },
+    {
+      "epoch": 0.49689637033140455,
+      "grad_norm": 1.4540914297103882,
+      "learning_rate": 0.00010587496131464019,
+      "loss": 1.3696,
+      "step": 4723
+    },
+    {
+      "epoch": 0.49700157811678064,
+      "grad_norm": 1.2410134077072144,
+      "learning_rate": 0.00010584094246526237,
+      "loss": 1.4659,
+      "step": 4724
+    },
+    {
+      "epoch": 0.49710678590215673,
+      "grad_norm": 1.1496258974075317,
+      "learning_rate": 0.00010580692293759513,
+      "loss": 1.7615,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4972119936875329,
+      "grad_norm": 1.969294548034668,
+      "learning_rate": 0.00010577290273558908,
+      "loss": 1.6735,
+      "step": 4726
+    },
+    {
+      "epoch": 0.497317201472909,
+      "grad_norm": 1.9359304904937744,
+      "learning_rate": 0.00010573888186319482,
+      "loss": 1.7351,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4974224092582851,
+      "grad_norm": 1.6216297149658203,
+      "learning_rate": 0.0001057048603243631,
+      "loss": 1.7732,
+      "step": 4728
+    },
+    {
+      "epoch": 0.4975276170436612,
+      "grad_norm": 1.4571208953857422,
+      "learning_rate": 0.00010567083812304477,
+      "loss": 1.5308,
+      "step": 4729
+    },
+    {
+      "epoch": 0.49763282482903737,
+      "grad_norm": 1.5610977411270142,
+      "learning_rate": 0.00010563681526319069,
+      "loss": 1.8377,
+      "step": 4730
+    },
+    {
+      "epoch": 0.49773803261441346,
+      "grad_norm": 1.9585388898849487,
+      "learning_rate": 0.00010560279174875179,
+      "loss": 2.1389,
+      "step": 4731
+    },
+    {
+      "epoch": 0.4978432403997896,
+      "grad_norm": 1.929347038269043,
+      "learning_rate": 0.0001055687675836791,
+      "loss": 0.9178,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4979484481851657,
+      "grad_norm": 1.9738050699234009,
+      "learning_rate": 0.00010553474277192381,
+      "loss": 1.9461,
+      "step": 4733
+    },
+    {
+      "epoch": 0.4980536559705418,
+      "grad_norm": 1.227402925491333,
+      "learning_rate": 0.00010550071731743707,
+      "loss": 2.15,
+      "step": 4734
+    },
+    {
+      "epoch": 0.49815886375591795,
+      "grad_norm": 1.7934569120407104,
+      "learning_rate": 0.00010546669122417013,
+      "loss": 1.7138,
+      "step": 4735
+    },
+    {
+      "epoch": 0.49826407154129404,
+      "grad_norm": 1.155175805091858,
+      "learning_rate": 0.00010543266449607432,
+      "loss": 1.708,
+      "step": 4736
+    },
+    {
+      "epoch": 0.4983692793266702,
+      "grad_norm": 1.2957451343536377,
+      "learning_rate": 0.0001053986371371011,
+      "loss": 1.8071,
+      "step": 4737
+    },
+    {
+      "epoch": 0.4984744871120463,
+      "grad_norm": 1.2430212497711182,
+      "learning_rate": 0.0001053646091512019,
+      "loss": 1.568,
+      "step": 4738
+    },
+    {
+      "epoch": 0.49857969489742243,
+      "grad_norm": 0.9808127880096436,
+      "learning_rate": 0.00010533058054232832,
+      "loss": 1.6683,
+      "step": 4739
+    },
+    {
+      "epoch": 0.4986849026827985,
+      "grad_norm": 1.0529319047927856,
+      "learning_rate": 0.00010529655131443199,
+      "loss": 1.5132,
+      "step": 4740
+    },
+    {
+      "epoch": 0.4987901104681746,
+      "grad_norm": 1.2187601327896118,
+      "learning_rate": 0.0001052625214714646,
+      "loss": 1.7863,
+      "step": 4741
+    },
+    {
+      "epoch": 0.49889531825355077,
+      "grad_norm": 1.5013796091079712,
+      "learning_rate": 0.00010522849101737788,
+      "loss": 1.8348,
+      "step": 4742
+    },
+    {
+      "epoch": 0.49900052603892686,
+      "grad_norm": 1.5242013931274414,
+      "learning_rate": 0.00010519445995612374,
+      "loss": 2.0386,
+      "step": 4743
+    },
+    {
+      "epoch": 0.499105733824303,
+      "grad_norm": 1.6240057945251465,
+      "learning_rate": 0.00010516042829165408,
+      "loss": 1.6951,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4992109416096791,
+      "grad_norm": 1.4725539684295654,
+      "learning_rate": 0.00010512639602792088,
+      "loss": 1.9254,
+      "step": 4745
+    },
+    {
+      "epoch": 0.49931614939505525,
+      "grad_norm": 1.0966986417770386,
+      "learning_rate": 0.00010509236316887615,
+      "loss": 1.7749,
+      "step": 4746
+    },
+    {
+      "epoch": 0.49942135718043135,
+      "grad_norm": 1.0299161672592163,
+      "learning_rate": 0.0001050583297184721,
+      "loss": 1.4928,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4995265649658075,
+      "grad_norm": 1.2240405082702637,
+      "learning_rate": 0.00010502429568066084,
+      "loss": 1.4958,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4996317727511836,
+      "grad_norm": 1.3169293403625488,
+      "learning_rate": 0.00010499026105939467,
+      "loss": 1.3972,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4997369805365597,
+      "grad_norm": 1.836531400680542,
+      "learning_rate": 0.00010495622585862594,
+      "loss": 2.0759,
+      "step": 4750
+    },
+    {
+      "epoch": 0.49984218832193583,
+      "grad_norm": 1.1490098237991333,
+      "learning_rate": 0.00010492219008230704,
+      "loss": 1.9707,
+      "step": 4751
+    },
+    {
+      "epoch": 0.4999473961073119,
+      "grad_norm": 1.809786319732666,
+      "learning_rate": 0.00010488815373439036,
+      "loss": 2.212,
+      "step": 4752
+    },
+    {
+      "epoch": 0.500052603892688,
+      "grad_norm": 1.3090404272079468,
+      "learning_rate": 0.0001048541168188285,
+      "loss": 1.7952,
+      "step": 4753
+    },
+    {
+      "epoch": 0.5001578116780642,
+      "grad_norm": 1.8462897539138794,
+      "learning_rate": 0.00010482007933957407,
+      "loss": 1.837,
+      "step": 4754
+    },
+    {
+      "epoch": 0.5002630194634403,
+      "grad_norm": 1.5334430932998657,
+      "learning_rate": 0.00010478604130057965,
+      "loss": 1.7316,
+      "step": 4755
+    },
+    {
+      "epoch": 0.5003682272488165,
+      "grad_norm": 1.2265576124191284,
+      "learning_rate": 0.00010475200270579803,
+      "loss": 1.6373,
+      "step": 4756
+    },
+    {
+      "epoch": 0.5004734350341925,
+      "grad_norm": 1.6965841054916382,
+      "learning_rate": 0.00010471796355918202,
+      "loss": 2.1889,
+      "step": 4757
+    },
+    {
+      "epoch": 0.5005786428195687,
+      "grad_norm": 1.6729693412780762,
+      "learning_rate": 0.0001046839238646844,
+      "loss": 1.4878,
+      "step": 4758
+    },
+    {
+      "epoch": 0.5006838506049448,
+      "grad_norm": 1.4285995960235596,
+      "learning_rate": 0.00010464988362625812,
+      "loss": 2.3533,
+      "step": 4759
+    },
+    {
+      "epoch": 0.5007890583903208,
+      "grad_norm": 0.9867104291915894,
+      "learning_rate": 0.00010461584284785617,
+      "loss": 2.2433,
+      "step": 4760
+    },
+    {
+      "epoch": 0.500894266175697,
+      "grad_norm": 1.198486566543579,
+      "learning_rate": 0.00010458180153343162,
+      "loss": 1.491,
+      "step": 4761
+    },
+    {
+      "epoch": 0.5009994739610731,
+      "grad_norm": 2.1748099327087402,
+      "learning_rate": 0.00010454775968693753,
+      "loss": 1.4489,
+      "step": 4762
+    },
+    {
+      "epoch": 0.5011046817464493,
+      "grad_norm": 1.7211601734161377,
+      "learning_rate": 0.00010451371731232708,
+      "loss": 1.4753,
+      "step": 4763
+    },
+    {
+      "epoch": 0.5012098895318253,
+      "grad_norm": 1.0159982442855835,
+      "learning_rate": 0.00010447967441355349,
+      "loss": 1.9713,
+      "step": 4764
+    },
+    {
+      "epoch": 0.5013150973172015,
+      "grad_norm": 1.20917546749115,
+      "learning_rate": 0.00010444563099457008,
+      "loss": 1.5139,
+      "step": 4765
+    },
+    {
+      "epoch": 0.5014203051025776,
+      "grad_norm": 1.7025082111358643,
+      "learning_rate": 0.00010441158705933016,
+      "loss": 1.5271,
+      "step": 4766
+    },
+    {
+      "epoch": 0.5015255128879537,
+      "grad_norm": 1.3307257890701294,
+      "learning_rate": 0.00010437754261178719,
+      "loss": 1.9943,
+      "step": 4767
+    },
+    {
+      "epoch": 0.5016307206733298,
+      "grad_norm": 2.893484592437744,
+      "learning_rate": 0.00010434349765589459,
+      "loss": 1.4271,
+      "step": 4768
+    },
+    {
+      "epoch": 0.501735928458706,
+      "grad_norm": 1.664506196975708,
+      "learning_rate": 0.0001043094521956059,
+      "loss": 1.5161,
+      "step": 4769
+    },
+    {
+      "epoch": 0.5018411362440821,
+      "grad_norm": 1.219966173171997,
+      "learning_rate": 0.00010427540623487475,
+      "loss": 1.3841,
+      "step": 4770
+    },
+    {
+      "epoch": 0.5019463440294581,
+      "grad_norm": 1.3663967847824097,
+      "learning_rate": 0.00010424135977765475,
+      "loss": 2.0504,
+      "step": 4771
+    },
+    {
+      "epoch": 0.5020515518148343,
+      "grad_norm": 1.2188875675201416,
+      "learning_rate": 0.00010420731282789957,
+      "loss": 1.7724,
+      "step": 4772
+    },
+    {
+      "epoch": 0.5021567596002104,
+      "grad_norm": 1.4278043508529663,
+      "learning_rate": 0.00010417326538956305,
+      "loss": 1.5938,
+      "step": 4773
+    },
+    {
+      "epoch": 0.5022619673855865,
+      "grad_norm": 1.2825640439987183,
+      "learning_rate": 0.00010413921746659894,
+      "loss": 2.0405,
+      "step": 4774
+    },
+    {
+      "epoch": 0.5023671751709626,
+      "grad_norm": 1.455748200416565,
+      "learning_rate": 0.00010410516906296115,
+      "loss": 1.3787,
+      "step": 4775
+    },
+    {
+      "epoch": 0.5024723829563388,
+      "grad_norm": 0.9623004198074341,
+      "learning_rate": 0.00010407112018260356,
+      "loss": 1.4602,
+      "step": 4776
+    },
+    {
+      "epoch": 0.5025775907417149,
+      "grad_norm": 1.078614592552185,
+      "learning_rate": 0.0001040370708294802,
+      "loss": 1.9882,
+      "step": 4777
+    },
+    {
+      "epoch": 0.502682798527091,
+      "grad_norm": 1.2258964776992798,
+      "learning_rate": 0.00010400302100754514,
+      "loss": 1.5276,
+      "step": 4778
+    },
+    {
+      "epoch": 0.5027880063124671,
+      "grad_norm": 1.2255494594573975,
+      "learning_rate": 0.00010396897072075237,
+      "loss": 1.2363,
+      "step": 4779
+    },
+    {
+      "epoch": 0.5028932140978433,
+      "grad_norm": 1.2698360681533813,
+      "learning_rate": 0.00010393491997305613,
+      "loss": 1.7825,
+      "step": 4780
+    },
+    {
+      "epoch": 0.5029984218832193,
+      "grad_norm": 1.997664451599121,
+      "learning_rate": 0.00010390086876841061,
+      "loss": 1.9237,
+      "step": 4781
+    },
+    {
+      "epoch": 0.5031036296685955,
+      "grad_norm": 1.0741022825241089,
+      "learning_rate": 0.00010386681711077002,
+      "loss": 1.8118,
+      "step": 4782
+    },
+    {
+      "epoch": 0.5032088374539716,
+      "grad_norm": 1.1076297760009766,
+      "learning_rate": 0.0001038327650040887,
+      "loss": 1.4896,
+      "step": 4783
+    },
+    {
+      "epoch": 0.5033140452393478,
+      "grad_norm": 2.157230854034424,
+      "learning_rate": 0.000103798712452321,
+      "loss": 1.6452,
+      "step": 4784
+    },
+    {
+      "epoch": 0.5034192530247238,
+      "grad_norm": 1.4255646467208862,
+      "learning_rate": 0.00010376465945942133,
+      "loss": 2.2098,
+      "step": 4785
+    },
+    {
+      "epoch": 0.5035244608100999,
+      "grad_norm": 1.7662725448608398,
+      "learning_rate": 0.00010373060602934415,
+      "loss": 2.1151,
+      "step": 4786
+    },
+    {
+      "epoch": 0.5036296685954761,
+      "grad_norm": 1.1463786363601685,
+      "learning_rate": 0.00010369655216604397,
+      "loss": 1.8305,
+      "step": 4787
+    },
+    {
+      "epoch": 0.5037348763808522,
+      "grad_norm": 1.4723010063171387,
+      "learning_rate": 0.00010366249787347537,
+      "loss": 2.3292,
+      "step": 4788
+    },
+    {
+      "epoch": 0.5038400841662283,
+      "grad_norm": 0.9463380575180054,
+      "learning_rate": 0.00010362844315559297,
+      "loss": 1.6957,
+      "step": 4789
+    },
+    {
+      "epoch": 0.5039452919516044,
+      "grad_norm": 1.907961368560791,
+      "learning_rate": 0.0001035943880163514,
+      "loss": 2.1228,
+      "step": 4790
+    },
+    {
+      "epoch": 0.5040504997369806,
+      "grad_norm": 1.6900379657745361,
+      "learning_rate": 0.00010356033245970536,
+      "loss": 2.1128,
+      "step": 4791
+    },
+    {
+      "epoch": 0.5041557075223566,
+      "grad_norm": 1.4977108240127563,
+      "learning_rate": 0.00010352627648960966,
+      "loss": 1.584,
+      "step": 4792
+    },
+    {
+      "epoch": 0.5042609153077328,
+      "grad_norm": 1.1327449083328247,
+      "learning_rate": 0.00010349222011001908,
+      "loss": 1.4022,
+      "step": 4793
+    },
+    {
+      "epoch": 0.5043661230931089,
+      "grad_norm": 1.6916382312774658,
+      "learning_rate": 0.0001034581633248885,
+      "loss": 1.8529,
+      "step": 4794
+    },
+    {
+      "epoch": 0.5044713308784851,
+      "grad_norm": 1.5584489107131958,
+      "learning_rate": 0.00010342410613817277,
+      "loss": 1.1992,
+      "step": 4795
+    },
+    {
+      "epoch": 0.5045765386638611,
+      "grad_norm": 1.77381432056427,
+      "learning_rate": 0.0001033900485538269,
+      "loss": 2.2351,
+      "step": 4796
+    },
+    {
+      "epoch": 0.5046817464492372,
+      "grad_norm": 1.6010788679122925,
+      "learning_rate": 0.00010335599057580583,
+      "loss": 1.8224,
+      "step": 4797
+    },
+    {
+      "epoch": 0.5047869542346134,
+      "grad_norm": 1.5694942474365234,
+      "learning_rate": 0.0001033219322080646,
+      "loss": 2.1278,
+      "step": 4798
+    },
+    {
+      "epoch": 0.5048921620199894,
+      "grad_norm": 1.6243647336959839,
+      "learning_rate": 0.00010328787345455837,
+      "loss": 1.913,
+      "step": 4799
+    },
+    {
+      "epoch": 0.5049973698053656,
+      "grad_norm": 2.164302110671997,
+      "learning_rate": 0.00010325381431924221,
+      "loss": 1.8029,
+      "step": 4800
+    },
+    {
+      "epoch": 0.5051025775907417,
+      "grad_norm": 0.8400641083717346,
+      "learning_rate": 0.00010321975480607129,
+      "loss": 1.3965,
+      "step": 4801
+    },
+    {
+      "epoch": 0.5052077853761179,
+      "grad_norm": 1.4535051584243774,
+      "learning_rate": 0.00010318569491900088,
+      "loss": 2.4847,
+      "step": 4802
+    },
+    {
+      "epoch": 0.5053129931614939,
+      "grad_norm": 1.4781876802444458,
+      "learning_rate": 0.00010315163466198616,
+      "loss": 1.9232,
+      "step": 4803
+    },
+    {
+      "epoch": 0.5054182009468701,
+      "grad_norm": 1.4223634004592896,
+      "learning_rate": 0.00010311757403898252,
+      "loss": 1.3876,
+      "step": 4804
+    },
+    {
+      "epoch": 0.5055234087322462,
+      "grad_norm": 2.039214849472046,
+      "learning_rate": 0.00010308351305394528,
+      "loss": 2.0856,
+      "step": 4805
+    },
+    {
+      "epoch": 0.5056286165176223,
+      "grad_norm": 1.5365289449691772,
+      "learning_rate": 0.0001030494517108298,
+      "loss": 1.9099,
+      "step": 4806
+    },
+    {
+      "epoch": 0.5057338243029984,
+      "grad_norm": 1.1576287746429443,
+      "learning_rate": 0.00010301539001359155,
+      "loss": 1.7402,
+      "step": 4807
+    },
+    {
+      "epoch": 0.5058390320883746,
+      "grad_norm": 1.7206387519836426,
+      "learning_rate": 0.00010298132796618596,
+      "loss": 1.9806,
+      "step": 4808
+    },
+    {
+      "epoch": 0.5059442398737507,
+      "grad_norm": 1.5616505146026611,
+      "learning_rate": 0.00010294726557256862,
+      "loss": 1.7111,
+      "step": 4809
+    },
+    {
+      "epoch": 0.5060494476591267,
+      "grad_norm": 1.6971865892410278,
+      "learning_rate": 0.00010291320283669499,
+      "loss": 2.0612,
+      "step": 4810
+    },
+    {
+      "epoch": 0.5061546554445029,
+      "grad_norm": 0.908016562461853,
+      "learning_rate": 0.0001028791397625207,
+      "loss": 2.0077,
+      "step": 4811
+    },
+    {
+      "epoch": 0.506259863229879,
+      "grad_norm": 0.8206803798675537,
+      "learning_rate": 0.00010284507635400142,
+      "loss": 1.7445,
+      "step": 4812
+    },
+    {
+      "epoch": 0.5063650710152551,
+      "grad_norm": 1.4862329959869385,
+      "learning_rate": 0.00010281101261509278,
+      "loss": 1.562,
+      "step": 4813
+    },
+    {
+      "epoch": 0.5064702788006312,
+      "grad_norm": 1.458016276359558,
+      "learning_rate": 0.00010277694854975051,
+      "loss": 1.0668,
+      "step": 4814
+    },
+    {
+      "epoch": 0.5065754865860074,
+      "grad_norm": 1.4686590433120728,
+      "learning_rate": 0.00010274288416193034,
+      "loss": 1.8496,
+      "step": 4815
+    },
+    {
+      "epoch": 0.5066806943713835,
+      "grad_norm": 1.2068148851394653,
+      "learning_rate": 0.00010270881945558808,
+      "loss": 2.0571,
+      "step": 4816
+    },
+    {
+      "epoch": 0.5067859021567596,
+      "grad_norm": 2.0828161239624023,
+      "learning_rate": 0.00010267475443467954,
+      "loss": 1.4989,
+      "step": 4817
+    },
+    {
+      "epoch": 0.5068911099421357,
+      "grad_norm": 1.2611733675003052,
+      "learning_rate": 0.00010264068910316055,
+      "loss": 1.7526,
+      "step": 4818
+    },
+    {
+      "epoch": 0.5069963177275119,
+      "grad_norm": 1.14738929271698,
+      "learning_rate": 0.00010260662346498703,
+      "loss": 2.0188,
+      "step": 4819
+    },
+    {
+      "epoch": 0.507101525512888,
+      "grad_norm": 1.0972294807434082,
+      "learning_rate": 0.00010257255752411495,
+      "loss": 1.8969,
+      "step": 4820
+    },
+    {
+      "epoch": 0.507206733298264,
+      "grad_norm": 1.5468823909759521,
+      "learning_rate": 0.0001025384912845002,
+      "loss": 1.5493,
+      "step": 4821
+    },
+    {
+      "epoch": 0.5073119410836402,
+      "grad_norm": 1.338287353515625,
+      "learning_rate": 0.0001025044247500988,
+      "loss": 1.9397,
+      "step": 4822
+    },
+    {
+      "epoch": 0.5074171488690163,
+      "grad_norm": 1.2431987524032593,
+      "learning_rate": 0.00010247035792486683,
+      "loss": 1.6526,
+      "step": 4823
+    },
+    {
+      "epoch": 0.5075223566543924,
+      "grad_norm": 3.7089953422546387,
+      "learning_rate": 0.00010243629081276031,
+      "loss": 2.1808,
+      "step": 4824
+    },
+    {
+      "epoch": 0.5076275644397685,
+      "grad_norm": 1.4418970346450806,
+      "learning_rate": 0.00010240222341773538,
+      "loss": 1.5014,
+      "step": 4825
+    },
+    {
+      "epoch": 0.5077327722251447,
+      "grad_norm": 1.5097178220748901,
+      "learning_rate": 0.00010236815574374816,
+      "loss": 1.9034,
+      "step": 4826
+    },
+    {
+      "epoch": 0.5078379800105208,
+      "grad_norm": 1.105833888053894,
+      "learning_rate": 0.00010233408779475482,
+      "loss": 1.6967,
+      "step": 4827
+    },
+    {
+      "epoch": 0.5079431877958969,
+      "grad_norm": 1.0300766229629517,
+      "learning_rate": 0.00010230001957471151,
+      "loss": 1.9713,
+      "step": 4828
+    },
+    {
+      "epoch": 0.508048395581273,
+      "grad_norm": 1.495307445526123,
+      "learning_rate": 0.00010226595108757451,
+      "loss": 1.7107,
+      "step": 4829
+    },
+    {
+      "epoch": 0.5081536033666492,
+      "grad_norm": 1.6375658512115479,
+      "learning_rate": 0.0001022318823373001,
+      "loss": 1.7321,
+      "step": 4830
+    },
+    {
+      "epoch": 0.5082588111520252,
+      "grad_norm": 1.7210208177566528,
+      "learning_rate": 0.00010219781332784451,
+      "loss": 1.7835,
+      "step": 4831
+    },
+    {
+      "epoch": 0.5083640189374014,
+      "grad_norm": 1.3353241682052612,
+      "learning_rate": 0.00010216374406316411,
+      "loss": 1.8193,
+      "step": 4832
+    },
+    {
+      "epoch": 0.5084692267227775,
+      "grad_norm": 1.2672992944717407,
+      "learning_rate": 0.00010212967454721523,
+      "loss": 2.0596,
+      "step": 4833
+    },
+    {
+      "epoch": 0.5085744345081536,
+      "grad_norm": 1.5790518522262573,
+      "learning_rate": 0.00010209560478395428,
+      "loss": 1.3901,
+      "step": 4834
+    },
+    {
+      "epoch": 0.5086796422935297,
+      "grad_norm": 1.039208173751831,
+      "learning_rate": 0.00010206153477733762,
+      "loss": 2.0043,
+      "step": 4835
+    },
+    {
+      "epoch": 0.5087848500789058,
+      "grad_norm": 1.4386025667190552,
+      "learning_rate": 0.00010202746453132172,
+      "loss": 1.807,
+      "step": 4836
+    },
+    {
+      "epoch": 0.508890057864282,
+      "grad_norm": 1.9724053144454956,
+      "learning_rate": 0.00010199339404986308,
+      "loss": 1.9082,
+      "step": 4837
+    },
+    {
+      "epoch": 0.508995265649658,
+      "grad_norm": 1.3470215797424316,
+      "learning_rate": 0.00010195932333691812,
+      "loss": 2.0544,
+      "step": 4838
+    },
+    {
+      "epoch": 0.5091004734350342,
+      "grad_norm": 2.0379178524017334,
+      "learning_rate": 0.0001019252523964434,
+      "loss": 1.5562,
+      "step": 4839
+    },
+    {
+      "epoch": 0.5092056812204103,
+      "grad_norm": 1.6759377717971802,
+      "learning_rate": 0.00010189118123239543,
+      "loss": 2.1569,
+      "step": 4840
+    },
+    {
+      "epoch": 0.5093108890057865,
+      "grad_norm": 1.4470930099487305,
+      "learning_rate": 0.00010185710984873084,
+      "loss": 2.3249,
+      "step": 4841
+    },
+    {
+      "epoch": 0.5094160967911625,
+      "grad_norm": 1.781419038772583,
+      "learning_rate": 0.0001018230382494062,
+      "loss": 2.0459,
+      "step": 4842
+    },
+    {
+      "epoch": 0.5095213045765387,
+      "grad_norm": 1.2743864059448242,
+      "learning_rate": 0.00010178896643837809,
+      "loss": 1.7856,
+      "step": 4843
+    },
+    {
+      "epoch": 0.5096265123619148,
+      "grad_norm": 1.4296742677688599,
+      "learning_rate": 0.00010175489441960327,
+      "loss": 1.943,
+      "step": 4844
+    },
+    {
+      "epoch": 0.5097317201472908,
+      "grad_norm": 2.150285005569458,
+      "learning_rate": 0.00010172082219703829,
+      "loss": 1.6565,
+      "step": 4845
+    },
+    {
+      "epoch": 0.509836927932667,
+      "grad_norm": 1.225191593170166,
+      "learning_rate": 0.0001016867497746399,
+      "loss": 1.8036,
+      "step": 4846
+    },
+    {
+      "epoch": 0.5099421357180431,
+      "grad_norm": 1.5737195014953613,
+      "learning_rate": 0.00010165267715636482,
+      "loss": 1.6792,
+      "step": 4847
+    },
+    {
+      "epoch": 0.5100473435034193,
+      "grad_norm": 1.7750451564788818,
+      "learning_rate": 0.00010161860434616982,
+      "loss": 1.3811,
+      "step": 4848
+    },
+    {
+      "epoch": 0.5101525512887953,
+      "grad_norm": 1.4361519813537598,
+      "learning_rate": 0.00010158453134801155,
+      "loss": 1.4103,
+      "step": 4849
+    },
+    {
+      "epoch": 0.5102577590741715,
+      "grad_norm": 1.3288068771362305,
+      "learning_rate": 0.00010155045816584691,
+      "loss": 2.1853,
+      "step": 4850
+    },
+    {
+      "epoch": 0.5103629668595476,
+      "grad_norm": 1.5870311260223389,
+      "learning_rate": 0.0001015163848036327,
+      "loss": 1.683,
+      "step": 4851
+    },
+    {
+      "epoch": 0.5104681746449238,
+      "grad_norm": 1.2659305334091187,
+      "learning_rate": 0.00010148231126532568,
+      "loss": 1.7193,
+      "step": 4852
+    },
+    {
+      "epoch": 0.5105733824302998,
+      "grad_norm": 1.333278775215149,
+      "learning_rate": 0.00010144823755488273,
+      "loss": 1.5757,
+      "step": 4853
+    },
+    {
+      "epoch": 0.510678590215676,
+      "grad_norm": 0.9039688110351562,
+      "learning_rate": 0.00010141416367626075,
+      "loss": 1.8356,
+      "step": 4854
+    },
+    {
+      "epoch": 0.5107837980010521,
+      "grad_norm": 0.9868873357772827,
+      "learning_rate": 0.00010138008963341657,
+      "loss": 1.6002,
+      "step": 4855
+    },
+    {
+      "epoch": 0.5108890057864282,
+      "grad_norm": 1.448551893234253,
+      "learning_rate": 0.00010134601543030713,
+      "loss": 2.0055,
+      "step": 4856
+    },
+    {
+      "epoch": 0.5109942135718043,
+      "grad_norm": 0.9456648230552673,
+      "learning_rate": 0.00010131194107088935,
+      "loss": 1.847,
+      "step": 4857
+    },
+    {
+      "epoch": 0.5110994213571805,
+      "grad_norm": 1.293251633644104,
+      "learning_rate": 0.00010127786655912021,
+      "loss": 1.235,
+      "step": 4858
+    },
+    {
+      "epoch": 0.5112046291425566,
+      "grad_norm": 1.0912578105926514,
+      "learning_rate": 0.00010124379189895661,
+      "loss": 2.0099,
+      "step": 4859
+    },
+    {
+      "epoch": 0.5113098369279326,
+      "grad_norm": 1.4057592153549194,
+      "learning_rate": 0.00010120971709435553,
+      "loss": 2.2696,
+      "step": 4860
+    },
+    {
+      "epoch": 0.5114150447133088,
+      "grad_norm": 0.9786197543144226,
+      "learning_rate": 0.000101175642149274,
+      "loss": 1.313,
+      "step": 4861
+    },
+    {
+      "epoch": 0.5115202524986849,
+      "grad_norm": 1.2093125581741333,
+      "learning_rate": 0.00010114156706766904,
+      "loss": 1.5894,
+      "step": 4862
+    },
+    {
+      "epoch": 0.511625460284061,
+      "grad_norm": 1.1187880039215088,
+      "learning_rate": 0.00010110749185349763,
+      "loss": 1.8596,
+      "step": 4863
+    },
+    {
+      "epoch": 0.5117306680694371,
+      "grad_norm": 1.3426660299301147,
+      "learning_rate": 0.00010107341651071684,
+      "loss": 1.6574,
+      "step": 4864
+    },
+    {
+      "epoch": 0.5118358758548133,
+      "grad_norm": 1.5116214752197266,
+      "learning_rate": 0.00010103934104328375,
+      "loss": 1.551,
+      "step": 4865
+    },
+    {
+      "epoch": 0.5119410836401894,
+      "grad_norm": 1.235568642616272,
+      "learning_rate": 0.00010100526545515539,
+      "loss": 1.8606,
+      "step": 4866
+    },
+    {
+      "epoch": 0.5120462914255655,
+      "grad_norm": 1.1877378225326538,
+      "learning_rate": 0.00010097118975028885,
+      "loss": 1.8721,
+      "step": 4867
+    },
+    {
+      "epoch": 0.5121514992109416,
+      "grad_norm": 1.012769103050232,
+      "learning_rate": 0.00010093711393264127,
+      "loss": 1.5946,
+      "step": 4868
+    },
+    {
+      "epoch": 0.5122567069963178,
+      "grad_norm": 1.3246477842330933,
+      "learning_rate": 0.00010090303800616974,
+      "loss": 1.3306,
+      "step": 4869
+    },
+    {
+      "epoch": 0.5123619147816938,
+      "grad_norm": 1.2476806640625,
+      "learning_rate": 0.00010086896197483136,
+      "loss": 1.8805,
+      "step": 4870
+    },
+    {
+      "epoch": 0.5124671225670699,
+      "grad_norm": 1.3036119937896729,
+      "learning_rate": 0.00010083488584258326,
+      "loss": 1.6369,
+      "step": 4871
+    },
+    {
+      "epoch": 0.5125723303524461,
+      "grad_norm": 1.2664161920547485,
+      "learning_rate": 0.00010080080961338265,
+      "loss": 1.5136,
+      "step": 4872
+    },
+    {
+      "epoch": 0.5126775381378222,
+      "grad_norm": 1.7038179636001587,
+      "learning_rate": 0.00010076673329118665,
+      "loss": 2.3886,
+      "step": 4873
+    },
+    {
+      "epoch": 0.5127827459231983,
+      "grad_norm": 1.4440287351608276,
+      "learning_rate": 0.00010073265687995243,
+      "loss": 1.9345,
+      "step": 4874
+    },
+    {
+      "epoch": 0.5128879537085744,
+      "grad_norm": 2.526855707168579,
+      "learning_rate": 0.0001006985803836372,
+      "loss": 1.3719,
+      "step": 4875
+    },
+    {
+      "epoch": 0.5129931614939506,
+      "grad_norm": 1.2801002264022827,
+      "learning_rate": 0.00010066450380619812,
+      "loss": 2.1389,
+      "step": 4876
+    },
+    {
+      "epoch": 0.5130983692793266,
+      "grad_norm": 1.681620478630066,
+      "learning_rate": 0.0001006304271515924,
+      "loss": 1.7829,
+      "step": 4877
+    },
+    {
+      "epoch": 0.5132035770647028,
+      "grad_norm": 1.5121160745620728,
+      "learning_rate": 0.00010059635042377725,
+      "loss": 2.1326,
+      "step": 4878
+    },
+    {
+      "epoch": 0.5133087848500789,
+      "grad_norm": 1.5406997203826904,
+      "learning_rate": 0.00010056227362670989,
+      "loss": 2.1427,
+      "step": 4879
+    },
+    {
+      "epoch": 0.5134139926354551,
+      "grad_norm": 1.1765246391296387,
+      "learning_rate": 0.00010052819676434754,
+      "loss": 1.8298,
+      "step": 4880
+    },
+    {
+      "epoch": 0.5135192004208311,
+      "grad_norm": 1.1417006254196167,
+      "learning_rate": 0.00010049411984064745,
+      "loss": 1.2332,
+      "step": 4881
+    },
+    {
+      "epoch": 0.5136244082062073,
+      "grad_norm": 0.9688271880149841,
+      "learning_rate": 0.00010046004285956684,
+      "loss": 1.8887,
+      "step": 4882
+    },
+    {
+      "epoch": 0.5137296159915834,
+      "grad_norm": 1.804365634918213,
+      "learning_rate": 0.00010042596582506298,
+      "loss": 1.6797,
+      "step": 4883
+    },
+    {
+      "epoch": 0.5138348237769595,
+      "grad_norm": 1.8476359844207764,
+      "learning_rate": 0.00010039188874109308,
+      "loss": 1.5778,
+      "step": 4884
+    },
+    {
+      "epoch": 0.5139400315623356,
+      "grad_norm": 1.6497973203659058,
+      "learning_rate": 0.00010035781161161446,
+      "loss": 1.8296,
+      "step": 4885
+    },
+    {
+      "epoch": 0.5140452393477117,
+      "grad_norm": 1.3590152263641357,
+      "learning_rate": 0.00010032373444058437,
+      "loss": 2.2775,
+      "step": 4886
+    },
+    {
+      "epoch": 0.5141504471330879,
+      "grad_norm": 1.5917162895202637,
+      "learning_rate": 0.00010028965723196002,
+      "loss": 2.0477,
+      "step": 4887
+    },
+    {
+      "epoch": 0.5142556549184639,
+      "grad_norm": 2.074033737182617,
+      "learning_rate": 0.00010025557998969875,
+      "loss": 1.708,
+      "step": 4888
+    },
+    {
+      "epoch": 0.5143608627038401,
+      "grad_norm": 1.4734264612197876,
+      "learning_rate": 0.00010022150271775783,
+      "loss": 1.8895,
+      "step": 4889
+    },
+    {
+      "epoch": 0.5144660704892162,
+      "grad_norm": 0.9418952465057373,
+      "learning_rate": 0.00010018742542009452,
+      "loss": 1.4237,
+      "step": 4890
+    },
+    {
+      "epoch": 0.5145712782745924,
+      "grad_norm": 1.5608463287353516,
+      "learning_rate": 0.00010015334810066612,
+      "loss": 1.7448,
+      "step": 4891
+    },
+    {
+      "epoch": 0.5146764860599684,
+      "grad_norm": 1.5609275102615356,
+      "learning_rate": 0.0001001192707634299,
+      "loss": 1.8851,
+      "step": 4892
+    },
+    {
+      "epoch": 0.5147816938453446,
+      "grad_norm": 1.515660047531128,
+      "learning_rate": 0.00010008519341234318,
+      "loss": 2.3635,
+      "step": 4893
+    },
+    {
+      "epoch": 0.5148869016307207,
+      "grad_norm": 1.6261802911758423,
+      "learning_rate": 0.00010005111605136319,
+      "loss": 1.8002,
+      "step": 4894
+    },
+    {
+      "epoch": 0.5149921094160967,
+      "grad_norm": 1.6840312480926514,
+      "learning_rate": 0.00010001703868444728,
+      "loss": 1.8234,
+      "step": 4895
+    },
+    {
+      "epoch": 0.5150973172014729,
+      "grad_norm": 1.0442866086959839,
+      "learning_rate": 9.998296131555273e-05,
+      "loss": 1.7835,
+      "step": 4896
+    },
+    {
+      "epoch": 0.515202524986849,
+      "grad_norm": 1.02167546749115,
+      "learning_rate": 9.994888394863683e-05,
+      "loss": 1.5165,
+      "step": 4897
+    },
+    {
+      "epoch": 0.5153077327722252,
+      "grad_norm": 1.9780964851379395,
+      "learning_rate": 9.991480658765685e-05,
+      "loss": 1.7533,
+      "step": 4898
+    },
+    {
+      "epoch": 0.5154129405576012,
+      "grad_norm": 2.685133695602417,
+      "learning_rate": 9.988072923657012e-05,
+      "loss": 2.181,
+      "step": 4899
+    },
+    {
+      "epoch": 0.5155181483429774,
+      "grad_norm": 1.0838717222213745,
+      "learning_rate": 9.98466518993339e-05,
+      "loss": 1.7872,
+      "step": 4900
+    },
+    {
+      "epoch": 0.5156233561283535,
+      "grad_norm": 2.0336694717407227,
+      "learning_rate": 9.981257457990548e-05,
+      "loss": 1.6419,
+      "step": 4901
+    },
+    {
+      "epoch": 0.5157285639137296,
+      "grad_norm": 1.6287788152694702,
+      "learning_rate": 9.977849728224219e-05,
+      "loss": 2.189,
+      "step": 4902
+    },
+    {
+      "epoch": 0.5158337716991057,
+      "grad_norm": 1.2882606983184814,
+      "learning_rate": 9.974442001030125e-05,
+      "loss": 1.7009,
+      "step": 4903
+    },
+    {
+      "epoch": 0.5159389794844819,
+      "grad_norm": 1.8696643114089966,
+      "learning_rate": 9.971034276803998e-05,
+      "loss": 2.0965,
+      "step": 4904
+    },
+    {
+      "epoch": 0.516044187269858,
+      "grad_norm": 1.5908252000808716,
+      "learning_rate": 9.967626555941564e-05,
+      "loss": 1.727,
+      "step": 4905
+    },
+    {
+      "epoch": 0.516149395055234,
+      "grad_norm": 1.1613487005233765,
+      "learning_rate": 9.964218838838554e-05,
+      "loss": 2.1126,
+      "step": 4906
+    },
+    {
+      "epoch": 0.5162546028406102,
+      "grad_norm": 1.0229045152664185,
+      "learning_rate": 9.960811125890695e-05,
+      "loss": 2.3455,
+      "step": 4907
+    },
+    {
+      "epoch": 0.5163598106259863,
+      "grad_norm": 1.273767113685608,
+      "learning_rate": 9.957403417493707e-05,
+      "loss": 1.6472,
+      "step": 4908
+    },
+    {
+      "epoch": 0.5164650184113624,
+      "grad_norm": 1.6049546003341675,
+      "learning_rate": 9.953995714043319e-05,
+      "loss": 1.6844,
+      "step": 4909
+    },
+    {
+      "epoch": 0.5165702261967385,
+      "grad_norm": 1.1211811304092407,
+      "learning_rate": 9.95058801593526e-05,
+      "loss": 2.0708,
+      "step": 4910
+    },
+    {
+      "epoch": 0.5166754339821147,
+      "grad_norm": 1.2839888334274292,
+      "learning_rate": 9.94718032356525e-05,
+      "loss": 1.8021,
+      "step": 4911
+    },
+    {
+      "epoch": 0.5167806417674908,
+      "grad_norm": 1.3963242769241333,
+      "learning_rate": 9.943772637329015e-05,
+      "loss": 2.2216,
+      "step": 4912
+    },
+    {
+      "epoch": 0.5168858495528669,
+      "grad_norm": 1.2890313863754272,
+      "learning_rate": 9.940364957622276e-05,
+      "loss": 1.7626,
+      "step": 4913
+    },
+    {
+      "epoch": 0.516991057338243,
+      "grad_norm": 1.017500400543213,
+      "learning_rate": 9.936957284840763e-05,
+      "loss": 1.8122,
+      "step": 4914
+    },
+    {
+      "epoch": 0.5170962651236192,
+      "grad_norm": 1.0217024087905884,
+      "learning_rate": 9.93354961938019e-05,
+      "loss": 2.2297,
+      "step": 4915
+    },
+    {
+      "epoch": 0.5172014729089953,
+      "grad_norm": 1.5468822717666626,
+      "learning_rate": 9.93014196163628e-05,
+      "loss": 2.0883,
+      "step": 4916
+    },
+    {
+      "epoch": 0.5173066806943714,
+      "grad_norm": 1.8743644952774048,
+      "learning_rate": 9.926734312004759e-05,
+      "loss": 1.6563,
+      "step": 4917
+    },
+    {
+      "epoch": 0.5174118884797475,
+      "grad_norm": 1.372307538986206,
+      "learning_rate": 9.923326670881336e-05,
+      "loss": 1.7315,
+      "step": 4918
+    },
+    {
+      "epoch": 0.5175170962651237,
+      "grad_norm": 1.082401156425476,
+      "learning_rate": 9.919919038661736e-05,
+      "loss": 1.8861,
+      "step": 4919
+    },
+    {
+      "epoch": 0.5176223040504997,
+      "grad_norm": 1.2021406888961792,
+      "learning_rate": 9.916511415741676e-05,
+      "loss": 2.2311,
+      "step": 4920
+    },
+    {
+      "epoch": 0.5177275118358758,
+      "grad_norm": 1.4951965808868408,
+      "learning_rate": 9.913103802516868e-05,
+      "loss": 2.1998,
+      "step": 4921
+    },
+    {
+      "epoch": 0.517832719621252,
+      "grad_norm": 1.098451852798462,
+      "learning_rate": 9.90969619938303e-05,
+      "loss": 2.01,
+      "step": 4922
+    },
+    {
+      "epoch": 0.5179379274066281,
+      "grad_norm": 1.2300734519958496,
+      "learning_rate": 9.906288606735875e-05,
+      "loss": 1.8119,
+      "step": 4923
+    },
+    {
+      "epoch": 0.5180431351920042,
+      "grad_norm": 1.2791273593902588,
+      "learning_rate": 9.902881024971116e-05,
+      "loss": 1.8411,
+      "step": 4924
+    },
+    {
+      "epoch": 0.5181483429773803,
+      "grad_norm": 1.747589111328125,
+      "learning_rate": 9.899473454484461e-05,
+      "loss": 1.8983,
+      "step": 4925
+    },
+    {
+      "epoch": 0.5182535507627565,
+      "grad_norm": 1.192456841468811,
+      "learning_rate": 9.896065895671625e-05,
+      "loss": 1.4597,
+      "step": 4926
+    },
+    {
+      "epoch": 0.5183587585481325,
+      "grad_norm": 1.400451898574829,
+      "learning_rate": 9.892658348928316e-05,
+      "loss": 2.15,
+      "step": 4927
+    },
+    {
+      "epoch": 0.5184639663335087,
+      "grad_norm": 1.4197369813919067,
+      "learning_rate": 9.88925081465024e-05,
+      "loss": 1.6032,
+      "step": 4928
+    },
+    {
+      "epoch": 0.5185691741188848,
+      "grad_norm": 1.509582757949829,
+      "learning_rate": 9.8858432932331e-05,
+      "loss": 1.9569,
+      "step": 4929
+    },
+    {
+      "epoch": 0.518674381904261,
+      "grad_norm": 0.9736523628234863,
+      "learning_rate": 9.882435785072601e-05,
+      "loss": 1.6721,
+      "step": 4930
+    },
+    {
+      "epoch": 0.518779589689637,
+      "grad_norm": 1.5283832550048828,
+      "learning_rate": 9.87902829056445e-05,
+      "loss": 2.0695,
+      "step": 4931
+    },
+    {
+      "epoch": 0.5188847974750131,
+      "grad_norm": 1.4057093858718872,
+      "learning_rate": 9.875620810104344e-05,
+      "loss": 1.4839,
+      "step": 4932
+    },
+    {
+      "epoch": 0.5189900052603893,
+      "grad_norm": 1.614888310432434,
+      "learning_rate": 9.872213344087983e-05,
+      "loss": 1.7947,
+      "step": 4933
+    },
+    {
+      "epoch": 0.5190952130457653,
+      "grad_norm": 1.5216361284255981,
+      "learning_rate": 9.868805892911067e-05,
+      "loss": 1.3341,
+      "step": 4934
+    },
+    {
+      "epoch": 0.5192004208311415,
+      "grad_norm": 1.41942298412323,
+      "learning_rate": 9.86539845696929e-05,
+      "loss": 1.8061,
+      "step": 4935
+    },
+    {
+      "epoch": 0.5193056286165176,
+      "grad_norm": 1.3126589059829712,
+      "learning_rate": 9.861991036658345e-05,
+      "loss": 2.092,
+      "step": 4936
+    },
+    {
+      "epoch": 0.5194108364018938,
+      "grad_norm": 2.1999170780181885,
+      "learning_rate": 9.858583632373927e-05,
+      "loss": 1.9388,
+      "step": 4937
+    },
+    {
+      "epoch": 0.5195160441872698,
+      "grad_norm": 1.6807349920272827,
+      "learning_rate": 9.85517624451173e-05,
+      "loss": 1.3968,
+      "step": 4938
+    },
+    {
+      "epoch": 0.519621251972646,
+      "grad_norm": 1.121988296508789,
+      "learning_rate": 9.851768873467435e-05,
+      "loss": 1.8882,
+      "step": 4939
+    },
+    {
+      "epoch": 0.5197264597580221,
+      "grad_norm": 1.7207212448120117,
+      "learning_rate": 9.848361519636733e-05,
+      "loss": 1.5875,
+      "step": 4940
+    },
+    {
+      "epoch": 0.5198316675433982,
+      "grad_norm": 1.8505074977874756,
+      "learning_rate": 9.84495418341531e-05,
+      "loss": 2.0285,
+      "step": 4941
+    },
+    {
+      "epoch": 0.5199368753287743,
+      "grad_norm": 1.3527370691299438,
+      "learning_rate": 9.841546865198846e-05,
+      "loss": 1.8778,
+      "step": 4942
+    },
+    {
+      "epoch": 0.5200420831141505,
+      "grad_norm": 1.1984349489212036,
+      "learning_rate": 9.838139565383022e-05,
+      "loss": 1.9808,
+      "step": 4943
+    },
+    {
+      "epoch": 0.5201472908995266,
+      "grad_norm": 1.9260704517364502,
+      "learning_rate": 9.834732284363519e-05,
+      "loss": 1.8141,
+      "step": 4944
+    },
+    {
+      "epoch": 0.5202524986849026,
+      "grad_norm": 1.8520371913909912,
+      "learning_rate": 9.83132502253601e-05,
+      "loss": 2.1056,
+      "step": 4945
+    },
+    {
+      "epoch": 0.5203577064702788,
+      "grad_norm": 1.291551113128662,
+      "learning_rate": 9.827917780296172e-05,
+      "loss": 1.7853,
+      "step": 4946
+    },
+    {
+      "epoch": 0.5204629142556549,
+      "grad_norm": 1.5085182189941406,
+      "learning_rate": 9.824510558039675e-05,
+      "loss": 1.8278,
+      "step": 4947
+    },
+    {
+      "epoch": 0.5205681220410311,
+      "grad_norm": 1.3979451656341553,
+      "learning_rate": 9.821103356162189e-05,
+      "loss": 1.2452,
+      "step": 4948
+    },
+    {
+      "epoch": 0.5206733298264071,
+      "grad_norm": 1.249040961265564,
+      "learning_rate": 9.817696175059381e-05,
+      "loss": 1.2894,
+      "step": 4949
+    },
+    {
+      "epoch": 0.5207785376117833,
+      "grad_norm": 1.443249225616455,
+      "learning_rate": 9.814289015126919e-05,
+      "loss": 1.8148,
+      "step": 4950
+    },
+    {
+      "epoch": 0.5208837453971594,
+      "grad_norm": 1.2137818336486816,
+      "learning_rate": 9.81088187676046e-05,
+      "loss": 2.0707,
+      "step": 4951
+    },
+    {
+      "epoch": 0.5209889531825355,
+      "grad_norm": 1.0801608562469482,
+      "learning_rate": 9.807474760355665e-05,
+      "loss": 1.5743,
+      "step": 4952
+    },
+    {
+      "epoch": 0.5210941609679116,
+      "grad_norm": 1.206287145614624,
+      "learning_rate": 9.804067666308192e-05,
+      "loss": 2.013,
+      "step": 4953
+    },
+    {
+      "epoch": 0.5211993687532878,
+      "grad_norm": 1.502138376235962,
+      "learning_rate": 9.800660595013696e-05,
+      "loss": 1.8673,
+      "step": 4954
+    },
+    {
+      "epoch": 0.5213045765386639,
+      "grad_norm": 1.6293638944625854,
+      "learning_rate": 9.797253546867831e-05,
+      "loss": 2.0488,
+      "step": 4955
+    },
+    {
+      "epoch": 0.52140978432404,
+      "grad_norm": 1.6563540697097778,
+      "learning_rate": 9.79384652226624e-05,
+      "loss": 1.6133,
+      "step": 4956
+    },
+    {
+      "epoch": 0.5215149921094161,
+      "grad_norm": 1.4221042394638062,
+      "learning_rate": 9.790439521604574e-05,
+      "loss": 1.7739,
+      "step": 4957
+    },
+    {
+      "epoch": 0.5216201998947922,
+      "grad_norm": 1.3993518352508545,
+      "learning_rate": 9.78703254527848e-05,
+      "loss": 1.2957,
+      "step": 4958
+    },
+    {
+      "epoch": 0.5217254076801683,
+      "grad_norm": 1.2816994190216064,
+      "learning_rate": 9.783625593683592e-05,
+      "loss": 1.9906,
+      "step": 4959
+    },
+    {
+      "epoch": 0.5218306154655444,
+      "grad_norm": 1.2781740427017212,
+      "learning_rate": 9.78021866721555e-05,
+      "loss": 1.5746,
+      "step": 4960
+    },
+    {
+      "epoch": 0.5219358232509206,
+      "grad_norm": 1.2438249588012695,
+      "learning_rate": 9.776811766269993e-05,
+      "loss": 1.978,
+      "step": 4961
+    },
+    {
+      "epoch": 0.5220410310362967,
+      "grad_norm": 2.31378436088562,
+      "learning_rate": 9.773404891242551e-05,
+      "loss": 1.8975,
+      "step": 4962
+    },
+    {
+      "epoch": 0.5221462388216728,
+      "grad_norm": 1.96122407913208,
+      "learning_rate": 9.769998042528852e-05,
+      "loss": 2.2517,
+      "step": 4963
+    },
+    {
+      "epoch": 0.5222514466070489,
+      "grad_norm": 1.7919749021530151,
+      "learning_rate": 9.766591220524521e-05,
+      "loss": 1.7129,
+      "step": 4964
+    },
+    {
+      "epoch": 0.5223566543924251,
+      "grad_norm": 1.509630799293518,
+      "learning_rate": 9.763184425625186e-05,
+      "loss": 1.589,
+      "step": 4965
+    },
+    {
+      "epoch": 0.5224618621778011,
+      "grad_norm": 1.3050771951675415,
+      "learning_rate": 9.759777658226462e-05,
+      "loss": 1.8745,
+      "step": 4966
+    },
+    {
+      "epoch": 0.5225670699631773,
+      "grad_norm": 2.0204238891601562,
+      "learning_rate": 9.756370918723968e-05,
+      "loss": 2.0122,
+      "step": 4967
+    },
+    {
+      "epoch": 0.5226722777485534,
+      "grad_norm": 1.1685175895690918,
+      "learning_rate": 9.752964207513318e-05,
+      "loss": 1.5934,
+      "step": 4968
+    },
+    {
+      "epoch": 0.5227774855339296,
+      "grad_norm": 2.0005977153778076,
+      "learning_rate": 9.749557524990121e-05,
+      "loss": 1.8383,
+      "step": 4969
+    },
+    {
+      "epoch": 0.5228826933193056,
+      "grad_norm": 1.3754639625549316,
+      "learning_rate": 9.746150871549981e-05,
+      "loss": 2.2025,
+      "step": 4970
+    },
+    {
+      "epoch": 0.5229879011046817,
+      "grad_norm": 1.4466034173965454,
+      "learning_rate": 9.742744247588512e-05,
+      "loss": 1.8922,
+      "step": 4971
+    },
+    {
+      "epoch": 0.5230931088900579,
+      "grad_norm": 1.596903920173645,
+      "learning_rate": 9.739337653501299e-05,
+      "loss": 1.8912,
+      "step": 4972
+    },
+    {
+      "epoch": 0.5231983166754339,
+      "grad_norm": 1.1891504526138306,
+      "learning_rate": 9.73593108968395e-05,
+      "loss": 1.5827,
+      "step": 4973
+    },
+    {
+      "epoch": 0.5233035244608101,
+      "grad_norm": 1.6135132312774658,
+      "learning_rate": 9.732524556532051e-05,
+      "loss": 2.2273,
+      "step": 4974
+    },
+    {
+      "epoch": 0.5234087322461862,
+      "grad_norm": 2.420681953430176,
+      "learning_rate": 9.729118054441194e-05,
+      "loss": 1.9405,
+      "step": 4975
+    },
+    {
+      "epoch": 0.5235139400315624,
+      "grad_norm": 1.0372477769851685,
+      "learning_rate": 9.72571158380697e-05,
+      "loss": 2.0549,
+      "step": 4976
+    },
+    {
+      "epoch": 0.5236191478169384,
+      "grad_norm": 1.2813150882720947,
+      "learning_rate": 9.722305145024951e-05,
+      "loss": 1.3903,
+      "step": 4977
+    },
+    {
+      "epoch": 0.5237243556023146,
+      "grad_norm": 1.1346882581710815,
+      "learning_rate": 9.718898738490723e-05,
+      "loss": 2.2729,
+      "step": 4978
+    },
+    {
+      "epoch": 0.5238295633876907,
+      "grad_norm": 1.6237995624542236,
+      "learning_rate": 9.71549236459986e-05,
+      "loss": 1.7977,
+      "step": 4979
+    },
+    {
+      "epoch": 0.5239347711730669,
+      "grad_norm": 1.280985713005066,
+      "learning_rate": 9.71208602374793e-05,
+      "loss": 1.543,
+      "step": 4980
+    },
+    {
+      "epoch": 0.5240399789584429,
+      "grad_norm": 1.655838131904602,
+      "learning_rate": 9.708679716330504e-05,
+      "loss": 2.5013,
+      "step": 4981
+    },
+    {
+      "epoch": 0.524145186743819,
+      "grad_norm": 1.0946263074874878,
+      "learning_rate": 9.705273442743142e-05,
+      "loss": 1.923,
+      "step": 4982
+    },
+    {
+      "epoch": 0.5242503945291952,
+      "grad_norm": 1.1894264221191406,
+      "learning_rate": 9.701867203381405e-05,
+      "loss": 1.7739,
+      "step": 4983
+    },
+    {
+      "epoch": 0.5243556023145712,
+      "grad_norm": 1.1565569639205933,
+      "learning_rate": 9.698460998640848e-05,
+      "loss": 1.5141,
+      "step": 4984
+    },
+    {
+      "epoch": 0.5244608100999474,
+      "grad_norm": 1.2816983461380005,
+      "learning_rate": 9.695054828917021e-05,
+      "loss": 1.6513,
+      "step": 4985
+    },
+    {
+      "epoch": 0.5245660178853235,
+      "grad_norm": 1.1288788318634033,
+      "learning_rate": 9.691648694605475e-05,
+      "loss": 1.9599,
+      "step": 4986
+    },
+    {
+      "epoch": 0.5246712256706997,
+      "grad_norm": 2.1675658226013184,
+      "learning_rate": 9.688242596101749e-05,
+      "loss": 1.8789,
+      "step": 4987
+    },
+    {
+      "epoch": 0.5247764334560757,
+      "grad_norm": 1.5547009706497192,
+      "learning_rate": 9.684836533801383e-05,
+      "loss": 1.6302,
+      "step": 4988
+    },
+    {
+      "epoch": 0.5248816412414519,
+      "grad_norm": 1.4101799726486206,
+      "learning_rate": 9.681430508099916e-05,
+      "loss": 1.6675,
+      "step": 4989
+    },
+    {
+      "epoch": 0.524986849026828,
+      "grad_norm": 1.8517366647720337,
+      "learning_rate": 9.678024519392871e-05,
+      "loss": 1.555,
+      "step": 4990
+    },
+    {
+      "epoch": 0.525092056812204,
+      "grad_norm": 1.7405954599380493,
+      "learning_rate": 9.67461856807578e-05,
+      "loss": 2.156,
+      "step": 4991
+    },
+    {
+      "epoch": 0.5251972645975802,
+      "grad_norm": 1.4130207300186157,
+      "learning_rate": 9.671212654544167e-05,
+      "loss": 1.2812,
+      "step": 4992
+    },
+    {
+      "epoch": 0.5253024723829564,
+      "grad_norm": 0.9680867791175842,
+      "learning_rate": 9.667806779193541e-05,
+      "loss": 1.6768,
+      "step": 4993
+    },
+    {
+      "epoch": 0.5254076801683325,
+      "grad_norm": 1.184627890586853,
+      "learning_rate": 9.664400942419423e-05,
+      "loss": 1.4777,
+      "step": 4994
+    },
+    {
+      "epoch": 0.5255128879537085,
+      "grad_norm": 1.430696964263916,
+      "learning_rate": 9.660995144617316e-05,
+      "loss": 1.8958,
+      "step": 4995
+    },
+    {
+      "epoch": 0.5256180957390847,
+      "grad_norm": 1.450882911682129,
+      "learning_rate": 9.657589386182725e-05,
+      "loss": 2.0249,
+      "step": 4996
+    },
+    {
+      "epoch": 0.5257233035244608,
+      "grad_norm": 1.3977291584014893,
+      "learning_rate": 9.654183667511154e-05,
+      "loss": 1.873,
+      "step": 4997
+    },
+    {
+      "epoch": 0.5258285113098369,
+      "grad_norm": 1.272571325302124,
+      "learning_rate": 9.650777988998093e-05,
+      "loss": 2.1763,
+      "step": 4998
+    },
+    {
+      "epoch": 0.525933719095213,
+      "grad_norm": 1.565454125404358,
+      "learning_rate": 9.647372351039035e-05,
+      "loss": 2.0786,
+      "step": 4999
+    },
+    {
+      "epoch": 0.5260389268805892,
+      "grad_norm": 1.4186714887619019,
+      "learning_rate": 9.643966754029466e-05,
+      "loss": 1.9929,
+      "step": 5000
+    },
+    {
+      "epoch": 0.5261441346659653,
+      "grad_norm": 1.9095489978790283,
+      "learning_rate": 9.640561198364864e-05,
+      "loss": 1.466,
+      "step": 5001
+    },
+    {
+      "epoch": 0.5262493424513414,
+      "grad_norm": 1.268289566040039,
+      "learning_rate": 9.637155684440705e-05,
+      "loss": 1.3982,
+      "step": 5002
+    },
+    {
+      "epoch": 0.5263545502367175,
+      "grad_norm": 1.090356469154358,
+      "learning_rate": 9.633750212652465e-05,
+      "loss": 2.1568,
+      "step": 5003
+    },
+    {
+      "epoch": 0.5264597580220937,
+      "grad_norm": 2.0192060470581055,
+      "learning_rate": 9.630344783395604e-05,
+      "loss": 1.8322,
+      "step": 5004
+    },
+    {
+      "epoch": 0.5265649658074697,
+      "grad_norm": 0.9348209500312805,
+      "learning_rate": 9.626939397065586e-05,
+      "loss": 1.8744,
+      "step": 5005
+    },
+    {
+      "epoch": 0.5266701735928458,
+      "grad_norm": 1.5496996641159058,
+      "learning_rate": 9.623534054057868e-05,
+      "loss": 1.703,
+      "step": 5006
+    },
+    {
+      "epoch": 0.526775381378222,
+      "grad_norm": 1.0677669048309326,
+      "learning_rate": 9.620128754767904e-05,
+      "loss": 1.8331,
+      "step": 5007
+    },
+    {
+      "epoch": 0.5268805891635981,
+      "grad_norm": 1.6514785289764404,
+      "learning_rate": 9.616723499591131e-05,
+      "loss": 1.8713,
+      "step": 5008
+    },
+    {
+      "epoch": 0.5269857969489742,
+      "grad_norm": 1.1393558979034424,
+      "learning_rate": 9.613318288922999e-05,
+      "loss": 1.6698,
+      "step": 5009
+    },
+    {
+      "epoch": 0.5270910047343503,
+      "grad_norm": 1.6217390298843384,
+      "learning_rate": 9.609913123158941e-05,
+      "loss": 1.7692,
+      "step": 5010
+    },
+    {
+      "epoch": 0.5271962125197265,
+      "grad_norm": 1.725000023841858,
+      "learning_rate": 9.606508002694386e-05,
+      "loss": 1.7788,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5273014203051026,
+      "grad_norm": 1.360186219215393,
+      "learning_rate": 9.603102927924762e-05,
+      "loss": 1.9891,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5274066280904787,
+      "grad_norm": 1.6628353595733643,
+      "learning_rate": 9.59969789924549e-05,
+      "loss": 1.7478,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5275118358758548,
+      "grad_norm": 1.2454009056091309,
+      "learning_rate": 9.596292917051985e-05,
+      "loss": 1.744,
+      "step": 5014
+    },
+    {
+      "epoch": 0.527617043661231,
+      "grad_norm": 1.4857524633407593,
+      "learning_rate": 9.592887981739648e-05,
+      "loss": 1.5022,
+      "step": 5015
+    },
+    {
+      "epoch": 0.527722251446607,
+      "grad_norm": 1.3809852600097656,
+      "learning_rate": 9.58948309370389e-05,
+      "loss": 1.8917,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5278274592319832,
+      "grad_norm": 0.873028039932251,
+      "learning_rate": 9.58607825334011e-05,
+      "loss": 1.7126,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5279326670173593,
+      "grad_norm": 1.1375359296798706,
+      "learning_rate": 9.5826734610437e-05,
+      "loss": 2.3039,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5280378748027355,
+      "grad_norm": 1.4480020999908447,
+      "learning_rate": 9.579268717210045e-05,
+      "loss": 1.6948,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5281430825881115,
+      "grad_norm": 1.870962381362915,
+      "learning_rate": 9.575864022234527e-05,
+      "loss": 1.8897,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5282482903734876,
+      "grad_norm": 1.3602070808410645,
+      "learning_rate": 9.572459376512528e-05,
+      "loss": 2.2099,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5283534981588638,
+      "grad_norm": 1.2275303602218628,
+      "learning_rate": 9.56905478043941e-05,
+      "loss": 1.9035,
+      "step": 5022
+    },
+    {
+      "epoch": 0.5284587059442398,
+      "grad_norm": 1.2830156087875366,
+      "learning_rate": 9.565650234410542e-05,
+      "loss": 1.6825,
+      "step": 5023
+    },
+    {
+      "epoch": 0.528563913729616,
+      "grad_norm": 1.4303338527679443,
+      "learning_rate": 9.562245738821285e-05,
+      "loss": 2.0779,
+      "step": 5024
+    },
+    {
+      "epoch": 0.5286691215149921,
+      "grad_norm": 1.9997981786727905,
+      "learning_rate": 9.558841294066985e-05,
+      "loss": 1.9707,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5287743293003683,
+      "grad_norm": 1.8480234146118164,
+      "learning_rate": 9.555436900542993e-05,
+      "loss": 1.656,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5288795370857443,
+      "grad_norm": 1.244927167892456,
+      "learning_rate": 9.552032558644654e-05,
+      "loss": 2.1726,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5289847448711205,
+      "grad_norm": 1.005563497543335,
+      "learning_rate": 9.548628268767294e-05,
+      "loss": 2.0138,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5290899526564966,
+      "grad_norm": 1.155867576599121,
+      "learning_rate": 9.545224031306249e-05,
+      "loss": 1.6304,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5291951604418726,
+      "grad_norm": 1.5326203107833862,
+      "learning_rate": 9.541819846656839e-05,
+      "loss": 1.7073,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5293003682272488,
+      "grad_norm": 1.6995837688446045,
+      "learning_rate": 9.538415715214383e-05,
+      "loss": 1.8962,
+      "step": 5031
+    },
+    {
+      "epoch": 0.529405576012625,
+      "grad_norm": 1.8225377798080444,
+      "learning_rate": 9.535011637374189e-05,
+      "loss": 2.0861,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5295107837980011,
+      "grad_norm": 1.634724736213684,
+      "learning_rate": 9.53160761353156e-05,
+      "loss": 1.4762,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5296159915833771,
+      "grad_norm": 0.9659777879714966,
+      "learning_rate": 9.528203644081801e-05,
+      "loss": 1.0181,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5297211993687533,
+      "grad_norm": 1.0323597192764282,
+      "learning_rate": 9.5247997294202e-05,
+      "loss": 1.7988,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5298264071541294,
+      "grad_norm": 1.1700800657272339,
+      "learning_rate": 9.521395869942039e-05,
+      "loss": 1.8157,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5299316149395055,
+      "grad_norm": 2.2417099475860596,
+      "learning_rate": 9.517992066042598e-05,
+      "loss": 2.1183,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5300368227248816,
+      "grad_norm": 1.6578564643859863,
+      "learning_rate": 9.514588318117152e-05,
+      "loss": 1.5281,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5301420305102578,
+      "grad_norm": 1.3541558980941772,
+      "learning_rate": 9.511184626560968e-05,
+      "loss": 2.1024,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5302472382956339,
+      "grad_norm": 1.1925020217895508,
+      "learning_rate": 9.507780991769302e-05,
+      "loss": 1.6266,
+      "step": 5040
+    },
+    {
+      "epoch": 0.53035244608101,
+      "grad_norm": 2.338503360748291,
+      "learning_rate": 9.504377414137407e-05,
+      "loss": 2.1032,
+      "step": 5041
+    },
+    {
+      "epoch": 0.5304576538663861,
+      "grad_norm": 1.6318100690841675,
+      "learning_rate": 9.500973894060534e-05,
+      "loss": 1.9588,
+      "step": 5042
+    },
+    {
+      "epoch": 0.5305628616517623,
+      "grad_norm": 1.629852533340454,
+      "learning_rate": 9.497570431933917e-05,
+      "loss": 1.8528,
+      "step": 5043
+    },
+    {
+      "epoch": 0.5306680694371384,
+      "grad_norm": 1.3625423908233643,
+      "learning_rate": 9.494167028152792e-05,
+      "loss": 2.5084,
+      "step": 5044
+    },
+    {
+      "epoch": 0.5307732772225144,
+      "grad_norm": 1.6486890316009521,
+      "learning_rate": 9.490763683112386e-05,
+      "loss": 1.3986,
+      "step": 5045
+    },
+    {
+      "epoch": 0.5308784850078906,
+      "grad_norm": 1.887699007987976,
+      "learning_rate": 9.487360397207916e-05,
+      "loss": 1.9863,
+      "step": 5046
+    },
+    {
+      "epoch": 0.5309836927932667,
+      "grad_norm": 1.5402462482452393,
+      "learning_rate": 9.483957170834593e-05,
+      "loss": 2.2311,
+      "step": 5047
+    },
+    {
+      "epoch": 0.5310889005786428,
+      "grad_norm": 1.120913028717041,
+      "learning_rate": 9.480554004387627e-05,
+      "loss": 1.6302,
+      "step": 5048
+    },
+    {
+      "epoch": 0.5311941083640189,
+      "grad_norm": 1.8108731508255005,
+      "learning_rate": 9.477150898262213e-05,
+      "loss": 1.8564,
+      "step": 5049
+    },
+    {
+      "epoch": 0.5312993161493951,
+      "grad_norm": 0.9820263385772705,
+      "learning_rate": 9.473747852853543e-05,
+      "loss": 2.0829,
+      "step": 5050
+    },
+    {
+      "epoch": 0.5314045239347712,
+      "grad_norm": 2.3030552864074707,
+      "learning_rate": 9.4703448685568e-05,
+      "loss": 1.5117,
+      "step": 5051
+    },
+    {
+      "epoch": 0.5315097317201473,
+      "grad_norm": 1.2382352352142334,
+      "learning_rate": 9.466941945767168e-05,
+      "loss": 1.4252,
+      "step": 5052
+    },
+    {
+      "epoch": 0.5316149395055234,
+      "grad_norm": 1.1163996458053589,
+      "learning_rate": 9.463539084879809e-05,
+      "loss": 1.8226,
+      "step": 5053
+    },
+    {
+      "epoch": 0.5317201472908996,
+      "grad_norm": 1.8982148170471191,
+      "learning_rate": 9.46013628628989e-05,
+      "loss": 1.6211,
+      "step": 5054
+    },
+    {
+      "epoch": 0.5318253550762756,
+      "grad_norm": 1.5627444982528687,
+      "learning_rate": 9.456733550392568e-05,
+      "loss": 1.846,
+      "step": 5055
+    },
+    {
+      "epoch": 0.5319305628616517,
+      "grad_norm": 3.1989519596099854,
+      "learning_rate": 9.453330877582988e-05,
+      "loss": 1.5141,
+      "step": 5056
+    },
+    {
+      "epoch": 0.5320357706470279,
+      "grad_norm": 1.2844103574752808,
+      "learning_rate": 9.449928268256299e-05,
+      "loss": 1.4024,
+      "step": 5057
+    },
+    {
+      "epoch": 0.532140978432404,
+      "grad_norm": 2.4570162296295166,
+      "learning_rate": 9.446525722807623e-05,
+      "loss": 1.7243,
+      "step": 5058
+    },
+    {
+      "epoch": 0.5322461862177801,
+      "grad_norm": 1.4320532083511353,
+      "learning_rate": 9.443123241632093e-05,
+      "loss": 1.5919,
+      "step": 5059
+    },
+    {
+      "epoch": 0.5323513940031562,
+      "grad_norm": 1.297221302986145,
+      "learning_rate": 9.439720825124827e-05,
+      "loss": 1.6516,
+      "step": 5060
+    },
+    {
+      "epoch": 0.5324566017885324,
+      "grad_norm": 1.2610849142074585,
+      "learning_rate": 9.436318473680936e-05,
+      "loss": 1.7887,
+      "step": 5061
+    },
+    {
+      "epoch": 0.5325618095739084,
+      "grad_norm": 1.0165259838104248,
+      "learning_rate": 9.432916187695525e-05,
+      "loss": 1.8535,
+      "step": 5062
+    },
+    {
+      "epoch": 0.5326670173592846,
+      "grad_norm": 1.0816254615783691,
+      "learning_rate": 9.42951396756369e-05,
+      "loss": 1.93,
+      "step": 5063
+    },
+    {
+      "epoch": 0.5327722251446607,
+      "grad_norm": 1.453770637512207,
+      "learning_rate": 9.42611181368052e-05,
+      "loss": 1.601,
+      "step": 5064
+    },
+    {
+      "epoch": 0.5328774329300369,
+      "grad_norm": 1.5351864099502563,
+      "learning_rate": 9.422709726441094e-05,
+      "loss": 2.3019,
+      "step": 5065
+    },
+    {
+      "epoch": 0.5329826407154129,
+      "grad_norm": 1.867706537246704,
+      "learning_rate": 9.419307706240489e-05,
+      "loss": 2.1457,
+      "step": 5066
+    },
+    {
+      "epoch": 0.533087848500789,
+      "grad_norm": 2.1947357654571533,
+      "learning_rate": 9.415905753473765e-05,
+      "loss": 1.4276,
+      "step": 5067
+    },
+    {
+      "epoch": 0.5331930562861652,
+      "grad_norm": 1.1833242177963257,
+      "learning_rate": 9.412503868535983e-05,
+      "loss": 1.9144,
+      "step": 5068
+    },
+    {
+      "epoch": 0.5332982640715412,
+      "grad_norm": 1.284825325012207,
+      "learning_rate": 9.409102051822195e-05,
+      "loss": 2.1141,
+      "step": 5069
+    },
+    {
+      "epoch": 0.5334034718569174,
+      "grad_norm": 1.165414571762085,
+      "learning_rate": 9.405700303727435e-05,
+      "loss": 1.8646,
+      "step": 5070
+    },
+    {
+      "epoch": 0.5335086796422935,
+      "grad_norm": 1.4519579410552979,
+      "learning_rate": 9.402298624646744e-05,
+      "loss": 2.0232,
+      "step": 5071
+    },
+    {
+      "epoch": 0.5336138874276697,
+      "grad_norm": 1.0977115631103516,
+      "learning_rate": 9.398897014975149e-05,
+      "loss": 1.4565,
+      "step": 5072
+    },
+    {
+      "epoch": 0.5337190952130457,
+      "grad_norm": 1.6109386682510376,
+      "learning_rate": 9.39549547510766e-05,
+      "loss": 2.0598,
+      "step": 5073
+    },
+    {
+      "epoch": 0.5338243029984219,
+      "grad_norm": 1.9979568719863892,
+      "learning_rate": 9.392094005439291e-05,
+      "loss": 2.0561,
+      "step": 5074
+    },
+    {
+      "epoch": 0.533929510783798,
+      "grad_norm": 1.9197112321853638,
+      "learning_rate": 9.388692606365043e-05,
+      "loss": 1.81,
+      "step": 5075
+    },
+    {
+      "epoch": 0.5340347185691742,
+      "grad_norm": 1.5256714820861816,
+      "learning_rate": 9.385291278279914e-05,
+      "loss": 1.4039,
+      "step": 5076
+    },
+    {
+      "epoch": 0.5341399263545502,
+      "grad_norm": 1.7796401977539062,
+      "learning_rate": 9.381890021578881e-05,
+      "loss": 2.1833,
+      "step": 5077
+    },
+    {
+      "epoch": 0.5342451341399264,
+      "grad_norm": 1.145056962966919,
+      "learning_rate": 9.37848883665693e-05,
+      "loss": 1.7855,
+      "step": 5078
+    },
+    {
+      "epoch": 0.5343503419253025,
+      "grad_norm": 2.008789300918579,
+      "learning_rate": 9.375087723909017e-05,
+      "loss": 1.6486,
+      "step": 5079
+    },
+    {
+      "epoch": 0.5344555497106785,
+      "grad_norm": 1.488021731376648,
+      "learning_rate": 9.371686683730113e-05,
+      "loss": 1.9833,
+      "step": 5080
+    },
+    {
+      "epoch": 0.5345607574960547,
+      "grad_norm": 1.6001816987991333,
+      "learning_rate": 9.368285716515162e-05,
+      "loss": 1.4451,
+      "step": 5081
+    },
+    {
+      "epoch": 0.5346659652814308,
+      "grad_norm": 1.6554683446884155,
+      "learning_rate": 9.36488482265911e-05,
+      "loss": 2.2062,
+      "step": 5082
+    },
+    {
+      "epoch": 0.534771173066807,
+      "grad_norm": 1.8193860054016113,
+      "learning_rate": 9.361484002556898e-05,
+      "loss": 1.4615,
+      "step": 5083
+    },
+    {
+      "epoch": 0.534876380852183,
+      "grad_norm": 1.3823652267456055,
+      "learning_rate": 9.35808325660344e-05,
+      "loss": 1.9892,
+      "step": 5084
+    },
+    {
+      "epoch": 0.5349815886375592,
+      "grad_norm": 1.5087515115737915,
+      "learning_rate": 9.354682585193662e-05,
+      "loss": 1.9344,
+      "step": 5085
+    },
+    {
+      "epoch": 0.5350867964229353,
+      "grad_norm": 1.761411190032959,
+      "learning_rate": 9.351281988722469e-05,
+      "loss": 1.7704,
+      "step": 5086
+    },
+    {
+      "epoch": 0.5351920042083114,
+      "grad_norm": 1.332608938217163,
+      "learning_rate": 9.347881467584764e-05,
+      "loss": 1.6883,
+      "step": 5087
+    },
+    {
+      "epoch": 0.5352972119936875,
+      "grad_norm": 1.6421012878417969,
+      "learning_rate": 9.344481022175436e-05,
+      "loss": 1.832,
+      "step": 5088
+    },
+    {
+      "epoch": 0.5354024197790637,
+      "grad_norm": 1.594934344291687,
+      "learning_rate": 9.34108065288937e-05,
+      "loss": 1.6753,
+      "step": 5089
+    },
+    {
+      "epoch": 0.5355076275644398,
+      "grad_norm": 1.367395043373108,
+      "learning_rate": 9.337680360121436e-05,
+      "loss": 1.1994,
+      "step": 5090
+    },
+    {
+      "epoch": 0.5356128353498159,
+      "grad_norm": 1.6289167404174805,
+      "learning_rate": 9.334280144266501e-05,
+      "loss": 1.8281,
+      "step": 5091
+    },
+    {
+      "epoch": 0.535718043135192,
+      "grad_norm": 2.1067452430725098,
+      "learning_rate": 9.330880005719422e-05,
+      "loss": 1.4705,
+      "step": 5092
+    },
+    {
+      "epoch": 0.5358232509205682,
+      "grad_norm": 1.5518765449523926,
+      "learning_rate": 9.327479944875045e-05,
+      "loss": 2.2224,
+      "step": 5093
+    },
+    {
+      "epoch": 0.5359284587059442,
+      "grad_norm": 1.0815091133117676,
+      "learning_rate": 9.324079962128207e-05,
+      "loss": 1.8141,
+      "step": 5094
+    },
+    {
+      "epoch": 0.5360336664913203,
+      "grad_norm": 1.461266279220581,
+      "learning_rate": 9.320680057873735e-05,
+      "loss": 2.0772,
+      "step": 5095
+    },
+    {
+      "epoch": 0.5361388742766965,
+      "grad_norm": 1.2924128770828247,
+      "learning_rate": 9.317280232506454e-05,
+      "loss": 1.8573,
+      "step": 5096
+    },
+    {
+      "epoch": 0.5362440820620726,
+      "grad_norm": 1.1167140007019043,
+      "learning_rate": 9.31388048642117e-05,
+      "loss": 2.0632,
+      "step": 5097
+    },
+    {
+      "epoch": 0.5363492898474487,
+      "grad_norm": 1.1252926588058472,
+      "learning_rate": 9.310480820012684e-05,
+      "loss": 2.7284,
+      "step": 5098
+    },
+    {
+      "epoch": 0.5364544976328248,
+      "grad_norm": 1.0790119171142578,
+      "learning_rate": 9.307081233675791e-05,
+      "loss": 1.9891,
+      "step": 5099
+    },
+    {
+      "epoch": 0.536559705418201,
+      "grad_norm": 1.325523018836975,
+      "learning_rate": 9.303681727805276e-05,
+      "loss": 2.269,
+      "step": 5100
+    },
+    {
+      "epoch": 0.536664913203577,
+      "grad_norm": 1.8128808736801147,
+      "learning_rate": 9.300282302795909e-05,
+      "loss": 1.2854,
+      "step": 5101
+    },
+    {
+      "epoch": 0.5367701209889532,
+      "grad_norm": 2.0237619876861572,
+      "learning_rate": 9.29688295904245e-05,
+      "loss": 1.7467,
+      "step": 5102
+    },
+    {
+      "epoch": 0.5368753287743293,
+      "grad_norm": 1.2668648958206177,
+      "learning_rate": 9.293483696939658e-05,
+      "loss": 1.72,
+      "step": 5103
+    },
+    {
+      "epoch": 0.5369805365597055,
+      "grad_norm": 1.171208143234253,
+      "learning_rate": 9.290084516882281e-05,
+      "loss": 1.9828,
+      "step": 5104
+    },
+    {
+      "epoch": 0.5370857443450815,
+      "grad_norm": 1.4353476762771606,
+      "learning_rate": 9.286685419265048e-05,
+      "loss": 1.9993,
+      "step": 5105
+    },
+    {
+      "epoch": 0.5371909521304576,
+      "grad_norm": 1.5309464931488037,
+      "learning_rate": 9.283286404482688e-05,
+      "loss": 2.002,
+      "step": 5106
+    },
+    {
+      "epoch": 0.5372961599158338,
+      "grad_norm": 1.001417636871338,
+      "learning_rate": 9.27988747292992e-05,
+      "loss": 2.2841,
+      "step": 5107
+    },
+    {
+      "epoch": 0.5374013677012099,
+      "grad_norm": 1.408814549446106,
+      "learning_rate": 9.276488625001448e-05,
+      "loss": 1.5659,
+      "step": 5108
+    },
+    {
+      "epoch": 0.537506575486586,
+      "grad_norm": 1.94085693359375,
+      "learning_rate": 9.273089861091969e-05,
+      "loss": 1.3672,
+      "step": 5109
+    },
+    {
+      "epoch": 0.5376117832719621,
+      "grad_norm": 1.7884807586669922,
+      "learning_rate": 9.269691181596169e-05,
+      "loss": 1.5813,
+      "step": 5110
+    },
+    {
+      "epoch": 0.5377169910573383,
+      "grad_norm": 1.3529471158981323,
+      "learning_rate": 9.266292586908732e-05,
+      "loss": 2.0667,
+      "step": 5111
+    },
+    {
+      "epoch": 0.5378221988427143,
+      "grad_norm": 0.8547163009643555,
+      "learning_rate": 9.262894077424317e-05,
+      "loss": 1.8774,
+      "step": 5112
+    },
+    {
+      "epoch": 0.5379274066280905,
+      "grad_norm": 1.3882497549057007,
+      "learning_rate": 9.259495653537586e-05,
+      "loss": 1.9437,
+      "step": 5113
+    },
+    {
+      "epoch": 0.5380326144134666,
+      "grad_norm": 1.9616116285324097,
+      "learning_rate": 9.256097315643188e-05,
+      "loss": 1.5366,
+      "step": 5114
+    },
+    {
+      "epoch": 0.5381378221988428,
+      "grad_norm": 0.89047771692276,
+      "learning_rate": 9.252699064135758e-05,
+      "loss": 2.124,
+      "step": 5115
+    },
+    {
+      "epoch": 0.5382430299842188,
+      "grad_norm": 1.3593010902404785,
+      "learning_rate": 9.249300899409924e-05,
+      "loss": 1.9311,
+      "step": 5116
+    },
+    {
+      "epoch": 0.538348237769595,
+      "grad_norm": 1.1994686126708984,
+      "learning_rate": 9.245902821860308e-05,
+      "loss": 1.8233,
+      "step": 5117
+    },
+    {
+      "epoch": 0.5384534455549711,
+      "grad_norm": 1.4323614835739136,
+      "learning_rate": 9.24250483188151e-05,
+      "loss": 1.5137,
+      "step": 5118
+    },
+    {
+      "epoch": 0.5385586533403471,
+      "grad_norm": 1.0706384181976318,
+      "learning_rate": 9.239106929868133e-05,
+      "loss": 1.4401,
+      "step": 5119
+    },
+    {
+      "epoch": 0.5386638611257233,
+      "grad_norm": 1.6682357788085938,
+      "learning_rate": 9.235709116214764e-05,
+      "loss": 1.8276,
+      "step": 5120
+    },
+    {
+      "epoch": 0.5387690689110994,
+      "grad_norm": 1.3352079391479492,
+      "learning_rate": 9.232311391315979e-05,
+      "loss": 2.0025,
+      "step": 5121
+    },
+    {
+      "epoch": 0.5388742766964756,
+      "grad_norm": 1.127171516418457,
+      "learning_rate": 9.228913755566344e-05,
+      "loss": 1.959,
+      "step": 5122
+    },
+    {
+      "epoch": 0.5389794844818516,
+      "grad_norm": 0.859086275100708,
+      "learning_rate": 9.225516209360413e-05,
+      "loss": 1.5351,
+      "step": 5123
+    },
+    {
+      "epoch": 0.5390846922672278,
+      "grad_norm": 1.3062644004821777,
+      "learning_rate": 9.222118753092735e-05,
+      "loss": 1.8726,
+      "step": 5124
+    },
+    {
+      "epoch": 0.5391899000526039,
+      "grad_norm": 1.8840057849884033,
+      "learning_rate": 9.218721387157846e-05,
+      "loss": 1.8806,
+      "step": 5125
+    },
+    {
+      "epoch": 0.53929510783798,
+      "grad_norm": 1.518869400024414,
+      "learning_rate": 9.215324111950267e-05,
+      "loss": 1.5398,
+      "step": 5126
+    },
+    {
+      "epoch": 0.5394003156233561,
+      "grad_norm": 1.1188104152679443,
+      "learning_rate": 9.211926927864518e-05,
+      "loss": 1.5614,
+      "step": 5127
+    },
+    {
+      "epoch": 0.5395055234087323,
+      "grad_norm": 1.0320559740066528,
+      "learning_rate": 9.208529835295098e-05,
+      "loss": 1.6083,
+      "step": 5128
+    },
+    {
+      "epoch": 0.5396107311941084,
+      "grad_norm": 2.2655322551727295,
+      "learning_rate": 9.205132834636502e-05,
+      "loss": 1.5238,
+      "step": 5129
+    },
+    {
+      "epoch": 0.5397159389794844,
+      "grad_norm": 1.1405690908432007,
+      "learning_rate": 9.201735926283213e-05,
+      "loss": 2.1654,
+      "step": 5130
+    },
+    {
+      "epoch": 0.5398211467648606,
+      "grad_norm": 1.9452557563781738,
+      "learning_rate": 9.198339110629701e-05,
+      "loss": 1.5046,
+      "step": 5131
+    },
+    {
+      "epoch": 0.5399263545502367,
+      "grad_norm": 1.15932035446167,
+      "learning_rate": 9.194942388070431e-05,
+      "loss": 1.6429,
+      "step": 5132
+    },
+    {
+      "epoch": 0.5400315623356128,
+      "grad_norm": 1.0874872207641602,
+      "learning_rate": 9.191545758999848e-05,
+      "loss": 1.3977,
+      "step": 5133
+    },
+    {
+      "epoch": 0.5401367701209889,
+      "grad_norm": 1.428829550743103,
+      "learning_rate": 9.188149223812393e-05,
+      "loss": 1.8418,
+      "step": 5134
+    },
+    {
+      "epoch": 0.5402419779063651,
+      "grad_norm": 1.535487174987793,
+      "learning_rate": 9.1847527829025e-05,
+      "loss": 1.6133,
+      "step": 5135
+    },
+    {
+      "epoch": 0.5403471856917412,
+      "grad_norm": 1.295906662940979,
+      "learning_rate": 9.181356436664578e-05,
+      "loss": 1.9259,
+      "step": 5136
+    },
+    {
+      "epoch": 0.5404523934771173,
+      "grad_norm": 0.8215249180793762,
+      "learning_rate": 9.177960185493036e-05,
+      "loss": 1.8883,
+      "step": 5137
+    },
+    {
+      "epoch": 0.5405576012624934,
+      "grad_norm": 1.4424512386322021,
+      "learning_rate": 9.174564029782275e-05,
+      "loss": 1.9409,
+      "step": 5138
+    },
+    {
+      "epoch": 0.5406628090478696,
+      "grad_norm": 1.4258966445922852,
+      "learning_rate": 9.171167969926672e-05,
+      "loss": 1.5655,
+      "step": 5139
+    },
+    {
+      "epoch": 0.5407680168332457,
+      "grad_norm": 1.681448221206665,
+      "learning_rate": 9.167772006320604e-05,
+      "loss": 1.6488,
+      "step": 5140
+    },
+    {
+      "epoch": 0.5408732246186218,
+      "grad_norm": 1.6626986265182495,
+      "learning_rate": 9.164376139358433e-05,
+      "loss": 1.7347,
+      "step": 5141
+    },
+    {
+      "epoch": 0.5409784324039979,
+      "grad_norm": 1.1256407499313354,
+      "learning_rate": 9.16098036943451e-05,
+      "loss": 2.2423,
+      "step": 5142
+    },
+    {
+      "epoch": 0.541083640189374,
+      "grad_norm": 2.0072197914123535,
+      "learning_rate": 9.157584696943175e-05,
+      "loss": 1.2106,
+      "step": 5143
+    },
+    {
+      "epoch": 0.5411888479747501,
+      "grad_norm": 1.7380682229995728,
+      "learning_rate": 9.154189122278754e-05,
+      "loss": 1.436,
+      "step": 5144
+    },
+    {
+      "epoch": 0.5412940557601262,
+      "grad_norm": 1.4622247219085693,
+      "learning_rate": 9.150793645835562e-05,
+      "loss": 1.0282,
+      "step": 5145
+    },
+    {
+      "epoch": 0.5413992635455024,
+      "grad_norm": 1.5467033386230469,
+      "learning_rate": 9.147398268007912e-05,
+      "loss": 1.4557,
+      "step": 5146
+    },
+    {
+      "epoch": 0.5415044713308785,
+      "grad_norm": 1.8523279428482056,
+      "learning_rate": 9.14400298919009e-05,
+      "loss": 1.7551,
+      "step": 5147
+    },
+    {
+      "epoch": 0.5416096791162546,
+      "grad_norm": 2.7987358570098877,
+      "learning_rate": 9.140607809776382e-05,
+      "loss": 1.4688,
+      "step": 5148
+    },
+    {
+      "epoch": 0.5417148869016307,
+      "grad_norm": 1.0487982034683228,
+      "learning_rate": 9.137212730161062e-05,
+      "loss": 2.1586,
+      "step": 5149
+    },
+    {
+      "epoch": 0.5418200946870069,
+      "grad_norm": 1.263087272644043,
+      "learning_rate": 9.133817750738384e-05,
+      "loss": 1.4165,
+      "step": 5150
+    },
+    {
+      "epoch": 0.5419253024723829,
+      "grad_norm": 1.5688707828521729,
+      "learning_rate": 9.1304228719026e-05,
+      "loss": 1.8595,
+      "step": 5151
+    },
+    {
+      "epoch": 0.5420305102577591,
+      "grad_norm": 1.5955549478530884,
+      "learning_rate": 9.127028094047944e-05,
+      "loss": 1.722,
+      "step": 5152
+    },
+    {
+      "epoch": 0.5421357180431352,
+      "grad_norm": 1.5457823276519775,
+      "learning_rate": 9.123633417568641e-05,
+      "loss": 1.8036,
+      "step": 5153
+    },
+    {
+      "epoch": 0.5422409258285114,
+      "grad_norm": 2.240226984024048,
+      "learning_rate": 9.120238842858903e-05,
+      "loss": 1.8054,
+      "step": 5154
+    },
+    {
+      "epoch": 0.5423461336138874,
+      "grad_norm": 1.7893513441085815,
+      "learning_rate": 9.11684437031293e-05,
+      "loss": 1.8167,
+      "step": 5155
+    },
+    {
+      "epoch": 0.5424513413992635,
+      "grad_norm": 1.6553772687911987,
+      "learning_rate": 9.113450000324914e-05,
+      "loss": 1.9113,
+      "step": 5156
+    },
+    {
+      "epoch": 0.5425565491846397,
+      "grad_norm": 1.1440198421478271,
+      "learning_rate": 9.110055733289029e-05,
+      "loss": 1.7322,
+      "step": 5157
+    },
+    {
+      "epoch": 0.5426617569700157,
+      "grad_norm": 1.0650477409362793,
+      "learning_rate": 9.106661569599442e-05,
+      "loss": 1.9446,
+      "step": 5158
+    },
+    {
+      "epoch": 0.5427669647553919,
+      "grad_norm": 1.3577461242675781,
+      "learning_rate": 9.103267509650305e-05,
+      "loss": 2.0171,
+      "step": 5159
+    },
+    {
+      "epoch": 0.542872172540768,
+      "grad_norm": 1.0990657806396484,
+      "learning_rate": 9.099873553835758e-05,
+      "loss": 1.5223,
+      "step": 5160
+    },
+    {
+      "epoch": 0.5429773803261442,
+      "grad_norm": 1.6959264278411865,
+      "learning_rate": 9.096479702549933e-05,
+      "loss": 2.3626,
+      "step": 5161
+    },
+    {
+      "epoch": 0.5430825881115202,
+      "grad_norm": 1.6704100370407104,
+      "learning_rate": 9.093085956186945e-05,
+      "loss": 2.2676,
+      "step": 5162
+    },
+    {
+      "epoch": 0.5431877958968964,
+      "grad_norm": 2.997234582901001,
+      "learning_rate": 9.089692315140896e-05,
+      "loss": 1.617,
+      "step": 5163
+    },
+    {
+      "epoch": 0.5432930036822725,
+      "grad_norm": 1.61349356174469,
+      "learning_rate": 9.086298779805887e-05,
+      "loss": 1.7279,
+      "step": 5164
+    },
+    {
+      "epoch": 0.5433982114676486,
+      "grad_norm": 1.6692132949829102,
+      "learning_rate": 9.082905350575986e-05,
+      "loss": 1.9036,
+      "step": 5165
+    },
+    {
+      "epoch": 0.5435034192530247,
+      "grad_norm": 1.7874563932418823,
+      "learning_rate": 9.079512027845268e-05,
+      "loss": 2.4346,
+      "step": 5166
+    },
+    {
+      "epoch": 0.5436086270384008,
+      "grad_norm": 1.3825801610946655,
+      "learning_rate": 9.076118812007789e-05,
+      "loss": 1.3331,
+      "step": 5167
+    },
+    {
+      "epoch": 0.543713834823777,
+      "grad_norm": 1.6000703573226929,
+      "learning_rate": 9.072725703457587e-05,
+      "loss": 1.7898,
+      "step": 5168
+    },
+    {
+      "epoch": 0.543819042609153,
+      "grad_norm": 1.600172996520996,
+      "learning_rate": 9.069332702588698e-05,
+      "loss": 1.4787,
+      "step": 5169
+    },
+    {
+      "epoch": 0.5439242503945292,
+      "grad_norm": 1.7228039503097534,
+      "learning_rate": 9.065939809795137e-05,
+      "loss": 2.2927,
+      "step": 5170
+    },
+    {
+      "epoch": 0.5440294581799053,
+      "grad_norm": 1.3640235662460327,
+      "learning_rate": 9.062547025470908e-05,
+      "loss": 1.8261,
+      "step": 5171
+    },
+    {
+      "epoch": 0.5441346659652815,
+      "grad_norm": 1.3940327167510986,
+      "learning_rate": 9.059154350010008e-05,
+      "loss": 1.997,
+      "step": 5172
+    },
+    {
+      "epoch": 0.5442398737506575,
+      "grad_norm": 1.7132809162139893,
+      "learning_rate": 9.055761783806416e-05,
+      "loss": 2.0858,
+      "step": 5173
+    },
+    {
+      "epoch": 0.5443450815360337,
+      "grad_norm": 1.438615322113037,
+      "learning_rate": 9.052369327254098e-05,
+      "loss": 1.7468,
+      "step": 5174
+    },
+    {
+      "epoch": 0.5444502893214098,
+      "grad_norm": 1.4938368797302246,
+      "learning_rate": 9.048976980747008e-05,
+      "loss": 1.7274,
+      "step": 5175
+    },
+    {
+      "epoch": 0.5445554971067859,
+      "grad_norm": 1.2716962099075317,
+      "learning_rate": 9.045584744679092e-05,
+      "loss": 1.5497,
+      "step": 5176
+    },
+    {
+      "epoch": 0.544660704892162,
+      "grad_norm": 1.7363189458847046,
+      "learning_rate": 9.042192619444275e-05,
+      "loss": 1.6291,
+      "step": 5177
+    },
+    {
+      "epoch": 0.5447659126775382,
+      "grad_norm": 1.5554457902908325,
+      "learning_rate": 9.038800605436475e-05,
+      "loss": 1.8769,
+      "step": 5178
+    },
+    {
+      "epoch": 0.5448711204629143,
+      "grad_norm": 1.222988247871399,
+      "learning_rate": 9.035408703049596e-05,
+      "loss": 1.9806,
+      "step": 5179
+    },
+    {
+      "epoch": 0.5449763282482903,
+      "grad_norm": 1.5716575384140015,
+      "learning_rate": 9.03201691267753e-05,
+      "loss": 2.0254,
+      "step": 5180
+    },
+    {
+      "epoch": 0.5450815360336665,
+      "grad_norm": 1.1745754480361938,
+      "learning_rate": 9.02862523471415e-05,
+      "loss": 2.1139,
+      "step": 5181
+    },
+    {
+      "epoch": 0.5451867438190426,
+      "grad_norm": 0.8382359147071838,
+      "learning_rate": 9.025233669553322e-05,
+      "loss": 1.5266,
+      "step": 5182
+    },
+    {
+      "epoch": 0.5452919516044187,
+      "grad_norm": 1.7974534034729004,
+      "learning_rate": 9.021842217588901e-05,
+      "loss": 1.7337,
+      "step": 5183
+    },
+    {
+      "epoch": 0.5453971593897948,
+      "grad_norm": 1.5042651891708374,
+      "learning_rate": 9.018450879214721e-05,
+      "loss": 2.0298,
+      "step": 5184
+    },
+    {
+      "epoch": 0.545502367175171,
+      "grad_norm": 1.2993311882019043,
+      "learning_rate": 9.015059654824611e-05,
+      "loss": 1.4338,
+      "step": 5185
+    },
+    {
+      "epoch": 0.5456075749605471,
+      "grad_norm": 1.5049631595611572,
+      "learning_rate": 9.011668544812377e-05,
+      "loss": 1.6596,
+      "step": 5186
+    },
+    {
+      "epoch": 0.5457127827459232,
+      "grad_norm": 1.4581938982009888,
+      "learning_rate": 9.00827754957182e-05,
+      "loss": 1.8009,
+      "step": 5187
+    },
+    {
+      "epoch": 0.5458179905312993,
+      "grad_norm": 1.5323725938796997,
+      "learning_rate": 9.004886669496728e-05,
+      "loss": 1.6577,
+      "step": 5188
+    },
+    {
+      "epoch": 0.5459231983166755,
+      "grad_norm": 1.0720787048339844,
+      "learning_rate": 9.001495904980867e-05,
+      "loss": 1.7775,
+      "step": 5189
+    },
+    {
+      "epoch": 0.5460284061020515,
+      "grad_norm": 1.4327290058135986,
+      "learning_rate": 8.998105256418e-05,
+      "loss": 1.7773,
+      "step": 5190
+    },
+    {
+      "epoch": 0.5461336138874276,
+      "grad_norm": 1.6901955604553223,
+      "learning_rate": 8.99471472420187e-05,
+      "loss": 1.8843,
+      "step": 5191
+    },
+    {
+      "epoch": 0.5462388216728038,
+      "grad_norm": 2.1214852333068848,
+      "learning_rate": 8.991324308726209e-05,
+      "loss": 2.1451,
+      "step": 5192
+    },
+    {
+      "epoch": 0.54634402945818,
+      "grad_norm": 1.875383734703064,
+      "learning_rate": 8.987934010384733e-05,
+      "loss": 1.6287,
+      "step": 5193
+    },
+    {
+      "epoch": 0.546449237243556,
+      "grad_norm": 1.438472867012024,
+      "learning_rate": 8.984543829571151e-05,
+      "loss": 1.7141,
+      "step": 5194
+    },
+    {
+      "epoch": 0.5465544450289321,
+      "grad_norm": 1.697603464126587,
+      "learning_rate": 8.981153766679149e-05,
+      "loss": 1.9385,
+      "step": 5195
+    },
+    {
+      "epoch": 0.5466596528143083,
+      "grad_norm": 1.5435723066329956,
+      "learning_rate": 8.977763822102404e-05,
+      "loss": 1.7889,
+      "step": 5196
+    },
+    {
+      "epoch": 0.5467648605996843,
+      "grad_norm": 1.3071941137313843,
+      "learning_rate": 8.97437399623458e-05,
+      "loss": 1.8384,
+      "step": 5197
+    },
+    {
+      "epoch": 0.5468700683850605,
+      "grad_norm": 1.1430269479751587,
+      "learning_rate": 8.970984289469327e-05,
+      "loss": 1.4976,
+      "step": 5198
+    },
+    {
+      "epoch": 0.5469752761704366,
+      "grad_norm": 2.691263198852539,
+      "learning_rate": 8.96759470220028e-05,
+      "loss": 1.39,
+      "step": 5199
+    },
+    {
+      "epoch": 0.5470804839558128,
+      "grad_norm": 1.5296295881271362,
+      "learning_rate": 8.96420523482106e-05,
+      "loss": 1.8981,
+      "step": 5200
+    },
+    {
+      "epoch": 0.5471856917411888,
+      "grad_norm": 2.23455810546875,
+      "learning_rate": 8.960815887725278e-05,
+      "loss": 1.573,
+      "step": 5201
+    },
+    {
+      "epoch": 0.547290899526565,
+      "grad_norm": 2.9234347343444824,
+      "learning_rate": 8.957426661306522e-05,
+      "loss": 1.4707,
+      "step": 5202
+    },
+    {
+      "epoch": 0.5473961073119411,
+      "grad_norm": 1.3259090185165405,
+      "learning_rate": 8.954037555958376e-05,
+      "loss": 1.2053,
+      "step": 5203
+    },
+    {
+      "epoch": 0.5475013150973173,
+      "grad_norm": 1.4656827449798584,
+      "learning_rate": 8.950648572074405e-05,
+      "loss": 1.3973,
+      "step": 5204
+    },
+    {
+      "epoch": 0.5476065228826933,
+      "grad_norm": 1.52262282371521,
+      "learning_rate": 8.947259710048158e-05,
+      "loss": 2.1761,
+      "step": 5205
+    },
+    {
+      "epoch": 0.5477117306680694,
+      "grad_norm": 1.1430667638778687,
+      "learning_rate": 8.943870970273174e-05,
+      "loss": 1.4505,
+      "step": 5206
+    },
+    {
+      "epoch": 0.5478169384534456,
+      "grad_norm": 1.6592637300491333,
+      "learning_rate": 8.940482353142983e-05,
+      "loss": 1.9043,
+      "step": 5207
+    },
+    {
+      "epoch": 0.5479221462388216,
+      "grad_norm": 1.875982642173767,
+      "learning_rate": 8.937093859051083e-05,
+      "loss": 1.9452,
+      "step": 5208
+    },
+    {
+      "epoch": 0.5480273540241978,
+      "grad_norm": 1.242350697517395,
+      "learning_rate": 8.933705488390972e-05,
+      "loss": 1.7839,
+      "step": 5209
+    },
+    {
+      "epoch": 0.5481325618095739,
+      "grad_norm": 1.8692058324813843,
+      "learning_rate": 8.930317241556132e-05,
+      "loss": 1.6776,
+      "step": 5210
+    },
+    {
+      "epoch": 0.5482377695949501,
+      "grad_norm": 1.910408854484558,
+      "learning_rate": 8.926929118940026e-05,
+      "loss": 1.588,
+      "step": 5211
+    },
+    {
+      "epoch": 0.5483429773803261,
+      "grad_norm": 1.608375906944275,
+      "learning_rate": 8.923541120936111e-05,
+      "loss": 1.525,
+      "step": 5212
+    },
+    {
+      "epoch": 0.5484481851657023,
+      "grad_norm": 1.786906123161316,
+      "learning_rate": 8.92015324793782e-05,
+      "loss": 1.5707,
+      "step": 5213
+    },
+    {
+      "epoch": 0.5485533929510784,
+      "grad_norm": 1.4942151308059692,
+      "learning_rate": 8.916765500338575e-05,
+      "loss": 2.0173,
+      "step": 5214
+    },
+    {
+      "epoch": 0.5486586007364544,
+      "grad_norm": 1.761673092842102,
+      "learning_rate": 8.913377878531789e-05,
+      "loss": 1.9173,
+      "step": 5215
+    },
+    {
+      "epoch": 0.5487638085218306,
+      "grad_norm": 1.1698923110961914,
+      "learning_rate": 8.909990382910849e-05,
+      "loss": 1.9928,
+      "step": 5216
+    },
+    {
+      "epoch": 0.5488690163072067,
+      "grad_norm": 1.3080923557281494,
+      "learning_rate": 8.906603013869136e-05,
+      "loss": 2.0938,
+      "step": 5217
+    },
+    {
+      "epoch": 0.5489742240925829,
+      "grad_norm": 1.443345546722412,
+      "learning_rate": 8.903215771800017e-05,
+      "loss": 2.204,
+      "step": 5218
+    },
+    {
+      "epoch": 0.5490794318779589,
+      "grad_norm": 1.6220980882644653,
+      "learning_rate": 8.899828657096838e-05,
+      "loss": 1.4677,
+      "step": 5219
+    },
+    {
+      "epoch": 0.5491846396633351,
+      "grad_norm": 2.169625759124756,
+      "learning_rate": 8.896441670152932e-05,
+      "loss": 1.6339,
+      "step": 5220
+    },
+    {
+      "epoch": 0.5492898474487112,
+      "grad_norm": 0.9750475287437439,
+      "learning_rate": 8.893054811361624e-05,
+      "loss": 1.76,
+      "step": 5221
+    },
+    {
+      "epoch": 0.5493950552340873,
+      "grad_norm": 1.8625184297561646,
+      "learning_rate": 8.889668081116214e-05,
+      "loss": 1.6533,
+      "step": 5222
+    },
+    {
+      "epoch": 0.5495002630194634,
+      "grad_norm": 1.3964723348617554,
+      "learning_rate": 8.886281479809993e-05,
+      "loss": 2.0727,
+      "step": 5223
+    },
+    {
+      "epoch": 0.5496054708048396,
+      "grad_norm": 1.3396962881088257,
+      "learning_rate": 8.882895007836236e-05,
+      "loss": 1.5633,
+      "step": 5224
+    },
+    {
+      "epoch": 0.5497106785902157,
+      "grad_norm": 1.373002290725708,
+      "learning_rate": 8.879508665588206e-05,
+      "loss": 1.6469,
+      "step": 5225
+    },
+    {
+      "epoch": 0.5498158863755918,
+      "grad_norm": 0.9774206876754761,
+      "learning_rate": 8.876122453459143e-05,
+      "loss": 1.7567,
+      "step": 5226
+    },
+    {
+      "epoch": 0.5499210941609679,
+      "grad_norm": 1.5650739669799805,
+      "learning_rate": 8.872736371842279e-05,
+      "loss": 1.8958,
+      "step": 5227
+    },
+    {
+      "epoch": 0.550026301946344,
+      "grad_norm": 1.5855236053466797,
+      "learning_rate": 8.869350421130831e-05,
+      "loss": 2.1006,
+      "step": 5228
+    },
+    {
+      "epoch": 0.5501315097317201,
+      "grad_norm": 1.8107531070709229,
+      "learning_rate": 8.865964601717994e-05,
+      "loss": 1.9863,
+      "step": 5229
+    },
+    {
+      "epoch": 0.5502367175170962,
+      "grad_norm": 1.2894846200942993,
+      "learning_rate": 8.862578913996952e-05,
+      "loss": 1.749,
+      "step": 5230
+    },
+    {
+      "epoch": 0.5503419253024724,
+      "grad_norm": 1.1580836772918701,
+      "learning_rate": 8.859193358360874e-05,
+      "loss": 1.3304,
+      "step": 5231
+    },
+    {
+      "epoch": 0.5504471330878485,
+      "grad_norm": 1.6148262023925781,
+      "learning_rate": 8.855807935202915e-05,
+      "loss": 1.5428,
+      "step": 5232
+    },
+    {
+      "epoch": 0.5505523408732246,
+      "grad_norm": 1.7478631734848022,
+      "learning_rate": 8.852422644916216e-05,
+      "loss": 1.6134,
+      "step": 5233
+    },
+    {
+      "epoch": 0.5506575486586007,
+      "grad_norm": 1.809343934059143,
+      "learning_rate": 8.849037487893893e-05,
+      "loss": 2.3462,
+      "step": 5234
+    },
+    {
+      "epoch": 0.5507627564439769,
+      "grad_norm": 1.5731574296951294,
+      "learning_rate": 8.845652464529057e-05,
+      "loss": 1.7138,
+      "step": 5235
+    },
+    {
+      "epoch": 0.550867964229353,
+      "grad_norm": 1.3270047903060913,
+      "learning_rate": 8.842267575214802e-05,
+      "loss": 2.0084,
+      "step": 5236
+    },
+    {
+      "epoch": 0.5509731720147291,
+      "grad_norm": 1.761651873588562,
+      "learning_rate": 8.838882820344198e-05,
+      "loss": 1.4156,
+      "step": 5237
+    },
+    {
+      "epoch": 0.5510783798001052,
+      "grad_norm": 1.6372829675674438,
+      "learning_rate": 8.835498200310309e-05,
+      "loss": 2.1276,
+      "step": 5238
+    },
+    {
+      "epoch": 0.5511835875854814,
+      "grad_norm": 1.144127368927002,
+      "learning_rate": 8.832113715506181e-05,
+      "loss": 1.574,
+      "step": 5239
+    },
+    {
+      "epoch": 0.5512887953708574,
+      "grad_norm": 1.8449068069458008,
+      "learning_rate": 8.82872936632484e-05,
+      "loss": 1.67,
+      "step": 5240
+    },
+    {
+      "epoch": 0.5513940031562335,
+      "grad_norm": 1.1163779497146606,
+      "learning_rate": 8.825345153159301e-05,
+      "loss": 1.6902,
+      "step": 5241
+    },
+    {
+      "epoch": 0.5514992109416097,
+      "grad_norm": 1.2700437307357788,
+      "learning_rate": 8.821961076402563e-05,
+      "loss": 2.3529,
+      "step": 5242
+    },
+    {
+      "epoch": 0.5516044187269858,
+      "grad_norm": 1.2813799381256104,
+      "learning_rate": 8.818577136447603e-05,
+      "loss": 1.6812,
+      "step": 5243
+    },
+    {
+      "epoch": 0.5517096265123619,
+      "grad_norm": 1.321010947227478,
+      "learning_rate": 8.815193333687391e-05,
+      "loss": 1.7545,
+      "step": 5244
+    },
+    {
+      "epoch": 0.551814834297738,
+      "grad_norm": 1.3672521114349365,
+      "learning_rate": 8.811809668514878e-05,
+      "loss": 1.5022,
+      "step": 5245
+    },
+    {
+      "epoch": 0.5519200420831142,
+      "grad_norm": 1.245273470878601,
+      "learning_rate": 8.808426141322994e-05,
+      "loss": 1.6858,
+      "step": 5246
+    },
+    {
+      "epoch": 0.5520252498684902,
+      "grad_norm": 2.0327329635620117,
+      "learning_rate": 8.805042752504656e-05,
+      "loss": 1.8182,
+      "step": 5247
+    },
+    {
+      "epoch": 0.5521304576538664,
+      "grad_norm": 1.9399306774139404,
+      "learning_rate": 8.801659502452769e-05,
+      "loss": 1.8135,
+      "step": 5248
+    },
+    {
+      "epoch": 0.5522356654392425,
+      "grad_norm": 1.4941593408584595,
+      "learning_rate": 8.79827639156022e-05,
+      "loss": 2.0688,
+      "step": 5249
+    },
+    {
+      "epoch": 0.5523408732246187,
+      "grad_norm": 1.5331664085388184,
+      "learning_rate": 8.794893420219881e-05,
+      "loss": 2.1115,
+      "step": 5250
+    },
+    {
+      "epoch": 0.5524460810099947,
+      "grad_norm": 2.205839157104492,
+      "learning_rate": 8.791510588824594e-05,
+      "loss": 1.3747,
+      "step": 5251
+    },
+    {
+      "epoch": 0.5525512887953709,
+      "grad_norm": 1.1780728101730347,
+      "learning_rate": 8.788127897767204e-05,
+      "loss": 1.7812,
+      "step": 5252
+    },
+    {
+      "epoch": 0.552656496580747,
+      "grad_norm": 1.1612539291381836,
+      "learning_rate": 8.784745347440533e-05,
+      "loss": 2.0561,
+      "step": 5253
+    },
+    {
+      "epoch": 0.552761704366123,
+      "grad_norm": 1.275850534439087,
+      "learning_rate": 8.78136293823738e-05,
+      "loss": 1.6603,
+      "step": 5254
+    },
+    {
+      "epoch": 0.5528669121514992,
+      "grad_norm": 1.7113317251205444,
+      "learning_rate": 8.777980670550536e-05,
+      "loss": 1.4027,
+      "step": 5255
+    },
+    {
+      "epoch": 0.5529721199368753,
+      "grad_norm": 1.0461790561676025,
+      "learning_rate": 8.774598544772774e-05,
+      "loss": 2.2595,
+      "step": 5256
+    },
+    {
+      "epoch": 0.5530773277222515,
+      "grad_norm": 1.1092756986618042,
+      "learning_rate": 8.771216561296849e-05,
+      "loss": 1.6448,
+      "step": 5257
+    },
+    {
+      "epoch": 0.5531825355076275,
+      "grad_norm": 1.1810952425003052,
+      "learning_rate": 8.767834720515496e-05,
+      "loss": 1.8668,
+      "step": 5258
+    },
+    {
+      "epoch": 0.5532877432930037,
+      "grad_norm": 1.1021080017089844,
+      "learning_rate": 8.76445302282144e-05,
+      "loss": 2.0023,
+      "step": 5259
+    },
+    {
+      "epoch": 0.5533929510783798,
+      "grad_norm": 1.4798521995544434,
+      "learning_rate": 8.761071468607388e-05,
+      "loss": 1.8638,
+      "step": 5260
+    },
+    {
+      "epoch": 0.5534981588637559,
+      "grad_norm": 1.3366169929504395,
+      "learning_rate": 8.757690058266025e-05,
+      "loss": 1.458,
+      "step": 5261
+    },
+    {
+      "epoch": 0.553603366649132,
+      "grad_norm": 1.5138647556304932,
+      "learning_rate": 8.754308792190024e-05,
+      "loss": 1.6667,
+      "step": 5262
+    },
+    {
+      "epoch": 0.5537085744345082,
+      "grad_norm": 1.3789781332015991,
+      "learning_rate": 8.750927670772044e-05,
+      "loss": 2.1057,
+      "step": 5263
+    },
+    {
+      "epoch": 0.5538137822198843,
+      "grad_norm": 1.4035643339157104,
+      "learning_rate": 8.747546694404717e-05,
+      "loss": 1.6416,
+      "step": 5264
+    },
+    {
+      "epoch": 0.5539189900052603,
+      "grad_norm": 1.30898916721344,
+      "learning_rate": 8.744165863480669e-05,
+      "loss": 1.9435,
+      "step": 5265
+    },
+    {
+      "epoch": 0.5540241977906365,
+      "grad_norm": 1.4918098449707031,
+      "learning_rate": 8.740785178392505e-05,
+      "loss": 1.9183,
+      "step": 5266
+    },
+    {
+      "epoch": 0.5541294055760126,
+      "grad_norm": 1.285561203956604,
+      "learning_rate": 8.737404639532811e-05,
+      "loss": 1.82,
+      "step": 5267
+    },
+    {
+      "epoch": 0.5542346133613888,
+      "grad_norm": 1.183773159980774,
+      "learning_rate": 8.734024247294157e-05,
+      "loss": 1.9191,
+      "step": 5268
+    },
+    {
+      "epoch": 0.5543398211467648,
+      "grad_norm": 1.6294718980789185,
+      "learning_rate": 8.7306440020691e-05,
+      "loss": 1.9698,
+      "step": 5269
+    },
+    {
+      "epoch": 0.554445028932141,
+      "grad_norm": 1.5693155527114868,
+      "learning_rate": 8.727263904250178e-05,
+      "loss": 1.8887,
+      "step": 5270
+    },
+    {
+      "epoch": 0.5545502367175171,
+      "grad_norm": 1.6532758474349976,
+      "learning_rate": 8.723883954229908e-05,
+      "loss": 1.8751,
+      "step": 5271
+    },
+    {
+      "epoch": 0.5546554445028932,
+      "grad_norm": 1.02011239528656,
+      "learning_rate": 8.72050415240079e-05,
+      "loss": 2.3712,
+      "step": 5272
+    },
+    {
+      "epoch": 0.5547606522882693,
+      "grad_norm": 1.6323237419128418,
+      "learning_rate": 8.71712449915531e-05,
+      "loss": 1.7668,
+      "step": 5273
+    },
+    {
+      "epoch": 0.5548658600736455,
+      "grad_norm": 1.436021327972412,
+      "learning_rate": 8.713744994885938e-05,
+      "loss": 1.8613,
+      "step": 5274
+    },
+    {
+      "epoch": 0.5549710678590216,
+      "grad_norm": 1.5757768154144287,
+      "learning_rate": 8.710365639985126e-05,
+      "loss": 2.0038,
+      "step": 5275
+    },
+    {
+      "epoch": 0.5550762756443977,
+      "grad_norm": 2.4132494926452637,
+      "learning_rate": 8.706986434845302e-05,
+      "loss": 1.5683,
+      "step": 5276
+    },
+    {
+      "epoch": 0.5551814834297738,
+      "grad_norm": 1.5619103908538818,
+      "learning_rate": 8.703607379858889e-05,
+      "loss": 1.2968,
+      "step": 5277
+    },
+    {
+      "epoch": 0.55528669121515,
+      "grad_norm": 1.2575358152389526,
+      "learning_rate": 8.70022847541828e-05,
+      "loss": 1.5361,
+      "step": 5278
+    },
+    {
+      "epoch": 0.555391899000526,
+      "grad_norm": 0.924577534198761,
+      "learning_rate": 8.696849721915859e-05,
+      "loss": 1.6288,
+      "step": 5279
+    },
+    {
+      "epoch": 0.5554971067859021,
+      "grad_norm": 1.1372270584106445,
+      "learning_rate": 8.693471119743987e-05,
+      "loss": 2.1464,
+      "step": 5280
+    },
+    {
+      "epoch": 0.5556023145712783,
+      "grad_norm": 1.3003607988357544,
+      "learning_rate": 8.690092669295014e-05,
+      "loss": 2.0537,
+      "step": 5281
+    },
+    {
+      "epoch": 0.5557075223566544,
+      "grad_norm": 1.147006869316101,
+      "learning_rate": 8.686714370961264e-05,
+      "loss": 2.0348,
+      "step": 5282
+    },
+    {
+      "epoch": 0.5558127301420305,
+      "grad_norm": 1.6022502183914185,
+      "learning_rate": 8.68333622513505e-05,
+      "loss": 1.6282,
+      "step": 5283
+    },
+    {
+      "epoch": 0.5559179379274066,
+      "grad_norm": 1.5122016668319702,
+      "learning_rate": 8.679958232208668e-05,
+      "loss": 1.2036,
+      "step": 5284
+    },
+    {
+      "epoch": 0.5560231457127828,
+      "grad_norm": 1.583200216293335,
+      "learning_rate": 8.676580392574385e-05,
+      "loss": 1.5517,
+      "step": 5285
+    },
+    {
+      "epoch": 0.5561283534981588,
+      "grad_norm": 1.0503642559051514,
+      "learning_rate": 8.673202706624464e-05,
+      "loss": 1.9161,
+      "step": 5286
+    },
+    {
+      "epoch": 0.556233561283535,
+      "grad_norm": 1.8765935897827148,
+      "learning_rate": 8.669825174751144e-05,
+      "loss": 2.1381,
+      "step": 5287
+    },
+    {
+      "epoch": 0.5563387690689111,
+      "grad_norm": 0.8225523233413696,
+      "learning_rate": 8.666447797346648e-05,
+      "loss": 1.781,
+      "step": 5288
+    },
+    {
+      "epoch": 0.5564439768542873,
+      "grad_norm": 1.6632425785064697,
+      "learning_rate": 8.663070574803175e-05,
+      "loss": 1.636,
+      "step": 5289
+    },
+    {
+      "epoch": 0.5565491846396633,
+      "grad_norm": 1.561035394668579,
+      "learning_rate": 8.659693507512917e-05,
+      "loss": 1.7061,
+      "step": 5290
+    },
+    {
+      "epoch": 0.5566543924250394,
+      "grad_norm": 1.1021442413330078,
+      "learning_rate": 8.656316595868037e-05,
+      "loss": 1.7368,
+      "step": 5291
+    },
+    {
+      "epoch": 0.5567596002104156,
+      "grad_norm": 1.2675588130950928,
+      "learning_rate": 8.652939840260686e-05,
+      "loss": 1.6662,
+      "step": 5292
+    },
+    {
+      "epoch": 0.5568648079957917,
+      "grad_norm": 1.4895219802856445,
+      "learning_rate": 8.649563241082998e-05,
+      "loss": 1.3631,
+      "step": 5293
+    },
+    {
+      "epoch": 0.5569700157811678,
+      "grad_norm": 1.1941972970962524,
+      "learning_rate": 8.64618679872708e-05,
+      "loss": 1.2894,
+      "step": 5294
+    },
+    {
+      "epoch": 0.5570752235665439,
+      "grad_norm": 1.6961616277694702,
+      "learning_rate": 8.642810513585035e-05,
+      "loss": 2.1377,
+      "step": 5295
+    },
+    {
+      "epoch": 0.5571804313519201,
+      "grad_norm": 1.7371389865875244,
+      "learning_rate": 8.639434386048932e-05,
+      "loss": 2.2303,
+      "step": 5296
+    },
+    {
+      "epoch": 0.5572856391372961,
+      "grad_norm": 1.3116867542266846,
+      "learning_rate": 8.636058416510836e-05,
+      "loss": 1.7173,
+      "step": 5297
+    },
+    {
+      "epoch": 0.5573908469226723,
+      "grad_norm": 1.2672263383865356,
+      "learning_rate": 8.632682605362784e-05,
+      "loss": 1.8597,
+      "step": 5298
+    },
+    {
+      "epoch": 0.5574960547080484,
+      "grad_norm": 1.0156499147415161,
+      "learning_rate": 8.629306952996797e-05,
+      "loss": 2.0679,
+      "step": 5299
+    },
+    {
+      "epoch": 0.5576012624934246,
+      "grad_norm": 1.6344144344329834,
+      "learning_rate": 8.625931459804881e-05,
+      "loss": 1.8525,
+      "step": 5300
+    },
+    {
+      "epoch": 0.5577064702788006,
+      "grad_norm": 1.6262001991271973,
+      "learning_rate": 8.622556126179023e-05,
+      "loss": 1.9396,
+      "step": 5301
+    },
+    {
+      "epoch": 0.5578116780641768,
+      "grad_norm": 1.4433337450027466,
+      "learning_rate": 8.619180952511181e-05,
+      "loss": 2.3541,
+      "step": 5302
+    },
+    {
+      "epoch": 0.5579168858495529,
+      "grad_norm": 1.4532362222671509,
+      "learning_rate": 8.61580593919331e-05,
+      "loss": 1.6864,
+      "step": 5303
+    },
+    {
+      "epoch": 0.5580220936349289,
+      "grad_norm": 1.5361088514328003,
+      "learning_rate": 8.612431086617337e-05,
+      "loss": 1.7575,
+      "step": 5304
+    },
+    {
+      "epoch": 0.5581273014203051,
+      "grad_norm": 1.3419499397277832,
+      "learning_rate": 8.609056395175175e-05,
+      "loss": 1.6097,
+      "step": 5305
+    },
+    {
+      "epoch": 0.5582325092056812,
+      "grad_norm": 1.3803333044052124,
+      "learning_rate": 8.605681865258712e-05,
+      "loss": 1.5305,
+      "step": 5306
+    },
+    {
+      "epoch": 0.5583377169910574,
+      "grad_norm": 1.5263311862945557,
+      "learning_rate": 8.602307497259821e-05,
+      "loss": 1.6276,
+      "step": 5307
+    },
+    {
+      "epoch": 0.5584429247764334,
+      "grad_norm": 1.4065600633621216,
+      "learning_rate": 8.598933291570361e-05,
+      "loss": 1.5332,
+      "step": 5308
+    },
+    {
+      "epoch": 0.5585481325618096,
+      "grad_norm": 1.0995397567749023,
+      "learning_rate": 8.595559248582161e-05,
+      "loss": 2.0225,
+      "step": 5309
+    },
+    {
+      "epoch": 0.5586533403471857,
+      "grad_norm": 1.0431246757507324,
+      "learning_rate": 8.592185368687043e-05,
+      "loss": 1.662,
+      "step": 5310
+    },
+    {
+      "epoch": 0.5587585481325618,
+      "grad_norm": 1.5991910696029663,
+      "learning_rate": 8.588811652276803e-05,
+      "loss": 1.2915,
+      "step": 5311
+    },
+    {
+      "epoch": 0.5588637559179379,
+      "grad_norm": 1.2698297500610352,
+      "learning_rate": 8.585438099743217e-05,
+      "loss": 2.1424,
+      "step": 5312
+    },
+    {
+      "epoch": 0.5589689637033141,
+      "grad_norm": 1.9178346395492554,
+      "learning_rate": 8.582064711478046e-05,
+      "loss": 1.7453,
+      "step": 5313
+    },
+    {
+      "epoch": 0.5590741714886902,
+      "grad_norm": 1.5896426439285278,
+      "learning_rate": 8.578691487873036e-05,
+      "loss": 1.0764,
+      "step": 5314
+    },
+    {
+      "epoch": 0.5591793792740662,
+      "grad_norm": 1.0997956991195679,
+      "learning_rate": 8.575318429319899e-05,
+      "loss": 1.588,
+      "step": 5315
+    },
+    {
+      "epoch": 0.5592845870594424,
+      "grad_norm": 1.4343537092208862,
+      "learning_rate": 8.571945536210342e-05,
+      "loss": 1.578,
+      "step": 5316
+    },
+    {
+      "epoch": 0.5593897948448185,
+      "grad_norm": 1.9470630884170532,
+      "learning_rate": 8.568572808936047e-05,
+      "loss": 1.905,
+      "step": 5317
+    },
+    {
+      "epoch": 0.5594950026301946,
+      "grad_norm": 1.999913215637207,
+      "learning_rate": 8.565200247888678e-05,
+      "loss": 2.0373,
+      "step": 5318
+    },
+    {
+      "epoch": 0.5596002104155707,
+      "grad_norm": 1.2631433010101318,
+      "learning_rate": 8.56182785345988e-05,
+      "loss": 2.4557,
+      "step": 5319
+    },
+    {
+      "epoch": 0.5597054182009469,
+      "grad_norm": 0.9738912582397461,
+      "learning_rate": 8.558455626041277e-05,
+      "loss": 1.4726,
+      "step": 5320
+    },
+    {
+      "epoch": 0.559810625986323,
+      "grad_norm": 1.5372016429901123,
+      "learning_rate": 8.555083566024474e-05,
+      "loss": 2.1864,
+      "step": 5321
+    },
+    {
+      "epoch": 0.5599158337716991,
+      "grad_norm": 1.7038294076919556,
+      "learning_rate": 8.551711673801062e-05,
+      "loss": 1.8453,
+      "step": 5322
+    },
+    {
+      "epoch": 0.5600210415570752,
+      "grad_norm": 1.2219566106796265,
+      "learning_rate": 8.548339949762601e-05,
+      "loss": 1.3113,
+      "step": 5323
+    },
+    {
+      "epoch": 0.5601262493424514,
+      "grad_norm": 2.9071662425994873,
+      "learning_rate": 8.544968394300642e-05,
+      "loss": 1.8082,
+      "step": 5324
+    },
+    {
+      "epoch": 0.5602314571278275,
+      "grad_norm": 2.0915303230285645,
+      "learning_rate": 8.541597007806712e-05,
+      "loss": 1.5558,
+      "step": 5325
+    },
+    {
+      "epoch": 0.5603366649132036,
+      "grad_norm": 1.6373728513717651,
+      "learning_rate": 8.538225790672322e-05,
+      "loss": 1.5953,
+      "step": 5326
+    },
+    {
+      "epoch": 0.5604418726985797,
+      "grad_norm": 1.2306228876113892,
+      "learning_rate": 8.534854743288954e-05,
+      "loss": 1.5297,
+      "step": 5327
+    },
+    {
+      "epoch": 0.5605470804839559,
+      "grad_norm": 1.990144968032837,
+      "learning_rate": 8.531483866048081e-05,
+      "loss": 1.6063,
+      "step": 5328
+    },
+    {
+      "epoch": 0.5606522882693319,
+      "grad_norm": 1.5562758445739746,
+      "learning_rate": 8.528113159341153e-05,
+      "loss": 1.6421,
+      "step": 5329
+    },
+    {
+      "epoch": 0.560757496054708,
+      "grad_norm": 1.1921128034591675,
+      "learning_rate": 8.524742623559594e-05,
+      "loss": 1.4861,
+      "step": 5330
+    },
+    {
+      "epoch": 0.5608627038400842,
+      "grad_norm": 1.379715085029602,
+      "learning_rate": 8.521372259094818e-05,
+      "loss": 1.4598,
+      "step": 5331
+    },
+    {
+      "epoch": 0.5609679116254603,
+      "grad_norm": 2.179091453552246,
+      "learning_rate": 8.518002066338212e-05,
+      "loss": 1.9094,
+      "step": 5332
+    },
+    {
+      "epoch": 0.5610731194108364,
+      "grad_norm": 1.716064214706421,
+      "learning_rate": 8.514632045681145e-05,
+      "loss": 1.6363,
+      "step": 5333
+    },
+    {
+      "epoch": 0.5611783271962125,
+      "grad_norm": 1.1381585597991943,
+      "learning_rate": 8.511262197514968e-05,
+      "loss": 1.5435,
+      "step": 5334
+    },
+    {
+      "epoch": 0.5612835349815887,
+      "grad_norm": 2.24033522605896,
+      "learning_rate": 8.507892522231012e-05,
+      "loss": 1.8934,
+      "step": 5335
+    },
+    {
+      "epoch": 0.5613887427669647,
+      "grad_norm": 1.3724555969238281,
+      "learning_rate": 8.504523020220583e-05,
+      "loss": 2.2629,
+      "step": 5336
+    },
+    {
+      "epoch": 0.5614939505523409,
+      "grad_norm": 1.862369418144226,
+      "learning_rate": 8.501153691874971e-05,
+      "loss": 2.0515,
+      "step": 5337
+    },
+    {
+      "epoch": 0.561599158337717,
+      "grad_norm": 2.476644992828369,
+      "learning_rate": 8.497784537585444e-05,
+      "loss": 1.9697,
+      "step": 5338
+    },
+    {
+      "epoch": 0.5617043661230932,
+      "grad_norm": 1.3707612752914429,
+      "learning_rate": 8.494415557743252e-05,
+      "loss": 2.1986,
+      "step": 5339
+    },
+    {
+      "epoch": 0.5618095739084692,
+      "grad_norm": 2.085458993911743,
+      "learning_rate": 8.491046752739624e-05,
+      "loss": 1.9783,
+      "step": 5340
+    },
+    {
+      "epoch": 0.5619147816938453,
+      "grad_norm": 1.2586299180984497,
+      "learning_rate": 8.487678122965767e-05,
+      "loss": 2.0315,
+      "step": 5341
+    },
+    {
+      "epoch": 0.5620199894792215,
+      "grad_norm": 1.0489764213562012,
+      "learning_rate": 8.48430966881287e-05,
+      "loss": 1.7353,
+      "step": 5342
+    },
+    {
+      "epoch": 0.5621251972645975,
+      "grad_norm": 1.1171035766601562,
+      "learning_rate": 8.480941390672101e-05,
+      "loss": 1.8519,
+      "step": 5343
+    },
+    {
+      "epoch": 0.5622304050499737,
+      "grad_norm": 1.3247672319412231,
+      "learning_rate": 8.477573288934605e-05,
+      "loss": 1.7206,
+      "step": 5344
+    },
+    {
+      "epoch": 0.5623356128353498,
+      "grad_norm": 2.1845767498016357,
+      "learning_rate": 8.47420536399151e-05,
+      "loss": 2.2207,
+      "step": 5345
+    },
+    {
+      "epoch": 0.562440820620726,
+      "grad_norm": 1.5519604682922363,
+      "learning_rate": 8.470837616233924e-05,
+      "loss": 1.7987,
+      "step": 5346
+    },
+    {
+      "epoch": 0.562546028406102,
+      "grad_norm": 1.6114767789840698,
+      "learning_rate": 8.467470046052927e-05,
+      "loss": 1.608,
+      "step": 5347
+    },
+    {
+      "epoch": 0.5626512361914782,
+      "grad_norm": 1.703809142112732,
+      "learning_rate": 8.464102653839588e-05,
+      "loss": 1.2808,
+      "step": 5348
+    },
+    {
+      "epoch": 0.5627564439768543,
+      "grad_norm": 1.431298851966858,
+      "learning_rate": 8.460735439984949e-05,
+      "loss": 2.1355,
+      "step": 5349
+    },
+    {
+      "epoch": 0.5628616517622304,
+      "grad_norm": 1.7516570091247559,
+      "learning_rate": 8.457368404880037e-05,
+      "loss": 2.1989,
+      "step": 5350
+    },
+    {
+      "epoch": 0.5629668595476065,
+      "grad_norm": 1.4657411575317383,
+      "learning_rate": 8.454001548915851e-05,
+      "loss": 1.5914,
+      "step": 5351
+    },
+    {
+      "epoch": 0.5630720673329827,
+      "grad_norm": 1.8436520099639893,
+      "learning_rate": 8.450634872483374e-05,
+      "loss": 2.0691,
+      "step": 5352
+    },
+    {
+      "epoch": 0.5631772751183588,
+      "grad_norm": 1.4249285459518433,
+      "learning_rate": 8.44726837597357e-05,
+      "loss": 1.4602,
+      "step": 5353
+    },
+    {
+      "epoch": 0.5632824829037348,
+      "grad_norm": 1.0019652843475342,
+      "learning_rate": 8.443902059777373e-05,
+      "loss": 2.0361,
+      "step": 5354
+    },
+    {
+      "epoch": 0.563387690689111,
+      "grad_norm": 2.051858901977539,
+      "learning_rate": 8.440535924285706e-05,
+      "loss": 1.5908,
+      "step": 5355
+    },
+    {
+      "epoch": 0.5634928984744871,
+      "grad_norm": 1.3849743604660034,
+      "learning_rate": 8.43716996988947e-05,
+      "loss": 1.8964,
+      "step": 5356
+    },
+    {
+      "epoch": 0.5635981062598633,
+      "grad_norm": 1.2312546968460083,
+      "learning_rate": 8.433804196979541e-05,
+      "loss": 1.8502,
+      "step": 5357
+    },
+    {
+      "epoch": 0.5637033140452393,
+      "grad_norm": 1.2379924058914185,
+      "learning_rate": 8.430438605946769e-05,
+      "loss": 2.0632,
+      "step": 5358
+    },
+    {
+      "epoch": 0.5638085218306155,
+      "grad_norm": 0.9644449353218079,
+      "learning_rate": 8.427073197181993e-05,
+      "loss": 1.6485,
+      "step": 5359
+    },
+    {
+      "epoch": 0.5639137296159916,
+      "grad_norm": 1.0192735195159912,
+      "learning_rate": 8.423707971076026e-05,
+      "loss": 1.4893,
+      "step": 5360
+    },
+    {
+      "epoch": 0.5640189374013677,
+      "grad_norm": 1.3611501455307007,
+      "learning_rate": 8.420342928019666e-05,
+      "loss": 1.7624,
+      "step": 5361
+    },
+    {
+      "epoch": 0.5641241451867438,
+      "grad_norm": 1.648772120475769,
+      "learning_rate": 8.416978068403676e-05,
+      "loss": 1.7394,
+      "step": 5362
+    },
+    {
+      "epoch": 0.56422935297212,
+      "grad_norm": 1.2542641162872314,
+      "learning_rate": 8.413613392618811e-05,
+      "loss": 2.0594,
+      "step": 5363
+    },
+    {
+      "epoch": 0.5643345607574961,
+      "grad_norm": 1.6775392293930054,
+      "learning_rate": 8.410248901055801e-05,
+      "loss": 2.0371,
+      "step": 5364
+    },
+    {
+      "epoch": 0.5644397685428721,
+      "grad_norm": 1.1697038412094116,
+      "learning_rate": 8.40688459410535e-05,
+      "loss": 1.8887,
+      "step": 5365
+    },
+    {
+      "epoch": 0.5645449763282483,
+      "grad_norm": 1.5925973653793335,
+      "learning_rate": 8.403520472158143e-05,
+      "loss": 1.8341,
+      "step": 5366
+    },
+    {
+      "epoch": 0.5646501841136244,
+      "grad_norm": 1.3041836023330688,
+      "learning_rate": 8.40015653560485e-05,
+      "loss": 1.5598,
+      "step": 5367
+    },
+    {
+      "epoch": 0.5647553918990005,
+      "grad_norm": 1.0811697244644165,
+      "learning_rate": 8.396792784836108e-05,
+      "loss": 1.3874,
+      "step": 5368
+    },
+    {
+      "epoch": 0.5648605996843766,
+      "grad_norm": 1.3267872333526611,
+      "learning_rate": 8.393429220242541e-05,
+      "loss": 2.1926,
+      "step": 5369
+    },
+    {
+      "epoch": 0.5649658074697528,
+      "grad_norm": 1.133404016494751,
+      "learning_rate": 8.39006584221475e-05,
+      "loss": 1.7818,
+      "step": 5370
+    },
+    {
+      "epoch": 0.5650710152551289,
+      "grad_norm": 2.1951334476470947,
+      "learning_rate": 8.38670265114331e-05,
+      "loss": 2.0216,
+      "step": 5371
+    },
+    {
+      "epoch": 0.565176223040505,
+      "grad_norm": 1.3310232162475586,
+      "learning_rate": 8.383339647418777e-05,
+      "loss": 1.5981,
+      "step": 5372
+    },
+    {
+      "epoch": 0.5652814308258811,
+      "grad_norm": 0.9504427313804626,
+      "learning_rate": 8.379976831431689e-05,
+      "loss": 1.7203,
+      "step": 5373
+    },
+    {
+      "epoch": 0.5653866386112573,
+      "grad_norm": 1.273130178451538,
+      "learning_rate": 8.376614203572559e-05,
+      "loss": 1.5829,
+      "step": 5374
+    },
+    {
+      "epoch": 0.5654918463966333,
+      "grad_norm": 1.1506158113479614,
+      "learning_rate": 8.373251764231872e-05,
+      "loss": 1.3818,
+      "step": 5375
+    },
+    {
+      "epoch": 0.5655970541820095,
+      "grad_norm": 1.0576092004776,
+      "learning_rate": 8.369889513800102e-05,
+      "loss": 1.9378,
+      "step": 5376
+    },
+    {
+      "epoch": 0.5657022619673856,
+      "grad_norm": 1.4205738306045532,
+      "learning_rate": 8.366527452667698e-05,
+      "loss": 1.9037,
+      "step": 5377
+    },
+    {
+      "epoch": 0.5658074697527617,
+      "grad_norm": 1.1215096712112427,
+      "learning_rate": 8.363165581225083e-05,
+      "loss": 1.6023,
+      "step": 5378
+    },
+    {
+      "epoch": 0.5659126775381378,
+      "grad_norm": 1.6068058013916016,
+      "learning_rate": 8.359803899862655e-05,
+      "loss": 2.4349,
+      "step": 5379
+    },
+    {
+      "epoch": 0.5660178853235139,
+      "grad_norm": 1.7272415161132812,
+      "learning_rate": 8.356442408970799e-05,
+      "loss": 1.3773,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5661230931088901,
+      "grad_norm": 1.9546716213226318,
+      "learning_rate": 8.353081108939874e-05,
+      "loss": 1.6571,
+      "step": 5381
+    },
+    {
+      "epoch": 0.5662283008942661,
+      "grad_norm": 1.531908392906189,
+      "learning_rate": 8.349720000160218e-05,
+      "loss": 1.9016,
+      "step": 5382
+    },
+    {
+      "epoch": 0.5663335086796423,
+      "grad_norm": 1.5750274658203125,
+      "learning_rate": 8.346359083022143e-05,
+      "loss": 2.044,
+      "step": 5383
+    },
+    {
+      "epoch": 0.5664387164650184,
+      "grad_norm": 1.3168236017227173,
+      "learning_rate": 8.342998357915942e-05,
+      "loss": 1.6516,
+      "step": 5384
+    },
+    {
+      "epoch": 0.5665439242503946,
+      "grad_norm": 1.2597466707229614,
+      "learning_rate": 8.339637825231887e-05,
+      "loss": 1.4705,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5666491320357706,
+      "grad_norm": 1.7140848636627197,
+      "learning_rate": 8.336277485360223e-05,
+      "loss": 1.9122,
+      "step": 5386
+    },
+    {
+      "epoch": 0.5667543398211468,
+      "grad_norm": 1.403868556022644,
+      "learning_rate": 8.332917338691175e-05,
+      "loss": 1.7833,
+      "step": 5387
+    },
+    {
+      "epoch": 0.5668595476065229,
+      "grad_norm": 1.871900200843811,
+      "learning_rate": 8.32955738561495e-05,
+      "loss": 1.7798,
+      "step": 5388
+    },
+    {
+      "epoch": 0.5669647553918991,
+      "grad_norm": 1.4764777421951294,
+      "learning_rate": 8.326197626521723e-05,
+      "loss": 1.9589,
+      "step": 5389
+    },
+    {
+      "epoch": 0.5670699631772751,
+      "grad_norm": 1.7492798566818237,
+      "learning_rate": 8.322838061801653e-05,
+      "loss": 2.1874,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5671751709626512,
+      "grad_norm": 1.9016979932785034,
+      "learning_rate": 8.319478691844878e-05,
+      "loss": 1.4618,
+      "step": 5391
+    },
+    {
+      "epoch": 0.5672803787480274,
+      "grad_norm": 1.8420443534851074,
+      "learning_rate": 8.316119517041508e-05,
+      "loss": 1.305,
+      "step": 5392
+    },
+    {
+      "epoch": 0.5673855865334034,
+      "grad_norm": 1.3935943841934204,
+      "learning_rate": 8.312760537781632e-05,
+      "loss": 1.9563,
+      "step": 5393
+    },
+    {
+      "epoch": 0.5674907943187796,
+      "grad_norm": 1.9522449970245361,
+      "learning_rate": 8.30940175445532e-05,
+      "loss": 1.5606,
+      "step": 5394
+    },
+    {
+      "epoch": 0.5675960021041557,
+      "grad_norm": 1.159475326538086,
+      "learning_rate": 8.306043167452617e-05,
+      "loss": 1.5782,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5677012098895319,
+      "grad_norm": 1.9530138969421387,
+      "learning_rate": 8.30268477716354e-05,
+      "loss": 1.8486,
+      "step": 5396
+    },
+    {
+      "epoch": 0.5678064176749079,
+      "grad_norm": 1.253212571144104,
+      "learning_rate": 8.299326583978092e-05,
+      "loss": 1.7982,
+      "step": 5397
+    },
+    {
+      "epoch": 0.5679116254602841,
+      "grad_norm": 1.5111079216003418,
+      "learning_rate": 8.29596858828625e-05,
+      "loss": 1.5008,
+      "step": 5398
+    },
+    {
+      "epoch": 0.5680168332456602,
+      "grad_norm": 1.4527966976165771,
+      "learning_rate": 8.292610790477962e-05,
+      "loss": 2.2226,
+      "step": 5399
+    },
+    {
+      "epoch": 0.5681220410310363,
+      "grad_norm": 1.5884594917297363,
+      "learning_rate": 8.289253190943164e-05,
+      "loss": 1.9317,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5682272488164124,
+      "grad_norm": 1.4073702096939087,
+      "learning_rate": 8.285895790071757e-05,
+      "loss": 1.4504,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5683324566017885,
+      "grad_norm": 1.079156756401062,
+      "learning_rate": 8.282538588253627e-05,
+      "loss": 1.2555,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5684376643871647,
+      "grad_norm": 1.4192665815353394,
+      "learning_rate": 8.279181585878635e-05,
+      "loss": 1.947,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5685428721725407,
+      "grad_norm": 1.4388915300369263,
+      "learning_rate": 8.275824783336618e-05,
+      "loss": 1.5275,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5686480799579169,
+      "grad_norm": 0.9725925326347351,
+      "learning_rate": 8.272468181017391e-05,
+      "loss": 1.8777,
+      "step": 5405
+    },
+    {
+      "epoch": 0.568753287743293,
+      "grad_norm": 2.05253005027771,
+      "learning_rate": 8.26911177931075e-05,
+      "loss": 1.1397,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5688584955286691,
+      "grad_norm": 1.8297544717788696,
+      "learning_rate": 8.265755578606456e-05,
+      "loss": 1.6646,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5689637033140452,
+      "grad_norm": 1.0543756484985352,
+      "learning_rate": 8.262399579294253e-05,
+      "loss": 2.0218,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5690689110994214,
+      "grad_norm": 1.6449626684188843,
+      "learning_rate": 8.259043781763869e-05,
+      "loss": 1.4181,
+      "step": 5409
+    },
+    {
+      "epoch": 0.5691741188847975,
+      "grad_norm": 1.7541320323944092,
+      "learning_rate": 8.255688186404996e-05,
+      "loss": 1.8855,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5692793266701736,
+      "grad_norm": 2.0391685962677,
+      "learning_rate": 8.25233279360731e-05,
+      "loss": 1.7674,
+      "step": 5411
+    },
+    {
+      "epoch": 0.5693845344555497,
+      "grad_norm": 1.9726836681365967,
+      "learning_rate": 8.248977603760464e-05,
+      "loss": 1.7674,
+      "step": 5412
+    },
+    {
+      "epoch": 0.5694897422409259,
+      "grad_norm": 1.3250395059585571,
+      "learning_rate": 8.245622617254079e-05,
+      "loss": 1.7336,
+      "step": 5413
+    },
+    {
+      "epoch": 0.5695949500263019,
+      "grad_norm": 1.848076581954956,
+      "learning_rate": 8.242267834477764e-05,
+      "loss": 1.9363,
+      "step": 5414
+    },
+    {
+      "epoch": 0.569700157811678,
+      "grad_norm": 1.8783268928527832,
+      "learning_rate": 8.238913255821099e-05,
+      "loss": 1.3876,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5698053655970542,
+      "grad_norm": 1.4165843725204468,
+      "learning_rate": 8.235558881673637e-05,
+      "loss": 2.0615,
+      "step": 5416
+    },
+    {
+      "epoch": 0.5699105733824303,
+      "grad_norm": 1.4009727239608765,
+      "learning_rate": 8.232204712424911e-05,
+      "loss": 1.9336,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5700157811678064,
+      "grad_norm": 1.3084132671356201,
+      "learning_rate": 8.228850748464431e-05,
+      "loss": 1.7258,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5701209889531825,
+      "grad_norm": 1.5547186136245728,
+      "learning_rate": 8.225496990181684e-05,
+      "loss": 1.4471,
+      "step": 5419
+    },
+    {
+      "epoch": 0.5702261967385587,
+      "grad_norm": 1.5895652770996094,
+      "learning_rate": 8.222143437966124e-05,
+      "loss": 1.815,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5703314045239348,
+      "grad_norm": 1.7180938720703125,
+      "learning_rate": 8.218790092207199e-05,
+      "loss": 1.4393,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5704366123093109,
+      "grad_norm": 1.1043965816497803,
+      "learning_rate": 8.21543695329431e-05,
+      "loss": 1.5099,
+      "step": 5422
+    },
+    {
+      "epoch": 0.570541820094687,
+      "grad_norm": 1.113439917564392,
+      "learning_rate": 8.212084021616852e-05,
+      "loss": 1.9908,
+      "step": 5423
+    },
+    {
+      "epoch": 0.5706470278800632,
+      "grad_norm": 1.640676736831665,
+      "learning_rate": 8.208731297564189e-05,
+      "loss": 1.6397,
+      "step": 5424
+    },
+    {
+      "epoch": 0.5707522356654392,
+      "grad_norm": 1.0427812337875366,
+      "learning_rate": 8.205378781525662e-05,
+      "loss": 1.6538,
+      "step": 5425
+    },
+    {
+      "epoch": 0.5708574434508153,
+      "grad_norm": 1.614121437072754,
+      "learning_rate": 8.202026473890588e-05,
+      "loss": 1.6651,
+      "step": 5426
+    },
+    {
+      "epoch": 0.5709626512361915,
+      "grad_norm": 1.6778579950332642,
+      "learning_rate": 8.198674375048257e-05,
+      "loss": 1.6189,
+      "step": 5427
+    },
+    {
+      "epoch": 0.5710678590215676,
+      "grad_norm": 1.0318078994750977,
+      "learning_rate": 8.195322485387939e-05,
+      "loss": 1.7629,
+      "step": 5428
+    },
+    {
+      "epoch": 0.5711730668069437,
+      "grad_norm": 1.1207058429718018,
+      "learning_rate": 8.191970805298881e-05,
+      "loss": 1.6471,
+      "step": 5429
+    },
+    {
+      "epoch": 0.5712782745923198,
+      "grad_norm": 1.0314679145812988,
+      "learning_rate": 8.188619335170298e-05,
+      "loss": 1.5387,
+      "step": 5430
+    },
+    {
+      "epoch": 0.571383482377696,
+      "grad_norm": 1.1926995515823364,
+      "learning_rate": 8.185268075391388e-05,
+      "loss": 1.5988,
+      "step": 5431
+    },
+    {
+      "epoch": 0.571488690163072,
+      "grad_norm": 1.2411372661590576,
+      "learning_rate": 8.181917026351318e-05,
+      "loss": 1.2326,
+      "step": 5432
+    },
+    {
+      "epoch": 0.5715938979484482,
+      "grad_norm": 1.1884357929229736,
+      "learning_rate": 8.17856618843924e-05,
+      "loss": 1.7397,
+      "step": 5433
+    },
+    {
+      "epoch": 0.5716991057338243,
+      "grad_norm": 1.4349546432495117,
+      "learning_rate": 8.175215562044272e-05,
+      "loss": 1.6069,
+      "step": 5434
+    },
+    {
+      "epoch": 0.5718043135192005,
+      "grad_norm": 2.4111011028289795,
+      "learning_rate": 8.17186514755551e-05,
+      "loss": 1.669,
+      "step": 5435
+    },
+    {
+      "epoch": 0.5719095213045765,
+      "grad_norm": 1.5937304496765137,
+      "learning_rate": 8.168514945362031e-05,
+      "loss": 1.7242,
+      "step": 5436
+    },
+    {
+      "epoch": 0.5720147290899527,
+      "grad_norm": 1.468326210975647,
+      "learning_rate": 8.165164955852879e-05,
+      "loss": 1.5671,
+      "step": 5437
+    },
+    {
+      "epoch": 0.5721199368753288,
+      "grad_norm": 1.6039079427719116,
+      "learning_rate": 8.161815179417078e-05,
+      "loss": 1.2596,
+      "step": 5438
+    },
+    {
+      "epoch": 0.5722251446607048,
+      "grad_norm": 1.7768357992172241,
+      "learning_rate": 8.15846561644363e-05,
+      "loss": 2.1289,
+      "step": 5439
+    },
+    {
+      "epoch": 0.572330352446081,
+      "grad_norm": 1.709168553352356,
+      "learning_rate": 8.155116267321503e-05,
+      "loss": 1.5969,
+      "step": 5440
+    },
+    {
+      "epoch": 0.5724355602314571,
+      "grad_norm": 2.2013728618621826,
+      "learning_rate": 8.151767132439649e-05,
+      "loss": 2.0476,
+      "step": 5441
+    },
+    {
+      "epoch": 0.5725407680168333,
+      "grad_norm": 1.7853853702545166,
+      "learning_rate": 8.148418212186992e-05,
+      "loss": 1.9721,
+      "step": 5442
+    },
+    {
+      "epoch": 0.5726459758022093,
+      "grad_norm": 1.9058358669281006,
+      "learning_rate": 8.145069506952436e-05,
+      "loss": 2.2435,
+      "step": 5443
+    },
+    {
+      "epoch": 0.5727511835875855,
+      "grad_norm": 2.1775949001312256,
+      "learning_rate": 8.141721017124847e-05,
+      "loss": 1.8347,
+      "step": 5444
+    },
+    {
+      "epoch": 0.5728563913729616,
+      "grad_norm": 1.0717917680740356,
+      "learning_rate": 8.138372743093076e-05,
+      "loss": 1.486,
+      "step": 5445
+    },
+    {
+      "epoch": 0.5729615991583377,
+      "grad_norm": 1.1905382871627808,
+      "learning_rate": 8.135024685245947e-05,
+      "loss": 1.5313,
+      "step": 5446
+    },
+    {
+      "epoch": 0.5730668069437138,
+      "grad_norm": 1.8441627025604248,
+      "learning_rate": 8.131676843972263e-05,
+      "loss": 1.5609,
+      "step": 5447
+    },
+    {
+      "epoch": 0.57317201472909,
+      "grad_norm": 1.70012366771698,
+      "learning_rate": 8.128329219660791e-05,
+      "loss": 1.4996,
+      "step": 5448
+    },
+    {
+      "epoch": 0.5732772225144661,
+      "grad_norm": 2.055849313735962,
+      "learning_rate": 8.124981812700285e-05,
+      "loss": 1.6694,
+      "step": 5449
+    },
+    {
+      "epoch": 0.5733824302998421,
+      "grad_norm": 1.5213593244552612,
+      "learning_rate": 8.121634623479466e-05,
+      "loss": 1.7859,
+      "step": 5450
+    },
+    {
+      "epoch": 0.5734876380852183,
+      "grad_norm": 1.240997314453125,
+      "learning_rate": 8.118287652387035e-05,
+      "loss": 1.9417,
+      "step": 5451
+    },
+    {
+      "epoch": 0.5735928458705944,
+      "grad_norm": 1.187052607536316,
+      "learning_rate": 8.114940899811662e-05,
+      "loss": 1.7307,
+      "step": 5452
+    },
+    {
+      "epoch": 0.5736980536559706,
+      "grad_norm": 2.3511910438537598,
+      "learning_rate": 8.111594366141993e-05,
+      "loss": 2.0432,
+      "step": 5453
+    },
+    {
+      "epoch": 0.5738032614413466,
+      "grad_norm": 1.187315583229065,
+      "learning_rate": 8.108248051766656e-05,
+      "loss": 1.6997,
+      "step": 5454
+    },
+    {
+      "epoch": 0.5739084692267228,
+      "grad_norm": 1.4285547733306885,
+      "learning_rate": 8.10490195707424e-05,
+      "loss": 2.0346,
+      "step": 5455
+    },
+    {
+      "epoch": 0.5740136770120989,
+      "grad_norm": 1.6566582918167114,
+      "learning_rate": 8.101556082453319e-05,
+      "loss": 2.0706,
+      "step": 5456
+    },
+    {
+      "epoch": 0.574118884797475,
+      "grad_norm": 1.2293633222579956,
+      "learning_rate": 8.098210428292441e-05,
+      "loss": 2.2512,
+      "step": 5457
+    },
+    {
+      "epoch": 0.5742240925828511,
+      "grad_norm": 1.1286085844039917,
+      "learning_rate": 8.094864994980123e-05,
+      "loss": 1.6408,
+      "step": 5458
+    },
+    {
+      "epoch": 0.5743293003682273,
+      "grad_norm": 1.2557892799377441,
+      "learning_rate": 8.091519782904857e-05,
+      "loss": 1.9494,
+      "step": 5459
+    },
+    {
+      "epoch": 0.5744345081536034,
+      "grad_norm": 1.3432778120040894,
+      "learning_rate": 8.088174792455119e-05,
+      "loss": 2.1635,
+      "step": 5460
+    },
+    {
+      "epoch": 0.5745397159389795,
+      "grad_norm": 2.0232627391815186,
+      "learning_rate": 8.084830024019343e-05,
+      "loss": 1.565,
+      "step": 5461
+    },
+    {
+      "epoch": 0.5746449237243556,
+      "grad_norm": 1.3621368408203125,
+      "learning_rate": 8.08148547798595e-05,
+      "loss": 1.755,
+      "step": 5462
+    },
+    {
+      "epoch": 0.5747501315097318,
+      "grad_norm": 1.1986950635910034,
+      "learning_rate": 8.078141154743332e-05,
+      "loss": 1.8158,
+      "step": 5463
+    },
+    {
+      "epoch": 0.5748553392951078,
+      "grad_norm": 1.457479476928711,
+      "learning_rate": 8.074797054679855e-05,
+      "loss": 1.6491,
+      "step": 5464
+    },
+    {
+      "epoch": 0.5749605470804839,
+      "grad_norm": 1.3137367963790894,
+      "learning_rate": 8.071453178183856e-05,
+      "loss": 1.7712,
+      "step": 5465
+    },
+    {
+      "epoch": 0.5750657548658601,
+      "grad_norm": 1.0634950399398804,
+      "learning_rate": 8.068109525643647e-05,
+      "loss": 1.9505,
+      "step": 5466
+    },
+    {
+      "epoch": 0.5751709626512362,
+      "grad_norm": 0.9608885645866394,
+      "learning_rate": 8.064766097447516e-05,
+      "loss": 1.57,
+      "step": 5467
+    },
+    {
+      "epoch": 0.5752761704366123,
+      "grad_norm": 1.5873912572860718,
+      "learning_rate": 8.061422893983729e-05,
+      "loss": 1.6664,
+      "step": 5468
+    },
+    {
+      "epoch": 0.5753813782219884,
+      "grad_norm": 1.4330425262451172,
+      "learning_rate": 8.058079915640515e-05,
+      "loss": 1.8861,
+      "step": 5469
+    },
+    {
+      "epoch": 0.5754865860073646,
+      "grad_norm": 1.3749375343322754,
+      "learning_rate": 8.054737162806086e-05,
+      "loss": 1.9692,
+      "step": 5470
+    },
+    {
+      "epoch": 0.5755917937927406,
+      "grad_norm": 1.9150141477584839,
+      "learning_rate": 8.051394635868626e-05,
+      "loss": 1.6297,
+      "step": 5471
+    },
+    {
+      "epoch": 0.5756970015781168,
+      "grad_norm": 1.4533182382583618,
+      "learning_rate": 8.048052335216289e-05,
+      "loss": 1.2963,
+      "step": 5472
+    },
+    {
+      "epoch": 0.5758022093634929,
+      "grad_norm": 2.1088831424713135,
+      "learning_rate": 8.044710261237207e-05,
+      "loss": 1.4709,
+      "step": 5473
+    },
+    {
+      "epoch": 0.5759074171488691,
+      "grad_norm": 1.277236819267273,
+      "learning_rate": 8.041368414319483e-05,
+      "loss": 1.8557,
+      "step": 5474
+    },
+    {
+      "epoch": 0.5760126249342451,
+      "grad_norm": 1.2016223669052124,
+      "learning_rate": 8.038026794851198e-05,
+      "loss": 1.6482,
+      "step": 5475
+    },
+    {
+      "epoch": 0.5761178327196212,
+      "grad_norm": 1.476108431816101,
+      "learning_rate": 8.034685403220398e-05,
+      "loss": 1.7097,
+      "step": 5476
+    },
+    {
+      "epoch": 0.5762230405049974,
+      "grad_norm": 2.2213351726531982,
+      "learning_rate": 8.031344239815111e-05,
+      "loss": 1.9822,
+      "step": 5477
+    },
+    {
+      "epoch": 0.5763282482903734,
+      "grad_norm": 1.5243347883224487,
+      "learning_rate": 8.028003305023338e-05,
+      "loss": 2.1552,
+      "step": 5478
+    },
+    {
+      "epoch": 0.5764334560757496,
+      "grad_norm": 2.157576322555542,
+      "learning_rate": 8.024662599233043e-05,
+      "loss": 2.0237,
+      "step": 5479
+    },
+    {
+      "epoch": 0.5765386638611257,
+      "grad_norm": 1.6475622653961182,
+      "learning_rate": 8.021322122832178e-05,
+      "loss": 2.1017,
+      "step": 5480
+    },
+    {
+      "epoch": 0.5766438716465019,
+      "grad_norm": 1.4928076267242432,
+      "learning_rate": 8.01798187620866e-05,
+      "loss": 1.6096,
+      "step": 5481
+    },
+    {
+      "epoch": 0.5767490794318779,
+      "grad_norm": 1.6426701545715332,
+      "learning_rate": 8.014641859750379e-05,
+      "loss": 1.8119,
+      "step": 5482
+    },
+    {
+      "epoch": 0.5768542872172541,
+      "grad_norm": 1.8036248683929443,
+      "learning_rate": 8.011302073845201e-05,
+      "loss": 1.9572,
+      "step": 5483
+    },
+    {
+      "epoch": 0.5769594950026302,
+      "grad_norm": 1.9108166694641113,
+      "learning_rate": 8.007962518880966e-05,
+      "loss": 2.202,
+      "step": 5484
+    },
+    {
+      "epoch": 0.5770647027880064,
+      "grad_norm": 1.3631569147109985,
+      "learning_rate": 8.004623195245481e-05,
+      "loss": 1.9852,
+      "step": 5485
+    },
+    {
+      "epoch": 0.5771699105733824,
+      "grad_norm": 1.4796253442764282,
+      "learning_rate": 8.001284103326539e-05,
+      "loss": 1.9356,
+      "step": 5486
+    },
+    {
+      "epoch": 0.5772751183587586,
+      "grad_norm": 1.3997398614883423,
+      "learning_rate": 7.99794524351189e-05,
+      "loss": 1.9229,
+      "step": 5487
+    },
+    {
+      "epoch": 0.5773803261441347,
+      "grad_norm": 0.9555890560150146,
+      "learning_rate": 7.994606616189264e-05,
+      "loss": 1.6695,
+      "step": 5488
+    },
+    {
+      "epoch": 0.5774855339295107,
+      "grad_norm": 3.1957201957702637,
+      "learning_rate": 7.991268221746373e-05,
+      "loss": 2.0273,
+      "step": 5489
+    },
+    {
+      "epoch": 0.5775907417148869,
+      "grad_norm": 1.504439115524292,
+      "learning_rate": 7.987930060570885e-05,
+      "loss": 2.1869,
+      "step": 5490
+    },
+    {
+      "epoch": 0.577695949500263,
+      "grad_norm": 1.781453251838684,
+      "learning_rate": 7.984592133050454e-05,
+      "loss": 1.8198,
+      "step": 5491
+    },
+    {
+      "epoch": 0.5778011572856392,
+      "grad_norm": 1.4022020101547241,
+      "learning_rate": 7.981254439572704e-05,
+      "loss": 1.6377,
+      "step": 5492
+    },
+    {
+      "epoch": 0.5779063650710152,
+      "grad_norm": 2.5264246463775635,
+      "learning_rate": 7.977916980525227e-05,
+      "loss": 1.8286,
+      "step": 5493
+    },
+    {
+      "epoch": 0.5780115728563914,
+      "grad_norm": 1.4550338983535767,
+      "learning_rate": 7.974579756295591e-05,
+      "loss": 1.9897,
+      "step": 5494
+    },
+    {
+      "epoch": 0.5781167806417675,
+      "grad_norm": 1.6223548650741577,
+      "learning_rate": 7.97124276727134e-05,
+      "loss": 1.7706,
+      "step": 5495
+    },
+    {
+      "epoch": 0.5782219884271436,
+      "grad_norm": 1.2290290594100952,
+      "learning_rate": 7.967906013839987e-05,
+      "loss": 1.8209,
+      "step": 5496
+    },
+    {
+      "epoch": 0.5783271962125197,
+      "grad_norm": 1.5976051092147827,
+      "learning_rate": 7.964569496389013e-05,
+      "loss": 2.1965,
+      "step": 5497
+    },
+    {
+      "epoch": 0.5784324039978959,
+      "grad_norm": 1.4536248445510864,
+      "learning_rate": 7.961233215305884e-05,
+      "loss": 1.8925,
+      "step": 5498
+    },
+    {
+      "epoch": 0.578537611783272,
+      "grad_norm": 0.8368136286735535,
+      "learning_rate": 7.957897170978031e-05,
+      "loss": 1.5314,
+      "step": 5499
+    },
+    {
+      "epoch": 0.578642819568648,
+      "grad_norm": 1.7957181930541992,
+      "learning_rate": 7.95456136379285e-05,
+      "loss": 2.1585,
+      "step": 5500
+    },
+    {
+      "epoch": 0.5787480273540242,
+      "grad_norm": 1.1656877994537354,
+      "learning_rate": 7.951225794137724e-05,
+      "loss": 1.6489,
+      "step": 5501
+    },
+    {
+      "epoch": 0.5788532351394003,
+      "grad_norm": 1.2200168371200562,
+      "learning_rate": 7.947890462400002e-05,
+      "loss": 1.6195,
+      "step": 5502
+    },
+    {
+      "epoch": 0.5789584429247764,
+      "grad_norm": 1.2942944765090942,
+      "learning_rate": 7.944555368967001e-05,
+      "loss": 1.8944,
+      "step": 5503
+    },
+    {
+      "epoch": 0.5790636507101525,
+      "grad_norm": 1.5925309658050537,
+      "learning_rate": 7.941220514226016e-05,
+      "loss": 1.5092,
+      "step": 5504
+    },
+    {
+      "epoch": 0.5791688584955287,
+      "grad_norm": 1.2316579818725586,
+      "learning_rate": 7.937885898564315e-05,
+      "loss": 1.7135,
+      "step": 5505
+    },
+    {
+      "epoch": 0.5792740662809048,
+      "grad_norm": 1.9800273180007935,
+      "learning_rate": 7.934551522369134e-05,
+      "loss": 1.6139,
+      "step": 5506
+    },
+    {
+      "epoch": 0.5793792740662809,
+      "grad_norm": 1.2461237907409668,
+      "learning_rate": 7.931217386027686e-05,
+      "loss": 1.6989,
+      "step": 5507
+    },
+    {
+      "epoch": 0.579484481851657,
+      "grad_norm": 1.102742075920105,
+      "learning_rate": 7.927883489927147e-05,
+      "loss": 1.9475,
+      "step": 5508
+    },
+    {
+      "epoch": 0.5795896896370332,
+      "grad_norm": 1.5722696781158447,
+      "learning_rate": 7.924549834454674e-05,
+      "loss": 1.6444,
+      "step": 5509
+    },
+    {
+      "epoch": 0.5796948974224092,
+      "grad_norm": 1.331183671951294,
+      "learning_rate": 7.921216419997398e-05,
+      "loss": 1.3104,
+      "step": 5510
+    },
+    {
+      "epoch": 0.5798001052077854,
+      "grad_norm": 1.825430989265442,
+      "learning_rate": 7.917883246942412e-05,
+      "loss": 2.1806,
+      "step": 5511
+    },
+    {
+      "epoch": 0.5799053129931615,
+      "grad_norm": 1.2514722347259521,
+      "learning_rate": 7.914550315676787e-05,
+      "loss": 1.7759,
+      "step": 5512
+    },
+    {
+      "epoch": 0.5800105207785377,
+      "grad_norm": 1.1857120990753174,
+      "learning_rate": 7.91121762658757e-05,
+      "loss": 1.59,
+      "step": 5513
+    },
+    {
+      "epoch": 0.5801157285639137,
+      "grad_norm": 1.2514647245407104,
+      "learning_rate": 7.907885180061767e-05,
+      "loss": 2.1173,
+      "step": 5514
+    },
+    {
+      "epoch": 0.5802209363492898,
+      "grad_norm": 2.3011891841888428,
+      "learning_rate": 7.904552976486372e-05,
+      "loss": 1.8243,
+      "step": 5515
+    },
+    {
+      "epoch": 0.580326144134666,
+      "grad_norm": 1.686474084854126,
+      "learning_rate": 7.90122101624834e-05,
+      "loss": 1.6543,
+      "step": 5516
+    },
+    {
+      "epoch": 0.5804313519200421,
+      "grad_norm": 1.496381163597107,
+      "learning_rate": 7.897889299734599e-05,
+      "loss": 1.6579,
+      "step": 5517
+    },
+    {
+      "epoch": 0.5805365597054182,
+      "grad_norm": 2.0382790565490723,
+      "learning_rate": 7.894557827332052e-05,
+      "loss": 1.0179,
+      "step": 5518
+    },
+    {
+      "epoch": 0.5806417674907943,
+      "grad_norm": 1.186845064163208,
+      "learning_rate": 7.891226599427572e-05,
+      "loss": 1.7989,
+      "step": 5519
+    },
+    {
+      "epoch": 0.5807469752761705,
+      "grad_norm": 1.6111037731170654,
+      "learning_rate": 7.887895616408001e-05,
+      "loss": 1.6084,
+      "step": 5520
+    },
+    {
+      "epoch": 0.5808521830615465,
+      "grad_norm": 1.4457639455795288,
+      "learning_rate": 7.884564878660159e-05,
+      "loss": 1.784,
+      "step": 5521
+    },
+    {
+      "epoch": 0.5809573908469227,
+      "grad_norm": 1.4394114017486572,
+      "learning_rate": 7.88123438657083e-05,
+      "loss": 1.555,
+      "step": 5522
+    },
+    {
+      "epoch": 0.5810625986322988,
+      "grad_norm": 2.1034014225006104,
+      "learning_rate": 7.877904140526778e-05,
+      "loss": 1.7858,
+      "step": 5523
+    },
+    {
+      "epoch": 0.581167806417675,
+      "grad_norm": 1.5533802509307861,
+      "learning_rate": 7.874574140914727e-05,
+      "loss": 1.9843,
+      "step": 5524
+    },
+    {
+      "epoch": 0.581273014203051,
+      "grad_norm": 1.3675698041915894,
+      "learning_rate": 7.871244388121381e-05,
+      "loss": 2.3865,
+      "step": 5525
+    },
+    {
+      "epoch": 0.5813782219884271,
+      "grad_norm": 1.3307299613952637,
+      "learning_rate": 7.867914882533419e-05,
+      "loss": 1.7025,
+      "step": 5526
+    },
+    {
+      "epoch": 0.5814834297738033,
+      "grad_norm": 1.7351833581924438,
+      "learning_rate": 7.864585624537478e-05,
+      "loss": 1.679,
+      "step": 5527
+    },
+    {
+      "epoch": 0.5815886375591793,
+      "grad_norm": 1.8544718027114868,
+      "learning_rate": 7.86125661452018e-05,
+      "loss": 2.0512,
+      "step": 5528
+    },
+    {
+      "epoch": 0.5816938453445555,
+      "grad_norm": 1.8268216848373413,
+      "learning_rate": 7.857927852868107e-05,
+      "loss": 1.7498,
+      "step": 5529
+    },
+    {
+      "epoch": 0.5817990531299316,
+      "grad_norm": 1.4137629270553589,
+      "learning_rate": 7.854599339967817e-05,
+      "loss": 1.8924,
+      "step": 5530
+    },
+    {
+      "epoch": 0.5819042609153078,
+      "grad_norm": 1.431031584739685,
+      "learning_rate": 7.851271076205843e-05,
+      "loss": 1.6172,
+      "step": 5531
+    },
+    {
+      "epoch": 0.5820094687006838,
+      "grad_norm": 1.0963670015335083,
+      "learning_rate": 7.84794306196868e-05,
+      "loss": 1.6189,
+      "step": 5532
+    },
+    {
+      "epoch": 0.58211467648606,
+      "grad_norm": 1.4139877557754517,
+      "learning_rate": 7.844615297642805e-05,
+      "loss": 1.8776,
+      "step": 5533
+    },
+    {
+      "epoch": 0.5822198842714361,
+      "grad_norm": 1.4871132373809814,
+      "learning_rate": 7.84128778361466e-05,
+      "loss": 1.6445,
+      "step": 5534
+    },
+    {
+      "epoch": 0.5823250920568122,
+      "grad_norm": 1.2888953685760498,
+      "learning_rate": 7.837960520270652e-05,
+      "loss": 1.4948,
+      "step": 5535
+    },
+    {
+      "epoch": 0.5824302998421883,
+      "grad_norm": 1.376495599746704,
+      "learning_rate": 7.83463350799717e-05,
+      "loss": 1.8704,
+      "step": 5536
+    },
+    {
+      "epoch": 0.5825355076275645,
+      "grad_norm": 1.1731458902359009,
+      "learning_rate": 7.831306747180571e-05,
+      "loss": 1.5219,
+      "step": 5537
+    },
+    {
+      "epoch": 0.5826407154129406,
+      "grad_norm": 1.5646111965179443,
+      "learning_rate": 7.827980238207177e-05,
+      "loss": 1.3992,
+      "step": 5538
+    },
+    {
+      "epoch": 0.5827459231983166,
+      "grad_norm": 1.2269337177276611,
+      "learning_rate": 7.824653981463284e-05,
+      "loss": 2.2809,
+      "step": 5539
+    },
+    {
+      "epoch": 0.5828511309836928,
+      "grad_norm": 1.4412834644317627,
+      "learning_rate": 7.821327977335164e-05,
+      "loss": 1.2724,
+      "step": 5540
+    },
+    {
+      "epoch": 0.5829563387690689,
+      "grad_norm": 2.3292593955993652,
+      "learning_rate": 7.818002226209049e-05,
+      "loss": 1.4849,
+      "step": 5541
+    },
+    {
+      "epoch": 0.583061546554445,
+      "grad_norm": 1.5618584156036377,
+      "learning_rate": 7.814676728471151e-05,
+      "loss": 2.0205,
+      "step": 5542
+    },
+    {
+      "epoch": 0.5831667543398211,
+      "grad_norm": 1.4830820560455322,
+      "learning_rate": 7.811351484507647e-05,
+      "loss": 1.7858,
+      "step": 5543
+    },
+    {
+      "epoch": 0.5832719621251973,
+      "grad_norm": 1.1995006799697876,
+      "learning_rate": 7.808026494704692e-05,
+      "loss": 2.016,
+      "step": 5544
+    },
+    {
+      "epoch": 0.5833771699105734,
+      "grad_norm": 1.5571500062942505,
+      "learning_rate": 7.804701759448398e-05,
+      "loss": 1.7179,
+      "step": 5545
+    },
+    {
+      "epoch": 0.5834823776959495,
+      "grad_norm": 1.301533579826355,
+      "learning_rate": 7.801377279124862e-05,
+      "loss": 2.0771,
+      "step": 5546
+    },
+    {
+      "epoch": 0.5835875854813256,
+      "grad_norm": 1.841912031173706,
+      "learning_rate": 7.798053054120143e-05,
+      "loss": 1.9201,
+      "step": 5547
+    },
+    {
+      "epoch": 0.5836927932667018,
+      "grad_norm": 2.079071044921875,
+      "learning_rate": 7.794729084820272e-05,
+      "loss": 2.0332,
+      "step": 5548
+    },
+    {
+      "epoch": 0.5837980010520779,
+      "grad_norm": 1.3038408756256104,
+      "learning_rate": 7.791405371611249e-05,
+      "loss": 1.6136,
+      "step": 5549
+    },
+    {
+      "epoch": 0.583903208837454,
+      "grad_norm": 1.194056510925293,
+      "learning_rate": 7.788081914879051e-05,
+      "loss": 1.6015,
+      "step": 5550
+    },
+    {
+      "epoch": 0.5840084166228301,
+      "grad_norm": 1.3977482318878174,
+      "learning_rate": 7.784758715009616e-05,
+      "loss": 1.8922,
+      "step": 5551
+    },
+    {
+      "epoch": 0.5841136244082062,
+      "grad_norm": 1.124822735786438,
+      "learning_rate": 7.781435772388854e-05,
+      "loss": 1.4993,
+      "step": 5552
+    },
+    {
+      "epoch": 0.5842188321935823,
+      "grad_norm": 1.9948351383209229,
+      "learning_rate": 7.778113087402649e-05,
+      "loss": 1.7896,
+      "step": 5553
+    },
+    {
+      "epoch": 0.5843240399789584,
+      "grad_norm": 1.693495273590088,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 1.8484,
+      "step": 5554
+    },
+    {
+      "epoch": 0.5844292477643346,
+      "grad_norm": 1.079588532447815,
+      "learning_rate": 7.771468491877299e-05,
+      "loss": 2.1711,
+      "step": 5555
+    },
+    {
+      "epoch": 0.5845344555497107,
+      "grad_norm": 2.974010467529297,
+      "learning_rate": 7.768146582109765e-05,
+      "loss": 1.8077,
+      "step": 5556
+    },
+    {
+      "epoch": 0.5846396633350868,
+      "grad_norm": 2.0303964614868164,
+      "learning_rate": 7.764824931520018e-05,
+      "loss": 1.8446,
+      "step": 5557
+    },
+    {
+      "epoch": 0.5847448711204629,
+      "grad_norm": 1.5578415393829346,
+      "learning_rate": 7.761503540493795e-05,
+      "loss": 1.1523,
+      "step": 5558
+    },
+    {
+      "epoch": 0.5848500789058391,
+      "grad_norm": 1.480180025100708,
+      "learning_rate": 7.758182409416792e-05,
+      "loss": 1.5651,
+      "step": 5559
+    },
+    {
+      "epoch": 0.5849552866912151,
+      "grad_norm": 1.6015815734863281,
+      "learning_rate": 7.754861538674683e-05,
+      "loss": 1.8731,
+      "step": 5560
+    },
+    {
+      "epoch": 0.5850604944765913,
+      "grad_norm": 2.9025611877441406,
+      "learning_rate": 7.751540928653113e-05,
+      "loss": 0.9335,
+      "step": 5561
+    },
+    {
+      "epoch": 0.5851657022619674,
+      "grad_norm": 1.393959403038025,
+      "learning_rate": 7.748220579737689e-05,
+      "loss": 2.0774,
+      "step": 5562
+    },
+    {
+      "epoch": 0.5852709100473436,
+      "grad_norm": 1.700785756111145,
+      "learning_rate": 7.744900492313995e-05,
+      "loss": 1.7171,
+      "step": 5563
+    },
+    {
+      "epoch": 0.5853761178327196,
+      "grad_norm": 1.6016006469726562,
+      "learning_rate": 7.741580666767583e-05,
+      "loss": 1.7215,
+      "step": 5564
+    },
+    {
+      "epoch": 0.5854813256180957,
+      "grad_norm": 1.1057853698730469,
+      "learning_rate": 7.738261103483968e-05,
+      "loss": 1.1364,
+      "step": 5565
+    },
+    {
+      "epoch": 0.5855865334034719,
+      "grad_norm": 1.9537837505340576,
+      "learning_rate": 7.734941802848643e-05,
+      "loss": 1.5892,
+      "step": 5566
+    },
+    {
+      "epoch": 0.5856917411888479,
+      "grad_norm": 1.664334774017334,
+      "learning_rate": 7.731622765247069e-05,
+      "loss": 1.7183,
+      "step": 5567
+    },
+    {
+      "epoch": 0.5857969489742241,
+      "grad_norm": 1.9867991209030151,
+      "learning_rate": 7.728303991064675e-05,
+      "loss": 1.5369,
+      "step": 5568
+    },
+    {
+      "epoch": 0.5859021567596002,
+      "grad_norm": 1.713571310043335,
+      "learning_rate": 7.724985480686856e-05,
+      "loss": 2.0291,
+      "step": 5569
+    },
+    {
+      "epoch": 0.5860073645449764,
+      "grad_norm": 1.1955255270004272,
+      "learning_rate": 7.721667234498982e-05,
+      "loss": 2.378,
+      "step": 5570
+    },
+    {
+      "epoch": 0.5861125723303524,
+      "grad_norm": 1.5413669347763062,
+      "learning_rate": 7.718349252886395e-05,
+      "loss": 2.0389,
+      "step": 5571
+    },
+    {
+      "epoch": 0.5862177801157286,
+      "grad_norm": 1.7773537635803223,
+      "learning_rate": 7.715031536234392e-05,
+      "loss": 1.7247,
+      "step": 5572
+    },
+    {
+      "epoch": 0.5863229879011047,
+      "grad_norm": 1.7139322757720947,
+      "learning_rate": 7.711714084928251e-05,
+      "loss": 1.7779,
+      "step": 5573
+    },
+    {
+      "epoch": 0.5864281956864807,
+      "grad_norm": 1.3501460552215576,
+      "learning_rate": 7.708396899353219e-05,
+      "loss": 1.6001,
+      "step": 5574
+    },
+    {
+      "epoch": 0.5865334034718569,
+      "grad_norm": 3.4829583168029785,
+      "learning_rate": 7.705079979894509e-05,
+      "loss": 1.5646,
+      "step": 5575
+    },
+    {
+      "epoch": 0.586638611257233,
+      "grad_norm": 1.7532110214233398,
+      "learning_rate": 7.701763326937304e-05,
+      "loss": 1.5209,
+      "step": 5576
+    },
+    {
+      "epoch": 0.5867438190426092,
+      "grad_norm": 1.371377944946289,
+      "learning_rate": 7.698446940866754e-05,
+      "loss": 1.5853,
+      "step": 5577
+    },
+    {
+      "epoch": 0.5868490268279852,
+      "grad_norm": 1.3144980669021606,
+      "learning_rate": 7.695130822067984e-05,
+      "loss": 1.5741,
+      "step": 5578
+    },
+    {
+      "epoch": 0.5869542346133614,
+      "grad_norm": 1.3463817834854126,
+      "learning_rate": 7.691814970926083e-05,
+      "loss": 1.7503,
+      "step": 5579
+    },
+    {
+      "epoch": 0.5870594423987375,
+      "grad_norm": 1.7925662994384766,
+      "learning_rate": 7.688499387826107e-05,
+      "loss": 1.3115,
+      "step": 5580
+    },
+    {
+      "epoch": 0.5871646501841137,
+      "grad_norm": 4.266043186187744,
+      "learning_rate": 7.685184073153085e-05,
+      "loss": 1.97,
+      "step": 5581
+    },
+    {
+      "epoch": 0.5872698579694897,
+      "grad_norm": 1.344207525253296,
+      "learning_rate": 7.681869027292016e-05,
+      "loss": 1.6598,
+      "step": 5582
+    },
+    {
+      "epoch": 0.5873750657548659,
+      "grad_norm": 1.2654930353164673,
+      "learning_rate": 7.67855425062786e-05,
+      "loss": 1.6116,
+      "step": 5583
+    },
+    {
+      "epoch": 0.587480273540242,
+      "grad_norm": 1.6905845403671265,
+      "learning_rate": 7.675239743545557e-05,
+      "loss": 2.0407,
+      "step": 5584
+    },
+    {
+      "epoch": 0.587585481325618,
+      "grad_norm": 1.8982789516448975,
+      "learning_rate": 7.67192550643001e-05,
+      "loss": 1.9079,
+      "step": 5585
+    },
+    {
+      "epoch": 0.5876906891109942,
+      "grad_norm": 1.407412052154541,
+      "learning_rate": 7.668611539666085e-05,
+      "loss": 1.8989,
+      "step": 5586
+    },
+    {
+      "epoch": 0.5877958968963704,
+      "grad_norm": 1.6381629705429077,
+      "learning_rate": 7.665297843638623e-05,
+      "loss": 1.3814,
+      "step": 5587
+    },
+    {
+      "epoch": 0.5879011046817465,
+      "grad_norm": 2.0129973888397217,
+      "learning_rate": 7.661984418732438e-05,
+      "loss": 2.0981,
+      "step": 5588
+    },
+    {
+      "epoch": 0.5880063124671225,
+      "grad_norm": 1.516152024269104,
+      "learning_rate": 7.6586712653323e-05,
+      "loss": 1.6191,
+      "step": 5589
+    },
+    {
+      "epoch": 0.5881115202524987,
+      "grad_norm": 1.2309309244155884,
+      "learning_rate": 7.655358383822959e-05,
+      "loss": 1.4041,
+      "step": 5590
+    },
+    {
+      "epoch": 0.5882167280378748,
+      "grad_norm": 1.7488607168197632,
+      "learning_rate": 7.652045774589129e-05,
+      "loss": 2.1543,
+      "step": 5591
+    },
+    {
+      "epoch": 0.5883219358232509,
+      "grad_norm": 1.7684071063995361,
+      "learning_rate": 7.648733438015493e-05,
+      "loss": 2.056,
+      "step": 5592
+    },
+    {
+      "epoch": 0.588427143608627,
+      "grad_norm": 1.56985342502594,
+      "learning_rate": 7.645421374486702e-05,
+      "loss": 1.8957,
+      "step": 5593
+    },
+    {
+      "epoch": 0.5885323513940032,
+      "grad_norm": 2.020207166671753,
+      "learning_rate": 7.64210958438737e-05,
+      "loss": 1.7548,
+      "step": 5594
+    },
+    {
+      "epoch": 0.5886375591793793,
+      "grad_norm": 1.3547688722610474,
+      "learning_rate": 7.638798068102086e-05,
+      "loss": 1.8783,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5887427669647554,
+      "grad_norm": 1.2740882635116577,
+      "learning_rate": 7.635486826015412e-05,
+      "loss": 2.1481,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5888479747501315,
+      "grad_norm": 1.9236841201782227,
+      "learning_rate": 7.632175858511863e-05,
+      "loss": 1.3462,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5889531825355077,
+      "grad_norm": 1.3053388595581055,
+      "learning_rate": 7.628865165975934e-05,
+      "loss": 1.6169,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5890583903208837,
+      "grad_norm": 1.8421201705932617,
+      "learning_rate": 7.625554748792085e-05,
+      "loss": 1.6345,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5891635981062598,
+      "grad_norm": 1.4617395401000977,
+      "learning_rate": 7.622244607344748e-05,
+      "loss": 1.9279,
+      "step": 5600
+    },
+    {
+      "epoch": 0.589268805891636,
+      "grad_norm": 1.3376179933547974,
+      "learning_rate": 7.618934742018312e-05,
+      "loss": 1.5823,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5893740136770121,
+      "grad_norm": 1.5780961513519287,
+      "learning_rate": 7.615625153197143e-05,
+      "loss": 2.1328,
+      "step": 5602
+    },
+    {
+      "epoch": 0.5894792214623882,
+      "grad_norm": 1.1965007781982422,
+      "learning_rate": 7.612315841265577e-05,
+      "loss": 1.136,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5895844292477643,
+      "grad_norm": 1.290488600730896,
+      "learning_rate": 7.609006806607907e-05,
+      "loss": 1.5277,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5896896370331405,
+      "grad_norm": 1.6050409078598022,
+      "learning_rate": 7.605698049608403e-05,
+      "loss": 1.8864,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5897948448185165,
+      "grad_norm": 1.6614772081375122,
+      "learning_rate": 7.602389570651303e-05,
+      "loss": 1.9447,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5899000526038927,
+      "grad_norm": 1.3600127696990967,
+      "learning_rate": 7.599081370120804e-05,
+      "loss": 1.6589,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5900052603892688,
+      "grad_norm": 1.4860687255859375,
+      "learning_rate": 7.595773448401081e-05,
+      "loss": 1.5742,
+      "step": 5608
+    },
+    {
+      "epoch": 0.590110468174645,
+      "grad_norm": 1.154513955116272,
+      "learning_rate": 7.59246580587627e-05,
+      "loss": 2.1251,
+      "step": 5609
+    },
+    {
+      "epoch": 0.590215675960021,
+      "grad_norm": 2.0505974292755127,
+      "learning_rate": 7.589158442930478e-05,
+      "loss": 1.799,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5903208837453972,
+      "grad_norm": 2.0056207180023193,
+      "learning_rate": 7.585851359947776e-05,
+      "loss": 1.6373,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5904260915307733,
+      "grad_norm": 1.5547747611999512,
+      "learning_rate": 7.582544557312205e-05,
+      "loss": 1.4256,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5905312993161494,
+      "grad_norm": 1.8376978635787964,
+      "learning_rate": 7.579238035407776e-05,
+      "loss": 1.3878,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5906365071015255,
+      "grad_norm": 1.0613032579421997,
+      "learning_rate": 7.575931794618466e-05,
+      "loss": 1.9392,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5907417148869016,
+      "grad_norm": 1.1583870649337769,
+      "learning_rate": 7.572625835328211e-05,
+      "loss": 1.7329,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5908469226722778,
+      "grad_norm": 1.783522367477417,
+      "learning_rate": 7.569320157920923e-05,
+      "loss": 1.6231,
+      "step": 5616
+    },
+    {
+      "epoch": 0.5909521304576538,
+      "grad_norm": 1.3971893787384033,
+      "learning_rate": 7.566014762780483e-05,
+      "loss": 1.6672,
+      "step": 5617
+    },
+    {
+      "epoch": 0.59105733824303,
+      "grad_norm": 1.681029200553894,
+      "learning_rate": 7.562709650290732e-05,
+      "loss": 1.5265,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5911625460284061,
+      "grad_norm": 1.8332031965255737,
+      "learning_rate": 7.559404820835484e-05,
+      "loss": 1.695,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5912677538137823,
+      "grad_norm": 1.792855143547058,
+      "learning_rate": 7.556100274798519e-05,
+      "loss": 1.7872,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5913729615991583,
+      "grad_norm": 1.1729624271392822,
+      "learning_rate": 7.55279601256358e-05,
+      "loss": 2.0353,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5914781693845345,
+      "grad_norm": 1.0647164583206177,
+      "learning_rate": 7.549492034514381e-05,
+      "loss": 1.7371,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5915833771699106,
+      "grad_norm": 1.6151602268218994,
+      "learning_rate": 7.546188341034603e-05,
+      "loss": 1.0778,
+      "step": 5623
+    },
+    {
+      "epoch": 0.5916885849552866,
+      "grad_norm": 1.2735862731933594,
+      "learning_rate": 7.542884932507896e-05,
+      "loss": 1.7771,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5917937927406628,
+      "grad_norm": 1.3248671293258667,
+      "learning_rate": 7.539581809317866e-05,
+      "loss": 1.3799,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5918990005260389,
+      "grad_norm": 1.618085265159607,
+      "learning_rate": 7.536278971848101e-05,
+      "loss": 1.9688,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5920042083114151,
+      "grad_norm": 1.132401704788208,
+      "learning_rate": 7.532976420482146e-05,
+      "loss": 1.8808,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5921094160967911,
+      "grad_norm": 1.2111188173294067,
+      "learning_rate": 7.529674155603516e-05,
+      "loss": 1.8875,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5922146238821673,
+      "grad_norm": 1.4924447536468506,
+      "learning_rate": 7.52637217759569e-05,
+      "loss": 1.5021,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5923198316675434,
+      "grad_norm": 1.672135353088379,
+      "learning_rate": 7.52307048684212e-05,
+      "loss": 1.4673,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5924250394529195,
+      "grad_norm": 1.491942048072815,
+      "learning_rate": 7.519769083726216e-05,
+      "loss": 1.5303,
+      "step": 5631
+    },
+    {
+      "epoch": 0.5925302472382956,
+      "grad_norm": 1.7676005363464355,
+      "learning_rate": 7.51646796863136e-05,
+      "loss": 1.7333,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5926354550236718,
+      "grad_norm": 0.917346179485321,
+      "learning_rate": 7.513167141940904e-05,
+      "loss": 1.8271,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5927406628090479,
+      "grad_norm": 1.5192989110946655,
+      "learning_rate": 7.509866604038157e-05,
+      "loss": 1.8536,
+      "step": 5634
+    },
+    {
+      "epoch": 0.592845870594424,
+      "grad_norm": 0.7993506789207458,
+      "learning_rate": 7.506566355306402e-05,
+      "loss": 1.6131,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5929510783798001,
+      "grad_norm": 1.2340017557144165,
+      "learning_rate": 7.503266396128887e-05,
+      "loss": 1.6488,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5930562861651762,
+      "grad_norm": 1.8110322952270508,
+      "learning_rate": 7.499966726888823e-05,
+      "loss": 2.1135,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5931614939505523,
+      "grad_norm": 1.2862292528152466,
+      "learning_rate": 7.49666734796939e-05,
+      "loss": 2.2786,
+      "step": 5638
+    },
+    {
+      "epoch": 0.5932667017359284,
+      "grad_norm": 2.9395089149475098,
+      "learning_rate": 7.493368259753734e-05,
+      "loss": 1.8967,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5933719095213046,
+      "grad_norm": 2.132434844970703,
+      "learning_rate": 7.490069462624967e-05,
+      "loss": 1.3319,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5934771173066807,
+      "grad_norm": 1.9779353141784668,
+      "learning_rate": 7.486770956966171e-05,
+      "loss": 2.0619,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5935823250920568,
+      "grad_norm": 1.540423035621643,
+      "learning_rate": 7.483472743160387e-05,
+      "loss": 1.6819,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5936875328774329,
+      "grad_norm": 1.2363083362579346,
+      "learning_rate": 7.480174821590624e-05,
+      "loss": 1.7712,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5937927406628091,
+      "grad_norm": 1.7649072408676147,
+      "learning_rate": 7.476877192639866e-05,
+      "loss": 1.4488,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5938979484481852,
+      "grad_norm": 1.634068489074707,
+      "learning_rate": 7.473579856691047e-05,
+      "loss": 2.044,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5940031562335613,
+      "grad_norm": 1.0379782915115356,
+      "learning_rate": 7.470282814127081e-05,
+      "loss": 1.8553,
+      "step": 5646
+    },
+    {
+      "epoch": 0.5941083640189374,
+      "grad_norm": 1.697240948677063,
+      "learning_rate": 7.466986065330841e-05,
+      "loss": 1.4889,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5942135718043136,
+      "grad_norm": 1.2661585807800293,
+      "learning_rate": 7.463689610685171e-05,
+      "loss": 2.0126,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5943187795896896,
+      "grad_norm": 1.558271884918213,
+      "learning_rate": 7.460393450572872e-05,
+      "loss": 1.5544,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5944239873750657,
+      "grad_norm": 1.657724142074585,
+      "learning_rate": 7.457097585376719e-05,
+      "loss": 2.0922,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5945291951604419,
+      "grad_norm": 1.5575658082962036,
+      "learning_rate": 7.453802015479452e-05,
+      "loss": 2.0446,
+      "step": 5651
+    },
+    {
+      "epoch": 0.594634402945818,
+      "grad_norm": 1.7816355228424072,
+      "learning_rate": 7.45050674126377e-05,
+      "loss": 1.5467,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5947396107311941,
+      "grad_norm": 1.14240562915802,
+      "learning_rate": 7.447211763112346e-05,
+      "loss": 1.4401,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5948448185165702,
+      "grad_norm": 1.691174030303955,
+      "learning_rate": 7.443917081407816e-05,
+      "loss": 1.5377,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5949500263019464,
+      "grad_norm": 2.1762633323669434,
+      "learning_rate": 7.440622696532775e-05,
+      "loss": 1.6951,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5950552340873224,
+      "grad_norm": 1.880930781364441,
+      "learning_rate": 7.437328608869793e-05,
+      "loss": 1.3951,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5951604418726986,
+      "grad_norm": 1.4270281791687012,
+      "learning_rate": 7.434034818801405e-05,
+      "loss": 2.372,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5952656496580747,
+      "grad_norm": 1.9598197937011719,
+      "learning_rate": 7.4307413267101e-05,
+      "loss": 1.3984,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5953708574434509,
+      "grad_norm": 1.6525239944458008,
+      "learning_rate": 7.427448132978346e-05,
+      "loss": 1.699,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5954760652288269,
+      "grad_norm": 1.7269477844238281,
+      "learning_rate": 7.424155237988567e-05,
+      "loss": 1.6827,
+      "step": 5660
+    },
+    {
+      "epoch": 0.595581273014203,
+      "grad_norm": 1.3311235904693604,
+      "learning_rate": 7.420862642123158e-05,
+      "loss": 1.6645,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5956864807995792,
+      "grad_norm": 1.446498990058899,
+      "learning_rate": 7.417570345764481e-05,
+      "loss": 1.3863,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5957916885849552,
+      "grad_norm": 1.93930184841156,
+      "learning_rate": 7.414278349294852e-05,
+      "loss": 1.7474,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5958968963703314,
+      "grad_norm": 1.7791824340820312,
+      "learning_rate": 7.410986653096565e-05,
+      "loss": 1.564,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5960021041557075,
+      "grad_norm": 1.487729549407959,
+      "learning_rate": 7.407695257551875e-05,
+      "loss": 1.54,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5961073119410837,
+      "grad_norm": 1.072798490524292,
+      "learning_rate": 7.404404163042995e-05,
+      "loss": 1.4969,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5962125197264597,
+      "grad_norm": 1.5234678983688354,
+      "learning_rate": 7.401113369952113e-05,
+      "loss": 2.2265,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5963177275118359,
+      "grad_norm": 1.3278653621673584,
+      "learning_rate": 7.397822878661377e-05,
+      "loss": 2.0208,
+      "step": 5668
+    },
+    {
+      "epoch": 0.596422935297212,
+      "grad_norm": 1.424271583557129,
+      "learning_rate": 7.394532689552905e-05,
+      "loss": 2.034,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5965281430825881,
+      "grad_norm": 1.7164738178253174,
+      "learning_rate": 7.391242803008768e-05,
+      "loss": 2.2689,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5966333508679642,
+      "grad_norm": 3.884221315383911,
+      "learning_rate": 7.387953219411015e-05,
+      "loss": 1.8426,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5967385586533404,
+      "grad_norm": 1.339316964149475,
+      "learning_rate": 7.384663939141656e-05,
+      "loss": 1.9429,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5968437664387165,
+      "grad_norm": 1.4739105701446533,
+      "learning_rate": 7.381374962582659e-05,
+      "loss": 1.3312,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5969489742240925,
+      "grad_norm": 1.0755393505096436,
+      "learning_rate": 7.378086290115964e-05,
+      "loss": 1.7607,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5970541820094687,
+      "grad_norm": 1.6911028623580933,
+      "learning_rate": 7.374797922123478e-05,
+      "loss": 1.4523,
+      "step": 5675
+    },
+    {
+      "epoch": 0.5971593897948448,
+      "grad_norm": 1.7561362981796265,
+      "learning_rate": 7.371509858987061e-05,
+      "loss": 2.0099,
+      "step": 5676
+    },
+    {
+      "epoch": 0.597264597580221,
+      "grad_norm": 1.560784935951233,
+      "learning_rate": 7.368222101088549e-05,
+      "loss": 1.5806,
+      "step": 5677
+    },
+    {
+      "epoch": 0.597369805365597,
+      "grad_norm": 1.1424559354782104,
+      "learning_rate": 7.364934648809741e-05,
+      "loss": 2.1778,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5974750131509732,
+      "grad_norm": 1.393723726272583,
+      "learning_rate": 7.361647502532395e-05,
+      "loss": 2.0167,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5975802209363493,
+      "grad_norm": 0.8226059079170227,
+      "learning_rate": 7.358360662638236e-05,
+      "loss": 1.9121,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5976854287217254,
+      "grad_norm": 1.6506311893463135,
+      "learning_rate": 7.355074129508953e-05,
+      "loss": 2.2151,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5977906365071015,
+      "grad_norm": 1.2521584033966064,
+      "learning_rate": 7.351787903526201e-05,
+      "loss": 2.056,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5978958442924777,
+      "grad_norm": 1.4110209941864014,
+      "learning_rate": 7.348501985071603e-05,
+      "loss": 1.7132,
+      "step": 5683
+    },
+    {
+      "epoch": 0.5980010520778538,
+      "grad_norm": 1.7280937433242798,
+      "learning_rate": 7.345216374526736e-05,
+      "loss": 1.8772,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5981062598632298,
+      "grad_norm": 1.448601484298706,
+      "learning_rate": 7.341931072273148e-05,
+      "loss": 1.9043,
+      "step": 5685
+    },
+    {
+      "epoch": 0.598211467648606,
+      "grad_norm": 1.5108412504196167,
+      "learning_rate": 7.338646078692356e-05,
+      "loss": 1.0912,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5983166754339821,
+      "grad_norm": 1.2896003723144531,
+      "learning_rate": 7.335361394165825e-05,
+      "loss": 1.8589,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5984218832193582,
+      "grad_norm": 1.537742257118225,
+      "learning_rate": 7.332077019075005e-05,
+      "loss": 1.4406,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5985270910047343,
+      "grad_norm": 1.5651648044586182,
+      "learning_rate": 7.328792953801296e-05,
+      "loss": 1.7907,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5986322987901105,
+      "grad_norm": 1.8168197870254517,
+      "learning_rate": 7.325509198726064e-05,
+      "loss": 2.1356,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5987375065754866,
+      "grad_norm": 1.6306514739990234,
+      "learning_rate": 7.322225754230641e-05,
+      "loss": 2.0181,
+      "step": 5691
+    },
+    {
+      "epoch": 0.5988427143608627,
+      "grad_norm": 1.733379602432251,
+      "learning_rate": 7.318942620696323e-05,
+      "loss": 1.3749,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5989479221462388,
+      "grad_norm": 2.751979351043701,
+      "learning_rate": 7.315659798504373e-05,
+      "loss": 1.2602,
+      "step": 5693
+    },
+    {
+      "epoch": 0.599053129931615,
+      "grad_norm": 1.8571940660476685,
+      "learning_rate": 7.312377288036009e-05,
+      "loss": 1.5986,
+      "step": 5694
+    },
+    {
+      "epoch": 0.599158337716991,
+      "grad_norm": 2.700819253921509,
+      "learning_rate": 7.30909508967242e-05,
+      "loss": 2.2613,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5992635455023672,
+      "grad_norm": 1.9650945663452148,
+      "learning_rate": 7.30581320379476e-05,
+      "loss": 1.4637,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5993687532877433,
+      "grad_norm": 2.2203495502471924,
+      "learning_rate": 7.302531630784137e-05,
+      "loss": 1.7075,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5994739610731195,
+      "grad_norm": 1.5563139915466309,
+      "learning_rate": 7.299250371021635e-05,
+      "loss": 1.9756,
+      "step": 5698
+    },
+    {
+      "epoch": 0.5995791688584955,
+      "grad_norm": 1.332543969154358,
+      "learning_rate": 7.295969424888295e-05,
+      "loss": 1.4028,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5996843766438716,
+      "grad_norm": 1.5485048294067383,
+      "learning_rate": 7.292688792765126e-05,
+      "loss": 1.5521,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5997895844292478,
+      "grad_norm": 1.584691047668457,
+      "learning_rate": 7.289408475033086e-05,
+      "loss": 1.5764,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5998947922146238,
+      "grad_norm": 1.7013684511184692,
+      "learning_rate": 7.286128472073114e-05,
+      "loss": 2.1275,
+      "step": 5702
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.437490701675415,
+      "learning_rate": 7.282848784266107e-05,
+      "loss": 2.214,
+      "step": 5703
+    },
+    {
+      "epoch": 0.6001052077853761,
+      "grad_norm": 1.080392837524414,
+      "learning_rate": 7.279569411992926e-05,
+      "loss": 1.4222,
+      "step": 5704
+    },
+    {
+      "epoch": 0.6002104155707523,
+      "grad_norm": 1.6214826107025146,
+      "learning_rate": 7.276290355634387e-05,
+      "loss": 2.1349,
+      "step": 5705
+    },
+    {
+      "epoch": 0.6003156233561283,
+      "grad_norm": 1.7013198137283325,
+      "learning_rate": 7.273011615571282e-05,
+      "loss": 1.5759,
+      "step": 5706
+    },
+    {
+      "epoch": 0.6004208311415045,
+      "grad_norm": 1.5219919681549072,
+      "learning_rate": 7.26973319218436e-05,
+      "loss": 2.1324,
+      "step": 5707
+    },
+    {
+      "epoch": 0.6005260389268806,
+      "grad_norm": 1.9878424406051636,
+      "learning_rate": 7.266455085854329e-05,
+      "loss": 2.4534,
+      "step": 5708
+    },
+    {
+      "epoch": 0.6006312467122568,
+      "grad_norm": 1.9398351907730103,
+      "learning_rate": 7.263177296961867e-05,
+      "loss": 1.2152,
+      "step": 5709
+    },
+    {
+      "epoch": 0.6007364544976328,
+      "grad_norm": 1.528769850730896,
+      "learning_rate": 7.259899825887617e-05,
+      "loss": 2.2585,
+      "step": 5710
+    },
+    {
+      "epoch": 0.600841662283009,
+      "grad_norm": 1.6075571775436401,
+      "learning_rate": 7.256622673012175e-05,
+      "loss": 2.076,
+      "step": 5711
+    },
+    {
+      "epoch": 0.6009468700683851,
+      "grad_norm": 1.6357545852661133,
+      "learning_rate": 7.253345838716108e-05,
+      "loss": 1.9542,
+      "step": 5712
+    },
+    {
+      "epoch": 0.6010520778537611,
+      "grad_norm": 1.8700578212738037,
+      "learning_rate": 7.250069323379945e-05,
+      "loss": 1.5798,
+      "step": 5713
+    },
+    {
+      "epoch": 0.6011572856391373,
+      "grad_norm": 2.4654650688171387,
+      "learning_rate": 7.246793127384174e-05,
+      "loss": 1.8467,
+      "step": 5714
+    },
+    {
+      "epoch": 0.6012624934245134,
+      "grad_norm": 2.0539462566375732,
+      "learning_rate": 7.243517251109254e-05,
+      "loss": 1.6161,
+      "step": 5715
+    },
+    {
+      "epoch": 0.6013677012098896,
+      "grad_norm": 2.258213520050049,
+      "learning_rate": 7.240241694935597e-05,
+      "loss": 1.9685,
+      "step": 5716
+    },
+    {
+      "epoch": 0.6014729089952656,
+      "grad_norm": 1.7232213020324707,
+      "learning_rate": 7.236966459243586e-05,
+      "loss": 2.4748,
+      "step": 5717
+    },
+    {
+      "epoch": 0.6015781167806418,
+      "grad_norm": 1.12027907371521,
+      "learning_rate": 7.233691544413558e-05,
+      "loss": 1.8557,
+      "step": 5718
+    },
+    {
+      "epoch": 0.6016833245660179,
+      "grad_norm": 1.5948566198349,
+      "learning_rate": 7.230416950825825e-05,
+      "loss": 1.2433,
+      "step": 5719
+    },
+    {
+      "epoch": 0.601788532351394,
+      "grad_norm": 1.1951110363006592,
+      "learning_rate": 7.227142678860652e-05,
+      "loss": 1.9291,
+      "step": 5720
+    },
+    {
+      "epoch": 0.6018937401367701,
+      "grad_norm": 1.3234333992004395,
+      "learning_rate": 7.22386872889827e-05,
+      "loss": 2.0288,
+      "step": 5721
+    },
+    {
+      "epoch": 0.6019989479221463,
+      "grad_norm": 1.8448857069015503,
+      "learning_rate": 7.22059510131887e-05,
+      "loss": 2.0855,
+      "step": 5722
+    },
+    {
+      "epoch": 0.6021041557075224,
+      "grad_norm": 1.6373887062072754,
+      "learning_rate": 7.217321796502605e-05,
+      "loss": 1.0645,
+      "step": 5723
+    },
+    {
+      "epoch": 0.6022093634928984,
+      "grad_norm": 2.3060243129730225,
+      "learning_rate": 7.214048814829598e-05,
+      "loss": 1.3767,
+      "step": 5724
+    },
+    {
+      "epoch": 0.6023145712782746,
+      "grad_norm": 1.55854332447052,
+      "learning_rate": 7.210776156679931e-05,
+      "loss": 1.5182,
+      "step": 5725
+    },
+    {
+      "epoch": 0.6024197790636507,
+      "grad_norm": 1.64441978931427,
+      "learning_rate": 7.20750382243364e-05,
+      "loss": 2.3997,
+      "step": 5726
+    },
+    {
+      "epoch": 0.6025249868490268,
+      "grad_norm": 0.9315603375434875,
+      "learning_rate": 7.204231812470736e-05,
+      "loss": 2.0199,
+      "step": 5727
+    },
+    {
+      "epoch": 0.6026301946344029,
+      "grad_norm": 1.946433663368225,
+      "learning_rate": 7.200960127171188e-05,
+      "loss": 2.2558,
+      "step": 5728
+    },
+    {
+      "epoch": 0.6027354024197791,
+      "grad_norm": 1.7605088949203491,
+      "learning_rate": 7.197688766914921e-05,
+      "loss": 1.8627,
+      "step": 5729
+    },
+    {
+      "epoch": 0.6028406102051552,
+      "grad_norm": 1.548949956893921,
+      "learning_rate": 7.19441773208183e-05,
+      "loss": 1.4795,
+      "step": 5730
+    },
+    {
+      "epoch": 0.6029458179905313,
+      "grad_norm": 2.702446222305298,
+      "learning_rate": 7.19114702305177e-05,
+      "loss": 1.6094,
+      "step": 5731
+    },
+    {
+      "epoch": 0.6030510257759074,
+      "grad_norm": 2.4522550106048584,
+      "learning_rate": 7.187876640204556e-05,
+      "loss": 1.7925,
+      "step": 5732
+    },
+    {
+      "epoch": 0.6031562335612836,
+      "grad_norm": 1.2998313903808594,
+      "learning_rate": 7.184606583919966e-05,
+      "loss": 1.4222,
+      "step": 5733
+    },
+    {
+      "epoch": 0.6032614413466596,
+      "grad_norm": 2.228344678878784,
+      "learning_rate": 7.181336854577747e-05,
+      "loss": 1.7094,
+      "step": 5734
+    },
+    {
+      "epoch": 0.6033666491320357,
+      "grad_norm": 1.8566731214523315,
+      "learning_rate": 7.178067452557595e-05,
+      "loss": 1.943,
+      "step": 5735
+    },
+    {
+      "epoch": 0.6034718569174119,
+      "grad_norm": 1.346657395362854,
+      "learning_rate": 7.174798378239176e-05,
+      "loss": 1.7142,
+      "step": 5736
+    },
+    {
+      "epoch": 0.603577064702788,
+      "grad_norm": 1.3058844804763794,
+      "learning_rate": 7.171529632002121e-05,
+      "loss": 1.9819,
+      "step": 5737
+    },
+    {
+      "epoch": 0.6036822724881641,
+      "grad_norm": 1.9217864274978638,
+      "learning_rate": 7.168261214226014e-05,
+      "loss": 1.6599,
+      "step": 5738
+    },
+    {
+      "epoch": 0.6037874802735402,
+      "grad_norm": 1.5166659355163574,
+      "learning_rate": 7.164993125290407e-05,
+      "loss": 1.9654,
+      "step": 5739
+    },
+    {
+      "epoch": 0.6038926880589164,
+      "grad_norm": 1.7919169664382935,
+      "learning_rate": 7.161725365574811e-05,
+      "loss": 1.7845,
+      "step": 5740
+    },
+    {
+      "epoch": 0.6039978958442925,
+      "grad_norm": 1.7721195220947266,
+      "learning_rate": 7.158457935458706e-05,
+      "loss": 1.7366,
+      "step": 5741
+    },
+    {
+      "epoch": 0.6041031036296686,
+      "grad_norm": 3.324281930923462,
+      "learning_rate": 7.155190835321523e-05,
+      "loss": 1.8447,
+      "step": 5742
+    },
+    {
+      "epoch": 0.6042083114150447,
+      "grad_norm": 1.3101855516433716,
+      "learning_rate": 7.151924065542665e-05,
+      "loss": 1.9632,
+      "step": 5743
+    },
+    {
+      "epoch": 0.6043135192004209,
+      "grad_norm": 1.5195070505142212,
+      "learning_rate": 7.14865762650148e-05,
+      "loss": 1.4127,
+      "step": 5744
+    },
+    {
+      "epoch": 0.6044187269857969,
+      "grad_norm": 1.4836463928222656,
+      "learning_rate": 7.1453915185773e-05,
+      "loss": 2.0167,
+      "step": 5745
+    },
+    {
+      "epoch": 0.604523934771173,
+      "grad_norm": 2.11446475982666,
+      "learning_rate": 7.1421257421494e-05,
+      "loss": 1.4259,
+      "step": 5746
+    },
+    {
+      "epoch": 0.6046291425565492,
+      "grad_norm": 1.1224156618118286,
+      "learning_rate": 7.138860297597026e-05,
+      "loss": 1.6661,
+      "step": 5747
+    },
+    {
+      "epoch": 0.6047343503419254,
+      "grad_norm": 0.8229581713676453,
+      "learning_rate": 7.135595185299386e-05,
+      "loss": 1.5664,
+      "step": 5748
+    },
+    {
+      "epoch": 0.6048395581273014,
+      "grad_norm": 1.6141940355300903,
+      "learning_rate": 7.132330405635645e-05,
+      "loss": 1.6177,
+      "step": 5749
+    },
+    {
+      "epoch": 0.6049447659126775,
+      "grad_norm": 1.415408968925476,
+      "learning_rate": 7.12906595898493e-05,
+      "loss": 1.1493,
+      "step": 5750
+    },
+    {
+      "epoch": 0.6050499736980537,
+      "grad_norm": 2.1405954360961914,
+      "learning_rate": 7.12580184572633e-05,
+      "loss": 2.1279,
+      "step": 5751
+    },
+    {
+      "epoch": 0.6051551814834297,
+      "grad_norm": 1.3183344602584839,
+      "learning_rate": 7.122538066238902e-05,
+      "loss": 1.8046,
+      "step": 5752
+    },
+    {
+      "epoch": 0.6052603892688059,
+      "grad_norm": 2.0879404544830322,
+      "learning_rate": 7.119274620901649e-05,
+      "loss": 1.4804,
+      "step": 5753
+    },
+    {
+      "epoch": 0.605365597054182,
+      "grad_norm": 1.4626708030700684,
+      "learning_rate": 7.116011510093547e-05,
+      "loss": 1.9303,
+      "step": 5754
+    },
+    {
+      "epoch": 0.6054708048395582,
+      "grad_norm": 1.3531354665756226,
+      "learning_rate": 7.112748734193537e-05,
+      "loss": 1.9332,
+      "step": 5755
+    },
+    {
+      "epoch": 0.6055760126249342,
+      "grad_norm": 2.5419464111328125,
+      "learning_rate": 7.109486293580505e-05,
+      "loss": 2.2981,
+      "step": 5756
+    },
+    {
+      "epoch": 0.6056812204103104,
+      "grad_norm": 1.8105266094207764,
+      "learning_rate": 7.106224188633311e-05,
+      "loss": 2.1267,
+      "step": 5757
+    },
+    {
+      "epoch": 0.6057864281956865,
+      "grad_norm": 1.5312819480895996,
+      "learning_rate": 7.102962419730776e-05,
+      "loss": 1.9916,
+      "step": 5758
+    },
+    {
+      "epoch": 0.6058916359810625,
+      "grad_norm": 2.3092429637908936,
+      "learning_rate": 7.099700987251674e-05,
+      "loss": 2.2213,
+      "step": 5759
+    },
+    {
+      "epoch": 0.6059968437664387,
+      "grad_norm": 1.4174106121063232,
+      "learning_rate": 7.096439891574745e-05,
+      "loss": 1.9643,
+      "step": 5760
+    },
+    {
+      "epoch": 0.6061020515518148,
+      "grad_norm": 2.2073097229003906,
+      "learning_rate": 7.09317913307869e-05,
+      "loss": 2.1425,
+      "step": 5761
+    },
+    {
+      "epoch": 0.606207259337191,
+      "grad_norm": 1.472976803779602,
+      "learning_rate": 7.089918712142172e-05,
+      "loss": 1.3008,
+      "step": 5762
+    },
+    {
+      "epoch": 0.606312467122567,
+      "grad_norm": 27.837261199951172,
+      "learning_rate": 7.086658629143811e-05,
+      "loss": 1.9546,
+      "step": 5763
+    },
+    {
+      "epoch": 0.6064176749079432,
+      "grad_norm": 1.4600732326507568,
+      "learning_rate": 7.083398884462194e-05,
+      "loss": 1.7937,
+      "step": 5764
+    },
+    {
+      "epoch": 0.6065228826933193,
+      "grad_norm": 1.1610373258590698,
+      "learning_rate": 7.080139478475853e-05,
+      "loss": 2.1027,
+      "step": 5765
+    },
+    {
+      "epoch": 0.6066280904786954,
+      "grad_norm": 1.7224698066711426,
+      "learning_rate": 7.076880411563305e-05,
+      "loss": 1.8358,
+      "step": 5766
+    },
+    {
+      "epoch": 0.6067332982640715,
+      "grad_norm": 1.7691140174865723,
+      "learning_rate": 7.073621684103007e-05,
+      "loss": 1.7801,
+      "step": 5767
+    },
+    {
+      "epoch": 0.6068385060494477,
+      "grad_norm": 1.6140860319137573,
+      "learning_rate": 7.070363296473384e-05,
+      "loss": 1.1727,
+      "step": 5768
+    },
+    {
+      "epoch": 0.6069437138348238,
+      "grad_norm": 1.1361531019210815,
+      "learning_rate": 7.067105249052828e-05,
+      "loss": 1.7667,
+      "step": 5769
+    },
+    {
+      "epoch": 0.6070489216201999,
+      "grad_norm": 1.355162501335144,
+      "learning_rate": 7.063847542219679e-05,
+      "loss": 1.0353,
+      "step": 5770
+    },
+    {
+      "epoch": 0.607154129405576,
+      "grad_norm": 0.9227995872497559,
+      "learning_rate": 7.060590176352248e-05,
+      "loss": 1.9074,
+      "step": 5771
+    },
+    {
+      "epoch": 0.6072593371909522,
+      "grad_norm": 1.033094882965088,
+      "learning_rate": 7.057333151828799e-05,
+      "loss": 2.163,
+      "step": 5772
+    },
+    {
+      "epoch": 0.6073645449763283,
+      "grad_norm": 1.4324654340744019,
+      "learning_rate": 7.054076469027565e-05,
+      "loss": 1.7172,
+      "step": 5773
+    },
+    {
+      "epoch": 0.6074697527617043,
+      "grad_norm": 2.334664821624756,
+      "learning_rate": 7.050820128326724e-05,
+      "loss": 2.092,
+      "step": 5774
+    },
+    {
+      "epoch": 0.6075749605470805,
+      "grad_norm": 2.054091453552246,
+      "learning_rate": 7.047564130104434e-05,
+      "loss": 1.0949,
+      "step": 5775
+    },
+    {
+      "epoch": 0.6076801683324566,
+      "grad_norm": 2.1884820461273193,
+      "learning_rate": 7.044308474738798e-05,
+      "loss": 1.7727,
+      "step": 5776
+    },
+    {
+      "epoch": 0.6077853761178327,
+      "grad_norm": 1.3433359861373901,
+      "learning_rate": 7.041053162607886e-05,
+      "loss": 1.8029,
+      "step": 5777
+    },
+    {
+      "epoch": 0.6078905839032088,
+      "grad_norm": 1.5086860656738281,
+      "learning_rate": 7.037798194089728e-05,
+      "loss": 1.7065,
+      "step": 5778
+    },
+    {
+      "epoch": 0.607995791688585,
+      "grad_norm": 1.6031910181045532,
+      "learning_rate": 7.034543569562313e-05,
+      "loss": 1.7895,
+      "step": 5779
+    },
+    {
+      "epoch": 0.6081009994739611,
+      "grad_norm": 1.8059049844741821,
+      "learning_rate": 7.031289289403584e-05,
+      "loss": 1.7179,
+      "step": 5780
+    },
+    {
+      "epoch": 0.6082062072593372,
+      "grad_norm": 1.4660775661468506,
+      "learning_rate": 7.028035353991456e-05,
+      "loss": 2.0142,
+      "step": 5781
+    },
+    {
+      "epoch": 0.6083114150447133,
+      "grad_norm": 1.5748193264007568,
+      "learning_rate": 7.024781763703797e-05,
+      "loss": 1.9199,
+      "step": 5782
+    },
+    {
+      "epoch": 0.6084166228300895,
+      "grad_norm": 1.2944796085357666,
+      "learning_rate": 7.021528518918433e-05,
+      "loss": 1.1247,
+      "step": 5783
+    },
+    {
+      "epoch": 0.6085218306154655,
+      "grad_norm": 1.12498140335083,
+      "learning_rate": 7.018275620013154e-05,
+      "loss": 1.3863,
+      "step": 5784
+    },
+    {
+      "epoch": 0.6086270384008416,
+      "grad_norm": 1.4992774724960327,
+      "learning_rate": 7.01502306736571e-05,
+      "loss": 1.9758,
+      "step": 5785
+    },
+    {
+      "epoch": 0.6087322461862178,
+      "grad_norm": 1.8976898193359375,
+      "learning_rate": 7.01177086135381e-05,
+      "loss": 1.6387,
+      "step": 5786
+    },
+    {
+      "epoch": 0.608837453971594,
+      "grad_norm": 1.2787110805511475,
+      "learning_rate": 7.008519002355118e-05,
+      "loss": 1.9901,
+      "step": 5787
+    },
+    {
+      "epoch": 0.60894266175697,
+      "grad_norm": 1.5045870542526245,
+      "learning_rate": 7.005267490747263e-05,
+      "loss": 1.5003,
+      "step": 5788
+    },
+    {
+      "epoch": 0.6090478695423461,
+      "grad_norm": 1.5272982120513916,
+      "learning_rate": 7.002016326907831e-05,
+      "loss": 2.1544,
+      "step": 5789
+    },
+    {
+      "epoch": 0.6091530773277223,
+      "grad_norm": 1.7053319215774536,
+      "learning_rate": 6.998765511214374e-05,
+      "loss": 1.8607,
+      "step": 5790
+    },
+    {
+      "epoch": 0.6092582851130983,
+      "grad_norm": 1.7263697385787964,
+      "learning_rate": 6.995515044044393e-05,
+      "loss": 2.009,
+      "step": 5791
+    },
+    {
+      "epoch": 0.6093634928984745,
+      "grad_norm": 1.4831238985061646,
+      "learning_rate": 6.992264925775356e-05,
+      "loss": 1.7212,
+      "step": 5792
+    },
+    {
+      "epoch": 0.6094687006838506,
+      "grad_norm": 1.557859182357788,
+      "learning_rate": 6.989015156784689e-05,
+      "loss": 1.6485,
+      "step": 5793
+    },
+    {
+      "epoch": 0.6095739084692268,
+      "grad_norm": 1.1065948009490967,
+      "learning_rate": 6.98576573744978e-05,
+      "loss": 1.7752,
+      "step": 5794
+    },
+    {
+      "epoch": 0.6096791162546028,
+      "grad_norm": 1.3376137018203735,
+      "learning_rate": 6.982516668147967e-05,
+      "loss": 1.8011,
+      "step": 5795
+    },
+    {
+      "epoch": 0.609784324039979,
+      "grad_norm": 1.223147988319397,
+      "learning_rate": 6.979267949256558e-05,
+      "loss": 1.2879,
+      "step": 5796
+    },
+    {
+      "epoch": 0.6098895318253551,
+      "grad_norm": 1.0986629724502563,
+      "learning_rate": 6.976019581152818e-05,
+      "loss": 2.0554,
+      "step": 5797
+    },
+    {
+      "epoch": 0.6099947396107311,
+      "grad_norm": 1.3327361345291138,
+      "learning_rate": 6.972771564213963e-05,
+      "loss": 1.5186,
+      "step": 5798
+    },
+    {
+      "epoch": 0.6100999473961073,
+      "grad_norm": 1.3183317184448242,
+      "learning_rate": 6.969523898817176e-05,
+      "loss": 1.9726,
+      "step": 5799
+    },
+    {
+      "epoch": 0.6102051551814834,
+      "grad_norm": 1.7785512208938599,
+      "learning_rate": 6.966276585339604e-05,
+      "loss": 2.1342,
+      "step": 5800
+    },
+    {
+      "epoch": 0.6103103629668596,
+      "grad_norm": 1.0888605117797852,
+      "learning_rate": 6.96302962415834e-05,
+      "loss": 1.4301,
+      "step": 5801
+    },
+    {
+      "epoch": 0.6104155707522356,
+      "grad_norm": 1.8897411823272705,
+      "learning_rate": 6.959783015650446e-05,
+      "loss": 2.3269,
+      "step": 5802
+    },
+    {
+      "epoch": 0.6105207785376118,
+      "grad_norm": 1.019864559173584,
+      "learning_rate": 6.956536760192938e-05,
+      "loss": 1.5989,
+      "step": 5803
+    },
+    {
+      "epoch": 0.6106259863229879,
+      "grad_norm": 1.5215952396392822,
+      "learning_rate": 6.953290858162794e-05,
+      "loss": 2.14,
+      "step": 5804
+    },
+    {
+      "epoch": 0.6107311941083641,
+      "grad_norm": 1.6711841821670532,
+      "learning_rate": 6.95004530993695e-05,
+      "loss": 2.1474,
+      "step": 5805
+    },
+    {
+      "epoch": 0.6108364018937401,
+      "grad_norm": 1.50629460811615,
+      "learning_rate": 6.946800115892305e-05,
+      "loss": 2.319,
+      "step": 5806
+    },
+    {
+      "epoch": 0.6109416096791163,
+      "grad_norm": 1.2559032440185547,
+      "learning_rate": 6.943555276405705e-05,
+      "loss": 1.4975,
+      "step": 5807
+    },
+    {
+      "epoch": 0.6110468174644924,
+      "grad_norm": 0.9776951670646667,
+      "learning_rate": 6.940310791853968e-05,
+      "loss": 2.0983,
+      "step": 5808
+    },
+    {
+      "epoch": 0.6111520252498684,
+      "grad_norm": 1.521141767501831,
+      "learning_rate": 6.937066662613863e-05,
+      "loss": 1.5946,
+      "step": 5809
+    },
+    {
+      "epoch": 0.6112572330352446,
+      "grad_norm": 1.377365231513977,
+      "learning_rate": 6.933822889062118e-05,
+      "loss": 1.4309,
+      "step": 5810
+    },
+    {
+      "epoch": 0.6113624408206207,
+      "grad_norm": 1.1052137613296509,
+      "learning_rate": 6.930579471575427e-05,
+      "loss": 2.1511,
+      "step": 5811
+    },
+    {
+      "epoch": 0.6114676486059969,
+      "grad_norm": 1.3557707071304321,
+      "learning_rate": 6.927336410530432e-05,
+      "loss": 1.6768,
+      "step": 5812
+    },
+    {
+      "epoch": 0.6115728563913729,
+      "grad_norm": 2.0096118450164795,
+      "learning_rate": 6.924093706303743e-05,
+      "loss": 1.9034,
+      "step": 5813
+    },
+    {
+      "epoch": 0.6116780641767491,
+      "grad_norm": 2.108721971511841,
+      "learning_rate": 6.920851359271922e-05,
+      "loss": 1.6447,
+      "step": 5814
+    },
+    {
+      "epoch": 0.6117832719621252,
+      "grad_norm": 1.1239184141159058,
+      "learning_rate": 6.917609369811496e-05,
+      "loss": 1.9551,
+      "step": 5815
+    },
+    {
+      "epoch": 0.6118884797475013,
+      "grad_norm": 1.2308555841445923,
+      "learning_rate": 6.914367738298941e-05,
+      "loss": 2.1679,
+      "step": 5816
+    },
+    {
+      "epoch": 0.6119936875328774,
+      "grad_norm": 1.2711541652679443,
+      "learning_rate": 6.9111264651107e-05,
+      "loss": 1.9665,
+      "step": 5817
+    },
+    {
+      "epoch": 0.6120988953182536,
+      "grad_norm": 1.186928629875183,
+      "learning_rate": 6.907885550623172e-05,
+      "loss": 1.644,
+      "step": 5818
+    },
+    {
+      "epoch": 0.6122041031036297,
+      "grad_norm": 1.484810709953308,
+      "learning_rate": 6.904644995212713e-05,
+      "loss": 1.3018,
+      "step": 5819
+    },
+    {
+      "epoch": 0.6123093108890058,
+      "grad_norm": 1.1081911325454712,
+      "learning_rate": 6.901404799255638e-05,
+      "loss": 1.6068,
+      "step": 5820
+    },
+    {
+      "epoch": 0.6124145186743819,
+      "grad_norm": 1.3067530393600464,
+      "learning_rate": 6.898164963128221e-05,
+      "loss": 1.6391,
+      "step": 5821
+    },
+    {
+      "epoch": 0.612519726459758,
+      "grad_norm": 0.9942488670349121,
+      "learning_rate": 6.894925487206691e-05,
+      "loss": 1.9811,
+      "step": 5822
+    },
+    {
+      "epoch": 0.6126249342451341,
+      "grad_norm": 1.2306791543960571,
+      "learning_rate": 6.891686371867239e-05,
+      "loss": 1.9109,
+      "step": 5823
+    },
+    {
+      "epoch": 0.6127301420305102,
+      "grad_norm": 1.1397932767868042,
+      "learning_rate": 6.888447617486016e-05,
+      "loss": 1.5098,
+      "step": 5824
+    },
+    {
+      "epoch": 0.6128353498158864,
+      "grad_norm": 1.0217232704162598,
+      "learning_rate": 6.885209224439123e-05,
+      "loss": 1.4489,
+      "step": 5825
+    },
+    {
+      "epoch": 0.6129405576012625,
+      "grad_norm": 2.21636700630188,
+      "learning_rate": 6.881971193102625e-05,
+      "loss": 1.5184,
+      "step": 5826
+    },
+    {
+      "epoch": 0.6130457653866386,
+      "grad_norm": 1.0205903053283691,
+      "learning_rate": 6.878733523852549e-05,
+      "loss": 1.8485,
+      "step": 5827
+    },
+    {
+      "epoch": 0.6131509731720147,
+      "grad_norm": 1.5595358610153198,
+      "learning_rate": 6.875496217064867e-05,
+      "loss": 1.5117,
+      "step": 5828
+    },
+    {
+      "epoch": 0.6132561809573909,
+      "grad_norm": 1.1299363374710083,
+      "learning_rate": 6.872259273115525e-05,
+      "loss": 1.9574,
+      "step": 5829
+    },
+    {
+      "epoch": 0.6133613887427669,
+      "grad_norm": 1.109512448310852,
+      "learning_rate": 6.869022692380411e-05,
+      "loss": 2.0785,
+      "step": 5830
+    },
+    {
+      "epoch": 0.6134665965281431,
+      "grad_norm": 1.951183795928955,
+      "learning_rate": 6.865786475235381e-05,
+      "loss": 2.1442,
+      "step": 5831
+    },
+    {
+      "epoch": 0.6135718043135192,
+      "grad_norm": 1.3576364517211914,
+      "learning_rate": 6.862550622056249e-05,
+      "loss": 1.6519,
+      "step": 5832
+    },
+    {
+      "epoch": 0.6136770120988954,
+      "grad_norm": 1.2096575498580933,
+      "learning_rate": 6.85931513321878e-05,
+      "loss": 1.4433,
+      "step": 5833
+    },
+    {
+      "epoch": 0.6137822198842714,
+      "grad_norm": 2.158343553543091,
+      "learning_rate": 6.856080009098701e-05,
+      "loss": 1.4268,
+      "step": 5834
+    },
+    {
+      "epoch": 0.6138874276696475,
+      "grad_norm": 1.6646260023117065,
+      "learning_rate": 6.852845250071702e-05,
+      "loss": 1.9002,
+      "step": 5835
+    },
+    {
+      "epoch": 0.6139926354550237,
+      "grad_norm": 1.3010480403900146,
+      "learning_rate": 6.849610856513418e-05,
+      "loss": 1.8029,
+      "step": 5836
+    },
+    {
+      "epoch": 0.6140978432403998,
+      "grad_norm": 1.888411283493042,
+      "learning_rate": 6.846376828799451e-05,
+      "loss": 1.7934,
+      "step": 5837
+    },
+    {
+      "epoch": 0.6142030510257759,
+      "grad_norm": 1.5725120306015015,
+      "learning_rate": 6.843143167305361e-05,
+      "loss": 1.2792,
+      "step": 5838
+    },
+    {
+      "epoch": 0.614308258811152,
+      "grad_norm": 1.0562975406646729,
+      "learning_rate": 6.839909872406657e-05,
+      "loss": 1.8781,
+      "step": 5839
+    },
+    {
+      "epoch": 0.6144134665965282,
+      "grad_norm": 1.287243127822876,
+      "learning_rate": 6.836676944478812e-05,
+      "loss": 1.6942,
+      "step": 5840
+    },
+    {
+      "epoch": 0.6145186743819042,
+      "grad_norm": 1.386404275894165,
+      "learning_rate": 6.83344438389726e-05,
+      "loss": 1.4059,
+      "step": 5841
+    },
+    {
+      "epoch": 0.6146238821672804,
+      "grad_norm": 1.312274694442749,
+      "learning_rate": 6.830212191037386e-05,
+      "loss": 1.6319,
+      "step": 5842
+    },
+    {
+      "epoch": 0.6147290899526565,
+      "grad_norm": 1.408747673034668,
+      "learning_rate": 6.826980366274529e-05,
+      "loss": 1.0995,
+      "step": 5843
+    },
+    {
+      "epoch": 0.6148342977380327,
+      "grad_norm": 1.132150650024414,
+      "learning_rate": 6.823748909983994e-05,
+      "loss": 1.7678,
+      "step": 5844
+    },
+    {
+      "epoch": 0.6149395055234087,
+      "grad_norm": 1.0861999988555908,
+      "learning_rate": 6.820517822541041e-05,
+      "loss": 1.2332,
+      "step": 5845
+    },
+    {
+      "epoch": 0.6150447133087849,
+      "grad_norm": 1.2827601432800293,
+      "learning_rate": 6.81728710432088e-05,
+      "loss": 2.0137,
+      "step": 5846
+    },
+    {
+      "epoch": 0.615149921094161,
+      "grad_norm": 1.5106278657913208,
+      "learning_rate": 6.81405675569869e-05,
+      "loss": 1.7959,
+      "step": 5847
+    },
+    {
+      "epoch": 0.615255128879537,
+      "grad_norm": 1.3239543437957764,
+      "learning_rate": 6.810826777049597e-05,
+      "loss": 1.6723,
+      "step": 5848
+    },
+    {
+      "epoch": 0.6153603366649132,
+      "grad_norm": 1.4004448652267456,
+      "learning_rate": 6.807597168748689e-05,
+      "loss": 1.9686,
+      "step": 5849
+    },
+    {
+      "epoch": 0.6154655444502893,
+      "grad_norm": 0.9085174798965454,
+      "learning_rate": 6.804367931171013e-05,
+      "loss": 1.5498,
+      "step": 5850
+    },
+    {
+      "epoch": 0.6155707522356655,
+      "grad_norm": 1.3527324199676514,
+      "learning_rate": 6.801139064691562e-05,
+      "loss": 1.4251,
+      "step": 5851
+    },
+    {
+      "epoch": 0.6156759600210415,
+      "grad_norm": 1.9584040641784668,
+      "learning_rate": 6.797910569685297e-05,
+      "loss": 1.6893,
+      "step": 5852
+    },
+    {
+      "epoch": 0.6157811678064177,
+      "grad_norm": 1.093604326248169,
+      "learning_rate": 6.794682446527137e-05,
+      "loss": 1.369,
+      "step": 5853
+    },
+    {
+      "epoch": 0.6158863755917938,
+      "grad_norm": 1.4752413034439087,
+      "learning_rate": 6.791454695591945e-05,
+      "loss": 1.6492,
+      "step": 5854
+    },
+    {
+      "epoch": 0.6159915833771699,
+      "grad_norm": 0.9944615960121155,
+      "learning_rate": 6.788227317254556e-05,
+      "loss": 1.9943,
+      "step": 5855
+    },
+    {
+      "epoch": 0.616096791162546,
+      "grad_norm": 1.6758373975753784,
+      "learning_rate": 6.785000311889754e-05,
+      "loss": 1.9175,
+      "step": 5856
+    },
+    {
+      "epoch": 0.6162019989479222,
+      "grad_norm": 2.0718863010406494,
+      "learning_rate": 6.781773679872276e-05,
+      "loss": 1.7659,
+      "step": 5857
+    },
+    {
+      "epoch": 0.6163072067332983,
+      "grad_norm": 1.6973742246627808,
+      "learning_rate": 6.778547421576825e-05,
+      "loss": 1.8969,
+      "step": 5858
+    },
+    {
+      "epoch": 0.6164124145186743,
+      "grad_norm": 1.7583503723144531,
+      "learning_rate": 6.775321537378054e-05,
+      "loss": 1.8226,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6165176223040505,
+      "grad_norm": 1.5600526332855225,
+      "learning_rate": 6.772096027650574e-05,
+      "loss": 1.6122,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6166228300894266,
+      "grad_norm": 1.8818323612213135,
+      "learning_rate": 6.768870892768952e-05,
+      "loss": 1.4151,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6167280378748027,
+      "grad_norm": 1.4384057521820068,
+      "learning_rate": 6.765646133107714e-05,
+      "loss": 2.2372,
+      "step": 5862
+    },
+    {
+      "epoch": 0.6168332456601788,
+      "grad_norm": 1.1038570404052734,
+      "learning_rate": 6.762421749041342e-05,
+      "loss": 2.0336,
+      "step": 5863
+    },
+    {
+      "epoch": 0.616938453445555,
+      "grad_norm": 1.2065256834030151,
+      "learning_rate": 6.759197740944267e-05,
+      "loss": 1.1006,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6170436612309311,
+      "grad_norm": 2.1591498851776123,
+      "learning_rate": 6.75597410919089e-05,
+      "loss": 1.3848,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6171488690163072,
+      "grad_norm": 1.131312608718872,
+      "learning_rate": 6.752750854155558e-05,
+      "loss": 1.5338,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6172540768016833,
+      "grad_norm": 2.397049903869629,
+      "learning_rate": 6.749527976212573e-05,
+      "loss": 1.3482,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6173592845870595,
+      "grad_norm": 1.6215959787368774,
+      "learning_rate": 6.746305475736202e-05,
+      "loss": 1.6116,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6174644923724356,
+      "grad_norm": 1.2333688735961914,
+      "learning_rate": 6.743083353100664e-05,
+      "loss": 1.7272,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6175697001578117,
+      "grad_norm": 2.3157310485839844,
+      "learning_rate": 6.739861608680129e-05,
+      "loss": 1.7432,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6176749079431878,
+      "grad_norm": 1.532241940498352,
+      "learning_rate": 6.736640242848735e-05,
+      "loss": 1.1707,
+      "step": 5871
+    },
+    {
+      "epoch": 0.617780115728564,
+      "grad_norm": 1.3878227472305298,
+      "learning_rate": 6.733419255980559e-05,
+      "loss": 2.2097,
+      "step": 5872
+    },
+    {
+      "epoch": 0.61788532351394,
+      "grad_norm": 1.6810948848724365,
+      "learning_rate": 6.730198648449648e-05,
+      "loss": 1.7966,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6179905312993161,
+      "grad_norm": 1.3429570198059082,
+      "learning_rate": 6.726978420630002e-05,
+      "loss": 1.7373,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6180957390846923,
+      "grad_norm": 1.960030436515808,
+      "learning_rate": 6.723758572895573e-05,
+      "loss": 2.0929,
+      "step": 5875
+    },
+    {
+      "epoch": 0.6182009468700684,
+      "grad_norm": 2.1462693214416504,
+      "learning_rate": 6.720539105620272e-05,
+      "loss": 1.7261,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6183061546554445,
+      "grad_norm": 1.1600507497787476,
+      "learning_rate": 6.717320019177969e-05,
+      "loss": 1.5239,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6184113624408206,
+      "grad_norm": 1.3913511037826538,
+      "learning_rate": 6.71410131394248e-05,
+      "loss": 1.8831,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6185165702261968,
+      "grad_norm": 1.2141426801681519,
+      "learning_rate": 6.710882990287585e-05,
+      "loss": 2.0194,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6186217780115728,
+      "grad_norm": 1.3260263204574585,
+      "learning_rate": 6.70766504858702e-05,
+      "loss": 1.4936,
+      "step": 5880
+    },
+    {
+      "epoch": 0.618726985796949,
+      "grad_norm": 1.1059815883636475,
+      "learning_rate": 6.704447489214468e-05,
+      "loss": 2.0024,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6188321935823251,
+      "grad_norm": 1.362945795059204,
+      "learning_rate": 6.701230312543578e-05,
+      "loss": 1.6022,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6189374013677013,
+      "grad_norm": 1.131805658340454,
+      "learning_rate": 6.698013518947952e-05,
+      "loss": 2.0098,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6190426091530773,
+      "grad_norm": 1.8835530281066895,
+      "learning_rate": 6.69479710880114e-05,
+      "loss": 1.5957,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6191478169384534,
+      "grad_norm": 1.4850059747695923,
+      "learning_rate": 6.691581082476656e-05,
+      "loss": 1.617,
+      "step": 5885
+    },
+    {
+      "epoch": 0.6192530247238296,
+      "grad_norm": 1.5727633237838745,
+      "learning_rate": 6.688365440347965e-05,
+      "loss": 1.9237,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6193582325092056,
+      "grad_norm": 1.0962116718292236,
+      "learning_rate": 6.685150182788495e-05,
+      "loss": 1.6268,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6194634402945818,
+      "grad_norm": 2.0313258171081543,
+      "learning_rate": 6.681935310171616e-05,
+      "loss": 1.8827,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6195686480799579,
+      "grad_norm": 1.543372631072998,
+      "learning_rate": 6.678720822870663e-05,
+      "loss": 1.8296,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6196738558653341,
+      "grad_norm": 1.0897983312606812,
+      "learning_rate": 6.675506721258926e-05,
+      "loss": 1.9225,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6197790636507101,
+      "grad_norm": 1.2541539669036865,
+      "learning_rate": 6.672293005709644e-05,
+      "loss": 1.6706,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6198842714360863,
+      "grad_norm": 1.780050277709961,
+      "learning_rate": 6.669079676596019e-05,
+      "loss": 1.6821,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6199894792214624,
+      "grad_norm": 1.5210779905319214,
+      "learning_rate": 6.665866734291205e-05,
+      "loss": 1.7069,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6200946870068385,
+      "grad_norm": 1.2912250757217407,
+      "learning_rate": 6.662654179168306e-05,
+      "loss": 1.7669,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6201998947922146,
+      "grad_norm": 1.736703634262085,
+      "learning_rate": 6.659442011600387e-05,
+      "loss": 1.457,
+      "step": 5895
+    },
+    {
+      "epoch": 0.6203051025775908,
+      "grad_norm": 1.1300891637802124,
+      "learning_rate": 6.656230231960466e-05,
+      "loss": 1.8526,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6204103103629669,
+      "grad_norm": 1.8092447519302368,
+      "learning_rate": 6.653018840621516e-05,
+      "loss": 2.3459,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6205155181483429,
+      "grad_norm": 1.1894848346710205,
+      "learning_rate": 6.649807837956472e-05,
+      "loss": 1.5572,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6206207259337191,
+      "grad_norm": 1.3011435270309448,
+      "learning_rate": 6.646597224338207e-05,
+      "loss": 1.6887,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6207259337190952,
+      "grad_norm": 2.1688480377197266,
+      "learning_rate": 6.643387000139565e-05,
+      "loss": 1.3811,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6208311415044714,
+      "grad_norm": 1.4247039556503296,
+      "learning_rate": 6.640177165733339e-05,
+      "loss": 1.7982,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6209363492898474,
+      "grad_norm": 0.9942802786827087,
+      "learning_rate": 6.636967721492274e-05,
+      "loss": 1.3522,
+      "step": 5902
+    },
+    {
+      "epoch": 0.6210415570752236,
+      "grad_norm": 1.3744851350784302,
+      "learning_rate": 6.633758667789074e-05,
+      "loss": 1.7667,
+      "step": 5903
+    },
+    {
+      "epoch": 0.6211467648605997,
+      "grad_norm": 1.2810285091400146,
+      "learning_rate": 6.630550004996396e-05,
+      "loss": 2.0339,
+      "step": 5904
+    },
+    {
+      "epoch": 0.6212519726459758,
+      "grad_norm": 1.1830532550811768,
+      "learning_rate": 6.627341733486847e-05,
+      "loss": 1.8695,
+      "step": 5905
+    },
+    {
+      "epoch": 0.6213571804313519,
+      "grad_norm": 1.8294991254806519,
+      "learning_rate": 6.624133853632998e-05,
+      "loss": 1.95,
+      "step": 5906
+    },
+    {
+      "epoch": 0.6214623882167281,
+      "grad_norm": 2.3274435997009277,
+      "learning_rate": 6.620926365807372e-05,
+      "loss": 1.9036,
+      "step": 5907
+    },
+    {
+      "epoch": 0.6215675960021042,
+      "grad_norm": 1.6169378757476807,
+      "learning_rate": 6.617719270382436e-05,
+      "loss": 2.0314,
+      "step": 5908
+    },
+    {
+      "epoch": 0.6216728037874802,
+      "grad_norm": 1.2674856185913086,
+      "learning_rate": 6.614512567730625e-05,
+      "loss": 1.1626,
+      "step": 5909
+    },
+    {
+      "epoch": 0.6217780115728564,
+      "grad_norm": 1.3875672817230225,
+      "learning_rate": 6.611306258224319e-05,
+      "loss": 1.5651,
+      "step": 5910
+    },
+    {
+      "epoch": 0.6218832193582325,
+      "grad_norm": 2.156421422958374,
+      "learning_rate": 6.608100342235861e-05,
+      "loss": 2.0087,
+      "step": 5911
+    },
+    {
+      "epoch": 0.6219884271436086,
+      "grad_norm": 1.7543668746948242,
+      "learning_rate": 6.604894820137541e-05,
+      "loss": 1.9096,
+      "step": 5912
+    },
+    {
+      "epoch": 0.6220936349289847,
+      "grad_norm": 1.749904990196228,
+      "learning_rate": 6.601689692301604e-05,
+      "loss": 1.5697,
+      "step": 5913
+    },
+    {
+      "epoch": 0.6221988427143609,
+      "grad_norm": 1.3653181791305542,
+      "learning_rate": 6.598484959100257e-05,
+      "loss": 1.8171,
+      "step": 5914
+    },
+    {
+      "epoch": 0.622304050499737,
+      "grad_norm": 1.5228540897369385,
+      "learning_rate": 6.59528062090565e-05,
+      "loss": 1.6228,
+      "step": 5915
+    },
+    {
+      "epoch": 0.6224092582851131,
+      "grad_norm": 1.0161175727844238,
+      "learning_rate": 6.592076678089889e-05,
+      "loss": 1.3673,
+      "step": 5916
+    },
+    {
+      "epoch": 0.6225144660704892,
+      "grad_norm": 1.3234752416610718,
+      "learning_rate": 6.588873131025042e-05,
+      "loss": 1.1581,
+      "step": 5917
+    },
+    {
+      "epoch": 0.6226196738558654,
+      "grad_norm": 1.3997480869293213,
+      "learning_rate": 6.585669980083128e-05,
+      "loss": 1.6416,
+      "step": 5918
+    },
+    {
+      "epoch": 0.6227248816412414,
+      "grad_norm": 2.785766124725342,
+      "learning_rate": 6.582467225636116e-05,
+      "loss": 1.6409,
+      "step": 5919
+    },
+    {
+      "epoch": 0.6228300894266176,
+      "grad_norm": 1.0819607973098755,
+      "learning_rate": 6.579264868055928e-05,
+      "loss": 1.4652,
+      "step": 5920
+    },
+    {
+      "epoch": 0.6229352972119937,
+      "grad_norm": 1.7267218828201294,
+      "learning_rate": 6.576062907714448e-05,
+      "loss": 2.1395,
+      "step": 5921
+    },
+    {
+      "epoch": 0.6230405049973698,
+      "grad_norm": 1.7600226402282715,
+      "learning_rate": 6.572861344983511e-05,
+      "loss": 2.0258,
+      "step": 5922
+    },
+    {
+      "epoch": 0.6231457127827459,
+      "grad_norm": 1.3963185548782349,
+      "learning_rate": 6.569660180234898e-05,
+      "loss": 1.7034,
+      "step": 5923
+    },
+    {
+      "epoch": 0.623250920568122,
+      "grad_norm": 1.8364591598510742,
+      "learning_rate": 6.566459413840351e-05,
+      "loss": 1.9706,
+      "step": 5924
+    },
+    {
+      "epoch": 0.6233561283534982,
+      "grad_norm": 1.4244076013565063,
+      "learning_rate": 6.56325904617157e-05,
+      "loss": 2.1374,
+      "step": 5925
+    },
+    {
+      "epoch": 0.6234613361388742,
+      "grad_norm": 1.3648873567581177,
+      "learning_rate": 6.560059077600195e-05,
+      "loss": 1.6418,
+      "step": 5926
+    },
+    {
+      "epoch": 0.6235665439242504,
+      "grad_norm": 1.695998191833496,
+      "learning_rate": 6.556859508497834e-05,
+      "loss": 1.712,
+      "step": 5927
+    },
+    {
+      "epoch": 0.6236717517096265,
+      "grad_norm": 2.2389862537384033,
+      "learning_rate": 6.553660339236041e-05,
+      "loss": 1.6126,
+      "step": 5928
+    },
+    {
+      "epoch": 0.6237769594950027,
+      "grad_norm": 1.46173095703125,
+      "learning_rate": 6.550461570186322e-05,
+      "loss": 2.1347,
+      "step": 5929
+    },
+    {
+      "epoch": 0.6238821672803787,
+      "grad_norm": 1.7502105236053467,
+      "learning_rate": 6.547263201720143e-05,
+      "loss": 1.4319,
+      "step": 5930
+    },
+    {
+      "epoch": 0.6239873750657549,
+      "grad_norm": 1.4961512088775635,
+      "learning_rate": 6.54406523420892e-05,
+      "loss": 2.2034,
+      "step": 5931
+    },
+    {
+      "epoch": 0.624092582851131,
+      "grad_norm": 1.5784069299697876,
+      "learning_rate": 6.54086766802402e-05,
+      "loss": 2.0528,
+      "step": 5932
+    },
+    {
+      "epoch": 0.6241977906365072,
+      "grad_norm": 1.485357642173767,
+      "learning_rate": 6.537670503536766e-05,
+      "loss": 2.0438,
+      "step": 5933
+    },
+    {
+      "epoch": 0.6243029984218832,
+      "grad_norm": 1.615233063697815,
+      "learning_rate": 6.534473741118434e-05,
+      "loss": 1.5522,
+      "step": 5934
+    },
+    {
+      "epoch": 0.6244082062072593,
+      "grad_norm": 1.7099108695983887,
+      "learning_rate": 6.53127738114026e-05,
+      "loss": 1.8988,
+      "step": 5935
+    },
+    {
+      "epoch": 0.6245134139926355,
+      "grad_norm": 1.4433890581130981,
+      "learning_rate": 6.528081423973422e-05,
+      "loss": 2.0022,
+      "step": 5936
+    },
+    {
+      "epoch": 0.6246186217780115,
+      "grad_norm": 1.8532782793045044,
+      "learning_rate": 6.52488586998905e-05,
+      "loss": 1.748,
+      "step": 5937
+    },
+    {
+      "epoch": 0.6247238295633877,
+      "grad_norm": 1.721971869468689,
+      "learning_rate": 6.52169071955824e-05,
+      "loss": 1.4468,
+      "step": 5938
+    },
+    {
+      "epoch": 0.6248290373487638,
+      "grad_norm": 1.340277910232544,
+      "learning_rate": 6.518495973052036e-05,
+      "loss": 1.4,
+      "step": 5939
+    },
+    {
+      "epoch": 0.62493424513414,
+      "grad_norm": 1.9042940139770508,
+      "learning_rate": 6.515301630841426e-05,
+      "loss": 1.6993,
+      "step": 5940
+    },
+    {
+      "epoch": 0.625039452919516,
+      "grad_norm": 2.6405863761901855,
+      "learning_rate": 6.512107693297365e-05,
+      "loss": 1.4162,
+      "step": 5941
+    },
+    {
+      "epoch": 0.6251446607048922,
+      "grad_norm": 1.6686121225357056,
+      "learning_rate": 6.508914160790752e-05,
+      "loss": 1.3212,
+      "step": 5942
+    },
+    {
+      "epoch": 0.6252498684902683,
+      "grad_norm": 1.4782112836837769,
+      "learning_rate": 6.505721033692443e-05,
+      "loss": 1.7841,
+      "step": 5943
+    },
+    {
+      "epoch": 0.6253550762756444,
+      "grad_norm": 1.8376623392105103,
+      "learning_rate": 6.502528312373241e-05,
+      "loss": 1.5778,
+      "step": 5944
+    },
+    {
+      "epoch": 0.6254602840610205,
+      "grad_norm": 1.424533724784851,
+      "learning_rate": 6.49933599720391e-05,
+      "loss": 1.4852,
+      "step": 5945
+    },
+    {
+      "epoch": 0.6255654918463966,
+      "grad_norm": 1.5104315280914307,
+      "learning_rate": 6.496144088555162e-05,
+      "loss": 1.6104,
+      "step": 5946
+    },
+    {
+      "epoch": 0.6256706996317728,
+      "grad_norm": 1.5233991146087646,
+      "learning_rate": 6.492952586797665e-05,
+      "loss": 1.2861,
+      "step": 5947
+    },
+    {
+      "epoch": 0.6257759074171488,
+      "grad_norm": 1.3382761478424072,
+      "learning_rate": 6.489761492302034e-05,
+      "loss": 2.0656,
+      "step": 5948
+    },
+    {
+      "epoch": 0.625881115202525,
+      "grad_norm": 1.0290378332138062,
+      "learning_rate": 6.486570805438843e-05,
+      "loss": 1.7404,
+      "step": 5949
+    },
+    {
+      "epoch": 0.6259863229879011,
+      "grad_norm": 1.8992608785629272,
+      "learning_rate": 6.483380526578615e-05,
+      "loss": 1.3986,
+      "step": 5950
+    },
+    {
+      "epoch": 0.6260915307732772,
+      "grad_norm": 1.5879778861999512,
+      "learning_rate": 6.480190656091825e-05,
+      "loss": 1.7387,
+      "step": 5951
+    },
+    {
+      "epoch": 0.6261967385586533,
+      "grad_norm": 1.1449265480041504,
+      "learning_rate": 6.477001194348906e-05,
+      "loss": 1.3209,
+      "step": 5952
+    },
+    {
+      "epoch": 0.6263019463440295,
+      "grad_norm": 1.9107697010040283,
+      "learning_rate": 6.473812141720234e-05,
+      "loss": 1.6479,
+      "step": 5953
+    },
+    {
+      "epoch": 0.6264071541294056,
+      "grad_norm": 2.4911386966705322,
+      "learning_rate": 6.47062349857615e-05,
+      "loss": 1.9478,
+      "step": 5954
+    },
+    {
+      "epoch": 0.6265123619147817,
+      "grad_norm": 1.8238067626953125,
+      "learning_rate": 6.467435265286935e-05,
+      "loss": 1.6787,
+      "step": 5955
+    },
+    {
+      "epoch": 0.6266175697001578,
+      "grad_norm": 1.4120367765426636,
+      "learning_rate": 6.46424744222283e-05,
+      "loss": 1.6102,
+      "step": 5956
+    },
+    {
+      "epoch": 0.626722777485534,
+      "grad_norm": 0.9109852910041809,
+      "learning_rate": 6.461060029754031e-05,
+      "loss": 1.982,
+      "step": 5957
+    },
+    {
+      "epoch": 0.62682798527091,
+      "grad_norm": 1.119532823562622,
+      "learning_rate": 6.457873028250674e-05,
+      "loss": 1.7647,
+      "step": 5958
+    },
+    {
+      "epoch": 0.6269331930562861,
+      "grad_norm": 1.667096734046936,
+      "learning_rate": 6.454686438082858e-05,
+      "loss": 1.2471,
+      "step": 5959
+    },
+    {
+      "epoch": 0.6270384008416623,
+      "grad_norm": 1.9557980298995972,
+      "learning_rate": 6.45150025962063e-05,
+      "loss": 1.553,
+      "step": 5960
+    },
+    {
+      "epoch": 0.6271436086270384,
+      "grad_norm": 1.4863619804382324,
+      "learning_rate": 6.448314493233995e-05,
+      "loss": 1.706,
+      "step": 5961
+    },
+    {
+      "epoch": 0.6272488164124145,
+      "grad_norm": 1.5268634557724,
+      "learning_rate": 6.445129139292899e-05,
+      "loss": 1.7509,
+      "step": 5962
+    },
+    {
+      "epoch": 0.6273540241977906,
+      "grad_norm": 1.5882983207702637,
+      "learning_rate": 6.441944198167253e-05,
+      "loss": 1.9265,
+      "step": 5963
+    },
+    {
+      "epoch": 0.6274592319831668,
+      "grad_norm": 1.2085886001586914,
+      "learning_rate": 6.43875967022691e-05,
+      "loss": 1.6107,
+      "step": 5964
+    },
+    {
+      "epoch": 0.6275644397685429,
+      "grad_norm": 2.4494829177856445,
+      "learning_rate": 6.435575555841679e-05,
+      "loss": 1.9172,
+      "step": 5965
+    },
+    {
+      "epoch": 0.627669647553919,
+      "grad_norm": 0.8796269297599792,
+      "learning_rate": 6.432391855381321e-05,
+      "loss": 0.9647,
+      "step": 5966
+    },
+    {
+      "epoch": 0.6277748553392951,
+      "grad_norm": 1.2112059593200684,
+      "learning_rate": 6.42920856921555e-05,
+      "loss": 1.734,
+      "step": 5967
+    },
+    {
+      "epoch": 0.6278800631246713,
+      "grad_norm": 1.265119194984436,
+      "learning_rate": 6.426025697714029e-05,
+      "loss": 1.9329,
+      "step": 5968
+    },
+    {
+      "epoch": 0.6279852709100473,
+      "grad_norm": 1.5609668493270874,
+      "learning_rate": 6.422843241246374e-05,
+      "loss": 1.5995,
+      "step": 5969
+    },
+    {
+      "epoch": 0.6280904786954234,
+      "grad_norm": 1.1384879350662231,
+      "learning_rate": 6.419661200182158e-05,
+      "loss": 2.1466,
+      "step": 5970
+    },
+    {
+      "epoch": 0.6281956864807996,
+      "grad_norm": 1.4794455766677856,
+      "learning_rate": 6.416479574890894e-05,
+      "loss": 2.2411,
+      "step": 5971
+    },
+    {
+      "epoch": 0.6283008942661757,
+      "grad_norm": 1.0125900506973267,
+      "learning_rate": 6.413298365742055e-05,
+      "loss": 1.6002,
+      "step": 5972
+    },
+    {
+      "epoch": 0.6284061020515518,
+      "grad_norm": 1.3207452297210693,
+      "learning_rate": 6.41011757310507e-05,
+      "loss": 1.6828,
+      "step": 5973
+    },
+    {
+      "epoch": 0.6285113098369279,
+      "grad_norm": 1.2414960861206055,
+      "learning_rate": 6.406937197349308e-05,
+      "loss": 1.6645,
+      "step": 5974
+    },
+    {
+      "epoch": 0.6286165176223041,
+      "grad_norm": 1.6514127254486084,
+      "learning_rate": 6.403757238844096e-05,
+      "loss": 1.9715,
+      "step": 5975
+    },
+    {
+      "epoch": 0.6287217254076801,
+      "grad_norm": 1.0575364828109741,
+      "learning_rate": 6.400577697958718e-05,
+      "loss": 2.2627,
+      "step": 5976
+    },
+    {
+      "epoch": 0.6288269331930563,
+      "grad_norm": 1.608067512512207,
+      "learning_rate": 6.397398575062396e-05,
+      "loss": 1.871,
+      "step": 5977
+    },
+    {
+      "epoch": 0.6289321409784324,
+      "grad_norm": 1.4367784261703491,
+      "learning_rate": 6.394219870524314e-05,
+      "loss": 1.8095,
+      "step": 5978
+    },
+    {
+      "epoch": 0.6290373487638086,
+      "grad_norm": 1.3168582916259766,
+      "learning_rate": 6.391041584713608e-05,
+      "loss": 1.6412,
+      "step": 5979
+    },
+    {
+      "epoch": 0.6291425565491846,
+      "grad_norm": 1.5024646520614624,
+      "learning_rate": 6.387863717999357e-05,
+      "loss": 2.0073,
+      "step": 5980
+    },
+    {
+      "epoch": 0.6292477643345608,
+      "grad_norm": 1.1878137588500977,
+      "learning_rate": 6.384686270750599e-05,
+      "loss": 1.0912,
+      "step": 5981
+    },
+    {
+      "epoch": 0.6293529721199369,
+      "grad_norm": 1.134263515472412,
+      "learning_rate": 6.381509243336318e-05,
+      "loss": 1.9806,
+      "step": 5982
+    },
+    {
+      "epoch": 0.6294581799053129,
+      "grad_norm": 1.093222737312317,
+      "learning_rate": 6.378332636125453e-05,
+      "loss": 1.8243,
+      "step": 5983
+    },
+    {
+      "epoch": 0.6295633876906891,
+      "grad_norm": 1.2039942741394043,
+      "learning_rate": 6.375156449486895e-05,
+      "loss": 1.6477,
+      "step": 5984
+    },
+    {
+      "epoch": 0.6296685954760652,
+      "grad_norm": 1.28606116771698,
+      "learning_rate": 6.371980683789479e-05,
+      "loss": 1.6914,
+      "step": 5985
+    },
+    {
+      "epoch": 0.6297738032614414,
+      "grad_norm": 1.0567982196807861,
+      "learning_rate": 6.368805339402e-05,
+      "loss": 1.9269,
+      "step": 5986
+    },
+    {
+      "epoch": 0.6298790110468174,
+      "grad_norm": 1.3032525777816772,
+      "learning_rate": 6.365630416693203e-05,
+      "loss": 1.9851,
+      "step": 5987
+    },
+    {
+      "epoch": 0.6299842188321936,
+      "grad_norm": 2.6468424797058105,
+      "learning_rate": 6.362455916031774e-05,
+      "loss": 1.8291,
+      "step": 5988
+    },
+    {
+      "epoch": 0.6300894266175697,
+      "grad_norm": 1.2577786445617676,
+      "learning_rate": 6.359281837786363e-05,
+      "loss": 1.6674,
+      "step": 5989
+    },
+    {
+      "epoch": 0.6301946344029458,
+      "grad_norm": 1.3022758960723877,
+      "learning_rate": 6.356108182325562e-05,
+      "loss": 1.4832,
+      "step": 5990
+    },
+    {
+      "epoch": 0.6302998421883219,
+      "grad_norm": 1.515007495880127,
+      "learning_rate": 6.352934950017921e-05,
+      "loss": 2.0965,
+      "step": 5991
+    },
+    {
+      "epoch": 0.6304050499736981,
+      "grad_norm": 1.462588906288147,
+      "learning_rate": 6.349762141231934e-05,
+      "loss": 1.5093,
+      "step": 5992
+    },
+    {
+      "epoch": 0.6305102577590742,
+      "grad_norm": 1.503005862236023,
+      "learning_rate": 6.34658975633605e-05,
+      "loss": 1.4638,
+      "step": 5993
+    },
+    {
+      "epoch": 0.6306154655444502,
+      "grad_norm": 1.4400442838668823,
+      "learning_rate": 6.343417795698667e-05,
+      "loss": 1.2181,
+      "step": 5994
+    },
+    {
+      "epoch": 0.6307206733298264,
+      "grad_norm": 1.6780505180358887,
+      "learning_rate": 6.340246259688133e-05,
+      "loss": 1.6997,
+      "step": 5995
+    },
+    {
+      "epoch": 0.6308258811152025,
+      "grad_norm": 1.3209447860717773,
+      "learning_rate": 6.337075148672751e-05,
+      "loss": 1.5113,
+      "step": 5996
+    },
+    {
+      "epoch": 0.6309310889005787,
+      "grad_norm": 1.7085509300231934,
+      "learning_rate": 6.333904463020772e-05,
+      "loss": 1.8627,
+      "step": 5997
+    },
+    {
+      "epoch": 0.6310362966859547,
+      "grad_norm": 1.7101259231567383,
+      "learning_rate": 6.330734203100394e-05,
+      "loss": 2.1645,
+      "step": 5998
+    },
+    {
+      "epoch": 0.6311415044713309,
+      "grad_norm": 1.1601276397705078,
+      "learning_rate": 6.327564369279768e-05,
+      "loss": 2.0211,
+      "step": 5999
+    },
+    {
+      "epoch": 0.631246712256707,
+      "grad_norm": 1.344951868057251,
+      "learning_rate": 6.324394961927005e-05,
+      "loss": 1.6263,
+      "step": 6000
+    },
+    {
+      "epoch": 0.6313519200420831,
+      "grad_norm": 1.1966255903244019,
+      "learning_rate": 6.321225981410147e-05,
+      "loss": 1.8129,
+      "step": 6001
+    },
+    {
+      "epoch": 0.6314571278274592,
+      "grad_norm": 1.8970527648925781,
+      "learning_rate": 6.318057428097203e-05,
+      "loss": 1.8969,
+      "step": 6002
+    },
+    {
+      "epoch": 0.6315623356128354,
+      "grad_norm": 1.4465181827545166,
+      "learning_rate": 6.314889302356125e-05,
+      "loss": 2.0013,
+      "step": 6003
+    },
+    {
+      "epoch": 0.6316675433982115,
+      "grad_norm": 1.2381423711776733,
+      "learning_rate": 6.311721604554816e-05,
+      "loss": 1.6398,
+      "step": 6004
+    },
+    {
+      "epoch": 0.6317727511835876,
+      "grad_norm": 1.656591773033142,
+      "learning_rate": 6.308554335061135e-05,
+      "loss": 1.757,
+      "step": 6005
+    },
+    {
+      "epoch": 0.6318779589689637,
+      "grad_norm": 1.8384569883346558,
+      "learning_rate": 6.305387494242882e-05,
+      "loss": 1.7847,
+      "step": 6006
+    },
+    {
+      "epoch": 0.6319831667543399,
+      "grad_norm": 1.5828437805175781,
+      "learning_rate": 6.302221082467812e-05,
+      "loss": 1.5665,
+      "step": 6007
+    },
+    {
+      "epoch": 0.6320883745397159,
+      "grad_norm": 1.4291951656341553,
+      "learning_rate": 6.299055100103632e-05,
+      "loss": 1.9382,
+      "step": 6008
+    },
+    {
+      "epoch": 0.632193582325092,
+      "grad_norm": 1.4509108066558838,
+      "learning_rate": 6.295889547517997e-05,
+      "loss": 2.2804,
+      "step": 6009
+    },
+    {
+      "epoch": 0.6322987901104682,
+      "grad_norm": 1.5382047891616821,
+      "learning_rate": 6.29272442507851e-05,
+      "loss": 0.9914,
+      "step": 6010
+    },
+    {
+      "epoch": 0.6324039978958443,
+      "grad_norm": 0.95433109998703,
+      "learning_rate": 6.289559733152727e-05,
+      "loss": 1.648,
+      "step": 6011
+    },
+    {
+      "epoch": 0.6325092056812204,
+      "grad_norm": 0.9161847829818726,
+      "learning_rate": 6.286395472108158e-05,
+      "loss": 1.5227,
+      "step": 6012
+    },
+    {
+      "epoch": 0.6326144134665965,
+      "grad_norm": 2.2820706367492676,
+      "learning_rate": 6.283231642312251e-05,
+      "loss": 1.4385,
+      "step": 6013
+    },
+    {
+      "epoch": 0.6327196212519727,
+      "grad_norm": 1.1651874780654907,
+      "learning_rate": 6.280068244132415e-05,
+      "loss": 1.3021,
+      "step": 6014
+    },
+    {
+      "epoch": 0.6328248290373487,
+      "grad_norm": 1.7624272108078003,
+      "learning_rate": 6.276905277936005e-05,
+      "loss": 1.7141,
+      "step": 6015
+    },
+    {
+      "epoch": 0.6329300368227249,
+      "grad_norm": 1.4675081968307495,
+      "learning_rate": 6.273742744090325e-05,
+      "loss": 1.8335,
+      "step": 6016
+    },
+    {
+      "epoch": 0.633035244608101,
+      "grad_norm": 1.489524245262146,
+      "learning_rate": 6.270580642962629e-05,
+      "loss": 1.8555,
+      "step": 6017
+    },
+    {
+      "epoch": 0.6331404523934772,
+      "grad_norm": 1.2529215812683105,
+      "learning_rate": 6.267418974920125e-05,
+      "loss": 1.7023,
+      "step": 6018
+    },
+    {
+      "epoch": 0.6332456601788532,
+      "grad_norm": 1.7118120193481445,
+      "learning_rate": 6.26425774032996e-05,
+      "loss": 1.7771,
+      "step": 6019
+    },
+    {
+      "epoch": 0.6333508679642293,
+      "grad_norm": 1.1630779504776,
+      "learning_rate": 6.261096939559243e-05,
+      "loss": 1.6504,
+      "step": 6020
+    },
+    {
+      "epoch": 0.6334560757496055,
+      "grad_norm": 1.5538249015808105,
+      "learning_rate": 6.257936572975029e-05,
+      "loss": 1.967,
+      "step": 6021
+    },
+    {
+      "epoch": 0.6335612835349815,
+      "grad_norm": 1.3640360832214355,
+      "learning_rate": 6.25477664094432e-05,
+      "loss": 1.583,
+      "step": 6022
+    },
+    {
+      "epoch": 0.6336664913203577,
+      "grad_norm": 1.6162277460098267,
+      "learning_rate": 6.251617143834065e-05,
+      "loss": 1.8734,
+      "step": 6023
+    },
+    {
+      "epoch": 0.6337716991057338,
+      "grad_norm": 1.6816718578338623,
+      "learning_rate": 6.248458082011167e-05,
+      "loss": 1.6501,
+      "step": 6024
+    },
+    {
+      "epoch": 0.63387690689111,
+      "grad_norm": 1.3452551364898682,
+      "learning_rate": 6.245299455842477e-05,
+      "loss": 1.609,
+      "step": 6025
+    },
+    {
+      "epoch": 0.633982114676486,
+      "grad_norm": 1.8827195167541504,
+      "learning_rate": 6.2421412656948e-05,
+      "loss": 1.2681,
+      "step": 6026
+    },
+    {
+      "epoch": 0.6340873224618622,
+      "grad_norm": 1.4761601686477661,
+      "learning_rate": 6.238983511934883e-05,
+      "loss": 1.4392,
+      "step": 6027
+    },
+    {
+      "epoch": 0.6341925302472383,
+      "grad_norm": 1.6614230871200562,
+      "learning_rate": 6.235826194929423e-05,
+      "loss": 1.8507,
+      "step": 6028
+    },
+    {
+      "epoch": 0.6342977380326145,
+      "grad_norm": 1.2700610160827637,
+      "learning_rate": 6.232669315045076e-05,
+      "loss": 1.9591,
+      "step": 6029
+    },
+    {
+      "epoch": 0.6344029458179905,
+      "grad_norm": 1.129776120185852,
+      "learning_rate": 6.229512872648435e-05,
+      "loss": 1.9298,
+      "step": 6030
+    },
+    {
+      "epoch": 0.6345081536033667,
+      "grad_norm": 1.5436091423034668,
+      "learning_rate": 6.226356868106046e-05,
+      "loss": 1.6272,
+      "step": 6031
+    },
+    {
+      "epoch": 0.6346133613887428,
+      "grad_norm": 1.54899001121521,
+      "learning_rate": 6.22320130178441e-05,
+      "loss": 1.6734,
+      "step": 6032
+    },
+    {
+      "epoch": 0.6347185691741188,
+      "grad_norm": 2.6774120330810547,
+      "learning_rate": 6.220046174049968e-05,
+      "loss": 1.5251,
+      "step": 6033
+    },
+    {
+      "epoch": 0.634823776959495,
+      "grad_norm": 1.2315189838409424,
+      "learning_rate": 6.216891485269118e-05,
+      "loss": 1.4497,
+      "step": 6034
+    },
+    {
+      "epoch": 0.6349289847448711,
+      "grad_norm": 0.9964236617088318,
+      "learning_rate": 6.213737235808201e-05,
+      "loss": 1.695,
+      "step": 6035
+    },
+    {
+      "epoch": 0.6350341925302473,
+      "grad_norm": 1.0956590175628662,
+      "learning_rate": 6.210583426033513e-05,
+      "loss": 2.1008,
+      "step": 6036
+    },
+    {
+      "epoch": 0.6351394003156233,
+      "grad_norm": 1.8140305280685425,
+      "learning_rate": 6.207430056311292e-05,
+      "loss": 1.6233,
+      "step": 6037
+    },
+    {
+      "epoch": 0.6352446081009995,
+      "grad_norm": 2.13716721534729,
+      "learning_rate": 6.204277127007729e-05,
+      "loss": 2.0819,
+      "step": 6038
+    },
+    {
+      "epoch": 0.6353498158863756,
+      "grad_norm": 1.4235343933105469,
+      "learning_rate": 6.201124638488968e-05,
+      "loss": 1.3355,
+      "step": 6039
+    },
+    {
+      "epoch": 0.6354550236717517,
+      "grad_norm": 1.6728320121765137,
+      "learning_rate": 6.19797259112109e-05,
+      "loss": 1.7395,
+      "step": 6040
+    },
+    {
+      "epoch": 0.6355602314571278,
+      "grad_norm": 1.455208659172058,
+      "learning_rate": 6.194820985270136e-05,
+      "loss": 1.9213,
+      "step": 6041
+    },
+    {
+      "epoch": 0.635665439242504,
+      "grad_norm": 1.577768325805664,
+      "learning_rate": 6.191669821302091e-05,
+      "loss": 1.9931,
+      "step": 6042
+    },
+    {
+      "epoch": 0.6357706470278801,
+      "grad_norm": 1.7181400060653687,
+      "learning_rate": 6.188519099582893e-05,
+      "loss": 1.69,
+      "step": 6043
+    },
+    {
+      "epoch": 0.6358758548132561,
+      "grad_norm": 1.1263937950134277,
+      "learning_rate": 6.185368820478417e-05,
+      "loss": 1.3001,
+      "step": 6044
+    },
+    {
+      "epoch": 0.6359810625986323,
+      "grad_norm": 1.8166402578353882,
+      "learning_rate": 6.182218984354497e-05,
+      "loss": 1.9147,
+      "step": 6045
+    },
+    {
+      "epoch": 0.6360862703840084,
+      "grad_norm": 1.7598438262939453,
+      "learning_rate": 6.179069591576916e-05,
+      "loss": 1.9621,
+      "step": 6046
+    },
+    {
+      "epoch": 0.6361914781693845,
+      "grad_norm": 1.1624400615692139,
+      "learning_rate": 6.175920642511404e-05,
+      "loss": 2.1653,
+      "step": 6047
+    },
+    {
+      "epoch": 0.6362966859547606,
+      "grad_norm": 2.103053092956543,
+      "learning_rate": 6.172772137523632e-05,
+      "loss": 1.9937,
+      "step": 6048
+    },
+    {
+      "epoch": 0.6364018937401368,
+      "grad_norm": 0.8931379914283752,
+      "learning_rate": 6.169624076979229e-05,
+      "loss": 1.8209,
+      "step": 6049
+    },
+    {
+      "epoch": 0.6365071015255129,
+      "grad_norm": 1.4275034666061401,
+      "learning_rate": 6.166476461243771e-05,
+      "loss": 1.3759,
+      "step": 6050
+    },
+    {
+      "epoch": 0.636612309310889,
+      "grad_norm": 1.7524127960205078,
+      "learning_rate": 6.163329290682775e-05,
+      "loss": 1.8575,
+      "step": 6051
+    },
+    {
+      "epoch": 0.6367175170962651,
+      "grad_norm": 1.2306782007217407,
+      "learning_rate": 6.160182565661717e-05,
+      "loss": 1.3555,
+      "step": 6052
+    },
+    {
+      "epoch": 0.6368227248816413,
+      "grad_norm": 1.199044942855835,
+      "learning_rate": 6.157036286546014e-05,
+      "loss": 1.6082,
+      "step": 6053
+    },
+    {
+      "epoch": 0.6369279326670173,
+      "grad_norm": 0.8733108043670654,
+      "learning_rate": 6.153890453701031e-05,
+      "loss": 1.8242,
+      "step": 6054
+    },
+    {
+      "epoch": 0.6370331404523935,
+      "grad_norm": 1.531132698059082,
+      "learning_rate": 6.150745067492085e-05,
+      "loss": 1.5927,
+      "step": 6055
+    },
+    {
+      "epoch": 0.6371383482377696,
+      "grad_norm": 2.1788547039031982,
+      "learning_rate": 6.14760012828444e-05,
+      "loss": 2.3346,
+      "step": 6056
+    },
+    {
+      "epoch": 0.6372435560231458,
+      "grad_norm": 1.90433931350708,
+      "learning_rate": 6.144455636443304e-05,
+      "loss": 1.5207,
+      "step": 6057
+    },
+    {
+      "epoch": 0.6373487638085218,
+      "grad_norm": 2.0594825744628906,
+      "learning_rate": 6.141311592333841e-05,
+      "loss": 1.23,
+      "step": 6058
+    },
+    {
+      "epoch": 0.6374539715938979,
+      "grad_norm": 1.3014582395553589,
+      "learning_rate": 6.138167996321155e-05,
+      "loss": 1.8238,
+      "step": 6059
+    },
+    {
+      "epoch": 0.6375591793792741,
+      "grad_norm": 1.347284197807312,
+      "learning_rate": 6.135024848770306e-05,
+      "loss": 1.1295,
+      "step": 6060
+    },
+    {
+      "epoch": 0.6376643871646502,
+      "grad_norm": 1.683262825012207,
+      "learning_rate": 6.131882150046291e-05,
+      "loss": 1.449,
+      "step": 6061
+    },
+    {
+      "epoch": 0.6377695949500263,
+      "grad_norm": 1.515728235244751,
+      "learning_rate": 6.128739900514064e-05,
+      "loss": 1.3518,
+      "step": 6062
+    },
+    {
+      "epoch": 0.6378748027354024,
+      "grad_norm": 1.24699068069458,
+      "learning_rate": 6.125598100538528e-05,
+      "loss": 1.9356,
+      "step": 6063
+    },
+    {
+      "epoch": 0.6379800105207786,
+      "grad_norm": 2.0184340476989746,
+      "learning_rate": 6.122456750484528e-05,
+      "loss": 1.4496,
+      "step": 6064
+    },
+    {
+      "epoch": 0.6380852183061546,
+      "grad_norm": 1.378085970878601,
+      "learning_rate": 6.119315850716853e-05,
+      "loss": 1.6465,
+      "step": 6065
+    },
+    {
+      "epoch": 0.6381904260915308,
+      "grad_norm": 1.6989691257476807,
+      "learning_rate": 6.116175401600249e-05,
+      "loss": 1.8344,
+      "step": 6066
+    },
+    {
+      "epoch": 0.6382956338769069,
+      "grad_norm": 1.1261394023895264,
+      "learning_rate": 6.113035403499408e-05,
+      "loss": 1.4539,
+      "step": 6067
+    },
+    {
+      "epoch": 0.6384008416622831,
+      "grad_norm": 1.6350476741790771,
+      "learning_rate": 6.109895856778967e-05,
+      "loss": 1.5789,
+      "step": 6068
+    },
+    {
+      "epoch": 0.6385060494476591,
+      "grad_norm": 1.9975515604019165,
+      "learning_rate": 6.106756761803507e-05,
+      "loss": 1.6606,
+      "step": 6069
+    },
+    {
+      "epoch": 0.6386112572330352,
+      "grad_norm": 1.355108618736267,
+      "learning_rate": 6.103618118937567e-05,
+      "loss": 1.4639,
+      "step": 6070
+    },
+    {
+      "epoch": 0.6387164650184114,
+      "grad_norm": 1.5188654661178589,
+      "learning_rate": 6.1004799285456235e-05,
+      "loss": 2.2806,
+      "step": 6071
+    },
+    {
+      "epoch": 0.6388216728037874,
+      "grad_norm": 1.9493741989135742,
+      "learning_rate": 6.097342190992105e-05,
+      "loss": 2.7309,
+      "step": 6072
+    },
+    {
+      "epoch": 0.6389268805891636,
+      "grad_norm": 1.889501690864563,
+      "learning_rate": 6.0942049066413855e-05,
+      "loss": 2.1665,
+      "step": 6073
+    },
+    {
+      "epoch": 0.6390320883745397,
+      "grad_norm": 1.2038894891738892,
+      "learning_rate": 6.091068075857791e-05,
+      "loss": 1.5338,
+      "step": 6074
+    },
+    {
+      "epoch": 0.6391372961599159,
+      "grad_norm": 1.072657823562622,
+      "learning_rate": 6.087931699005588e-05,
+      "loss": 1.9299,
+      "step": 6075
+    },
+    {
+      "epoch": 0.6392425039452919,
+      "grad_norm": 1.8095741271972656,
+      "learning_rate": 6.084795776448995e-05,
+      "loss": 1.8311,
+      "step": 6076
+    },
+    {
+      "epoch": 0.6393477117306681,
+      "grad_norm": 1.5067007541656494,
+      "learning_rate": 6.0816603085521764e-05,
+      "loss": 1.3399,
+      "step": 6077
+    },
+    {
+      "epoch": 0.6394529195160442,
+      "grad_norm": 1.420461654663086,
+      "learning_rate": 6.078525295679243e-05,
+      "loss": 2.0262,
+      "step": 6078
+    },
+    {
+      "epoch": 0.6395581273014203,
+      "grad_norm": 1.3096047639846802,
+      "learning_rate": 6.075390738194252e-05,
+      "loss": 1.7605,
+      "step": 6079
+    },
+    {
+      "epoch": 0.6396633350867964,
+      "grad_norm": 1.2972394227981567,
+      "learning_rate": 6.072256636461214e-05,
+      "loss": 1.6696,
+      "step": 6080
+    },
+    {
+      "epoch": 0.6397685428721726,
+      "grad_norm": 1.2217929363250732,
+      "learning_rate": 6.0691229908440775e-05,
+      "loss": 2.151,
+      "step": 6081
+    },
+    {
+      "epoch": 0.6398737506575487,
+      "grad_norm": 1.5966821908950806,
+      "learning_rate": 6.065989801706744e-05,
+      "loss": 1.2176,
+      "step": 6082
+    },
+    {
+      "epoch": 0.6399789584429247,
+      "grad_norm": 1.4299871921539307,
+      "learning_rate": 6.0628570694130594e-05,
+      "loss": 1.6411,
+      "step": 6083
+    },
+    {
+      "epoch": 0.6400841662283009,
+      "grad_norm": 1.8349710702896118,
+      "learning_rate": 6.059724794326822e-05,
+      "loss": 1.697,
+      "step": 6084
+    },
+    {
+      "epoch": 0.640189374013677,
+      "grad_norm": 0.9039126038551331,
+      "learning_rate": 6.056592976811766e-05,
+      "loss": 1.509,
+      "step": 6085
+    },
+    {
+      "epoch": 0.6402945817990531,
+      "grad_norm": 1.0073765516281128,
+      "learning_rate": 6.053461617231586e-05,
+      "loss": 1.7619,
+      "step": 6086
+    },
+    {
+      "epoch": 0.6403997895844292,
+      "grad_norm": 1.246187448501587,
+      "learning_rate": 6.05033071594991e-05,
+      "loss": 2.2623,
+      "step": 6087
+    },
+    {
+      "epoch": 0.6405049973698054,
+      "grad_norm": 1.0114855766296387,
+      "learning_rate": 6.047200273330325e-05,
+      "loss": 1.8475,
+      "step": 6088
+    },
+    {
+      "epoch": 0.6406102051551815,
+      "grad_norm": 1.7710341215133667,
+      "learning_rate": 6.044070289736352e-05,
+      "loss": 1.2952,
+      "step": 6089
+    },
+    {
+      "epoch": 0.6407154129405576,
+      "grad_norm": 1.3931844234466553,
+      "learning_rate": 6.04094076553147e-05,
+      "loss": 1.7707,
+      "step": 6090
+    },
+    {
+      "epoch": 0.6408206207259337,
+      "grad_norm": 1.1191129684448242,
+      "learning_rate": 6.0378117010791e-05,
+      "loss": 1.9957,
+      "step": 6091
+    },
+    {
+      "epoch": 0.6409258285113099,
+      "grad_norm": 1.1291834115982056,
+      "learning_rate": 6.034683096742613e-05,
+      "loss": 1.8528,
+      "step": 6092
+    },
+    {
+      "epoch": 0.641031036296686,
+      "grad_norm": 1.390347957611084,
+      "learning_rate": 6.031554952885317e-05,
+      "loss": 1.6936,
+      "step": 6093
+    },
+    {
+      "epoch": 0.641136244082062,
+      "grad_norm": 1.2321089506149292,
+      "learning_rate": 6.028427269870478e-05,
+      "loss": 1.4666,
+      "step": 6094
+    },
+    {
+      "epoch": 0.6412414518674382,
+      "grad_norm": 1.7802302837371826,
+      "learning_rate": 6.025300048061302e-05,
+      "loss": 2.4148,
+      "step": 6095
+    },
+    {
+      "epoch": 0.6413466596528143,
+      "grad_norm": 1.3233754634857178,
+      "learning_rate": 6.0221732878209425e-05,
+      "loss": 1.2631,
+      "step": 6096
+    },
+    {
+      "epoch": 0.6414518674381904,
+      "grad_norm": 1.9571051597595215,
+      "learning_rate": 6.0190469895125e-05,
+      "loss": 1.8066,
+      "step": 6097
+    },
+    {
+      "epoch": 0.6415570752235665,
+      "grad_norm": 1.236015796661377,
+      "learning_rate": 6.0159211534990246e-05,
+      "loss": 1.7285,
+      "step": 6098
+    },
+    {
+      "epoch": 0.6416622830089427,
+      "grad_norm": 2.017487049102783,
+      "learning_rate": 6.012795780143503e-05,
+      "loss": 2.0849,
+      "step": 6099
+    },
+    {
+      "epoch": 0.6417674907943188,
+      "grad_norm": 1.846824049949646,
+      "learning_rate": 6.009670869808879e-05,
+      "loss": 1.7874,
+      "step": 6100
+    },
+    {
+      "epoch": 0.6418726985796949,
+      "grad_norm": 2.1392126083374023,
+      "learning_rate": 6.006546422858039e-05,
+      "loss": 1.5903,
+      "step": 6101
+    },
+    {
+      "epoch": 0.641977906365071,
+      "grad_norm": 2.1033949851989746,
+      "learning_rate": 6.003422439653811e-05,
+      "loss": 1.7481,
+      "step": 6102
+    },
+    {
+      "epoch": 0.6420831141504472,
+      "grad_norm": 1.5869107246398926,
+      "learning_rate": 6.0002989205589734e-05,
+      "loss": 1.3625,
+      "step": 6103
+    },
+    {
+      "epoch": 0.6421883219358232,
+      "grad_norm": 1.291629433631897,
+      "learning_rate": 5.997175865936253e-05,
+      "loss": 1.7495,
+      "step": 6104
+    },
+    {
+      "epoch": 0.6422935297211994,
+      "grad_norm": 1.8969913721084595,
+      "learning_rate": 5.994053276148319e-05,
+      "loss": 1.7404,
+      "step": 6105
+    },
+    {
+      "epoch": 0.6423987375065755,
+      "grad_norm": 1.4604777097702026,
+      "learning_rate": 5.990931151557786e-05,
+      "loss": 1.749,
+      "step": 6106
+    },
+    {
+      "epoch": 0.6425039452919517,
+      "grad_norm": 1.6946632862091064,
+      "learning_rate": 5.987809492527219e-05,
+      "loss": 1.9089,
+      "step": 6107
+    },
+    {
+      "epoch": 0.6426091530773277,
+      "grad_norm": 1.3629951477050781,
+      "learning_rate": 5.984688299419121e-05,
+      "loss": 1.7441,
+      "step": 6108
+    },
+    {
+      "epoch": 0.6427143608627038,
+      "grad_norm": 1.3938630819320679,
+      "learning_rate": 5.981567572595951e-05,
+      "loss": 1.412,
+      "step": 6109
+    },
+    {
+      "epoch": 0.64281956864808,
+      "grad_norm": 1.3842976093292236,
+      "learning_rate": 5.978447312420103e-05,
+      "loss": 1.7809,
+      "step": 6110
+    },
+    {
+      "epoch": 0.642924776433456,
+      "grad_norm": 1.5348026752471924,
+      "learning_rate": 5.9753275192539284e-05,
+      "loss": 1.398,
+      "step": 6111
+    },
+    {
+      "epoch": 0.6430299842188322,
+      "grad_norm": 1.968120813369751,
+      "learning_rate": 5.9722081934597164e-05,
+      "loss": 1.6602,
+      "step": 6112
+    },
+    {
+      "epoch": 0.6431351920042083,
+      "grad_norm": 1.070837378501892,
+      "learning_rate": 5.9690893353997e-05,
+      "loss": 1.6361,
+      "step": 6113
+    },
+    {
+      "epoch": 0.6432403997895845,
+      "grad_norm": 2.2223477363586426,
+      "learning_rate": 5.965970945436068e-05,
+      "loss": 1.5202,
+      "step": 6114
+    },
+    {
+      "epoch": 0.6433456075749605,
+      "grad_norm": 1.0596355199813843,
+      "learning_rate": 5.9628530239309456e-05,
+      "loss": 1.4806,
+      "step": 6115
+    },
+    {
+      "epoch": 0.6434508153603367,
+      "grad_norm": 1.7896836996078491,
+      "learning_rate": 5.95973557124641e-05,
+      "loss": 1.3566,
+      "step": 6116
+    },
+    {
+      "epoch": 0.6435560231457128,
+      "grad_norm": 1.657974362373352,
+      "learning_rate": 5.9566185877444755e-05,
+      "loss": 1.7571,
+      "step": 6117
+    },
+    {
+      "epoch": 0.6436612309310888,
+      "grad_norm": 1.8864045143127441,
+      "learning_rate": 5.9535020737871115e-05,
+      "loss": 1.9383,
+      "step": 6118
+    },
+    {
+      "epoch": 0.643766438716465,
+      "grad_norm": 1.6928189992904663,
+      "learning_rate": 5.950386029736228e-05,
+      "loss": 1.394,
+      "step": 6119
+    },
+    {
+      "epoch": 0.6438716465018411,
+      "grad_norm": 1.4859063625335693,
+      "learning_rate": 5.947270455953677e-05,
+      "loss": 1.7629,
+      "step": 6120
+    },
+    {
+      "epoch": 0.6439768542872173,
+      "grad_norm": 2.7389304637908936,
+      "learning_rate": 5.9441553528012626e-05,
+      "loss": 1.9255,
+      "step": 6121
+    },
+    {
+      "epoch": 0.6440820620725933,
+      "grad_norm": 2.5364770889282227,
+      "learning_rate": 5.9410407206407335e-05,
+      "loss": 1.8894,
+      "step": 6122
+    },
+    {
+      "epoch": 0.6441872698579695,
+      "grad_norm": 1.21177077293396,
+      "learning_rate": 5.9379265598337786e-05,
+      "loss": 1.7663,
+      "step": 6123
+    },
+    {
+      "epoch": 0.6442924776433456,
+      "grad_norm": 1.6818538904190063,
+      "learning_rate": 5.934812870742036e-05,
+      "loss": 2.2091,
+      "step": 6124
+    },
+    {
+      "epoch": 0.6443976854287218,
+      "grad_norm": 1.6835134029388428,
+      "learning_rate": 5.9316996537270895e-05,
+      "loss": 1.6155,
+      "step": 6125
+    },
+    {
+      "epoch": 0.6445028932140978,
+      "grad_norm": 1.7322865724563599,
+      "learning_rate": 5.928586909150464e-05,
+      "loss": 1.8218,
+      "step": 6126
+    },
+    {
+      "epoch": 0.644608100999474,
+      "grad_norm": 2.4374451637268066,
+      "learning_rate": 5.925474637373635e-05,
+      "loss": 1.2015,
+      "step": 6127
+    },
+    {
+      "epoch": 0.6447133087848501,
+      "grad_norm": 1.3041414022445679,
+      "learning_rate": 5.9223628387580176e-05,
+      "loss": 2.0628,
+      "step": 6128
+    },
+    {
+      "epoch": 0.6448185165702262,
+      "grad_norm": 1.4549946784973145,
+      "learning_rate": 5.919251513664982e-05,
+      "loss": 1.3928,
+      "step": 6129
+    },
+    {
+      "epoch": 0.6449237243556023,
+      "grad_norm": 2.7458503246307373,
+      "learning_rate": 5.916140662455828e-05,
+      "loss": 1.7509,
+      "step": 6130
+    },
+    {
+      "epoch": 0.6450289321409785,
+      "grad_norm": 1.7431880235671997,
+      "learning_rate": 5.913030285491808e-05,
+      "loss": 1.6354,
+      "step": 6131
+    },
+    {
+      "epoch": 0.6451341399263546,
+      "grad_norm": 1.597713828086853,
+      "learning_rate": 5.909920383134124e-05,
+      "loss": 1.6892,
+      "step": 6132
+    },
+    {
+      "epoch": 0.6452393477117306,
+      "grad_norm": 1.7393560409545898,
+      "learning_rate": 5.90681095574392e-05,
+      "loss": 1.5215,
+      "step": 6133
+    },
+    {
+      "epoch": 0.6453445554971068,
+      "grad_norm": 1.863715648651123,
+      "learning_rate": 5.903702003682278e-05,
+      "loss": 2.0865,
+      "step": 6134
+    },
+    {
+      "epoch": 0.6454497632824829,
+      "grad_norm": 1.6684576272964478,
+      "learning_rate": 5.900593527310234e-05,
+      "loss": 2.2619,
+      "step": 6135
+    },
+    {
+      "epoch": 0.645554971067859,
+      "grad_norm": 1.9676207304000854,
+      "learning_rate": 5.897485526988766e-05,
+      "loss": 1.9868,
+      "step": 6136
+    },
+    {
+      "epoch": 0.6456601788532351,
+      "grad_norm": 2.0112955570220947,
+      "learning_rate": 5.8943780030787935e-05,
+      "loss": 1.1787,
+      "step": 6137
+    },
+    {
+      "epoch": 0.6457653866386113,
+      "grad_norm": 1.3841115236282349,
+      "learning_rate": 5.891270955941184e-05,
+      "loss": 1.5988,
+      "step": 6138
+    },
+    {
+      "epoch": 0.6458705944239874,
+      "grad_norm": 2.0346131324768066,
+      "learning_rate": 5.8881643859367475e-05,
+      "loss": 1.7077,
+      "step": 6139
+    },
+    {
+      "epoch": 0.6459758022093635,
+      "grad_norm": 1.9194620847702026,
+      "learning_rate": 5.8850582934262446e-05,
+      "loss": 1.7822,
+      "step": 6140
+    },
+    {
+      "epoch": 0.6460810099947396,
+      "grad_norm": 1.3288987874984741,
+      "learning_rate": 5.881952678770368e-05,
+      "loss": 1.7559,
+      "step": 6141
+    },
+    {
+      "epoch": 0.6461862177801158,
+      "grad_norm": 1.0419418811798096,
+      "learning_rate": 5.8788475423297674e-05,
+      "loss": 2.037,
+      "step": 6142
+    },
+    {
+      "epoch": 0.6462914255654918,
+      "grad_norm": 1.2322132587432861,
+      "learning_rate": 5.875742884465033e-05,
+      "loss": 2.2489,
+      "step": 6143
+    },
+    {
+      "epoch": 0.6463966333508679,
+      "grad_norm": 1.2339824438095093,
+      "learning_rate": 5.872638705536695e-05,
+      "loss": 1.3035,
+      "step": 6144
+    },
+    {
+      "epoch": 0.6465018411362441,
+      "grad_norm": 1.2765202522277832,
+      "learning_rate": 5.869535005905232e-05,
+      "loss": 1.6457,
+      "step": 6145
+    },
+    {
+      "epoch": 0.6466070489216202,
+      "grad_norm": 1.7985990047454834,
+      "learning_rate": 5.86643178593107e-05,
+      "loss": 1.2573,
+      "step": 6146
+    },
+    {
+      "epoch": 0.6467122567069963,
+      "grad_norm": 1.2002838850021362,
+      "learning_rate": 5.863329045974572e-05,
+      "loss": 1.7499,
+      "step": 6147
+    },
+    {
+      "epoch": 0.6468174644923724,
+      "grad_norm": 1.2847400903701782,
+      "learning_rate": 5.860226786396049e-05,
+      "loss": 2.0853,
+      "step": 6148
+    },
+    {
+      "epoch": 0.6469226722777486,
+      "grad_norm": 1.2351857423782349,
+      "learning_rate": 5.857125007555759e-05,
+      "loss": 1.9727,
+      "step": 6149
+    },
+    {
+      "epoch": 0.6470278800631246,
+      "grad_norm": 1.335922360420227,
+      "learning_rate": 5.8540237098139006e-05,
+      "loss": 1.3194,
+      "step": 6150
+    },
+    {
+      "epoch": 0.6471330878485008,
+      "grad_norm": 2.39359188079834,
+      "learning_rate": 5.850922893530617e-05,
+      "loss": 1.9385,
+      "step": 6151
+    },
+    {
+      "epoch": 0.6472382956338769,
+      "grad_norm": 1.5656342506408691,
+      "learning_rate": 5.847822559065992e-05,
+      "loss": 2.1779,
+      "step": 6152
+    },
+    {
+      "epoch": 0.6473435034192531,
+      "grad_norm": 1.4513521194458008,
+      "learning_rate": 5.8447227067800616e-05,
+      "loss": 1.644,
+      "step": 6153
+    },
+    {
+      "epoch": 0.6474487112046291,
+      "grad_norm": 1.374786138534546,
+      "learning_rate": 5.841623337032801e-05,
+      "loss": 1.4539,
+      "step": 6154
+    },
+    {
+      "epoch": 0.6475539189900053,
+      "grad_norm": 1.5855339765548706,
+      "learning_rate": 5.838524450184126e-05,
+      "loss": 1.6174,
+      "step": 6155
+    },
+    {
+      "epoch": 0.6476591267753814,
+      "grad_norm": 1.3137321472167969,
+      "learning_rate": 5.835426046593906e-05,
+      "loss": 1.5084,
+      "step": 6156
+    },
+    {
+      "epoch": 0.6477643345607575,
+      "grad_norm": 1.2792092561721802,
+      "learning_rate": 5.8323281266219466e-05,
+      "loss": 1.8137,
+      "step": 6157
+    },
+    {
+      "epoch": 0.6478695423461336,
+      "grad_norm": 1.172930121421814,
+      "learning_rate": 5.8292306906279935e-05,
+      "loss": 1.7933,
+      "step": 6158
+    },
+    {
+      "epoch": 0.6479747501315097,
+      "grad_norm": 1.8277498483657837,
+      "learning_rate": 5.8261337389717506e-05,
+      "loss": 2.0978,
+      "step": 6159
+    },
+    {
+      "epoch": 0.6480799579168859,
+      "grad_norm": 1.2689887285232544,
+      "learning_rate": 5.823037272012852e-05,
+      "loss": 1.3535,
+      "step": 6160
+    },
+    {
+      "epoch": 0.6481851657022619,
+      "grad_norm": 1.2342352867126465,
+      "learning_rate": 5.8199412901108774e-05,
+      "loss": 2.1196,
+      "step": 6161
+    },
+    {
+      "epoch": 0.6482903734876381,
+      "grad_norm": 1.4118258953094482,
+      "learning_rate": 5.8168457936253604e-05,
+      "loss": 2.2767,
+      "step": 6162
+    },
+    {
+      "epoch": 0.6483955812730142,
+      "grad_norm": 2.2059452533721924,
+      "learning_rate": 5.8137507829157655e-05,
+      "loss": 1.887,
+      "step": 6163
+    },
+    {
+      "epoch": 0.6485007890583904,
+      "grad_norm": 1.9310505390167236,
+      "learning_rate": 5.8106562583415037e-05,
+      "loss": 1.9649,
+      "step": 6164
+    },
+    {
+      "epoch": 0.6486059968437664,
+      "grad_norm": 2.263485908508301,
+      "learning_rate": 5.807562220261939e-05,
+      "loss": 2.1794,
+      "step": 6165
+    },
+    {
+      "epoch": 0.6487112046291426,
+      "grad_norm": 1.5354760885238647,
+      "learning_rate": 5.804468669036369e-05,
+      "loss": 1.8551,
+      "step": 6166
+    },
+    {
+      "epoch": 0.6488164124145187,
+      "grad_norm": 2.3956634998321533,
+      "learning_rate": 5.801375605024033e-05,
+      "loss": 1.8773,
+      "step": 6167
+    },
+    {
+      "epoch": 0.6489216201998947,
+      "grad_norm": 1.603357195854187,
+      "learning_rate": 5.798283028584126e-05,
+      "loss": 1.9813,
+      "step": 6168
+    },
+    {
+      "epoch": 0.6490268279852709,
+      "grad_norm": 1.1227657794952393,
+      "learning_rate": 5.795190940075774e-05,
+      "loss": 1.279,
+      "step": 6169
+    },
+    {
+      "epoch": 0.649132035770647,
+      "grad_norm": 1.2054638862609863,
+      "learning_rate": 5.792099339858048e-05,
+      "loss": 1.933,
+      "step": 6170
+    },
+    {
+      "epoch": 0.6492372435560232,
+      "grad_norm": 1.7264693975448608,
+      "learning_rate": 5.789008228289975e-05,
+      "loss": 1.4221,
+      "step": 6171
+    },
+    {
+      "epoch": 0.6493424513413992,
+      "grad_norm": 1.8171310424804688,
+      "learning_rate": 5.785917605730509e-05,
+      "loss": 2.2143,
+      "step": 6172
+    },
+    {
+      "epoch": 0.6494476591267754,
+      "grad_norm": 1.2820940017700195,
+      "learning_rate": 5.7828274725385544e-05,
+      "loss": 1.1761,
+      "step": 6173
+    },
+    {
+      "epoch": 0.6495528669121515,
+      "grad_norm": 2.7752187252044678,
+      "learning_rate": 5.7797378290729595e-05,
+      "loss": 1.6048,
+      "step": 6174
+    },
+    {
+      "epoch": 0.6496580746975276,
+      "grad_norm": 2.0099592208862305,
+      "learning_rate": 5.7766486756925086e-05,
+      "loss": 2.1283,
+      "step": 6175
+    },
+    {
+      "epoch": 0.6497632824829037,
+      "grad_norm": 1.9316635131835938,
+      "learning_rate": 5.773560012755945e-05,
+      "loss": 2.0693,
+      "step": 6176
+    },
+    {
+      "epoch": 0.6498684902682799,
+      "grad_norm": 1.1823091506958008,
+      "learning_rate": 5.770471840621938e-05,
+      "loss": 2.015,
+      "step": 6177
+    },
+    {
+      "epoch": 0.649973698053656,
+      "grad_norm": 1.3414077758789062,
+      "learning_rate": 5.767384159649107e-05,
+      "loss": 1.5643,
+      "step": 6178
+    },
+    {
+      "epoch": 0.650078905839032,
+      "grad_norm": 1.3535269498825073,
+      "learning_rate": 5.764296970196018e-05,
+      "loss": 1.6854,
+      "step": 6179
+    },
+    {
+      "epoch": 0.6501841136244082,
+      "grad_norm": 1.28633451461792,
+      "learning_rate": 5.761210272621175e-05,
+      "loss": 1.9889,
+      "step": 6180
+    },
+    {
+      "epoch": 0.6502893214097843,
+      "grad_norm": 1.7541005611419678,
+      "learning_rate": 5.7581240672830206e-05,
+      "loss": 2.1746,
+      "step": 6181
+    },
+    {
+      "epoch": 0.6503945291951604,
+      "grad_norm": 1.0183143615722656,
+      "learning_rate": 5.7550383545399545e-05,
+      "loss": 1.1908,
+      "step": 6182
+    },
+    {
+      "epoch": 0.6504997369805365,
+      "grad_norm": 1.4759997129440308,
+      "learning_rate": 5.7519531347503054e-05,
+      "loss": 1.4782,
+      "step": 6183
+    },
+    {
+      "epoch": 0.6506049447659127,
+      "grad_norm": 1.615997076034546,
+      "learning_rate": 5.7488684082723454e-05,
+      "loss": 1.966,
+      "step": 6184
+    },
+    {
+      "epoch": 0.6507101525512888,
+      "grad_norm": 1.790147304534912,
+      "learning_rate": 5.745784175464304e-05,
+      "loss": 1.5509,
+      "step": 6185
+    },
+    {
+      "epoch": 0.6508153603366649,
+      "grad_norm": 1.5848060846328735,
+      "learning_rate": 5.742700436684337e-05,
+      "loss": 1.7556,
+      "step": 6186
+    },
+    {
+      "epoch": 0.650920568122041,
+      "grad_norm": 1.6461782455444336,
+      "learning_rate": 5.739617192290545e-05,
+      "loss": 1.6253,
+      "step": 6187
+    },
+    {
+      "epoch": 0.6510257759074172,
+      "grad_norm": 1.7758451700210571,
+      "learning_rate": 5.736534442640984e-05,
+      "loss": 1.7024,
+      "step": 6188
+    },
+    {
+      "epoch": 0.6511309836927933,
+      "grad_norm": 1.5912493467330933,
+      "learning_rate": 5.73345218809364e-05,
+      "loss": 1.5275,
+      "step": 6189
+    },
+    {
+      "epoch": 0.6512361914781694,
+      "grad_norm": 0.911089301109314,
+      "learning_rate": 5.7303704290064375e-05,
+      "loss": 1.5513,
+      "step": 6190
+    },
+    {
+      "epoch": 0.6513413992635455,
+      "grad_norm": 1.3374494314193726,
+      "learning_rate": 5.727289165737263e-05,
+      "loss": 1.6507,
+      "step": 6191
+    },
+    {
+      "epoch": 0.6514466070489217,
+      "grad_norm": 1.4440089464187622,
+      "learning_rate": 5.724208398643924e-05,
+      "loss": 1.9977,
+      "step": 6192
+    },
+    {
+      "epoch": 0.6515518148342977,
+      "grad_norm": 2.0308356285095215,
+      "learning_rate": 5.721128128084191e-05,
+      "loss": 1.7098,
+      "step": 6193
+    },
+    {
+      "epoch": 0.6516570226196738,
+      "grad_norm": 1.5039621591567993,
+      "learning_rate": 5.7180483544157546e-05,
+      "loss": 1.539,
+      "step": 6194
+    },
+    {
+      "epoch": 0.65176223040505,
+      "grad_norm": 1.9608867168426514,
+      "learning_rate": 5.7149690779962594e-05,
+      "loss": 1.5212,
+      "step": 6195
+    },
+    {
+      "epoch": 0.6518674381904261,
+      "grad_norm": 2.1692678928375244,
+      "learning_rate": 5.711890299183298e-05,
+      "loss": 1.9479,
+      "step": 6196
+    },
+    {
+      "epoch": 0.6519726459758022,
+      "grad_norm": 1.2889890670776367,
+      "learning_rate": 5.7088120183343976e-05,
+      "loss": 1.9426,
+      "step": 6197
+    },
+    {
+      "epoch": 0.6520778537611783,
+      "grad_norm": 1.587933897972107,
+      "learning_rate": 5.705734235807021e-05,
+      "loss": 1.6135,
+      "step": 6198
+    },
+    {
+      "epoch": 0.6521830615465545,
+      "grad_norm": 2.34407639503479,
+      "learning_rate": 5.7026569519585916e-05,
+      "loss": 1.769,
+      "step": 6199
+    },
+    {
+      "epoch": 0.6522882693319305,
+      "grad_norm": 1.3627784252166748,
+      "learning_rate": 5.6995801671464556e-05,
+      "loss": 1.8221,
+      "step": 6200
+    },
+    {
+      "epoch": 0.6523934771173067,
+      "grad_norm": 1.99224054813385,
+      "learning_rate": 5.696503881727917e-05,
+      "loss": 1.7675,
+      "step": 6201
+    },
+    {
+      "epoch": 0.6524986849026828,
+      "grad_norm": 1.4435157775878906,
+      "learning_rate": 5.693428096060213e-05,
+      "loss": 1.2552,
+      "step": 6202
+    },
+    {
+      "epoch": 0.652603892688059,
+      "grad_norm": 1.6768929958343506,
+      "learning_rate": 5.6903528105005175e-05,
+      "loss": 2.1129,
+      "step": 6203
+    },
+    {
+      "epoch": 0.652709100473435,
+      "grad_norm": 1.7496428489685059,
+      "learning_rate": 5.6872780254059646e-05,
+      "loss": 1.8461,
+      "step": 6204
+    },
+    {
+      "epoch": 0.6528143082588111,
+      "grad_norm": 2.0133605003356934,
+      "learning_rate": 5.6842037411336116e-05,
+      "loss": 1.4858,
+      "step": 6205
+    },
+    {
+      "epoch": 0.6529195160441873,
+      "grad_norm": 1.2797411680221558,
+      "learning_rate": 5.6811299580404634e-05,
+      "loss": 1.4251,
+      "step": 6206
+    },
+    {
+      "epoch": 0.6530247238295633,
+      "grad_norm": 1.3031213283538818,
+      "learning_rate": 5.678056676483474e-05,
+      "loss": 1.9861,
+      "step": 6207
+    },
+    {
+      "epoch": 0.6531299316149395,
+      "grad_norm": 0.9973400831222534,
+      "learning_rate": 5.6749838968195326e-05,
+      "loss": 1.543,
+      "step": 6208
+    },
+    {
+      "epoch": 0.6532351394003156,
+      "grad_norm": 1.331733226776123,
+      "learning_rate": 5.671911619405465e-05,
+      "loss": 1.9739,
+      "step": 6209
+    },
+    {
+      "epoch": 0.6533403471856918,
+      "grad_norm": 2.5411765575408936,
+      "learning_rate": 5.668839844598053e-05,
+      "loss": 2.0505,
+      "step": 6210
+    },
+    {
+      "epoch": 0.6534455549710678,
+      "grad_norm": 1.2405493259429932,
+      "learning_rate": 5.665768572754007e-05,
+      "loss": 1.9774,
+      "step": 6211
+    },
+    {
+      "epoch": 0.653550762756444,
+      "grad_norm": 1.3992949724197388,
+      "learning_rate": 5.6626978042299814e-05,
+      "loss": 1.735,
+      "step": 6212
+    },
+    {
+      "epoch": 0.6536559705418201,
+      "grad_norm": 1.305866003036499,
+      "learning_rate": 5.6596275393825804e-05,
+      "loss": 2.2097,
+      "step": 6213
+    },
+    {
+      "epoch": 0.6537611783271962,
+      "grad_norm": 1.8051230907440186,
+      "learning_rate": 5.656557778568341e-05,
+      "loss": 0.9921,
+      "step": 6214
+    },
+    {
+      "epoch": 0.6538663861125723,
+      "grad_norm": 1.1146427392959595,
+      "learning_rate": 5.653488522143744e-05,
+      "loss": 1.6378,
+      "step": 6215
+    },
+    {
+      "epoch": 0.6539715938979485,
+      "grad_norm": 1.5699180364608765,
+      "learning_rate": 5.650419770465213e-05,
+      "loss": 1.5805,
+      "step": 6216
+    },
+    {
+      "epoch": 0.6540768016833246,
+      "grad_norm": 1.5119853019714355,
+      "learning_rate": 5.647351523889106e-05,
+      "loss": 1.5158,
+      "step": 6217
+    },
+    {
+      "epoch": 0.6541820094687006,
+      "grad_norm": 1.7245757579803467,
+      "learning_rate": 5.6442837827717386e-05,
+      "loss": 1.2592,
+      "step": 6218
+    },
+    {
+      "epoch": 0.6542872172540768,
+      "grad_norm": 1.8998960256576538,
+      "learning_rate": 5.6412165474693526e-05,
+      "loss": 1.7007,
+      "step": 6219
+    },
+    {
+      "epoch": 0.6543924250394529,
+      "grad_norm": 1.3357900381088257,
+      "learning_rate": 5.638149818338131e-05,
+      "loss": 1.5266,
+      "step": 6220
+    },
+    {
+      "epoch": 0.6544976328248291,
+      "grad_norm": 1.37607741355896,
+      "learning_rate": 5.635083595734212e-05,
+      "loss": 1.6224,
+      "step": 6221
+    },
+    {
+      "epoch": 0.6546028406102051,
+      "grad_norm": 1.2907686233520508,
+      "learning_rate": 5.6320178800136626e-05,
+      "loss": 1.5942,
+      "step": 6222
+    },
+    {
+      "epoch": 0.6547080483955813,
+      "grad_norm": 1.0506342649459839,
+      "learning_rate": 5.628952671532488e-05,
+      "loss": 1.6633,
+      "step": 6223
+    },
+    {
+      "epoch": 0.6548132561809574,
+      "grad_norm": 1.3889976739883423,
+      "learning_rate": 5.625887970646651e-05,
+      "loss": 1.5176,
+      "step": 6224
+    },
+    {
+      "epoch": 0.6549184639663335,
+      "grad_norm": 1.7190347909927368,
+      "learning_rate": 5.6228237777120406e-05,
+      "loss": 1.4025,
+      "step": 6225
+    },
+    {
+      "epoch": 0.6550236717517096,
+      "grad_norm": 2.0763463973999023,
+      "learning_rate": 5.6197600930844864e-05,
+      "loss": 1.6678,
+      "step": 6226
+    },
+    {
+      "epoch": 0.6551288795370858,
+      "grad_norm": 1.7063326835632324,
+      "learning_rate": 5.616696917119773e-05,
+      "loss": 1.6968,
+      "step": 6227
+    },
+    {
+      "epoch": 0.6552340873224619,
+      "grad_norm": 1.3571560382843018,
+      "learning_rate": 5.6136342501736126e-05,
+      "loss": 1.7551,
+      "step": 6228
+    },
+    {
+      "epoch": 0.655339295107838,
+      "grad_norm": 1.795301079750061,
+      "learning_rate": 5.610572092601659e-05,
+      "loss": 1.7408,
+      "step": 6229
+    },
+    {
+      "epoch": 0.6554445028932141,
+      "grad_norm": 1.3519048690795898,
+      "learning_rate": 5.6075104447595186e-05,
+      "loss": 1.8361,
+      "step": 6230
+    },
+    {
+      "epoch": 0.6555497106785902,
+      "grad_norm": 1.5597832202911377,
+      "learning_rate": 5.604449307002726e-05,
+      "loss": 2.2093,
+      "step": 6231
+    },
+    {
+      "epoch": 0.6556549184639663,
+      "grad_norm": 1.250503420829773,
+      "learning_rate": 5.601388679686757e-05,
+      "loss": 1.7934,
+      "step": 6232
+    },
+    {
+      "epoch": 0.6557601262493424,
+      "grad_norm": 1.4893306493759155,
+      "learning_rate": 5.598328563167039e-05,
+      "loss": 1.7416,
+      "step": 6233
+    },
+    {
+      "epoch": 0.6558653340347186,
+      "grad_norm": 2.179457902908325,
+      "learning_rate": 5.5952689577989324e-05,
+      "loss": 1.2715,
+      "step": 6234
+    },
+    {
+      "epoch": 0.6559705418200947,
+      "grad_norm": 2.1621286869049072,
+      "learning_rate": 5.592209863937733e-05,
+      "loss": 1.8506,
+      "step": 6235
+    },
+    {
+      "epoch": 0.6560757496054708,
+      "grad_norm": 1.5253031253814697,
+      "learning_rate": 5.589151281938695e-05,
+      "loss": 1.5244,
+      "step": 6236
+    },
+    {
+      "epoch": 0.6561809573908469,
+      "grad_norm": 2.098914384841919,
+      "learning_rate": 5.586093212156986e-05,
+      "loss": 1.7779,
+      "step": 6237
+    },
+    {
+      "epoch": 0.6562861651762231,
+      "grad_norm": 1.6869080066680908,
+      "learning_rate": 5.583035654947743e-05,
+      "loss": 1.5032,
+      "step": 6238
+    },
+    {
+      "epoch": 0.6563913729615991,
+      "grad_norm": 1.3635993003845215,
+      "learning_rate": 5.5799786106660234e-05,
+      "loss": 1.3241,
+      "step": 6239
+    },
+    {
+      "epoch": 0.6564965807469753,
+      "grad_norm": 1.3757116794586182,
+      "learning_rate": 5.576922079666829e-05,
+      "loss": 1.5449,
+      "step": 6240
+    },
+    {
+      "epoch": 0.6566017885323514,
+      "grad_norm": 1.2479023933410645,
+      "learning_rate": 5.573866062305113e-05,
+      "loss": 1.9185,
+      "step": 6241
+    },
+    {
+      "epoch": 0.6567069963177276,
+      "grad_norm": 1.2713433504104614,
+      "learning_rate": 5.570810558935756e-05,
+      "loss": 1.955,
+      "step": 6242
+    },
+    {
+      "epoch": 0.6568122041031036,
+      "grad_norm": 1.468234658241272,
+      "learning_rate": 5.56775556991358e-05,
+      "loss": 1.5705,
+      "step": 6243
+    },
+    {
+      "epoch": 0.6569174118884797,
+      "grad_norm": 1.4933536052703857,
+      "learning_rate": 5.5647010955933586e-05,
+      "loss": 1.5719,
+      "step": 6244
+    },
+    {
+      "epoch": 0.6570226196738559,
+      "grad_norm": 1.548898458480835,
+      "learning_rate": 5.561647136329789e-05,
+      "loss": 1.3829,
+      "step": 6245
+    },
+    {
+      "epoch": 0.6571278274592319,
+      "grad_norm": 1.346425175666809,
+      "learning_rate": 5.5585936924775275e-05,
+      "loss": 2.2525,
+      "step": 6246
+    },
+    {
+      "epoch": 0.6572330352446081,
+      "grad_norm": 1.4224555492401123,
+      "learning_rate": 5.555540764391156e-05,
+      "loss": 1.5195,
+      "step": 6247
+    },
+    {
+      "epoch": 0.6573382430299842,
+      "grad_norm": 1.1593234539031982,
+      "learning_rate": 5.552488352425195e-05,
+      "loss": 1.8726,
+      "step": 6248
+    },
+    {
+      "epoch": 0.6574434508153604,
+      "grad_norm": 1.3007352352142334,
+      "learning_rate": 5.549436456934121e-05,
+      "loss": 1.6618,
+      "step": 6249
+    },
+    {
+      "epoch": 0.6575486586007364,
+      "grad_norm": 1.5232993364334106,
+      "learning_rate": 5.5463850782723346e-05,
+      "loss": 1.1512,
+      "step": 6250
+    },
+    {
+      "epoch": 0.6576538663861126,
+      "grad_norm": 1.0250799655914307,
+      "learning_rate": 5.5433342167941803e-05,
+      "loss": 1.8019,
+      "step": 6251
+    },
+    {
+      "epoch": 0.6577590741714887,
+      "grad_norm": 1.5869766473770142,
+      "learning_rate": 5.540283872853953e-05,
+      "loss": 1.0935,
+      "step": 6252
+    },
+    {
+      "epoch": 0.6578642819568649,
+      "grad_norm": 1.3457857370376587,
+      "learning_rate": 5.5372340468058726e-05,
+      "loss": 1.8767,
+      "step": 6253
+    },
+    {
+      "epoch": 0.6579694897422409,
+      "grad_norm": 1.3926914930343628,
+      "learning_rate": 5.5341847390041035e-05,
+      "loss": 1.7272,
+      "step": 6254
+    },
+    {
+      "epoch": 0.658074697527617,
+      "grad_norm": 1.3577390909194946,
+      "learning_rate": 5.531135949802759e-05,
+      "loss": 1.9582,
+      "step": 6255
+    },
+    {
+      "epoch": 0.6581799053129932,
+      "grad_norm": 1.4779307842254639,
+      "learning_rate": 5.52808767955588e-05,
+      "loss": 1.578,
+      "step": 6256
+    },
+    {
+      "epoch": 0.6582851130983692,
+      "grad_norm": 1.3357998132705688,
+      "learning_rate": 5.5250399286174546e-05,
+      "loss": 1.6586,
+      "step": 6257
+    },
+    {
+      "epoch": 0.6583903208837454,
+      "grad_norm": 1.462868094444275,
+      "learning_rate": 5.521992697341407e-05,
+      "loss": 1.3614,
+      "step": 6258
+    },
+    {
+      "epoch": 0.6584955286691215,
+      "grad_norm": 1.4444942474365234,
+      "learning_rate": 5.518945986081596e-05,
+      "loss": 1.1518,
+      "step": 6259
+    },
+    {
+      "epoch": 0.6586007364544977,
+      "grad_norm": 1.828895092010498,
+      "learning_rate": 5.515899795191837e-05,
+      "loss": 1.2489,
+      "step": 6260
+    },
+    {
+      "epoch": 0.6587059442398737,
+      "grad_norm": 1.7447843551635742,
+      "learning_rate": 5.512854125025868e-05,
+      "loss": 1.7433,
+      "step": 6261
+    },
+    {
+      "epoch": 0.6588111520252499,
+      "grad_norm": 1.3825078010559082,
+      "learning_rate": 5.5098089759373714e-05,
+      "loss": 2.2669,
+      "step": 6262
+    },
+    {
+      "epoch": 0.658916359810626,
+      "grad_norm": 1.5063486099243164,
+      "learning_rate": 5.5067643482799746e-05,
+      "loss": 1.814,
+      "step": 6263
+    },
+    {
+      "epoch": 0.6590215675960021,
+      "grad_norm": 1.3307785987854004,
+      "learning_rate": 5.50372024240724e-05,
+      "loss": 1.5017,
+      "step": 6264
+    },
+    {
+      "epoch": 0.6591267753813782,
+      "grad_norm": 1.2832993268966675,
+      "learning_rate": 5.500676658672662e-05,
+      "loss": 1.5099,
+      "step": 6265
+    },
+    {
+      "epoch": 0.6592319831667544,
+      "grad_norm": 1.9960840940475464,
+      "learning_rate": 5.4976335974296923e-05,
+      "loss": 1.546,
+      "step": 6266
+    },
+    {
+      "epoch": 0.6593371909521305,
+      "grad_norm": 2.2886388301849365,
+      "learning_rate": 5.4945910590317074e-05,
+      "loss": 1.4721,
+      "step": 6267
+    },
+    {
+      "epoch": 0.6594423987375065,
+      "grad_norm": 1.4624122381210327,
+      "learning_rate": 5.491549043832023e-05,
+      "loss": 1.8469,
+      "step": 6268
+    },
+    {
+      "epoch": 0.6595476065228827,
+      "grad_norm": 1.8550081253051758,
+      "learning_rate": 5.488507552183906e-05,
+      "loss": 1.6482,
+      "step": 6269
+    },
+    {
+      "epoch": 0.6596528143082588,
+      "grad_norm": 1.4133846759796143,
+      "learning_rate": 5.4854665844405505e-05,
+      "loss": 1.7455,
+      "step": 6270
+    },
+    {
+      "epoch": 0.6597580220936349,
+      "grad_norm": 1.177406668663025,
+      "learning_rate": 5.48242614095509e-05,
+      "loss": 1.6862,
+      "step": 6271
+    },
+    {
+      "epoch": 0.659863229879011,
+      "grad_norm": 1.473006010055542,
+      "learning_rate": 5.4793862220806114e-05,
+      "loss": 1.888,
+      "step": 6272
+    },
+    {
+      "epoch": 0.6599684376643872,
+      "grad_norm": 1.623795986175537,
+      "learning_rate": 5.4763468281701235e-05,
+      "loss": 1.6365,
+      "step": 6273
+    },
+    {
+      "epoch": 0.6600736454497633,
+      "grad_norm": 1.2633938789367676,
+      "learning_rate": 5.473307959576579e-05,
+      "loss": 1.6566,
+      "step": 6274
+    },
+    {
+      "epoch": 0.6601788532351394,
+      "grad_norm": 1.473612904548645,
+      "learning_rate": 5.470269616652879e-05,
+      "loss": 2.0853,
+      "step": 6275
+    },
+    {
+      "epoch": 0.6602840610205155,
+      "grad_norm": 1.1714433431625366,
+      "learning_rate": 5.467231799751853e-05,
+      "loss": 1.3471,
+      "step": 6276
+    },
+    {
+      "epoch": 0.6603892688058917,
+      "grad_norm": 1.7451331615447998,
+      "learning_rate": 5.464194509226267e-05,
+      "loss": 1.5852,
+      "step": 6277
+    },
+    {
+      "epoch": 0.6604944765912677,
+      "grad_norm": 1.2476874589920044,
+      "learning_rate": 5.461157745428841e-05,
+      "loss": 2.072,
+      "step": 6278
+    },
+    {
+      "epoch": 0.6605996843766438,
+      "grad_norm": 1.997950553894043,
+      "learning_rate": 5.45812150871222e-05,
+      "loss": 1.5509,
+      "step": 6279
+    },
+    {
+      "epoch": 0.66070489216202,
+      "grad_norm": 1.408400058746338,
+      "learning_rate": 5.455085799428992e-05,
+      "loss": 1.9504,
+      "step": 6280
+    },
+    {
+      "epoch": 0.6608100999473961,
+      "grad_norm": 1.8198645114898682,
+      "learning_rate": 5.452050617931683e-05,
+      "loss": 2.1562,
+      "step": 6281
+    },
+    {
+      "epoch": 0.6609153077327722,
+      "grad_norm": 1.5449212789535522,
+      "learning_rate": 5.449015964572758e-05,
+      "loss": 1.5864,
+      "step": 6282
+    },
+    {
+      "epoch": 0.6610205155181483,
+      "grad_norm": 1.971395492553711,
+      "learning_rate": 5.445981839704626e-05,
+      "loss": 1.6987,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6611257233035245,
+      "grad_norm": 1.50814950466156,
+      "learning_rate": 5.4429482436796265e-05,
+      "loss": 1.385,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6612309310889006,
+      "grad_norm": 0.9933773279190063,
+      "learning_rate": 5.439915176850037e-05,
+      "loss": 1.5161,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6613361388742767,
+      "grad_norm": 1.3976936340332031,
+      "learning_rate": 5.4368826395680875e-05,
+      "loss": 1.037,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6614413466596528,
+      "grad_norm": 1.625756859779358,
+      "learning_rate": 5.4338506321859304e-05,
+      "loss": 1.6669,
+      "step": 6287
+    },
+    {
+      "epoch": 0.661546554445029,
+      "grad_norm": 1.8389314413070679,
+      "learning_rate": 5.430819155055659e-05,
+      "loss": 2.227,
+      "step": 6288
+    },
+    {
+      "epoch": 0.661651762230405,
+      "grad_norm": 1.3766262531280518,
+      "learning_rate": 5.427788208529318e-05,
+      "loss": 2.0434,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6617569700157812,
+      "grad_norm": 1.752624750137329,
+      "learning_rate": 5.4247577929588745e-05,
+      "loss": 1.7892,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6618621778011573,
+      "grad_norm": 1.2732161283493042,
+      "learning_rate": 5.4217279086962416e-05,
+      "loss": 1.6846,
+      "step": 6291
+    },
+    {
+      "epoch": 0.6619673855865335,
+      "grad_norm": 1.762704849243164,
+      "learning_rate": 5.418698556093271e-05,
+      "loss": 1.3161,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6620725933719095,
+      "grad_norm": 1.4395411014556885,
+      "learning_rate": 5.41566973550175e-05,
+      "loss": 2.1022,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6621778011572856,
+      "grad_norm": 1.3859562873840332,
+      "learning_rate": 5.41264144727341e-05,
+      "loss": 1.3643,
+      "step": 6294
+    },
+    {
+      "epoch": 0.6622830089426618,
+      "grad_norm": 1.1399462223052979,
+      "learning_rate": 5.409613691759914e-05,
+      "loss": 2.0001,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6623882167280378,
+      "grad_norm": 2.5576746463775635,
+      "learning_rate": 5.406586469312859e-05,
+      "loss": 1.2268,
+      "step": 6296
+    },
+    {
+      "epoch": 0.662493424513414,
+      "grad_norm": 1.959189534187317,
+      "learning_rate": 5.403559780283795e-05,
+      "loss": 0.9835,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6625986322987901,
+      "grad_norm": 1.6175181865692139,
+      "learning_rate": 5.400533625024199e-05,
+      "loss": 1.8782,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6627038400841663,
+      "grad_norm": 1.2457187175750732,
+      "learning_rate": 5.397508003885483e-05,
+      "loss": 1.7866,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6628090478695423,
+      "grad_norm": 1.5392271280288696,
+      "learning_rate": 5.394482917219015e-05,
+      "loss": 1.664,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6629142556549185,
+      "grad_norm": 2.1098384857177734,
+      "learning_rate": 5.391458365376072e-05,
+      "loss": 2.0587,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6630194634402946,
+      "grad_norm": 1.1697083711624146,
+      "learning_rate": 5.3884343487078984e-05,
+      "loss": 2.0644,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6631246712256706,
+      "grad_norm": 1.4275423288345337,
+      "learning_rate": 5.385410867565658e-05,
+      "loss": 2.0324,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6632298790110468,
+      "grad_norm": 1.6790199279785156,
+      "learning_rate": 5.382387922300454e-05,
+      "loss": 1.5068,
+      "step": 6304
+    },
+    {
+      "epoch": 0.663335086796423,
+      "grad_norm": 1.8416249752044678,
+      "learning_rate": 5.379365513263338e-05,
+      "loss": 1.9226,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6634402945817991,
+      "grad_norm": 1.2681195735931396,
+      "learning_rate": 5.3763436408052904e-05,
+      "loss": 2.0517,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6635455023671751,
+      "grad_norm": 1.1384334564208984,
+      "learning_rate": 5.3733223052772265e-05,
+      "loss": 1.3946,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6636507101525513,
+      "grad_norm": 0.976108968257904,
+      "learning_rate": 5.370301507030012e-05,
+      "loss": 1.823,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6637559179379274,
+      "grad_norm": 1.8622645139694214,
+      "learning_rate": 5.367281246414439e-05,
+      "loss": 1.3448,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6638611257233035,
+      "grad_norm": 1.617946982383728,
+      "learning_rate": 5.364261523781234e-05,
+      "loss": 1.6499,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6639663335086796,
+      "grad_norm": 1.8407984972000122,
+      "learning_rate": 5.361242339481078e-05,
+      "loss": 2.258,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6640715412940558,
+      "grad_norm": 1.4248191118240356,
+      "learning_rate": 5.358223693864575e-05,
+      "loss": 1.4035,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6641767490794319,
+      "grad_norm": 1.5487806797027588,
+      "learning_rate": 5.3552055872822636e-05,
+      "loss": 1.6018,
+      "step": 6313
+    },
+    {
+      "epoch": 0.664281956864808,
+      "grad_norm": 1.2339686155319214,
+      "learning_rate": 5.352188020084638e-05,
+      "loss": 2.0545,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6643871646501841,
+      "grad_norm": 1.4714442491531372,
+      "learning_rate": 5.349170992622112e-05,
+      "loss": 1.9142,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6644923724355603,
+      "grad_norm": 1.6540426015853882,
+      "learning_rate": 5.34615450524504e-05,
+      "loss": 1.6512,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6645975802209364,
+      "grad_norm": 1.5326392650604248,
+      "learning_rate": 5.3431385583037244e-05,
+      "loss": 1.5426,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6647027880063124,
+      "grad_norm": 1.2688010931015015,
+      "learning_rate": 5.340123152148393e-05,
+      "loss": 1.5912,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6648079957916886,
+      "grad_norm": 1.2774553298950195,
+      "learning_rate": 5.337108287129211e-05,
+      "loss": 1.7242,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6649132035770647,
+      "grad_norm": 1.8071551322937012,
+      "learning_rate": 5.334093963596294e-05,
+      "loss": 1.8998,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6650184113624408,
+      "grad_norm": 1.0663738250732422,
+      "learning_rate": 5.33108018189968e-05,
+      "loss": 1.7188,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6651236191478169,
+      "grad_norm": 1.3195592164993286,
+      "learning_rate": 5.328066942389351e-05,
+      "loss": 1.8181,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6652288269331931,
+      "grad_norm": 1.5998344421386719,
+      "learning_rate": 5.325054245415223e-05,
+      "loss": 1.454,
+      "step": 6323
+    },
+    {
+      "epoch": 0.6653340347185692,
+      "grad_norm": 1.0443423986434937,
+      "learning_rate": 5.322042091327148e-05,
+      "loss": 1.8442,
+      "step": 6324
+    },
+    {
+      "epoch": 0.6654392425039453,
+      "grad_norm": 1.2177835702896118,
+      "learning_rate": 5.319030480474923e-05,
+      "loss": 1.6729,
+      "step": 6325
+    },
+    {
+      "epoch": 0.6655444502893214,
+      "grad_norm": 2.082824945449829,
+      "learning_rate": 5.316019413208275e-05,
+      "loss": 1.8512,
+      "step": 6326
+    },
+    {
+      "epoch": 0.6656496580746976,
+      "grad_norm": 1.6515694856643677,
+      "learning_rate": 5.313008889876865e-05,
+      "loss": 1.9645,
+      "step": 6327
+    },
+    {
+      "epoch": 0.6657548658600736,
+      "grad_norm": 1.6280758380889893,
+      "learning_rate": 5.309998910830303e-05,
+      "loss": 2.4568,
+      "step": 6328
+    },
+    {
+      "epoch": 0.6658600736454497,
+      "grad_norm": 1.4808907508850098,
+      "learning_rate": 5.306989476418123e-05,
+      "loss": 1.864,
+      "step": 6329
+    },
+    {
+      "epoch": 0.6659652814308259,
+      "grad_norm": 1.735153079032898,
+      "learning_rate": 5.3039805869897985e-05,
+      "loss": 2.0455,
+      "step": 6330
+    },
+    {
+      "epoch": 0.666070489216202,
+      "grad_norm": 1.4877818822860718,
+      "learning_rate": 5.3009722428947475e-05,
+      "loss": 1.5956,
+      "step": 6331
+    },
+    {
+      "epoch": 0.6661756970015781,
+      "grad_norm": 1.3403220176696777,
+      "learning_rate": 5.297964444482317e-05,
+      "loss": 1.806,
+      "step": 6332
+    },
+    {
+      "epoch": 0.6662809047869542,
+      "grad_norm": 1.4084241390228271,
+      "learning_rate": 5.294957192101788e-05,
+      "loss": 1.4097,
+      "step": 6333
+    },
+    {
+      "epoch": 0.6663861125723304,
+      "grad_norm": 1.148952603340149,
+      "learning_rate": 5.2919504861023903e-05,
+      "loss": 2.0219,
+      "step": 6334
+    },
+    {
+      "epoch": 0.6664913203577064,
+      "grad_norm": 1.5325878858566284,
+      "learning_rate": 5.288944326833281e-05,
+      "loss": 2.2054,
+      "step": 6335
+    },
+    {
+      "epoch": 0.6665965281430826,
+      "grad_norm": 1.375962495803833,
+      "learning_rate": 5.285938714643548e-05,
+      "loss": 1.1267,
+      "step": 6336
+    },
+    {
+      "epoch": 0.6667017359284587,
+      "grad_norm": 1.8014800548553467,
+      "learning_rate": 5.2829336498822335e-05,
+      "loss": 1.6149,
+      "step": 6337
+    },
+    {
+      "epoch": 0.6668069437138349,
+      "grad_norm": 1.3949754238128662,
+      "learning_rate": 5.279929132898298e-05,
+      "loss": 1.2513,
+      "step": 6338
+    },
+    {
+      "epoch": 0.6669121514992109,
+      "grad_norm": 1.4356800317764282,
+      "learning_rate": 5.276925164040653e-05,
+      "loss": 1.763,
+      "step": 6339
+    },
+    {
+      "epoch": 0.667017359284587,
+      "grad_norm": 1.1746344566345215,
+      "learning_rate": 5.2739217436581365e-05,
+      "loss": 1.503,
+      "step": 6340
+    },
+    {
+      "epoch": 0.6671225670699632,
+      "grad_norm": 2.2012033462524414,
+      "learning_rate": 5.270918872099522e-05,
+      "loss": 2.013,
+      "step": 6341
+    },
+    {
+      "epoch": 0.6672277748553394,
+      "grad_norm": 2.013820171356201,
+      "learning_rate": 5.2679165497135285e-05,
+      "loss": 1.8092,
+      "step": 6342
+    },
+    {
+      "epoch": 0.6673329826407154,
+      "grad_norm": 1.2771358489990234,
+      "learning_rate": 5.264914776848808e-05,
+      "loss": 1.6848,
+      "step": 6343
+    },
+    {
+      "epoch": 0.6674381904260915,
+      "grad_norm": 1.3562238216400146,
+      "learning_rate": 5.2619135538539355e-05,
+      "loss": 1.7494,
+      "step": 6344
+    },
+    {
+      "epoch": 0.6675433982114677,
+      "grad_norm": 2.059318780899048,
+      "learning_rate": 5.2589128810774426e-05,
+      "loss": 1.9948,
+      "step": 6345
+    },
+    {
+      "epoch": 0.6676486059968437,
+      "grad_norm": 1.2015169858932495,
+      "learning_rate": 5.2559127588677846e-05,
+      "loss": 1.5834,
+      "step": 6346
+    },
+    {
+      "epoch": 0.6677538137822199,
+      "grad_norm": 1.4178433418273926,
+      "learning_rate": 5.252913187573354e-05,
+      "loss": 1.76,
+      "step": 6347
+    },
+    {
+      "epoch": 0.667859021567596,
+      "grad_norm": 1.4683200120925903,
+      "learning_rate": 5.249914167542486e-05,
+      "loss": 1.7822,
+      "step": 6348
+    },
+    {
+      "epoch": 0.6679642293529722,
+      "grad_norm": 1.260953664779663,
+      "learning_rate": 5.246915699123439e-05,
+      "loss": 0.8526,
+      "step": 6349
+    },
+    {
+      "epoch": 0.6680694371383482,
+      "grad_norm": 1.269636869430542,
+      "learning_rate": 5.243917782664425e-05,
+      "loss": 1.7709,
+      "step": 6350
+    },
+    {
+      "epoch": 0.6681746449237244,
+      "grad_norm": 1.585425615310669,
+      "learning_rate": 5.240920418513577e-05,
+      "loss": 1.542,
+      "step": 6351
+    },
+    {
+      "epoch": 0.6682798527091005,
+      "grad_norm": 2.510784149169922,
+      "learning_rate": 5.2379236070189677e-05,
+      "loss": 1.4086,
+      "step": 6352
+    },
+    {
+      "epoch": 0.6683850604944765,
+      "grad_norm": 1.2472009658813477,
+      "learning_rate": 5.234927348528611e-05,
+      "loss": 1.7158,
+      "step": 6353
+    },
+    {
+      "epoch": 0.6684902682798527,
+      "grad_norm": 1.3211387395858765,
+      "learning_rate": 5.231931643390451e-05,
+      "loss": 1.8004,
+      "step": 6354
+    },
+    {
+      "epoch": 0.6685954760652288,
+      "grad_norm": 1.3025023937225342,
+      "learning_rate": 5.228936491952363e-05,
+      "loss": 2.2033,
+      "step": 6355
+    },
+    {
+      "epoch": 0.668700683850605,
+      "grad_norm": 2.7773120403289795,
+      "learning_rate": 5.2259418945621754e-05,
+      "loss": 2.0115,
+      "step": 6356
+    },
+    {
+      "epoch": 0.668805891635981,
+      "grad_norm": 1.2266188859939575,
+      "learning_rate": 5.222947851567633e-05,
+      "loss": 1.3392,
+      "step": 6357
+    },
+    {
+      "epoch": 0.6689110994213572,
+      "grad_norm": 1.2715981006622314,
+      "learning_rate": 5.219954363316424e-05,
+      "loss": 1.6663,
+      "step": 6358
+    },
+    {
+      "epoch": 0.6690163072067333,
+      "grad_norm": 2.1951401233673096,
+      "learning_rate": 5.2169614301561775e-05,
+      "loss": 1.6716,
+      "step": 6359
+    },
+    {
+      "epoch": 0.6691215149921094,
+      "grad_norm": 1.2387498617172241,
+      "learning_rate": 5.2139690524344495e-05,
+      "loss": 1.6217,
+      "step": 6360
+    },
+    {
+      "epoch": 0.6692267227774855,
+      "grad_norm": 1.4151759147644043,
+      "learning_rate": 5.210977230498733e-05,
+      "loss": 2.2641,
+      "step": 6361
+    },
+    {
+      "epoch": 0.6693319305628617,
+      "grad_norm": 1.9566744565963745,
+      "learning_rate": 5.207985964696462e-05,
+      "loss": 1.018,
+      "step": 6362
+    },
+    {
+      "epoch": 0.6694371383482378,
+      "grad_norm": 1.473006010055542,
+      "learning_rate": 5.2049952553750046e-05,
+      "loss": 1.7412,
+      "step": 6363
+    },
+    {
+      "epoch": 0.6695423461336139,
+      "grad_norm": 1.857047438621521,
+      "learning_rate": 5.202005102881653e-05,
+      "loss": 1.3064,
+      "step": 6364
+    },
+    {
+      "epoch": 0.66964755391899,
+      "grad_norm": 1.2658127546310425,
+      "learning_rate": 5.199015507563656e-05,
+      "loss": 1.7702,
+      "step": 6365
+    },
+    {
+      "epoch": 0.6697527617043662,
+      "grad_norm": 1.1785234212875366,
+      "learning_rate": 5.1960264697681726e-05,
+      "loss": 2.3375,
+      "step": 6366
+    },
+    {
+      "epoch": 0.6698579694897422,
+      "grad_norm": 1.1640081405639648,
+      "learning_rate": 5.19303798984232e-05,
+      "loss": 2.1538,
+      "step": 6367
+    },
+    {
+      "epoch": 0.6699631772751183,
+      "grad_norm": 1.3098970651626587,
+      "learning_rate": 5.1900500681331363e-05,
+      "loss": 2.0698,
+      "step": 6368
+    },
+    {
+      "epoch": 0.6700683850604945,
+      "grad_norm": 1.9889769554138184,
+      "learning_rate": 5.1870627049875954e-05,
+      "loss": 1.8563,
+      "step": 6369
+    },
+    {
+      "epoch": 0.6701735928458706,
+      "grad_norm": 1.559300184249878,
+      "learning_rate": 5.184075900752619e-05,
+      "loss": 1.4302,
+      "step": 6370
+    },
+    {
+      "epoch": 0.6702788006312467,
+      "grad_norm": 1.8356385231018066,
+      "learning_rate": 5.1810896557750485e-05,
+      "loss": 1.4985,
+      "step": 6371
+    },
+    {
+      "epoch": 0.6703840084166228,
+      "grad_norm": 1.3510042428970337,
+      "learning_rate": 5.178103970401664e-05,
+      "loss": 1.2977,
+      "step": 6372
+    },
+    {
+      "epoch": 0.670489216201999,
+      "grad_norm": 1.291181206703186,
+      "learning_rate": 5.1751188449791924e-05,
+      "loss": 1.4883,
+      "step": 6373
+    },
+    {
+      "epoch": 0.6705944239873751,
+      "grad_norm": 1.7430511713027954,
+      "learning_rate": 5.1721342798542795e-05,
+      "loss": 1.609,
+      "step": 6374
+    },
+    {
+      "epoch": 0.6706996317727512,
+      "grad_norm": 1.0107218027114868,
+      "learning_rate": 5.169150275373513e-05,
+      "loss": 1.5741,
+      "step": 6375
+    },
+    {
+      "epoch": 0.6708048395581273,
+      "grad_norm": 1.1406487226486206,
+      "learning_rate": 5.16616683188342e-05,
+      "loss": 1.1794,
+      "step": 6376
+    },
+    {
+      "epoch": 0.6709100473435035,
+      "grad_norm": 1.4461215734481812,
+      "learning_rate": 5.163183949730456e-05,
+      "loss": 1.4193,
+      "step": 6377
+    },
+    {
+      "epoch": 0.6710152551288795,
+      "grad_norm": 1.2850546836853027,
+      "learning_rate": 5.1602016292610075e-05,
+      "loss": 1.5627,
+      "step": 6378
+    },
+    {
+      "epoch": 0.6711204629142556,
+      "grad_norm": 1.6368341445922852,
+      "learning_rate": 5.157219870821413e-05,
+      "loss": 1.6965,
+      "step": 6379
+    },
+    {
+      "epoch": 0.6712256706996318,
+      "grad_norm": 1.221516489982605,
+      "learning_rate": 5.154238674757925e-05,
+      "loss": 1.3989,
+      "step": 6380
+    },
+    {
+      "epoch": 0.6713308784850079,
+      "grad_norm": 1.8983935117721558,
+      "learning_rate": 5.151258041416742e-05,
+      "loss": 2.2616,
+      "step": 6381
+    },
+    {
+      "epoch": 0.671436086270384,
+      "grad_norm": 1.487594485282898,
+      "learning_rate": 5.148277971143998e-05,
+      "loss": 2.0648,
+      "step": 6382
+    },
+    {
+      "epoch": 0.6715412940557601,
+      "grad_norm": 1.1469671726226807,
+      "learning_rate": 5.145298464285757e-05,
+      "loss": 1.7304,
+      "step": 6383
+    },
+    {
+      "epoch": 0.6716465018411363,
+      "grad_norm": 1.3301132917404175,
+      "learning_rate": 5.142319521188017e-05,
+      "loss": 1.673,
+      "step": 6384
+    },
+    {
+      "epoch": 0.6717517096265123,
+      "grad_norm": 1.2145934104919434,
+      "learning_rate": 5.1393411421967174e-05,
+      "loss": 1.6685,
+      "step": 6385
+    },
+    {
+      "epoch": 0.6718569174118885,
+      "grad_norm": 1.669686198234558,
+      "learning_rate": 5.136363327657725e-05,
+      "loss": 1.9549,
+      "step": 6386
+    },
+    {
+      "epoch": 0.6719621251972646,
+      "grad_norm": 1.2095863819122314,
+      "learning_rate": 5.1333860779168455e-05,
+      "loss": 1.6674,
+      "step": 6387
+    },
+    {
+      "epoch": 0.6720673329826408,
+      "grad_norm": 1.208296537399292,
+      "learning_rate": 5.1304093933198136e-05,
+      "loss": 1.8812,
+      "step": 6388
+    },
+    {
+      "epoch": 0.6721725407680168,
+      "grad_norm": 1.2008256912231445,
+      "learning_rate": 5.127433274212301e-05,
+      "loss": 1.8069,
+      "step": 6389
+    },
+    {
+      "epoch": 0.672277748553393,
+      "grad_norm": 1.6719691753387451,
+      "learning_rate": 5.12445772093992e-05,
+      "loss": 1.8409,
+      "step": 6390
+    },
+    {
+      "epoch": 0.6723829563387691,
+      "grad_norm": 1.1335160732269287,
+      "learning_rate": 5.1214827338482094e-05,
+      "loss": 1.9089,
+      "step": 6391
+    },
+    {
+      "epoch": 0.6724881641241451,
+      "grad_norm": 1.1994752883911133,
+      "learning_rate": 5.1185083132826414e-05,
+      "loss": 1.6759,
+      "step": 6392
+    },
+    {
+      "epoch": 0.6725933719095213,
+      "grad_norm": 1.675121784210205,
+      "learning_rate": 5.115534459588631e-05,
+      "loss": 1.1881,
+      "step": 6393
+    },
+    {
+      "epoch": 0.6726985796948974,
+      "grad_norm": 2.1219699382781982,
+      "learning_rate": 5.1125611731115174e-05,
+      "loss": 1.4658,
+      "step": 6394
+    },
+    {
+      "epoch": 0.6728037874802736,
+      "grad_norm": 1.6067357063293457,
+      "learning_rate": 5.1095884541965835e-05,
+      "loss": 1.4778,
+      "step": 6395
+    },
+    {
+      "epoch": 0.6729089952656496,
+      "grad_norm": 1.3906060457229614,
+      "learning_rate": 5.106616303189039e-05,
+      "loss": 1.727,
+      "step": 6396
+    },
+    {
+      "epoch": 0.6730142030510258,
+      "grad_norm": 1.3164790868759155,
+      "learning_rate": 5.103644720434027e-05,
+      "loss": 1.8831,
+      "step": 6397
+    },
+    {
+      "epoch": 0.6731194108364019,
+      "grad_norm": 2.2263622283935547,
+      "learning_rate": 5.100673706276633e-05,
+      "loss": 1.7309,
+      "step": 6398
+    },
+    {
+      "epoch": 0.673224618621778,
+      "grad_norm": 2.4222521781921387,
+      "learning_rate": 5.097703261061868e-05,
+      "loss": 1.2371,
+      "step": 6399
+    },
+    {
+      "epoch": 0.6733298264071541,
+      "grad_norm": 1.8905748128890991,
+      "learning_rate": 5.094733385134677e-05,
+      "loss": 1.5675,
+      "step": 6400
+    },
+    {
+      "epoch": 0.6734350341925303,
+      "grad_norm": 1.5610456466674805,
+      "learning_rate": 5.091764078839949e-05,
+      "loss": 1.4919,
+      "step": 6401
+    },
+    {
+      "epoch": 0.6735402419779064,
+      "grad_norm": 1.896572232246399,
+      "learning_rate": 5.088795342522497e-05,
+      "loss": 1.4104,
+      "step": 6402
+    },
+    {
+      "epoch": 0.6736454497632824,
+      "grad_norm": 1.2327032089233398,
+      "learning_rate": 5.085827176527064e-05,
+      "loss": 1.3777,
+      "step": 6403
+    },
+    {
+      "epoch": 0.6737506575486586,
+      "grad_norm": 1.7423434257507324,
+      "learning_rate": 5.082859581198344e-05,
+      "loss": 1.6206,
+      "step": 6404
+    },
+    {
+      "epoch": 0.6738558653340347,
+      "grad_norm": 1.2722043991088867,
+      "learning_rate": 5.0798925568809486e-05,
+      "loss": 1.4621,
+      "step": 6405
+    },
+    {
+      "epoch": 0.6739610731194109,
+      "grad_norm": 1.320483922958374,
+      "learning_rate": 5.076926103919426e-05,
+      "loss": 2.0733,
+      "step": 6406
+    },
+    {
+      "epoch": 0.6740662809047869,
+      "grad_norm": 1.2347058057785034,
+      "learning_rate": 5.0739602226582706e-05,
+      "loss": 1.6421,
+      "step": 6407
+    },
+    {
+      "epoch": 0.6741714886901631,
+      "grad_norm": 1.6619998216629028,
+      "learning_rate": 5.0709949134418865e-05,
+      "loss": 2.2148,
+      "step": 6408
+    },
+    {
+      "epoch": 0.6742766964755392,
+      "grad_norm": 2.012146234512329,
+      "learning_rate": 5.0680301766146355e-05,
+      "loss": 1.6082,
+      "step": 6409
+    },
+    {
+      "epoch": 0.6743819042609153,
+      "grad_norm": 1.2811518907546997,
+      "learning_rate": 5.0650660125207994e-05,
+      "loss": 1.6369,
+      "step": 6410
+    },
+    {
+      "epoch": 0.6744871120462914,
+      "grad_norm": 1.504457712173462,
+      "learning_rate": 5.062102421504593e-05,
+      "loss": 1.9906,
+      "step": 6411
+    },
+    {
+      "epoch": 0.6745923198316676,
+      "grad_norm": 1.718910574913025,
+      "learning_rate": 5.059139403910177e-05,
+      "loss": 1.5476,
+      "step": 6412
+    },
+    {
+      "epoch": 0.6746975276170437,
+      "grad_norm": 1.8171217441558838,
+      "learning_rate": 5.056176960081631e-05,
+      "loss": 1.605,
+      "step": 6413
+    },
+    {
+      "epoch": 0.6748027354024198,
+      "grad_norm": 1.2599389553070068,
+      "learning_rate": 5.0532150903629724e-05,
+      "loss": 1.892,
+      "step": 6414
+    },
+    {
+      "epoch": 0.6749079431877959,
+      "grad_norm": 1.823708415031433,
+      "learning_rate": 5.050253795098159e-05,
+      "loss": 1.8375,
+      "step": 6415
+    },
+    {
+      "epoch": 0.675013150973172,
+      "grad_norm": 1.2546007633209229,
+      "learning_rate": 5.047293074631074e-05,
+      "loss": 1.4461,
+      "step": 6416
+    },
+    {
+      "epoch": 0.6751183587585481,
+      "grad_norm": 1.1789088249206543,
+      "learning_rate": 5.0443329293055305e-05,
+      "loss": 1.7132,
+      "step": 6417
+    },
+    {
+      "epoch": 0.6752235665439242,
+      "grad_norm": 1.3993732929229736,
+      "learning_rate": 5.041373359465289e-05,
+      "loss": 1.5287,
+      "step": 6418
+    },
+    {
+      "epoch": 0.6753287743293004,
+      "grad_norm": 1.6638224124908447,
+      "learning_rate": 5.0384143654540314e-05,
+      "loss": 1.1025,
+      "step": 6419
+    },
+    {
+      "epoch": 0.6754339821146765,
+      "grad_norm": 1.2212295532226562,
+      "learning_rate": 5.035455947615373e-05,
+      "loss": 2.0747,
+      "step": 6420
+    },
+    {
+      "epoch": 0.6755391899000526,
+      "grad_norm": 1.5106666088104248,
+      "learning_rate": 5.032498106292869e-05,
+      "loss": 1.3987,
+      "step": 6421
+    },
+    {
+      "epoch": 0.6756443976854287,
+      "grad_norm": 1.0953783988952637,
+      "learning_rate": 5.029540841830004e-05,
+      "loss": 1.4668,
+      "step": 6422
+    },
+    {
+      "epoch": 0.6757496054708049,
+      "grad_norm": 1.4428025484085083,
+      "learning_rate": 5.0265841545701886e-05,
+      "loss": 1.8213,
+      "step": 6423
+    },
+    {
+      "epoch": 0.6758548132561809,
+      "grad_norm": 1.1565446853637695,
+      "learning_rate": 5.023628044856783e-05,
+      "loss": 1.9793,
+      "step": 6424
+    },
+    {
+      "epoch": 0.6759600210415571,
+      "grad_norm": 1.158007025718689,
+      "learning_rate": 5.020672513033066e-05,
+      "loss": 1.5613,
+      "step": 6425
+    },
+    {
+      "epoch": 0.6760652288269332,
+      "grad_norm": 1.8131704330444336,
+      "learning_rate": 5.017717559442249e-05,
+      "loss": 2.1706,
+      "step": 6426
+    },
+    {
+      "epoch": 0.6761704366123094,
+      "grad_norm": 1.257173776626587,
+      "learning_rate": 5.014763184427489e-05,
+      "loss": 1.7876,
+      "step": 6427
+    },
+    {
+      "epoch": 0.6762756443976854,
+      "grad_norm": 1.4548490047454834,
+      "learning_rate": 5.011809388331865e-05,
+      "loss": 2.1786,
+      "step": 6428
+    },
+    {
+      "epoch": 0.6763808521830615,
+      "grad_norm": 1.3961397409439087,
+      "learning_rate": 5.0088561714983906e-05,
+      "loss": 1.6834,
+      "step": 6429
+    },
+    {
+      "epoch": 0.6764860599684377,
+      "grad_norm": 1.5321639776229858,
+      "learning_rate": 5.0059035342700144e-05,
+      "loss": 1.0357,
+      "step": 6430
+    },
+    {
+      "epoch": 0.6765912677538137,
+      "grad_norm": 1.6330370903015137,
+      "learning_rate": 5.0029514769896114e-05,
+      "loss": 1.6928,
+      "step": 6431
+    },
+    {
+      "epoch": 0.6766964755391899,
+      "grad_norm": 1.063062310218811,
+      "learning_rate": 5.000000000000002e-05,
+      "loss": 1.5404,
+      "step": 6432
+    },
+    {
+      "epoch": 0.676801683324566,
+      "grad_norm": 0.9866712093353271,
+      "learning_rate": 4.9970491036439284e-05,
+      "loss": 1.7373,
+      "step": 6433
+    },
+    {
+      "epoch": 0.6769068911099422,
+      "grad_norm": 1.9362293481826782,
+      "learning_rate": 4.9940987882640647e-05,
+      "loss": 2.12,
+      "step": 6434
+    },
+    {
+      "epoch": 0.6770120988953182,
+      "grad_norm": 1.8418405055999756,
+      "learning_rate": 4.991149054203027e-05,
+      "loss": 1.6442,
+      "step": 6435
+    },
+    {
+      "epoch": 0.6771173066806944,
+      "grad_norm": 1.7248798608779907,
+      "learning_rate": 4.988199901803357e-05,
+      "loss": 1.6048,
+      "step": 6436
+    },
+    {
+      "epoch": 0.6772225144660705,
+      "grad_norm": 1.2888562679290771,
+      "learning_rate": 4.985251331407524e-05,
+      "loss": 1.7667,
+      "step": 6437
+    },
+    {
+      "epoch": 0.6773277222514467,
+      "grad_norm": 1.4221709966659546,
+      "learning_rate": 4.982303343357946e-05,
+      "loss": 1.7982,
+      "step": 6438
+    },
+    {
+      "epoch": 0.6774329300368227,
+      "grad_norm": 1.8176785707473755,
+      "learning_rate": 4.9793559379969566e-05,
+      "loss": 1.9093,
+      "step": 6439
+    },
+    {
+      "epoch": 0.6775381378221988,
+      "grad_norm": 1.803604245185852,
+      "learning_rate": 4.9764091156668266e-05,
+      "loss": 1.9377,
+      "step": 6440
+    },
+    {
+      "epoch": 0.677643345607575,
+      "grad_norm": 1.8046791553497314,
+      "learning_rate": 4.973462876709767e-05,
+      "loss": 1.357,
+      "step": 6441
+    },
+    {
+      "epoch": 0.677748553392951,
+      "grad_norm": 2.0363752841949463,
+      "learning_rate": 4.970517221467909e-05,
+      "loss": 1.4152,
+      "step": 6442
+    },
+    {
+      "epoch": 0.6778537611783272,
+      "grad_norm": 1.5641330480575562,
+      "learning_rate": 4.967572150283326e-05,
+      "loss": 1.8567,
+      "step": 6443
+    },
+    {
+      "epoch": 0.6779589689637033,
+      "grad_norm": 1.910548448562622,
+      "learning_rate": 4.9646276634980194e-05,
+      "loss": 2.0569,
+      "step": 6444
+    },
+    {
+      "epoch": 0.6780641767490795,
+      "grad_norm": 1.1518720388412476,
+      "learning_rate": 4.961683761453917e-05,
+      "loss": 1.8144,
+      "step": 6445
+    },
+    {
+      "epoch": 0.6781693845344555,
+      "grad_norm": 2.0624051094055176,
+      "learning_rate": 4.958740444492892e-05,
+      "loss": 1.5392,
+      "step": 6446
+    },
+    {
+      "epoch": 0.6782745923198317,
+      "grad_norm": 1.4468212127685547,
+      "learning_rate": 4.955797712956739e-05,
+      "loss": 1.4592,
+      "step": 6447
+    },
+    {
+      "epoch": 0.6783798001052078,
+      "grad_norm": 1.4319233894348145,
+      "learning_rate": 4.9528555671871835e-05,
+      "loss": 1.6202,
+      "step": 6448
+    },
+    {
+      "epoch": 0.6784850078905839,
+      "grad_norm": 2.070284605026245,
+      "learning_rate": 4.9499140075258957e-05,
+      "loss": 1.5695,
+      "step": 6449
+    },
+    {
+      "epoch": 0.67859021567596,
+      "grad_norm": 1.7108412981033325,
+      "learning_rate": 4.9469730343144635e-05,
+      "loss": 1.3484,
+      "step": 6450
+    },
+    {
+      "epoch": 0.6786954234613362,
+      "grad_norm": 1.4341391324996948,
+      "learning_rate": 4.944032647894414e-05,
+      "loss": 1.2341,
+      "step": 6451
+    },
+    {
+      "epoch": 0.6788006312467123,
+      "grad_norm": 1.3790180683135986,
+      "learning_rate": 4.941092848607204e-05,
+      "loss": 1.4654,
+      "step": 6452
+    },
+    {
+      "epoch": 0.6789058390320883,
+      "grad_norm": 2.0244486331939697,
+      "learning_rate": 4.9381536367942195e-05,
+      "loss": 1.7548,
+      "step": 6453
+    },
+    {
+      "epoch": 0.6790110468174645,
+      "grad_norm": 1.3763808012008667,
+      "learning_rate": 4.935215012796789e-05,
+      "loss": 1.2737,
+      "step": 6454
+    },
+    {
+      "epoch": 0.6791162546028406,
+      "grad_norm": 1.4411334991455078,
+      "learning_rate": 4.93227697695616e-05,
+      "loss": 1.1812,
+      "step": 6455
+    },
+    {
+      "epoch": 0.6792214623882167,
+      "grad_norm": 1.3906333446502686,
+      "learning_rate": 4.929339529613515e-05,
+      "loss": 1.8517,
+      "step": 6456
+    },
+    {
+      "epoch": 0.6793266701735928,
+      "grad_norm": 1.6477974653244019,
+      "learning_rate": 4.9264026711099764e-05,
+      "loss": 1.6151,
+      "step": 6457
+    },
+    {
+      "epoch": 0.679431877958969,
+      "grad_norm": 1.5431394577026367,
+      "learning_rate": 4.9234664017865896e-05,
+      "loss": 2.044,
+      "step": 6458
+    },
+    {
+      "epoch": 0.6795370857443451,
+      "grad_norm": 1.9720886945724487,
+      "learning_rate": 4.920530721984329e-05,
+      "loss": 1.6605,
+      "step": 6459
+    },
+    {
+      "epoch": 0.6796422935297212,
+      "grad_norm": 1.2280595302581787,
+      "learning_rate": 4.917595632044113e-05,
+      "loss": 1.6364,
+      "step": 6460
+    },
+    {
+      "epoch": 0.6797475013150973,
+      "grad_norm": 1.3215405941009521,
+      "learning_rate": 4.914661132306779e-05,
+      "loss": 1.275,
+      "step": 6461
+    },
+    {
+      "epoch": 0.6798527091004735,
+      "grad_norm": 1.446702241897583,
+      "learning_rate": 4.911727223113099e-05,
+      "loss": 2.1152,
+      "step": 6462
+    },
+    {
+      "epoch": 0.6799579168858495,
+      "grad_norm": 1.3523727655410767,
+      "learning_rate": 4.908793904803787e-05,
+      "loss": 1.2915,
+      "step": 6463
+    },
+    {
+      "epoch": 0.6800631246712256,
+      "grad_norm": 1.8233401775360107,
+      "learning_rate": 4.9058611777194716e-05,
+      "loss": 1.0423,
+      "step": 6464
+    },
+    {
+      "epoch": 0.6801683324566018,
+      "grad_norm": 1.9023778438568115,
+      "learning_rate": 4.9029290422007204e-05,
+      "loss": 1.5998,
+      "step": 6465
+    },
+    {
+      "epoch": 0.680273540241978,
+      "grad_norm": 1.1877436637878418,
+      "learning_rate": 4.8999974985880384e-05,
+      "loss": 1.9294,
+      "step": 6466
+    },
+    {
+      "epoch": 0.680378748027354,
+      "grad_norm": 1.5044838190078735,
+      "learning_rate": 4.8970665472218537e-05,
+      "loss": 1.5298,
+      "step": 6467
+    },
+    {
+      "epoch": 0.6804839558127301,
+      "grad_norm": 1.9050568342208862,
+      "learning_rate": 4.8941361884425215e-05,
+      "loss": 1.5252,
+      "step": 6468
+    },
+    {
+      "epoch": 0.6805891635981063,
+      "grad_norm": 1.1263233423233032,
+      "learning_rate": 4.891206422590347e-05,
+      "loss": 1.6088,
+      "step": 6469
+    },
+    {
+      "epoch": 0.6806943713834824,
+      "grad_norm": 1.9254624843597412,
+      "learning_rate": 4.8882772500055464e-05,
+      "loss": 2.1593,
+      "step": 6470
+    },
+    {
+      "epoch": 0.6807995791688585,
+      "grad_norm": 2.4042844772338867,
+      "learning_rate": 4.885348671028273e-05,
+      "loss": 1.6122,
+      "step": 6471
+    },
+    {
+      "epoch": 0.6809047869542346,
+      "grad_norm": 1.4427109956741333,
+      "learning_rate": 4.882420685998623e-05,
+      "loss": 1.304,
+      "step": 6472
+    },
+    {
+      "epoch": 0.6810099947396108,
+      "grad_norm": 1.020768404006958,
+      "learning_rate": 4.8794932952566e-05,
+      "loss": 1.9123,
+      "step": 6473
+    },
+    {
+      "epoch": 0.6811152025249868,
+      "grad_norm": 1.6786848306655884,
+      "learning_rate": 4.8765664991421634e-05,
+      "loss": 1.9933,
+      "step": 6474
+    },
+    {
+      "epoch": 0.681220410310363,
+      "grad_norm": 1.6853234767913818,
+      "learning_rate": 4.8736402979951867e-05,
+      "loss": 1.9755,
+      "step": 6475
+    },
+    {
+      "epoch": 0.6813256180957391,
+      "grad_norm": 1.9556273221969604,
+      "learning_rate": 4.870714692155479e-05,
+      "loss": 1.5554,
+      "step": 6476
+    },
+    {
+      "epoch": 0.6814308258811153,
+      "grad_norm": 1.1733824014663696,
+      "learning_rate": 4.867789681962788e-05,
+      "loss": 1.7569,
+      "step": 6477
+    },
+    {
+      "epoch": 0.6815360336664913,
+      "grad_norm": 1.0935428142547607,
+      "learning_rate": 4.864865267756779e-05,
+      "loss": 1.993,
+      "step": 6478
+    },
+    {
+      "epoch": 0.6816412414518674,
+      "grad_norm": 1.0769548416137695,
+      "learning_rate": 4.8619414498770556e-05,
+      "loss": 1.8029,
+      "step": 6479
+    },
+    {
+      "epoch": 0.6817464492372436,
+      "grad_norm": 1.2517523765563965,
+      "learning_rate": 4.859018228663155e-05,
+      "loss": 2.1288,
+      "step": 6480
+    },
+    {
+      "epoch": 0.6818516570226196,
+      "grad_norm": 1.2191598415374756,
+      "learning_rate": 4.856095604454539e-05,
+      "loss": 1.2368,
+      "step": 6481
+    },
+    {
+      "epoch": 0.6819568648079958,
+      "grad_norm": 1.3996628522872925,
+      "learning_rate": 4.8531735775905975e-05,
+      "loss": 1.7399,
+      "step": 6482
+    },
+    {
+      "epoch": 0.6820620725933719,
+      "grad_norm": 1.9159706830978394,
+      "learning_rate": 4.850252148410665e-05,
+      "loss": 1.5246,
+      "step": 6483
+    },
+    {
+      "epoch": 0.6821672803787481,
+      "grad_norm": 1.7578186988830566,
+      "learning_rate": 4.8473313172539925e-05,
+      "loss": 1.8513,
+      "step": 6484
+    },
+    {
+      "epoch": 0.6822724881641241,
+      "grad_norm": 1.3070460557937622,
+      "learning_rate": 4.8444110844597626e-05,
+      "loss": 1.6536,
+      "step": 6485
+    },
+    {
+      "epoch": 0.6823776959495003,
+      "grad_norm": 2.009951591491699,
+      "learning_rate": 4.8414914503671006e-05,
+      "loss": 1.1448,
+      "step": 6486
+    },
+    {
+      "epoch": 0.6824829037348764,
+      "grad_norm": 2.186450719833374,
+      "learning_rate": 4.838572415315046e-05,
+      "loss": 1.8157,
+      "step": 6487
+    },
+    {
+      "epoch": 0.6825881115202524,
+      "grad_norm": 1.4713935852050781,
+      "learning_rate": 4.835653979642585e-05,
+      "loss": 1.0556,
+      "step": 6488
+    },
+    {
+      "epoch": 0.6826933193056286,
+      "grad_norm": 1.4540444612503052,
+      "learning_rate": 4.832736143688621e-05,
+      "loss": 0.9736,
+      "step": 6489
+    },
+    {
+      "epoch": 0.6827985270910047,
+      "grad_norm": 1.8345993757247925,
+      "learning_rate": 4.829818907791988e-05,
+      "loss": 1.4606,
+      "step": 6490
+    },
+    {
+      "epoch": 0.6829037348763809,
+      "grad_norm": 1.7492965459823608,
+      "learning_rate": 4.826902272291467e-05,
+      "loss": 2.0597,
+      "step": 6491
+    },
+    {
+      "epoch": 0.6830089426617569,
+      "grad_norm": 1.3600223064422607,
+      "learning_rate": 4.8239862375257484e-05,
+      "loss": 1.4011,
+      "step": 6492
+    },
+    {
+      "epoch": 0.6831141504471331,
+      "grad_norm": 1.2309643030166626,
+      "learning_rate": 4.821070803833464e-05,
+      "loss": 1.3888,
+      "step": 6493
+    },
+    {
+      "epoch": 0.6832193582325092,
+      "grad_norm": 1.445009708404541,
+      "learning_rate": 4.818155971553174e-05,
+      "loss": 2.1089,
+      "step": 6494
+    },
+    {
+      "epoch": 0.6833245660178853,
+      "grad_norm": 1.5966815948486328,
+      "learning_rate": 4.815241741023367e-05,
+      "loss": 1.3903,
+      "step": 6495
+    },
+    {
+      "epoch": 0.6834297738032614,
+      "grad_norm": 2.34873104095459,
+      "learning_rate": 4.8123281125824605e-05,
+      "loss": 1.5625,
+      "step": 6496
+    },
+    {
+      "epoch": 0.6835349815886376,
+      "grad_norm": 1.5460422039031982,
+      "learning_rate": 4.809415086568812e-05,
+      "loss": 1.5584,
+      "step": 6497
+    },
+    {
+      "epoch": 0.6836401893740137,
+      "grad_norm": 1.2038516998291016,
+      "learning_rate": 4.806502663320692e-05,
+      "loss": 1.5281,
+      "step": 6498
+    },
+    {
+      "epoch": 0.6837453971593898,
+      "grad_norm": 1.3815281391143799,
+      "learning_rate": 4.803590843176321e-05,
+      "loss": 1.5034,
+      "step": 6499
+    },
+    {
+      "epoch": 0.6838506049447659,
+      "grad_norm": 0.981395959854126,
+      "learning_rate": 4.800679626473833e-05,
+      "loss": 1.9422,
+      "step": 6500
+    },
+    {
+      "epoch": 0.683955812730142,
+      "grad_norm": 1.7010056972503662,
+      "learning_rate": 4.797769013551295e-05,
+      "loss": 1.8847,
+      "step": 6501
+    },
+    {
+      "epoch": 0.6840610205155182,
+      "grad_norm": 1.381854772567749,
+      "learning_rate": 4.7948590047467153e-05,
+      "loss": 1.5431,
+      "step": 6502
+    },
+    {
+      "epoch": 0.6841662283008942,
+      "grad_norm": 1.5627236366271973,
+      "learning_rate": 4.7919496003980204e-05,
+      "loss": 2.1837,
+      "step": 6503
+    },
+    {
+      "epoch": 0.6842714360862704,
+      "grad_norm": 1.6263660192489624,
+      "learning_rate": 4.7890408008430634e-05,
+      "loss": 1.7972,
+      "step": 6504
+    },
+    {
+      "epoch": 0.6843766438716465,
+      "grad_norm": 1.3480515480041504,
+      "learning_rate": 4.786132606419643e-05,
+      "loss": 2.0057,
+      "step": 6505
+    },
+    {
+      "epoch": 0.6844818516570226,
+      "grad_norm": 1.6764869689941406,
+      "learning_rate": 4.783225017465475e-05,
+      "loss": 1.7601,
+      "step": 6506
+    },
+    {
+      "epoch": 0.6845870594423987,
+      "grad_norm": 1.2646028995513916,
+      "learning_rate": 4.780318034318202e-05,
+      "loss": 1.5116,
+      "step": 6507
+    },
+    {
+      "epoch": 0.6846922672277749,
+      "grad_norm": 1.2866673469543457,
+      "learning_rate": 4.7774116573154125e-05,
+      "loss": 1.5891,
+      "step": 6508
+    },
+    {
+      "epoch": 0.684797475013151,
+      "grad_norm": 1.5979012250900269,
+      "learning_rate": 4.774505886794609e-05,
+      "loss": 1.6371,
+      "step": 6509
+    },
+    {
+      "epoch": 0.6849026827985271,
+      "grad_norm": 1.9245673418045044,
+      "learning_rate": 4.771600723093227e-05,
+      "loss": 1.6979,
+      "step": 6510
+    },
+    {
+      "epoch": 0.6850078905839032,
+      "grad_norm": 1.8134851455688477,
+      "learning_rate": 4.7686961665486396e-05,
+      "loss": 1.9395,
+      "step": 6511
+    },
+    {
+      "epoch": 0.6851130983692794,
+      "grad_norm": 1.7789303064346313,
+      "learning_rate": 4.76579221749814e-05,
+      "loss": 1.915,
+      "step": 6512
+    },
+    {
+      "epoch": 0.6852183061546554,
+      "grad_norm": 1.5152759552001953,
+      "learning_rate": 4.7628888762789504e-05,
+      "loss": 1.4219,
+      "step": 6513
+    },
+    {
+      "epoch": 0.6853235139400315,
+      "grad_norm": 1.3812377452850342,
+      "learning_rate": 4.7599861432282334e-05,
+      "loss": 1.9226,
+      "step": 6514
+    },
+    {
+      "epoch": 0.6854287217254077,
+      "grad_norm": 0.8945170044898987,
+      "learning_rate": 4.757084018683071e-05,
+      "loss": 1.481,
+      "step": 6515
+    },
+    {
+      "epoch": 0.6855339295107838,
+      "grad_norm": 1.3005067110061646,
+      "learning_rate": 4.754182502980477e-05,
+      "loss": 1.8098,
+      "step": 6516
+    },
+    {
+      "epoch": 0.6856391372961599,
+      "grad_norm": 2.368222951889038,
+      "learning_rate": 4.7512815964573966e-05,
+      "loss": 1.1709,
+      "step": 6517
+    },
+    {
+      "epoch": 0.685744345081536,
+      "grad_norm": 1.7931416034698486,
+      "learning_rate": 4.748381299450695e-05,
+      "loss": 1.9524,
+      "step": 6518
+    },
+    {
+      "epoch": 0.6858495528669122,
+      "grad_norm": 1.7992273569107056,
+      "learning_rate": 4.7454816122971846e-05,
+      "loss": 0.7749,
+      "step": 6519
+    },
+    {
+      "epoch": 0.6859547606522882,
+      "grad_norm": 1.1509588956832886,
+      "learning_rate": 4.7425825353335915e-05,
+      "loss": 1.5281,
+      "step": 6520
+    },
+    {
+      "epoch": 0.6860599684376644,
+      "grad_norm": 2.1700100898742676,
+      "learning_rate": 4.7396840688965726e-05,
+      "loss": 1.7834,
+      "step": 6521
+    },
+    {
+      "epoch": 0.6861651762230405,
+      "grad_norm": 1.983696699142456,
+      "learning_rate": 4.7367862133227244e-05,
+      "loss": 1.9279,
+      "step": 6522
+    },
+    {
+      "epoch": 0.6862703840084167,
+      "grad_norm": 1.1107069253921509,
+      "learning_rate": 4.7338889689485624e-05,
+      "loss": 1.4591,
+      "step": 6523
+    },
+    {
+      "epoch": 0.6863755917937927,
+      "grad_norm": 1.1690186262130737,
+      "learning_rate": 4.730992336110529e-05,
+      "loss": 1.857,
+      "step": 6524
+    },
+    {
+      "epoch": 0.6864807995791689,
+      "grad_norm": 1.3441321849822998,
+      "learning_rate": 4.7280963151450096e-05,
+      "loss": 1.4621,
+      "step": 6525
+    },
+    {
+      "epoch": 0.686586007364545,
+      "grad_norm": 2.181993246078491,
+      "learning_rate": 4.7252009063883054e-05,
+      "loss": 1.483,
+      "step": 6526
+    },
+    {
+      "epoch": 0.686691215149921,
+      "grad_norm": 1.914829134941101,
+      "learning_rate": 4.722306110176647e-05,
+      "loss": 0.7554,
+      "step": 6527
+    },
+    {
+      "epoch": 0.6867964229352972,
+      "grad_norm": 2.0463366508483887,
+      "learning_rate": 4.719411926846203e-05,
+      "loss": 1.722,
+      "step": 6528
+    },
+    {
+      "epoch": 0.6869016307206733,
+      "grad_norm": 1.2413219213485718,
+      "learning_rate": 4.716518356733064e-05,
+      "loss": 1.53,
+      "step": 6529
+    },
+    {
+      "epoch": 0.6870068385060495,
+      "grad_norm": 1.5359559059143066,
+      "learning_rate": 4.713625400173247e-05,
+      "loss": 1.1939,
+      "step": 6530
+    },
+    {
+      "epoch": 0.6871120462914255,
+      "grad_norm": 1.2304325103759766,
+      "learning_rate": 4.7107330575027084e-05,
+      "loss": 1.6209,
+      "step": 6531
+    },
+    {
+      "epoch": 0.6872172540768017,
+      "grad_norm": 1.448864221572876,
+      "learning_rate": 4.707841329057322e-05,
+      "loss": 1.6243,
+      "step": 6532
+    },
+    {
+      "epoch": 0.6873224618621778,
+      "grad_norm": 1.8410067558288574,
+      "learning_rate": 4.7049502151728933e-05,
+      "loss": 1.5567,
+      "step": 6533
+    },
+    {
+      "epoch": 0.687427669647554,
+      "grad_norm": 2.130892753601074,
+      "learning_rate": 4.702059716185162e-05,
+      "loss": 1.9438,
+      "step": 6534
+    },
+    {
+      "epoch": 0.68753287743293,
+      "grad_norm": 1.4536875486373901,
+      "learning_rate": 4.6991698324297874e-05,
+      "loss": 1.7863,
+      "step": 6535
+    },
+    {
+      "epoch": 0.6876380852183062,
+      "grad_norm": 1.5897191762924194,
+      "learning_rate": 4.696280564242371e-05,
+      "loss": 0.9989,
+      "step": 6536
+    },
+    {
+      "epoch": 0.6877432930036823,
+      "grad_norm": 1.4431312084197998,
+      "learning_rate": 4.693391911958426e-05,
+      "loss": 1.7681,
+      "step": 6537
+    },
+    {
+      "epoch": 0.6878485007890583,
+      "grad_norm": 1.8787363767623901,
+      "learning_rate": 4.690503875913399e-05,
+      "loss": 1.8113,
+      "step": 6538
+    },
+    {
+      "epoch": 0.6879537085744345,
+      "grad_norm": 1.407496452331543,
+      "learning_rate": 4.687616456442677e-05,
+      "loss": 1.7591,
+      "step": 6539
+    },
+    {
+      "epoch": 0.6880589163598106,
+      "grad_norm": 1.6579171419143677,
+      "learning_rate": 4.684729653881563e-05,
+      "loss": 1.686,
+      "step": 6540
+    },
+    {
+      "epoch": 0.6881641241451868,
+      "grad_norm": 2.2351560592651367,
+      "learning_rate": 4.681843468565288e-05,
+      "loss": 1.7476,
+      "step": 6541
+    },
+    {
+      "epoch": 0.6882693319305628,
+      "grad_norm": 2.1286520957946777,
+      "learning_rate": 4.6789579008290216e-05,
+      "loss": 1.9967,
+      "step": 6542
+    },
+    {
+      "epoch": 0.688374539715939,
+      "grad_norm": 1.540056586265564,
+      "learning_rate": 4.676072951007849e-05,
+      "loss": 1.0093,
+      "step": 6543
+    },
+    {
+      "epoch": 0.6884797475013151,
+      "grad_norm": 1.2352454662322998,
+      "learning_rate": 4.673188619436798e-05,
+      "loss": 1.7239,
+      "step": 6544
+    },
+    {
+      "epoch": 0.6885849552866912,
+      "grad_norm": 1.0865057706832886,
+      "learning_rate": 4.670304906450811e-05,
+      "loss": 2.0791,
+      "step": 6545
+    },
+    {
+      "epoch": 0.6886901630720673,
+      "grad_norm": 1.6150873899459839,
+      "learning_rate": 4.667421812384761e-05,
+      "loss": 1.813,
+      "step": 6546
+    },
+    {
+      "epoch": 0.6887953708574435,
+      "grad_norm": 1.4798706769943237,
+      "learning_rate": 4.66453933757346e-05,
+      "loss": 1.634,
+      "step": 6547
+    },
+    {
+      "epoch": 0.6889005786428196,
+      "grad_norm": 1.3703545331954956,
+      "learning_rate": 4.661657482351637e-05,
+      "loss": 1.7388,
+      "step": 6548
+    },
+    {
+      "epoch": 0.6890057864281957,
+      "grad_norm": 1.6316715478897095,
+      "learning_rate": 4.658776247053948e-05,
+      "loss": 2.3265,
+      "step": 6549
+    },
+    {
+      "epoch": 0.6891109942135718,
+      "grad_norm": 1.0946226119995117,
+      "learning_rate": 4.6558956320149884e-05,
+      "loss": 1.2866,
+      "step": 6550
+    },
+    {
+      "epoch": 0.689216201998948,
+      "grad_norm": 1.6693997383117676,
+      "learning_rate": 4.6530156375692726e-05,
+      "loss": 1.583,
+      "step": 6551
+    },
+    {
+      "epoch": 0.689321409784324,
+      "grad_norm": 1.315307378768921,
+      "learning_rate": 4.650136264051238e-05,
+      "loss": 2.1853,
+      "step": 6552
+    },
+    {
+      "epoch": 0.6894266175697001,
+      "grad_norm": 1.1383665800094604,
+      "learning_rate": 4.6472575117952676e-05,
+      "loss": 1.7014,
+      "step": 6553
+    },
+    {
+      "epoch": 0.6895318253550763,
+      "grad_norm": 1.3326300382614136,
+      "learning_rate": 4.644379381135655e-05,
+      "loss": 1.7168,
+      "step": 6554
+    },
+    {
+      "epoch": 0.6896370331404524,
+      "grad_norm": 1.85318922996521,
+      "learning_rate": 4.641501872406626e-05,
+      "loss": 1.7368,
+      "step": 6555
+    },
+    {
+      "epoch": 0.6897422409258285,
+      "grad_norm": 1.3762325048446655,
+      "learning_rate": 4.6386249859423434e-05,
+      "loss": 2.0301,
+      "step": 6556
+    },
+    {
+      "epoch": 0.6898474487112046,
+      "grad_norm": 1.573012351989746,
+      "learning_rate": 4.635748722076887e-05,
+      "loss": 2.1703,
+      "step": 6557
+    },
+    {
+      "epoch": 0.6899526564965808,
+      "grad_norm": 1.4833385944366455,
+      "learning_rate": 4.632873081144267e-05,
+      "loss": 1.4837,
+      "step": 6558
+    },
+    {
+      "epoch": 0.6900578642819568,
+      "grad_norm": 1.669286847114563,
+      "learning_rate": 4.629998063478422e-05,
+      "loss": 1.6238,
+      "step": 6559
+    },
+    {
+      "epoch": 0.690163072067333,
+      "grad_norm": 1.670414924621582,
+      "learning_rate": 4.627123669413216e-05,
+      "loss": 1.5925,
+      "step": 6560
+    },
+    {
+      "epoch": 0.6902682798527091,
+      "grad_norm": 1.836533784866333,
+      "learning_rate": 4.624249899282449e-05,
+      "loss": 1.5332,
+      "step": 6561
+    },
+    {
+      "epoch": 0.6903734876380853,
+      "grad_norm": 1.0955008268356323,
+      "learning_rate": 4.6213767534198395e-05,
+      "loss": 1.9474,
+      "step": 6562
+    },
+    {
+      "epoch": 0.6904786954234613,
+      "grad_norm": 1.3523670434951782,
+      "learning_rate": 4.618504232159032e-05,
+      "loss": 1.3039,
+      "step": 6563
+    },
+    {
+      "epoch": 0.6905839032088374,
+      "grad_norm": 1.6386244297027588,
+      "learning_rate": 4.6156323358336116e-05,
+      "loss": 1.7856,
+      "step": 6564
+    },
+    {
+      "epoch": 0.6906891109942136,
+      "grad_norm": 1.7706154584884644,
+      "learning_rate": 4.6127610647770767e-05,
+      "loss": 2.0213,
+      "step": 6565
+    },
+    {
+      "epoch": 0.6907943187795897,
+      "grad_norm": 1.2387391328811646,
+      "learning_rate": 4.6098904193228576e-05,
+      "loss": 1.5836,
+      "step": 6566
+    },
+    {
+      "epoch": 0.6908995265649658,
+      "grad_norm": 1.4333043098449707,
+      "learning_rate": 4.6070203998043173e-05,
+      "loss": 1.6886,
+      "step": 6567
+    },
+    {
+      "epoch": 0.6910047343503419,
+      "grad_norm": 1.220526933670044,
+      "learning_rate": 4.60415100655474e-05,
+      "loss": 1.6656,
+      "step": 6568
+    },
+    {
+      "epoch": 0.6911099421357181,
+      "grad_norm": 1.2803022861480713,
+      "learning_rate": 4.601282239907334e-05,
+      "loss": 1.6475,
+      "step": 6569
+    },
+    {
+      "epoch": 0.6912151499210941,
+      "grad_norm": 1.9146497249603271,
+      "learning_rate": 4.5984141001952477e-05,
+      "loss": 1.638,
+      "step": 6570
+    },
+    {
+      "epoch": 0.6913203577064703,
+      "grad_norm": 0.9470534324645996,
+      "learning_rate": 4.595546587751545e-05,
+      "loss": 1.6093,
+      "step": 6571
+    },
+    {
+      "epoch": 0.6914255654918464,
+      "grad_norm": 1.471522569656372,
+      "learning_rate": 4.592679702909216e-05,
+      "loss": 1.1011,
+      "step": 6572
+    },
+    {
+      "epoch": 0.6915307732772226,
+      "grad_norm": 1.4972786903381348,
+      "learning_rate": 4.589813446001192e-05,
+      "loss": 1.4328,
+      "step": 6573
+    },
+    {
+      "epoch": 0.6916359810625986,
+      "grad_norm": 1.2175990343093872,
+      "learning_rate": 4.5869478173603175e-05,
+      "loss": 1.8741,
+      "step": 6574
+    },
+    {
+      "epoch": 0.6917411888479748,
+      "grad_norm": 1.4953633546829224,
+      "learning_rate": 4.584082817319364e-05,
+      "loss": 1.8677,
+      "step": 6575
+    },
+    {
+      "epoch": 0.6918463966333509,
+      "grad_norm": 1.126177191734314,
+      "learning_rate": 4.581218446211043e-05,
+      "loss": 1.9968,
+      "step": 6576
+    },
+    {
+      "epoch": 0.6919516044187269,
+      "grad_norm": 1.063721776008606,
+      "learning_rate": 4.578354704367978e-05,
+      "loss": 1.6822,
+      "step": 6577
+    },
+    {
+      "epoch": 0.6920568122041031,
+      "grad_norm": 1.747952938079834,
+      "learning_rate": 4.575491592122727e-05,
+      "loss": 1.8211,
+      "step": 6578
+    },
+    {
+      "epoch": 0.6921620199894792,
+      "grad_norm": 1.776321291923523,
+      "learning_rate": 4.572629109807782e-05,
+      "loss": 2.0429,
+      "step": 6579
+    },
+    {
+      "epoch": 0.6922672277748554,
+      "grad_norm": 1.6665074825286865,
+      "learning_rate": 4.569767257755538e-05,
+      "loss": 1.7637,
+      "step": 6580
+    },
+    {
+      "epoch": 0.6923724355602314,
+      "grad_norm": 1.534345030784607,
+      "learning_rate": 4.566906036298345e-05,
+      "loss": 1.4153,
+      "step": 6581
+    },
+    {
+      "epoch": 0.6924776433456076,
+      "grad_norm": 1.4633084535598755,
+      "learning_rate": 4.564045445768464e-05,
+      "loss": 1.7596,
+      "step": 6582
+    },
+    {
+      "epoch": 0.6925828511309837,
+      "grad_norm": 1.559431552886963,
+      "learning_rate": 4.561185486498081e-05,
+      "loss": 1.2046,
+      "step": 6583
+    },
+    {
+      "epoch": 0.6926880589163598,
+      "grad_norm": 1.8445146083831787,
+      "learning_rate": 4.558326158819322e-05,
+      "loss": 1.4508,
+      "step": 6584
+    },
+    {
+      "epoch": 0.6927932667017359,
+      "grad_norm": 1.142551064491272,
+      "learning_rate": 4.555467463064227e-05,
+      "loss": 1.0419,
+      "step": 6585
+    },
+    {
+      "epoch": 0.6928984744871121,
+      "grad_norm": 2.6292953491210938,
+      "learning_rate": 4.552609399564762e-05,
+      "loss": 1.4199,
+      "step": 6586
+    },
+    {
+      "epoch": 0.6930036822724882,
+      "grad_norm": 0.9221459627151489,
+      "learning_rate": 4.549751968652836e-05,
+      "loss": 1.4626,
+      "step": 6587
+    },
+    {
+      "epoch": 0.6931088900578642,
+      "grad_norm": 1.3929451704025269,
+      "learning_rate": 4.5468951706602644e-05,
+      "loss": 2.0308,
+      "step": 6588
+    },
+    {
+      "epoch": 0.6932140978432404,
+      "grad_norm": 1.388808012008667,
+      "learning_rate": 4.5440390059187964e-05,
+      "loss": 1.8868,
+      "step": 6589
+    },
+    {
+      "epoch": 0.6933193056286165,
+      "grad_norm": 1.4524317979812622,
+      "learning_rate": 4.541183474760118e-05,
+      "loss": 1.747,
+      "step": 6590
+    },
+    {
+      "epoch": 0.6934245134139926,
+      "grad_norm": 1.3403692245483398,
+      "learning_rate": 4.538328577515821e-05,
+      "loss": 1.5627,
+      "step": 6591
+    },
+    {
+      "epoch": 0.6935297211993687,
+      "grad_norm": 1.3830885887145996,
+      "learning_rate": 4.535474314517447e-05,
+      "loss": 1.6002,
+      "step": 6592
+    },
+    {
+      "epoch": 0.6936349289847449,
+      "grad_norm": 2.1278774738311768,
+      "learning_rate": 4.532620686096446e-05,
+      "loss": 1.9424,
+      "step": 6593
+    },
+    {
+      "epoch": 0.693740136770121,
+      "grad_norm": 1.8062392473220825,
+      "learning_rate": 4.529767692584198e-05,
+      "loss": 2.012,
+      "step": 6594
+    },
+    {
+      "epoch": 0.6938453445554971,
+      "grad_norm": 1.8589589595794678,
+      "learning_rate": 4.5269153343120174e-05,
+      "loss": 1.4928,
+      "step": 6595
+    },
+    {
+      "epoch": 0.6939505523408732,
+      "grad_norm": 1.7464429140090942,
+      "learning_rate": 4.524063611611138e-05,
+      "loss": 1.1396,
+      "step": 6596
+    },
+    {
+      "epoch": 0.6940557601262494,
+      "grad_norm": 1.3345447778701782,
+      "learning_rate": 4.5212125248127143e-05,
+      "loss": 1.5639,
+      "step": 6597
+    },
+    {
+      "epoch": 0.6941609679116255,
+      "grad_norm": 1.2210328578948975,
+      "learning_rate": 4.518362074247844e-05,
+      "loss": 1.8796,
+      "step": 6598
+    },
+    {
+      "epoch": 0.6942661756970016,
+      "grad_norm": 1.5386043787002563,
+      "learning_rate": 4.515512260247534e-05,
+      "loss": 1.1067,
+      "step": 6599
+    },
+    {
+      "epoch": 0.6943713834823777,
+      "grad_norm": 1.575567364692688,
+      "learning_rate": 4.5126630831427264e-05,
+      "loss": 1.0896,
+      "step": 6600
+    },
+    {
+      "epoch": 0.6944765912677539,
+      "grad_norm": 1.171311855316162,
+      "learning_rate": 4.5098145432642845e-05,
+      "loss": 1.6938,
+      "step": 6601
+    },
+    {
+      "epoch": 0.6945817990531299,
+      "grad_norm": 1.322606086730957,
+      "learning_rate": 4.506966640942999e-05,
+      "loss": 1.9014,
+      "step": 6602
+    },
+    {
+      "epoch": 0.694687006838506,
+      "grad_norm": 1.8430259227752686,
+      "learning_rate": 4.504119376509591e-05,
+      "loss": 1.916,
+      "step": 6603
+    },
+    {
+      "epoch": 0.6947922146238822,
+      "grad_norm": 1.57040536403656,
+      "learning_rate": 4.501272750294704e-05,
+      "loss": 1.1643,
+      "step": 6604
+    },
+    {
+      "epoch": 0.6948974224092583,
+      "grad_norm": 1.510570764541626,
+      "learning_rate": 4.4984267626289e-05,
+      "loss": 1.4288,
+      "step": 6605
+    },
+    {
+      "epoch": 0.6950026301946344,
+      "grad_norm": 1.7676854133605957,
+      "learning_rate": 4.495581413842685e-05,
+      "loss": 1.0889,
+      "step": 6606
+    },
+    {
+      "epoch": 0.6951078379800105,
+      "grad_norm": 1.2651560306549072,
+      "learning_rate": 4.492736704266475e-05,
+      "loss": 1.6327,
+      "step": 6607
+    },
+    {
+      "epoch": 0.6952130457653867,
+      "grad_norm": 1.7821506261825562,
+      "learning_rate": 4.4898926342306115e-05,
+      "loss": 1.7645,
+      "step": 6608
+    },
+    {
+      "epoch": 0.6953182535507627,
+      "grad_norm": 1.2521933317184448,
+      "learning_rate": 4.487049204065377e-05,
+      "loss": 2.0037,
+      "step": 6609
+    },
+    {
+      "epoch": 0.6954234613361389,
+      "grad_norm": 1.3124496936798096,
+      "learning_rate": 4.4842064141009644e-05,
+      "loss": 1.5326,
+      "step": 6610
+    },
+    {
+      "epoch": 0.695528669121515,
+      "grad_norm": 1.9574371576309204,
+      "learning_rate": 4.4813642646674936e-05,
+      "loss": 1.8915,
+      "step": 6611
+    },
+    {
+      "epoch": 0.6956338769068912,
+      "grad_norm": 1.5282095670700073,
+      "learning_rate": 4.4785227560950226e-05,
+      "loss": 1.9569,
+      "step": 6612
+    },
+    {
+      "epoch": 0.6957390846922672,
+      "grad_norm": 1.1745871305465698,
+      "learning_rate": 4.475681888713522e-05,
+      "loss": 1.8232,
+      "step": 6613
+    },
+    {
+      "epoch": 0.6958442924776433,
+      "grad_norm": 1.0485895872116089,
+      "learning_rate": 4.472841662852888e-05,
+      "loss": 1.814,
+      "step": 6614
+    },
+    {
+      "epoch": 0.6959495002630195,
+      "grad_norm": 1.5237230062484741,
+      "learning_rate": 4.470002078842957e-05,
+      "loss": 1.4868,
+      "step": 6615
+    },
+    {
+      "epoch": 0.6960547080483955,
+      "grad_norm": 1.6121946573257446,
+      "learning_rate": 4.467163137013473e-05,
+      "loss": 1.727,
+      "step": 6616
+    },
+    {
+      "epoch": 0.6961599158337717,
+      "grad_norm": 1.475019097328186,
+      "learning_rate": 4.4643248376941104e-05,
+      "loss": 1.5386,
+      "step": 6617
+    },
+    {
+      "epoch": 0.6962651236191478,
+      "grad_norm": 1.4994443655014038,
+      "learning_rate": 4.461487181214481e-05,
+      "loss": 1.6624,
+      "step": 6618
+    },
+    {
+      "epoch": 0.696370331404524,
+      "grad_norm": 2.619147300720215,
+      "learning_rate": 4.458650167904106e-05,
+      "loss": 1.5808,
+      "step": 6619
+    },
+    {
+      "epoch": 0.6964755391899,
+      "grad_norm": 1.311523675918579,
+      "learning_rate": 4.455813798092438e-05,
+      "loss": 1.7372,
+      "step": 6620
+    },
+    {
+      "epoch": 0.6965807469752762,
+      "grad_norm": 1.382138967514038,
+      "learning_rate": 4.452978072108859e-05,
+      "loss": 1.5582,
+      "step": 6621
+    },
+    {
+      "epoch": 0.6966859547606523,
+      "grad_norm": 1.6938810348510742,
+      "learning_rate": 4.450142990282671e-05,
+      "loss": 1.7523,
+      "step": 6622
+    },
+    {
+      "epoch": 0.6967911625460284,
+      "grad_norm": 1.1859630346298218,
+      "learning_rate": 4.4473085529431024e-05,
+      "loss": 1.7407,
+      "step": 6623
+    },
+    {
+      "epoch": 0.6968963703314045,
+      "grad_norm": 1.2967774868011475,
+      "learning_rate": 4.4444747604193074e-05,
+      "loss": 1.9172,
+      "step": 6624
+    },
+    {
+      "epoch": 0.6970015781167807,
+      "grad_norm": 1.2925198078155518,
+      "learning_rate": 4.44164161304036e-05,
+      "loss": 1.9331,
+      "step": 6625
+    },
+    {
+      "epoch": 0.6971067859021568,
+      "grad_norm": 1.6732769012451172,
+      "learning_rate": 4.438809111135274e-05,
+      "loss": 1.8172,
+      "step": 6626
+    },
+    {
+      "epoch": 0.6972119936875328,
+      "grad_norm": 1.5457383394241333,
+      "learning_rate": 4.435977255032971e-05,
+      "loss": 1.7216,
+      "step": 6627
+    },
+    {
+      "epoch": 0.697317201472909,
+      "grad_norm": 1.5493268966674805,
+      "learning_rate": 4.4331460450623064e-05,
+      "loss": 1.2244,
+      "step": 6628
+    },
+    {
+      "epoch": 0.6974224092582851,
+      "grad_norm": 1.3660433292388916,
+      "learning_rate": 4.430315481552063e-05,
+      "loss": 1.4947,
+      "step": 6629
+    },
+    {
+      "epoch": 0.6975276170436613,
+      "grad_norm": 1.6390910148620605,
+      "learning_rate": 4.427485564830942e-05,
+      "loss": 1.8469,
+      "step": 6630
+    },
+    {
+      "epoch": 0.6976328248290373,
+      "grad_norm": 2.015472173690796,
+      "learning_rate": 4.424656295227568e-05,
+      "loss": 1.7447,
+      "step": 6631
+    },
+    {
+      "epoch": 0.6977380326144135,
+      "grad_norm": 1.0427132844924927,
+      "learning_rate": 4.4218276730705045e-05,
+      "loss": 1.5705,
+      "step": 6632
+    },
+    {
+      "epoch": 0.6978432403997896,
+      "grad_norm": 1.1284888982772827,
+      "learning_rate": 4.4189996986882245e-05,
+      "loss": 1.869,
+      "step": 6633
+    },
+    {
+      "epoch": 0.6979484481851657,
+      "grad_norm": 1.4589309692382812,
+      "learning_rate": 4.4161723724091276e-05,
+      "loss": 1.6884,
+      "step": 6634
+    },
+    {
+      "epoch": 0.6980536559705418,
+      "grad_norm": 0.9497972726821899,
+      "learning_rate": 4.413345694561549e-05,
+      "loss": 1.9814,
+      "step": 6635
+    },
+    {
+      "epoch": 0.698158863755918,
+      "grad_norm": 1.5925368070602417,
+      "learning_rate": 4.410519665473736e-05,
+      "loss": 1.5224,
+      "step": 6636
+    },
+    {
+      "epoch": 0.6982640715412941,
+      "grad_norm": 2.054982900619507,
+      "learning_rate": 4.407694285473871e-05,
+      "loss": 1.4793,
+      "step": 6637
+    },
+    {
+      "epoch": 0.6983692793266701,
+      "grad_norm": 1.2141071557998657,
+      "learning_rate": 4.404869554890054e-05,
+      "loss": 1.2192,
+      "step": 6638
+    },
+    {
+      "epoch": 0.6984744871120463,
+      "grad_norm": 1.3335728645324707,
+      "learning_rate": 4.402045474050308e-05,
+      "loss": 1.9024,
+      "step": 6639
+    },
+    {
+      "epoch": 0.6985796948974224,
+      "grad_norm": 1.548563838005066,
+      "learning_rate": 4.399222043282591e-05,
+      "loss": 2.0828,
+      "step": 6640
+    },
+    {
+      "epoch": 0.6986849026827985,
+      "grad_norm": 2.5695066452026367,
+      "learning_rate": 4.3963992629147755e-05,
+      "loss": 1.768,
+      "step": 6641
+    },
+    {
+      "epoch": 0.6987901104681746,
+      "grad_norm": 1.7723878622055054,
+      "learning_rate": 4.393577133274658e-05,
+      "loss": 1.7614,
+      "step": 6642
+    },
+    {
+      "epoch": 0.6988953182535508,
+      "grad_norm": 1.082208275794983,
+      "learning_rate": 4.390755654689973e-05,
+      "loss": 2.1651,
+      "step": 6643
+    },
+    {
+      "epoch": 0.6990005260389269,
+      "grad_norm": 1.8237231969833374,
+      "learning_rate": 4.3879348274883594e-05,
+      "loss": 1.8408,
+      "step": 6644
+    },
+    {
+      "epoch": 0.699105733824303,
+      "grad_norm": 1.6238867044448853,
+      "learning_rate": 4.3851146519973906e-05,
+      "loss": 1.2499,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6992109416096791,
+      "grad_norm": 1.6902471780776978,
+      "learning_rate": 4.382295128544572e-05,
+      "loss": 1.916,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6993161493950553,
+      "grad_norm": 1.5541309118270874,
+      "learning_rate": 4.379476257457318e-05,
+      "loss": 1.6904,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6994213571804313,
+      "grad_norm": 1.7713851928710938,
+      "learning_rate": 4.376658039062981e-05,
+      "loss": 1.7081,
+      "step": 6648
+    },
+    {
+      "epoch": 0.6995265649658075,
+      "grad_norm": 1.4880828857421875,
+      "learning_rate": 4.373840473688829e-05,
+      "loss": 1.865,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6996317727511836,
+      "grad_norm": 1.3452609777450562,
+      "learning_rate": 4.371023561662052e-05,
+      "loss": 2.0241,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6997369805365597,
+      "grad_norm": 1.048627495765686,
+      "learning_rate": 4.3682073033097785e-05,
+      "loss": 1.8618,
+      "step": 6651
+    },
+    {
+      "epoch": 0.6998421883219358,
+      "grad_norm": 1.0086387395858765,
+      "learning_rate": 4.365391698959044e-05,
+      "loss": 1.8462,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6999473961073119,
+      "grad_norm": 1.9956762790679932,
+      "learning_rate": 4.3625767489368143e-05,
+      "loss": 1.7446,
+      "step": 6653
+    },
+    {
+      "epoch": 0.7000526038926881,
+      "grad_norm": 1.3309284448623657,
+      "learning_rate": 4.3597624535699865e-05,
+      "loss": 1.8599,
+      "step": 6654
+    },
+    {
+      "epoch": 0.7001578116780641,
+      "grad_norm": 1.1695083379745483,
+      "learning_rate": 4.3569488131853733e-05,
+      "loss": 1.3716,
+      "step": 6655
+    },
+    {
+      "epoch": 0.7002630194634403,
+      "grad_norm": 1.7402067184448242,
+      "learning_rate": 4.354135828109707e-05,
+      "loss": 1.3841,
+      "step": 6656
+    },
+    {
+      "epoch": 0.7003682272488164,
+      "grad_norm": 1.5015475749969482,
+      "learning_rate": 4.351323498669659e-05,
+      "loss": 1.2452,
+      "step": 6657
+    },
+    {
+      "epoch": 0.7004734350341926,
+      "grad_norm": 1.4218254089355469,
+      "learning_rate": 4.3485118251918146e-05,
+      "loss": 1.9119,
+      "step": 6658
+    },
+    {
+      "epoch": 0.7005786428195686,
+      "grad_norm": 1.1995227336883545,
+      "learning_rate": 4.345700808002676e-05,
+      "loss": 1.6662,
+      "step": 6659
+    },
+    {
+      "epoch": 0.7006838506049448,
+      "grad_norm": 1.214152455329895,
+      "learning_rate": 4.3428904474286894e-05,
+      "loss": 1.6176,
+      "step": 6660
+    },
+    {
+      "epoch": 0.7007890583903209,
+      "grad_norm": 1.6113765239715576,
+      "learning_rate": 4.340080743796204e-05,
+      "loss": 1.8252,
+      "step": 6661
+    },
+    {
+      "epoch": 0.7008942661756971,
+      "grad_norm": 1.3512358665466309,
+      "learning_rate": 4.337271697431503e-05,
+      "loss": 1.5226,
+      "step": 6662
+    },
+    {
+      "epoch": 0.7009994739610731,
+      "grad_norm": 1.1863796710968018,
+      "learning_rate": 4.3344633086607955e-05,
+      "loss": 1.6714,
+      "step": 6663
+    },
+    {
+      "epoch": 0.7011046817464492,
+      "grad_norm": 1.354353904724121,
+      "learning_rate": 4.331655577810207e-05,
+      "loss": 1.7832,
+      "step": 6664
+    },
+    {
+      "epoch": 0.7012098895318254,
+      "grad_norm": 1.0943442583084106,
+      "learning_rate": 4.328848505205792e-05,
+      "loss": 1.9705,
+      "step": 6665
+    },
+    {
+      "epoch": 0.7013150973172014,
+      "grad_norm": 1.4298934936523438,
+      "learning_rate": 4.326042091173526e-05,
+      "loss": 1.2593,
+      "step": 6666
+    },
+    {
+      "epoch": 0.7014203051025776,
+      "grad_norm": 2.2146830558776855,
+      "learning_rate": 4.3232363360393026e-05,
+      "loss": 1.7405,
+      "step": 6667
+    },
+    {
+      "epoch": 0.7015255128879537,
+      "grad_norm": 1.1296732425689697,
+      "learning_rate": 4.320431240128955e-05,
+      "loss": 1.8832,
+      "step": 6668
+    },
+    {
+      "epoch": 0.7016307206733299,
+      "grad_norm": 1.771040439605713,
+      "learning_rate": 4.317626803768224e-05,
+      "loss": 1.9983,
+      "step": 6669
+    },
+    {
+      "epoch": 0.7017359284587059,
+      "grad_norm": 1.8517831563949585,
+      "learning_rate": 4.3148230272827784e-05,
+      "loss": 1.7378,
+      "step": 6670
+    },
+    {
+      "epoch": 0.7018411362440821,
+      "grad_norm": 1.6702775955200195,
+      "learning_rate": 4.312019910998217e-05,
+      "loss": 1.7574,
+      "step": 6671
+    },
+    {
+      "epoch": 0.7019463440294582,
+      "grad_norm": 1.3865050077438354,
+      "learning_rate": 4.3092174552400535e-05,
+      "loss": 1.5659,
+      "step": 6672
+    },
+    {
+      "epoch": 0.7020515518148343,
+      "grad_norm": 1.095326542854309,
+      "learning_rate": 4.306415660333724e-05,
+      "loss": 1.0576,
+      "step": 6673
+    },
+    {
+      "epoch": 0.7021567596002104,
+      "grad_norm": 1.3745871782302856,
+      "learning_rate": 4.303614526604598e-05,
+      "loss": 1.9348,
+      "step": 6674
+    },
+    {
+      "epoch": 0.7022619673855865,
+      "grad_norm": 1.3735612630844116,
+      "learning_rate": 4.300814054377961e-05,
+      "loss": 1.3382,
+      "step": 6675
+    },
+    {
+      "epoch": 0.7023671751709627,
+      "grad_norm": 1.8741841316223145,
+      "learning_rate": 4.298014243979016e-05,
+      "loss": 1.7592,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7024723829563387,
+      "grad_norm": 1.2442336082458496,
+      "learning_rate": 4.295215095732904e-05,
+      "loss": 1.4593,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7025775907417149,
+      "grad_norm": 1.6263591051101685,
+      "learning_rate": 4.292416609964678e-05,
+      "loss": 1.3543,
+      "step": 6678
+    },
+    {
+      "epoch": 0.702682798527091,
+      "grad_norm": 1.3969215154647827,
+      "learning_rate": 4.289618786999313e-05,
+      "loss": 1.6856,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7027880063124671,
+      "grad_norm": 1.4358292818069458,
+      "learning_rate": 4.2868216271617175e-05,
+      "loss": 1.7041,
+      "step": 6680
+    },
+    {
+      "epoch": 0.7028932140978432,
+      "grad_norm": 1.739772081375122,
+      "learning_rate": 4.284025130776711e-05,
+      "loss": 2.2992,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7029984218832194,
+      "grad_norm": 1.508034586906433,
+      "learning_rate": 4.281229298169046e-05,
+      "loss": 1.5951,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7031036296685955,
+      "grad_norm": 1.9585610628128052,
+      "learning_rate": 4.278434129663392e-05,
+      "loss": 1.8921,
+      "step": 6683
+    },
+    {
+      "epoch": 0.7032088374539716,
+      "grad_norm": 1.0624831914901733,
+      "learning_rate": 4.275639625584338e-05,
+      "loss": 1.7377,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7033140452393477,
+      "grad_norm": 1.728935956954956,
+      "learning_rate": 4.2728457862564074e-05,
+      "loss": 1.6178,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7034192530247239,
+      "grad_norm": 1.3254528045654297,
+      "learning_rate": 4.2700526120040405e-05,
+      "loss": 1.6511,
+      "step": 6686
+    },
+    {
+      "epoch": 0.7035244608100999,
+      "grad_norm": 1.6863447427749634,
+      "learning_rate": 4.267260103151589e-05,
+      "loss": 1.4184,
+      "step": 6687
+    },
+    {
+      "epoch": 0.703629668595476,
+      "grad_norm": 1.829534888267517,
+      "learning_rate": 4.264468260023348e-05,
+      "loss": 1.9914,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7037348763808522,
+      "grad_norm": 1.3819561004638672,
+      "learning_rate": 4.261677082943521e-05,
+      "loss": 1.7343,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7038400841662283,
+      "grad_norm": 1.2711637020111084,
+      "learning_rate": 4.2588865722362334e-05,
+      "loss": 1.9204,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7039452919516044,
+      "grad_norm": 1.641611099243164,
+      "learning_rate": 4.256096728225548e-05,
+      "loss": 1.5314,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7040504997369805,
+      "grad_norm": 1.4409161806106567,
+      "learning_rate": 4.253307551235431e-05,
+      "loss": 1.745,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7041557075223567,
+      "grad_norm": 1.646813988685608,
+      "learning_rate": 4.25051904158979e-05,
+      "loss": 1.9297,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7042609153077328,
+      "grad_norm": 1.5589001178741455,
+      "learning_rate": 4.247731199612439e-05,
+      "loss": 1.5343,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7043661230931089,
+      "grad_norm": 2.051630973815918,
+      "learning_rate": 4.244944025627118e-05,
+      "loss": 1.4001,
+      "step": 6695
+    },
+    {
+      "epoch": 0.704471330878485,
+      "grad_norm": 1.5140016078948975,
+      "learning_rate": 4.2421575199575e-05,
+      "loss": 1.8798,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7045765386638612,
+      "grad_norm": 1.471285104751587,
+      "learning_rate": 4.23937168292717e-05,
+      "loss": 1.5356,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7046817464492372,
+      "grad_norm": 1.4824402332305908,
+      "learning_rate": 4.236586514859633e-05,
+      "loss": 1.5193,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7047869542346133,
+      "grad_norm": 1.7141013145446777,
+      "learning_rate": 4.233802016078329e-05,
+      "loss": 1.4694,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7048921620199895,
+      "grad_norm": 1.1465423107147217,
+      "learning_rate": 4.23101818690661e-05,
+      "loss": 1.6485,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7049973698053656,
+      "grad_norm": 1.4803214073181152,
+      "learning_rate": 4.2282350276677475e-05,
+      "loss": 1.6299,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7051025775907417,
+      "grad_norm": 1.4447064399719238,
+      "learning_rate": 4.2254525386849497e-05,
+      "loss": 1.7181,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7052077853761178,
+      "grad_norm": 2.032977819442749,
+      "learning_rate": 4.222670720281333e-05,
+      "loss": 0.969,
+      "step": 6703
+    },
+    {
+      "epoch": 0.705312993161494,
+      "grad_norm": 1.6613085269927979,
+      "learning_rate": 4.219889572779937e-05,
+      "loss": 1.4189,
+      "step": 6704
+    },
+    {
+      "epoch": 0.70541820094687,
+      "grad_norm": 1.3568087816238403,
+      "learning_rate": 4.217109096503736e-05,
+      "loss": 1.8463,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7055234087322462,
+      "grad_norm": 1.6580621004104614,
+      "learning_rate": 4.214329291775613e-05,
+      "loss": 1.6142,
+      "step": 6706
+    },
+    {
+      "epoch": 0.7056286165176223,
+      "grad_norm": 1.5139204263687134,
+      "learning_rate": 4.2115501589183734e-05,
+      "loss": 1.7448,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7057338243029985,
+      "grad_norm": 2.026707649230957,
+      "learning_rate": 4.208771698254761e-05,
+      "loss": 1.5281,
+      "step": 6708
+    },
+    {
+      "epoch": 0.7058390320883745,
+      "grad_norm": 1.6222214698791504,
+      "learning_rate": 4.205993910107413e-05,
+      "loss": 1.3807,
+      "step": 6709
+    },
+    {
+      "epoch": 0.7059442398737507,
+      "grad_norm": 1.521702766418457,
+      "learning_rate": 4.2032167947989175e-05,
+      "loss": 1.706,
+      "step": 6710
+    },
+    {
+      "epoch": 0.7060494476591268,
+      "grad_norm": 1.6336759328842163,
+      "learning_rate": 4.200440352651767e-05,
+      "loss": 1.8752,
+      "step": 6711
+    },
+    {
+      "epoch": 0.7061546554445028,
+      "grad_norm": 1.2385070323944092,
+      "learning_rate": 4.197664583988376e-05,
+      "loss": 1.8389,
+      "step": 6712
+    },
+    {
+      "epoch": 0.706259863229879,
+      "grad_norm": 1.7074710130691528,
+      "learning_rate": 4.1948894891310955e-05,
+      "loss": 2.3842,
+      "step": 6713
+    },
+    {
+      "epoch": 0.7063650710152551,
+      "grad_norm": 1.0736308097839355,
+      "learning_rate": 4.192115068402183e-05,
+      "loss": 1.5353,
+      "step": 6714
+    },
+    {
+      "epoch": 0.7064702788006313,
+      "grad_norm": 2.215698719024658,
+      "learning_rate": 4.189341322123818e-05,
+      "loss": 1.0607,
+      "step": 6715
+    },
+    {
+      "epoch": 0.7065754865860073,
+      "grad_norm": 2.0693204402923584,
+      "learning_rate": 4.186568250618115e-05,
+      "loss": 1.8865,
+      "step": 6716
+    },
+    {
+      "epoch": 0.7066806943713835,
+      "grad_norm": 2.8470065593719482,
+      "learning_rate": 4.183795854207098e-05,
+      "loss": 1.8147,
+      "step": 6717
+    },
+    {
+      "epoch": 0.7067859021567596,
+      "grad_norm": 1.833847999572754,
+      "learning_rate": 4.181024133212713e-05,
+      "loss": 1.4404,
+      "step": 6718
+    },
+    {
+      "epoch": 0.7068911099421357,
+      "grad_norm": 1.5717699527740479,
+      "learning_rate": 4.1782530879568374e-05,
+      "loss": 1.5635,
+      "step": 6719
+    },
+    {
+      "epoch": 0.7069963177275118,
+      "grad_norm": 1.44779372215271,
+      "learning_rate": 4.175482718761259e-05,
+      "loss": 1.9344,
+      "step": 6720
+    },
+    {
+      "epoch": 0.707101525512888,
+      "grad_norm": 1.8596739768981934,
+      "learning_rate": 4.172713025947691e-05,
+      "loss": 2.044,
+      "step": 6721
+    },
+    {
+      "epoch": 0.7072067332982641,
+      "grad_norm": 1.5004318952560425,
+      "learning_rate": 4.169944009837773e-05,
+      "loss": 1.8553,
+      "step": 6722
+    },
+    {
+      "epoch": 0.7073119410836401,
+      "grad_norm": 1.9273593425750732,
+      "learning_rate": 4.1671756707530594e-05,
+      "loss": 1.2872,
+      "step": 6723
+    },
+    {
+      "epoch": 0.7074171488690163,
+      "grad_norm": 1.7082390785217285,
+      "learning_rate": 4.164408009015024e-05,
+      "loss": 1.8351,
+      "step": 6724
+    },
+    {
+      "epoch": 0.7075223566543924,
+      "grad_norm": 2.25231671333313,
+      "learning_rate": 4.1616410249450746e-05,
+      "loss": 1.7099,
+      "step": 6725
+    },
+    {
+      "epoch": 0.7076275644397686,
+      "grad_norm": 1.2057231664657593,
+      "learning_rate": 4.1588747188645275e-05,
+      "loss": 2.0397,
+      "step": 6726
+    },
+    {
+      "epoch": 0.7077327722251446,
+      "grad_norm": 1.857649564743042,
+      "learning_rate": 4.156109091094622e-05,
+      "loss": 1.8514,
+      "step": 6727
+    },
+    {
+      "epoch": 0.7078379800105208,
+      "grad_norm": 1.4656842947006226,
+      "learning_rate": 4.1533441419565265e-05,
+      "loss": 1.2047,
+      "step": 6728
+    },
+    {
+      "epoch": 0.7079431877958969,
+      "grad_norm": 1.4205302000045776,
+      "learning_rate": 4.150579871771324e-05,
+      "loss": 1.2295,
+      "step": 6729
+    },
+    {
+      "epoch": 0.708048395581273,
+      "grad_norm": 1.5245219469070435,
+      "learning_rate": 4.14781628086002e-05,
+      "loss": 1.7168,
+      "step": 6730
+    },
+    {
+      "epoch": 0.7081536033666491,
+      "grad_norm": 1.4015170335769653,
+      "learning_rate": 4.145053369543539e-05,
+      "loss": 1.6294,
+      "step": 6731
+    },
+    {
+      "epoch": 0.7082588111520253,
+      "grad_norm": 1.7922075986862183,
+      "learning_rate": 4.1422911381427274e-05,
+      "loss": 1.4777,
+      "step": 6732
+    },
+    {
+      "epoch": 0.7083640189374014,
+      "grad_norm": 1.0388702154159546,
+      "learning_rate": 4.1395295869783615e-05,
+      "loss": 1.8166,
+      "step": 6733
+    },
+    {
+      "epoch": 0.7084692267227775,
+      "grad_norm": 1.3340606689453125,
+      "learning_rate": 4.136768716371125e-05,
+      "loss": 1.4049,
+      "step": 6734
+    },
+    {
+      "epoch": 0.7085744345081536,
+      "grad_norm": 1.485906720161438,
+      "learning_rate": 4.134008526641628e-05,
+      "loss": 1.9171,
+      "step": 6735
+    },
+    {
+      "epoch": 0.7086796422935298,
+      "grad_norm": 1.4040229320526123,
+      "learning_rate": 4.131249018110408e-05,
+      "loss": 1.3881,
+      "step": 6736
+    },
+    {
+      "epoch": 0.7087848500789058,
+      "grad_norm": 2.063403844833374,
+      "learning_rate": 4.12849019109791e-05,
+      "loss": 1.8729,
+      "step": 6737
+    },
+    {
+      "epoch": 0.7088900578642819,
+      "grad_norm": 1.6362403631210327,
+      "learning_rate": 4.1257320459245154e-05,
+      "loss": 1.7883,
+      "step": 6738
+    },
+    {
+      "epoch": 0.7089952656496581,
+      "grad_norm": 1.6882222890853882,
+      "learning_rate": 4.122974582910515e-05,
+      "loss": 1.3546,
+      "step": 6739
+    },
+    {
+      "epoch": 0.7091004734350342,
+      "grad_norm": 1.3080660104751587,
+      "learning_rate": 4.1202178023761195e-05,
+      "loss": 1.4403,
+      "step": 6740
+    },
+    {
+      "epoch": 0.7092056812204103,
+      "grad_norm": 2.508234739303589,
+      "learning_rate": 4.117461704641473e-05,
+      "loss": 1.5945,
+      "step": 6741
+    },
+    {
+      "epoch": 0.7093108890057864,
+      "grad_norm": 1.9061312675476074,
+      "learning_rate": 4.1147062900266285e-05,
+      "loss": 1.9146,
+      "step": 6742
+    },
+    {
+      "epoch": 0.7094160967911626,
+      "grad_norm": 1.108188271522522,
+      "learning_rate": 4.111951558851559e-05,
+      "loss": 1.9639,
+      "step": 6743
+    },
+    {
+      "epoch": 0.7095213045765386,
+      "grad_norm": 1.4938721656799316,
+      "learning_rate": 4.109197511436169e-05,
+      "loss": 2.0542,
+      "step": 6744
+    },
+    {
+      "epoch": 0.7096265123619148,
+      "grad_norm": 1.5593281984329224,
+      "learning_rate": 4.106444148100275e-05,
+      "loss": 1.6947,
+      "step": 6745
+    },
+    {
+      "epoch": 0.7097317201472909,
+      "grad_norm": 1.297189712524414,
+      "learning_rate": 4.10369146916361e-05,
+      "loss": 1.4036,
+      "step": 6746
+    },
+    {
+      "epoch": 0.7098369279326671,
+      "grad_norm": 0.986399233341217,
+      "learning_rate": 4.100939474945843e-05,
+      "loss": 1.5753,
+      "step": 6747
+    },
+    {
+      "epoch": 0.7099421357180431,
+      "grad_norm": 1.9157464504241943,
+      "learning_rate": 4.09818816576655e-05,
+      "loss": 1.4842,
+      "step": 6748
+    },
+    {
+      "epoch": 0.7100473435034192,
+      "grad_norm": 1.0592901706695557,
+      "learning_rate": 4.095437541945226e-05,
+      "loss": 2.2621,
+      "step": 6749
+    },
+    {
+      "epoch": 0.7101525512887954,
+      "grad_norm": 1.2856818437576294,
+      "learning_rate": 4.0926876038013026e-05,
+      "loss": 1.7017,
+      "step": 6750
+    },
+    {
+      "epoch": 0.7102577590741714,
+      "grad_norm": 1.17574942111969,
+      "learning_rate": 4.0899383516541146e-05,
+      "loss": 1.4187,
+      "step": 6751
+    },
+    {
+      "epoch": 0.7103629668595476,
+      "grad_norm": 3.1928186416625977,
+      "learning_rate": 4.087189785822925e-05,
+      "loss": 1.4191,
+      "step": 6752
+    },
+    {
+      "epoch": 0.7104681746449237,
+      "grad_norm": 1.1456485986709595,
+      "learning_rate": 4.084441906626914e-05,
+      "loss": 1.5175,
+      "step": 6753
+    },
+    {
+      "epoch": 0.7105733824302999,
+      "grad_norm": 1.2433396577835083,
+      "learning_rate": 4.0816947143851816e-05,
+      "loss": 1.2899,
+      "step": 6754
+    },
+    {
+      "epoch": 0.7106785902156759,
+      "grad_norm": 1.4701285362243652,
+      "learning_rate": 4.078948209416758e-05,
+      "loss": 1.6928,
+      "step": 6755
+    },
+    {
+      "epoch": 0.7107837980010521,
+      "grad_norm": 1.0511400699615479,
+      "learning_rate": 4.0762023920405804e-05,
+      "loss": 1.7949,
+      "step": 6756
+    },
+    {
+      "epoch": 0.7108890057864282,
+      "grad_norm": 1.6696053743362427,
+      "learning_rate": 4.073457262575509e-05,
+      "loss": 1.991,
+      "step": 6757
+    },
+    {
+      "epoch": 0.7109942135718044,
+      "grad_norm": 1.6411384344100952,
+      "learning_rate": 4.0707128213403336e-05,
+      "loss": 1.3082,
+      "step": 6758
+    },
+    {
+      "epoch": 0.7110994213571804,
+      "grad_norm": 2.3022947311401367,
+      "learning_rate": 4.067969068653754e-05,
+      "loss": 1.8167,
+      "step": 6759
+    },
+    {
+      "epoch": 0.7112046291425566,
+      "grad_norm": 1.5651785135269165,
+      "learning_rate": 4.0652260048343885e-05,
+      "loss": 1.7786,
+      "step": 6760
+    },
+    {
+      "epoch": 0.7113098369279327,
+      "grad_norm": 1.68247389793396,
+      "learning_rate": 4.0624836302007886e-05,
+      "loss": 2.5257,
+      "step": 6761
+    },
+    {
+      "epoch": 0.7114150447133087,
+      "grad_norm": 1.1986908912658691,
+      "learning_rate": 4.059741945071412e-05,
+      "loss": 1.7782,
+      "step": 6762
+    },
+    {
+      "epoch": 0.7115202524986849,
+      "grad_norm": 1.7932888269424438,
+      "learning_rate": 4.05700094976464e-05,
+      "loss": 1.3013,
+      "step": 6763
+    },
+    {
+      "epoch": 0.711625460284061,
+      "grad_norm": 1.4314543008804321,
+      "learning_rate": 4.054260644598781e-05,
+      "loss": 1.3228,
+      "step": 6764
+    },
+    {
+      "epoch": 0.7117306680694372,
+      "grad_norm": 1.5555866956710815,
+      "learning_rate": 4.0515210298920545e-05,
+      "loss": 1.5795,
+      "step": 6765
+    },
+    {
+      "epoch": 0.7118358758548132,
+      "grad_norm": 1.2650176286697388,
+      "learning_rate": 4.048782105962598e-05,
+      "loss": 2.0545,
+      "step": 6766
+    },
+    {
+      "epoch": 0.7119410836401894,
+      "grad_norm": 1.1688004732131958,
+      "learning_rate": 4.0460438731284845e-05,
+      "loss": 1.5544,
+      "step": 6767
+    },
+    {
+      "epoch": 0.7120462914255655,
+      "grad_norm": 1.4285098314285278,
+      "learning_rate": 4.0433063317076893e-05,
+      "loss": 1.4884,
+      "step": 6768
+    },
+    {
+      "epoch": 0.7121514992109416,
+      "grad_norm": 2.4605870246887207,
+      "learning_rate": 4.0405694820181104e-05,
+      "loss": 1.0857,
+      "step": 6769
+    },
+    {
+      "epoch": 0.7122567069963177,
+      "grad_norm": 1.3824515342712402,
+      "learning_rate": 4.037833324377578e-05,
+      "loss": 1.6449,
+      "step": 6770
+    },
+    {
+      "epoch": 0.7123619147816939,
+      "grad_norm": 2.0455679893493652,
+      "learning_rate": 4.035097859103829e-05,
+      "loss": 1.7399,
+      "step": 6771
+    },
+    {
+      "epoch": 0.71246712256707,
+      "grad_norm": 1.9680933952331543,
+      "learning_rate": 4.032363086514523e-05,
+      "loss": 2.0078,
+      "step": 6772
+    },
+    {
+      "epoch": 0.712572330352446,
+      "grad_norm": 1.2317272424697876,
+      "learning_rate": 4.0296290069272416e-05,
+      "loss": 1.2191,
+      "step": 6773
+    },
+    {
+      "epoch": 0.7126775381378222,
+      "grad_norm": 1.2015796899795532,
+      "learning_rate": 4.026895620659479e-05,
+      "loss": 1.8345,
+      "step": 6774
+    },
+    {
+      "epoch": 0.7127827459231983,
+      "grad_norm": 2.4510412216186523,
+      "learning_rate": 4.024162928028663e-05,
+      "loss": 1.554,
+      "step": 6775
+    },
+    {
+      "epoch": 0.7128879537085744,
+      "grad_norm": 1.147552251815796,
+      "learning_rate": 4.021430929352128e-05,
+      "loss": 1.9372,
+      "step": 6776
+    },
+    {
+      "epoch": 0.7129931614939505,
+      "grad_norm": 1.3203489780426025,
+      "learning_rate": 4.01869962494713e-05,
+      "loss": 1.7651,
+      "step": 6777
+    },
+    {
+      "epoch": 0.7130983692793267,
+      "grad_norm": 2.268237829208374,
+      "learning_rate": 4.0159690151308504e-05,
+      "loss": 1.2124,
+      "step": 6778
+    },
+    {
+      "epoch": 0.7132035770647028,
+      "grad_norm": 1.9930332899093628,
+      "learning_rate": 4.013239100220385e-05,
+      "loss": 1.3614,
+      "step": 6779
+    },
+    {
+      "epoch": 0.7133087848500789,
+      "grad_norm": 1.6461526155471802,
+      "learning_rate": 4.010509880532745e-05,
+      "loss": 1.8164,
+      "step": 6780
+    },
+    {
+      "epoch": 0.713413992635455,
+      "grad_norm": 1.5002760887145996,
+      "learning_rate": 4.007781356384873e-05,
+      "loss": 2.0341,
+      "step": 6781
+    },
+    {
+      "epoch": 0.7135192004208312,
+      "grad_norm": 2.1459338665008545,
+      "learning_rate": 4.0050535280936205e-05,
+      "loss": 1.5856,
+      "step": 6782
+    },
+    {
+      "epoch": 0.7136244082062072,
+      "grad_norm": 1.3347117900848389,
+      "learning_rate": 4.002326395975758e-05,
+      "loss": 1.6578,
+      "step": 6783
+    },
+    {
+      "epoch": 0.7137296159915834,
+      "grad_norm": 2.3518266677856445,
+      "learning_rate": 3.9995999603479836e-05,
+      "loss": 1.6109,
+      "step": 6784
+    },
+    {
+      "epoch": 0.7138348237769595,
+      "grad_norm": 1.601401686668396,
+      "learning_rate": 3.996874221526905e-05,
+      "loss": 1.5894,
+      "step": 6785
+    },
+    {
+      "epoch": 0.7139400315623357,
+      "grad_norm": 2.32871150970459,
+      "learning_rate": 3.994149179829058e-05,
+      "loss": 1.2623,
+      "step": 6786
+    },
+    {
+      "epoch": 0.7140452393477117,
+      "grad_norm": 1.1024394035339355,
+      "learning_rate": 3.9914248355708894e-05,
+      "loss": 2.0739,
+      "step": 6787
+    },
+    {
+      "epoch": 0.7141504471330878,
+      "grad_norm": 1.1708279848098755,
+      "learning_rate": 3.988701189068765e-05,
+      "loss": 1.6601,
+      "step": 6788
+    },
+    {
+      "epoch": 0.714255654918464,
+      "grad_norm": 1.7237240076065063,
+      "learning_rate": 3.985978240638981e-05,
+      "loss": 1.7179,
+      "step": 6789
+    },
+    {
+      "epoch": 0.7143608627038401,
+      "grad_norm": 1.493378758430481,
+      "learning_rate": 3.9832559905977404e-05,
+      "loss": 1.97,
+      "step": 6790
+    },
+    {
+      "epoch": 0.7144660704892162,
+      "grad_norm": 1.8581169843673706,
+      "learning_rate": 3.9805344392611653e-05,
+      "loss": 1.7689,
+      "step": 6791
+    },
+    {
+      "epoch": 0.7145712782745923,
+      "grad_norm": 1.4340283870697021,
+      "learning_rate": 3.977813586945308e-05,
+      "loss": 1.2957,
+      "step": 6792
+    },
+    {
+      "epoch": 0.7146764860599685,
+      "grad_norm": 1.2262341976165771,
+      "learning_rate": 3.9750934339661275e-05,
+      "loss": 1.7558,
+      "step": 6793
+    },
+    {
+      "epoch": 0.7147816938453445,
+      "grad_norm": 2.0445914268493652,
+      "learning_rate": 3.972373980639508e-05,
+      "loss": 2.3494,
+      "step": 6794
+    },
+    {
+      "epoch": 0.7148869016307207,
+      "grad_norm": 2.4502785205841064,
+      "learning_rate": 3.9696552272812484e-05,
+      "loss": 2.2171,
+      "step": 6795
+    },
+    {
+      "epoch": 0.7149921094160968,
+      "grad_norm": 1.6572961807250977,
+      "learning_rate": 3.966937174207066e-05,
+      "loss": 1.8123,
+      "step": 6796
+    },
+    {
+      "epoch": 0.715097317201473,
+      "grad_norm": 2.4192516803741455,
+      "learning_rate": 3.9642198217326075e-05,
+      "loss": 1.5431,
+      "step": 6797
+    },
+    {
+      "epoch": 0.715202524986849,
+      "grad_norm": 2.080559253692627,
+      "learning_rate": 3.961503170173426e-05,
+      "loss": 1.6864,
+      "step": 6798
+    },
+    {
+      "epoch": 0.7153077327722251,
+      "grad_norm": 2.1489667892456055,
+      "learning_rate": 3.958787219844994e-05,
+      "loss": 1.4183,
+      "step": 6799
+    },
+    {
+      "epoch": 0.7154129405576013,
+      "grad_norm": 0.9947004914283752,
+      "learning_rate": 3.9560719710627115e-05,
+      "loss": 1.6476,
+      "step": 6800
+    },
+    {
+      "epoch": 0.7155181483429773,
+      "grad_norm": 1.2469807863235474,
+      "learning_rate": 3.9533574241418884e-05,
+      "loss": 1.5677,
+      "step": 6801
+    },
+    {
+      "epoch": 0.7156233561283535,
+      "grad_norm": 1.3769633769989014,
+      "learning_rate": 3.9506435793977535e-05,
+      "loss": 1.7477,
+      "step": 6802
+    },
+    {
+      "epoch": 0.7157285639137296,
+      "grad_norm": 1.638057827949524,
+      "learning_rate": 3.947930437145464e-05,
+      "loss": 1.2805,
+      "step": 6803
+    },
+    {
+      "epoch": 0.7158337716991058,
+      "grad_norm": 1.301336646080017,
+      "learning_rate": 3.9452179977000826e-05,
+      "loss": 1.4646,
+      "step": 6804
+    },
+    {
+      "epoch": 0.7159389794844818,
+      "grad_norm": 1.5804367065429688,
+      "learning_rate": 3.942506261376594e-05,
+      "loss": 2.0871,
+      "step": 6805
+    },
+    {
+      "epoch": 0.716044187269858,
+      "grad_norm": 1.6908665895462036,
+      "learning_rate": 3.93979522848991e-05,
+      "loss": 1.9848,
+      "step": 6806
+    },
+    {
+      "epoch": 0.7161493950552341,
+      "grad_norm": 1.653495192527771,
+      "learning_rate": 3.93708489935485e-05,
+      "loss": 2.2619,
+      "step": 6807
+    },
+    {
+      "epoch": 0.7162546028406102,
+      "grad_norm": 1.4094468355178833,
+      "learning_rate": 3.934375274286154e-05,
+      "loss": 1.3118,
+      "step": 6808
+    },
+    {
+      "epoch": 0.7163598106259863,
+      "grad_norm": 1.1481539011001587,
+      "learning_rate": 3.931666353598485e-05,
+      "loss": 1.5423,
+      "step": 6809
+    },
+    {
+      "epoch": 0.7164650184113625,
+      "grad_norm": 1.4447335004806519,
+      "learning_rate": 3.928958137606421e-05,
+      "loss": 1.7084,
+      "step": 6810
+    },
+    {
+      "epoch": 0.7165702261967386,
+      "grad_norm": 1.6430801153182983,
+      "learning_rate": 3.9262506266244534e-05,
+      "loss": 1.5448,
+      "step": 6811
+    },
+    {
+      "epoch": 0.7166754339821146,
+      "grad_norm": 1.7120585441589355,
+      "learning_rate": 3.923543820967004e-05,
+      "loss": 1.8249,
+      "step": 6812
+    },
+    {
+      "epoch": 0.7167806417674908,
+      "grad_norm": 1.6759033203125,
+      "learning_rate": 3.9208377209484014e-05,
+      "loss": 1.4926,
+      "step": 6813
+    },
+    {
+      "epoch": 0.7168858495528669,
+      "grad_norm": 1.6250011920928955,
+      "learning_rate": 3.918132326882892e-05,
+      "loss": 1.7885,
+      "step": 6814
+    },
+    {
+      "epoch": 0.716991057338243,
+      "grad_norm": 1.3283613920211792,
+      "learning_rate": 3.9154276390846555e-05,
+      "loss": 1.2566,
+      "step": 6815
+    },
+    {
+      "epoch": 0.7170962651236191,
+      "grad_norm": 1.6472797393798828,
+      "learning_rate": 3.912723657867764e-05,
+      "loss": 1.2617,
+      "step": 6816
+    },
+    {
+      "epoch": 0.7172014729089953,
+      "grad_norm": 1.686281442642212,
+      "learning_rate": 3.910020383546233e-05,
+      "loss": 1.7821,
+      "step": 6817
+    },
+    {
+      "epoch": 0.7173066806943714,
+      "grad_norm": 1.3786574602127075,
+      "learning_rate": 3.907317816433982e-05,
+      "loss": 1.6652,
+      "step": 6818
+    },
+    {
+      "epoch": 0.7174118884797475,
+      "grad_norm": 1.1315668821334839,
+      "learning_rate": 3.9046159568448446e-05,
+      "loss": 1.6919,
+      "step": 6819
+    },
+    {
+      "epoch": 0.7175170962651236,
+      "grad_norm": 1.1525264978408813,
+      "learning_rate": 3.90191480509259e-05,
+      "loss": 2.0687,
+      "step": 6820
+    },
+    {
+      "epoch": 0.7176223040504998,
+      "grad_norm": 2.229443311691284,
+      "learning_rate": 3.8992143614908874e-05,
+      "loss": 1.6294,
+      "step": 6821
+    },
+    {
+      "epoch": 0.7177275118358759,
+      "grad_norm": 1.3578799962997437,
+      "learning_rate": 3.896514626353328e-05,
+      "loss": 1.0833,
+      "step": 6822
+    },
+    {
+      "epoch": 0.717832719621252,
+      "grad_norm": 1.2835088968276978,
+      "learning_rate": 3.893815599993431e-05,
+      "loss": 2.0935,
+      "step": 6823
+    },
+    {
+      "epoch": 0.7179379274066281,
+      "grad_norm": 1.247178077697754,
+      "learning_rate": 3.8911172827246215e-05,
+      "loss": 1.4496,
+      "step": 6824
+    },
+    {
+      "epoch": 0.7180431351920042,
+      "grad_norm": 1.1701315641403198,
+      "learning_rate": 3.888419674860241e-05,
+      "loss": 1.8054,
+      "step": 6825
+    },
+    {
+      "epoch": 0.7181483429773803,
+      "grad_norm": 1.4860901832580566,
+      "learning_rate": 3.885722776713563e-05,
+      "loss": 1.6664,
+      "step": 6826
+    },
+    {
+      "epoch": 0.7182535507627564,
+      "grad_norm": 2.2401113510131836,
+      "learning_rate": 3.8830265885977656e-05,
+      "loss": 1.5906,
+      "step": 6827
+    },
+    {
+      "epoch": 0.7183587585481326,
+      "grad_norm": 1.4506216049194336,
+      "learning_rate": 3.8803311108259435e-05,
+      "loss": 1.4774,
+      "step": 6828
+    },
+    {
+      "epoch": 0.7184639663335087,
+      "grad_norm": 1.3430507183074951,
+      "learning_rate": 3.877636343711122e-05,
+      "loss": 1.604,
+      "step": 6829
+    },
+    {
+      "epoch": 0.7185691741188848,
+      "grad_norm": 1.241802453994751,
+      "learning_rate": 3.874942287566227e-05,
+      "loss": 1.391,
+      "step": 6830
+    },
+    {
+      "epoch": 0.7186743819042609,
+      "grad_norm": 1.9131029844284058,
+      "learning_rate": 3.8722489427041185e-05,
+      "loss": 2.0294,
+      "step": 6831
+    },
+    {
+      "epoch": 0.7187795896896371,
+      "grad_norm": 1.5295307636260986,
+      "learning_rate": 3.869556309437563e-05,
+      "loss": 1.2301,
+      "step": 6832
+    },
+    {
+      "epoch": 0.7188847974750131,
+      "grad_norm": 1.5379595756530762,
+      "learning_rate": 3.866864388079242e-05,
+      "loss": 1.8451,
+      "step": 6833
+    },
+    {
+      "epoch": 0.7189900052603893,
+      "grad_norm": 1.5962355136871338,
+      "learning_rate": 3.864173178941767e-05,
+      "loss": 1.1769,
+      "step": 6834
+    },
+    {
+      "epoch": 0.7190952130457654,
+      "grad_norm": 1.6853702068328857,
+      "learning_rate": 3.8614826823376557e-05,
+      "loss": 1.1957,
+      "step": 6835
+    },
+    {
+      "epoch": 0.7192004208311416,
+      "grad_norm": 1.4030197858810425,
+      "learning_rate": 3.858792898579348e-05,
+      "loss": 1.0337,
+      "step": 6836
+    },
+    {
+      "epoch": 0.7193056286165176,
+      "grad_norm": 1.4564900398254395,
+      "learning_rate": 3.8561038279791974e-05,
+      "loss": 1.8006,
+      "step": 6837
+    },
+    {
+      "epoch": 0.7194108364018937,
+      "grad_norm": 1.0711987018585205,
+      "learning_rate": 3.853415470849479e-05,
+      "loss": 1.8206,
+      "step": 6838
+    },
+    {
+      "epoch": 0.7195160441872699,
+      "grad_norm": 1.8588398694992065,
+      "learning_rate": 3.8507278275023774e-05,
+      "loss": 1.295,
+      "step": 6839
+    },
+    {
+      "epoch": 0.7196212519726459,
+      "grad_norm": 2.0701515674591064,
+      "learning_rate": 3.848040898250007e-05,
+      "loss": 1.3578,
+      "step": 6840
+    },
+    {
+      "epoch": 0.7197264597580221,
+      "grad_norm": 1.5314396619796753,
+      "learning_rate": 3.8453546834043866e-05,
+      "loss": 1.7377,
+      "step": 6841
+    },
+    {
+      "epoch": 0.7198316675433982,
+      "grad_norm": 1.5056813955307007,
+      "learning_rate": 3.842669183277463e-05,
+      "loss": 1.7067,
+      "step": 6842
+    },
+    {
+      "epoch": 0.7199368753287744,
+      "grad_norm": 2.3147194385528564,
+      "learning_rate": 3.839984398181092e-05,
+      "loss": 1.9101,
+      "step": 6843
+    },
+    {
+      "epoch": 0.7200420831141504,
+      "grad_norm": 2.0493955612182617,
+      "learning_rate": 3.8373003284270445e-05,
+      "loss": 1.6542,
+      "step": 6844
+    },
+    {
+      "epoch": 0.7201472908995266,
+      "grad_norm": 1.2979793548583984,
+      "learning_rate": 3.834616974327021e-05,
+      "loss": 1.5933,
+      "step": 6845
+    },
+    {
+      "epoch": 0.7202524986849027,
+      "grad_norm": 1.7987275123596191,
+      "learning_rate": 3.831934336192625e-05,
+      "loss": 1.5552,
+      "step": 6846
+    },
+    {
+      "epoch": 0.7203577064702787,
+      "grad_norm": 1.7454094886779785,
+      "learning_rate": 3.829252414335381e-05,
+      "loss": 1.7823,
+      "step": 6847
+    },
+    {
+      "epoch": 0.7204629142556549,
+      "grad_norm": 1.5051946640014648,
+      "learning_rate": 3.826571209066737e-05,
+      "loss": 1.4311,
+      "step": 6848
+    },
+    {
+      "epoch": 0.720568122041031,
+      "grad_norm": 1.2123810052871704,
+      "learning_rate": 3.8238907206980513e-05,
+      "loss": 1.8447,
+      "step": 6849
+    },
+    {
+      "epoch": 0.7206733298264072,
+      "grad_norm": 1.1441395282745361,
+      "learning_rate": 3.821210949540593e-05,
+      "loss": 1.6574,
+      "step": 6850
+    },
+    {
+      "epoch": 0.7207785376117832,
+      "grad_norm": 1.625401258468628,
+      "learning_rate": 3.818531895905566e-05,
+      "loss": 1.9151,
+      "step": 6851
+    },
+    {
+      "epoch": 0.7208837453971594,
+      "grad_norm": 2.0465898513793945,
+      "learning_rate": 3.815853560104075e-05,
+      "loss": 1.8502,
+      "step": 6852
+    },
+    {
+      "epoch": 0.7209889531825355,
+      "grad_norm": 1.7587473392486572,
+      "learning_rate": 3.813175942447141e-05,
+      "loss": 1.6247,
+      "step": 6853
+    },
+    {
+      "epoch": 0.7210941609679117,
+      "grad_norm": 3.3340325355529785,
+      "learning_rate": 3.810499043245718e-05,
+      "loss": 1.1276,
+      "step": 6854
+    },
+    {
+      "epoch": 0.7211993687532877,
+      "grad_norm": 1.4002957344055176,
+      "learning_rate": 3.807822862810657e-05,
+      "loss": 1.452,
+      "step": 6855
+    },
+    {
+      "epoch": 0.7213045765386639,
+      "grad_norm": 1.8713266849517822,
+      "learning_rate": 3.805147401452734e-05,
+      "loss": 1.9372,
+      "step": 6856
+    },
+    {
+      "epoch": 0.72140978432404,
+      "grad_norm": 2.1480062007904053,
+      "learning_rate": 3.802472659482649e-05,
+      "loss": 1.6592,
+      "step": 6857
+    },
+    {
+      "epoch": 0.721514992109416,
+      "grad_norm": 0.9366277456283569,
+      "learning_rate": 3.799798637211005e-05,
+      "loss": 1.5559,
+      "step": 6858
+    },
+    {
+      "epoch": 0.7216201998947922,
+      "grad_norm": 1.6962757110595703,
+      "learning_rate": 3.7971253349483285e-05,
+      "loss": 1.6838,
+      "step": 6859
+    },
+    {
+      "epoch": 0.7217254076801684,
+      "grad_norm": 1.3592772483825684,
+      "learning_rate": 3.794452753005061e-05,
+      "loss": 1.8806,
+      "step": 6860
+    },
+    {
+      "epoch": 0.7218306154655445,
+      "grad_norm": 2.0401268005371094,
+      "learning_rate": 3.791780891691558e-05,
+      "loss": 2.4508,
+      "step": 6861
+    },
+    {
+      "epoch": 0.7219358232509205,
+      "grad_norm": 1.0858590602874756,
+      "learning_rate": 3.7891097513180995e-05,
+      "loss": 2.0971,
+      "step": 6862
+    },
+    {
+      "epoch": 0.7220410310362967,
+      "grad_norm": 1.913306713104248,
+      "learning_rate": 3.786439332194874e-05,
+      "loss": 1.1957,
+      "step": 6863
+    },
+    {
+      "epoch": 0.7221462388216728,
+      "grad_norm": 2.756697416305542,
+      "learning_rate": 3.783769634631985e-05,
+      "loss": 1.5992,
+      "step": 6864
+    },
+    {
+      "epoch": 0.7222514466070489,
+      "grad_norm": 1.9628227949142456,
+      "learning_rate": 3.781100658939461e-05,
+      "loss": 2.0914,
+      "step": 6865
+    },
+    {
+      "epoch": 0.722356654392425,
+      "grad_norm": 1.16515052318573,
+      "learning_rate": 3.7784324054272405e-05,
+      "loss": 1.7869,
+      "step": 6866
+    },
+    {
+      "epoch": 0.7224618621778012,
+      "grad_norm": 1.3066214323043823,
+      "learning_rate": 3.7757648744051736e-05,
+      "loss": 2.3816,
+      "step": 6867
+    },
+    {
+      "epoch": 0.7225670699631773,
+      "grad_norm": 1.6055272817611694,
+      "learning_rate": 3.773098066183039e-05,
+      "loss": 1.3618,
+      "step": 6868
+    },
+    {
+      "epoch": 0.7226722777485534,
+      "grad_norm": 1.2972073554992676,
+      "learning_rate": 3.7704319810705225e-05,
+      "loss": 1.9931,
+      "step": 6869
+    },
+    {
+      "epoch": 0.7227774855339295,
+      "grad_norm": 1.3764903545379639,
+      "learning_rate": 3.767766619377222e-05,
+      "loss": 2.0662,
+      "step": 6870
+    },
+    {
+      "epoch": 0.7228826933193057,
+      "grad_norm": 1.6971714496612549,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 1.4372,
+      "step": 6871
+    },
+    {
+      "epoch": 0.7229879011046817,
+      "grad_norm": 2.1044673919677734,
+      "learning_rate": 3.7624380674862845e-05,
+      "loss": 2.0204,
+      "step": 6872
+    },
+    {
+      "epoch": 0.7230931088900578,
+      "grad_norm": 1.2413538694381714,
+      "learning_rate": 3.759774877907428e-05,
+      "loss": 1.8452,
+      "step": 6873
+    },
+    {
+      "epoch": 0.723198316675434,
+      "grad_norm": 1.1550841331481934,
+      "learning_rate": 3.75711241298537e-05,
+      "loss": 1.655,
+      "step": 6874
+    },
+    {
+      "epoch": 0.7233035244608101,
+      "grad_norm": 1.314095139503479,
+      "learning_rate": 3.75445067302929e-05,
+      "loss": 1.648,
+      "step": 6875
+    },
+    {
+      "epoch": 0.7234087322461862,
+      "grad_norm": 1.6076027154922485,
+      "learning_rate": 3.751789658348284e-05,
+      "loss": 1.8427,
+      "step": 6876
+    },
+    {
+      "epoch": 0.7235139400315623,
+      "grad_norm": 1.512660264968872,
+      "learning_rate": 3.749129369251372e-05,
+      "loss": 2.0951,
+      "step": 6877
+    },
+    {
+      "epoch": 0.7236191478169385,
+      "grad_norm": 1.5200482606887817,
+      "learning_rate": 3.7464698060474814e-05,
+      "loss": 1.3188,
+      "step": 6878
+    },
+    {
+      "epoch": 0.7237243556023145,
+      "grad_norm": 1.276044249534607,
+      "learning_rate": 3.7438109690454646e-05,
+      "loss": 1.5648,
+      "step": 6879
+    },
+    {
+      "epoch": 0.7238295633876907,
+      "grad_norm": 1.4830344915390015,
+      "learning_rate": 3.741152858554077e-05,
+      "loss": 1.3373,
+      "step": 6880
+    },
+    {
+      "epoch": 0.7239347711730668,
+      "grad_norm": 1.0072591304779053,
+      "learning_rate": 3.7384954748819934e-05,
+      "loss": 1.9321,
+      "step": 6881
+    },
+    {
+      "epoch": 0.724039978958443,
+      "grad_norm": 1.2842603921890259,
+      "learning_rate": 3.7358388183378166e-05,
+      "loss": 1.791,
+      "step": 6882
+    },
+    {
+      "epoch": 0.724145186743819,
+      "grad_norm": 2.4097936153411865,
+      "learning_rate": 3.733182889230049e-05,
+      "loss": 1.6026,
+      "step": 6883
+    },
+    {
+      "epoch": 0.7242503945291952,
+      "grad_norm": 1.9151344299316406,
+      "learning_rate": 3.730527687867114e-05,
+      "loss": 1.7359,
+      "step": 6884
+    },
+    {
+      "epoch": 0.7243556023145713,
+      "grad_norm": 1.9665045738220215,
+      "learning_rate": 3.727873214557357e-05,
+      "loss": 2.2739,
+      "step": 6885
+    },
+    {
+      "epoch": 0.7244608100999474,
+      "grad_norm": 1.395699143409729,
+      "learning_rate": 3.725219469609026e-05,
+      "loss": 1.382,
+      "step": 6886
+    },
+    {
+      "epoch": 0.7245660178853235,
+      "grad_norm": 2.1957099437713623,
+      "learning_rate": 3.722566453330298e-05,
+      "loss": 1.5297,
+      "step": 6887
+    },
+    {
+      "epoch": 0.7246712256706996,
+      "grad_norm": 2.5647594928741455,
+      "learning_rate": 3.7199141660292594e-05,
+      "loss": 1.7129,
+      "step": 6888
+    },
+    {
+      "epoch": 0.7247764334560758,
+      "grad_norm": 1.1315737962722778,
+      "learning_rate": 3.717262608013903e-05,
+      "loss": 1.2686,
+      "step": 6889
+    },
+    {
+      "epoch": 0.7248816412414518,
+      "grad_norm": 2.1699776649475098,
+      "learning_rate": 3.714611779592156e-05,
+      "loss": 2.1215,
+      "step": 6890
+    },
+    {
+      "epoch": 0.724986849026828,
+      "grad_norm": 1.1801518201828003,
+      "learning_rate": 3.711961681071845e-05,
+      "loss": 2.093,
+      "step": 6891
+    },
+    {
+      "epoch": 0.7250920568122041,
+      "grad_norm": 1.2889292240142822,
+      "learning_rate": 3.7093123127607155e-05,
+      "loss": 1.5338,
+      "step": 6892
+    },
+    {
+      "epoch": 0.7251972645975803,
+      "grad_norm": 1.0799230337142944,
+      "learning_rate": 3.706663674966435e-05,
+      "loss": 1.2338,
+      "step": 6893
+    },
+    {
+      "epoch": 0.7253024723829563,
+      "grad_norm": 0.9696769714355469,
+      "learning_rate": 3.7040157679965796e-05,
+      "loss": 1.7135,
+      "step": 6894
+    },
+    {
+      "epoch": 0.7254076801683325,
+      "grad_norm": 1.161821961402893,
+      "learning_rate": 3.701368592158636e-05,
+      "loss": 1.9596,
+      "step": 6895
+    },
+    {
+      "epoch": 0.7255128879537086,
+      "grad_norm": 1.847954511642456,
+      "learning_rate": 3.6987221477600206e-05,
+      "loss": 1.9575,
+      "step": 6896
+    },
+    {
+      "epoch": 0.7256180957390846,
+      "grad_norm": 1.5871635675430298,
+      "learning_rate": 3.6960764351080535e-05,
+      "loss": 1.4081,
+      "step": 6897
+    },
+    {
+      "epoch": 0.7257233035244608,
+      "grad_norm": 2.1948516368865967,
+      "learning_rate": 3.6934314545099666e-05,
+      "loss": 1.8729,
+      "step": 6898
+    },
+    {
+      "epoch": 0.7258285113098369,
+      "grad_norm": 1.237642765045166,
+      "learning_rate": 3.690787206272923e-05,
+      "loss": 1.4971,
+      "step": 6899
+    },
+    {
+      "epoch": 0.7259337190952131,
+      "grad_norm": 2.1544601917266846,
+      "learning_rate": 3.6881436907039845e-05,
+      "loss": 1.6227,
+      "step": 6900
+    },
+    {
+      "epoch": 0.7260389268805891,
+      "grad_norm": 1.725161075592041,
+      "learning_rate": 3.6855009081101355e-05,
+      "loss": 2.0619,
+      "step": 6901
+    },
+    {
+      "epoch": 0.7261441346659653,
+      "grad_norm": 2.4666850566864014,
+      "learning_rate": 3.6828588587982715e-05,
+      "loss": 1.9632,
+      "step": 6902
+    },
+    {
+      "epoch": 0.7262493424513414,
+      "grad_norm": 1.0997231006622314,
+      "learning_rate": 3.680217543075204e-05,
+      "loss": 1.6217,
+      "step": 6903
+    },
+    {
+      "epoch": 0.7263545502367175,
+      "grad_norm": 1.5529903173446655,
+      "learning_rate": 3.6775769612476666e-05,
+      "loss": 1.8783,
+      "step": 6904
+    },
+    {
+      "epoch": 0.7264597580220936,
+      "grad_norm": 1.4422789812088013,
+      "learning_rate": 3.674937113622297e-05,
+      "loss": 1.4074,
+      "step": 6905
+    },
+    {
+      "epoch": 0.7265649658074698,
+      "grad_norm": 1.7190214395523071,
+      "learning_rate": 3.6722980005056474e-05,
+      "loss": 1.3818,
+      "step": 6906
+    },
+    {
+      "epoch": 0.7266701735928459,
+      "grad_norm": 1.7955530881881714,
+      "learning_rate": 3.669659622204199e-05,
+      "loss": 2.1845,
+      "step": 6907
+    },
+    {
+      "epoch": 0.726775381378222,
+      "grad_norm": 2.1686935424804688,
+      "learning_rate": 3.6670219790243344e-05,
+      "loss": 1.2339,
+      "step": 6908
+    },
+    {
+      "epoch": 0.7268805891635981,
+      "grad_norm": 1.332140326499939,
+      "learning_rate": 3.664385071272348e-05,
+      "loss": 1.6447,
+      "step": 6909
+    },
+    {
+      "epoch": 0.7269857969489742,
+      "grad_norm": 2.21612548828125,
+      "learning_rate": 3.6617488992544656e-05,
+      "loss": 1.5974,
+      "step": 6910
+    },
+    {
+      "epoch": 0.7270910047343503,
+      "grad_norm": 1.5317012071609497,
+      "learning_rate": 3.659113463276812e-05,
+      "loss": 1.6033,
+      "step": 6911
+    },
+    {
+      "epoch": 0.7271962125197264,
+      "grad_norm": 1.3678877353668213,
+      "learning_rate": 3.656478763645428e-05,
+      "loss": 1.4386,
+      "step": 6912
+    },
+    {
+      "epoch": 0.7273014203051026,
+      "grad_norm": 1.0950294733047485,
+      "learning_rate": 3.6538448006662795e-05,
+      "loss": 1.7116,
+      "step": 6913
+    },
+    {
+      "epoch": 0.7274066280904787,
+      "grad_norm": 1.2202073335647583,
+      "learning_rate": 3.6512115746452366e-05,
+      "loss": 1.0636,
+      "step": 6914
+    },
+    {
+      "epoch": 0.7275118358758548,
+      "grad_norm": 1.1715128421783447,
+      "learning_rate": 3.648579085888085e-05,
+      "loss": 1.5944,
+      "step": 6915
+    },
+    {
+      "epoch": 0.7276170436612309,
+      "grad_norm": 0.9897493720054626,
+      "learning_rate": 3.6459473347005334e-05,
+      "loss": 1.6186,
+      "step": 6916
+    },
+    {
+      "epoch": 0.7277222514466071,
+      "grad_norm": 1.9510959386825562,
+      "learning_rate": 3.643316321388194e-05,
+      "loss": 1.1786,
+      "step": 6917
+    },
+    {
+      "epoch": 0.7278274592319832,
+      "grad_norm": 1.5720961093902588,
+      "learning_rate": 3.640686046256594e-05,
+      "loss": 1.6419,
+      "step": 6918
+    },
+    {
+      "epoch": 0.7279326670173593,
+      "grad_norm": 1.484616994857788,
+      "learning_rate": 3.6380565096111866e-05,
+      "loss": 1.5632,
+      "step": 6919
+    },
+    {
+      "epoch": 0.7280378748027354,
+      "grad_norm": 1.5460129976272583,
+      "learning_rate": 3.6354277117573264e-05,
+      "loss": 1.5821,
+      "step": 6920
+    },
+    {
+      "epoch": 0.7281430825881116,
+      "grad_norm": 1.5257236957550049,
+      "learning_rate": 3.632799653000286e-05,
+      "loss": 1.3601,
+      "step": 6921
+    },
+    {
+      "epoch": 0.7282482903734876,
+      "grad_norm": 1.349097490310669,
+      "learning_rate": 3.630172333645261e-05,
+      "loss": 1.8427,
+      "step": 6922
+    },
+    {
+      "epoch": 0.7283534981588637,
+      "grad_norm": 1.3090531826019287,
+      "learning_rate": 3.627545753997341e-05,
+      "loss": 1.4297,
+      "step": 6923
+    },
+    {
+      "epoch": 0.7284587059442399,
+      "grad_norm": 1.86859130859375,
+      "learning_rate": 3.624919914361552e-05,
+      "loss": 1.3909,
+      "step": 6924
+    },
+    {
+      "epoch": 0.728563913729616,
+      "grad_norm": 2.1052424907684326,
+      "learning_rate": 3.622294815042821e-05,
+      "loss": 1.4364,
+      "step": 6925
+    },
+    {
+      "epoch": 0.7286691215149921,
+      "grad_norm": 1.1871145963668823,
+      "learning_rate": 3.6196704563459885e-05,
+      "loss": 1.5061,
+      "step": 6926
+    },
+    {
+      "epoch": 0.7287743293003682,
+      "grad_norm": 2.0205655097961426,
+      "learning_rate": 3.617046838575819e-05,
+      "loss": 1.5433,
+      "step": 6927
+    },
+    {
+      "epoch": 0.7288795370857444,
+      "grad_norm": 1.3786308765411377,
+      "learning_rate": 3.6144239620369816e-05,
+      "loss": 1.7252,
+      "step": 6928
+    },
+    {
+      "epoch": 0.7289847448711204,
+      "grad_norm": 1.8365287780761719,
+      "learning_rate": 3.611801827034059e-05,
+      "loss": 1.3477,
+      "step": 6929
+    },
+    {
+      "epoch": 0.7290899526564966,
+      "grad_norm": 2.479865789413452,
+      "learning_rate": 3.609180433871558e-05,
+      "loss": 1.362,
+      "step": 6930
+    },
+    {
+      "epoch": 0.7291951604418727,
+      "grad_norm": 1.4172632694244385,
+      "learning_rate": 3.606559782853889e-05,
+      "loss": 1.4992,
+      "step": 6931
+    },
+    {
+      "epoch": 0.7293003682272489,
+      "grad_norm": 1.7480148077011108,
+      "learning_rate": 3.603939874285375e-05,
+      "loss": 1.7928,
+      "step": 6932
+    },
+    {
+      "epoch": 0.7294055760126249,
+      "grad_norm": 2.404942512512207,
+      "learning_rate": 3.6013207084702646e-05,
+      "loss": 1.1866,
+      "step": 6933
+    },
+    {
+      "epoch": 0.729510783798001,
+      "grad_norm": 1.3613299131393433,
+      "learning_rate": 3.5987022857127086e-05,
+      "loss": 1.1066,
+      "step": 6934
+    },
+    {
+      "epoch": 0.7296159915833772,
+      "grad_norm": 2.4102582931518555,
+      "learning_rate": 3.596084606316778e-05,
+      "loss": 1.5663,
+      "step": 6935
+    },
+    {
+      "epoch": 0.7297211993687532,
+      "grad_norm": 1.294023036956787,
+      "learning_rate": 3.593467670586457e-05,
+      "loss": 1.8498,
+      "step": 6936
+    },
+    {
+      "epoch": 0.7298264071541294,
+      "grad_norm": 1.4353657960891724,
+      "learning_rate": 3.5908514788256344e-05,
+      "loss": 2.0498,
+      "step": 6937
+    },
+    {
+      "epoch": 0.7299316149395055,
+      "grad_norm": 2.242220401763916,
+      "learning_rate": 3.588236031338129e-05,
+      "loss": 1.8686,
+      "step": 6938
+    },
+    {
+      "epoch": 0.7300368227248817,
+      "grad_norm": 1.6816960573196411,
+      "learning_rate": 3.585621328427658e-05,
+      "loss": 1.8161,
+      "step": 6939
+    },
+    {
+      "epoch": 0.7301420305102577,
+      "grad_norm": 1.5004388093948364,
+      "learning_rate": 3.58300737039786e-05,
+      "loss": 1.7814,
+      "step": 6940
+    },
+    {
+      "epoch": 0.7302472382956339,
+      "grad_norm": 2.273651599884033,
+      "learning_rate": 3.580394157552286e-05,
+      "loss": 1.9734,
+      "step": 6941
+    },
+    {
+      "epoch": 0.73035244608101,
+      "grad_norm": 1.1475566625595093,
+      "learning_rate": 3.577781690194399e-05,
+      "loss": 1.4193,
+      "step": 6942
+    },
+    {
+      "epoch": 0.7304576538663861,
+      "grad_norm": 1.948401927947998,
+      "learning_rate": 3.5751699686275786e-05,
+      "loss": 1.8889,
+      "step": 6943
+    },
+    {
+      "epoch": 0.7305628616517622,
+      "grad_norm": 2.96512770652771,
+      "learning_rate": 3.572558993155112e-05,
+      "loss": 2.0799,
+      "step": 6944
+    },
+    {
+      "epoch": 0.7306680694371384,
+      "grad_norm": 1.6030654907226562,
+      "learning_rate": 3.569948764080201e-05,
+      "loss": 1.497,
+      "step": 6945
+    },
+    {
+      "epoch": 0.7307732772225145,
+      "grad_norm": 1.6341030597686768,
+      "learning_rate": 3.5673392817059705e-05,
+      "loss": 1.5135,
+      "step": 6946
+    },
+    {
+      "epoch": 0.7308784850078905,
+      "grad_norm": 1.9366779327392578,
+      "learning_rate": 3.5647305463354466e-05,
+      "loss": 1.916,
+      "step": 6947
+    },
+    {
+      "epoch": 0.7309836927932667,
+      "grad_norm": 1.9645856618881226,
+      "learning_rate": 3.562122558271569e-05,
+      "loss": 1.4412,
+      "step": 6948
+    },
+    {
+      "epoch": 0.7310889005786428,
+      "grad_norm": 1.8454338312149048,
+      "learning_rate": 3.559515317817204e-05,
+      "loss": 1.4741,
+      "step": 6949
+    },
+    {
+      "epoch": 0.731194108364019,
+      "grad_norm": 1.1774706840515137,
+      "learning_rate": 3.556908825275117e-05,
+      "loss": 1.5745,
+      "step": 6950
+    },
+    {
+      "epoch": 0.731299316149395,
+      "grad_norm": 1.3022416830062866,
+      "learning_rate": 3.5543030809479874e-05,
+      "loss": 1.7731,
+      "step": 6951
+    },
+    {
+      "epoch": 0.7314045239347712,
+      "grad_norm": 0.9204453229904175,
+      "learning_rate": 3.5516980851384194e-05,
+      "loss": 1.5551,
+      "step": 6952
+    },
+    {
+      "epoch": 0.7315097317201473,
+      "grad_norm": 1.7485677003860474,
+      "learning_rate": 3.549093838148919e-05,
+      "loss": 1.6053,
+      "step": 6953
+    },
+    {
+      "epoch": 0.7316149395055234,
+      "grad_norm": 1.1012275218963623,
+      "learning_rate": 3.546490340281906e-05,
+      "loss": 1.6423,
+      "step": 6954
+    },
+    {
+      "epoch": 0.7317201472908995,
+      "grad_norm": 1.050255537033081,
+      "learning_rate": 3.543887591839723e-05,
+      "loss": 1.4111,
+      "step": 6955
+    },
+    {
+      "epoch": 0.7318253550762757,
+      "grad_norm": 1.184462070465088,
+      "learning_rate": 3.5412855931246126e-05,
+      "loss": 1.9403,
+      "step": 6956
+    },
+    {
+      "epoch": 0.7319305628616518,
+      "grad_norm": 1.2454978227615356,
+      "learning_rate": 3.538684344438736e-05,
+      "loss": 1.5388,
+      "step": 6957
+    },
+    {
+      "epoch": 0.7320357706470279,
+      "grad_norm": 1.4453421831130981,
+      "learning_rate": 3.5360838460841725e-05,
+      "loss": 1.9107,
+      "step": 6958
+    },
+    {
+      "epoch": 0.732140978432404,
+      "grad_norm": 1.2171581983566284,
+      "learning_rate": 3.533484098362908e-05,
+      "loss": 1.5818,
+      "step": 6959
+    },
+    {
+      "epoch": 0.7322461862177801,
+      "grad_norm": 2.109985113143921,
+      "learning_rate": 3.5308851015768375e-05,
+      "loss": 2.0371,
+      "step": 6960
+    },
+    {
+      "epoch": 0.7323513940031562,
+      "grad_norm": 2.1604323387145996,
+      "learning_rate": 3.52828685602778e-05,
+      "loss": 1.6088,
+      "step": 6961
+    },
+    {
+      "epoch": 0.7324566017885323,
+      "grad_norm": 1.6006240844726562,
+      "learning_rate": 3.525689362017461e-05,
+      "loss": 1.353,
+      "step": 6962
+    },
+    {
+      "epoch": 0.7325618095739085,
+      "grad_norm": 3.5379319190979004,
+      "learning_rate": 3.523092619847512e-05,
+      "loss": 1.9756,
+      "step": 6963
+    },
+    {
+      "epoch": 0.7326670173592846,
+      "grad_norm": 1.7658113241195679,
+      "learning_rate": 3.520496629819494e-05,
+      "loss": 1.8514,
+      "step": 6964
+    },
+    {
+      "epoch": 0.7327722251446607,
+      "grad_norm": 2.1337103843688965,
+      "learning_rate": 3.517901392234865e-05,
+      "loss": 1.785,
+      "step": 6965
+    },
+    {
+      "epoch": 0.7328774329300368,
+      "grad_norm": 1.1942282915115356,
+      "learning_rate": 3.515306907395002e-05,
+      "loss": 1.2969,
+      "step": 6966
+    },
+    {
+      "epoch": 0.732982640715413,
+      "grad_norm": 1.3888230323791504,
+      "learning_rate": 3.512713175601194e-05,
+      "loss": 1.5518,
+      "step": 6967
+    },
+    {
+      "epoch": 0.733087848500789,
+      "grad_norm": 1.5503004789352417,
+      "learning_rate": 3.51012019715464e-05,
+      "loss": 1.4631,
+      "step": 6968
+    },
+    {
+      "epoch": 0.7331930562861652,
+      "grad_norm": 2.311354160308838,
+      "learning_rate": 3.5075279723564616e-05,
+      "loss": 1.0551,
+      "step": 6969
+    },
+    {
+      "epoch": 0.7332982640715413,
+      "grad_norm": 2.0432231426239014,
+      "learning_rate": 3.504936501507679e-05,
+      "loss": 2.2122,
+      "step": 6970
+    },
+    {
+      "epoch": 0.7334034718569175,
+      "grad_norm": 1.2789908647537231,
+      "learning_rate": 3.502345784909229e-05,
+      "loss": 1.6112,
+      "step": 6971
+    },
+    {
+      "epoch": 0.7335086796422935,
+      "grad_norm": 1.3882858753204346,
+      "learning_rate": 3.499755822861971e-05,
+      "loss": 1.2767,
+      "step": 6972
+    },
+    {
+      "epoch": 0.7336138874276696,
+      "grad_norm": 1.920723795890808,
+      "learning_rate": 3.497166615666664e-05,
+      "loss": 1.8132,
+      "step": 6973
+    },
+    {
+      "epoch": 0.7337190952130458,
+      "grad_norm": 1.692369818687439,
+      "learning_rate": 3.4945781636239814e-05,
+      "loss": 1.7823,
+      "step": 6974
+    },
+    {
+      "epoch": 0.7338243029984218,
+      "grad_norm": 1.6629142761230469,
+      "learning_rate": 3.491990467034518e-05,
+      "loss": 1.7026,
+      "step": 6975
+    },
+    {
+      "epoch": 0.733929510783798,
+      "grad_norm": 1.6108320951461792,
+      "learning_rate": 3.4894035261987715e-05,
+      "loss": 1.2472,
+      "step": 6976
+    },
+    {
+      "epoch": 0.7340347185691741,
+      "grad_norm": 2.1136841773986816,
+      "learning_rate": 3.486817341417151e-05,
+      "loss": 1.0959,
+      "step": 6977
+    },
+    {
+      "epoch": 0.7341399263545503,
+      "grad_norm": 2.271759510040283,
+      "learning_rate": 3.484231912989989e-05,
+      "loss": 1.7992,
+      "step": 6978
+    },
+    {
+      "epoch": 0.7342451341399263,
+      "grad_norm": 1.7375454902648926,
+      "learning_rate": 3.481647241217516e-05,
+      "loss": 1.3712,
+      "step": 6979
+    },
+    {
+      "epoch": 0.7343503419253025,
+      "grad_norm": 1.6927552223205566,
+      "learning_rate": 3.4790633263998874e-05,
+      "loss": 1.3709,
+      "step": 6980
+    },
+    {
+      "epoch": 0.7344555497106786,
+      "grad_norm": 1.4837682247161865,
+      "learning_rate": 3.476480168837161e-05,
+      "loss": 1.4844,
+      "step": 6981
+    },
+    {
+      "epoch": 0.7345607574960548,
+      "grad_norm": 1.7805767059326172,
+      "learning_rate": 3.473897768829308e-05,
+      "loss": 1.9579,
+      "step": 6982
+    },
+    {
+      "epoch": 0.7346659652814308,
+      "grad_norm": 1.7141696214675903,
+      "learning_rate": 3.471316126676222e-05,
+      "loss": 1.1072,
+      "step": 6983
+    },
+    {
+      "epoch": 0.734771173066807,
+      "grad_norm": 2.4495108127593994,
+      "learning_rate": 3.4687352426776945e-05,
+      "loss": 1.4545,
+      "step": 6984
+    },
+    {
+      "epoch": 0.7348763808521831,
+      "grad_norm": 1.993420124053955,
+      "learning_rate": 3.466155117133433e-05,
+      "loss": 2.0231,
+      "step": 6985
+    },
+    {
+      "epoch": 0.7349815886375591,
+      "grad_norm": 1.47242271900177,
+      "learning_rate": 3.4635757503430685e-05,
+      "loss": 1.5673,
+      "step": 6986
+    },
+    {
+      "epoch": 0.7350867964229353,
+      "grad_norm": 2.614037036895752,
+      "learning_rate": 3.460997142606126e-05,
+      "loss": 2.0294,
+      "step": 6987
+    },
+    {
+      "epoch": 0.7351920042083114,
+      "grad_norm": 1.551087737083435,
+      "learning_rate": 3.4584192942220495e-05,
+      "loss": 1.889,
+      "step": 6988
+    },
+    {
+      "epoch": 0.7352972119936876,
+      "grad_norm": 1.1275653839111328,
+      "learning_rate": 3.455842205490202e-05,
+      "loss": 1.9836,
+      "step": 6989
+    },
+    {
+      "epoch": 0.7354024197790636,
+      "grad_norm": 2.598032236099243,
+      "learning_rate": 3.453265876709847e-05,
+      "loss": 1.7154,
+      "step": 6990
+    },
+    {
+      "epoch": 0.7355076275644398,
+      "grad_norm": 2.10746431350708,
+      "learning_rate": 3.45069030818017e-05,
+      "loss": 1.7135,
+      "step": 6991
+    },
+    {
+      "epoch": 0.7356128353498159,
+      "grad_norm": 1.3205444812774658,
+      "learning_rate": 3.448115500200263e-05,
+      "loss": 1.7395,
+      "step": 6992
+    },
+    {
+      "epoch": 0.735718043135192,
+      "grad_norm": 2.8115272521972656,
+      "learning_rate": 3.4455414530691234e-05,
+      "loss": 1.9317,
+      "step": 6993
+    },
+    {
+      "epoch": 0.7358232509205681,
+      "grad_norm": 2.6412436962127686,
+      "learning_rate": 3.442968167085675e-05,
+      "loss": 1.5553,
+      "step": 6994
+    },
+    {
+      "epoch": 0.7359284587059443,
+      "grad_norm": 1.441660761833191,
+      "learning_rate": 3.440395642548743e-05,
+      "loss": 1.4242,
+      "step": 6995
+    },
+    {
+      "epoch": 0.7360336664913204,
+      "grad_norm": 1.848536729812622,
+      "learning_rate": 3.437823879757059e-05,
+      "loss": 1.2416,
+      "step": 6996
+    },
+    {
+      "epoch": 0.7361388742766964,
+      "grad_norm": 1.5345301628112793,
+      "learning_rate": 3.435252879009284e-05,
+      "loss": 2.0876,
+      "step": 6997
+    },
+    {
+      "epoch": 0.7362440820620726,
+      "grad_norm": 1.2060550451278687,
+      "learning_rate": 3.432682640603975e-05,
+      "loss": 2.075,
+      "step": 6998
+    },
+    {
+      "epoch": 0.7363492898474487,
+      "grad_norm": 1.2402889728546143,
+      "learning_rate": 3.430113164839601e-05,
+      "loss": 2.3051,
+      "step": 6999
+    },
+    {
+      "epoch": 0.7364544976328248,
+      "grad_norm": 1.3935623168945312,
+      "learning_rate": 3.427544452014556e-05,
+      "loss": 1.045,
+      "step": 7000
+    },
+    {
+      "epoch": 0.7365597054182009,
+      "grad_norm": 1.647644281387329,
+      "learning_rate": 3.42497650242713e-05,
+      "loss": 1.5662,
+      "step": 7001
+    },
+    {
+      "epoch": 0.7366649132035771,
+      "grad_norm": 1.3244524002075195,
+      "learning_rate": 3.422409316375529e-05,
+      "loss": 1.6658,
+      "step": 7002
+    },
+    {
+      "epoch": 0.7367701209889532,
+      "grad_norm": 2.1039204597473145,
+      "learning_rate": 3.4198428941578776e-05,
+      "loss": 1.6034,
+      "step": 7003
+    },
+    {
+      "epoch": 0.7368753287743293,
+      "grad_norm": 1.919314980506897,
+      "learning_rate": 3.417277236072203e-05,
+      "loss": 1.7035,
+      "step": 7004
+    },
+    {
+      "epoch": 0.7369805365597054,
+      "grad_norm": 1.365127682685852,
+      "learning_rate": 3.414712342416443e-05,
+      "loss": 1.7206,
+      "step": 7005
+    },
+    {
+      "epoch": 0.7370857443450816,
+      "grad_norm": 1.5440795421600342,
+      "learning_rate": 3.4121482134884575e-05,
+      "loss": 0.6072,
+      "step": 7006
+    },
+    {
+      "epoch": 0.7371909521304576,
+      "grad_norm": 1.7925177812576294,
+      "learning_rate": 3.409584849586006e-05,
+      "loss": 1.7965,
+      "step": 7007
+    },
+    {
+      "epoch": 0.7372961599158337,
+      "grad_norm": 1.45937979221344,
+      "learning_rate": 3.4070222510067653e-05,
+      "loss": 1.5833,
+      "step": 7008
+    },
+    {
+      "epoch": 0.7374013677012099,
+      "grad_norm": 1.4615490436553955,
+      "learning_rate": 3.40446041804832e-05,
+      "loss": 1.5594,
+      "step": 7009
+    },
+    {
+      "epoch": 0.737506575486586,
+      "grad_norm": 2.050783395767212,
+      "learning_rate": 3.401899351008163e-05,
+      "loss": 1.8929,
+      "step": 7010
+    },
+    {
+      "epoch": 0.7376117832719621,
+      "grad_norm": 1.2625833749771118,
+      "learning_rate": 3.3993390501837116e-05,
+      "loss": 1.7766,
+      "step": 7011
+    },
+    {
+      "epoch": 0.7377169910573382,
+      "grad_norm": 1.5997296571731567,
+      "learning_rate": 3.396779515872282e-05,
+      "loss": 2.3762,
+      "step": 7012
+    },
+    {
+      "epoch": 0.7378221988427144,
+      "grad_norm": 1.210913062095642,
+      "learning_rate": 3.3942207483710986e-05,
+      "loss": 1.93,
+      "step": 7013
+    },
+    {
+      "epoch": 0.7379274066280905,
+      "grad_norm": 1.6027042865753174,
+      "learning_rate": 3.391662747977312e-05,
+      "loss": 2.1033,
+      "step": 7014
+    },
+    {
+      "epoch": 0.7380326144134666,
+      "grad_norm": 1.183409571647644,
+      "learning_rate": 3.389105514987969e-05,
+      "loss": 2.1993,
+      "step": 7015
+    },
+    {
+      "epoch": 0.7381378221988427,
+      "grad_norm": 1.665959358215332,
+      "learning_rate": 3.38654904970003e-05,
+      "loss": 1.7956,
+      "step": 7016
+    },
+    {
+      "epoch": 0.7382430299842189,
+      "grad_norm": 1.5837026834487915,
+      "learning_rate": 3.383993352410375e-05,
+      "loss": 1.4741,
+      "step": 7017
+    },
+    {
+      "epoch": 0.7383482377695949,
+      "grad_norm": 1.7137260437011719,
+      "learning_rate": 3.381438423415787e-05,
+      "loss": 1.6929,
+      "step": 7018
+    },
+    {
+      "epoch": 0.738453445554971,
+      "grad_norm": 1.120949625968933,
+      "learning_rate": 3.378884263012957e-05,
+      "loss": 1.3851,
+      "step": 7019
+    },
+    {
+      "epoch": 0.7385586533403472,
+      "grad_norm": 1.4287333488464355,
+      "learning_rate": 3.3763308714984974e-05,
+      "loss": 2.1058,
+      "step": 7020
+    },
+    {
+      "epoch": 0.7386638611257234,
+      "grad_norm": 1.9645508527755737,
+      "learning_rate": 3.373778249168923e-05,
+      "loss": 1.747,
+      "step": 7021
+    },
+    {
+      "epoch": 0.7387690689110994,
+      "grad_norm": 1.1778647899627686,
+      "learning_rate": 3.3712263963206583e-05,
+      "loss": 2.0621,
+      "step": 7022
+    },
+    {
+      "epoch": 0.7388742766964755,
+      "grad_norm": 2.0654361248016357,
+      "learning_rate": 3.368675313250046e-05,
+      "loss": 1.6014,
+      "step": 7023
+    },
+    {
+      "epoch": 0.7389794844818517,
+      "grad_norm": 1.6172906160354614,
+      "learning_rate": 3.3661250002533305e-05,
+      "loss": 1.6385,
+      "step": 7024
+    },
+    {
+      "epoch": 0.7390846922672277,
+      "grad_norm": 1.6722594499588013,
+      "learning_rate": 3.363575457626678e-05,
+      "loss": 1.7566,
+      "step": 7025
+    },
+    {
+      "epoch": 0.7391899000526039,
+      "grad_norm": 1.1120942831039429,
+      "learning_rate": 3.361026685666152e-05,
+      "loss": 1.7478,
+      "step": 7026
+    },
+    {
+      "epoch": 0.73929510783798,
+      "grad_norm": 1.4074763059616089,
+      "learning_rate": 3.358478684667734e-05,
+      "loss": 1.9311,
+      "step": 7027
+    },
+    {
+      "epoch": 0.7394003156233562,
+      "grad_norm": 1.5094165802001953,
+      "learning_rate": 3.355931454927319e-05,
+      "loss": 1.4962,
+      "step": 7028
+    },
+    {
+      "epoch": 0.7395055234087322,
+      "grad_norm": 1.2964214086532593,
+      "learning_rate": 3.353384996740709e-05,
+      "loss": 2.0931,
+      "step": 7029
+    },
+    {
+      "epoch": 0.7396107311941084,
+      "grad_norm": 1.637769103050232,
+      "learning_rate": 3.3508393104036076e-05,
+      "loss": 1.556,
+      "step": 7030
+    },
+    {
+      "epoch": 0.7397159389794845,
+      "grad_norm": 1.4458996057510376,
+      "learning_rate": 3.348294396211643e-05,
+      "loss": 1.7264,
+      "step": 7031
+    },
+    {
+      "epoch": 0.7398211467648605,
+      "grad_norm": 1.4599863290786743,
+      "learning_rate": 3.345750254460348e-05,
+      "loss": 1.3761,
+      "step": 7032
+    },
+    {
+      "epoch": 0.7399263545502367,
+      "grad_norm": 1.3754396438598633,
+      "learning_rate": 3.3432068854451594e-05,
+      "loss": 1.8374,
+      "step": 7033
+    },
+    {
+      "epoch": 0.7400315623356128,
+      "grad_norm": 1.6223207712173462,
+      "learning_rate": 3.3406642894614394e-05,
+      "loss": 1.5258,
+      "step": 7034
+    },
+    {
+      "epoch": 0.740136770120989,
+      "grad_norm": 2.438016414642334,
+      "learning_rate": 3.3381224668044434e-05,
+      "loss": 1.9796,
+      "step": 7035
+    },
+    {
+      "epoch": 0.740241977906365,
+      "grad_norm": 1.7482492923736572,
+      "learning_rate": 3.3355814177693514e-05,
+      "loss": 2.0305,
+      "step": 7036
+    },
+    {
+      "epoch": 0.7403471856917412,
+      "grad_norm": 2.278324842453003,
+      "learning_rate": 3.3330411426512435e-05,
+      "loss": 1.6287,
+      "step": 7037
+    },
+    {
+      "epoch": 0.7404523934771173,
+      "grad_norm": 1.471503734588623,
+      "learning_rate": 3.3305016417451105e-05,
+      "loss": 1.3077,
+      "step": 7038
+    },
+    {
+      "epoch": 0.7405576012624934,
+      "grad_norm": 1.7559709548950195,
+      "learning_rate": 3.327962915345864e-05,
+      "loss": 1.2328,
+      "step": 7039
+    },
+    {
+      "epoch": 0.7406628090478695,
+      "grad_norm": 1.2783993482589722,
+      "learning_rate": 3.325424963748313e-05,
+      "loss": 1.6599,
+      "step": 7040
+    },
+    {
+      "epoch": 0.7407680168332457,
+      "grad_norm": 1.633362054824829,
+      "learning_rate": 3.3228877872471786e-05,
+      "loss": 1.8694,
+      "step": 7041
+    },
+    {
+      "epoch": 0.7408732246186218,
+      "grad_norm": 1.3471500873565674,
+      "learning_rate": 3.3203513861371017e-05,
+      "loss": 1.4373,
+      "step": 7042
+    },
+    {
+      "epoch": 0.7409784324039979,
+      "grad_norm": 1.6833635568618774,
+      "learning_rate": 3.317815760712622e-05,
+      "loss": 1.2696,
+      "step": 7043
+    },
+    {
+      "epoch": 0.741083640189374,
+      "grad_norm": 1.1694504022598267,
+      "learning_rate": 3.315280911268193e-05,
+      "loss": 1.5157,
+      "step": 7044
+    },
+    {
+      "epoch": 0.7411888479747502,
+      "grad_norm": 1.1271198987960815,
+      "learning_rate": 3.312746838098181e-05,
+      "loss": 1.4756,
+      "step": 7045
+    },
+    {
+      "epoch": 0.7412940557601263,
+      "grad_norm": 1.275770664215088,
+      "learning_rate": 3.310213541496858e-05,
+      "loss": 1.4592,
+      "step": 7046
+    },
+    {
+      "epoch": 0.7413992635455023,
+      "grad_norm": 1.2125309705734253,
+      "learning_rate": 3.307681021758405e-05,
+      "loss": 1.9089,
+      "step": 7047
+    },
+    {
+      "epoch": 0.7415044713308785,
+      "grad_norm": 1.1902837753295898,
+      "learning_rate": 3.305149279176921e-05,
+      "loss": 2.156,
+      "step": 7048
+    },
+    {
+      "epoch": 0.7416096791162546,
+      "grad_norm": 1.931227684020996,
+      "learning_rate": 3.302618314046405e-05,
+      "loss": 1.6646,
+      "step": 7049
+    },
+    {
+      "epoch": 0.7417148869016307,
+      "grad_norm": 1.360236644744873,
+      "learning_rate": 3.300088126660768e-05,
+      "loss": 0.9917,
+      "step": 7050
+    },
+    {
+      "epoch": 0.7418200946870068,
+      "grad_norm": 1.1433465480804443,
+      "learning_rate": 3.297558717313839e-05,
+      "loss": 1.7004,
+      "step": 7051
+    },
+    {
+      "epoch": 0.741925302472383,
+      "grad_norm": 1.5731385946273804,
+      "learning_rate": 3.295030086299341e-05,
+      "loss": 1.8221,
+      "step": 7052
+    },
+    {
+      "epoch": 0.7420305102577591,
+      "grad_norm": 1.5698689222335815,
+      "learning_rate": 3.292502233910922e-05,
+      "loss": 2.1785,
+      "step": 7053
+    },
+    {
+      "epoch": 0.7421357180431352,
+      "grad_norm": 2.299558162689209,
+      "learning_rate": 3.2899751604421324e-05,
+      "loss": 1.2728,
+      "step": 7054
+    },
+    {
+      "epoch": 0.7422409258285113,
+      "grad_norm": 1.727777123451233,
+      "learning_rate": 3.287448866186428e-05,
+      "loss": 1.5016,
+      "step": 7055
+    },
+    {
+      "epoch": 0.7423461336138875,
+      "grad_norm": 1.1530907154083252,
+      "learning_rate": 3.284923351437187e-05,
+      "loss": 1.8296,
+      "step": 7056
+    },
+    {
+      "epoch": 0.7424513413992635,
+      "grad_norm": 1.4847744703292847,
+      "learning_rate": 3.282398616487684e-05,
+      "loss": 1.5375,
+      "step": 7057
+    },
+    {
+      "epoch": 0.7425565491846396,
+      "grad_norm": 1.4450937509536743,
+      "learning_rate": 3.279874661631106e-05,
+      "loss": 2.0141,
+      "step": 7058
+    },
+    {
+      "epoch": 0.7426617569700158,
+      "grad_norm": 1.7158372402191162,
+      "learning_rate": 3.277351487160558e-05,
+      "loss": 1.7055,
+      "step": 7059
+    },
+    {
+      "epoch": 0.742766964755392,
+      "grad_norm": 1.9537938833236694,
+      "learning_rate": 3.2748290933690454e-05,
+      "loss": 1.7621,
+      "step": 7060
+    },
+    {
+      "epoch": 0.742872172540768,
+      "grad_norm": 1.7717899084091187,
+      "learning_rate": 3.2723074805494805e-05,
+      "loss": 1.357,
+      "step": 7061
+    },
+    {
+      "epoch": 0.7429773803261441,
+      "grad_norm": 1.749004602432251,
+      "learning_rate": 3.269786648994697e-05,
+      "loss": 1.998,
+      "step": 7062
+    },
+    {
+      "epoch": 0.7430825881115203,
+      "grad_norm": 1.6080665588378906,
+      "learning_rate": 3.267266598997429e-05,
+      "loss": 1.6851,
+      "step": 7063
+    },
+    {
+      "epoch": 0.7431877958968963,
+      "grad_norm": 1.6543947458267212,
+      "learning_rate": 3.2647473308503164e-05,
+      "loss": 1.483,
+      "step": 7064
+    },
+    {
+      "epoch": 0.7432930036822725,
+      "grad_norm": 1.3643039464950562,
+      "learning_rate": 3.262228844845922e-05,
+      "loss": 1.7009,
+      "step": 7065
+    },
+    {
+      "epoch": 0.7433982114676486,
+      "grad_norm": 1.9166144132614136,
+      "learning_rate": 3.259711141276703e-05,
+      "loss": 2.1545,
+      "step": 7066
+    },
+    {
+      "epoch": 0.7435034192530248,
+      "grad_norm": 1.5741544961929321,
+      "learning_rate": 3.2571942204350324e-05,
+      "loss": 1.9652,
+      "step": 7067
+    },
+    {
+      "epoch": 0.7436086270384008,
+      "grad_norm": 1.287895679473877,
+      "learning_rate": 3.254678082613196e-05,
+      "loss": 1.753,
+      "step": 7068
+    },
+    {
+      "epoch": 0.743713834823777,
+      "grad_norm": 1.658474326133728,
+      "learning_rate": 3.252162728103382e-05,
+      "loss": 1.4791,
+      "step": 7069
+    },
+    {
+      "epoch": 0.7438190426091531,
+      "grad_norm": 1.8779799938201904,
+      "learning_rate": 3.249648157197688e-05,
+      "loss": 1.3374,
+      "step": 7070
+    },
+    {
+      "epoch": 0.7439242503945291,
+      "grad_norm": 1.7601832151412964,
+      "learning_rate": 3.2471343701881275e-05,
+      "loss": 1.4406,
+      "step": 7071
+    },
+    {
+      "epoch": 0.7440294581799053,
+      "grad_norm": 1.6893101930618286,
+      "learning_rate": 3.244621367366616e-05,
+      "loss": 1.0462,
+      "step": 7072
+    },
+    {
+      "epoch": 0.7441346659652814,
+      "grad_norm": 1.2970998287200928,
+      "learning_rate": 3.24210914902498e-05,
+      "loss": 1.7947,
+      "step": 7073
+    },
+    {
+      "epoch": 0.7442398737506576,
+      "grad_norm": 1.8401261568069458,
+      "learning_rate": 3.2395977154549554e-05,
+      "loss": 2.1458,
+      "step": 7074
+    },
+    {
+      "epoch": 0.7443450815360336,
+      "grad_norm": 2.419576406478882,
+      "learning_rate": 3.2370870669481834e-05,
+      "loss": 2.044,
+      "step": 7075
+    },
+    {
+      "epoch": 0.7444502893214098,
+      "grad_norm": 1.1183511018753052,
+      "learning_rate": 3.234577203796223e-05,
+      "loss": 1.7889,
+      "step": 7076
+    },
+    {
+      "epoch": 0.7445554971067859,
+      "grad_norm": 1.3616372346878052,
+      "learning_rate": 3.232068126290535e-05,
+      "loss": 1.8218,
+      "step": 7077
+    },
+    {
+      "epoch": 0.7446607048921621,
+      "grad_norm": 1.127997875213623,
+      "learning_rate": 3.229559834722485e-05,
+      "loss": 2.2213,
+      "step": 7078
+    },
+    {
+      "epoch": 0.7447659126775381,
+      "grad_norm": 1.5394973754882812,
+      "learning_rate": 3.227052329383362e-05,
+      "loss": 1.6531,
+      "step": 7079
+    },
+    {
+      "epoch": 0.7448711204629143,
+      "grad_norm": 1.4563318490982056,
+      "learning_rate": 3.224545610564345e-05,
+      "loss": 2.0104,
+      "step": 7080
+    },
+    {
+      "epoch": 0.7449763282482904,
+      "grad_norm": 1.76889169216156,
+      "learning_rate": 3.2220396785565375e-05,
+      "loss": 1.372,
+      "step": 7081
+    },
+    {
+      "epoch": 0.7450815360336664,
+      "grad_norm": 1.5505903959274292,
+      "learning_rate": 3.219534533650944e-05,
+      "loss": 2.0011,
+      "step": 7082
+    },
+    {
+      "epoch": 0.7451867438190426,
+      "grad_norm": 1.7687225341796875,
+      "learning_rate": 3.217030176138474e-05,
+      "loss": 1.8728,
+      "step": 7083
+    },
+    {
+      "epoch": 0.7452919516044187,
+      "grad_norm": 1.5347532033920288,
+      "learning_rate": 3.214526606309957e-05,
+      "loss": 1.6496,
+      "step": 7084
+    },
+    {
+      "epoch": 0.7453971593897949,
+      "grad_norm": 1.3686238527297974,
+      "learning_rate": 3.212023824456121e-05,
+      "loss": 1.7251,
+      "step": 7085
+    },
+    {
+      "epoch": 0.7455023671751709,
+      "grad_norm": 1.5368924140930176,
+      "learning_rate": 3.2095218308676024e-05,
+      "loss": 1.6474,
+      "step": 7086
+    },
+    {
+      "epoch": 0.7456075749605471,
+      "grad_norm": 1.4504789113998413,
+      "learning_rate": 3.207020625834957e-05,
+      "loss": 1.838,
+      "step": 7087
+    },
+    {
+      "epoch": 0.7457127827459232,
+      "grad_norm": 1.296593427658081,
+      "learning_rate": 3.204520209648637e-05,
+      "loss": 1.4902,
+      "step": 7088
+    },
+    {
+      "epoch": 0.7458179905312993,
+      "grad_norm": 1.3226828575134277,
+      "learning_rate": 3.2020205825990056e-05,
+      "loss": 1.504,
+      "step": 7089
+    },
+    {
+      "epoch": 0.7459231983166754,
+      "grad_norm": 1.6971758604049683,
+      "learning_rate": 3.199521744976342e-05,
+      "loss": 1.4026,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7460284061020516,
+      "grad_norm": 2.515711545944214,
+      "learning_rate": 3.1970236970708234e-05,
+      "loss": 1.9895,
+      "step": 7091
+    },
+    {
+      "epoch": 0.7461336138874277,
+      "grad_norm": 2.220277786254883,
+      "learning_rate": 3.194526439172539e-05,
+      "loss": 1.8306,
+      "step": 7092
+    },
+    {
+      "epoch": 0.7462388216728038,
+      "grad_norm": 1.5911275148391724,
+      "learning_rate": 3.192029971571492e-05,
+      "loss": 2.0089,
+      "step": 7093
+    },
+    {
+      "epoch": 0.7463440294581799,
+      "grad_norm": 1.8149131536483765,
+      "learning_rate": 3.189534294557587e-05,
+      "loss": 1.6243,
+      "step": 7094
+    },
+    {
+      "epoch": 0.746449237243556,
+      "grad_norm": 1.3475899696350098,
+      "learning_rate": 3.187039408420638e-05,
+      "loss": 1.5206,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7465544450289321,
+      "grad_norm": 1.1558003425598145,
+      "learning_rate": 3.184545313450368e-05,
+      "loss": 2.1106,
+      "step": 7096
+    },
+    {
+      "epoch": 0.7466596528143082,
+      "grad_norm": 1.5029397010803223,
+      "learning_rate": 3.182052009936404e-05,
+      "loss": 1.6606,
+      "step": 7097
+    },
+    {
+      "epoch": 0.7467648605996844,
+      "grad_norm": 1.187970519065857,
+      "learning_rate": 3.1795594981682917e-05,
+      "loss": 1.9256,
+      "step": 7098
+    },
+    {
+      "epoch": 0.7468700683850605,
+      "grad_norm": 1.2517000436782837,
+      "learning_rate": 3.1770677784354773e-05,
+      "loss": 1.4797,
+      "step": 7099
+    },
+    {
+      "epoch": 0.7469752761704366,
+      "grad_norm": 1.3138651847839355,
+      "learning_rate": 3.174576851027311e-05,
+      "loss": 2.2821,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7470804839558127,
+      "grad_norm": 1.6332142353057861,
+      "learning_rate": 3.172086716233061e-05,
+      "loss": 1.5739,
+      "step": 7101
+    },
+    {
+      "epoch": 0.7471856917411889,
+      "grad_norm": 1.3998867273330688,
+      "learning_rate": 3.1695973743418994e-05,
+      "loss": 1.7157,
+      "step": 7102
+    },
+    {
+      "epoch": 0.7472908995265649,
+      "grad_norm": 1.38740873336792,
+      "learning_rate": 3.167108825642897e-05,
+      "loss": 1.3737,
+      "step": 7103
+    },
+    {
+      "epoch": 0.7473961073119411,
+      "grad_norm": 1.5622433423995972,
+      "learning_rate": 3.164621070425051e-05,
+      "loss": 1.449,
+      "step": 7104
+    },
+    {
+      "epoch": 0.7475013150973172,
+      "grad_norm": 1.275417685508728,
+      "learning_rate": 3.162134108977253e-05,
+      "loss": 1.6909,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7476065228826934,
+      "grad_norm": 1.0392786264419556,
+      "learning_rate": 3.159647941588298e-05,
+      "loss": 1.8211,
+      "step": 7106
+    },
+    {
+      "epoch": 0.7477117306680694,
+      "grad_norm": 1.3279063701629639,
+      "learning_rate": 3.1571625685469086e-05,
+      "loss": 1.6451,
+      "step": 7107
+    },
+    {
+      "epoch": 0.7478169384534455,
+      "grad_norm": 1.5140197277069092,
+      "learning_rate": 3.1546779901416965e-05,
+      "loss": 2.0696,
+      "step": 7108
+    },
+    {
+      "epoch": 0.7479221462388217,
+      "grad_norm": 2.04419207572937,
+      "learning_rate": 3.152194206661185e-05,
+      "loss": 1.963,
+      "step": 7109
+    },
+    {
+      "epoch": 0.7480273540241978,
+      "grad_norm": 1.3939889669418335,
+      "learning_rate": 3.149711218393814e-05,
+      "loss": 1.2607,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7481325618095739,
+      "grad_norm": 1.5365245342254639,
+      "learning_rate": 3.147229025627922e-05,
+      "loss": 1.5934,
+      "step": 7111
+    },
+    {
+      "epoch": 0.74823776959495,
+      "grad_norm": 1.3564833402633667,
+      "learning_rate": 3.144747628651754e-05,
+      "loss": 1.3875,
+      "step": 7112
+    },
+    {
+      "epoch": 0.7483429773803262,
+      "grad_norm": 1.9279536008834839,
+      "learning_rate": 3.142267027753474e-05,
+      "loss": 2.0533,
+      "step": 7113
+    },
+    {
+      "epoch": 0.7484481851657022,
+      "grad_norm": 1.5397069454193115,
+      "learning_rate": 3.139787223221141e-05,
+      "loss": 1.8882,
+      "step": 7114
+    },
+    {
+      "epoch": 0.7485533929510784,
+      "grad_norm": 2.6162831783294678,
+      "learning_rate": 3.137308215342729e-05,
+      "loss": 2.1646,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7486586007364545,
+      "grad_norm": 1.33000910282135,
+      "learning_rate": 3.134830004406114e-05,
+      "loss": 1.243,
+      "step": 7116
+    },
+    {
+      "epoch": 0.7487638085218307,
+      "grad_norm": 1.1963053941726685,
+      "learning_rate": 3.1323525906990826e-05,
+      "loss": 1.856,
+      "step": 7117
+    },
+    {
+      "epoch": 0.7488690163072067,
+      "grad_norm": 1.3302133083343506,
+      "learning_rate": 3.129875974509332e-05,
+      "loss": 1.9321,
+      "step": 7118
+    },
+    {
+      "epoch": 0.7489742240925829,
+      "grad_norm": 1.8992441892623901,
+      "learning_rate": 3.127400156124463e-05,
+      "loss": 1.5662,
+      "step": 7119
+    },
+    {
+      "epoch": 0.749079431877959,
+      "grad_norm": 1.2559592723846436,
+      "learning_rate": 3.1249251358319784e-05,
+      "loss": 1.9587,
+      "step": 7120
+    },
+    {
+      "epoch": 0.749184639663335,
+      "grad_norm": 1.8781805038452148,
+      "learning_rate": 3.122450913919302e-05,
+      "loss": 1.5198,
+      "step": 7121
+    },
+    {
+      "epoch": 0.7492898474487112,
+      "grad_norm": 1.254428505897522,
+      "learning_rate": 3.1199774906737557e-05,
+      "loss": 2.0107,
+      "step": 7122
+    },
+    {
+      "epoch": 0.7493950552340873,
+      "grad_norm": 1.7342922687530518,
+      "learning_rate": 3.1175048663825626e-05,
+      "loss": 1.8242,
+      "step": 7123
+    },
+    {
+      "epoch": 0.7495002630194635,
+      "grad_norm": 1.5174115896224976,
+      "learning_rate": 3.11503304133287e-05,
+      "loss": 1.2619,
+      "step": 7124
+    },
+    {
+      "epoch": 0.7496054708048395,
+      "grad_norm": 1.4857186079025269,
+      "learning_rate": 3.1125620158117186e-05,
+      "loss": 1.4888,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7497106785902157,
+      "grad_norm": 1.7400273084640503,
+      "learning_rate": 3.110091790106057e-05,
+      "loss": 1.5356,
+      "step": 7126
+    },
+    {
+      "epoch": 0.7498158863755918,
+      "grad_norm": 1.744299292564392,
+      "learning_rate": 3.107622364502751e-05,
+      "loss": 2.2264,
+      "step": 7127
+    },
+    {
+      "epoch": 0.7499210941609679,
+      "grad_norm": 2.1220436096191406,
+      "learning_rate": 3.105153739288561e-05,
+      "loss": 1.645,
+      "step": 7128
+    },
+    {
+      "epoch": 0.750026301946344,
+      "grad_norm": 1.1671090126037598,
+      "learning_rate": 3.102685914750166e-05,
+      "loss": 1.389,
+      "step": 7129
+    },
+    {
+      "epoch": 0.7501315097317202,
+      "grad_norm": 1.858191728591919,
+      "learning_rate": 3.100218891174144e-05,
+      "loss": 1.56,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7502367175170963,
+      "grad_norm": 1.2349094152450562,
+      "learning_rate": 3.097752668846977e-05,
+      "loss": 1.6298,
+      "step": 7131
+    },
+    {
+      "epoch": 0.7503419253024723,
+      "grad_norm": 1.369010329246521,
+      "learning_rate": 3.095287248055069e-05,
+      "loss": 1.7109,
+      "step": 7132
+    },
+    {
+      "epoch": 0.7504471330878485,
+      "grad_norm": 1.2926762104034424,
+      "learning_rate": 3.092822629084715e-05,
+      "loss": 1.5841,
+      "step": 7133
+    },
+    {
+      "epoch": 0.7505523408732246,
+      "grad_norm": 1.9117960929870605,
+      "learning_rate": 3.090358812222122e-05,
+      "loss": 2.0065,
+      "step": 7134
+    },
+    {
+      "epoch": 0.7506575486586007,
+      "grad_norm": 1.1057976484298706,
+      "learning_rate": 3.0878957977534096e-05,
+      "loss": 1.5541,
+      "step": 7135
+    },
+    {
+      "epoch": 0.7507627564439768,
+      "grad_norm": 1.4562286138534546,
+      "learning_rate": 3.0854335859645975e-05,
+      "loss": 1.7463,
+      "step": 7136
+    },
+    {
+      "epoch": 0.750867964229353,
+      "grad_norm": 1.2036335468292236,
+      "learning_rate": 3.0829721771416146e-05,
+      "loss": 1.3571,
+      "step": 7137
+    },
+    {
+      "epoch": 0.7509731720147291,
+      "grad_norm": 1.2491194009780884,
+      "learning_rate": 3.080511571570294e-05,
+      "loss": 1.834,
+      "step": 7138
+    },
+    {
+      "epoch": 0.7510783798001052,
+      "grad_norm": 1.3567789793014526,
+      "learning_rate": 3.078051769536378e-05,
+      "loss": 1.8808,
+      "step": 7139
+    },
+    {
+      "epoch": 0.7511835875854813,
+      "grad_norm": 1.4406784772872925,
+      "learning_rate": 3.0755927713255174e-05,
+      "loss": 1.8301,
+      "step": 7140
+    },
+    {
+      "epoch": 0.7512887953708575,
+      "grad_norm": 1.7407746315002441,
+      "learning_rate": 3.073134577223268e-05,
+      "loss": 1.3971,
+      "step": 7141
+    },
+    {
+      "epoch": 0.7513940031562336,
+      "grad_norm": 1.3878988027572632,
+      "learning_rate": 3.070677187515087e-05,
+      "loss": 1.412,
+      "step": 7142
+    },
+    {
+      "epoch": 0.7514992109416097,
+      "grad_norm": 2.31683611869812,
+      "learning_rate": 3.06822060248635e-05,
+      "loss": 2.0428,
+      "step": 7143
+    },
+    {
+      "epoch": 0.7516044187269858,
+      "grad_norm": 1.463760495185852,
+      "learning_rate": 3.065764822422329e-05,
+      "loss": 1.7479,
+      "step": 7144
+    },
+    {
+      "epoch": 0.751709626512362,
+      "grad_norm": 1.3902920484542847,
+      "learning_rate": 3.063309847608202e-05,
+      "loss": 1.5929,
+      "step": 7145
+    },
+    {
+      "epoch": 0.751814834297738,
+      "grad_norm": 1.9475911855697632,
+      "learning_rate": 3.060855678329063e-05,
+      "loss": 1.6469,
+      "step": 7146
+    },
+    {
+      "epoch": 0.7519200420831141,
+      "grad_norm": 1.611803650856018,
+      "learning_rate": 3.0584023148699046e-05,
+      "loss": 1.775,
+      "step": 7147
+    },
+    {
+      "epoch": 0.7520252498684903,
+      "grad_norm": 1.7494837045669556,
+      "learning_rate": 3.055949757515624e-05,
+      "loss": 1.7225,
+      "step": 7148
+    },
+    {
+      "epoch": 0.7521304576538664,
+      "grad_norm": 1.4257961511611938,
+      "learning_rate": 3.0534980065510345e-05,
+      "loss": 1.7894,
+      "step": 7149
+    },
+    {
+      "epoch": 0.7522356654392425,
+      "grad_norm": 1.4211279153823853,
+      "learning_rate": 3.0510470622608478e-05,
+      "loss": 1.7914,
+      "step": 7150
+    },
+    {
+      "epoch": 0.7523408732246186,
+      "grad_norm": 1.7004531621932983,
+      "learning_rate": 3.0485969249296797e-05,
+      "loss": 2.1648,
+      "step": 7151
+    },
+    {
+      "epoch": 0.7524460810099948,
+      "grad_norm": 1.0059231519699097,
+      "learning_rate": 3.0461475948420637e-05,
+      "loss": 1.7455,
+      "step": 7152
+    },
+    {
+      "epoch": 0.7525512887953708,
+      "grad_norm": 1.09248685836792,
+      "learning_rate": 3.043699072282429e-05,
+      "loss": 1.839,
+      "step": 7153
+    },
+    {
+      "epoch": 0.752656496580747,
+      "grad_norm": 2.1308209896087646,
+      "learning_rate": 3.041251357535111e-05,
+      "loss": 1.5851,
+      "step": 7154
+    },
+    {
+      "epoch": 0.7527617043661231,
+      "grad_norm": 1.225701928138733,
+      "learning_rate": 3.0388044508843616e-05,
+      "loss": 2.0519,
+      "step": 7155
+    },
+    {
+      "epoch": 0.7528669121514993,
+      "grad_norm": 1.9547263383865356,
+      "learning_rate": 3.0363583526143292e-05,
+      "loss": 1.7462,
+      "step": 7156
+    },
+    {
+      "epoch": 0.7529721199368753,
+      "grad_norm": 1.4787694215774536,
+      "learning_rate": 3.0339130630090673e-05,
+      "loss": 1.0761,
+      "step": 7157
+    },
+    {
+      "epoch": 0.7530773277222514,
+      "grad_norm": 1.6599515676498413,
+      "learning_rate": 3.031468582352548e-05,
+      "loss": 1.8568,
+      "step": 7158
+    },
+    {
+      "epoch": 0.7531825355076276,
+      "grad_norm": 2.056112766265869,
+      "learning_rate": 3.0290249109286296e-05,
+      "loss": 1.7289,
+      "step": 7159
+    },
+    {
+      "epoch": 0.7532877432930036,
+      "grad_norm": 1.7584683895111084,
+      "learning_rate": 3.0265820490210973e-05,
+      "loss": 1.9411,
+      "step": 7160
+    },
+    {
+      "epoch": 0.7533929510783798,
+      "grad_norm": 1.8417630195617676,
+      "learning_rate": 3.0241399969136276e-05,
+      "loss": 1.5764,
+      "step": 7161
+    },
+    {
+      "epoch": 0.7534981588637559,
+      "grad_norm": 1.1658049821853638,
+      "learning_rate": 3.0216987548898068e-05,
+      "loss": 1.7206,
+      "step": 7162
+    },
+    {
+      "epoch": 0.7536033666491321,
+      "grad_norm": 1.9770032167434692,
+      "learning_rate": 3.019258323233133e-05,
+      "loss": 1.5526,
+      "step": 7163
+    },
+    {
+      "epoch": 0.7537085744345081,
+      "grad_norm": 1.6163029670715332,
+      "learning_rate": 3.0168187022270032e-05,
+      "loss": 1.8265,
+      "step": 7164
+    },
+    {
+      "epoch": 0.7538137822198843,
+      "grad_norm": 1.530423641204834,
+      "learning_rate": 3.0143798921547193e-05,
+      "loss": 1.5206,
+      "step": 7165
+    },
+    {
+      "epoch": 0.7539189900052604,
+      "grad_norm": 1.2398508787155151,
+      "learning_rate": 3.011941893299499e-05,
+      "loss": 1.5016,
+      "step": 7166
+    },
+    {
+      "epoch": 0.7540241977906365,
+      "grad_norm": 1.3902400732040405,
+      "learning_rate": 3.0095047059444546e-05,
+      "loss": 1.2807,
+      "step": 7167
+    },
+    {
+      "epoch": 0.7541294055760126,
+      "grad_norm": 1.967103362083435,
+      "learning_rate": 3.0070683303726076e-05,
+      "loss": 1.3802,
+      "step": 7168
+    },
+    {
+      "epoch": 0.7542346133613888,
+      "grad_norm": 1.6876822710037231,
+      "learning_rate": 3.0046327668668904e-05,
+      "loss": 1.5044,
+      "step": 7169
+    },
+    {
+      "epoch": 0.7543398211467649,
+      "grad_norm": 1.5691266059875488,
+      "learning_rate": 3.002198015710136e-05,
+      "loss": 1.4267,
+      "step": 7170
+    },
+    {
+      "epoch": 0.7544450289321409,
+      "grad_norm": 1.312497615814209,
+      "learning_rate": 2.999764077185079e-05,
+      "loss": 1.4682,
+      "step": 7171
+    },
+    {
+      "epoch": 0.7545502367175171,
+      "grad_norm": 1.056235432624817,
+      "learning_rate": 2.997330951574371e-05,
+      "loss": 1.8569,
+      "step": 7172
+    },
+    {
+      "epoch": 0.7546554445028932,
+      "grad_norm": 1.192795753479004,
+      "learning_rate": 2.9948986391605584e-05,
+      "loss": 1.3137,
+      "step": 7173
+    },
+    {
+      "epoch": 0.7547606522882694,
+      "grad_norm": 2.208214282989502,
+      "learning_rate": 2.9924671402261018e-05,
+      "loss": 1.7247,
+      "step": 7174
+    },
+    {
+      "epoch": 0.7548658600736454,
+      "grad_norm": 1.3944358825683594,
+      "learning_rate": 2.9900364550533612e-05,
+      "loss": 1.6949,
+      "step": 7175
+    },
+    {
+      "epoch": 0.7549710678590216,
+      "grad_norm": 1.4793987274169922,
+      "learning_rate": 2.9876065839246005e-05,
+      "loss": 1.8035,
+      "step": 7176
+    },
+    {
+      "epoch": 0.7550762756443977,
+      "grad_norm": 1.936575174331665,
+      "learning_rate": 2.9851775271219996e-05,
+      "loss": 1.279,
+      "step": 7177
+    },
+    {
+      "epoch": 0.7551814834297738,
+      "grad_norm": 1.4418457746505737,
+      "learning_rate": 2.9827492849276317e-05,
+      "loss": 1.6786,
+      "step": 7178
+    },
+    {
+      "epoch": 0.7552866912151499,
+      "grad_norm": 1.314530849456787,
+      "learning_rate": 2.9803218576234836e-05,
+      "loss": 1.5827,
+      "step": 7179
+    },
+    {
+      "epoch": 0.7553918990005261,
+      "grad_norm": 1.114014983177185,
+      "learning_rate": 2.9778952454914422e-05,
+      "loss": 1.484,
+      "step": 7180
+    },
+    {
+      "epoch": 0.7554971067859022,
+      "grad_norm": 1.7452201843261719,
+      "learning_rate": 2.9754694488133038e-05,
+      "loss": 2.1129,
+      "step": 7181
+    },
+    {
+      "epoch": 0.7556023145712782,
+      "grad_norm": 1.7121224403381348,
+      "learning_rate": 2.9730444678707625e-05,
+      "loss": 1.7277,
+      "step": 7182
+    },
+    {
+      "epoch": 0.7557075223566544,
+      "grad_norm": 1.2562663555145264,
+      "learning_rate": 2.9706203029454316e-05,
+      "loss": 1.6911,
+      "step": 7183
+    },
+    {
+      "epoch": 0.7558127301420305,
+      "grad_norm": 1.4814766645431519,
+      "learning_rate": 2.9681969543188148e-05,
+      "loss": 1.5525,
+      "step": 7184
+    },
+    {
+      "epoch": 0.7559179379274066,
+      "grad_norm": 0.9775087833404541,
+      "learning_rate": 2.965774422272334e-05,
+      "loss": 1.9295,
+      "step": 7185
+    },
+    {
+      "epoch": 0.7560231457127827,
+      "grad_norm": 1.847643494606018,
+      "learning_rate": 2.9633527070873057e-05,
+      "loss": 1.5732,
+      "step": 7186
+    },
+    {
+      "epoch": 0.7561283534981589,
+      "grad_norm": 1.5069630146026611,
+      "learning_rate": 2.9609318090449533e-05,
+      "loss": 1.3961,
+      "step": 7187
+    },
+    {
+      "epoch": 0.756233561283535,
+      "grad_norm": 1.6327673196792603,
+      "learning_rate": 2.958511728426414e-05,
+      "loss": 1.0892,
+      "step": 7188
+    },
+    {
+      "epoch": 0.7563387690689111,
+      "grad_norm": 1.8135273456573486,
+      "learning_rate": 2.956092465512721e-05,
+      "loss": 1.354,
+      "step": 7189
+    },
+    {
+      "epoch": 0.7564439768542872,
+      "grad_norm": 2.080944538116455,
+      "learning_rate": 2.9536740205848113e-05,
+      "loss": 1.5985,
+      "step": 7190
+    },
+    {
+      "epoch": 0.7565491846396634,
+      "grad_norm": 1.4439144134521484,
+      "learning_rate": 2.9512563939235382e-05,
+      "loss": 1.5032,
+      "step": 7191
+    },
+    {
+      "epoch": 0.7566543924250394,
+      "grad_norm": 1.6564826965332031,
+      "learning_rate": 2.9488395858096485e-05,
+      "loss": 1.9067,
+      "step": 7192
+    },
+    {
+      "epoch": 0.7567596002104156,
+      "grad_norm": 1.779802918434143,
+      "learning_rate": 2.9464235965237964e-05,
+      "loss": 1.8031,
+      "step": 7193
+    },
+    {
+      "epoch": 0.7568648079957917,
+      "grad_norm": 1.4289071559906006,
+      "learning_rate": 2.9440084263465484e-05,
+      "loss": 1.616,
+      "step": 7194
+    },
+    {
+      "epoch": 0.7569700157811678,
+      "grad_norm": 1.3593642711639404,
+      "learning_rate": 2.941594075558366e-05,
+      "loss": 0.9734,
+      "step": 7195
+    },
+    {
+      "epoch": 0.7570752235665439,
+      "grad_norm": 2.13067889213562,
+      "learning_rate": 2.939180544439618e-05,
+      "loss": 1.4978,
+      "step": 7196
+    },
+    {
+      "epoch": 0.75718043135192,
+      "grad_norm": 1.6457139253616333,
+      "learning_rate": 2.936767833270586e-05,
+      "loss": 1.8645,
+      "step": 7197
+    },
+    {
+      "epoch": 0.7572856391372962,
+      "grad_norm": 1.727137565612793,
+      "learning_rate": 2.934355942331447e-05,
+      "loss": 1.7537,
+      "step": 7198
+    },
+    {
+      "epoch": 0.7573908469226722,
+      "grad_norm": 1.2516162395477295,
+      "learning_rate": 2.9319448719022824e-05,
+      "loss": 1.9984,
+      "step": 7199
+    },
+    {
+      "epoch": 0.7574960547080484,
+      "grad_norm": 1.7397236824035645,
+      "learning_rate": 2.929534622263088e-05,
+      "loss": 1.6422,
+      "step": 7200
+    },
+    {
+      "epoch": 0.7576012624934245,
+      "grad_norm": 1.722465991973877,
+      "learning_rate": 2.927125193693756e-05,
+      "loss": 1.745,
+      "step": 7201
+    },
+    {
+      "epoch": 0.7577064702788007,
+      "grad_norm": 1.5950039625167847,
+      "learning_rate": 2.9247165864740856e-05,
+      "loss": 1.6426,
+      "step": 7202
+    },
+    {
+      "epoch": 0.7578116780641767,
+      "grad_norm": 1.0130795240402222,
+      "learning_rate": 2.9223088008837785e-05,
+      "loss": 1.5676,
+      "step": 7203
+    },
+    {
+      "epoch": 0.7579168858495529,
+      "grad_norm": 2.0876479148864746,
+      "learning_rate": 2.919901837202441e-05,
+      "loss": 1.7984,
+      "step": 7204
+    },
+    {
+      "epoch": 0.758022093634929,
+      "grad_norm": 2.0102639198303223,
+      "learning_rate": 2.917495695709592e-05,
+      "loss": 1.2913,
+      "step": 7205
+    },
+    {
+      "epoch": 0.7581273014203052,
+      "grad_norm": 1.934104561805725,
+      "learning_rate": 2.915090376684646e-05,
+      "loss": 0.9719,
+      "step": 7206
+    },
+    {
+      "epoch": 0.7582325092056812,
+      "grad_norm": 1.983300805091858,
+      "learning_rate": 2.9126858804069223e-05,
+      "loss": 1.83,
+      "step": 7207
+    },
+    {
+      "epoch": 0.7583377169910573,
+      "grad_norm": 2.1567113399505615,
+      "learning_rate": 2.9102822071556512e-05,
+      "loss": 1.7204,
+      "step": 7208
+    },
+    {
+      "epoch": 0.7584429247764335,
+      "grad_norm": 1.1860876083374023,
+      "learning_rate": 2.9078793572099616e-05,
+      "loss": 1.9178,
+      "step": 7209
+    },
+    {
+      "epoch": 0.7585481325618095,
+      "grad_norm": 2.058955669403076,
+      "learning_rate": 2.905477330848886e-05,
+      "loss": 1.7382,
+      "step": 7210
+    },
+    {
+      "epoch": 0.7586533403471857,
+      "grad_norm": 1.8354625701904297,
+      "learning_rate": 2.9030761283513684e-05,
+      "loss": 2.3162,
+      "step": 7211
+    },
+    {
+      "epoch": 0.7587585481325618,
+      "grad_norm": 1.469107747077942,
+      "learning_rate": 2.9006757499962512e-05,
+      "loss": 1.5963,
+      "step": 7212
+    },
+    {
+      "epoch": 0.758863755917938,
+      "grad_norm": 1.9937857389450073,
+      "learning_rate": 2.8982761960622782e-05,
+      "loss": 1.22,
+      "step": 7213
+    },
+    {
+      "epoch": 0.758968963703314,
+      "grad_norm": 1.7028518915176392,
+      "learning_rate": 2.8958774668281084e-05,
+      "loss": 1.4993,
+      "step": 7214
+    },
+    {
+      "epoch": 0.7590741714886902,
+      "grad_norm": 1.4997313022613525,
+      "learning_rate": 2.8934795625722943e-05,
+      "loss": 1.2415,
+      "step": 7215
+    },
+    {
+      "epoch": 0.7591793792740663,
+      "grad_norm": 2.3115732669830322,
+      "learning_rate": 2.8910824835732952e-05,
+      "loss": 1.4934,
+      "step": 7216
+    },
+    {
+      "epoch": 0.7592845870594424,
+      "grad_norm": 1.127637267112732,
+      "learning_rate": 2.8886862301094807e-05,
+      "loss": 1.6622,
+      "step": 7217
+    },
+    {
+      "epoch": 0.7593897948448185,
+      "grad_norm": 1.6619749069213867,
+      "learning_rate": 2.8862908024591186e-05,
+      "loss": 1.9157,
+      "step": 7218
+    },
+    {
+      "epoch": 0.7594950026301946,
+      "grad_norm": 1.5014777183532715,
+      "learning_rate": 2.8838962009003756e-05,
+      "loss": 1.8232,
+      "step": 7219
+    },
+    {
+      "epoch": 0.7596002104155708,
+      "grad_norm": 1.5407613515853882,
+      "learning_rate": 2.881502425711339e-05,
+      "loss": 1.6237,
+      "step": 7220
+    },
+    {
+      "epoch": 0.7597054182009468,
+      "grad_norm": 1.672674298286438,
+      "learning_rate": 2.8791094771699802e-05,
+      "loss": 1.9443,
+      "step": 7221
+    },
+    {
+      "epoch": 0.759810625986323,
+      "grad_norm": 1.8555692434310913,
+      "learning_rate": 2.8767173555541972e-05,
+      "loss": 1.3203,
+      "step": 7222
+    },
+    {
+      "epoch": 0.7599158337716991,
+      "grad_norm": 1.1525888442993164,
+      "learning_rate": 2.8743260611417665e-05,
+      "loss": 1.4575,
+      "step": 7223
+    },
+    {
+      "epoch": 0.7600210415570752,
+      "grad_norm": 1.2628978490829468,
+      "learning_rate": 2.8719355942103842e-05,
+      "loss": 1.5078,
+      "step": 7224
+    },
+    {
+      "epoch": 0.7601262493424513,
+      "grad_norm": 1.7043256759643555,
+      "learning_rate": 2.8695459550376515e-05,
+      "loss": 1.2751,
+      "step": 7225
+    },
+    {
+      "epoch": 0.7602314571278275,
+      "grad_norm": 2.068817615509033,
+      "learning_rate": 2.867157143901067e-05,
+      "loss": 1.9544,
+      "step": 7226
+    },
+    {
+      "epoch": 0.7603366649132036,
+      "grad_norm": 1.3836297988891602,
+      "learning_rate": 2.8647691610780326e-05,
+      "loss": 1.5782,
+      "step": 7227
+    },
+    {
+      "epoch": 0.7604418726985797,
+      "grad_norm": 1.6043787002563477,
+      "learning_rate": 2.862382006845863e-05,
+      "loss": 1.122,
+      "step": 7228
+    },
+    {
+      "epoch": 0.7605470804839558,
+      "grad_norm": 1.506122350692749,
+      "learning_rate": 2.8599956814817642e-05,
+      "loss": 1.8322,
+      "step": 7229
+    },
+    {
+      "epoch": 0.760652288269332,
+      "grad_norm": 1.124856948852539,
+      "learning_rate": 2.857610185262859e-05,
+      "loss": 1.8696,
+      "step": 7230
+    },
+    {
+      "epoch": 0.760757496054708,
+      "grad_norm": 1.269283652305603,
+      "learning_rate": 2.8552255184661624e-05,
+      "loss": 1.8721,
+      "step": 7231
+    },
+    {
+      "epoch": 0.7608627038400841,
+      "grad_norm": 1.4874763488769531,
+      "learning_rate": 2.8528416813685975e-05,
+      "loss": 1.6838,
+      "step": 7232
+    },
+    {
+      "epoch": 0.7609679116254603,
+      "grad_norm": 1.3948214054107666,
+      "learning_rate": 2.8504586742469952e-05,
+      "loss": 1.6012,
+      "step": 7233
+    },
+    {
+      "epoch": 0.7610731194108364,
+      "grad_norm": 1.520309329032898,
+      "learning_rate": 2.848076497378085e-05,
+      "loss": 2.2024,
+      "step": 7234
+    },
+    {
+      "epoch": 0.7611783271962125,
+      "grad_norm": 1.4063392877578735,
+      "learning_rate": 2.8456951510384966e-05,
+      "loss": 1.7189,
+      "step": 7235
+    },
+    {
+      "epoch": 0.7612835349815886,
+      "grad_norm": 1.1339725255966187,
+      "learning_rate": 2.8433146355047748e-05,
+      "loss": 1.6356,
+      "step": 7236
+    },
+    {
+      "epoch": 0.7613887427669648,
+      "grad_norm": 1.4083565473556519,
+      "learning_rate": 2.8409349510533578e-05,
+      "loss": 1.9028,
+      "step": 7237
+    },
+    {
+      "epoch": 0.7614939505523409,
+      "grad_norm": 1.4033015966415405,
+      "learning_rate": 2.8385560979605884e-05,
+      "loss": 1.6745,
+      "step": 7238
+    },
+    {
+      "epoch": 0.761599158337717,
+      "grad_norm": 1.8718355894088745,
+      "learning_rate": 2.83617807650272e-05,
+      "loss": 1.8888,
+      "step": 7239
+    },
+    {
+      "epoch": 0.7617043661230931,
+      "grad_norm": 1.4770317077636719,
+      "learning_rate": 2.833800886955902e-05,
+      "loss": 1.5846,
+      "step": 7240
+    },
+    {
+      "epoch": 0.7618095739084693,
+      "grad_norm": 1.7958834171295166,
+      "learning_rate": 2.831424529596185e-05,
+      "loss": 1.2971,
+      "step": 7241
+    },
+    {
+      "epoch": 0.7619147816938453,
+      "grad_norm": 1.5714819431304932,
+      "learning_rate": 2.8290490046995365e-05,
+      "loss": 1.4623,
+      "step": 7242
+    },
+    {
+      "epoch": 0.7620199894792214,
+      "grad_norm": 1.6446837186813354,
+      "learning_rate": 2.8266743125418128e-05,
+      "loss": 1.7678,
+      "step": 7243
+    },
+    {
+      "epoch": 0.7621251972645976,
+      "grad_norm": 1.509402871131897,
+      "learning_rate": 2.8243004533987793e-05,
+      "loss": 1.8722,
+      "step": 7244
+    },
+    {
+      "epoch": 0.7622304050499737,
+      "grad_norm": 1.4639089107513428,
+      "learning_rate": 2.8219274275461062e-05,
+      "loss": 1.5998,
+      "step": 7245
+    },
+    {
+      "epoch": 0.7623356128353498,
+      "grad_norm": 1.3890831470489502,
+      "learning_rate": 2.8195552352593612e-05,
+      "loss": 1.7413,
+      "step": 7246
+    },
+    {
+      "epoch": 0.7624408206207259,
+      "grad_norm": 1.8161264657974243,
+      "learning_rate": 2.8171838768140245e-05,
+      "loss": 1.8833,
+      "step": 7247
+    },
+    {
+      "epoch": 0.7625460284061021,
+      "grad_norm": 1.8306972980499268,
+      "learning_rate": 2.8148133524854716e-05,
+      "loss": 1.9024,
+      "step": 7248
+    },
+    {
+      "epoch": 0.7626512361914781,
+      "grad_norm": 1.03922438621521,
+      "learning_rate": 2.8124436625489813e-05,
+      "loss": 1.7287,
+      "step": 7249
+    },
+    {
+      "epoch": 0.7627564439768543,
+      "grad_norm": 2.0580270290374756,
+      "learning_rate": 2.8100748072797435e-05,
+      "loss": 1.5174,
+      "step": 7250
+    },
+    {
+      "epoch": 0.7628616517622304,
+      "grad_norm": 1.4314181804656982,
+      "learning_rate": 2.8077067869528417e-05,
+      "loss": 1.6338,
+      "step": 7251
+    },
+    {
+      "epoch": 0.7629668595476066,
+      "grad_norm": 1.2152514457702637,
+      "learning_rate": 2.8053396018432644e-05,
+      "loss": 1.6196,
+      "step": 7252
+    },
+    {
+      "epoch": 0.7630720673329826,
+      "grad_norm": 2.0151803493499756,
+      "learning_rate": 2.802973252225911e-05,
+      "loss": 1.5984,
+      "step": 7253
+    },
+    {
+      "epoch": 0.7631772751183588,
+      "grad_norm": 1.595259428024292,
+      "learning_rate": 2.8006077383755747e-05,
+      "loss": 1.4817,
+      "step": 7254
+    },
+    {
+      "epoch": 0.7632824829037349,
+      "grad_norm": 0.9341708421707153,
+      "learning_rate": 2.7982430605669517e-05,
+      "loss": 1.7419,
+      "step": 7255
+    },
+    {
+      "epoch": 0.7633876906891109,
+      "grad_norm": 1.5944820642471313,
+      "learning_rate": 2.7958792190746496e-05,
+      "loss": 1.6313,
+      "step": 7256
+    },
+    {
+      "epoch": 0.7634928984744871,
+      "grad_norm": 1.1736741065979004,
+      "learning_rate": 2.7935162141731718e-05,
+      "loss": 1.6742,
+      "step": 7257
+    },
+    {
+      "epoch": 0.7635981062598632,
+      "grad_norm": 1.133220911026001,
+      "learning_rate": 2.7911540461369222e-05,
+      "loss": 1.443,
+      "step": 7258
+    },
+    {
+      "epoch": 0.7637033140452394,
+      "grad_norm": 1.820003867149353,
+      "learning_rate": 2.788792715240218e-05,
+      "loss": 1.9558,
+      "step": 7259
+    },
+    {
+      "epoch": 0.7638085218306154,
+      "grad_norm": 1.6289899349212646,
+      "learning_rate": 2.78643222175727e-05,
+      "loss": 1.863,
+      "step": 7260
+    },
+    {
+      "epoch": 0.7639137296159916,
+      "grad_norm": 1.1900887489318848,
+      "learning_rate": 2.7840725659621915e-05,
+      "loss": 2.0473,
+      "step": 7261
+    },
+    {
+      "epoch": 0.7640189374013677,
+      "grad_norm": 2.775047779083252,
+      "learning_rate": 2.781713748129008e-05,
+      "loss": 1.7868,
+      "step": 7262
+    },
+    {
+      "epoch": 0.7641241451867438,
+      "grad_norm": 1.3283895254135132,
+      "learning_rate": 2.7793557685316363e-05,
+      "loss": 1.4167,
+      "step": 7263
+    },
+    {
+      "epoch": 0.7642293529721199,
+      "grad_norm": 1.6619638204574585,
+      "learning_rate": 2.7769986274439e-05,
+      "loss": 1.7392,
+      "step": 7264
+    },
+    {
+      "epoch": 0.7643345607574961,
+      "grad_norm": 1.9690237045288086,
+      "learning_rate": 2.774642325139535e-05,
+      "loss": 1.3192,
+      "step": 7265
+    },
+    {
+      "epoch": 0.7644397685428722,
+      "grad_norm": 1.4869492053985596,
+      "learning_rate": 2.772286861892157e-05,
+      "loss": 1.8371,
+      "step": 7266
+    },
+    {
+      "epoch": 0.7645449763282482,
+      "grad_norm": 2.642908811569214,
+      "learning_rate": 2.769932237975309e-05,
+      "loss": 1.3455,
+      "step": 7267
+    },
+    {
+      "epoch": 0.7646501841136244,
+      "grad_norm": 2.0182862281799316,
+      "learning_rate": 2.7675784536624215e-05,
+      "loss": 1.8982,
+      "step": 7268
+    },
+    {
+      "epoch": 0.7647553918990005,
+      "grad_norm": 1.7238906621932983,
+      "learning_rate": 2.7652255092268298e-05,
+      "loss": 1.21,
+      "step": 7269
+    },
+    {
+      "epoch": 0.7648605996843767,
+      "grad_norm": 1.1936345100402832,
+      "learning_rate": 2.762873404941779e-05,
+      "loss": 2.1362,
+      "step": 7270
+    },
+    {
+      "epoch": 0.7649658074697527,
+      "grad_norm": 1.5153876543045044,
+      "learning_rate": 2.7605221410804093e-05,
+      "loss": 1.9925,
+      "step": 7271
+    },
+    {
+      "epoch": 0.7650710152551289,
+      "grad_norm": 1.9088298082351685,
+      "learning_rate": 2.7581717179157606e-05,
+      "loss": 1.6018,
+      "step": 7272
+    },
+    {
+      "epoch": 0.765176223040505,
+      "grad_norm": 1.2692924737930298,
+      "learning_rate": 2.755822135720787e-05,
+      "loss": 1.8706,
+      "step": 7273
+    },
+    {
+      "epoch": 0.7652814308258811,
+      "grad_norm": 1.2598018646240234,
+      "learning_rate": 2.7534733947683334e-05,
+      "loss": 1.7554,
+      "step": 7274
+    },
+    {
+      "epoch": 0.7653866386112572,
+      "grad_norm": 2.6628575325012207,
+      "learning_rate": 2.7511254953311495e-05,
+      "loss": 1.4819,
+      "step": 7275
+    },
+    {
+      "epoch": 0.7654918463966334,
+      "grad_norm": 1.1039395332336426,
+      "learning_rate": 2.748778437681895e-05,
+      "loss": 1.0194,
+      "step": 7276
+    },
+    {
+      "epoch": 0.7655970541820095,
+      "grad_norm": 1.6342849731445312,
+      "learning_rate": 2.7464322220931205e-05,
+      "loss": 1.455,
+      "step": 7277
+    },
+    {
+      "epoch": 0.7657022619673856,
+      "grad_norm": 1.5950639247894287,
+      "learning_rate": 2.7440868488372884e-05,
+      "loss": 1.3383,
+      "step": 7278
+    },
+    {
+      "epoch": 0.7658074697527617,
+      "grad_norm": 1.440382719039917,
+      "learning_rate": 2.7417423181867585e-05,
+      "loss": 1.8153,
+      "step": 7279
+    },
+    {
+      "epoch": 0.7659126775381379,
+      "grad_norm": 2.1018497943878174,
+      "learning_rate": 2.7393986304137887e-05,
+      "loss": 1.6111,
+      "step": 7280
+    },
+    {
+      "epoch": 0.7660178853235139,
+      "grad_norm": 1.6211340427398682,
+      "learning_rate": 2.73705578579055e-05,
+      "loss": 1.7447,
+      "step": 7281
+    },
+    {
+      "epoch": 0.76612309310889,
+      "grad_norm": 1.3533055782318115,
+      "learning_rate": 2.7347137845891068e-05,
+      "loss": 1.5142,
+      "step": 7282
+    },
+    {
+      "epoch": 0.7662283008942662,
+      "grad_norm": 1.5132229328155518,
+      "learning_rate": 2.7323726270814253e-05,
+      "loss": 1.6848,
+      "step": 7283
+    },
+    {
+      "epoch": 0.7663335086796423,
+      "grad_norm": 1.4339923858642578,
+      "learning_rate": 2.7300323135393812e-05,
+      "loss": 1.6556,
+      "step": 7284
+    },
+    {
+      "epoch": 0.7664387164650184,
+      "grad_norm": 2.367436170578003,
+      "learning_rate": 2.727692844234746e-05,
+      "loss": 1.9757,
+      "step": 7285
+    },
+    {
+      "epoch": 0.7665439242503945,
+      "grad_norm": 1.1163835525512695,
+      "learning_rate": 2.72535421943919e-05,
+      "loss": 1.7622,
+      "step": 7286
+    },
+    {
+      "epoch": 0.7666491320357707,
+      "grad_norm": 1.5662150382995605,
+      "learning_rate": 2.7230164394242995e-05,
+      "loss": 1.6492,
+      "step": 7287
+    },
+    {
+      "epoch": 0.7667543398211467,
+      "grad_norm": 1.4997937679290771,
+      "learning_rate": 2.720679504461542e-05,
+      "loss": 1.3393,
+      "step": 7288
+    },
+    {
+      "epoch": 0.7668595476065229,
+      "grad_norm": 2.0090813636779785,
+      "learning_rate": 2.7183434148223052e-05,
+      "loss": 1.1286,
+      "step": 7289
+    },
+    {
+      "epoch": 0.766964755391899,
+      "grad_norm": 1.4088441133499146,
+      "learning_rate": 2.7160081707778707e-05,
+      "loss": 1.2337,
+      "step": 7290
+    },
+    {
+      "epoch": 0.7670699631772752,
+      "grad_norm": 1.4657981395721436,
+      "learning_rate": 2.7136737725994187e-05,
+      "loss": 1.166,
+      "step": 7291
+    },
+    {
+      "epoch": 0.7671751709626512,
+      "grad_norm": 2.0377955436706543,
+      "learning_rate": 2.7113402205580408e-05,
+      "loss": 1.8475,
+      "step": 7292
+    },
+    {
+      "epoch": 0.7672803787480273,
+      "grad_norm": 1.972943663597107,
+      "learning_rate": 2.7090075149247217e-05,
+      "loss": 1.7478,
+      "step": 7293
+    },
+    {
+      "epoch": 0.7673855865334035,
+      "grad_norm": 1.5420266389846802,
+      "learning_rate": 2.7066756559703498e-05,
+      "loss": 1.8444,
+      "step": 7294
+    },
+    {
+      "epoch": 0.7674907943187795,
+      "grad_norm": 1.34401535987854,
+      "learning_rate": 2.7043446439657193e-05,
+      "loss": 1.4609,
+      "step": 7295
+    },
+    {
+      "epoch": 0.7675960021041557,
+      "grad_norm": 1.3176796436309814,
+      "learning_rate": 2.7020144791815218e-05,
+      "loss": 1.3681,
+      "step": 7296
+    },
+    {
+      "epoch": 0.7677012098895318,
+      "grad_norm": 1.1592514514923096,
+      "learning_rate": 2.699685161888348e-05,
+      "loss": 1.6415,
+      "step": 7297
+    },
+    {
+      "epoch": 0.767806417674908,
+      "grad_norm": 1.5379077196121216,
+      "learning_rate": 2.6973566923566994e-05,
+      "loss": 1.543,
+      "step": 7298
+    },
+    {
+      "epoch": 0.767911625460284,
+      "grad_norm": 2.4584033489227295,
+      "learning_rate": 2.6950290708569716e-05,
+      "loss": 1.558,
+      "step": 7299
+    },
+    {
+      "epoch": 0.7680168332456602,
+      "grad_norm": 1.853628396987915,
+      "learning_rate": 2.6927022976594607e-05,
+      "loss": 2.2037,
+      "step": 7300
+    },
+    {
+      "epoch": 0.7681220410310363,
+      "grad_norm": 1.1601117849349976,
+      "learning_rate": 2.6903763730343713e-05,
+      "loss": 1.8498,
+      "step": 7301
+    },
+    {
+      "epoch": 0.7682272488164125,
+      "grad_norm": 1.5906758308410645,
+      "learning_rate": 2.6880512972518047e-05,
+      "loss": 2.0151,
+      "step": 7302
+    },
+    {
+      "epoch": 0.7683324566017885,
+      "grad_norm": 1.081691026687622,
+      "learning_rate": 2.6857270705817595e-05,
+      "loss": 1.7113,
+      "step": 7303
+    },
+    {
+      "epoch": 0.7684376643871647,
+      "grad_norm": 1.651273250579834,
+      "learning_rate": 2.6834036932941474e-05,
+      "loss": 1.6251,
+      "step": 7304
+    },
+    {
+      "epoch": 0.7685428721725408,
+      "grad_norm": 1.15084969997406,
+      "learning_rate": 2.681081165658772e-05,
+      "loss": 1.5665,
+      "step": 7305
+    },
+    {
+      "epoch": 0.7686480799579168,
+      "grad_norm": 1.6445162296295166,
+      "learning_rate": 2.678759487945337e-05,
+      "loss": 1.7392,
+      "step": 7306
+    },
+    {
+      "epoch": 0.768753287743293,
+      "grad_norm": 1.2424595355987549,
+      "learning_rate": 2.676438660423457e-05,
+      "loss": 2.0125,
+      "step": 7307
+    },
+    {
+      "epoch": 0.7688584955286691,
+      "grad_norm": 2.284137487411499,
+      "learning_rate": 2.6741186833626407e-05,
+      "loss": 1.8287,
+      "step": 7308
+    },
+    {
+      "epoch": 0.7689637033140453,
+      "grad_norm": 2.1580452919006348,
+      "learning_rate": 2.6717995570322964e-05,
+      "loss": 1.9503,
+      "step": 7309
+    },
+    {
+      "epoch": 0.7690689110994213,
+      "grad_norm": 1.483157753944397,
+      "learning_rate": 2.669481281701739e-05,
+      "loss": 1.3867,
+      "step": 7310
+    },
+    {
+      "epoch": 0.7691741188847975,
+      "grad_norm": 1.2321768999099731,
+      "learning_rate": 2.667163857640179e-05,
+      "loss": 1.7758,
+      "step": 7311
+    },
+    {
+      "epoch": 0.7692793266701736,
+      "grad_norm": 1.4711391925811768,
+      "learning_rate": 2.664847285116736e-05,
+      "loss": 1.371,
+      "step": 7312
+    },
+    {
+      "epoch": 0.7693845344555497,
+      "grad_norm": 1.1832042932510376,
+      "learning_rate": 2.6625315644004244e-05,
+      "loss": 1.6168,
+      "step": 7313
+    },
+    {
+      "epoch": 0.7694897422409258,
+      "grad_norm": 1.5034308433532715,
+      "learning_rate": 2.660216695760157e-05,
+      "loss": 1.6241,
+      "step": 7314
+    },
+    {
+      "epoch": 0.769594950026302,
+      "grad_norm": 1.575340747833252,
+      "learning_rate": 2.6579026794647587e-05,
+      "loss": 1.6202,
+      "step": 7315
+    },
+    {
+      "epoch": 0.7697001578116781,
+      "grad_norm": 1.1293988227844238,
+      "learning_rate": 2.655589515782946e-05,
+      "loss": 1.7913,
+      "step": 7316
+    },
+    {
+      "epoch": 0.7698053655970541,
+      "grad_norm": 2.4325718879699707,
+      "learning_rate": 2.653277204983334e-05,
+      "loss": 1.902,
+      "step": 7317
+    },
+    {
+      "epoch": 0.7699105733824303,
+      "grad_norm": 1.1691190004348755,
+      "learning_rate": 2.650965747334452e-05,
+      "loss": 1.5993,
+      "step": 7318
+    },
+    {
+      "epoch": 0.7700157811678064,
+      "grad_norm": 1.888068675994873,
+      "learning_rate": 2.648655143104717e-05,
+      "loss": 1.406,
+      "step": 7319
+    },
+    {
+      "epoch": 0.7701209889531825,
+      "grad_norm": 1.8982930183410645,
+      "learning_rate": 2.6463453925624503e-05,
+      "loss": 1.584,
+      "step": 7320
+    },
+    {
+      "epoch": 0.7702261967385586,
+      "grad_norm": 1.547112226486206,
+      "learning_rate": 2.6440364959758813e-05,
+      "loss": 1.9128,
+      "step": 7321
+    },
+    {
+      "epoch": 0.7703314045239348,
+      "grad_norm": 2.7643849849700928,
+      "learning_rate": 2.641728453613127e-05,
+      "loss": 2.1104,
+      "step": 7322
+    },
+    {
+      "epoch": 0.7704366123093109,
+      "grad_norm": 1.3012899160385132,
+      "learning_rate": 2.6394212657422225e-05,
+      "loss": 1.4724,
+      "step": 7323
+    },
+    {
+      "epoch": 0.770541820094687,
+      "grad_norm": 1.8508931398391724,
+      "learning_rate": 2.6371149326310874e-05,
+      "loss": 1.8537,
+      "step": 7324
+    },
+    {
+      "epoch": 0.7706470278800631,
+      "grad_norm": 2.359945774078369,
+      "learning_rate": 2.6348094545475465e-05,
+      "loss": 1.7603,
+      "step": 7325
+    },
+    {
+      "epoch": 0.7707522356654393,
+      "grad_norm": 1.6346330642700195,
+      "learning_rate": 2.6325048317593337e-05,
+      "loss": 1.9917,
+      "step": 7326
+    },
+    {
+      "epoch": 0.7708574434508153,
+      "grad_norm": 1.718336820602417,
+      "learning_rate": 2.6302010645340746e-05,
+      "loss": 1.9148,
+      "step": 7327
+    },
+    {
+      "epoch": 0.7709626512361915,
+      "grad_norm": 2.0984418392181396,
+      "learning_rate": 2.6278981531392945e-05,
+      "loss": 1.852,
+      "step": 7328
+    },
+    {
+      "epoch": 0.7710678590215676,
+      "grad_norm": 1.1272097826004028,
+      "learning_rate": 2.625596097842432e-05,
+      "loss": 1.2315,
+      "step": 7329
+    },
+    {
+      "epoch": 0.7711730668069438,
+      "grad_norm": 1.9899998903274536,
+      "learning_rate": 2.6232948989108086e-05,
+      "loss": 1.5241,
+      "step": 7330
+    },
+    {
+      "epoch": 0.7712782745923198,
+      "grad_norm": 1.6918717622756958,
+      "learning_rate": 2.6209945566116545e-05,
+      "loss": 1.5811,
+      "step": 7331
+    },
+    {
+      "epoch": 0.7713834823776959,
+      "grad_norm": 2.1108553409576416,
+      "learning_rate": 2.618695071212107e-05,
+      "loss": 1.7367,
+      "step": 7332
+    },
+    {
+      "epoch": 0.7714886901630721,
+      "grad_norm": 2.049133062362671,
+      "learning_rate": 2.616396442979192e-05,
+      "loss": 1.7588,
+      "step": 7333
+    },
+    {
+      "epoch": 0.7715938979484482,
+      "grad_norm": 2.400456428527832,
+      "learning_rate": 2.6140986721798466e-05,
+      "loss": 1.5582,
+      "step": 7334
+    },
+    {
+      "epoch": 0.7716991057338243,
+      "grad_norm": 1.7933460474014282,
+      "learning_rate": 2.6118017590809017e-05,
+      "loss": 2.0447,
+      "step": 7335
+    },
+    {
+      "epoch": 0.7718043135192004,
+      "grad_norm": 1.4793000221252441,
+      "learning_rate": 2.6095057039490878e-05,
+      "loss": 1.9915,
+      "step": 7336
+    },
+    {
+      "epoch": 0.7719095213045766,
+      "grad_norm": 1.532442569732666,
+      "learning_rate": 2.6072105070510422e-05,
+      "loss": 1.7986,
+      "step": 7337
+    },
+    {
+      "epoch": 0.7720147290899526,
+      "grad_norm": 1.4576950073242188,
+      "learning_rate": 2.6049161686532965e-05,
+      "loss": 1.3017,
+      "step": 7338
+    },
+    {
+      "epoch": 0.7721199368753288,
+      "grad_norm": 1.2867484092712402,
+      "learning_rate": 2.6026226890222814e-05,
+      "loss": 1.3474,
+      "step": 7339
+    },
+    {
+      "epoch": 0.7722251446607049,
+      "grad_norm": 1.4471523761749268,
+      "learning_rate": 2.600330068424338e-05,
+      "loss": 1.8549,
+      "step": 7340
+    },
+    {
+      "epoch": 0.7723303524460811,
+      "grad_norm": 1.353736400604248,
+      "learning_rate": 2.5980383071256975e-05,
+      "loss": 1.4423,
+      "step": 7341
+    },
+    {
+      "epoch": 0.7724355602314571,
+      "grad_norm": 1.9009381532669067,
+      "learning_rate": 2.595747405392491e-05,
+      "loss": 1.3345,
+      "step": 7342
+    },
+    {
+      "epoch": 0.7725407680168332,
+      "grad_norm": 1.2530128955841064,
+      "learning_rate": 2.59345736349076e-05,
+      "loss": 1.5174,
+      "step": 7343
+    },
+    {
+      "epoch": 0.7726459758022094,
+      "grad_norm": 1.770195722579956,
+      "learning_rate": 2.5911681816864354e-05,
+      "loss": 1.6167,
+      "step": 7344
+    },
+    {
+      "epoch": 0.7727511835875854,
+      "grad_norm": 2.0301904678344727,
+      "learning_rate": 2.588879860245351e-05,
+      "loss": 1.6029,
+      "step": 7345
+    },
+    {
+      "epoch": 0.7728563913729616,
+      "grad_norm": 2.069531202316284,
+      "learning_rate": 2.5865923994332463e-05,
+      "loss": 1.8537,
+      "step": 7346
+    },
+    {
+      "epoch": 0.7729615991583377,
+      "grad_norm": 1.2178399562835693,
+      "learning_rate": 2.5843057995157548e-05,
+      "loss": 1.4361,
+      "step": 7347
+    },
+    {
+      "epoch": 0.7730668069437139,
+      "grad_norm": 1.5581814050674438,
+      "learning_rate": 2.582020060758409e-05,
+      "loss": 1.5575,
+      "step": 7348
+    },
+    {
+      "epoch": 0.7731720147290899,
+      "grad_norm": 1.6995668411254883,
+      "learning_rate": 2.579735183426649e-05,
+      "loss": 1.515,
+      "step": 7349
+    },
+    {
+      "epoch": 0.7732772225144661,
+      "grad_norm": 1.9909636974334717,
+      "learning_rate": 2.577451167785808e-05,
+      "loss": 2.097,
+      "step": 7350
+    },
+    {
+      "epoch": 0.7733824302998422,
+      "grad_norm": 1.49967622756958,
+      "learning_rate": 2.5751680141011214e-05,
+      "loss": 1.6389,
+      "step": 7351
+    },
+    {
+      "epoch": 0.7734876380852183,
+      "grad_norm": 1.4037280082702637,
+      "learning_rate": 2.5728857226377246e-05,
+      "loss": 1.6968,
+      "step": 7352
+    },
+    {
+      "epoch": 0.7735928458705944,
+      "grad_norm": 1.8199878931045532,
+      "learning_rate": 2.5706042936606477e-05,
+      "loss": 1.9098,
+      "step": 7353
+    },
+    {
+      "epoch": 0.7736980536559706,
+      "grad_norm": 1.3631402254104614,
+      "learning_rate": 2.5683237274348327e-05,
+      "loss": 1.8551,
+      "step": 7354
+    },
+    {
+      "epoch": 0.7738032614413467,
+      "grad_norm": 1.259263038635254,
+      "learning_rate": 2.5660440242251117e-05,
+      "loss": 1.6699,
+      "step": 7355
+    },
+    {
+      "epoch": 0.7739084692267227,
+      "grad_norm": 1.1155054569244385,
+      "learning_rate": 2.5637651842962164e-05,
+      "loss": 1.5838,
+      "step": 7356
+    },
+    {
+      "epoch": 0.7740136770120989,
+      "grad_norm": 1.8859355449676514,
+      "learning_rate": 2.561487207912785e-05,
+      "loss": 1.7185,
+      "step": 7357
+    },
+    {
+      "epoch": 0.774118884797475,
+      "grad_norm": 1.8426570892333984,
+      "learning_rate": 2.5592100953393504e-05,
+      "loss": 1.9184,
+      "step": 7358
+    },
+    {
+      "epoch": 0.7742240925828511,
+      "grad_norm": 1.5915162563323975,
+      "learning_rate": 2.5569338468403426e-05,
+      "loss": 1.8337,
+      "step": 7359
+    },
+    {
+      "epoch": 0.7743293003682272,
+      "grad_norm": 1.3176912069320679,
+      "learning_rate": 2.5546584626801006e-05,
+      "loss": 1.5611,
+      "step": 7360
+    },
+    {
+      "epoch": 0.7744345081536034,
+      "grad_norm": 1.1959328651428223,
+      "learning_rate": 2.5523839431228537e-05,
+      "loss": 2.022,
+      "step": 7361
+    },
+    {
+      "epoch": 0.7745397159389795,
+      "grad_norm": 1.117728590965271,
+      "learning_rate": 2.550110288432733e-05,
+      "loss": 2.0816,
+      "step": 7362
+    },
+    {
+      "epoch": 0.7746449237243556,
+      "grad_norm": 1.091355800628662,
+      "learning_rate": 2.5478374988737753e-05,
+      "loss": 1.4844,
+      "step": 7363
+    },
+    {
+      "epoch": 0.7747501315097317,
+      "grad_norm": 1.6453313827514648,
+      "learning_rate": 2.5455655747099093e-05,
+      "loss": 1.8608,
+      "step": 7364
+    },
+    {
+      "epoch": 0.7748553392951079,
+      "grad_norm": 1.3026262521743774,
+      "learning_rate": 2.543294516204964e-05,
+      "loss": 1.9639,
+      "step": 7365
+    },
+    {
+      "epoch": 0.774960547080484,
+      "grad_norm": 1.868897795677185,
+      "learning_rate": 2.541024323622674e-05,
+      "loss": 1.6702,
+      "step": 7366
+    },
+    {
+      "epoch": 0.77506575486586,
+      "grad_norm": 2.3049232959747314,
+      "learning_rate": 2.5387549972266678e-05,
+      "loss": 1.1612,
+      "step": 7367
+    },
+    {
+      "epoch": 0.7751709626512362,
+      "grad_norm": 1.9099220037460327,
+      "learning_rate": 2.5364865372804712e-05,
+      "loss": 1.3566,
+      "step": 7368
+    },
+    {
+      "epoch": 0.7752761704366123,
+      "grad_norm": 1.5426534414291382,
+      "learning_rate": 2.5342189440475204e-05,
+      "loss": 1.7221,
+      "step": 7369
+    },
+    {
+      "epoch": 0.7753813782219884,
+      "grad_norm": 1.1355007886886597,
+      "learning_rate": 2.531952217791136e-05,
+      "loss": 1.6406,
+      "step": 7370
+    },
+    {
+      "epoch": 0.7754865860073645,
+      "grad_norm": 1.2040497064590454,
+      "learning_rate": 2.529686358774551e-05,
+      "loss": 1.8762,
+      "step": 7371
+    },
+    {
+      "epoch": 0.7755917937927407,
+      "grad_norm": 1.0842825174331665,
+      "learning_rate": 2.5274213672608936e-05,
+      "loss": 1.9909,
+      "step": 7372
+    },
+    {
+      "epoch": 0.7756970015781168,
+      "grad_norm": 2.0430164337158203,
+      "learning_rate": 2.52515724351318e-05,
+      "loss": 1.8172,
+      "step": 7373
+    },
+    {
+      "epoch": 0.7758022093634929,
+      "grad_norm": 1.6244868040084839,
+      "learning_rate": 2.5228939877943448e-05,
+      "loss": 1.9493,
+      "step": 7374
+    },
+    {
+      "epoch": 0.775907417148869,
+      "grad_norm": 2.7612531185150146,
+      "learning_rate": 2.520631600367209e-05,
+      "loss": 2.1905,
+      "step": 7375
+    },
+    {
+      "epoch": 0.7760126249342452,
+      "grad_norm": 1.8194602727890015,
+      "learning_rate": 2.5183700814944945e-05,
+      "loss": 1.7323,
+      "step": 7376
+    },
+    {
+      "epoch": 0.7761178327196212,
+      "grad_norm": 1.4295552968978882,
+      "learning_rate": 2.5161094314388278e-05,
+      "loss": 1.9286,
+      "step": 7377
+    },
+    {
+      "epoch": 0.7762230405049974,
+      "grad_norm": 1.6510827541351318,
+      "learning_rate": 2.5138496504627263e-05,
+      "loss": 1.4062,
+      "step": 7378
+    },
+    {
+      "epoch": 0.7763282482903735,
+      "grad_norm": 1.3370580673217773,
+      "learning_rate": 2.5115907388286165e-05,
+      "loss": 0.8727,
+      "step": 7379
+    },
+    {
+      "epoch": 0.7764334560757497,
+      "grad_norm": 1.845016598701477,
+      "learning_rate": 2.509332696798816e-05,
+      "loss": 2.2291,
+      "step": 7380
+    },
+    {
+      "epoch": 0.7765386638611257,
+      "grad_norm": 2.652451515197754,
+      "learning_rate": 2.5070755246355393e-05,
+      "loss": 1.9059,
+      "step": 7381
+    },
+    {
+      "epoch": 0.7766438716465018,
+      "grad_norm": 1.796697974205017,
+      "learning_rate": 2.5048192226009126e-05,
+      "loss": 0.9683,
+      "step": 7382
+    },
+    {
+      "epoch": 0.776749079431878,
+      "grad_norm": 2.484699010848999,
+      "learning_rate": 2.5025637909569475e-05,
+      "loss": 1.9713,
+      "step": 7383
+    },
+    {
+      "epoch": 0.776854287217254,
+      "grad_norm": 1.428265929222107,
+      "learning_rate": 2.5003092299655584e-05,
+      "loss": 1.8807,
+      "step": 7384
+    },
+    {
+      "epoch": 0.7769594950026302,
+      "grad_norm": 1.6806254386901855,
+      "learning_rate": 2.4980555398885653e-05,
+      "loss": 2.3045,
+      "step": 7385
+    },
+    {
+      "epoch": 0.7770647027880063,
+      "grad_norm": 1.6273754835128784,
+      "learning_rate": 2.4958027209876788e-05,
+      "loss": 1.5729,
+      "step": 7386
+    },
+    {
+      "epoch": 0.7771699105733825,
+      "grad_norm": 1.8025354146957397,
+      "learning_rate": 2.493550773524509e-05,
+      "loss": 1.7046,
+      "step": 7387
+    },
+    {
+      "epoch": 0.7772751183587585,
+      "grad_norm": 1.5753192901611328,
+      "learning_rate": 2.4912996977605718e-05,
+      "loss": 1.2373,
+      "step": 7388
+    },
+    {
+      "epoch": 0.7773803261441347,
+      "grad_norm": 1.7229233980178833,
+      "learning_rate": 2.4890494939572762e-05,
+      "loss": 1.7167,
+      "step": 7389
+    },
+    {
+      "epoch": 0.7774855339295108,
+      "grad_norm": 2.0432682037353516,
+      "learning_rate": 2.4868001623759263e-05,
+      "loss": 1.456,
+      "step": 7390
+    },
+    {
+      "epoch": 0.7775907417148868,
+      "grad_norm": 1.1946852207183838,
+      "learning_rate": 2.4845517032777364e-05,
+      "loss": 1.5981,
+      "step": 7391
+    },
+    {
+      "epoch": 0.777695949500263,
+      "grad_norm": 1.7232786417007446,
+      "learning_rate": 2.4823041169238092e-05,
+      "loss": 1.8003,
+      "step": 7392
+    },
+    {
+      "epoch": 0.7778011572856391,
+      "grad_norm": 2.1846988201141357,
+      "learning_rate": 2.480057403575148e-05,
+      "loss": 1.6615,
+      "step": 7393
+    },
+    {
+      "epoch": 0.7779063650710153,
+      "grad_norm": 1.523494839668274,
+      "learning_rate": 2.4778115634926624e-05,
+      "loss": 1.9023,
+      "step": 7394
+    },
+    {
+      "epoch": 0.7780115728563913,
+      "grad_norm": 1.5766888856887817,
+      "learning_rate": 2.4755665969371446e-05,
+      "loss": 1.6448,
+      "step": 7395
+    },
+    {
+      "epoch": 0.7781167806417675,
+      "grad_norm": 1.718214750289917,
+      "learning_rate": 2.4733225041693033e-05,
+      "loss": 1.8259,
+      "step": 7396
+    },
+    {
+      "epoch": 0.7782219884271436,
+      "grad_norm": 1.3229570388793945,
+      "learning_rate": 2.4710792854497346e-05,
+      "loss": 1.775,
+      "step": 7397
+    },
+    {
+      "epoch": 0.7783271962125198,
+      "grad_norm": 1.1155831813812256,
+      "learning_rate": 2.4688369410389334e-05,
+      "loss": 1.6853,
+      "step": 7398
+    },
+    {
+      "epoch": 0.7784324039978958,
+      "grad_norm": 1.7406342029571533,
+      "learning_rate": 2.4665954711973017e-05,
+      "loss": 1.4841,
+      "step": 7399
+    },
+    {
+      "epoch": 0.778537611783272,
+      "grad_norm": 1.489790439605713,
+      "learning_rate": 2.464354876185131e-05,
+      "loss": 1.4955,
+      "step": 7400
+    },
+    {
+      "epoch": 0.7786428195686481,
+      "grad_norm": 1.8641880750656128,
+      "learning_rate": 2.462115156262612e-05,
+      "loss": 1.6168,
+      "step": 7401
+    },
+    {
+      "epoch": 0.7787480273540242,
+      "grad_norm": 1.3012237548828125,
+      "learning_rate": 2.4598763116898405e-05,
+      "loss": 2.0688,
+      "step": 7402
+    },
+    {
+      "epoch": 0.7788532351394003,
+      "grad_norm": 2.4502408504486084,
+      "learning_rate": 2.4576383427268034e-05,
+      "loss": 1.3047,
+      "step": 7403
+    },
+    {
+      "epoch": 0.7789584429247765,
+      "grad_norm": 1.7557679414749146,
+      "learning_rate": 2.455401249633387e-05,
+      "loss": 1.3024,
+      "step": 7404
+    },
+    {
+      "epoch": 0.7790636507101526,
+      "grad_norm": 2.018825054168701,
+      "learning_rate": 2.4531650326693822e-05,
+      "loss": 2.3045,
+      "step": 7405
+    },
+    {
+      "epoch": 0.7791688584955286,
+      "grad_norm": 1.1796764135360718,
+      "learning_rate": 2.4509296920944712e-05,
+      "loss": 1.772,
+      "step": 7406
+    },
+    {
+      "epoch": 0.7792740662809048,
+      "grad_norm": 1.2721928358078003,
+      "learning_rate": 2.448695228168234e-05,
+      "loss": 1.2919,
+      "step": 7407
+    },
+    {
+      "epoch": 0.7793792740662809,
+      "grad_norm": 1.6079884767532349,
+      "learning_rate": 2.4464616411501572e-05,
+      "loss": 1.4524,
+      "step": 7408
+    },
+    {
+      "epoch": 0.779484481851657,
+      "grad_norm": 1.7307249307632446,
+      "learning_rate": 2.4442289312996158e-05,
+      "loss": 1.9304,
+      "step": 7409
+    },
+    {
+      "epoch": 0.7795896896370331,
+      "grad_norm": 1.4798263311386108,
+      "learning_rate": 2.4419970988758857e-05,
+      "loss": 1.9729,
+      "step": 7410
+    },
+    {
+      "epoch": 0.7796948974224093,
+      "grad_norm": 1.2387443780899048,
+      "learning_rate": 2.439766144138148e-05,
+      "loss": 1.5655,
+      "step": 7411
+    },
+    {
+      "epoch": 0.7798001052077854,
+      "grad_norm": 1.4330637454986572,
+      "learning_rate": 2.4375360673454718e-05,
+      "loss": 1.3443,
+      "step": 7412
+    },
+    {
+      "epoch": 0.7799053129931615,
+      "grad_norm": 1.1398248672485352,
+      "learning_rate": 2.435306868756827e-05,
+      "loss": 1.8242,
+      "step": 7413
+    },
+    {
+      "epoch": 0.7800105207785376,
+      "grad_norm": 2.0635628700256348,
+      "learning_rate": 2.433078548631088e-05,
+      "loss": 1.3481,
+      "step": 7414
+    },
+    {
+      "epoch": 0.7801157285639138,
+      "grad_norm": 1.374840259552002,
+      "learning_rate": 2.4308511072270202e-05,
+      "loss": 1.5717,
+      "step": 7415
+    },
+    {
+      "epoch": 0.7802209363492898,
+      "grad_norm": 1.5022417306900024,
+      "learning_rate": 2.4286245448032895e-05,
+      "loss": 1.6766,
+      "step": 7416
+    },
+    {
+      "epoch": 0.7803261441346659,
+      "grad_norm": 1.2573763132095337,
+      "learning_rate": 2.4263988616184574e-05,
+      "loss": 1.5103,
+      "step": 7417
+    },
+    {
+      "epoch": 0.7804313519200421,
+      "grad_norm": 1.6818538904190063,
+      "learning_rate": 2.4241740579309836e-05,
+      "loss": 0.8336,
+      "step": 7418
+    },
+    {
+      "epoch": 0.7805365597054182,
+      "grad_norm": 1.3387900590896606,
+      "learning_rate": 2.4219501339992334e-05,
+      "loss": 1.8443,
+      "step": 7419
+    },
+    {
+      "epoch": 0.7806417674907943,
+      "grad_norm": 1.4211339950561523,
+      "learning_rate": 2.4197270900814594e-05,
+      "loss": 1.7912,
+      "step": 7420
+    },
+    {
+      "epoch": 0.7807469752761704,
+      "grad_norm": 1.4029207229614258,
+      "learning_rate": 2.417504926435814e-05,
+      "loss": 1.2693,
+      "step": 7421
+    },
+    {
+      "epoch": 0.7808521830615466,
+      "grad_norm": 1.4126625061035156,
+      "learning_rate": 2.415283643320356e-05,
+      "loss": 1.4289,
+      "step": 7422
+    },
+    {
+      "epoch": 0.7809573908469227,
+      "grad_norm": 1.4173007011413574,
+      "learning_rate": 2.413063240993031e-05,
+      "loss": 1.8065,
+      "step": 7423
+    },
+    {
+      "epoch": 0.7810625986322988,
+      "grad_norm": 1.166293740272522,
+      "learning_rate": 2.4108437197116905e-05,
+      "loss": 1.9179,
+      "step": 7424
+    },
+    {
+      "epoch": 0.7811678064176749,
+      "grad_norm": 1.7653440237045288,
+      "learning_rate": 2.408625079734078e-05,
+      "loss": 1.5022,
+      "step": 7425
+    },
+    {
+      "epoch": 0.7812730142030511,
+      "grad_norm": 1.2314149141311646,
+      "learning_rate": 2.406407321317835e-05,
+      "loss": 1.5374,
+      "step": 7426
+    },
+    {
+      "epoch": 0.7813782219884271,
+      "grad_norm": 1.553765058517456,
+      "learning_rate": 2.4041904447205067e-05,
+      "loss": 1.6092,
+      "step": 7427
+    },
+    {
+      "epoch": 0.7814834297738033,
+      "grad_norm": 1.543643593788147,
+      "learning_rate": 2.40197445019953e-05,
+      "loss": 1.2104,
+      "step": 7428
+    },
+    {
+      "epoch": 0.7815886375591794,
+      "grad_norm": 1.2574193477630615,
+      "learning_rate": 2.3997593380122386e-05,
+      "loss": 1.6868,
+      "step": 7429
+    },
+    {
+      "epoch": 0.7816938453445555,
+      "grad_norm": 1.3462152481079102,
+      "learning_rate": 2.3975451084158707e-05,
+      "loss": 1.5889,
+      "step": 7430
+    },
+    {
+      "epoch": 0.7817990531299316,
+      "grad_norm": 1.5161739587783813,
+      "learning_rate": 2.395331761667554e-05,
+      "loss": 1.805,
+      "step": 7431
+    },
+    {
+      "epoch": 0.7819042609153077,
+      "grad_norm": 1.5476092100143433,
+      "learning_rate": 2.3931192980243166e-05,
+      "loss": 1.4936,
+      "step": 7432
+    },
+    {
+      "epoch": 0.7820094687006839,
+      "grad_norm": 1.412503957748413,
+      "learning_rate": 2.3909077177430893e-05,
+      "loss": 1.5105,
+      "step": 7433
+    },
+    {
+      "epoch": 0.7821146764860599,
+      "grad_norm": 1.7502893209457397,
+      "learning_rate": 2.3886970210806915e-05,
+      "loss": 1.1703,
+      "step": 7434
+    },
+    {
+      "epoch": 0.7822198842714361,
+      "grad_norm": 1.8714232444763184,
+      "learning_rate": 2.3864872082938426e-05,
+      "loss": 1.2081,
+      "step": 7435
+    },
+    {
+      "epoch": 0.7823250920568122,
+      "grad_norm": 1.5723828077316284,
+      "learning_rate": 2.3842782796391672e-05,
+      "loss": 1.4935,
+      "step": 7436
+    },
+    {
+      "epoch": 0.7824302998421884,
+      "grad_norm": 1.6067105531692505,
+      "learning_rate": 2.3820702353731773e-05,
+      "loss": 1.9455,
+      "step": 7437
+    },
+    {
+      "epoch": 0.7825355076275644,
+      "grad_norm": 1.2671598196029663,
+      "learning_rate": 2.3798630757522844e-05,
+      "loss": 1.6182,
+      "step": 7438
+    },
+    {
+      "epoch": 0.7826407154129406,
+      "grad_norm": 1.860741138458252,
+      "learning_rate": 2.3776568010328003e-05,
+      "loss": 1.8387,
+      "step": 7439
+    },
+    {
+      "epoch": 0.7827459231983167,
+      "grad_norm": 1.9345104694366455,
+      "learning_rate": 2.3754514114709304e-05,
+      "loss": 1.4232,
+      "step": 7440
+    },
+    {
+      "epoch": 0.7828511309836927,
+      "grad_norm": 2.1556923389434814,
+      "learning_rate": 2.3732469073227827e-05,
+      "loss": 1.588,
+      "step": 7441
+    },
+    {
+      "epoch": 0.7829563387690689,
+      "grad_norm": 1.9754849672317505,
+      "learning_rate": 2.371043288844358e-05,
+      "loss": 1.6852,
+      "step": 7442
+    },
+    {
+      "epoch": 0.783061546554445,
+      "grad_norm": 1.2634505033493042,
+      "learning_rate": 2.3688405562915517e-05,
+      "loss": 1.8126,
+      "step": 7443
+    },
+    {
+      "epoch": 0.7831667543398212,
+      "grad_norm": 2.078622579574585,
+      "learning_rate": 2.3666387099201648e-05,
+      "loss": 0.9941,
+      "step": 7444
+    },
+    {
+      "epoch": 0.7832719621251972,
+      "grad_norm": 1.6752347946166992,
+      "learning_rate": 2.3644377499858893e-05,
+      "loss": 1.8038,
+      "step": 7445
+    },
+    {
+      "epoch": 0.7833771699105734,
+      "grad_norm": 1.0644302368164062,
+      "learning_rate": 2.3622376767443123e-05,
+      "loss": 1.8299,
+      "step": 7446
+    },
+    {
+      "epoch": 0.7834823776959495,
+      "grad_norm": 1.3895219564437866,
+      "learning_rate": 2.3600384904509254e-05,
+      "loss": 1.8465,
+      "step": 7447
+    },
+    {
+      "epoch": 0.7835875854813256,
+      "grad_norm": 1.346755862236023,
+      "learning_rate": 2.3578401913611103e-05,
+      "loss": 1.5028,
+      "step": 7448
+    },
+    {
+      "epoch": 0.7836927932667017,
+      "grad_norm": 1.92898428440094,
+      "learning_rate": 2.3556427797301462e-05,
+      "loss": 1.1488,
+      "step": 7449
+    },
+    {
+      "epoch": 0.7837980010520779,
+      "grad_norm": 1.4644855260849,
+      "learning_rate": 2.3534462558132177e-05,
+      "loss": 1.8202,
+      "step": 7450
+    },
+    {
+      "epoch": 0.783903208837454,
+      "grad_norm": 1.326379656791687,
+      "learning_rate": 2.3512506198653948e-05,
+      "loss": 1.3529,
+      "step": 7451
+    },
+    {
+      "epoch": 0.78400841662283,
+      "grad_norm": 1.445070505142212,
+      "learning_rate": 2.3490558721416477e-05,
+      "loss": 1.7824,
+      "step": 7452
+    },
+    {
+      "epoch": 0.7841136244082062,
+      "grad_norm": 1.5502876043319702,
+      "learning_rate": 2.346862012896852e-05,
+      "loss": 1.8986,
+      "step": 7453
+    },
+    {
+      "epoch": 0.7842188321935823,
+      "grad_norm": 1.3673009872436523,
+      "learning_rate": 2.3446690423857685e-05,
+      "loss": 1.4538,
+      "step": 7454
+    },
+    {
+      "epoch": 0.7843240399789585,
+      "grad_norm": 2.0649657249450684,
+      "learning_rate": 2.3424769608630593e-05,
+      "loss": 1.7845,
+      "step": 7455
+    },
+    {
+      "epoch": 0.7844292477643345,
+      "grad_norm": 1.5386171340942383,
+      "learning_rate": 2.340285768583287e-05,
+      "loss": 2.1823,
+      "step": 7456
+    },
+    {
+      "epoch": 0.7845344555497107,
+      "grad_norm": 2.3255977630615234,
+      "learning_rate": 2.3380954658009057e-05,
+      "loss": 1.5615,
+      "step": 7457
+    },
+    {
+      "epoch": 0.7846396633350868,
+      "grad_norm": 2.3962433338165283,
+      "learning_rate": 2.335906052770267e-05,
+      "loss": 1.4474,
+      "step": 7458
+    },
+    {
+      "epoch": 0.7847448711204629,
+      "grad_norm": 1.4439377784729004,
+      "learning_rate": 2.3337175297456225e-05,
+      "loss": 2.055,
+      "step": 7459
+    },
+    {
+      "epoch": 0.784850078905839,
+      "grad_norm": 1.2580662965774536,
+      "learning_rate": 2.3315298969811127e-05,
+      "loss": 1.9116,
+      "step": 7460
+    },
+    {
+      "epoch": 0.7849552866912152,
+      "grad_norm": 2.165815830230713,
+      "learning_rate": 2.3293431547307887e-05,
+      "loss": 1.5773,
+      "step": 7461
+    },
+    {
+      "epoch": 0.7850604944765913,
+      "grad_norm": 1.3377245664596558,
+      "learning_rate": 2.327157303248584e-05,
+      "loss": 1.23,
+      "step": 7462
+    },
+    {
+      "epoch": 0.7851657022619674,
+      "grad_norm": 1.4234360456466675,
+      "learning_rate": 2.324972342788333e-05,
+      "loss": 1.5741,
+      "step": 7463
+    },
+    {
+      "epoch": 0.7852709100473435,
+      "grad_norm": 1.4981180429458618,
+      "learning_rate": 2.3227882736037732e-05,
+      "loss": 1.6655,
+      "step": 7464
+    },
+    {
+      "epoch": 0.7853761178327197,
+      "grad_norm": 1.7403837442398071,
+      "learning_rate": 2.3206050959485314e-05,
+      "loss": 1.9563,
+      "step": 7465
+    },
+    {
+      "epoch": 0.7854813256180957,
+      "grad_norm": 1.0999480485916138,
+      "learning_rate": 2.3184228100761285e-05,
+      "loss": 1.9103,
+      "step": 7466
+    },
+    {
+      "epoch": 0.7855865334034718,
+      "grad_norm": 1.1709569692611694,
+      "learning_rate": 2.316241416239994e-05,
+      "loss": 1.8448,
+      "step": 7467
+    },
+    {
+      "epoch": 0.785691741188848,
+      "grad_norm": 2.044626235961914,
+      "learning_rate": 2.31406091469344e-05,
+      "loss": 1.7845,
+      "step": 7468
+    },
+    {
+      "epoch": 0.7857969489742241,
+      "grad_norm": 1.8200260400772095,
+      "learning_rate": 2.3118813056896814e-05,
+      "loss": 1.8616,
+      "step": 7469
+    },
+    {
+      "epoch": 0.7859021567596002,
+      "grad_norm": 1.0556297302246094,
+      "learning_rate": 2.3097025894818326e-05,
+      "loss": 1.9248,
+      "step": 7470
+    },
+    {
+      "epoch": 0.7860073645449763,
+      "grad_norm": 1.9120993614196777,
+      "learning_rate": 2.307524766322896e-05,
+      "loss": 1.1505,
+      "step": 7471
+    },
+    {
+      "epoch": 0.7861125723303525,
+      "grad_norm": 1.3910388946533203,
+      "learning_rate": 2.30534783646578e-05,
+      "loss": 2.0402,
+      "step": 7472
+    },
+    {
+      "epoch": 0.7862177801157285,
+      "grad_norm": 2.19036602973938,
+      "learning_rate": 2.303171800163282e-05,
+      "loss": 1.7415,
+      "step": 7473
+    },
+    {
+      "epoch": 0.7863229879011047,
+      "grad_norm": 1.8024297952651978,
+      "learning_rate": 2.300996657668095e-05,
+      "loss": 1.6391,
+      "step": 7474
+    },
+    {
+      "epoch": 0.7864281956864808,
+      "grad_norm": 1.367393136024475,
+      "learning_rate": 2.298822409232817e-05,
+      "loss": 1.9299,
+      "step": 7475
+    },
+    {
+      "epoch": 0.786533403471857,
+      "grad_norm": 1.068220615386963,
+      "learning_rate": 2.2966490551099328e-05,
+      "loss": 1.2364,
+      "step": 7476
+    },
+    {
+      "epoch": 0.786638611257233,
+      "grad_norm": 1.452365756034851,
+      "learning_rate": 2.2944765955518242e-05,
+      "loss": 1.0474,
+      "step": 7477
+    },
+    {
+      "epoch": 0.7867438190426091,
+      "grad_norm": 1.1206425428390503,
+      "learning_rate": 2.2923050308107785e-05,
+      "loss": 1.9693,
+      "step": 7478
+    },
+    {
+      "epoch": 0.7868490268279853,
+      "grad_norm": 1.8541259765625,
+      "learning_rate": 2.290134361138968e-05,
+      "loss": 1.7089,
+      "step": 7479
+    },
+    {
+      "epoch": 0.7869542346133613,
+      "grad_norm": 1.5118783712387085,
+      "learning_rate": 2.287964586788467e-05,
+      "loss": 1.7571,
+      "step": 7480
+    },
+    {
+      "epoch": 0.7870594423987375,
+      "grad_norm": 1.5137947797775269,
+      "learning_rate": 2.2857957080112423e-05,
+      "loss": 1.0681,
+      "step": 7481
+    },
+    {
+      "epoch": 0.7871646501841136,
+      "grad_norm": 1.3139784336090088,
+      "learning_rate": 2.2836277250591574e-05,
+      "loss": 1.2861,
+      "step": 7482
+    },
+    {
+      "epoch": 0.7872698579694898,
+      "grad_norm": 2.15798020362854,
+      "learning_rate": 2.2814606381839786e-05,
+      "loss": 1.4572,
+      "step": 7483
+    },
+    {
+      "epoch": 0.7873750657548658,
+      "grad_norm": 1.4466586112976074,
+      "learning_rate": 2.279294447637359e-05,
+      "loss": 1.554,
+      "step": 7484
+    },
+    {
+      "epoch": 0.787480273540242,
+      "grad_norm": 2.4875476360321045,
+      "learning_rate": 2.2771291536708494e-05,
+      "loss": 2.1975,
+      "step": 7485
+    },
+    {
+      "epoch": 0.7875854813256181,
+      "grad_norm": 1.745013952255249,
+      "learning_rate": 2.2749647565359024e-05,
+      "loss": 1.6658,
+      "step": 7486
+    },
+    {
+      "epoch": 0.7876906891109943,
+      "grad_norm": 0.8909004926681519,
+      "learning_rate": 2.2728012564838608e-05,
+      "loss": 1.9497,
+      "step": 7487
+    },
+    {
+      "epoch": 0.7877958968963703,
+      "grad_norm": 1.4795335531234741,
+      "learning_rate": 2.2706386537659606e-05,
+      "loss": 1.4597,
+      "step": 7488
+    },
+    {
+      "epoch": 0.7879011046817465,
+      "grad_norm": 1.7913676500320435,
+      "learning_rate": 2.2684769486333445e-05,
+      "loss": 1.4416,
+      "step": 7489
+    },
+    {
+      "epoch": 0.7880063124671226,
+      "grad_norm": 1.4431122541427612,
+      "learning_rate": 2.2663161413370415e-05,
+      "loss": 1.8649,
+      "step": 7490
+    },
+    {
+      "epoch": 0.7881115202524986,
+      "grad_norm": 1.1279875040054321,
+      "learning_rate": 2.2641562321279752e-05,
+      "loss": 1.5093,
+      "step": 7491
+    },
+    {
+      "epoch": 0.7882167280378748,
+      "grad_norm": 1.534596562385559,
+      "learning_rate": 2.2619972212569752e-05,
+      "loss": 1.663,
+      "step": 7492
+    },
+    {
+      "epoch": 0.7883219358232509,
+      "grad_norm": 1.2349904775619507,
+      "learning_rate": 2.259839108974757e-05,
+      "loss": 1.8006,
+      "step": 7493
+    },
+    {
+      "epoch": 0.7884271436086271,
+      "grad_norm": 1.3401429653167725,
+      "learning_rate": 2.2576818955319333e-05,
+      "loss": 1.2232,
+      "step": 7494
+    },
+    {
+      "epoch": 0.7885323513940031,
+      "grad_norm": 1.3276652097702026,
+      "learning_rate": 2.2555255811790177e-05,
+      "loss": 1.3863,
+      "step": 7495
+    },
+    {
+      "epoch": 0.7886375591793793,
+      "grad_norm": 2.36476731300354,
+      "learning_rate": 2.2533701661664154e-05,
+      "loss": 1.9646,
+      "step": 7496
+    },
+    {
+      "epoch": 0.7887427669647554,
+      "grad_norm": 1.2848833799362183,
+      "learning_rate": 2.251215650744424e-05,
+      "loss": 1.5951,
+      "step": 7497
+    },
+    {
+      "epoch": 0.7888479747501315,
+      "grad_norm": 1.5567632913589478,
+      "learning_rate": 2.2490620351632452e-05,
+      "loss": 1.1698,
+      "step": 7498
+    },
+    {
+      "epoch": 0.7889531825355076,
+      "grad_norm": 2.238565444946289,
+      "learning_rate": 2.2469093196729696e-05,
+      "loss": 1.9644,
+      "step": 7499
+    },
+    {
+      "epoch": 0.7890583903208838,
+      "grad_norm": 1.1401231288909912,
+      "learning_rate": 2.24475750452358e-05,
+      "loss": 1.5368,
+      "step": 7500
+    },
+    {
+      "epoch": 0.7891635981062599,
+      "grad_norm": 2.0586628913879395,
+      "learning_rate": 2.242606589964972e-05,
+      "loss": 1.9507,
+      "step": 7501
+    },
+    {
+      "epoch": 0.789268805891636,
+      "grad_norm": 1.760940670967102,
+      "learning_rate": 2.2404565762469088e-05,
+      "loss": 1.8004,
+      "step": 7502
+    },
+    {
+      "epoch": 0.7893740136770121,
+      "grad_norm": 1.8332325220108032,
+      "learning_rate": 2.2383074636190748e-05,
+      "loss": 1.1507,
+      "step": 7503
+    },
+    {
+      "epoch": 0.7894792214623882,
+      "grad_norm": 1.1777528524398804,
+      "learning_rate": 2.236159252331037e-05,
+      "loss": 1.8222,
+      "step": 7504
+    },
+    {
+      "epoch": 0.7895844292477643,
+      "grad_norm": 2.514356851577759,
+      "learning_rate": 2.234011942632257e-05,
+      "loss": 1.0901,
+      "step": 7505
+    },
+    {
+      "epoch": 0.7896896370331404,
+      "grad_norm": 1.4754233360290527,
+      "learning_rate": 2.2318655347720995e-05,
+      "loss": 1.3529,
+      "step": 7506
+    },
+    {
+      "epoch": 0.7897948448185166,
+      "grad_norm": 1.38016939163208,
+      "learning_rate": 2.2297200289998176e-05,
+      "loss": 1.4848,
+      "step": 7507
+    },
+    {
+      "epoch": 0.7899000526038927,
+      "grad_norm": 2.304391384124756,
+      "learning_rate": 2.2275754255645587e-05,
+      "loss": 1.2804,
+      "step": 7508
+    },
+    {
+      "epoch": 0.7900052603892688,
+      "grad_norm": 2.2973196506500244,
+      "learning_rate": 2.2254317247153746e-05,
+      "loss": 1.3045,
+      "step": 7509
+    },
+    {
+      "epoch": 0.7901104681746449,
+      "grad_norm": 1.2090044021606445,
+      "learning_rate": 2.2232889267012038e-05,
+      "loss": 2.2979,
+      "step": 7510
+    },
+    {
+      "epoch": 0.7902156759600211,
+      "grad_norm": 1.6775548458099365,
+      "learning_rate": 2.221147031770878e-05,
+      "loss": 1.7299,
+      "step": 7511
+    },
+    {
+      "epoch": 0.7903208837453971,
+      "grad_norm": 1.9138872623443604,
+      "learning_rate": 2.2190060401731362e-05,
+      "loss": 1.9317,
+      "step": 7512
+    },
+    {
+      "epoch": 0.7904260915307733,
+      "grad_norm": 1.4541093111038208,
+      "learning_rate": 2.2168659521566004e-05,
+      "loss": 1.3817,
+      "step": 7513
+    },
+    {
+      "epoch": 0.7905312993161494,
+      "grad_norm": 2.349430799484253,
+      "learning_rate": 2.2147267679697892e-05,
+      "loss": 1.1925,
+      "step": 7514
+    },
+    {
+      "epoch": 0.7906365071015256,
+      "grad_norm": 2.0313339233398438,
+      "learning_rate": 2.2125884878611258e-05,
+      "loss": 1.643,
+      "step": 7515
+    },
+    {
+      "epoch": 0.7907417148869016,
+      "grad_norm": 1.2877901792526245,
+      "learning_rate": 2.210451112078914e-05,
+      "loss": 1.2234,
+      "step": 7516
+    },
+    {
+      "epoch": 0.7908469226722777,
+      "grad_norm": 1.4288432598114014,
+      "learning_rate": 2.2083146408713673e-05,
+      "loss": 1.4857,
+      "step": 7517
+    },
+    {
+      "epoch": 0.7909521304576539,
+      "grad_norm": 1.7954027652740479,
+      "learning_rate": 2.206179074486584e-05,
+      "loss": 1.359,
+      "step": 7518
+    },
+    {
+      "epoch": 0.79105733824303,
+      "grad_norm": 1.427760124206543,
+      "learning_rate": 2.204044413172558e-05,
+      "loss": 2.6701,
+      "step": 7519
+    },
+    {
+      "epoch": 0.7911625460284061,
+      "grad_norm": 1.8676557540893555,
+      "learning_rate": 2.201910657177185e-05,
+      "loss": 1.5542,
+      "step": 7520
+    },
+    {
+      "epoch": 0.7912677538137822,
+      "grad_norm": 2.1520488262176514,
+      "learning_rate": 2.19977780674825e-05,
+      "loss": 1.6475,
+      "step": 7521
+    },
+    {
+      "epoch": 0.7913729615991584,
+      "grad_norm": 2.3536930084228516,
+      "learning_rate": 2.1976458621334317e-05,
+      "loss": 1.2516,
+      "step": 7522
+    },
+    {
+      "epoch": 0.7914781693845344,
+      "grad_norm": 1.8606661558151245,
+      "learning_rate": 2.195514823580307e-05,
+      "loss": 1.6912,
+      "step": 7523
+    },
+    {
+      "epoch": 0.7915833771699106,
+      "grad_norm": 1.4110456705093384,
+      "learning_rate": 2.1933846913363466e-05,
+      "loss": 1.6379,
+      "step": 7524
+    },
+    {
+      "epoch": 0.7916885849552867,
+      "grad_norm": 1.5939291715621948,
+      "learning_rate": 2.1912554656489127e-05,
+      "loss": 1.5101,
+      "step": 7525
+    },
+    {
+      "epoch": 0.7917937927406629,
+      "grad_norm": 1.7965341806411743,
+      "learning_rate": 2.1891271467652696e-05,
+      "loss": 1.6246,
+      "step": 7526
+    },
+    {
+      "epoch": 0.7918990005260389,
+      "grad_norm": 1.8288224935531616,
+      "learning_rate": 2.186999734932569e-05,
+      "loss": 1.5909,
+      "step": 7527
+    },
+    {
+      "epoch": 0.792004208311415,
+      "grad_norm": 1.5028138160705566,
+      "learning_rate": 2.1848732303978638e-05,
+      "loss": 1.6557,
+      "step": 7528
+    },
+    {
+      "epoch": 0.7921094160967912,
+      "grad_norm": 1.5027766227722168,
+      "learning_rate": 2.1827476334080953e-05,
+      "loss": 1.3482,
+      "step": 7529
+    },
+    {
+      "epoch": 0.7922146238821672,
+      "grad_norm": 1.7562695741653442,
+      "learning_rate": 2.1806229442101e-05,
+      "loss": 2.0568,
+      "step": 7530
+    },
+    {
+      "epoch": 0.7923198316675434,
+      "grad_norm": 1.675010085105896,
+      "learning_rate": 2.178499163050617e-05,
+      "loss": 1.6856,
+      "step": 7531
+    },
+    {
+      "epoch": 0.7924250394529195,
+      "grad_norm": 1.9306681156158447,
+      "learning_rate": 2.1763762901762696e-05,
+      "loss": 1.4367,
+      "step": 7532
+    },
+    {
+      "epoch": 0.7925302472382957,
+      "grad_norm": 1.7994093894958496,
+      "learning_rate": 2.174254325833579e-05,
+      "loss": 1.6346,
+      "step": 7533
+    },
+    {
+      "epoch": 0.7926354550236717,
+      "grad_norm": 1.6323491334915161,
+      "learning_rate": 2.172133270268967e-05,
+      "loss": 1.7997,
+      "step": 7534
+    },
+    {
+      "epoch": 0.7927406628090479,
+      "grad_norm": 1.3499654531478882,
+      "learning_rate": 2.1700131237287414e-05,
+      "loss": 1.4139,
+      "step": 7535
+    },
+    {
+      "epoch": 0.792845870594424,
+      "grad_norm": 1.7880961894989014,
+      "learning_rate": 2.1678938864591046e-05,
+      "loss": 1.2011,
+      "step": 7536
+    },
+    {
+      "epoch": 0.7929510783798001,
+      "grad_norm": 1.5255318880081177,
+      "learning_rate": 2.1657755587061644e-05,
+      "loss": 1.9614,
+      "step": 7537
+    },
+    {
+      "epoch": 0.7930562861651762,
+      "grad_norm": 1.1118429899215698,
+      "learning_rate": 2.1636581407159105e-05,
+      "loss": 1.2734,
+      "step": 7538
+    },
+    {
+      "epoch": 0.7931614939505524,
+      "grad_norm": 1.6367865800857544,
+      "learning_rate": 2.1615416327342296e-05,
+      "loss": 2.0369,
+      "step": 7539
+    },
+    {
+      "epoch": 0.7932667017359285,
+      "grad_norm": 1.7008229494094849,
+      "learning_rate": 2.1594260350069096e-05,
+      "loss": 1.5876,
+      "step": 7540
+    },
+    {
+      "epoch": 0.7933719095213045,
+      "grad_norm": 1.5725849866867065,
+      "learning_rate": 2.157311347779626e-05,
+      "loss": 1.5209,
+      "step": 7541
+    },
+    {
+      "epoch": 0.7934771173066807,
+      "grad_norm": 1.7557533979415894,
+      "learning_rate": 2.1551975712979478e-05,
+      "loss": 1.2324,
+      "step": 7542
+    },
+    {
+      "epoch": 0.7935823250920568,
+      "grad_norm": 1.1576135158538818,
+      "learning_rate": 2.1530847058073466e-05,
+      "loss": 2.2651,
+      "step": 7543
+    },
+    {
+      "epoch": 0.7936875328774329,
+      "grad_norm": 1.6503269672393799,
+      "learning_rate": 2.1509727515531786e-05,
+      "loss": 1.9303,
+      "step": 7544
+    },
+    {
+      "epoch": 0.793792740662809,
+      "grad_norm": 2.0574920177459717,
+      "learning_rate": 2.1488617087806982e-05,
+      "loss": 1.6262,
+      "step": 7545
+    },
+    {
+      "epoch": 0.7938979484481852,
+      "grad_norm": 1.3746944665908813,
+      "learning_rate": 2.1467515777350544e-05,
+      "loss": 1.2937,
+      "step": 7546
+    },
+    {
+      "epoch": 0.7940031562335613,
+      "grad_norm": 1.4370603561401367,
+      "learning_rate": 2.1446423586612886e-05,
+      "loss": 1.3888,
+      "step": 7547
+    },
+    {
+      "epoch": 0.7941083640189374,
+      "grad_norm": 1.5636769533157349,
+      "learning_rate": 2.142534051804339e-05,
+      "loss": 1.312,
+      "step": 7548
+    },
+    {
+      "epoch": 0.7942135718043135,
+      "grad_norm": 1.7676070928573608,
+      "learning_rate": 2.140426657409038e-05,
+      "loss": 1.97,
+      "step": 7549
+    },
+    {
+      "epoch": 0.7943187795896897,
+      "grad_norm": 1.89109468460083,
+      "learning_rate": 2.1383201757201042e-05,
+      "loss": 1.271,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7944239873750658,
+      "grad_norm": 0.966954231262207,
+      "learning_rate": 2.1362146069821643e-05,
+      "loss": 1.6096,
+      "step": 7551
+    },
+    {
+      "epoch": 0.7945291951604418,
+      "grad_norm": 1.17153799533844,
+      "learning_rate": 2.1341099514397266e-05,
+      "loss": 1.5011,
+      "step": 7552
+    },
+    {
+      "epoch": 0.794634402945818,
+      "grad_norm": 1.4526724815368652,
+      "learning_rate": 2.132006209337195e-05,
+      "loss": 1.4198,
+      "step": 7553
+    },
+    {
+      "epoch": 0.7947396107311941,
+      "grad_norm": 2.1010799407958984,
+      "learning_rate": 2.1299033809188773e-05,
+      "loss": 1.6424,
+      "step": 7554
+    },
+    {
+      "epoch": 0.7948448185165702,
+      "grad_norm": 1.633360505104065,
+      "learning_rate": 2.1278014664289648e-05,
+      "loss": 1.9107,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7949500263019463,
+      "grad_norm": 1.9874669313430786,
+      "learning_rate": 2.125700466111542e-05,
+      "loss": 1.8695,
+      "step": 7556
+    },
+    {
+      "epoch": 0.7950552340873225,
+      "grad_norm": 1.591950535774231,
+      "learning_rate": 2.1236003802105974e-05,
+      "loss": 1.7396,
+      "step": 7557
+    },
+    {
+      "epoch": 0.7951604418726986,
+      "grad_norm": 1.1268935203552246,
+      "learning_rate": 2.121501208970005e-05,
+      "loss": 1.3255,
+      "step": 7558
+    },
+    {
+      "epoch": 0.7952656496580747,
+      "grad_norm": 1.344058871269226,
+      "learning_rate": 2.1194029526335303e-05,
+      "loss": 1.7558,
+      "step": 7559
+    },
+    {
+      "epoch": 0.7953708574434508,
+      "grad_norm": 1.8388925790786743,
+      "learning_rate": 2.117305611444843e-05,
+      "loss": 1.4927,
+      "step": 7560
+    },
+    {
+      "epoch": 0.795476065228827,
+      "grad_norm": 0.9249711632728577,
+      "learning_rate": 2.115209185647499e-05,
+      "loss": 1.4039,
+      "step": 7561
+    },
+    {
+      "epoch": 0.795581273014203,
+      "grad_norm": 1.0573887825012207,
+      "learning_rate": 2.1131136754849447e-05,
+      "loss": 1.5272,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7956864807995792,
+      "grad_norm": 1.5813140869140625,
+      "learning_rate": 2.1110190812005315e-05,
+      "loss": 1.2803,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7957916885849553,
+      "grad_norm": 2.2133073806762695,
+      "learning_rate": 2.1089254030374916e-05,
+      "loss": 2.0426,
+      "step": 7564
+    },
+    {
+      "epoch": 0.7958968963703315,
+      "grad_norm": 1.3776949644088745,
+      "learning_rate": 2.106832641238966e-05,
+      "loss": 1.5064,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7960021041557075,
+      "grad_norm": 2.1420438289642334,
+      "learning_rate": 2.1047407960479702e-05,
+      "loss": 1.888,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7961073119410836,
+      "grad_norm": 0.984514594078064,
+      "learning_rate": 2.102649867707426e-05,
+      "loss": 1.5134,
+      "step": 7567
+    },
+    {
+      "epoch": 0.7962125197264598,
+      "grad_norm": 1.1638802289962769,
+      "learning_rate": 2.1005598564601492e-05,
+      "loss": 1.907,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7963177275118358,
+      "grad_norm": 1.849143385887146,
+      "learning_rate": 2.0984707625488442e-05,
+      "loss": 2.197,
+      "step": 7569
+    },
+    {
+      "epoch": 0.796422935297212,
+      "grad_norm": 1.8121823072433472,
+      "learning_rate": 2.096382586216108e-05,
+      "loss": 1.6732,
+      "step": 7570
+    },
+    {
+      "epoch": 0.7965281430825881,
+      "grad_norm": 1.8981144428253174,
+      "learning_rate": 2.0942953277044386e-05,
+      "loss": 1.9267,
+      "step": 7571
+    },
+    {
+      "epoch": 0.7966333508679643,
+      "grad_norm": 1.1137375831604004,
+      "learning_rate": 2.092208987256217e-05,
+      "loss": 1.5532,
+      "step": 7572
+    },
+    {
+      "epoch": 0.7967385586533403,
+      "grad_norm": 1.7999489307403564,
+      "learning_rate": 2.0901235651137284e-05,
+      "loss": 2.235,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7968437664387165,
+      "grad_norm": 1.321373462677002,
+      "learning_rate": 2.0880390615191448e-05,
+      "loss": 1.6331,
+      "step": 7574
+    },
+    {
+      "epoch": 0.7969489742240926,
+      "grad_norm": 2.581979274749756,
+      "learning_rate": 2.0859554767145272e-05,
+      "loss": 1.2734,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7970541820094686,
+      "grad_norm": 1.2444729804992676,
+      "learning_rate": 2.0838728109418436e-05,
+      "loss": 1.7502,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7971593897948448,
+      "grad_norm": 1.8155887126922607,
+      "learning_rate": 2.081791064442943e-05,
+      "loss": 1.1768,
+      "step": 7577
+    },
+    {
+      "epoch": 0.797264597580221,
+      "grad_norm": 1.7820537090301514,
+      "learning_rate": 2.079710237459569e-05,
+      "loss": 1.7911,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7973698053655971,
+      "grad_norm": 1.5820388793945312,
+      "learning_rate": 2.0776303302333677e-05,
+      "loss": 1.8153,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7974750131509731,
+      "grad_norm": 2.486478090286255,
+      "learning_rate": 2.0755513430058672e-05,
+      "loss": 0.9953,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7975802209363493,
+      "grad_norm": 1.5562423467636108,
+      "learning_rate": 2.073473276018493e-05,
+      "loss": 1.8432,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7976854287217254,
+      "grad_norm": 1.674886703491211,
+      "learning_rate": 2.0713961295125685e-05,
+      "loss": 0.5957,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7977906365071016,
+      "grad_norm": 1.56822669506073,
+      "learning_rate": 2.0693199037293022e-05,
+      "loss": 1.8226,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7978958442924776,
+      "grad_norm": 1.492409110069275,
+      "learning_rate": 2.067244598909799e-05,
+      "loss": 1.4119,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7980010520778538,
+      "grad_norm": 1.627830147743225,
+      "learning_rate": 2.0651702152950602e-05,
+      "loss": 1.837,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7981062598632299,
+      "grad_norm": 1.6998107433319092,
+      "learning_rate": 2.0630967531259758e-05,
+      "loss": 1.6075,
+      "step": 7586
+    },
+    {
+      "epoch": 0.798211467648606,
+      "grad_norm": 1.3308358192443848,
+      "learning_rate": 2.0610242126433297e-05,
+      "loss": 1.7592,
+      "step": 7587
+    },
+    {
+      "epoch": 0.7983166754339821,
+      "grad_norm": 1.7519946098327637,
+      "learning_rate": 2.0589525940877996e-05,
+      "loss": 1.5667,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7984218832193583,
+      "grad_norm": 2.321702003479004,
+      "learning_rate": 2.0568818976999526e-05,
+      "loss": 1.3998,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7985270910047344,
+      "grad_norm": 1.7129671573638916,
+      "learning_rate": 2.0548121237202576e-05,
+      "loss": 1.7467,
+      "step": 7590
+    },
+    {
+      "epoch": 0.7986322987901104,
+      "grad_norm": 1.7081111669540405,
+      "learning_rate": 2.0527432723890684e-05,
+      "loss": 1.6383,
+      "step": 7591
+    },
+    {
+      "epoch": 0.7987375065754866,
+      "grad_norm": 1.5592128038406372,
+      "learning_rate": 2.0506753439466297e-05,
+      "loss": 1.7673,
+      "step": 7592
+    },
+    {
+      "epoch": 0.7988427143608627,
+      "grad_norm": 2.919283390045166,
+      "learning_rate": 2.04860833863309e-05,
+      "loss": 1.7805,
+      "step": 7593
+    },
+    {
+      "epoch": 0.7989479221462388,
+      "grad_norm": 2.3130650520324707,
+      "learning_rate": 2.0465422566884805e-05,
+      "loss": 1.2299,
+      "step": 7594
+    },
+    {
+      "epoch": 0.7990531299316149,
+      "grad_norm": 1.2341796159744263,
+      "learning_rate": 2.044477098352726e-05,
+      "loss": 1.4645,
+      "step": 7595
+    },
+    {
+      "epoch": 0.7991583377169911,
+      "grad_norm": 2.093064069747925,
+      "learning_rate": 2.0424128638656513e-05,
+      "loss": 1.6587,
+      "step": 7596
+    },
+    {
+      "epoch": 0.7992635455023672,
+      "grad_norm": 2.029285430908203,
+      "learning_rate": 2.040349553466967e-05,
+      "loss": 1.9854,
+      "step": 7597
+    },
+    {
+      "epoch": 0.7993687532877433,
+      "grad_norm": 1.8652676343917847,
+      "learning_rate": 2.0382871673962766e-05,
+      "loss": 1.6049,
+      "step": 7598
+    },
+    {
+      "epoch": 0.7994739610731194,
+      "grad_norm": 1.369429588317871,
+      "learning_rate": 2.0362257058930822e-05,
+      "loss": 1.6057,
+      "step": 7599
+    },
+    {
+      "epoch": 0.7995791688584956,
+      "grad_norm": 1.3833380937576294,
+      "learning_rate": 2.0341651691967735e-05,
+      "loss": 2.1065,
+      "step": 7600
+    },
+    {
+      "epoch": 0.7996843766438716,
+      "grad_norm": 2.4091103076934814,
+      "learning_rate": 2.0321055575466284e-05,
+      "loss": 2.0793,
+      "step": 7601
+    },
+    {
+      "epoch": 0.7997895844292477,
+      "grad_norm": 1.5020534992218018,
+      "learning_rate": 2.0300468711818322e-05,
+      "loss": 1.3944,
+      "step": 7602
+    },
+    {
+      "epoch": 0.7998947922146239,
+      "grad_norm": 1.9810582399368286,
+      "learning_rate": 2.027989110341446e-05,
+      "loss": 2.023,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.6195448637008667,
+      "learning_rate": 2.0259322752644327e-05,
+      "loss": 1.1899,
+      "step": 7604
+    },
+    {
+      "epoch": 0.8001052077853761,
+      "grad_norm": 1.4536365270614624,
+      "learning_rate": 2.0238763661896477e-05,
+      "loss": 1.6324,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8002104155707522,
+      "grad_norm": 1.5313684940338135,
+      "learning_rate": 2.0218213833558352e-05,
+      "loss": 1.6158,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8003156233561284,
+      "grad_norm": 1.2193766832351685,
+      "learning_rate": 2.0197673270016327e-05,
+      "loss": 1.7051,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8004208311415044,
+      "grad_norm": 1.6945223808288574,
+      "learning_rate": 2.0177141973655766e-05,
+      "loss": 1.9545,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8005260389268806,
+      "grad_norm": 1.5775741338729858,
+      "learning_rate": 2.01566199468608e-05,
+      "loss": 1.5485,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8006312467122567,
+      "grad_norm": 2.2319235801696777,
+      "learning_rate": 2.0136107192014676e-05,
+      "loss": 1.1478,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8007364544976329,
+      "grad_norm": 2.3744618892669678,
+      "learning_rate": 2.011560371149943e-05,
+      "loss": 1.3589,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8008416622830089,
+      "grad_norm": 1.1278767585754395,
+      "learning_rate": 2.0095109507696053e-05,
+      "loss": 1.7969,
+      "step": 7612
+    },
+    {
+      "epoch": 0.800946870068385,
+      "grad_norm": 2.358863353729248,
+      "learning_rate": 2.0074624582984512e-05,
+      "loss": 1.5369,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8010520778537612,
+      "grad_norm": 1.1637340784072876,
+      "learning_rate": 2.0054148939743634e-05,
+      "loss": 1.6914,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8011572856391374,
+      "grad_norm": 1.5420373678207397,
+      "learning_rate": 2.0033682580351144e-05,
+      "loss": 1.3169,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8012624934245134,
+      "grad_norm": 2.321443557739258,
+      "learning_rate": 2.001322550718382e-05,
+      "loss": 1.8365,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8013677012098895,
+      "grad_norm": 1.631739616394043,
+      "learning_rate": 1.9992777722617207e-05,
+      "loss": 1.0929,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8014729089952657,
+      "grad_norm": 1.5631780624389648,
+      "learning_rate": 1.997233922902585e-05,
+      "loss": 1.6305,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8015781167806417,
+      "grad_norm": 1.8600497245788574,
+      "learning_rate": 1.995191002878323e-05,
+      "loss": 1.323,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8016833245660179,
+      "grad_norm": 1.6327234506607056,
+      "learning_rate": 1.9931490124261688e-05,
+      "loss": 1.7434,
+      "step": 7620
+    },
+    {
+      "epoch": 0.801788532351394,
+      "grad_norm": 1.5391840934753418,
+      "learning_rate": 1.9911079517832555e-05,
+      "loss": 1.564,
+      "step": 7621
+    },
+    {
+      "epoch": 0.8018937401367702,
+      "grad_norm": 1.508278727531433,
+      "learning_rate": 1.9890678211866033e-05,
+      "loss": 1.7802,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8019989479221462,
+      "grad_norm": 1.6780965328216553,
+      "learning_rate": 1.9870286208731236e-05,
+      "loss": 1.6782,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8021041557075224,
+      "grad_norm": 1.9837194681167603,
+      "learning_rate": 1.9849903510796262e-05,
+      "loss": 1.9724,
+      "step": 7624
+    },
+    {
+      "epoch": 0.8022093634928985,
+      "grad_norm": 1.8240439891815186,
+      "learning_rate": 1.9829530120428064e-05,
+      "loss": 1.7938,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8023145712782745,
+      "grad_norm": 1.649254560470581,
+      "learning_rate": 1.9809166039992522e-05,
+      "loss": 1.2282,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8024197790636507,
+      "grad_norm": 2.882291793823242,
+      "learning_rate": 1.978881127185448e-05,
+      "loss": 1.2825,
+      "step": 7627
+    },
+    {
+      "epoch": 0.8025249868490268,
+      "grad_norm": 1.6689751148223877,
+      "learning_rate": 1.9768465818377656e-05,
+      "loss": 1.7849,
+      "step": 7628
+    },
+    {
+      "epoch": 0.802630194634403,
+      "grad_norm": 1.3756548166275024,
+      "learning_rate": 1.9748129681924675e-05,
+      "loss": 1.9856,
+      "step": 7629
+    },
+    {
+      "epoch": 0.802735402419779,
+      "grad_norm": 1.1730574369430542,
+      "learning_rate": 1.9727802864857194e-05,
+      "loss": 1.7476,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8028406102051552,
+      "grad_norm": 1.5033677816390991,
+      "learning_rate": 1.970748536953557e-05,
+      "loss": 1.54,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8029458179905313,
+      "grad_norm": 2.147925615310669,
+      "learning_rate": 1.9687177198319308e-05,
+      "loss": 1.521,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8030510257759074,
+      "grad_norm": 1.849800944328308,
+      "learning_rate": 1.9666878353566697e-05,
+      "loss": 1.9367,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8031562335612835,
+      "grad_norm": 1.3101515769958496,
+      "learning_rate": 1.9646588837634937e-05,
+      "loss": 1.9278,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8032614413466597,
+      "grad_norm": 2.0604212284088135,
+      "learning_rate": 1.9626308652880243e-05,
+      "loss": 1.7948,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8033666491320358,
+      "grad_norm": 1.3913450241088867,
+      "learning_rate": 1.9606037801657673e-05,
+      "loss": 1.5491,
+      "step": 7636
+    },
+    {
+      "epoch": 0.8034718569174119,
+      "grad_norm": 1.931066870689392,
+      "learning_rate": 1.9585776286321167e-05,
+      "loss": 1.4223,
+      "step": 7637
+    },
+    {
+      "epoch": 0.803577064702788,
+      "grad_norm": 2.459156036376953,
+      "learning_rate": 1.956552410922369e-05,
+      "loss": 1.1812,
+      "step": 7638
+    },
+    {
+      "epoch": 0.8036822724881642,
+      "grad_norm": 2.2327475547790527,
+      "learning_rate": 1.9545281272717032e-05,
+      "loss": 1.8111,
+      "step": 7639
+    },
+    {
+      "epoch": 0.8037874802735402,
+      "grad_norm": 1.7723302841186523,
+      "learning_rate": 1.9525047779151905e-05,
+      "loss": 2.2684,
+      "step": 7640
+    },
+    {
+      "epoch": 0.8038926880589163,
+      "grad_norm": 1.0957763195037842,
+      "learning_rate": 1.950482363087801e-05,
+      "loss": 1.3338,
+      "step": 7641
+    },
+    {
+      "epoch": 0.8039978958442925,
+      "grad_norm": 1.4228318929672241,
+      "learning_rate": 1.948460883024388e-05,
+      "loss": 1.2643,
+      "step": 7642
+    },
+    {
+      "epoch": 0.8041031036296686,
+      "grad_norm": 2.340993642807007,
+      "learning_rate": 1.9464403379596963e-05,
+      "loss": 2.2087,
+      "step": 7643
+    },
+    {
+      "epoch": 0.8042083114150447,
+      "grad_norm": 1.3735971450805664,
+      "learning_rate": 1.9444207281283723e-05,
+      "loss": 1.388,
+      "step": 7644
+    },
+    {
+      "epoch": 0.8043135192004208,
+      "grad_norm": 1.704942226409912,
+      "learning_rate": 1.9424020537649414e-05,
+      "loss": 1.6904,
+      "step": 7645
+    },
+    {
+      "epoch": 0.804418726985797,
+      "grad_norm": 1.3440220355987549,
+      "learning_rate": 1.940384315103825e-05,
+      "loss": 2.1261,
+      "step": 7646
+    },
+    {
+      "epoch": 0.8045239347711731,
+      "grad_norm": 1.9415428638458252,
+      "learning_rate": 1.938367512379341e-05,
+      "loss": 2.098,
+      "step": 7647
+    },
+    {
+      "epoch": 0.8046291425565492,
+      "grad_norm": 1.430979609489441,
+      "learning_rate": 1.9363516458256916e-05,
+      "loss": 1.1349,
+      "step": 7648
+    },
+    {
+      "epoch": 0.8047343503419253,
+      "grad_norm": 1.5660107135772705,
+      "learning_rate": 1.93433671567697e-05,
+      "loss": 2.0174,
+      "step": 7649
+    },
+    {
+      "epoch": 0.8048395581273015,
+      "grad_norm": 1.210718035697937,
+      "learning_rate": 1.932322722167168e-05,
+      "loss": 1.4806,
+      "step": 7650
+    },
+    {
+      "epoch": 0.8049447659126775,
+      "grad_norm": 1.4528659582138062,
+      "learning_rate": 1.9303096655301633e-05,
+      "loss": 1.9193,
+      "step": 7651
+    },
+    {
+      "epoch": 0.8050499736980536,
+      "grad_norm": 1.8522127866744995,
+      "learning_rate": 1.9282975459997234e-05,
+      "loss": 1.4749,
+      "step": 7652
+    },
+    {
+      "epoch": 0.8051551814834298,
+      "grad_norm": 1.7069746255874634,
+      "learning_rate": 1.9262863638095097e-05,
+      "loss": 1.7067,
+      "step": 7653
+    },
+    {
+      "epoch": 0.8052603892688059,
+      "grad_norm": 1.1644920110702515,
+      "learning_rate": 1.9242761191930725e-05,
+      "loss": 1.9264,
+      "step": 7654
+    },
+    {
+      "epoch": 0.805365597054182,
+      "grad_norm": 1.9184256792068481,
+      "learning_rate": 1.9222668123838593e-05,
+      "loss": 2.0388,
+      "step": 7655
+    },
+    {
+      "epoch": 0.8054708048395581,
+      "grad_norm": 2.2873263359069824,
+      "learning_rate": 1.9202584436152014e-05,
+      "loss": 1.719,
+      "step": 7656
+    },
+    {
+      "epoch": 0.8055760126249343,
+      "grad_norm": 1.712296724319458,
+      "learning_rate": 1.9182510131203224e-05,
+      "loss": 1.6544,
+      "step": 7657
+    },
+    {
+      "epoch": 0.8056812204103103,
+      "grad_norm": 2.636054515838623,
+      "learning_rate": 1.9162445211323432e-05,
+      "loss": 1.9039,
+      "step": 7658
+    },
+    {
+      "epoch": 0.8057864281956865,
+      "grad_norm": 1.4774726629257202,
+      "learning_rate": 1.914238967884269e-05,
+      "loss": 1.6511,
+      "step": 7659
+    },
+    {
+      "epoch": 0.8058916359810626,
+      "grad_norm": 1.71102774143219,
+      "learning_rate": 1.9122343536089937e-05,
+      "loss": 1.2127,
+      "step": 7660
+    },
+    {
+      "epoch": 0.8059968437664388,
+      "grad_norm": 1.1164214611053467,
+      "learning_rate": 1.910230678539314e-05,
+      "loss": 1.5642,
+      "step": 7661
+    },
+    {
+      "epoch": 0.8061020515518148,
+      "grad_norm": 1.0777497291564941,
+      "learning_rate": 1.9082279429079065e-05,
+      "loss": 1.6877,
+      "step": 7662
+    },
+    {
+      "epoch": 0.806207259337191,
+      "grad_norm": 1.3562448024749756,
+      "learning_rate": 1.9062261469473397e-05,
+      "loss": 1.9616,
+      "step": 7663
+    },
+    {
+      "epoch": 0.8063124671225671,
+      "grad_norm": 1.527605652809143,
+      "learning_rate": 1.9042252908900814e-05,
+      "loss": 1.4711,
+      "step": 7664
+    },
+    {
+      "epoch": 0.8064176749079431,
+      "grad_norm": 1.3711210489273071,
+      "learning_rate": 1.9022253749684783e-05,
+      "loss": 1.6572,
+      "step": 7665
+    },
+    {
+      "epoch": 0.8065228826933193,
+      "grad_norm": 2.562347888946533,
+      "learning_rate": 1.9002263994147796e-05,
+      "loss": 1.9137,
+      "step": 7666
+    },
+    {
+      "epoch": 0.8066280904786954,
+      "grad_norm": 1.8726879358291626,
+      "learning_rate": 1.8982283644611175e-05,
+      "loss": 1.1054,
+      "step": 7667
+    },
+    {
+      "epoch": 0.8067332982640716,
+      "grad_norm": 2.17808198928833,
+      "learning_rate": 1.8962312703395146e-05,
+      "loss": 1.6859,
+      "step": 7668
+    },
+    {
+      "epoch": 0.8068385060494476,
+      "grad_norm": 1.516634225845337,
+      "learning_rate": 1.8942351172818905e-05,
+      "loss": 1.5421,
+      "step": 7669
+    },
+    {
+      "epoch": 0.8069437138348238,
+      "grad_norm": 1.740494966506958,
+      "learning_rate": 1.8922399055200525e-05,
+      "loss": 1.5834,
+      "step": 7670
+    },
+    {
+      "epoch": 0.8070489216201999,
+      "grad_norm": 1.4491758346557617,
+      "learning_rate": 1.8902456352856925e-05,
+      "loss": 1.4972,
+      "step": 7671
+    },
+    {
+      "epoch": 0.807154129405576,
+      "grad_norm": 1.3840901851654053,
+      "learning_rate": 1.888252306810406e-05,
+      "loss": 1.7152,
+      "step": 7672
+    },
+    {
+      "epoch": 0.8072593371909521,
+      "grad_norm": 1.5224378108978271,
+      "learning_rate": 1.886259920325667e-05,
+      "loss": 1.2755,
+      "step": 7673
+    },
+    {
+      "epoch": 0.8073645449763283,
+      "grad_norm": 1.3467313051223755,
+      "learning_rate": 1.8842684760628425e-05,
+      "loss": 1.9975,
+      "step": 7674
+    },
+    {
+      "epoch": 0.8074697527617044,
+      "grad_norm": 1.9911725521087646,
+      "learning_rate": 1.882277974253197e-05,
+      "loss": 2.0025,
+      "step": 7675
+    },
+    {
+      "epoch": 0.8075749605470804,
+      "grad_norm": 1.4871760606765747,
+      "learning_rate": 1.8802884151278775e-05,
+      "loss": 1.4795,
+      "step": 7676
+    },
+    {
+      "epoch": 0.8076801683324566,
+      "grad_norm": 1.2734004259109497,
+      "learning_rate": 1.8782997989179297e-05,
+      "loss": 1.6471,
+      "step": 7677
+    },
+    {
+      "epoch": 0.8077853761178327,
+      "grad_norm": 1.3544495105743408,
+      "learning_rate": 1.8763121258542815e-05,
+      "loss": 1.744,
+      "step": 7678
+    },
+    {
+      "epoch": 0.8078905839032089,
+      "grad_norm": 1.0602861642837524,
+      "learning_rate": 1.874325396167753e-05,
+      "loss": 2.062,
+      "step": 7679
+    },
+    {
+      "epoch": 0.8079957916885849,
+      "grad_norm": 1.5346481800079346,
+      "learning_rate": 1.8723396100890623e-05,
+      "loss": 0.9999,
+      "step": 7680
+    },
+    {
+      "epoch": 0.8081009994739611,
+      "grad_norm": 1.847928762435913,
+      "learning_rate": 1.870354767848809e-05,
+      "loss": 1.5345,
+      "step": 7681
+    },
+    {
+      "epoch": 0.8082062072593372,
+      "grad_norm": 1.852360725402832,
+      "learning_rate": 1.8683708696774826e-05,
+      "loss": 1.8659,
+      "step": 7682
+    },
+    {
+      "epoch": 0.8083114150447133,
+      "grad_norm": 1.2460297346115112,
+      "learning_rate": 1.866387915805473e-05,
+      "loss": 1.8196,
+      "step": 7683
+    },
+    {
+      "epoch": 0.8084166228300894,
+      "grad_norm": 1.7130165100097656,
+      "learning_rate": 1.8644059064630516e-05,
+      "loss": 1.652,
+      "step": 7684
+    },
+    {
+      "epoch": 0.8085218306154656,
+      "grad_norm": 1.4033414125442505,
+      "learning_rate": 1.86242484188038e-05,
+      "loss": 1.5075,
+      "step": 7685
+    },
+    {
+      "epoch": 0.8086270384008417,
+      "grad_norm": 0.9263932108879089,
+      "learning_rate": 1.8604447222875178e-05,
+      "loss": 1.8016,
+      "step": 7686
+    },
+    {
+      "epoch": 0.8087322461862178,
+      "grad_norm": 1.5973109006881714,
+      "learning_rate": 1.8584655479144063e-05,
+      "loss": 1.5324,
+      "step": 7687
+    },
+    {
+      "epoch": 0.8088374539715939,
+      "grad_norm": 1.074812889099121,
+      "learning_rate": 1.8564873189908793e-05,
+      "loss": 1.6024,
+      "step": 7688
+    },
+    {
+      "epoch": 0.80894266175697,
+      "grad_norm": 1.3290114402770996,
+      "learning_rate": 1.854510035746667e-05,
+      "loss": 1.786,
+      "step": 7689
+    },
+    {
+      "epoch": 0.8090478695423461,
+      "grad_norm": 2.054743766784668,
+      "learning_rate": 1.8525336984113807e-05,
+      "loss": 1.2776,
+      "step": 7690
+    },
+    {
+      "epoch": 0.8091530773277222,
+      "grad_norm": 1.4225199222564697,
+      "learning_rate": 1.850558307214525e-05,
+      "loss": 1.4513,
+      "step": 7691
+    },
+    {
+      "epoch": 0.8092582851130984,
+      "grad_norm": 1.4592626094818115,
+      "learning_rate": 1.848583862385501e-05,
+      "loss": 1.4772,
+      "step": 7692
+    },
+    {
+      "epoch": 0.8093634928984745,
+      "grad_norm": 1.3948107957839966,
+      "learning_rate": 1.8466103641535904e-05,
+      "loss": 1.6123,
+      "step": 7693
+    },
+    {
+      "epoch": 0.8094687006838506,
+      "grad_norm": 1.5830365419387817,
+      "learning_rate": 1.8446378127479692e-05,
+      "loss": 1.663,
+      "step": 7694
+    },
+    {
+      "epoch": 0.8095739084692267,
+      "grad_norm": 1.350455403327942,
+      "learning_rate": 1.8426662083977042e-05,
+      "loss": 1.6457,
+      "step": 7695
+    },
+    {
+      "epoch": 0.8096791162546029,
+      "grad_norm": 1.3711130619049072,
+      "learning_rate": 1.8406955513317482e-05,
+      "loss": 1.8406,
+      "step": 7696
+    },
+    {
+      "epoch": 0.8097843240399789,
+      "grad_norm": 2.136914014816284,
+      "learning_rate": 1.8387258417789532e-05,
+      "loss": 1.7469,
+      "step": 7697
+    },
+    {
+      "epoch": 0.8098895318253551,
+      "grad_norm": 2.448425531387329,
+      "learning_rate": 1.836757079968051e-05,
+      "loss": 1.3565,
+      "step": 7698
+    },
+    {
+      "epoch": 0.8099947396107312,
+      "grad_norm": 1.5618599653244019,
+      "learning_rate": 1.8347892661276656e-05,
+      "loss": 1.587,
+      "step": 7699
+    },
+    {
+      "epoch": 0.8100999473961074,
+      "grad_norm": 1.5216078758239746,
+      "learning_rate": 1.8328224004863183e-05,
+      "loss": 2.1681,
+      "step": 7700
+    },
+    {
+      "epoch": 0.8102051551814834,
+      "grad_norm": 1.2957028150558472,
+      "learning_rate": 1.830856483272412e-05,
+      "loss": 1.0984,
+      "step": 7701
+    },
+    {
+      "epoch": 0.8103103629668595,
+      "grad_norm": 1.9072808027267456,
+      "learning_rate": 1.8288915147142384e-05,
+      "loss": 1.4575,
+      "step": 7702
+    },
+    {
+      "epoch": 0.8104155707522357,
+      "grad_norm": 1.6128511428833008,
+      "learning_rate": 1.8269274950399895e-05,
+      "loss": 1.3623,
+      "step": 7703
+    },
+    {
+      "epoch": 0.8105207785376117,
+      "grad_norm": 1.4067853689193726,
+      "learning_rate": 1.8249644244777376e-05,
+      "loss": 1.5455,
+      "step": 7704
+    },
+    {
+      "epoch": 0.8106259863229879,
+      "grad_norm": 1.5415167808532715,
+      "learning_rate": 1.8230023032554444e-05,
+      "loss": 1.7658,
+      "step": 7705
+    },
+    {
+      "epoch": 0.810731194108364,
+      "grad_norm": 1.5726876258850098,
+      "learning_rate": 1.82104113160097e-05,
+      "loss": 1.7449,
+      "step": 7706
+    },
+    {
+      "epoch": 0.8108364018937402,
+      "grad_norm": 1.307512879371643,
+      "learning_rate": 1.819080909742057e-05,
+      "loss": 1.7211,
+      "step": 7707
+    },
+    {
+      "epoch": 0.8109416096791162,
+      "grad_norm": 1.4828598499298096,
+      "learning_rate": 1.8171216379063348e-05,
+      "loss": 1.6675,
+      "step": 7708
+    },
+    {
+      "epoch": 0.8110468174644924,
+      "grad_norm": 1.5087251663208008,
+      "learning_rate": 1.8151633163213357e-05,
+      "loss": 2.0347,
+      "step": 7709
+    },
+    {
+      "epoch": 0.8111520252498685,
+      "grad_norm": 1.217799186706543,
+      "learning_rate": 1.8132059452144666e-05,
+      "loss": 1.4303,
+      "step": 7710
+    },
+    {
+      "epoch": 0.8112572330352447,
+      "grad_norm": 1.011033058166504,
+      "learning_rate": 1.8112495248130312e-05,
+      "loss": 1.6298,
+      "step": 7711
+    },
+    {
+      "epoch": 0.8113624408206207,
+      "grad_norm": 1.5476887226104736,
+      "learning_rate": 1.809294055344226e-05,
+      "loss": 2.5314,
+      "step": 7712
+    },
+    {
+      "epoch": 0.8114676486059968,
+      "grad_norm": 1.962207317352295,
+      "learning_rate": 1.8073395370351287e-05,
+      "loss": 1.6565,
+      "step": 7713
+    },
+    {
+      "epoch": 0.811572856391373,
+      "grad_norm": 1.5745935440063477,
+      "learning_rate": 1.8053859701127153e-05,
+      "loss": 1.4709,
+      "step": 7714
+    },
+    {
+      "epoch": 0.811678064176749,
+      "grad_norm": 1.5660057067871094,
+      "learning_rate": 1.8034333548038474e-05,
+      "loss": 2.3473,
+      "step": 7715
+    },
+    {
+      "epoch": 0.8117832719621252,
+      "grad_norm": 1.5668786764144897,
+      "learning_rate": 1.8014816913352682e-05,
+      "loss": 1.6102,
+      "step": 7716
+    },
+    {
+      "epoch": 0.8118884797475013,
+      "grad_norm": 1.458255648612976,
+      "learning_rate": 1.7995309799336256e-05,
+      "loss": 1.5805,
+      "step": 7717
+    },
+    {
+      "epoch": 0.8119936875328775,
+      "grad_norm": 1.485870599746704,
+      "learning_rate": 1.7975812208254473e-05,
+      "loss": 1.7534,
+      "step": 7718
+    },
+    {
+      "epoch": 0.8120988953182535,
+      "grad_norm": 1.127456784248352,
+      "learning_rate": 1.7956324142371485e-05,
+      "loss": 1.7223,
+      "step": 7719
+    },
+    {
+      "epoch": 0.8122041031036297,
+      "grad_norm": 1.6504329442977905,
+      "learning_rate": 1.7936845603950447e-05,
+      "loss": 1.1134,
+      "step": 7720
+    },
+    {
+      "epoch": 0.8123093108890058,
+      "grad_norm": 2.026771068572998,
+      "learning_rate": 1.7917376595253264e-05,
+      "loss": 1.236,
+      "step": 7721
+    },
+    {
+      "epoch": 0.8124145186743819,
+      "grad_norm": 1.5272623300552368,
+      "learning_rate": 1.7897917118540875e-05,
+      "loss": 1.6421,
+      "step": 7722
+    },
+    {
+      "epoch": 0.812519726459758,
+      "grad_norm": 1.7590751647949219,
+      "learning_rate": 1.787846717607302e-05,
+      "loss": 1.2447,
+      "step": 7723
+    },
+    {
+      "epoch": 0.8126249342451342,
+      "grad_norm": 1.131596326828003,
+      "learning_rate": 1.7859026770108323e-05,
+      "loss": 1.3865,
+      "step": 7724
+    },
+    {
+      "epoch": 0.8127301420305103,
+      "grad_norm": 1.4612594842910767,
+      "learning_rate": 1.7839595902904393e-05,
+      "loss": 1.7474,
+      "step": 7725
+    },
+    {
+      "epoch": 0.8128353498158863,
+      "grad_norm": 1.4456712007522583,
+      "learning_rate": 1.782017457671764e-05,
+      "loss": 1.4284,
+      "step": 7726
+    },
+    {
+      "epoch": 0.8129405576012625,
+      "grad_norm": 2.546482801437378,
+      "learning_rate": 1.780076279380337e-05,
+      "loss": 1.9752,
+      "step": 7727
+    },
+    {
+      "epoch": 0.8130457653866386,
+      "grad_norm": 2.082400321960449,
+      "learning_rate": 1.778136055641587e-05,
+      "loss": 1.666,
+      "step": 7728
+    },
+    {
+      "epoch": 0.8131509731720147,
+      "grad_norm": 1.367645025253296,
+      "learning_rate": 1.776196786680824e-05,
+      "loss": 1.4718,
+      "step": 7729
+    },
+    {
+      "epoch": 0.8132561809573908,
+      "grad_norm": 2.418299436569214,
+      "learning_rate": 1.7742584727232437e-05,
+      "loss": 1.7791,
+      "step": 7730
+    },
+    {
+      "epoch": 0.813361388742767,
+      "grad_norm": 1.4075639247894287,
+      "learning_rate": 1.7723211139939445e-05,
+      "loss": 1.9541,
+      "step": 7731
+    },
+    {
+      "epoch": 0.8134665965281431,
+      "grad_norm": 1.8531979322433472,
+      "learning_rate": 1.7703847107178996e-05,
+      "loss": 1.3266,
+      "step": 7732
+    },
+    {
+      "epoch": 0.8135718043135192,
+      "grad_norm": 1.2879350185394287,
+      "learning_rate": 1.768449263119977e-05,
+      "loss": 1.3593,
+      "step": 7733
+    },
+    {
+      "epoch": 0.8136770120988953,
+      "grad_norm": 1.4324150085449219,
+      "learning_rate": 1.7665147714249376e-05,
+      "loss": 1.1906,
+      "step": 7734
+    },
+    {
+      "epoch": 0.8137822198842715,
+      "grad_norm": 1.2460278272628784,
+      "learning_rate": 1.7645812358574264e-05,
+      "loss": 1.7535,
+      "step": 7735
+    },
+    {
+      "epoch": 0.8138874276696475,
+      "grad_norm": 1.501090407371521,
+      "learning_rate": 1.762648656641974e-05,
+      "loss": 1.507,
+      "step": 7736
+    },
+    {
+      "epoch": 0.8139926354550236,
+      "grad_norm": 2.3010196685791016,
+      "learning_rate": 1.7607170340030143e-05,
+      "loss": 1.1804,
+      "step": 7737
+    },
+    {
+      "epoch": 0.8140978432403998,
+      "grad_norm": 1.1160740852355957,
+      "learning_rate": 1.7587863681648487e-05,
+      "loss": 1.7082,
+      "step": 7738
+    },
+    {
+      "epoch": 0.814203051025776,
+      "grad_norm": 1.3494609594345093,
+      "learning_rate": 1.7568566593516867e-05,
+      "loss": 1.5496,
+      "step": 7739
+    },
+    {
+      "epoch": 0.814308258811152,
+      "grad_norm": 1.8527060747146606,
+      "learning_rate": 1.754927907787617e-05,
+      "loss": 1.4893,
+      "step": 7740
+    },
+    {
+      "epoch": 0.8144134665965281,
+      "grad_norm": 2.120997428894043,
+      "learning_rate": 1.753000113696617e-05,
+      "loss": 1.9963,
+      "step": 7741
+    },
+    {
+      "epoch": 0.8145186743819043,
+      "grad_norm": 2.1121816635131836,
+      "learning_rate": 1.7510732773025584e-05,
+      "loss": 2.1124,
+      "step": 7742
+    },
+    {
+      "epoch": 0.8146238821672804,
+      "grad_norm": 2.398660659790039,
+      "learning_rate": 1.7491473988291984e-05,
+      "loss": 1.6849,
+      "step": 7743
+    },
+    {
+      "epoch": 0.8147290899526565,
+      "grad_norm": 1.6509634256362915,
+      "learning_rate": 1.7472224785001778e-05,
+      "loss": 1.4914,
+      "step": 7744
+    },
+    {
+      "epoch": 0.8148342977380326,
+      "grad_norm": 1.1640690565109253,
+      "learning_rate": 1.745298516539039e-05,
+      "loss": 1.7646,
+      "step": 7745
+    },
+    {
+      "epoch": 0.8149395055234088,
+      "grad_norm": 0.9993892312049866,
+      "learning_rate": 1.7433755131692e-05,
+      "loss": 1.5413,
+      "step": 7746
+    },
+    {
+      "epoch": 0.8150447133087848,
+      "grad_norm": 2.1200225353240967,
+      "learning_rate": 1.7414534686139717e-05,
+      "loss": 1.3748,
+      "step": 7747
+    },
+    {
+      "epoch": 0.815149921094161,
+      "grad_norm": 1.2785836458206177,
+      "learning_rate": 1.7395323830965605e-05,
+      "loss": 1.9115,
+      "step": 7748
+    },
+    {
+      "epoch": 0.8152551288795371,
+      "grad_norm": 1.8023689985275269,
+      "learning_rate": 1.7376122568400532e-05,
+      "loss": 1.432,
+      "step": 7749
+    },
+    {
+      "epoch": 0.8153603366649133,
+      "grad_norm": 1.6197617053985596,
+      "learning_rate": 1.7356930900674228e-05,
+      "loss": 1.2918,
+      "step": 7750
+    },
+    {
+      "epoch": 0.8154655444502893,
+      "grad_norm": 1.9692368507385254,
+      "learning_rate": 1.7337748830015442e-05,
+      "loss": 1.4366,
+      "step": 7751
+    },
+    {
+      "epoch": 0.8155707522356654,
+      "grad_norm": 1.7193232774734497,
+      "learning_rate": 1.7318576358651685e-05,
+      "loss": 1.9742,
+      "step": 7752
+    },
+    {
+      "epoch": 0.8156759600210416,
+      "grad_norm": 1.3902307748794556,
+      "learning_rate": 1.7299413488809356e-05,
+      "loss": 1.6264,
+      "step": 7753
+    },
+    {
+      "epoch": 0.8157811678064176,
+      "grad_norm": 1.7483313083648682,
+      "learning_rate": 1.728026022271384e-05,
+      "loss": 1.9915,
+      "step": 7754
+    },
+    {
+      "epoch": 0.8158863755917938,
+      "grad_norm": 1.3935078382492065,
+      "learning_rate": 1.726111656258932e-05,
+      "loss": 1.5677,
+      "step": 7755
+    },
+    {
+      "epoch": 0.8159915833771699,
+      "grad_norm": 1.7369499206542969,
+      "learning_rate": 1.724198251065885e-05,
+      "loss": 1.5543,
+      "step": 7756
+    },
+    {
+      "epoch": 0.8160967911625461,
+      "grad_norm": 1.9357125759124756,
+      "learning_rate": 1.7222858069144464e-05,
+      "loss": 1.8799,
+      "step": 7757
+    },
+    {
+      "epoch": 0.8162019989479221,
+      "grad_norm": 1.534250020980835,
+      "learning_rate": 1.720374324026699e-05,
+      "loss": 1.5193,
+      "step": 7758
+    },
+    {
+      "epoch": 0.8163072067332983,
+      "grad_norm": 1.6834253072738647,
+      "learning_rate": 1.718463802624617e-05,
+      "loss": 1.3028,
+      "step": 7759
+    },
+    {
+      "epoch": 0.8164124145186744,
+      "grad_norm": 1.60200834274292,
+      "learning_rate": 1.716554242930063e-05,
+      "loss": 1.5029,
+      "step": 7760
+    },
+    {
+      "epoch": 0.8165176223040504,
+      "grad_norm": 1.758754849433899,
+      "learning_rate": 1.7146456451647863e-05,
+      "loss": 1.7036,
+      "step": 7761
+    },
+    {
+      "epoch": 0.8166228300894266,
+      "grad_norm": 1.4392300844192505,
+      "learning_rate": 1.7127380095504296e-05,
+      "loss": 1.312,
+      "step": 7762
+    },
+    {
+      "epoch": 0.8167280378748027,
+      "grad_norm": 1.1349120140075684,
+      "learning_rate": 1.710831336308519e-05,
+      "loss": 1.618,
+      "step": 7763
+    },
+    {
+      "epoch": 0.8168332456601789,
+      "grad_norm": 2.290132999420166,
+      "learning_rate": 1.708925625660467e-05,
+      "loss": 1.3436,
+      "step": 7764
+    },
+    {
+      "epoch": 0.8169384534455549,
+      "grad_norm": 1.6770775318145752,
+      "learning_rate": 1.707020877827583e-05,
+      "loss": 1.3586,
+      "step": 7765
+    },
+    {
+      "epoch": 0.8170436612309311,
+      "grad_norm": 1.3523870706558228,
+      "learning_rate": 1.7051170930310555e-05,
+      "loss": 1.717,
+      "step": 7766
+    },
+    {
+      "epoch": 0.8171488690163072,
+      "grad_norm": 1.853025197982788,
+      "learning_rate": 1.703214271491964e-05,
+      "loss": 1.5571,
+      "step": 7767
+    },
+    {
+      "epoch": 0.8172540768016833,
+      "grad_norm": 1.83201003074646,
+      "learning_rate": 1.701312413431281e-05,
+      "loss": 1.9148,
+      "step": 7768
+    },
+    {
+      "epoch": 0.8173592845870594,
+      "grad_norm": 1.2569760084152222,
+      "learning_rate": 1.699411519069858e-05,
+      "loss": 2.3183,
+      "step": 7769
+    },
+    {
+      "epoch": 0.8174644923724356,
+      "grad_norm": 2.7303202152252197,
+      "learning_rate": 1.697511588628443e-05,
+      "loss": 1.6214,
+      "step": 7770
+    },
+    {
+      "epoch": 0.8175697001578117,
+      "grad_norm": 2.05470609664917,
+      "learning_rate": 1.6956126223276692e-05,
+      "loss": 1.3433,
+      "step": 7771
+    },
+    {
+      "epoch": 0.8176749079431878,
+      "grad_norm": 1.919611930847168,
+      "learning_rate": 1.693714620388054e-05,
+      "loss": 1.4729,
+      "step": 7772
+    },
+    {
+      "epoch": 0.8177801157285639,
+      "grad_norm": 1.4189374446868896,
+      "learning_rate": 1.6918175830300088e-05,
+      "loss": 1.756,
+      "step": 7773
+    },
+    {
+      "epoch": 0.81788532351394,
+      "grad_norm": 1.5107403993606567,
+      "learning_rate": 1.6899215104738307e-05,
+      "loss": 1.9558,
+      "step": 7774
+    },
+    {
+      "epoch": 0.8179905312993162,
+      "grad_norm": 1.0100902318954468,
+      "learning_rate": 1.6880264029396997e-05,
+      "loss": 1.6087,
+      "step": 7775
+    },
+    {
+      "epoch": 0.8180957390846922,
+      "grad_norm": 1.7361371517181396,
+      "learning_rate": 1.686132260647696e-05,
+      "loss": 1.3897,
+      "step": 7776
+    },
+    {
+      "epoch": 0.8182009468700684,
+      "grad_norm": 1.6154359579086304,
+      "learning_rate": 1.6842390838177746e-05,
+      "loss": 1.8557,
+      "step": 7777
+    },
+    {
+      "epoch": 0.8183061546554445,
+      "grad_norm": 1.9448010921478271,
+      "learning_rate": 1.682346872669782e-05,
+      "loss": 1.7147,
+      "step": 7778
+    },
+    {
+      "epoch": 0.8184113624408206,
+      "grad_norm": 1.653695821762085,
+      "learning_rate": 1.6804556274234616e-05,
+      "loss": 1.7196,
+      "step": 7779
+    },
+    {
+      "epoch": 0.8185165702261967,
+      "grad_norm": 1.5844122171401978,
+      "learning_rate": 1.678565348298433e-05,
+      "loss": 1.4943,
+      "step": 7780
+    },
+    {
+      "epoch": 0.8186217780115729,
+      "grad_norm": 2.1291873455047607,
+      "learning_rate": 1.6766760355142098e-05,
+      "loss": 1.0847,
+      "step": 7781
+    },
+    {
+      "epoch": 0.818726985796949,
+      "grad_norm": 1.2485053539276123,
+      "learning_rate": 1.6747876892901893e-05,
+      "loss": 1.5608,
+      "step": 7782
+    },
+    {
+      "epoch": 0.8188321935823251,
+      "grad_norm": 2.34297776222229,
+      "learning_rate": 1.6729003098456576e-05,
+      "loss": 1.5364,
+      "step": 7783
+    },
+    {
+      "epoch": 0.8189374013677012,
+      "grad_norm": 1.2006014585494995,
+      "learning_rate": 1.6710138973997957e-05,
+      "loss": 1.8377,
+      "step": 7784
+    },
+    {
+      "epoch": 0.8190426091530774,
+      "grad_norm": 2.0813822746276855,
+      "learning_rate": 1.6691284521716622e-05,
+      "loss": 2.0049,
+      "step": 7785
+    },
+    {
+      "epoch": 0.8191478169384534,
+      "grad_norm": 1.6687533855438232,
+      "learning_rate": 1.667243974380207e-05,
+      "loss": 1.6737,
+      "step": 7786
+    },
+    {
+      "epoch": 0.8192530247238295,
+      "grad_norm": 1.5384453535079956,
+      "learning_rate": 1.665360464244272e-05,
+      "loss": 1.6079,
+      "step": 7787
+    },
+    {
+      "epoch": 0.8193582325092057,
+      "grad_norm": 1.5753425359725952,
+      "learning_rate": 1.6634779219825814e-05,
+      "loss": 1.8783,
+      "step": 7788
+    },
+    {
+      "epoch": 0.8194634402945818,
+      "grad_norm": 1.79296875,
+      "learning_rate": 1.6615963478137454e-05,
+      "loss": 1.8779,
+      "step": 7789
+    },
+    {
+      "epoch": 0.8195686480799579,
+      "grad_norm": 2.0262389183044434,
+      "learning_rate": 1.6597157419562703e-05,
+      "loss": 1.7647,
+      "step": 7790
+    },
+    {
+      "epoch": 0.819673855865334,
+      "grad_norm": 1.363146185874939,
+      "learning_rate": 1.6578361046285418e-05,
+      "loss": 1.7111,
+      "step": 7791
+    },
+    {
+      "epoch": 0.8197790636507102,
+      "grad_norm": 2.63222599029541,
+      "learning_rate": 1.655957436048834e-05,
+      "loss": 1.0101,
+      "step": 7792
+    },
+    {
+      "epoch": 0.8198842714360862,
+      "grad_norm": 1.6749835014343262,
+      "learning_rate": 1.6540797364353155e-05,
+      "loss": 1.5263,
+      "step": 7793
+    },
+    {
+      "epoch": 0.8199894792214624,
+      "grad_norm": 1.5727925300598145,
+      "learning_rate": 1.6522030060060356e-05,
+      "loss": 1.6823,
+      "step": 7794
+    },
+    {
+      "epoch": 0.8200946870068385,
+      "grad_norm": 1.6061309576034546,
+      "learning_rate": 1.6503272449789286e-05,
+      "loss": 1.4122,
+      "step": 7795
+    },
+    {
+      "epoch": 0.8201998947922147,
+      "grad_norm": 2.4078192710876465,
+      "learning_rate": 1.648452453571826e-05,
+      "loss": 1.9501,
+      "step": 7796
+    },
+    {
+      "epoch": 0.8203051025775907,
+      "grad_norm": 1.4480916261672974,
+      "learning_rate": 1.646578632002439e-05,
+      "loss": 1.7646,
+      "step": 7797
+    },
+    {
+      "epoch": 0.8204103103629669,
+      "grad_norm": 2.407717227935791,
+      "learning_rate": 1.644705780488367e-05,
+      "loss": 1.4102,
+      "step": 7798
+    },
+    {
+      "epoch": 0.820515518148343,
+      "grad_norm": 1.7151941061019897,
+      "learning_rate": 1.6428338992471005e-05,
+      "loss": 1.57,
+      "step": 7799
+    },
+    {
+      "epoch": 0.820620725933719,
+      "grad_norm": 1.3426469564437866,
+      "learning_rate": 1.6409629884960144e-05,
+      "loss": 1.7375,
+      "step": 7800
+    },
+    {
+      "epoch": 0.8207259337190952,
+      "grad_norm": 1.4567279815673828,
+      "learning_rate": 1.6390930484523704e-05,
+      "loss": 1.5916,
+      "step": 7801
+    },
+    {
+      "epoch": 0.8208311415044713,
+      "grad_norm": 2.2961177825927734,
+      "learning_rate": 1.6372240793333195e-05,
+      "loss": 1.5445,
+      "step": 7802
+    },
+    {
+      "epoch": 0.8209363492898475,
+      "grad_norm": 1.0729035139083862,
+      "learning_rate": 1.6353560813558953e-05,
+      "loss": 1.5018,
+      "step": 7803
+    },
+    {
+      "epoch": 0.8210415570752235,
+      "grad_norm": 1.8320657014846802,
+      "learning_rate": 1.6334890547370286e-05,
+      "loss": 1.4641,
+      "step": 7804
+    },
+    {
+      "epoch": 0.8211467648605997,
+      "grad_norm": 1.700055480003357,
+      "learning_rate": 1.6316229996935262e-05,
+      "loss": 2.0822,
+      "step": 7805
+    },
+    {
+      "epoch": 0.8212519726459758,
+      "grad_norm": 1.5183660984039307,
+      "learning_rate": 1.6297579164420873e-05,
+      "loss": 1.8721,
+      "step": 7806
+    },
+    {
+      "epoch": 0.821357180431352,
+      "grad_norm": 1.26556396484375,
+      "learning_rate": 1.6278938051993008e-05,
+      "loss": 1.7277,
+      "step": 7807
+    },
+    {
+      "epoch": 0.821462388216728,
+      "grad_norm": 2.1678805351257324,
+      "learning_rate": 1.6260306661816383e-05,
+      "loss": 1.5226,
+      "step": 7808
+    },
+    {
+      "epoch": 0.8215675960021042,
+      "grad_norm": 1.2085503339767456,
+      "learning_rate": 1.6241684996054574e-05,
+      "loss": 1.6556,
+      "step": 7809
+    },
+    {
+      "epoch": 0.8216728037874803,
+      "grad_norm": 1.5834039449691772,
+      "learning_rate": 1.6223073056870097e-05,
+      "loss": 1.7377,
+      "step": 7810
+    },
+    {
+      "epoch": 0.8217780115728563,
+      "grad_norm": 1.1570148468017578,
+      "learning_rate": 1.6204470846424268e-05,
+      "loss": 1.5493,
+      "step": 7811
+    },
+    {
+      "epoch": 0.8218832193582325,
+      "grad_norm": 2.1178195476531982,
+      "learning_rate": 1.6185878366877295e-05,
+      "loss": 1.6665,
+      "step": 7812
+    },
+    {
+      "epoch": 0.8219884271436086,
+      "grad_norm": 2.094390392303467,
+      "learning_rate": 1.61672956203883e-05,
+      "loss": 1.0295,
+      "step": 7813
+    },
+    {
+      "epoch": 0.8220936349289848,
+      "grad_norm": 1.8944170475006104,
+      "learning_rate": 1.6148722609115176e-05,
+      "loss": 1.8045,
+      "step": 7814
+    },
+    {
+      "epoch": 0.8221988427143608,
+      "grad_norm": 1.5587190389633179,
+      "learning_rate": 1.613015933521481e-05,
+      "loss": 1.6959,
+      "step": 7815
+    },
+    {
+      "epoch": 0.822304050499737,
+      "grad_norm": 1.878710150718689,
+      "learning_rate": 1.6111605800842865e-05,
+      "loss": 1.62,
+      "step": 7816
+    },
+    {
+      "epoch": 0.8224092582851131,
+      "grad_norm": 1.67520272731781,
+      "learning_rate": 1.609306200815387e-05,
+      "loss": 1.3253,
+      "step": 7817
+    },
+    {
+      "epoch": 0.8225144660704892,
+      "grad_norm": 1.488664984703064,
+      "learning_rate": 1.607452795930131e-05,
+      "loss": 1.8211,
+      "step": 7818
+    },
+    {
+      "epoch": 0.8226196738558653,
+      "grad_norm": 1.5285340547561646,
+      "learning_rate": 1.6056003656437458e-05,
+      "loss": 1.2902,
+      "step": 7819
+    },
+    {
+      "epoch": 0.8227248816412415,
+      "grad_norm": 1.4336509704589844,
+      "learning_rate": 1.6037489101713465e-05,
+      "loss": 1.866,
+      "step": 7820
+    },
+    {
+      "epoch": 0.8228300894266176,
+      "grad_norm": 1.3473023176193237,
+      "learning_rate": 1.6018984297279393e-05,
+      "loss": 1.4556,
+      "step": 7821
+    },
+    {
+      "epoch": 0.8229352972119937,
+      "grad_norm": 1.2973744869232178,
+      "learning_rate": 1.6000489245284133e-05,
+      "loss": 1.9248,
+      "step": 7822
+    },
+    {
+      "epoch": 0.8230405049973698,
+      "grad_norm": 1.9646800756454468,
+      "learning_rate": 1.5982003947875467e-05,
+      "loss": 1.7111,
+      "step": 7823
+    },
+    {
+      "epoch": 0.823145712782746,
+      "grad_norm": 1.3596656322479248,
+      "learning_rate": 1.59635284072e-05,
+      "loss": 1.5034,
+      "step": 7824
+    },
+    {
+      "epoch": 0.823250920568122,
+      "grad_norm": 2.065723180770874,
+      "learning_rate": 1.594506262540324e-05,
+      "loss": 1.876,
+      "step": 7825
+    },
+    {
+      "epoch": 0.8233561283534981,
+      "grad_norm": 1.7020695209503174,
+      "learning_rate": 1.592660660462959e-05,
+      "loss": 1.237,
+      "step": 7826
+    },
+    {
+      "epoch": 0.8234613361388743,
+      "grad_norm": 1.4010860919952393,
+      "learning_rate": 1.5908160347022272e-05,
+      "loss": 1.9249,
+      "step": 7827
+    },
+    {
+      "epoch": 0.8235665439242504,
+      "grad_norm": 1.8709789514541626,
+      "learning_rate": 1.588972385472336e-05,
+      "loss": 2.0006,
+      "step": 7828
+    },
+    {
+      "epoch": 0.8236717517096265,
+      "grad_norm": 1.6557965278625488,
+      "learning_rate": 1.5871297129873864e-05,
+      "loss": 1.9393,
+      "step": 7829
+    },
+    {
+      "epoch": 0.8237769594950026,
+      "grad_norm": 1.690056562423706,
+      "learning_rate": 1.5852880174613617e-05,
+      "loss": 1.6349,
+      "step": 7830
+    },
+    {
+      "epoch": 0.8238821672803788,
+      "grad_norm": 1.9580641984939575,
+      "learning_rate": 1.5834472991081273e-05,
+      "loss": 1.3244,
+      "step": 7831
+    },
+    {
+      "epoch": 0.8239873750657548,
+      "grad_norm": 1.4145424365997314,
+      "learning_rate": 1.5816075581414458e-05,
+      "loss": 1.5102,
+      "step": 7832
+    },
+    {
+      "epoch": 0.824092582851131,
+      "grad_norm": 1.6314736604690552,
+      "learning_rate": 1.5797687947749563e-05,
+      "loss": 1.424,
+      "step": 7833
+    },
+    {
+      "epoch": 0.8241977906365071,
+      "grad_norm": 1.2429150342941284,
+      "learning_rate": 1.577931009222189e-05,
+      "loss": 1.2783,
+      "step": 7834
+    },
+    {
+      "epoch": 0.8243029984218833,
+      "grad_norm": 1.750223994255066,
+      "learning_rate": 1.5760942016965608e-05,
+      "loss": 1.892,
+      "step": 7835
+    },
+    {
+      "epoch": 0.8244082062072593,
+      "grad_norm": 1.6230374574661255,
+      "learning_rate": 1.5742583724113746e-05,
+      "loss": 1.8513,
+      "step": 7836
+    },
+    {
+      "epoch": 0.8245134139926354,
+      "grad_norm": 1.2478396892547607,
+      "learning_rate": 1.5724235215798167e-05,
+      "loss": 1.8733,
+      "step": 7837
+    },
+    {
+      "epoch": 0.8246186217780116,
+      "grad_norm": 1.6397819519042969,
+      "learning_rate": 1.5705896494149654e-05,
+      "loss": 1.4,
+      "step": 7838
+    },
+    {
+      "epoch": 0.8247238295633877,
+      "grad_norm": 1.4806641340255737,
+      "learning_rate": 1.56875675612978e-05,
+      "loss": 1.4183,
+      "step": 7839
+    },
+    {
+      "epoch": 0.8248290373487638,
+      "grad_norm": 1.4073078632354736,
+      "learning_rate": 1.5669248419371085e-05,
+      "loss": 1.7526,
+      "step": 7840
+    },
+    {
+      "epoch": 0.8249342451341399,
+      "grad_norm": 1.6545143127441406,
+      "learning_rate": 1.565093907049686e-05,
+      "loss": 1.8446,
+      "step": 7841
+    },
+    {
+      "epoch": 0.8250394529195161,
+      "grad_norm": 1.461369276046753,
+      "learning_rate": 1.563263951680134e-05,
+      "loss": 1.4983,
+      "step": 7842
+    },
+    {
+      "epoch": 0.8251446607048921,
+      "grad_norm": 1.4910365343093872,
+      "learning_rate": 1.5614349760409552e-05,
+      "loss": 1.63,
+      "step": 7843
+    },
+    {
+      "epoch": 0.8252498684902683,
+      "grad_norm": 1.4851601123809814,
+      "learning_rate": 1.5596069803445502e-05,
+      "loss": 1.3496,
+      "step": 7844
+    },
+    {
+      "epoch": 0.8253550762756444,
+      "grad_norm": 1.3788039684295654,
+      "learning_rate": 1.5577799648031876e-05,
+      "loss": 1.8372,
+      "step": 7845
+    },
+    {
+      "epoch": 0.8254602840610206,
+      "grad_norm": 1.6159765720367432,
+      "learning_rate": 1.5559539296290403e-05,
+      "loss": 1.5477,
+      "step": 7846
+    },
+    {
+      "epoch": 0.8255654918463966,
+      "grad_norm": 2.3057737350463867,
+      "learning_rate": 1.5541288750341575e-05,
+      "loss": 1.866,
+      "step": 7847
+    },
+    {
+      "epoch": 0.8256706996317728,
+      "grad_norm": 2.4619557857513428,
+      "learning_rate": 1.5523048012304754e-05,
+      "loss": 1.6384,
+      "step": 7848
+    },
+    {
+      "epoch": 0.8257759074171489,
+      "grad_norm": 1.4026074409484863,
+      "learning_rate": 1.5504817084298207e-05,
+      "loss": 1.2915,
+      "step": 7849
+    },
+    {
+      "epoch": 0.8258811152025249,
+      "grad_norm": 1.2069450616836548,
+      "learning_rate": 1.548659596843902e-05,
+      "loss": 1.6054,
+      "step": 7850
+    },
+    {
+      "epoch": 0.8259863229879011,
+      "grad_norm": 1.7536677122116089,
+      "learning_rate": 1.5468384666843115e-05,
+      "loss": 1.7181,
+      "step": 7851
+    },
+    {
+      "epoch": 0.8260915307732772,
+      "grad_norm": 1.5129624605178833,
+      "learning_rate": 1.545018318162538e-05,
+      "loss": 2.1859,
+      "step": 7852
+    },
+    {
+      "epoch": 0.8261967385586534,
+      "grad_norm": 1.5461325645446777,
+      "learning_rate": 1.5431991514899446e-05,
+      "loss": 2.0414,
+      "step": 7853
+    },
+    {
+      "epoch": 0.8263019463440294,
+      "grad_norm": 1.6245604753494263,
+      "learning_rate": 1.5413809668777844e-05,
+      "loss": 1.2136,
+      "step": 7854
+    },
+    {
+      "epoch": 0.8264071541294056,
+      "grad_norm": 1.2918306589126587,
+      "learning_rate": 1.5395637645372007e-05,
+      "loss": 1.3098,
+      "step": 7855
+    },
+    {
+      "epoch": 0.8265123619147817,
+      "grad_norm": 1.5559884309768677,
+      "learning_rate": 1.5377475446792178e-05,
+      "loss": 1.5964,
+      "step": 7856
+    },
+    {
+      "epoch": 0.8266175697001578,
+      "grad_norm": 1.8957018852233887,
+      "learning_rate": 1.535932307514745e-05,
+      "loss": 1.8956,
+      "step": 7857
+    },
+    {
+      "epoch": 0.8267227774855339,
+      "grad_norm": 2.0008089542388916,
+      "learning_rate": 1.534118053254584e-05,
+      "loss": 1.6321,
+      "step": 7858
+    },
+    {
+      "epoch": 0.8268279852709101,
+      "grad_norm": 1.7587536573410034,
+      "learning_rate": 1.5323047821094126e-05,
+      "loss": 1.8533,
+      "step": 7859
+    },
+    {
+      "epoch": 0.8269331930562862,
+      "grad_norm": 2.5158214569091797,
+      "learning_rate": 1.5304924942898068e-05,
+      "loss": 1.1862,
+      "step": 7860
+    },
+    {
+      "epoch": 0.8270384008416622,
+      "grad_norm": 1.6874955892562866,
+      "learning_rate": 1.528681190006218e-05,
+      "loss": 1.5662,
+      "step": 7861
+    },
+    {
+      "epoch": 0.8271436086270384,
+      "grad_norm": 2.3075296878814697,
+      "learning_rate": 1.5268708694689847e-05,
+      "loss": 1.449,
+      "step": 7862
+    },
+    {
+      "epoch": 0.8272488164124145,
+      "grad_norm": 1.3436137437820435,
+      "learning_rate": 1.5250615328883388e-05,
+      "loss": 1.3811,
+      "step": 7863
+    },
+    {
+      "epoch": 0.8273540241977906,
+      "grad_norm": 2.0536916255950928,
+      "learning_rate": 1.5232531804743899e-05,
+      "loss": 1.6885,
+      "step": 7864
+    },
+    {
+      "epoch": 0.8274592319831667,
+      "grad_norm": 1.4442811012268066,
+      "learning_rate": 1.5214458124371345e-05,
+      "loss": 1.5104,
+      "step": 7865
+    },
+    {
+      "epoch": 0.8275644397685429,
+      "grad_norm": 2.0475945472717285,
+      "learning_rate": 1.5196394289864591e-05,
+      "loss": 1.7676,
+      "step": 7866
+    },
+    {
+      "epoch": 0.827669647553919,
+      "grad_norm": 1.2523587942123413,
+      "learning_rate": 1.5178340303321314e-05,
+      "loss": 1.5507,
+      "step": 7867
+    },
+    {
+      "epoch": 0.8277748553392951,
+      "grad_norm": 2.2663662433624268,
+      "learning_rate": 1.516029616683804e-05,
+      "loss": 1.4212,
+      "step": 7868
+    },
+    {
+      "epoch": 0.8278800631246712,
+      "grad_norm": 1.719728946685791,
+      "learning_rate": 1.5142261882510223e-05,
+      "loss": 2.0605,
+      "step": 7869
+    },
+    {
+      "epoch": 0.8279852709100474,
+      "grad_norm": 1.38359534740448,
+      "learning_rate": 1.5124237452432077e-05,
+      "loss": 1.6727,
+      "step": 7870
+    },
+    {
+      "epoch": 0.8280904786954235,
+      "grad_norm": 2.143702507019043,
+      "learning_rate": 1.5106222878696775e-05,
+      "loss": 1.3658,
+      "step": 7871
+    },
+    {
+      "epoch": 0.8281956864807996,
+      "grad_norm": 1.9734598398208618,
+      "learning_rate": 1.5088218163396262e-05,
+      "loss": 1.5329,
+      "step": 7872
+    },
+    {
+      "epoch": 0.8283008942661757,
+      "grad_norm": 1.5374348163604736,
+      "learning_rate": 1.5070223308621345e-05,
+      "loss": 1.5764,
+      "step": 7873
+    },
+    {
+      "epoch": 0.8284061020515519,
+      "grad_norm": 1.664512038230896,
+      "learning_rate": 1.5052238316461753e-05,
+      "loss": 1.4934,
+      "step": 7874
+    },
+    {
+      "epoch": 0.8285113098369279,
+      "grad_norm": 1.5587176084518433,
+      "learning_rate": 1.5034263189005993e-05,
+      "loss": 1.8861,
+      "step": 7875
+    },
+    {
+      "epoch": 0.828616517622304,
+      "grad_norm": 1.4705950021743774,
+      "learning_rate": 1.5016297928341438e-05,
+      "loss": 1.7146,
+      "step": 7876
+    },
+    {
+      "epoch": 0.8287217254076802,
+      "grad_norm": 1.6136786937713623,
+      "learning_rate": 1.499834253655439e-05,
+      "loss": 1.3576,
+      "step": 7877
+    },
+    {
+      "epoch": 0.8288269331930563,
+      "grad_norm": 1.1220991611480713,
+      "learning_rate": 1.4980397015729908e-05,
+      "loss": 1.9862,
+      "step": 7878
+    },
+    {
+      "epoch": 0.8289321409784324,
+      "grad_norm": 1.931833267211914,
+      "learning_rate": 1.496246136795194e-05,
+      "loss": 1.7605,
+      "step": 7879
+    },
+    {
+      "epoch": 0.8290373487638085,
+      "grad_norm": 3.0278358459472656,
+      "learning_rate": 1.4944535595303344e-05,
+      "loss": 1.4795,
+      "step": 7880
+    },
+    {
+      "epoch": 0.8291425565491847,
+      "grad_norm": 2.9382994174957275,
+      "learning_rate": 1.492661969986574e-05,
+      "loss": 1.4424,
+      "step": 7881
+    },
+    {
+      "epoch": 0.8292477643345607,
+      "grad_norm": 1.7495644092559814,
+      "learning_rate": 1.4908713683719632e-05,
+      "loss": 1.7025,
+      "step": 7882
+    },
+    {
+      "epoch": 0.8293529721199369,
+      "grad_norm": 1.588752269744873,
+      "learning_rate": 1.4890817548944424e-05,
+      "loss": 0.9935,
+      "step": 7883
+    },
+    {
+      "epoch": 0.829458179905313,
+      "grad_norm": 1.3464096784591675,
+      "learning_rate": 1.4872931297618308e-05,
+      "loss": 1.6566,
+      "step": 7884
+    },
+    {
+      "epoch": 0.8295633876906892,
+      "grad_norm": 1.621984601020813,
+      "learning_rate": 1.4855054931818357e-05,
+      "loss": 1.4483,
+      "step": 7885
+    },
+    {
+      "epoch": 0.8296685954760652,
+      "grad_norm": 1.6145195960998535,
+      "learning_rate": 1.483718845362051e-05,
+      "loss": 1.6128,
+      "step": 7886
+    },
+    {
+      "epoch": 0.8297738032614413,
+      "grad_norm": 1.5069514513015747,
+      "learning_rate": 1.4819331865099539e-05,
+      "loss": 1.5058,
+      "step": 7887
+    },
+    {
+      "epoch": 0.8298790110468175,
+      "grad_norm": 2.006563425064087,
+      "learning_rate": 1.4801485168329066e-05,
+      "loss": 1.7359,
+      "step": 7888
+    },
+    {
+      "epoch": 0.8299842188321935,
+      "grad_norm": 1.1885156631469727,
+      "learning_rate": 1.4783648365381563e-05,
+      "loss": 1.1542,
+      "step": 7889
+    },
+    {
+      "epoch": 0.8300894266175697,
+      "grad_norm": 2.4285728931427,
+      "learning_rate": 1.4765821458328355e-05,
+      "loss": 1.7459,
+      "step": 7890
+    },
+    {
+      "epoch": 0.8301946344029458,
+      "grad_norm": 1.307138442993164,
+      "learning_rate": 1.4748004449239639e-05,
+      "loss": 1.1976,
+      "step": 7891
+    },
+    {
+      "epoch": 0.830299842188322,
+      "grad_norm": 2.049842357635498,
+      "learning_rate": 1.473019734018445e-05,
+      "loss": 1.3167,
+      "step": 7892
+    },
+    {
+      "epoch": 0.830405049973698,
+      "grad_norm": 2.1878762245178223,
+      "learning_rate": 1.471240013323063e-05,
+      "loss": 1.5895,
+      "step": 7893
+    },
+    {
+      "epoch": 0.8305102577590742,
+      "grad_norm": 1.3817623853683472,
+      "learning_rate": 1.4694612830444953e-05,
+      "loss": 1.8225,
+      "step": 7894
+    },
+    {
+      "epoch": 0.8306154655444503,
+      "grad_norm": 1.5477038621902466,
+      "learning_rate": 1.4676835433892989e-05,
+      "loss": 1.5813,
+      "step": 7895
+    },
+    {
+      "epoch": 0.8307206733298264,
+      "grad_norm": 2.0993480682373047,
+      "learning_rate": 1.4659067945639137e-05,
+      "loss": 2.0678,
+      "step": 7896
+    },
+    {
+      "epoch": 0.8308258811152025,
+      "grad_norm": 1.446555733680725,
+      "learning_rate": 1.464131036774674e-05,
+      "loss": 1.5909,
+      "step": 7897
+    },
+    {
+      "epoch": 0.8309310889005787,
+      "grad_norm": 1.8362712860107422,
+      "learning_rate": 1.4623562702277882e-05,
+      "loss": 1.1406,
+      "step": 7898
+    },
+    {
+      "epoch": 0.8310362966859548,
+      "grad_norm": 1.4514873027801514,
+      "learning_rate": 1.4605824951293524e-05,
+      "loss": 1.7013,
+      "step": 7899
+    },
+    {
+      "epoch": 0.8311415044713308,
+      "grad_norm": 1.1335458755493164,
+      "learning_rate": 1.4588097116853538e-05,
+      "loss": 1.2316,
+      "step": 7900
+    },
+    {
+      "epoch": 0.831246712256707,
+      "grad_norm": 1.7612231969833374,
+      "learning_rate": 1.4570379201016581e-05,
+      "loss": 1.456,
+      "step": 7901
+    },
+    {
+      "epoch": 0.8313519200420831,
+      "grad_norm": 2.277941942214966,
+      "learning_rate": 1.4552671205840163e-05,
+      "loss": 2.1247,
+      "step": 7902
+    },
+    {
+      "epoch": 0.8314571278274593,
+      "grad_norm": 1.3662701845169067,
+      "learning_rate": 1.4534973133380669e-05,
+      "loss": 1.7872,
+      "step": 7903
+    },
+    {
+      "epoch": 0.8315623356128353,
+      "grad_norm": 1.494207501411438,
+      "learning_rate": 1.4517284985693335e-05,
+      "loss": 1.7331,
+      "step": 7904
+    },
+    {
+      "epoch": 0.8316675433982115,
+      "grad_norm": 1.8795514106750488,
+      "learning_rate": 1.4499606764832175e-05,
+      "loss": 1.3668,
+      "step": 7905
+    },
+    {
+      "epoch": 0.8317727511835876,
+      "grad_norm": 1.7564284801483154,
+      "learning_rate": 1.4481938472850154e-05,
+      "loss": 1.5335,
+      "step": 7906
+    },
+    {
+      "epoch": 0.8318779589689637,
+      "grad_norm": 2.4675612449645996,
+      "learning_rate": 1.4464280111799e-05,
+      "loss": 1.1034,
+      "step": 7907
+    },
+    {
+      "epoch": 0.8319831667543398,
+      "grad_norm": 1.5577418804168701,
+      "learning_rate": 1.4446631683729372e-05,
+      "loss": 2.0499,
+      "step": 7908
+    },
+    {
+      "epoch": 0.832088374539716,
+      "grad_norm": 2.425018787384033,
+      "learning_rate": 1.4428993190690677e-05,
+      "loss": 1.6102,
+      "step": 7909
+    },
+    {
+      "epoch": 0.8321935823250921,
+      "grad_norm": 1.142414927482605,
+      "learning_rate": 1.4411364634731183e-05,
+      "loss": 1.52,
+      "step": 7910
+    },
+    {
+      "epoch": 0.8322987901104681,
+      "grad_norm": 1.444345474243164,
+      "learning_rate": 1.4393746017898113e-05,
+      "loss": 1.5863,
+      "step": 7911
+    },
+    {
+      "epoch": 0.8324039978958443,
+      "grad_norm": 1.6111942529678345,
+      "learning_rate": 1.4376137342237417e-05,
+      "loss": 1.4745,
+      "step": 7912
+    },
+    {
+      "epoch": 0.8325092056812204,
+      "grad_norm": 2.1231207847595215,
+      "learning_rate": 1.4358538609793915e-05,
+      "loss": 1.719,
+      "step": 7913
+    },
+    {
+      "epoch": 0.8326144134665965,
+      "grad_norm": 1.4172590970993042,
+      "learning_rate": 1.4340949822611338e-05,
+      "loss": 1.4065,
+      "step": 7914
+    },
+    {
+      "epoch": 0.8327196212519726,
+      "grad_norm": 1.674170970916748,
+      "learning_rate": 1.4323370982732176e-05,
+      "loss": 1.8991,
+      "step": 7915
+    },
+    {
+      "epoch": 0.8328248290373488,
+      "grad_norm": 1.4009698629379272,
+      "learning_rate": 1.4305802092197829e-05,
+      "loss": 1.5887,
+      "step": 7916
+    },
+    {
+      "epoch": 0.8329300368227249,
+      "grad_norm": 1.580376148223877,
+      "learning_rate": 1.4288243153048497e-05,
+      "loss": 2.2614,
+      "step": 7917
+    },
+    {
+      "epoch": 0.833035244608101,
+      "grad_norm": 2.047874689102173,
+      "learning_rate": 1.4270694167323228e-05,
+      "loss": 1.3469,
+      "step": 7918
+    },
+    {
+      "epoch": 0.8331404523934771,
+      "grad_norm": 2.807647705078125,
+      "learning_rate": 1.425315513705997e-05,
+      "loss": 1.8353,
+      "step": 7919
+    },
+    {
+      "epoch": 0.8332456601788533,
+      "grad_norm": 1.343658208847046,
+      "learning_rate": 1.4235626064295438e-05,
+      "loss": 1.6872,
+      "step": 7920
+    },
+    {
+      "epoch": 0.8333508679642293,
+      "grad_norm": 1.3211522102355957,
+      "learning_rate": 1.4218106951065224e-05,
+      "loss": 1.6872,
+      "step": 7921
+    },
+    {
+      "epoch": 0.8334560757496055,
+      "grad_norm": 2.710646867752075,
+      "learning_rate": 1.4200597799403793e-05,
+      "loss": 1.703,
+      "step": 7922
+    },
+    {
+      "epoch": 0.8335612835349816,
+      "grad_norm": 1.5261638164520264,
+      "learning_rate": 1.4183098611344415e-05,
+      "loss": 1.5745,
+      "step": 7923
+    },
+    {
+      "epoch": 0.8336664913203577,
+      "grad_norm": 1.4623990058898926,
+      "learning_rate": 1.4165609388919176e-05,
+      "loss": 1.606,
+      "step": 7924
+    },
+    {
+      "epoch": 0.8337716991057338,
+      "grad_norm": 1.5466994047164917,
+      "learning_rate": 1.414813013415911e-05,
+      "loss": 1.7387,
+      "step": 7925
+    },
+    {
+      "epoch": 0.8338769068911099,
+      "grad_norm": 1.6613030433654785,
+      "learning_rate": 1.4130660849093969e-05,
+      "loss": 1.8445,
+      "step": 7926
+    },
+    {
+      "epoch": 0.8339821146764861,
+      "grad_norm": 1.4952675104141235,
+      "learning_rate": 1.4113201535752407e-05,
+      "loss": 1.7024,
+      "step": 7927
+    },
+    {
+      "epoch": 0.8340873224618621,
+      "grad_norm": 1.191802740097046,
+      "learning_rate": 1.409575219616196e-05,
+      "loss": 1.56,
+      "step": 7928
+    },
+    {
+      "epoch": 0.8341925302472383,
+      "grad_norm": 1.860492467880249,
+      "learning_rate": 1.4078312832348938e-05,
+      "loss": 0.9361,
+      "step": 7929
+    },
+    {
+      "epoch": 0.8342977380326144,
+      "grad_norm": 1.4754762649536133,
+      "learning_rate": 1.4060883446338502e-05,
+      "loss": 1.5502,
+      "step": 7930
+    },
+    {
+      "epoch": 0.8344029458179906,
+      "grad_norm": 1.2694263458251953,
+      "learning_rate": 1.4043464040154686e-05,
+      "loss": 1.761,
+      "step": 7931
+    },
+    {
+      "epoch": 0.8345081536033666,
+      "grad_norm": 2.5978357791900635,
+      "learning_rate": 1.4026054615820317e-05,
+      "loss": 0.6868,
+      "step": 7932
+    },
+    {
+      "epoch": 0.8346133613887428,
+      "grad_norm": 1.792586326599121,
+      "learning_rate": 1.4008655175357144e-05,
+      "loss": 1.808,
+      "step": 7933
+    },
+    {
+      "epoch": 0.8347185691741189,
+      "grad_norm": 1.2222882509231567,
+      "learning_rate": 1.3991265720785685e-05,
+      "loss": 1.7221,
+      "step": 7934
+    },
+    {
+      "epoch": 0.8348237769594951,
+      "grad_norm": 1.4182674884796143,
+      "learning_rate": 1.397388625412529e-05,
+      "loss": 1.4725,
+      "step": 7935
+    },
+    {
+      "epoch": 0.8349289847448711,
+      "grad_norm": 1.2701834440231323,
+      "learning_rate": 1.3956516777394235e-05,
+      "loss": 1.3707,
+      "step": 7936
+    },
+    {
+      "epoch": 0.8350341925302472,
+      "grad_norm": 1.180488109588623,
+      "learning_rate": 1.393915729260955e-05,
+      "loss": 1.5807,
+      "step": 7937
+    },
+    {
+      "epoch": 0.8351394003156234,
+      "grad_norm": 2.2820448875427246,
+      "learning_rate": 1.3921807801787112e-05,
+      "loss": 1.4147,
+      "step": 7938
+    },
+    {
+      "epoch": 0.8352446081009994,
+      "grad_norm": 2.877859354019165,
+      "learning_rate": 1.390446830694172e-05,
+      "loss": 1.2865,
+      "step": 7939
+    },
+    {
+      "epoch": 0.8353498158863756,
+      "grad_norm": 2.0427865982055664,
+      "learning_rate": 1.3887138810086908e-05,
+      "loss": 1.4107,
+      "step": 7940
+    },
+    {
+      "epoch": 0.8354550236717517,
+      "grad_norm": 1.6468256711959839,
+      "learning_rate": 1.3869819313235077e-05,
+      "loss": 1.4342,
+      "step": 7941
+    },
+    {
+      "epoch": 0.8355602314571279,
+      "grad_norm": 2.2093889713287354,
+      "learning_rate": 1.385250981839753e-05,
+      "loss": 1.8815,
+      "step": 7942
+    },
+    {
+      "epoch": 0.8356654392425039,
+      "grad_norm": 1.257178783416748,
+      "learning_rate": 1.3835210327584348e-05,
+      "loss": 1.3339,
+      "step": 7943
+    },
+    {
+      "epoch": 0.8357706470278801,
+      "grad_norm": 1.405913233757019,
+      "learning_rate": 1.3817920842804433e-05,
+      "loss": 1.3401,
+      "step": 7944
+    },
+    {
+      "epoch": 0.8358758548132562,
+      "grad_norm": 2.065035820007324,
+      "learning_rate": 1.3800641366065604e-05,
+      "loss": 1.1262,
+      "step": 7945
+    },
+    {
+      "epoch": 0.8359810625986323,
+      "grad_norm": 1.2551019191741943,
+      "learning_rate": 1.3783371899374442e-05,
+      "loss": 1.651,
+      "step": 7946
+    },
+    {
+      "epoch": 0.8360862703840084,
+      "grad_norm": 1.498842477798462,
+      "learning_rate": 1.3766112444736368e-05,
+      "loss": 1.2671,
+      "step": 7947
+    },
+    {
+      "epoch": 0.8361914781693845,
+      "grad_norm": 1.7398850917816162,
+      "learning_rate": 1.3748863004155732e-05,
+      "loss": 1.9418,
+      "step": 7948
+    },
+    {
+      "epoch": 0.8362966859547607,
+      "grad_norm": 1.3350509405136108,
+      "learning_rate": 1.37316235796356e-05,
+      "loss": 1.7554,
+      "step": 7949
+    },
+    {
+      "epoch": 0.8364018937401367,
+      "grad_norm": 1.206315279006958,
+      "learning_rate": 1.3714394173177936e-05,
+      "loss": 1.4759,
+      "step": 7950
+    },
+    {
+      "epoch": 0.8365071015255129,
+      "grad_norm": 2.4317026138305664,
+      "learning_rate": 1.3697174786783584e-05,
+      "loss": 1.921,
+      "step": 7951
+    },
+    {
+      "epoch": 0.836612309310889,
+      "grad_norm": 1.54344642162323,
+      "learning_rate": 1.3679965422452101e-05,
+      "loss": 2.177,
+      "step": 7952
+    },
+    {
+      "epoch": 0.8367175170962651,
+      "grad_norm": 1.8043365478515625,
+      "learning_rate": 1.3662766082181999e-05,
+      "loss": 1.2102,
+      "step": 7953
+    },
+    {
+      "epoch": 0.8368227248816412,
+      "grad_norm": 1.4540766477584839,
+      "learning_rate": 1.364557676797057e-05,
+      "loss": 0.7462,
+      "step": 7954
+    },
+    {
+      "epoch": 0.8369279326670174,
+      "grad_norm": 1.8703304529190063,
+      "learning_rate": 1.3628397481813936e-05,
+      "loss": 1.6058,
+      "step": 7955
+    },
+    {
+      "epoch": 0.8370331404523935,
+      "grad_norm": 1.9202097654342651,
+      "learning_rate": 1.3611228225707107e-05,
+      "loss": 1.4126,
+      "step": 7956
+    },
+    {
+      "epoch": 0.8371383482377696,
+      "grad_norm": 6.3809990882873535,
+      "learning_rate": 1.3594069001643872e-05,
+      "loss": 1.2723,
+      "step": 7957
+    },
+    {
+      "epoch": 0.8372435560231457,
+      "grad_norm": 1.363297462463379,
+      "learning_rate": 1.3576919811616862e-05,
+      "loss": 1.6663,
+      "step": 7958
+    },
+    {
+      "epoch": 0.8373487638085219,
+      "grad_norm": 4.482295513153076,
+      "learning_rate": 1.3559780657617582e-05,
+      "loss": 1.5827,
+      "step": 7959
+    },
+    {
+      "epoch": 0.8374539715938979,
+      "grad_norm": 2.7441794872283936,
+      "learning_rate": 1.354265154163633e-05,
+      "loss": 1.9887,
+      "step": 7960
+    },
+    {
+      "epoch": 0.837559179379274,
+      "grad_norm": 1.571427583694458,
+      "learning_rate": 1.352553246566225e-05,
+      "loss": 1.7604,
+      "step": 7961
+    },
+    {
+      "epoch": 0.8376643871646502,
+      "grad_norm": 1.7176398038864136,
+      "learning_rate": 1.3508423431683337e-05,
+      "loss": 1.6065,
+      "step": 7962
+    },
+    {
+      "epoch": 0.8377695949500263,
+      "grad_norm": 1.1875994205474854,
+      "learning_rate": 1.3491324441686392e-05,
+      "loss": 1.7916,
+      "step": 7963
+    },
+    {
+      "epoch": 0.8378748027354024,
+      "grad_norm": 1.7343966960906982,
+      "learning_rate": 1.3474235497657084e-05,
+      "loss": 1.7253,
+      "step": 7964
+    },
+    {
+      "epoch": 0.8379800105207785,
+      "grad_norm": 2.453002452850342,
+      "learning_rate": 1.345715660157989e-05,
+      "loss": 2.2406,
+      "step": 7965
+    },
+    {
+      "epoch": 0.8380852183061547,
+      "grad_norm": 2.0220048427581787,
+      "learning_rate": 1.3440087755438102e-05,
+      "loss": 1.5343,
+      "step": 7966
+    },
+    {
+      "epoch": 0.8381904260915308,
+      "grad_norm": 2.1328377723693848,
+      "learning_rate": 1.3423028961213912e-05,
+      "loss": 2.0797,
+      "step": 7967
+    },
+    {
+      "epoch": 0.8382956338769069,
+      "grad_norm": 2.1333796977996826,
+      "learning_rate": 1.340598022088827e-05,
+      "loss": 1.8357,
+      "step": 7968
+    },
+    {
+      "epoch": 0.838400841662283,
+      "grad_norm": 1.2837255001068115,
+      "learning_rate": 1.338894153644098e-05,
+      "loss": 1.6283,
+      "step": 7969
+    },
+    {
+      "epoch": 0.8385060494476592,
+      "grad_norm": 1.9028080701828003,
+      "learning_rate": 1.3371912909850726e-05,
+      "loss": 1.2375,
+      "step": 7970
+    },
+    {
+      "epoch": 0.8386112572330352,
+      "grad_norm": 1.260032296180725,
+      "learning_rate": 1.335489434309496e-05,
+      "loss": 1.5313,
+      "step": 7971
+    },
+    {
+      "epoch": 0.8387164650184114,
+      "grad_norm": 1.5635043382644653,
+      "learning_rate": 1.3337885838149988e-05,
+      "loss": 1.5254,
+      "step": 7972
+    },
+    {
+      "epoch": 0.8388216728037875,
+      "grad_norm": 1.1617438793182373,
+      "learning_rate": 1.3320887396990999e-05,
+      "loss": 1.3332,
+      "step": 7973
+    },
+    {
+      "epoch": 0.8389268805891636,
+      "grad_norm": 1.6330403089523315,
+      "learning_rate": 1.3303899021591882e-05,
+      "loss": 1.3151,
+      "step": 7974
+    },
+    {
+      "epoch": 0.8390320883745397,
+      "grad_norm": 1.6695667505264282,
+      "learning_rate": 1.328692071392552e-05,
+      "loss": 1.3915,
+      "step": 7975
+    },
+    {
+      "epoch": 0.8391372961599158,
+      "grad_norm": 1.507834553718567,
+      "learning_rate": 1.3269952475963509e-05,
+      "loss": 1.7375,
+      "step": 7976
+    },
+    {
+      "epoch": 0.839242503945292,
+      "grad_norm": 2.011200428009033,
+      "learning_rate": 1.3252994309676303e-05,
+      "loss": 1.5429,
+      "step": 7977
+    },
+    {
+      "epoch": 0.839347711730668,
+      "grad_norm": 1.5089296102523804,
+      "learning_rate": 1.3236046217033237e-05,
+      "loss": 1.9026,
+      "step": 7978
+    },
+    {
+      "epoch": 0.8394529195160442,
+      "grad_norm": 2.053021192550659,
+      "learning_rate": 1.3219108200002418e-05,
+      "loss": 1.2597,
+      "step": 7979
+    },
+    {
+      "epoch": 0.8395581273014203,
+      "grad_norm": 2.294508457183838,
+      "learning_rate": 1.3202180260550778e-05,
+      "loss": 2.0704,
+      "step": 7980
+    },
+    {
+      "epoch": 0.8396633350867965,
+      "grad_norm": 1.4443711042404175,
+      "learning_rate": 1.318526240064415e-05,
+      "loss": 1.6057,
+      "step": 7981
+    },
+    {
+      "epoch": 0.8397685428721725,
+      "grad_norm": 1.4994503259658813,
+      "learning_rate": 1.3168354622247115e-05,
+      "loss": 1.7615,
+      "step": 7982
+    },
+    {
+      "epoch": 0.8398737506575487,
+      "grad_norm": 2.1308159828186035,
+      "learning_rate": 1.3151456927323113e-05,
+      "loss": 1.8955,
+      "step": 7983
+    },
+    {
+      "epoch": 0.8399789584429248,
+      "grad_norm": 1.4624441862106323,
+      "learning_rate": 1.3134569317834454e-05,
+      "loss": 1.7515,
+      "step": 7984
+    },
+    {
+      "epoch": 0.8400841662283008,
+      "grad_norm": 1.2271381616592407,
+      "learning_rate": 1.3117691795742226e-05,
+      "loss": 1.494,
+      "step": 7985
+    },
+    {
+      "epoch": 0.840189374013677,
+      "grad_norm": 1.1306962966918945,
+      "learning_rate": 1.3100824363006326e-05,
+      "loss": 1.977,
+      "step": 7986
+    },
+    {
+      "epoch": 0.8402945817990531,
+      "grad_norm": 1.7996954917907715,
+      "learning_rate": 1.3083967021585564e-05,
+      "loss": 1.9938,
+      "step": 7987
+    },
+    {
+      "epoch": 0.8403997895844293,
+      "grad_norm": 1.3190232515335083,
+      "learning_rate": 1.3067119773437498e-05,
+      "loss": 1.3396,
+      "step": 7988
+    },
+    {
+      "epoch": 0.8405049973698053,
+      "grad_norm": 1.5940443277359009,
+      "learning_rate": 1.3050282620518528e-05,
+      "loss": 1.1903,
+      "step": 7989
+    },
+    {
+      "epoch": 0.8406102051551815,
+      "grad_norm": 1.4561878442764282,
+      "learning_rate": 1.3033455564783948e-05,
+      "loss": 1.3508,
+      "step": 7990
+    },
+    {
+      "epoch": 0.8407154129405576,
+      "grad_norm": 1.4680160284042358,
+      "learning_rate": 1.3016638608187792e-05,
+      "loss": 1.8019,
+      "step": 7991
+    },
+    {
+      "epoch": 0.8408206207259337,
+      "grad_norm": 1.9364701509475708,
+      "learning_rate": 1.2999831752682955e-05,
+      "loss": 1.4724,
+      "step": 7992
+    },
+    {
+      "epoch": 0.8409258285113098,
+      "grad_norm": 2.371671199798584,
+      "learning_rate": 1.2983035000221177e-05,
+      "loss": 2.2839,
+      "step": 7993
+    },
+    {
+      "epoch": 0.841031036296686,
+      "grad_norm": 1.6771042346954346,
+      "learning_rate": 1.2966248352753018e-05,
+      "loss": 1.2167,
+      "step": 7994
+    },
+    {
+      "epoch": 0.8411362440820621,
+      "grad_norm": 2.0259828567504883,
+      "learning_rate": 1.294947181222783e-05,
+      "loss": 1.7223,
+      "step": 7995
+    },
+    {
+      "epoch": 0.8412414518674382,
+      "grad_norm": 1.4655014276504517,
+      "learning_rate": 1.2932705380593846e-05,
+      "loss": 1.8478,
+      "step": 7996
+    },
+    {
+      "epoch": 0.8413466596528143,
+      "grad_norm": 1.9223905801773071,
+      "learning_rate": 1.2915949059798038e-05,
+      "loss": 1.5038,
+      "step": 7997
+    },
+    {
+      "epoch": 0.8414518674381904,
+      "grad_norm": 1.4577714204788208,
+      "learning_rate": 1.2899202851786341e-05,
+      "loss": 1.5354,
+      "step": 7998
+    },
+    {
+      "epoch": 0.8415570752235666,
+      "grad_norm": 1.3747444152832031,
+      "learning_rate": 1.2882466758503397e-05,
+      "loss": 2.3442,
+      "step": 7999
+    },
+    {
+      "epoch": 0.8416622830089426,
+      "grad_norm": 1.4338715076446533,
+      "learning_rate": 1.2865740781892699e-05,
+      "loss": 1.7268,
+      "step": 8000
+    },
+    {
+      "epoch": 0.8417674907943188,
+      "grad_norm": 1.9332977533340454,
+      "learning_rate": 1.2849024923896612e-05,
+      "loss": 2.0012,
+      "step": 8001
+    },
+    {
+      "epoch": 0.8418726985796949,
+      "grad_norm": 2.2032244205474854,
+      "learning_rate": 1.2832319186456288e-05,
+      "loss": 1.7438,
+      "step": 8002
+    },
+    {
+      "epoch": 0.841977906365071,
+      "grad_norm": 2.1048672199249268,
+      "learning_rate": 1.281562357151167e-05,
+      "loss": 0.9579,
+      "step": 8003
+    },
+    {
+      "epoch": 0.8420831141504471,
+      "grad_norm": 1.7806977033615112,
+      "learning_rate": 1.2798938081001621e-05,
+      "loss": 1.4379,
+      "step": 8004
+    },
+    {
+      "epoch": 0.8421883219358233,
+      "grad_norm": 2.0925915241241455,
+      "learning_rate": 1.2782262716863747e-05,
+      "loss": 2.3933,
+      "step": 8005
+    },
+    {
+      "epoch": 0.8422935297211994,
+      "grad_norm": 2.210217237472534,
+      "learning_rate": 1.2765597481034475e-05,
+      "loss": 1.8541,
+      "step": 8006
+    },
+    {
+      "epoch": 0.8423987375065755,
+      "grad_norm": 1.8628958463668823,
+      "learning_rate": 1.2748942375449135e-05,
+      "loss": 0.738,
+      "step": 8007
+    },
+    {
+      "epoch": 0.8425039452919516,
+      "grad_norm": 1.774337649345398,
+      "learning_rate": 1.2732297402041793e-05,
+      "loss": 1.3924,
+      "step": 8008
+    },
+    {
+      "epoch": 0.8426091530773278,
+      "grad_norm": 1.3823806047439575,
+      "learning_rate": 1.27156625627454e-05,
+      "loss": 1.7851,
+      "step": 8009
+    },
+    {
+      "epoch": 0.8427143608627038,
+      "grad_norm": 1.3597822189331055,
+      "learning_rate": 1.2699037859491702e-05,
+      "loss": 1.7565,
+      "step": 8010
+    },
+    {
+      "epoch": 0.8428195686480799,
+      "grad_norm": 1.7457836866378784,
+      "learning_rate": 1.2682423294211231e-05,
+      "loss": 2.0374,
+      "step": 8011
+    },
+    {
+      "epoch": 0.8429247764334561,
+      "grad_norm": 1.9241029024124146,
+      "learning_rate": 1.2665818868833445e-05,
+      "loss": 2.1648,
+      "step": 8012
+    },
+    {
+      "epoch": 0.8430299842188322,
+      "grad_norm": 1.4577206373214722,
+      "learning_rate": 1.2649224585286524e-05,
+      "loss": 1.5166,
+      "step": 8013
+    },
+    {
+      "epoch": 0.8431351920042083,
+      "grad_norm": 1.7863467931747437,
+      "learning_rate": 1.263264044549748e-05,
+      "loss": 1.5639,
+      "step": 8014
+    },
+    {
+      "epoch": 0.8432403997895844,
+      "grad_norm": 1.8786143064498901,
+      "learning_rate": 1.2616066451392262e-05,
+      "loss": 2.2911,
+      "step": 8015
+    },
+    {
+      "epoch": 0.8433456075749606,
+      "grad_norm": 1.9232258796691895,
+      "learning_rate": 1.2599502604895475e-05,
+      "loss": 1.5803,
+      "step": 8016
+    },
+    {
+      "epoch": 0.8434508153603366,
+      "grad_norm": 1.5343316793441772,
+      "learning_rate": 1.2582948907930626e-05,
+      "loss": 1.6762,
+      "step": 8017
+    },
+    {
+      "epoch": 0.8435560231457128,
+      "grad_norm": 1.7372281551361084,
+      "learning_rate": 1.2566405362420086e-05,
+      "loss": 1.2532,
+      "step": 8018
+    },
+    {
+      "epoch": 0.8436612309310889,
+      "grad_norm": 1.6992888450622559,
+      "learning_rate": 1.2549871970284954e-05,
+      "loss": 2.2251,
+      "step": 8019
+    },
+    {
+      "epoch": 0.8437664387164651,
+      "grad_norm": 1.0523444414138794,
+      "learning_rate": 1.253334873344525e-05,
+      "loss": 1.7233,
+      "step": 8020
+    },
+    {
+      "epoch": 0.8438716465018411,
+      "grad_norm": 2.2978196144104004,
+      "learning_rate": 1.2516835653819725e-05,
+      "loss": 1.1533,
+      "step": 8021
+    },
+    {
+      "epoch": 0.8439768542872172,
+      "grad_norm": 1.5797593593597412,
+      "learning_rate": 1.2500332733325993e-05,
+      "loss": 1.2688,
+      "step": 8022
+    },
+    {
+      "epoch": 0.8440820620725934,
+      "grad_norm": 2.2443854808807373,
+      "learning_rate": 1.2483839973880508e-05,
+      "loss": 1.2121,
+      "step": 8023
+    },
+    {
+      "epoch": 0.8441872698579694,
+      "grad_norm": 1.3470818996429443,
+      "learning_rate": 1.2467357377398504e-05,
+      "loss": 1.6369,
+      "step": 8024
+    },
+    {
+      "epoch": 0.8442924776433456,
+      "grad_norm": 1.4854938983917236,
+      "learning_rate": 1.2450884945794017e-05,
+      "loss": 1.9805,
+      "step": 8025
+    },
+    {
+      "epoch": 0.8443976854287217,
+      "grad_norm": 1.5202157497406006,
+      "learning_rate": 1.2434422680980006e-05,
+      "loss": 1.9186,
+      "step": 8026
+    },
+    {
+      "epoch": 0.8445028932140979,
+      "grad_norm": 1.5458056926727295,
+      "learning_rate": 1.2417970584868132e-05,
+      "loss": 1.3289,
+      "step": 8027
+    },
+    {
+      "epoch": 0.8446081009994739,
+      "grad_norm": 1.4900844097137451,
+      "learning_rate": 1.2401528659368911e-05,
+      "loss": 1.3999,
+      "step": 8028
+    },
+    {
+      "epoch": 0.8447133087848501,
+      "grad_norm": 1.6544053554534912,
+      "learning_rate": 1.2385096906391746e-05,
+      "loss": 1.7014,
+      "step": 8029
+    },
+    {
+      "epoch": 0.8448185165702262,
+      "grad_norm": 1.7936450242996216,
+      "learning_rate": 1.2368675327844758e-05,
+      "loss": 1.373,
+      "step": 8030
+    },
+    {
+      "epoch": 0.8449237243556024,
+      "grad_norm": 1.2108367681503296,
+      "learning_rate": 1.2352263925634922e-05,
+      "loss": 1.4582,
+      "step": 8031
+    },
+    {
+      "epoch": 0.8450289321409784,
+      "grad_norm": 1.679121732711792,
+      "learning_rate": 1.2335862701668078e-05,
+      "loss": 1.4045,
+      "step": 8032
+    },
+    {
+      "epoch": 0.8451341399263546,
+      "grad_norm": 1.2412477731704712,
+      "learning_rate": 1.2319471657848825e-05,
+      "loss": 1.9746,
+      "step": 8033
+    },
+    {
+      "epoch": 0.8452393477117307,
+      "grad_norm": 1.6706655025482178,
+      "learning_rate": 1.2303090796080585e-05,
+      "loss": 1.6583,
+      "step": 8034
+    },
+    {
+      "epoch": 0.8453445554971067,
+      "grad_norm": 2.606658935546875,
+      "learning_rate": 1.2286720118265659e-05,
+      "loss": 0.7691,
+      "step": 8035
+    },
+    {
+      "epoch": 0.8454497632824829,
+      "grad_norm": 1.2284669876098633,
+      "learning_rate": 1.2270359626305084e-05,
+      "loss": 1.4472,
+      "step": 8036
+    },
+    {
+      "epoch": 0.845554971067859,
+      "grad_norm": 1.9849002361297607,
+      "learning_rate": 1.2254009322098759e-05,
+      "loss": 1.8249,
+      "step": 8037
+    },
+    {
+      "epoch": 0.8456601788532352,
+      "grad_norm": 2.283884286880493,
+      "learning_rate": 1.22376692075454e-05,
+      "loss": 1.3119,
+      "step": 8038
+    },
+    {
+      "epoch": 0.8457653866386112,
+      "grad_norm": 1.697540283203125,
+      "learning_rate": 1.2221339284542488e-05,
+      "loss": 1.223,
+      "step": 8039
+    },
+    {
+      "epoch": 0.8458705944239874,
+      "grad_norm": 2.4393091201782227,
+      "learning_rate": 1.220501955498643e-05,
+      "loss": 1.8292,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8459758022093635,
+      "grad_norm": 1.4089748859405518,
+      "learning_rate": 1.2188710020772343e-05,
+      "loss": 1.5042,
+      "step": 8041
+    },
+    {
+      "epoch": 0.8460810099947396,
+      "grad_norm": 2.5239205360412598,
+      "learning_rate": 1.2172410683794177e-05,
+      "loss": 1.4657,
+      "step": 8042
+    },
+    {
+      "epoch": 0.8461862177801157,
+      "grad_norm": 2.2481822967529297,
+      "learning_rate": 1.2156121545944776e-05,
+      "loss": 1.74,
+      "step": 8043
+    },
+    {
+      "epoch": 0.8462914255654919,
+      "grad_norm": 1.7899489402770996,
+      "learning_rate": 1.2139842609115726e-05,
+      "loss": 1.798,
+      "step": 8044
+    },
+    {
+      "epoch": 0.846396633350868,
+      "grad_norm": 2.1522939205169678,
+      "learning_rate": 1.2123573875197402e-05,
+      "loss": 2.2505,
+      "step": 8045
+    },
+    {
+      "epoch": 0.846501841136244,
+      "grad_norm": 1.7058902978897095,
+      "learning_rate": 1.2107315346079107e-05,
+      "loss": 1.4551,
+      "step": 8046
+    },
+    {
+      "epoch": 0.8466070489216202,
+      "grad_norm": 1.6210920810699463,
+      "learning_rate": 1.209106702364885e-05,
+      "loss": 1.5557,
+      "step": 8047
+    },
+    {
+      "epoch": 0.8467122567069963,
+      "grad_norm": 2.2149577140808105,
+      "learning_rate": 1.2074828909793479e-05,
+      "loss": 1.0683,
+      "step": 8048
+    },
+    {
+      "epoch": 0.8468174644923724,
+      "grad_norm": 1.5407854318618774,
+      "learning_rate": 1.2058601006398718e-05,
+      "loss": 1.813,
+      "step": 8049
+    },
+    {
+      "epoch": 0.8469226722777485,
+      "grad_norm": 1.5162616968154907,
+      "learning_rate": 1.2042383315349037e-05,
+      "loss": 1.1297,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8470278800631247,
+      "grad_norm": 1.622755527496338,
+      "learning_rate": 1.2026175838527732e-05,
+      "loss": 1.7975,
+      "step": 8051
+    },
+    {
+      "epoch": 0.8471330878485008,
+      "grad_norm": 1.260178565979004,
+      "learning_rate": 1.2009978577816949e-05,
+      "loss": 1.3295,
+      "step": 8052
+    },
+    {
+      "epoch": 0.8472382956338769,
+      "grad_norm": 2.4802284240722656,
+      "learning_rate": 1.199379153509761e-05,
+      "loss": 1.7802,
+      "step": 8053
+    },
+    {
+      "epoch": 0.847343503419253,
+      "grad_norm": 1.4476691484451294,
+      "learning_rate": 1.1977614712249441e-05,
+      "loss": 1.2737,
+      "step": 8054
+    },
+    {
+      "epoch": 0.8474487112046292,
+      "grad_norm": 1.3926537036895752,
+      "learning_rate": 1.1961448111151053e-05,
+      "loss": 1.7071,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8475539189900052,
+      "grad_norm": 1.6615149974822998,
+      "learning_rate": 1.1945291733679764e-05,
+      "loss": 1.4381,
+      "step": 8056
+    },
+    {
+      "epoch": 0.8476591267753814,
+      "grad_norm": 1.7258626222610474,
+      "learning_rate": 1.192914558171181e-05,
+      "loss": 1.9516,
+      "step": 8057
+    },
+    {
+      "epoch": 0.8477643345607575,
+      "grad_norm": 2.5182111263275146,
+      "learning_rate": 1.1913009657122188e-05,
+      "loss": 1.4435,
+      "step": 8058
+    },
+    {
+      "epoch": 0.8478695423461337,
+      "grad_norm": 1.2161062955856323,
+      "learning_rate": 1.1896883961784656e-05,
+      "loss": 1.7092,
+      "step": 8059
+    },
+    {
+      "epoch": 0.8479747501315097,
+      "grad_norm": 1.7733458280563354,
+      "learning_rate": 1.1880768497571882e-05,
+      "loss": 1.5346,
+      "step": 8060
+    },
+    {
+      "epoch": 0.8480799579168858,
+      "grad_norm": 1.790793776512146,
+      "learning_rate": 1.1864663266355303e-05,
+      "loss": 1.3009,
+      "step": 8061
+    },
+    {
+      "epoch": 0.848185165702262,
+      "grad_norm": 1.6531035900115967,
+      "learning_rate": 1.1848568270005135e-05,
+      "loss": 1.4765,
+      "step": 8062
+    },
+    {
+      "epoch": 0.8482903734876381,
+      "grad_norm": 1.2282698154449463,
+      "learning_rate": 1.1832483510390469e-05,
+      "loss": 1.9069,
+      "step": 8063
+    },
+    {
+      "epoch": 0.8483955812730142,
+      "grad_norm": 1.9797649383544922,
+      "learning_rate": 1.1816408989379158e-05,
+      "loss": 1.4846,
+      "step": 8064
+    },
+    {
+      "epoch": 0.8485007890583903,
+      "grad_norm": 1.0080591440200806,
+      "learning_rate": 1.1800344708837895e-05,
+      "loss": 1.6405,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8486059968437665,
+      "grad_norm": 1.6383482217788696,
+      "learning_rate": 1.178429067063217e-05,
+      "loss": 1.3591,
+      "step": 8066
+    },
+    {
+      "epoch": 0.8487112046291425,
+      "grad_norm": 1.3764389753341675,
+      "learning_rate": 1.1768246876626265e-05,
+      "loss": 1.0397,
+      "step": 8067
+    },
+    {
+      "epoch": 0.8488164124145187,
+      "grad_norm": 1.7834566831588745,
+      "learning_rate": 1.1752213328683337e-05,
+      "loss": 0.6507,
+      "step": 8068
+    },
+    {
+      "epoch": 0.8489216201998948,
+      "grad_norm": 1.8441131114959717,
+      "learning_rate": 1.1736190028665273e-05,
+      "loss": 1.811,
+      "step": 8069
+    },
+    {
+      "epoch": 0.849026827985271,
+      "grad_norm": 1.3411214351654053,
+      "learning_rate": 1.1720176978432795e-05,
+      "loss": 1.6706,
+      "step": 8070
+    },
+    {
+      "epoch": 0.849132035770647,
+      "grad_norm": 1.3901194334030151,
+      "learning_rate": 1.1704174179845496e-05,
+      "loss": 1.4316,
+      "step": 8071
+    },
+    {
+      "epoch": 0.8492372435560231,
+      "grad_norm": 1.992873191833496,
+      "learning_rate": 1.1688181634761685e-05,
+      "loss": 2.0151,
+      "step": 8072
+    },
+    {
+      "epoch": 0.8493424513413993,
+      "grad_norm": 1.5536729097366333,
+      "learning_rate": 1.1672199345038526e-05,
+      "loss": 1.6268,
+      "step": 8073
+    },
+    {
+      "epoch": 0.8494476591267753,
+      "grad_norm": 1.7125334739685059,
+      "learning_rate": 1.1656227312532009e-05,
+      "loss": 1.4057,
+      "step": 8074
+    },
+    {
+      "epoch": 0.8495528669121515,
+      "grad_norm": 1.3038568496704102,
+      "learning_rate": 1.1640265539096918e-05,
+      "loss": 1.768,
+      "step": 8075
+    },
+    {
+      "epoch": 0.8496580746975276,
+      "grad_norm": 1.536347508430481,
+      "learning_rate": 1.1624314026586802e-05,
+      "loss": 1.634,
+      "step": 8076
+    },
+    {
+      "epoch": 0.8497632824829038,
+      "grad_norm": 1.6387101411819458,
+      "learning_rate": 1.1608372776854103e-05,
+      "loss": 1.7797,
+      "step": 8077
+    },
+    {
+      "epoch": 0.8498684902682798,
+      "grad_norm": 1.4296340942382812,
+      "learning_rate": 1.1592441791750009e-05,
+      "loss": 1.8383,
+      "step": 8078
+    },
+    {
+      "epoch": 0.849973698053656,
+      "grad_norm": 1.9562045335769653,
+      "learning_rate": 1.1576521073124513e-05,
+      "loss": 1.4364,
+      "step": 8079
+    },
+    {
+      "epoch": 0.8500789058390321,
+      "grad_norm": 2.0499372482299805,
+      "learning_rate": 1.1560610622826484e-05,
+      "loss": 1.2246,
+      "step": 8080
+    },
+    {
+      "epoch": 0.8501841136244082,
+      "grad_norm": 2.626941442489624,
+      "learning_rate": 1.1544710442703488e-05,
+      "loss": 1.1103,
+      "step": 8081
+    },
+    {
+      "epoch": 0.8502893214097843,
+      "grad_norm": 1.4844361543655396,
+      "learning_rate": 1.1528820534602002e-05,
+      "loss": 1.3648,
+      "step": 8082
+    },
+    {
+      "epoch": 0.8503945291951605,
+      "grad_norm": 1.9580860137939453,
+      "learning_rate": 1.1512940900367275e-05,
+      "loss": 1.21,
+      "step": 8083
+    },
+    {
+      "epoch": 0.8504997369805366,
+      "grad_norm": 1.9782524108886719,
+      "learning_rate": 1.1497071541843306e-05,
+      "loss": 1.9595,
+      "step": 8084
+    },
+    {
+      "epoch": 0.8506049447659126,
+      "grad_norm": 1.1815834045410156,
+      "learning_rate": 1.1481212460873014e-05,
+      "loss": 1.5414,
+      "step": 8085
+    },
+    {
+      "epoch": 0.8507101525512888,
+      "grad_norm": 1.4871257543563843,
+      "learning_rate": 1.1465363659298023e-05,
+      "loss": 1.876,
+      "step": 8086
+    },
+    {
+      "epoch": 0.8508153603366649,
+      "grad_norm": 1.5685290098190308,
+      "learning_rate": 1.1449525138958805e-05,
+      "loss": 1.6721,
+      "step": 8087
+    },
+    {
+      "epoch": 0.850920568122041,
+      "grad_norm": 1.5825947523117065,
+      "learning_rate": 1.1433696901694658e-05,
+      "loss": 1.4306,
+      "step": 8088
+    },
+    {
+      "epoch": 0.8510257759074171,
+      "grad_norm": 1.341077208518982,
+      "learning_rate": 1.141787894934364e-05,
+      "loss": 1.7303,
+      "step": 8089
+    },
+    {
+      "epoch": 0.8511309836927933,
+      "grad_norm": 1.389611840248108,
+      "learning_rate": 1.1402071283742632e-05,
+      "loss": 1.4725,
+      "step": 8090
+    },
+    {
+      "epoch": 0.8512361914781694,
+      "grad_norm": 2.9490091800689697,
+      "learning_rate": 1.1386273906727363e-05,
+      "loss": 1.549,
+      "step": 8091
+    },
+    {
+      "epoch": 0.8513413992635455,
+      "grad_norm": 1.6228629350662231,
+      "learning_rate": 1.1370486820132308e-05,
+      "loss": 1.2005,
+      "step": 8092
+    },
+    {
+      "epoch": 0.8514466070489216,
+      "grad_norm": 1.9842947721481323,
+      "learning_rate": 1.1354710025790738e-05,
+      "loss": 1.4727,
+      "step": 8093
+    },
+    {
+      "epoch": 0.8515518148342978,
+      "grad_norm": 2.101422071456909,
+      "learning_rate": 1.1338943525534818e-05,
+      "loss": 1.7552,
+      "step": 8094
+    },
+    {
+      "epoch": 0.8516570226196739,
+      "grad_norm": 1.4584681987762451,
+      "learning_rate": 1.1323187321195439e-05,
+      "loss": 1.9696,
+      "step": 8095
+    },
+    {
+      "epoch": 0.85176223040505,
+      "grad_norm": 2.264514446258545,
+      "learning_rate": 1.1307441414602282e-05,
+      "loss": 1.3978,
+      "step": 8096
+    },
+    {
+      "epoch": 0.8518674381904261,
+      "grad_norm": 1.4211277961730957,
+      "learning_rate": 1.1291705807583918e-05,
+      "loss": 1.4239,
+      "step": 8097
+    },
+    {
+      "epoch": 0.8519726459758022,
+      "grad_norm": 2.931905508041382,
+      "learning_rate": 1.1275980501967642e-05,
+      "loss": 1.3678,
+      "step": 8098
+    },
+    {
+      "epoch": 0.8520778537611783,
+      "grad_norm": 1.608646273612976,
+      "learning_rate": 1.126026549957958e-05,
+      "loss": 1.7463,
+      "step": 8099
+    },
+    {
+      "epoch": 0.8521830615465544,
+      "grad_norm": 2.0848610401153564,
+      "learning_rate": 1.1244560802244686e-05,
+      "loss": 1.5804,
+      "step": 8100
+    },
+    {
+      "epoch": 0.8522882693319306,
+      "grad_norm": 1.426628589630127,
+      "learning_rate": 1.1228866411786687e-05,
+      "loss": 1.608,
+      "step": 8101
+    },
+    {
+      "epoch": 0.8523934771173067,
+      "grad_norm": 1.9477870464324951,
+      "learning_rate": 1.1213182330028104e-05,
+      "loss": 1.4227,
+      "step": 8102
+    },
+    {
+      "epoch": 0.8524986849026828,
+      "grad_norm": 1.8772066831588745,
+      "learning_rate": 1.1197508558790304e-05,
+      "loss": 1.3724,
+      "step": 8103
+    },
+    {
+      "epoch": 0.8526038926880589,
+      "grad_norm": 1.8329577445983887,
+      "learning_rate": 1.1181845099893384e-05,
+      "loss": 1.495,
+      "step": 8104
+    },
+    {
+      "epoch": 0.8527091004734351,
+      "grad_norm": 1.718479871749878,
+      "learning_rate": 1.1166191955156346e-05,
+      "loss": 1.5246,
+      "step": 8105
+    },
+    {
+      "epoch": 0.8528143082588111,
+      "grad_norm": 2.2497758865356445,
+      "learning_rate": 1.1150549126396914e-05,
+      "loss": 1.773,
+      "step": 8106
+    },
+    {
+      "epoch": 0.8529195160441873,
+      "grad_norm": 1.177225947380066,
+      "learning_rate": 1.1134916615431611e-05,
+      "loss": 1.7879,
+      "step": 8107
+    },
+    {
+      "epoch": 0.8530247238295634,
+      "grad_norm": 1.9921637773513794,
+      "learning_rate": 1.1119294424075843e-05,
+      "loss": 1.4187,
+      "step": 8108
+    },
+    {
+      "epoch": 0.8531299316149396,
+      "grad_norm": 2.8252499103546143,
+      "learning_rate": 1.1103682554143736e-05,
+      "loss": 1.5446,
+      "step": 8109
+    },
+    {
+      "epoch": 0.8532351394003156,
+      "grad_norm": 1.615925669670105,
+      "learning_rate": 1.1088081007448214e-05,
+      "loss": 1.9703,
+      "step": 8110
+    },
+    {
+      "epoch": 0.8533403471856917,
+      "grad_norm": 1.6888537406921387,
+      "learning_rate": 1.10724897858011e-05,
+      "loss": 1.6726,
+      "step": 8111
+    },
+    {
+      "epoch": 0.8534455549710679,
+      "grad_norm": 1.7525259256362915,
+      "learning_rate": 1.1056908891012884e-05,
+      "loss": 1.2526,
+      "step": 8112
+    },
+    {
+      "epoch": 0.8535507627564439,
+      "grad_norm": 1.5043063163757324,
+      "learning_rate": 1.104133832489298e-05,
+      "loss": 1.8318,
+      "step": 8113
+    },
+    {
+      "epoch": 0.8536559705418201,
+      "grad_norm": 1.4781932830810547,
+      "learning_rate": 1.1025778089249527e-05,
+      "loss": 1.7238,
+      "step": 8114
+    },
+    {
+      "epoch": 0.8537611783271962,
+      "grad_norm": 1.9893333911895752,
+      "learning_rate": 1.1010228185889449e-05,
+      "loss": 1.5819,
+      "step": 8115
+    },
+    {
+      "epoch": 0.8538663861125724,
+      "grad_norm": 2.0442187786102295,
+      "learning_rate": 1.0994688616618565e-05,
+      "loss": 1.9624,
+      "step": 8116
+    },
+    {
+      "epoch": 0.8539715938979484,
+      "grad_norm": 1.369695782661438,
+      "learning_rate": 1.09791593832414e-05,
+      "loss": 2.0449,
+      "step": 8117
+    },
+    {
+      "epoch": 0.8540768016833246,
+      "grad_norm": 2.589287042617798,
+      "learning_rate": 1.09636404875613e-05,
+      "loss": 1.9639,
+      "step": 8118
+    },
+    {
+      "epoch": 0.8541820094687007,
+      "grad_norm": 1.5649428367614746,
+      "learning_rate": 1.0948131931380457e-05,
+      "loss": 1.2722,
+      "step": 8119
+    },
+    {
+      "epoch": 0.8542872172540767,
+      "grad_norm": 1.4404014348983765,
+      "learning_rate": 1.0932633716499818e-05,
+      "loss": 1.7971,
+      "step": 8120
+    },
+    {
+      "epoch": 0.8543924250394529,
+      "grad_norm": 1.283900260925293,
+      "learning_rate": 1.0917145844719101e-05,
+      "loss": 1.5276,
+      "step": 8121
+    },
+    {
+      "epoch": 0.854497632824829,
+      "grad_norm": 1.41376531124115,
+      "learning_rate": 1.0901668317836933e-05,
+      "loss": 1.4866,
+      "step": 8122
+    },
+    {
+      "epoch": 0.8546028406102052,
+      "grad_norm": 1.4160473346710205,
+      "learning_rate": 1.088620113765061e-05,
+      "loss": 1.8107,
+      "step": 8123
+    },
+    {
+      "epoch": 0.8547080483955812,
+      "grad_norm": 1.5706056356430054,
+      "learning_rate": 1.0870744305956315e-05,
+      "loss": 1.4278,
+      "step": 8124
+    },
+    {
+      "epoch": 0.8548132561809574,
+      "grad_norm": 1.7656512260437012,
+      "learning_rate": 1.0855297824548982e-05,
+      "loss": 1.6278,
+      "step": 8125
+    },
+    {
+      "epoch": 0.8549184639663335,
+      "grad_norm": 1.3173682689666748,
+      "learning_rate": 1.0839861695222354e-05,
+      "loss": 1.8172,
+      "step": 8126
+    },
+    {
+      "epoch": 0.8550236717517097,
+      "grad_norm": 1.5881686210632324,
+      "learning_rate": 1.0824435919769005e-05,
+      "loss": 2.1094,
+      "step": 8127
+    },
+    {
+      "epoch": 0.8551288795370857,
+      "grad_norm": 2.1221399307250977,
+      "learning_rate": 1.080902049998026e-05,
+      "loss": 1.5415,
+      "step": 8128
+    },
+    {
+      "epoch": 0.8552340873224619,
+      "grad_norm": 2.114959955215454,
+      "learning_rate": 1.0793615437646254e-05,
+      "loss": 0.6612,
+      "step": 8129
+    },
+    {
+      "epoch": 0.855339295107838,
+      "grad_norm": 1.5936521291732788,
+      "learning_rate": 1.0778220734555955e-05,
+      "loss": 1.7225,
+      "step": 8130
+    },
+    {
+      "epoch": 0.855444502893214,
+      "grad_norm": 1.4308537244796753,
+      "learning_rate": 1.0762836392497078e-05,
+      "loss": 1.5506,
+      "step": 8131
+    },
+    {
+      "epoch": 0.8555497106785902,
+      "grad_norm": 1.9329140186309814,
+      "learning_rate": 1.0747462413256148e-05,
+      "loss": 1.2109,
+      "step": 8132
+    },
+    {
+      "epoch": 0.8556549184639664,
+      "grad_norm": 1.7347162961959839,
+      "learning_rate": 1.0732098798618517e-05,
+      "loss": 1.8518,
+      "step": 8133
+    },
+    {
+      "epoch": 0.8557601262493425,
+      "grad_norm": 1.4052585363388062,
+      "learning_rate": 1.07167455503683e-05,
+      "loss": 1.7604,
+      "step": 8134
+    },
+    {
+      "epoch": 0.8558653340347185,
+      "grad_norm": 2.470710039138794,
+      "learning_rate": 1.0701402670288407e-05,
+      "loss": 1.5943,
+      "step": 8135
+    },
+    {
+      "epoch": 0.8559705418200947,
+      "grad_norm": 1.489478349685669,
+      "learning_rate": 1.0686070160160588e-05,
+      "loss": 1.3159,
+      "step": 8136
+    },
+    {
+      "epoch": 0.8560757496054708,
+      "grad_norm": 1.2157397270202637,
+      "learning_rate": 1.067074802176533e-05,
+      "loss": 1.3678,
+      "step": 8137
+    },
+    {
+      "epoch": 0.8561809573908469,
+      "grad_norm": 1.8080435991287231,
+      "learning_rate": 1.0655436256881935e-05,
+      "loss": 1.366,
+      "step": 8138
+    },
+    {
+      "epoch": 0.856286165176223,
+      "grad_norm": 1.8651374578475952,
+      "learning_rate": 1.0640134867288542e-05,
+      "loss": 1.5744,
+      "step": 8139
+    },
+    {
+      "epoch": 0.8563913729615992,
+      "grad_norm": 1.4366014003753662,
+      "learning_rate": 1.0624843854762034e-05,
+      "loss": 1.7384,
+      "step": 8140
+    },
+    {
+      "epoch": 0.8564965807469753,
+      "grad_norm": 1.5148162841796875,
+      "learning_rate": 1.0609563221078079e-05,
+      "loss": 1.2506,
+      "step": 8141
+    },
+    {
+      "epoch": 0.8566017885323514,
+      "grad_norm": 1.3511996269226074,
+      "learning_rate": 1.059429296801121e-05,
+      "loss": 1.7989,
+      "step": 8142
+    },
+    {
+      "epoch": 0.8567069963177275,
+      "grad_norm": 1.3725030422210693,
+      "learning_rate": 1.05790330973347e-05,
+      "loss": 1.2092,
+      "step": 8143
+    },
+    {
+      "epoch": 0.8568122041031037,
+      "grad_norm": 1.480331301689148,
+      "learning_rate": 1.056378361082062e-05,
+      "loss": 1.7247,
+      "step": 8144
+    },
+    {
+      "epoch": 0.8569174118884797,
+      "grad_norm": 1.1043062210083008,
+      "learning_rate": 1.0548544510239833e-05,
+      "loss": 2.1902,
+      "step": 8145
+    },
+    {
+      "epoch": 0.8570226196738558,
+      "grad_norm": 1.8570284843444824,
+      "learning_rate": 1.053331579736201e-05,
+      "loss": 1.2777,
+      "step": 8146
+    },
+    {
+      "epoch": 0.857127827459232,
+      "grad_norm": 1.4823728799819946,
+      "learning_rate": 1.0518097473955624e-05,
+      "loss": 1.4131,
+      "step": 8147
+    },
+    {
+      "epoch": 0.8572330352446081,
+      "grad_norm": 1.5833871364593506,
+      "learning_rate": 1.0502889541787918e-05,
+      "loss": 1.4025,
+      "step": 8148
+    },
+    {
+      "epoch": 0.8573382430299842,
+      "grad_norm": 1.5487840175628662,
+      "learning_rate": 1.0487692002624937e-05,
+      "loss": 2.1141,
+      "step": 8149
+    },
+    {
+      "epoch": 0.8574434508153603,
+      "grad_norm": 2.5426361560821533,
+      "learning_rate": 1.0472504858231535e-05,
+      "loss": 1.4438,
+      "step": 8150
+    },
+    {
+      "epoch": 0.8575486586007365,
+      "grad_norm": 1.742750644683838,
+      "learning_rate": 1.0457328110371345e-05,
+      "loss": 1.3274,
+      "step": 8151
+    },
+    {
+      "epoch": 0.8576538663861125,
+      "grad_norm": 1.529963731765747,
+      "learning_rate": 1.0442161760806756e-05,
+      "loss": 1.545,
+      "step": 8152
+    },
+    {
+      "epoch": 0.8577590741714887,
+      "grad_norm": 1.5266475677490234,
+      "learning_rate": 1.042700581129904e-05,
+      "loss": 1.4656,
+      "step": 8153
+    },
+    {
+      "epoch": 0.8578642819568648,
+      "grad_norm": 2.2451913356781006,
+      "learning_rate": 1.0411860263608186e-05,
+      "loss": 1.5722,
+      "step": 8154
+    },
+    {
+      "epoch": 0.857969489742241,
+      "grad_norm": 1.2465153932571411,
+      "learning_rate": 1.0396725119492967e-05,
+      "loss": 1.718,
+      "step": 8155
+    },
+    {
+      "epoch": 0.858074697527617,
+      "grad_norm": 1.9138121604919434,
+      "learning_rate": 1.038160038071102e-05,
+      "loss": 1.7302,
+      "step": 8156
+    },
+    {
+      "epoch": 0.8581799053129932,
+      "grad_norm": 1.7178139686584473,
+      "learning_rate": 1.036648604901871e-05,
+      "loss": 2.008,
+      "step": 8157
+    },
+    {
+      "epoch": 0.8582851130983693,
+      "grad_norm": 1.064705729484558,
+      "learning_rate": 1.0351382126171227e-05,
+      "loss": 1.4776,
+      "step": 8158
+    },
+    {
+      "epoch": 0.8583903208837455,
+      "grad_norm": 2.823317766189575,
+      "learning_rate": 1.033628861392253e-05,
+      "loss": 1.6744,
+      "step": 8159
+    },
+    {
+      "epoch": 0.8584955286691215,
+      "grad_norm": 1.7388535737991333,
+      "learning_rate": 1.0321205514025357e-05,
+      "loss": 1.3214,
+      "step": 8160
+    },
+    {
+      "epoch": 0.8586007364544976,
+      "grad_norm": 2.498295783996582,
+      "learning_rate": 1.0306132828231318e-05,
+      "loss": 1.5151,
+      "step": 8161
+    },
+    {
+      "epoch": 0.8587059442398738,
+      "grad_norm": 2.1892495155334473,
+      "learning_rate": 1.0291070558290705e-05,
+      "loss": 1.1706,
+      "step": 8162
+    },
+    {
+      "epoch": 0.8588111520252498,
+      "grad_norm": 1.274142861366272,
+      "learning_rate": 1.027601870595265e-05,
+      "loss": 1.5027,
+      "step": 8163
+    },
+    {
+      "epoch": 0.858916359810626,
+      "grad_norm": 1.8496899604797363,
+      "learning_rate": 1.02609772729651e-05,
+      "loss": 1.3466,
+      "step": 8164
+    },
+    {
+      "epoch": 0.8590215675960021,
+      "grad_norm": 1.2990412712097168,
+      "learning_rate": 1.0245946261074769e-05,
+      "loss": 1.7513,
+      "step": 8165
+    },
+    {
+      "epoch": 0.8591267753813783,
+      "grad_norm": 1.4909148216247559,
+      "learning_rate": 1.0230925672027137e-05,
+      "loss": 1.5437,
+      "step": 8166
+    },
+    {
+      "epoch": 0.8592319831667543,
+      "grad_norm": 1.508446455001831,
+      "learning_rate": 1.0215915507566499e-05,
+      "loss": 1.4974,
+      "step": 8167
+    },
+    {
+      "epoch": 0.8593371909521305,
+      "grad_norm": 1.729293942451477,
+      "learning_rate": 1.0200915769435937e-05,
+      "loss": 1.8774,
+      "step": 8168
+    },
+    {
+      "epoch": 0.8594423987375066,
+      "grad_norm": 2.2102980613708496,
+      "learning_rate": 1.0185926459377326e-05,
+      "loss": 0.6269,
+      "step": 8169
+    },
+    {
+      "epoch": 0.8595476065228826,
+      "grad_norm": 1.7405325174331665,
+      "learning_rate": 1.017094757913134e-05,
+      "loss": 1.7546,
+      "step": 8170
+    },
+    {
+      "epoch": 0.8596528143082588,
+      "grad_norm": 1.6175463199615479,
+      "learning_rate": 1.0155979130437387e-05,
+      "loss": 2.0148,
+      "step": 8171
+    },
+    {
+      "epoch": 0.8597580220936349,
+      "grad_norm": 1.628763198852539,
+      "learning_rate": 1.0141021115033745e-05,
+      "loss": 1.0579,
+      "step": 8172
+    },
+    {
+      "epoch": 0.8598632298790111,
+      "grad_norm": 1.998671293258667,
+      "learning_rate": 1.012607353465742e-05,
+      "loss": 1.5527,
+      "step": 8173
+    },
+    {
+      "epoch": 0.8599684376643871,
+      "grad_norm": 2.1311984062194824,
+      "learning_rate": 1.0111136391044218e-05,
+      "loss": 1.4062,
+      "step": 8174
+    },
+    {
+      "epoch": 0.8600736454497633,
+      "grad_norm": 2.120063304901123,
+      "learning_rate": 1.009620968592876e-05,
+      "loss": 1.3496,
+      "step": 8175
+    },
+    {
+      "epoch": 0.8601788532351394,
+      "grad_norm": 1.719994306564331,
+      "learning_rate": 1.0081293421044435e-05,
+      "loss": 1.2293,
+      "step": 8176
+    },
+    {
+      "epoch": 0.8602840610205155,
+      "grad_norm": 1.8668345212936401,
+      "learning_rate": 1.0066387598123383e-05,
+      "loss": 1.7347,
+      "step": 8177
+    },
+    {
+      "epoch": 0.8603892688058916,
+      "grad_norm": 1.9415462017059326,
+      "learning_rate": 1.0051492218896619e-05,
+      "loss": 1.2794,
+      "step": 8178
+    },
+    {
+      "epoch": 0.8604944765912678,
+      "grad_norm": 2.177868604660034,
+      "learning_rate": 1.0036607285093857e-05,
+      "loss": 1.6339,
+      "step": 8179
+    },
+    {
+      "epoch": 0.8605996843766439,
+      "grad_norm": 1.627882719039917,
+      "learning_rate": 1.002173279844364e-05,
+      "loss": 1.5512,
+      "step": 8180
+    },
+    {
+      "epoch": 0.86070489216202,
+      "grad_norm": 1.8268787860870361,
+      "learning_rate": 1.0006868760673327e-05,
+      "loss": 2.4057,
+      "step": 8181
+    },
+    {
+      "epoch": 0.8608100999473961,
+      "grad_norm": 1.1629137992858887,
+      "learning_rate": 9.992015173508995e-06,
+      "loss": 1.5108,
+      "step": 8182
+    },
+    {
+      "epoch": 0.8609153077327723,
+      "grad_norm": 1.567643404006958,
+      "learning_rate": 9.977172038675531e-06,
+      "loss": 1.5915,
+      "step": 8183
+    },
+    {
+      "epoch": 0.8610205155181483,
+      "grad_norm": 1.6565831899642944,
+      "learning_rate": 9.962339357896666e-06,
+      "loss": 1.6012,
+      "step": 8184
+    },
+    {
+      "epoch": 0.8611257233035244,
+      "grad_norm": 1.5608383417129517,
+      "learning_rate": 9.947517132894835e-06,
+      "loss": 1.5782,
+      "step": 8185
+    },
+    {
+      "epoch": 0.8612309310889006,
+      "grad_norm": 1.3910160064697266,
+      "learning_rate": 9.932705365391293e-06,
+      "loss": 1.7121,
+      "step": 8186
+    },
+    {
+      "epoch": 0.8613361388742767,
+      "grad_norm": 1.9219480752944946,
+      "learning_rate": 9.91790405710613e-06,
+      "loss": 1.6089,
+      "step": 8187
+    },
+    {
+      "epoch": 0.8614413466596528,
+      "grad_norm": 1.804298996925354,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 1.7155,
+      "step": 8188
+    },
+    {
+      "epoch": 0.8615465544450289,
+      "grad_norm": 0.9984854459762573,
+      "learning_rate": 9.88833282506486e-06,
+      "loss": 1.7646,
+      "step": 8189
+    },
+    {
+      "epoch": 0.8616517622304051,
+      "grad_norm": 1.9518897533416748,
+      "learning_rate": 9.873562904742805e-06,
+      "loss": 1.1884,
+      "step": 8190
+    },
+    {
+      "epoch": 0.8617569700157812,
+      "grad_norm": 1.4349753856658936,
+      "learning_rate": 9.858803450507081e-06,
+      "loss": 1.8053,
+      "step": 8191
+    },
+    {
+      "epoch": 0.8618621778011573,
+      "grad_norm": 1.8369276523590088,
+      "learning_rate": 9.844054464071717e-06,
+      "loss": 1.9263,
+      "step": 8192
+    },
+    {
+      "epoch": 0.8619673855865334,
+      "grad_norm": 1.3132811784744263,
+      "learning_rate": 9.829315947149431e-06,
+      "loss": 1.5865,
+      "step": 8193
+    },
+    {
+      "epoch": 0.8620725933719096,
+      "grad_norm": 2.2313907146453857,
+      "learning_rate": 9.814587901451733e-06,
+      "loss": 0.9993,
+      "step": 8194
+    },
+    {
+      "epoch": 0.8621778011572856,
+      "grad_norm": 1.455952525138855,
+      "learning_rate": 9.799870328688988e-06,
+      "loss": 1.4218,
+      "step": 8195
+    },
+    {
+      "epoch": 0.8622830089426617,
+      "grad_norm": 1.646437644958496,
+      "learning_rate": 9.785163230570282e-06,
+      "loss": 1.7,
+      "step": 8196
+    },
+    {
+      "epoch": 0.8623882167280379,
+      "grad_norm": 1.9158176183700562,
+      "learning_rate": 9.770466608803475e-06,
+      "loss": 2.0093,
+      "step": 8197
+    },
+    {
+      "epoch": 0.862493424513414,
+      "grad_norm": 1.696004033088684,
+      "learning_rate": 9.755780465095287e-06,
+      "loss": 1.4588,
+      "step": 8198
+    },
+    {
+      "epoch": 0.8625986322987901,
+      "grad_norm": 1.7793548107147217,
+      "learning_rate": 9.741104801151146e-06,
+      "loss": 1.8082,
+      "step": 8199
+    },
+    {
+      "epoch": 0.8627038400841662,
+      "grad_norm": 2.2487435340881348,
+      "learning_rate": 9.726439618675276e-06,
+      "loss": 1.5888,
+      "step": 8200
+    },
+    {
+      "epoch": 0.8628090478695424,
+      "grad_norm": 1.4217240810394287,
+      "learning_rate": 9.711784919370715e-06,
+      "loss": 1.4924,
+      "step": 8201
+    },
+    {
+      "epoch": 0.8629142556549184,
+      "grad_norm": 1.571770191192627,
+      "learning_rate": 9.697140704939245e-06,
+      "loss": 1.2066,
+      "step": 8202
+    },
+    {
+      "epoch": 0.8630194634402946,
+      "grad_norm": 1.2527103424072266,
+      "learning_rate": 9.682506977081496e-06,
+      "loss": 1.5845,
+      "step": 8203
+    },
+    {
+      "epoch": 0.8631246712256707,
+      "grad_norm": 1.499109148979187,
+      "learning_rate": 9.667883737496786e-06,
+      "loss": 0.9014,
+      "step": 8204
+    },
+    {
+      "epoch": 0.8632298790110469,
+      "grad_norm": 2.3821563720703125,
+      "learning_rate": 9.653270987883267e-06,
+      "loss": 1.6248,
+      "step": 8205
+    },
+    {
+      "epoch": 0.8633350867964229,
+      "grad_norm": 1.5363044738769531,
+      "learning_rate": 9.638668729937905e-06,
+      "loss": 1.9727,
+      "step": 8206
+    },
+    {
+      "epoch": 0.863440294581799,
+      "grad_norm": 1.1496707201004028,
+      "learning_rate": 9.624076965356388e-06,
+      "loss": 1.8011,
+      "step": 8207
+    },
+    {
+      "epoch": 0.8635455023671752,
+      "grad_norm": 1.7943934202194214,
+      "learning_rate": 9.609495695833216e-06,
+      "loss": 1.4859,
+      "step": 8208
+    },
+    {
+      "epoch": 0.8636507101525512,
+      "grad_norm": 2.865455389022827,
+      "learning_rate": 9.594924923061655e-06,
+      "loss": 1.6245,
+      "step": 8209
+    },
+    {
+      "epoch": 0.8637559179379274,
+      "grad_norm": 1.4639524221420288,
+      "learning_rate": 9.580364648733775e-06,
+      "loss": 1.8505,
+      "step": 8210
+    },
+    {
+      "epoch": 0.8638611257233035,
+      "grad_norm": 1.4049338102340698,
+      "learning_rate": 9.56581487454038e-06,
+      "loss": 1.5261,
+      "step": 8211
+    },
+    {
+      "epoch": 0.8639663335086797,
+      "grad_norm": 1.3107516765594482,
+      "learning_rate": 9.551275602171127e-06,
+      "loss": 1.493,
+      "step": 8212
+    },
+    {
+      "epoch": 0.8640715412940557,
+      "grad_norm": 1.295406460762024,
+      "learning_rate": 9.53674683331438e-06,
+      "loss": 1.3942,
+      "step": 8213
+    },
+    {
+      "epoch": 0.8641767490794319,
+      "grad_norm": 1.5618852376937866,
+      "learning_rate": 9.522228569657343e-06,
+      "loss": 2.2107,
+      "step": 8214
+    },
+    {
+      "epoch": 0.864281956864808,
+      "grad_norm": 1.705565094947815,
+      "learning_rate": 9.507720812885978e-06,
+      "loss": 1.8339,
+      "step": 8215
+    },
+    {
+      "epoch": 0.8643871646501841,
+      "grad_norm": 2.110238552093506,
+      "learning_rate": 9.493223564684994e-06,
+      "loss": 1.6849,
+      "step": 8216
+    },
+    {
+      "epoch": 0.8644923724355602,
+      "grad_norm": 1.721220850944519,
+      "learning_rate": 9.478736826737944e-06,
+      "loss": 1.3985,
+      "step": 8217
+    },
+    {
+      "epoch": 0.8645975802209364,
+      "grad_norm": 1.481402039527893,
+      "learning_rate": 9.464260600727104e-06,
+      "loss": 1.7841,
+      "step": 8218
+    },
+    {
+      "epoch": 0.8647027880063125,
+      "grad_norm": 2.818570852279663,
+      "learning_rate": 9.44979488833353e-06,
+      "loss": 1.3507,
+      "step": 8219
+    },
+    {
+      "epoch": 0.8648079957916885,
+      "grad_norm": 1.7059146165847778,
+      "learning_rate": 9.435339691237121e-06,
+      "loss": 1.4943,
+      "step": 8220
+    },
+    {
+      "epoch": 0.8649132035770647,
+      "grad_norm": 1.5640789270401,
+      "learning_rate": 9.420895011116492e-06,
+      "loss": 1.7849,
+      "step": 8221
+    },
+    {
+      "epoch": 0.8650184113624408,
+      "grad_norm": 2.26628041267395,
+      "learning_rate": 9.406460849649045e-06,
+      "loss": 1.6047,
+      "step": 8222
+    },
+    {
+      "epoch": 0.865123619147817,
+      "grad_norm": 1.201790452003479,
+      "learning_rate": 9.392037208510996e-06,
+      "loss": 1.8968,
+      "step": 8223
+    },
+    {
+      "epoch": 0.865228826933193,
+      "grad_norm": 1.8288311958312988,
+      "learning_rate": 9.37762408937729e-06,
+      "loss": 0.9009,
+      "step": 8224
+    },
+    {
+      "epoch": 0.8653340347185692,
+      "grad_norm": 1.5502636432647705,
+      "learning_rate": 9.36322149392168e-06,
+      "loss": 1.1525,
+      "step": 8225
+    },
+    {
+      "epoch": 0.8654392425039453,
+      "grad_norm": 1.8618223667144775,
+      "learning_rate": 9.348829423816718e-06,
+      "loss": 1.6285,
+      "step": 8226
+    },
+    {
+      "epoch": 0.8655444502893214,
+      "grad_norm": 1.6962296962738037,
+      "learning_rate": 9.334447880733676e-06,
+      "loss": 2.1117,
+      "step": 8227
+    },
+    {
+      "epoch": 0.8656496580746975,
+      "grad_norm": 1.1905288696289062,
+      "learning_rate": 9.320076866342642e-06,
+      "loss": 1.6631,
+      "step": 8228
+    },
+    {
+      "epoch": 0.8657548658600737,
+      "grad_norm": 1.2927476167678833,
+      "learning_rate": 9.30571638231249e-06,
+      "loss": 2.1606,
+      "step": 8229
+    },
+    {
+      "epoch": 0.8658600736454498,
+      "grad_norm": 1.4459666013717651,
+      "learning_rate": 9.291366430310844e-06,
+      "loss": 1.3559,
+      "step": 8230
+    },
+    {
+      "epoch": 0.8659652814308259,
+      "grad_norm": 1.827365517616272,
+      "learning_rate": 9.277027012004125e-06,
+      "loss": 1.6095,
+      "step": 8231
+    },
+    {
+      "epoch": 0.866070489216202,
+      "grad_norm": 1.9253352880477905,
+      "learning_rate": 9.262698129057512e-06,
+      "loss": 1.2613,
+      "step": 8232
+    },
+    {
+      "epoch": 0.8661756970015781,
+      "grad_norm": 1.1855809688568115,
+      "learning_rate": 9.248379783134952e-06,
+      "loss": 1.8954,
+      "step": 8233
+    },
+    {
+      "epoch": 0.8662809047869542,
+      "grad_norm": 2.3653783798217773,
+      "learning_rate": 9.234071975899228e-06,
+      "loss": 1.2972,
+      "step": 8234
+    },
+    {
+      "epoch": 0.8663861125723303,
+      "grad_norm": 1.508872389793396,
+      "learning_rate": 9.21977470901184e-06,
+      "loss": 1.6728,
+      "step": 8235
+    },
+    {
+      "epoch": 0.8664913203577065,
+      "grad_norm": 1.531275749206543,
+      "learning_rate": 9.205487984133076e-06,
+      "loss": 1.2766,
+      "step": 8236
+    },
+    {
+      "epoch": 0.8665965281430826,
+      "grad_norm": 1.4290529489517212,
+      "learning_rate": 9.191211802922017e-06,
+      "loss": 1.7471,
+      "step": 8237
+    },
+    {
+      "epoch": 0.8667017359284587,
+      "grad_norm": 1.9806660413742065,
+      "learning_rate": 9.176946167036516e-06,
+      "loss": 1.0977,
+      "step": 8238
+    },
+    {
+      "epoch": 0.8668069437138348,
+      "grad_norm": 1.909472107887268,
+      "learning_rate": 9.162691078133157e-06,
+      "loss": 1.8081,
+      "step": 8239
+    },
+    {
+      "epoch": 0.866912151499211,
+      "grad_norm": 2.2242584228515625,
+      "learning_rate": 9.148446537867383e-06,
+      "loss": 1.6033,
+      "step": 8240
+    },
+    {
+      "epoch": 0.867017359284587,
+      "grad_norm": 1.3341119289398193,
+      "learning_rate": 9.134212547893351e-06,
+      "loss": 1.9168,
+      "step": 8241
+    },
+    {
+      "epoch": 0.8671225670699632,
+      "grad_norm": 1.836234211921692,
+      "learning_rate": 9.11998910986398e-06,
+      "loss": 2.2142,
+      "step": 8242
+    },
+    {
+      "epoch": 0.8672277748553393,
+      "grad_norm": 1.667237401008606,
+      "learning_rate": 9.105776225431029e-06,
+      "loss": 1.3457,
+      "step": 8243
+    },
+    {
+      "epoch": 0.8673329826407155,
+      "grad_norm": 1.6611175537109375,
+      "learning_rate": 9.091573896244976e-06,
+      "loss": 1.3769,
+      "step": 8244
+    },
+    {
+      "epoch": 0.8674381904260915,
+      "grad_norm": 1.894445776939392,
+      "learning_rate": 9.07738212395508e-06,
+      "loss": 1.6441,
+      "step": 8245
+    },
+    {
+      "epoch": 0.8675433982114676,
+      "grad_norm": 1.9377069473266602,
+      "learning_rate": 9.063200910209413e-06,
+      "loss": 1.4411,
+      "step": 8246
+    },
+    {
+      "epoch": 0.8676486059968438,
+      "grad_norm": 1.5607261657714844,
+      "learning_rate": 9.049030256654777e-06,
+      "loss": 1.5803,
+      "step": 8247
+    },
+    {
+      "epoch": 0.8677538137822198,
+      "grad_norm": 1.2471551895141602,
+      "learning_rate": 9.034870164936737e-06,
+      "loss": 1.4342,
+      "step": 8248
+    },
+    {
+      "epoch": 0.867859021567596,
+      "grad_norm": 1.3888238668441772,
+      "learning_rate": 9.020720636699709e-06,
+      "loss": 1.5969,
+      "step": 8249
+    },
+    {
+      "epoch": 0.8679642293529721,
+      "grad_norm": 3.419867753982544,
+      "learning_rate": 9.006581673586789e-06,
+      "loss": 1.5199,
+      "step": 8250
+    },
+    {
+      "epoch": 0.8680694371383483,
+      "grad_norm": 1.7983578443527222,
+      "learning_rate": 8.992453277239942e-06,
+      "loss": 1.5565,
+      "step": 8251
+    },
+    {
+      "epoch": 0.8681746449237243,
+      "grad_norm": 1.5577338933944702,
+      "learning_rate": 8.978335449299791e-06,
+      "loss": 1.5991,
+      "step": 8252
+    },
+    {
+      "epoch": 0.8682798527091005,
+      "grad_norm": 2.433664321899414,
+      "learning_rate": 8.9642281914058e-06,
+      "loss": 1.3975,
+      "step": 8253
+    },
+    {
+      "epoch": 0.8683850604944766,
+      "grad_norm": 1.6767748594284058,
+      "learning_rate": 8.950131505196236e-06,
+      "loss": 1.9497,
+      "step": 8254
+    },
+    {
+      "epoch": 0.8684902682798528,
+      "grad_norm": 1.431541919708252,
+      "learning_rate": 8.936045392308079e-06,
+      "loss": 2.1185,
+      "step": 8255
+    },
+    {
+      "epoch": 0.8685954760652288,
+      "grad_norm": 2.1313490867614746,
+      "learning_rate": 8.921969854377088e-06,
+      "loss": 1.7141,
+      "step": 8256
+    },
+    {
+      "epoch": 0.868700683850605,
+      "grad_norm": 1.8773490190505981,
+      "learning_rate": 8.907904893037833e-06,
+      "loss": 1.592,
+      "step": 8257
+    },
+    {
+      "epoch": 0.8688058916359811,
+      "grad_norm": 1.4465880393981934,
+      "learning_rate": 8.893850509923619e-06,
+      "loss": 1.3952,
+      "step": 8258
+    },
+    {
+      "epoch": 0.8689110994213571,
+      "grad_norm": 1.6788699626922607,
+      "learning_rate": 8.87980670666655e-06,
+      "loss": 1.2683,
+      "step": 8259
+    },
+    {
+      "epoch": 0.8690163072067333,
+      "grad_norm": 1.4873356819152832,
+      "learning_rate": 8.865773484897477e-06,
+      "loss": 2.0945,
+      "step": 8260
+    },
+    {
+      "epoch": 0.8691215149921094,
+      "grad_norm": 1.6656405925750732,
+      "learning_rate": 8.85175084624602e-06,
+      "loss": 1.4234,
+      "step": 8261
+    },
+    {
+      "epoch": 0.8692267227774856,
+      "grad_norm": 1.3971835374832153,
+      "learning_rate": 8.8377387923406e-06,
+      "loss": 1.4474,
+      "step": 8262
+    },
+    {
+      "epoch": 0.8693319305628616,
+      "grad_norm": 1.494502305984497,
+      "learning_rate": 8.82373732480839e-06,
+      "loss": 1.8342,
+      "step": 8263
+    },
+    {
+      "epoch": 0.8694371383482378,
+      "grad_norm": 1.7074618339538574,
+      "learning_rate": 8.809746445275312e-06,
+      "loss": 1.4596,
+      "step": 8264
+    },
+    {
+      "epoch": 0.8695423461336139,
+      "grad_norm": 2.2928879261016846,
+      "learning_rate": 8.795766155366114e-06,
+      "loss": 1.4602,
+      "step": 8265
+    },
+    {
+      "epoch": 0.86964755391899,
+      "grad_norm": 1.632634162902832,
+      "learning_rate": 8.781796456704262e-06,
+      "loss": 1.6407,
+      "step": 8266
+    },
+    {
+      "epoch": 0.8697527617043661,
+      "grad_norm": 1.7646962404251099,
+      "learning_rate": 8.767837350912e-06,
+      "loss": 1.6187,
+      "step": 8267
+    },
+    {
+      "epoch": 0.8698579694897423,
+      "grad_norm": 1.2554585933685303,
+      "learning_rate": 8.75388883961038e-06,
+      "loss": 1.3565,
+      "step": 8268
+    },
+    {
+      "epoch": 0.8699631772751184,
+      "grad_norm": 2.188851833343506,
+      "learning_rate": 8.739950924419183e-06,
+      "loss": 1.4143,
+      "step": 8269
+    },
+    {
+      "epoch": 0.8700683850604944,
+      "grad_norm": 1.2379424571990967,
+      "learning_rate": 8.726023606956956e-06,
+      "loss": 2.0142,
+      "step": 8270
+    },
+    {
+      "epoch": 0.8701735928458706,
+      "grad_norm": 1.5064557790756226,
+      "learning_rate": 8.712106888841064e-06,
+      "loss": 1.8644,
+      "step": 8271
+    },
+    {
+      "epoch": 0.8702788006312467,
+      "grad_norm": 2.4556884765625,
+      "learning_rate": 8.698200771687592e-06,
+      "loss": 1.1037,
+      "step": 8272
+    },
+    {
+      "epoch": 0.8703840084166228,
+      "grad_norm": 0.9739543199539185,
+      "learning_rate": 8.684305257111425e-06,
+      "loss": 1.5688,
+      "step": 8273
+    },
+    {
+      "epoch": 0.8704892162019989,
+      "grad_norm": 1.589955449104309,
+      "learning_rate": 8.670420346726182e-06,
+      "loss": 1.9209,
+      "step": 8274
+    },
+    {
+      "epoch": 0.8705944239873751,
+      "grad_norm": 1.278948187828064,
+      "learning_rate": 8.656546042144275e-06,
+      "loss": 1.406,
+      "step": 8275
+    },
+    {
+      "epoch": 0.8706996317727512,
+      "grad_norm": 1.640868067741394,
+      "learning_rate": 8.642682344976904e-06,
+      "loss": 1.4179,
+      "step": 8276
+    },
+    {
+      "epoch": 0.8708048395581273,
+      "grad_norm": 1.3409028053283691,
+      "learning_rate": 8.628829256833992e-06,
+      "loss": 1.4,
+      "step": 8277
+    },
+    {
+      "epoch": 0.8709100473435034,
+      "grad_norm": 1.4325164556503296,
+      "learning_rate": 8.614986779324252e-06,
+      "loss": 1.6548,
+      "step": 8278
+    },
+    {
+      "epoch": 0.8710152551288796,
+      "grad_norm": 2.271646022796631,
+      "learning_rate": 8.601154914055187e-06,
+      "loss": 1.3482,
+      "step": 8279
+    },
+    {
+      "epoch": 0.8711204629142556,
+      "grad_norm": 1.5409026145935059,
+      "learning_rate": 8.587333662633035e-06,
+      "loss": 1.5707,
+      "step": 8280
+    },
+    {
+      "epoch": 0.8712256706996317,
+      "grad_norm": 1.1222127676010132,
+      "learning_rate": 8.57352302666279e-06,
+      "loss": 1.4839,
+      "step": 8281
+    },
+    {
+      "epoch": 0.8713308784850079,
+      "grad_norm": 1.5568034648895264,
+      "learning_rate": 8.559723007748278e-06,
+      "loss": 1.5636,
+      "step": 8282
+    },
+    {
+      "epoch": 0.871436086270384,
+      "grad_norm": 1.717877984046936,
+      "learning_rate": 8.545933607492019e-06,
+      "loss": 1.3919,
+      "step": 8283
+    },
+    {
+      "epoch": 0.8715412940557601,
+      "grad_norm": 1.7419726848602295,
+      "learning_rate": 8.53215482749532e-06,
+      "loss": 1.164,
+      "step": 8284
+    },
+    {
+      "epoch": 0.8716465018411362,
+      "grad_norm": 1.5329818725585938,
+      "learning_rate": 8.518386669358313e-06,
+      "loss": 1.5617,
+      "step": 8285
+    },
+    {
+      "epoch": 0.8717517096265124,
+      "grad_norm": 1.6898797750473022,
+      "learning_rate": 8.50462913467982e-06,
+      "loss": 1.2344,
+      "step": 8286
+    },
+    {
+      "epoch": 0.8718569174118885,
+      "grad_norm": 2.2260582447052,
+      "learning_rate": 8.490882225057428e-06,
+      "loss": 1.6568,
+      "step": 8287
+    },
+    {
+      "epoch": 0.8719621251972646,
+      "grad_norm": 1.553295612335205,
+      "learning_rate": 8.477145942087583e-06,
+      "loss": 1.0574,
+      "step": 8288
+    },
+    {
+      "epoch": 0.8720673329826407,
+      "grad_norm": 1.8379864692687988,
+      "learning_rate": 8.463420287365386e-06,
+      "loss": 1.7292,
+      "step": 8289
+    },
+    {
+      "epoch": 0.8721725407680169,
+      "grad_norm": 2.359562873840332,
+      "learning_rate": 8.449705262484763e-06,
+      "loss": 1.2984,
+      "step": 8290
+    },
+    {
+      "epoch": 0.8722777485533929,
+      "grad_norm": 1.2469983100891113,
+      "learning_rate": 8.436000869038418e-06,
+      "loss": 1.3601,
+      "step": 8291
+    },
+    {
+      "epoch": 0.8723829563387691,
+      "grad_norm": 1.5775099992752075,
+      "learning_rate": 8.422307108617777e-06,
+      "loss": 1.3026,
+      "step": 8292
+    },
+    {
+      "epoch": 0.8724881641241452,
+      "grad_norm": 1.6676642894744873,
+      "learning_rate": 8.408623982813036e-06,
+      "loss": 1.1921,
+      "step": 8293
+    },
+    {
+      "epoch": 0.8725933719095214,
+      "grad_norm": 1.9857598543167114,
+      "learning_rate": 8.39495149321322e-06,
+      "loss": 1.7544,
+      "step": 8294
+    },
+    {
+      "epoch": 0.8726985796948974,
+      "grad_norm": 2.608306407928467,
+      "learning_rate": 8.381289641405998e-06,
+      "loss": 1.2208,
+      "step": 8295
+    },
+    {
+      "epoch": 0.8728037874802735,
+      "grad_norm": 1.3763883113861084,
+      "learning_rate": 8.367638428977942e-06,
+      "loss": 1.5372,
+      "step": 8296
+    },
+    {
+      "epoch": 0.8729089952656497,
+      "grad_norm": 1.8553149700164795,
+      "learning_rate": 8.353997857514296e-06,
+      "loss": 1.6343,
+      "step": 8297
+    },
+    {
+      "epoch": 0.8730142030510257,
+      "grad_norm": 2.202342987060547,
+      "learning_rate": 8.34036792859908e-06,
+      "loss": 1.5843,
+      "step": 8298
+    },
+    {
+      "epoch": 0.8731194108364019,
+      "grad_norm": 2.3540780544281006,
+      "learning_rate": 8.32674864381513e-06,
+      "loss": 1.5165,
+      "step": 8299
+    },
+    {
+      "epoch": 0.873224618621778,
+      "grad_norm": 2.0896689891815186,
+      "learning_rate": 8.31314000474398e-06,
+      "loss": 1.613,
+      "step": 8300
+    },
+    {
+      "epoch": 0.8733298264071542,
+      "grad_norm": 1.7080516815185547,
+      "learning_rate": 8.299542012965944e-06,
+      "loss": 1.3788,
+      "step": 8301
+    },
+    {
+      "epoch": 0.8734350341925302,
+      "grad_norm": 1.593510627746582,
+      "learning_rate": 8.285954670060159e-06,
+      "loss": 1.6948,
+      "step": 8302
+    },
+    {
+      "epoch": 0.8735402419779064,
+      "grad_norm": 2.1248574256896973,
+      "learning_rate": 8.272377977604439e-06,
+      "loss": 1.6842,
+      "step": 8303
+    },
+    {
+      "epoch": 0.8736454497632825,
+      "grad_norm": 1.1334164142608643,
+      "learning_rate": 8.258811937175403e-06,
+      "loss": 2.0723,
+      "step": 8304
+    },
+    {
+      "epoch": 0.8737506575486585,
+      "grad_norm": 1.6127103567123413,
+      "learning_rate": 8.245256550348456e-06,
+      "loss": 1.4466,
+      "step": 8305
+    },
+    {
+      "epoch": 0.8738558653340347,
+      "grad_norm": 1.4904117584228516,
+      "learning_rate": 8.231711818697708e-06,
+      "loss": 1.4292,
+      "step": 8306
+    },
+    {
+      "epoch": 0.8739610731194108,
+      "grad_norm": 1.7392334938049316,
+      "learning_rate": 8.218177743796096e-06,
+      "loss": 2.2474,
+      "step": 8307
+    },
+    {
+      "epoch": 0.874066280904787,
+      "grad_norm": 2.0337204933166504,
+      "learning_rate": 8.204654327215267e-06,
+      "loss": 1.369,
+      "step": 8308
+    },
+    {
+      "epoch": 0.874171488690163,
+      "grad_norm": 1.8506286144256592,
+      "learning_rate": 8.19114157052564e-06,
+      "loss": 1.587,
+      "step": 8309
+    },
+    {
+      "epoch": 0.8742766964755392,
+      "grad_norm": 1.861567497253418,
+      "learning_rate": 8.177639475296451e-06,
+      "loss": 0.9129,
+      "step": 8310
+    },
+    {
+      "epoch": 0.8743819042609153,
+      "grad_norm": 1.4691027402877808,
+      "learning_rate": 8.16414804309562e-06,
+      "loss": 1.5196,
+      "step": 8311
+    },
+    {
+      "epoch": 0.8744871120462914,
+      "grad_norm": 1.7774051427841187,
+      "learning_rate": 8.150667275489842e-06,
+      "loss": 2.4965,
+      "step": 8312
+    },
+    {
+      "epoch": 0.8745923198316675,
+      "grad_norm": 1.6832690238952637,
+      "learning_rate": 8.137197174044653e-06,
+      "loss": 1.4504,
+      "step": 8313
+    },
+    {
+      "epoch": 0.8746975276170437,
+      "grad_norm": 1.254486083984375,
+      "learning_rate": 8.123737740324256e-06,
+      "loss": 1.5565,
+      "step": 8314
+    },
+    {
+      "epoch": 0.8748027354024198,
+      "grad_norm": 2.1630659103393555,
+      "learning_rate": 8.110288975891634e-06,
+      "loss": 1.232,
+      "step": 8315
+    },
+    {
+      "epoch": 0.8749079431877959,
+      "grad_norm": 1.3778032064437866,
+      "learning_rate": 8.096850882308593e-06,
+      "loss": 1.517,
+      "step": 8316
+    },
+    {
+      "epoch": 0.875013150973172,
+      "grad_norm": 2.6462600231170654,
+      "learning_rate": 8.083423461135608e-06,
+      "loss": 1.6253,
+      "step": 8317
+    },
+    {
+      "epoch": 0.8751183587585482,
+      "grad_norm": 2.1237175464630127,
+      "learning_rate": 8.070006713931988e-06,
+      "loss": 1.6068,
+      "step": 8318
+    },
+    {
+      "epoch": 0.8752235665439243,
+      "grad_norm": 1.620760202407837,
+      "learning_rate": 8.056600642255773e-06,
+      "loss": 1.4239,
+      "step": 8319
+    },
+    {
+      "epoch": 0.8753287743293003,
+      "grad_norm": 1.3363070487976074,
+      "learning_rate": 8.043205247663755e-06,
+      "loss": 1.7418,
+      "step": 8320
+    },
+    {
+      "epoch": 0.8754339821146765,
+      "grad_norm": 1.1547250747680664,
+      "learning_rate": 8.029820531711518e-06,
+      "loss": 1.7182,
+      "step": 8321
+    },
+    {
+      "epoch": 0.8755391899000526,
+      "grad_norm": 1.5865908861160278,
+      "learning_rate": 8.016446495953367e-06,
+      "loss": 2.1783,
+      "step": 8322
+    },
+    {
+      "epoch": 0.8756443976854287,
+      "grad_norm": 1.7956304550170898,
+      "learning_rate": 8.00308314194238e-06,
+      "loss": 1.5709,
+      "step": 8323
+    },
+    {
+      "epoch": 0.8757496054708048,
+      "grad_norm": 1.6148009300231934,
+      "learning_rate": 7.989730471230417e-06,
+      "loss": 1.8065,
+      "step": 8324
+    },
+    {
+      "epoch": 0.875854813256181,
+      "grad_norm": 1.6082184314727783,
+      "learning_rate": 7.97638848536808e-06,
+      "loss": 1.824,
+      "step": 8325
+    },
+    {
+      "epoch": 0.8759600210415571,
+      "grad_norm": 1.5092358589172363,
+      "learning_rate": 7.963057185904698e-06,
+      "loss": 1.1529,
+      "step": 8326
+    },
+    {
+      "epoch": 0.8760652288269332,
+      "grad_norm": 1.4924912452697754,
+      "learning_rate": 7.949736574388433e-06,
+      "loss": 1.3965,
+      "step": 8327
+    },
+    {
+      "epoch": 0.8761704366123093,
+      "grad_norm": 1.484432339668274,
+      "learning_rate": 7.936426652366147e-06,
+      "loss": 1.5392,
+      "step": 8328
+    },
+    {
+      "epoch": 0.8762756443976855,
+      "grad_norm": 1.8063507080078125,
+      "learning_rate": 7.923127421383458e-06,
+      "loss": 1.7397,
+      "step": 8329
+    },
+    {
+      "epoch": 0.8763808521830615,
+      "grad_norm": 1.7138848304748535,
+      "learning_rate": 7.909838882984799e-06,
+      "loss": 1.9147,
+      "step": 8330
+    },
+    {
+      "epoch": 0.8764860599684376,
+      "grad_norm": 1.6054270267486572,
+      "learning_rate": 7.896561038713302e-06,
+      "loss": 1.7124,
+      "step": 8331
+    },
+    {
+      "epoch": 0.8765912677538138,
+      "grad_norm": 1.4371368885040283,
+      "learning_rate": 7.883293890110865e-06,
+      "loss": 1.3358,
+      "step": 8332
+    },
+    {
+      "epoch": 0.87669647553919,
+      "grad_norm": 1.3768408298492432,
+      "learning_rate": 7.870037438718191e-06,
+      "loss": 1.4805,
+      "step": 8333
+    },
+    {
+      "epoch": 0.876801683324566,
+      "grad_norm": 1.5090663433074951,
+      "learning_rate": 7.856791686074694e-06,
+      "loss": 1.819,
+      "step": 8334
+    },
+    {
+      "epoch": 0.8769068911099421,
+      "grad_norm": 1.7968672513961792,
+      "learning_rate": 7.84355663371854e-06,
+      "loss": 1.5492,
+      "step": 8335
+    },
+    {
+      "epoch": 0.8770120988953183,
+      "grad_norm": 1.5424824953079224,
+      "learning_rate": 7.830332283186714e-06,
+      "loss": 1.8034,
+      "step": 8336
+    },
+    {
+      "epoch": 0.8771173066806943,
+      "grad_norm": 1.5214895009994507,
+      "learning_rate": 7.817118636014886e-06,
+      "loss": 1.3005,
+      "step": 8337
+    },
+    {
+      "epoch": 0.8772225144660705,
+      "grad_norm": 1.3257561922073364,
+      "learning_rate": 7.803915693737518e-06,
+      "loss": 1.3967,
+      "step": 8338
+    },
+    {
+      "epoch": 0.8773277222514466,
+      "grad_norm": 2.690847635269165,
+      "learning_rate": 7.790723457887828e-06,
+      "loss": 1.3393,
+      "step": 8339
+    },
+    {
+      "epoch": 0.8774329300368228,
+      "grad_norm": 1.336316466331482,
+      "learning_rate": 7.777541929997766e-06,
+      "loss": 1.6733,
+      "step": 8340
+    },
+    {
+      "epoch": 0.8775381378221988,
+      "grad_norm": 1.7110776901245117,
+      "learning_rate": 7.7643711115981e-06,
+      "loss": 1.7544,
+      "step": 8341
+    },
+    {
+      "epoch": 0.877643345607575,
+      "grad_norm": 1.3267966508865356,
+      "learning_rate": 7.751211004218295e-06,
+      "loss": 1.7771,
+      "step": 8342
+    },
+    {
+      "epoch": 0.8777485533929511,
+      "grad_norm": 1.5423023700714111,
+      "learning_rate": 7.73806160938656e-06,
+      "loss": 1.3971,
+      "step": 8343
+    },
+    {
+      "epoch": 0.8778537611783271,
+      "grad_norm": 1.562389850616455,
+      "learning_rate": 7.724922928629941e-06,
+      "loss": 1.2311,
+      "step": 8344
+    },
+    {
+      "epoch": 0.8779589689637033,
+      "grad_norm": 1.6092841625213623,
+      "learning_rate": 7.711794963474173e-06,
+      "loss": 2.0053,
+      "step": 8345
+    },
+    {
+      "epoch": 0.8780641767490794,
+      "grad_norm": 1.8974779844284058,
+      "learning_rate": 7.698677715443736e-06,
+      "loss": 1.2428,
+      "step": 8346
+    },
+    {
+      "epoch": 0.8781693845344556,
+      "grad_norm": 1.8776745796203613,
+      "learning_rate": 7.685571186061934e-06,
+      "loss": 1.9842,
+      "step": 8347
+    },
+    {
+      "epoch": 0.8782745923198316,
+      "grad_norm": 1.4398527145385742,
+      "learning_rate": 7.672475376850764e-06,
+      "loss": 1.878,
+      "step": 8348
+    },
+    {
+      "epoch": 0.8783798001052078,
+      "grad_norm": 1.1596499681472778,
+      "learning_rate": 7.65939028933098e-06,
+      "loss": 1.6491,
+      "step": 8349
+    },
+    {
+      "epoch": 0.8784850078905839,
+      "grad_norm": 2.233144760131836,
+      "learning_rate": 7.646315925022152e-06,
+      "loss": 1.5677,
+      "step": 8350
+    },
+    {
+      "epoch": 0.8785902156759601,
+      "grad_norm": 2.088862419128418,
+      "learning_rate": 7.633252285442526e-06,
+      "loss": 1.6353,
+      "step": 8351
+    },
+    {
+      "epoch": 0.8786954234613361,
+      "grad_norm": 1.5038179159164429,
+      "learning_rate": 7.620199372109172e-06,
+      "loss": 2.0619,
+      "step": 8352
+    },
+    {
+      "epoch": 0.8788006312467123,
+      "grad_norm": 1.6128473281860352,
+      "learning_rate": 7.607157186537872e-06,
+      "loss": 1.9283,
+      "step": 8353
+    },
+    {
+      "epoch": 0.8789058390320884,
+      "grad_norm": 1.4330729246139526,
+      "learning_rate": 7.59412573024314e-06,
+      "loss": 1.757,
+      "step": 8354
+    },
+    {
+      "epoch": 0.8790110468174644,
+      "grad_norm": 1.7641589641571045,
+      "learning_rate": 7.581105004738321e-06,
+      "loss": 1.8615,
+      "step": 8355
+    },
+    {
+      "epoch": 0.8791162546028406,
+      "grad_norm": 1.5351738929748535,
+      "learning_rate": 7.568095011535448e-06,
+      "loss": 1.5014,
+      "step": 8356
+    },
+    {
+      "epoch": 0.8792214623882167,
+      "grad_norm": 1.7756725549697876,
+      "learning_rate": 7.555095752145313e-06,
+      "loss": 1.4319,
+      "step": 8357
+    },
+    {
+      "epoch": 0.8793266701735929,
+      "grad_norm": 1.6346051692962646,
+      "learning_rate": 7.542107228077533e-06,
+      "loss": 1.2158,
+      "step": 8358
+    },
+    {
+      "epoch": 0.8794318779589689,
+      "grad_norm": 2.2152860164642334,
+      "learning_rate": 7.529129440840355e-06,
+      "loss": 1.389,
+      "step": 8359
+    },
+    {
+      "epoch": 0.8795370857443451,
+      "grad_norm": 1.4864938259124756,
+      "learning_rate": 7.516162391940873e-06,
+      "loss": 0.8488,
+      "step": 8360
+    },
+    {
+      "epoch": 0.8796422935297212,
+      "grad_norm": 2.238067626953125,
+      "learning_rate": 7.503206082884917e-06,
+      "loss": 1.8396,
+      "step": 8361
+    },
+    {
+      "epoch": 0.8797475013150973,
+      "grad_norm": 1.6795562505722046,
+      "learning_rate": 7.4902605151770385e-06,
+      "loss": 1.9979,
+      "step": 8362
+    },
+    {
+      "epoch": 0.8798527091004734,
+      "grad_norm": 2.297811269760132,
+      "learning_rate": 7.477325690320602e-06,
+      "loss": 1.9526,
+      "step": 8363
+    },
+    {
+      "epoch": 0.8799579168858496,
+      "grad_norm": 2.567331075668335,
+      "learning_rate": 7.4644016098176615e-06,
+      "loss": 1.5925,
+      "step": 8364
+    },
+    {
+      "epoch": 0.8800631246712257,
+      "grad_norm": 1.6319959163665771,
+      "learning_rate": 7.451488275169028e-06,
+      "loss": 1.9061,
+      "step": 8365
+    },
+    {
+      "epoch": 0.8801683324566018,
+      "grad_norm": 2.1503756046295166,
+      "learning_rate": 7.438585687874333e-06,
+      "loss": 1.4737,
+      "step": 8366
+    },
+    {
+      "epoch": 0.8802735402419779,
+      "grad_norm": 1.6706979274749756,
+      "learning_rate": 7.42569384943187e-06,
+      "loss": 1.822,
+      "step": 8367
+    },
+    {
+      "epoch": 0.880378748027354,
+      "grad_norm": 1.9475339651107788,
+      "learning_rate": 7.412812761338739e-06,
+      "loss": 1.8185,
+      "step": 8368
+    },
+    {
+      "epoch": 0.8804839558127301,
+      "grad_norm": 1.8608582019805908,
+      "learning_rate": 7.3999424250907775e-06,
+      "loss": 2.2257,
+      "step": 8369
+    },
+    {
+      "epoch": 0.8805891635981062,
+      "grad_norm": 1.6136201620101929,
+      "learning_rate": 7.387082842182591e-06,
+      "loss": 1.6696,
+      "step": 8370
+    },
+    {
+      "epoch": 0.8806943713834824,
+      "grad_norm": 1.2842328548431396,
+      "learning_rate": 7.374234014107484e-06,
+      "loss": 1.8681,
+      "step": 8371
+    },
+    {
+      "epoch": 0.8807995791688585,
+      "grad_norm": 2.344787120819092,
+      "learning_rate": 7.361395942357596e-06,
+      "loss": 1.4332,
+      "step": 8372
+    },
+    {
+      "epoch": 0.8809047869542346,
+      "grad_norm": 3.1230101585388184,
+      "learning_rate": 7.348568628423746e-06,
+      "loss": 1.5007,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8810099947396107,
+      "grad_norm": 1.2945367097854614,
+      "learning_rate": 7.335752073795499e-06,
+      "loss": 1.7119,
+      "step": 8374
+    },
+    {
+      "epoch": 0.8811152025249869,
+      "grad_norm": 1.189368486404419,
+      "learning_rate": 7.322946279961252e-06,
+      "loss": 1.8333,
+      "step": 8375
+    },
+    {
+      "epoch": 0.8812204103103629,
+      "grad_norm": 1.435043454170227,
+      "learning_rate": 7.31015124840807e-06,
+      "loss": 1.6932,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8813256180957391,
+      "grad_norm": 2.0978446006774902,
+      "learning_rate": 7.297366980621789e-06,
+      "loss": 1.2705,
+      "step": 8377
+    },
+    {
+      "epoch": 0.8814308258811152,
+      "grad_norm": 1.0726866722106934,
+      "learning_rate": 7.284593478087043e-06,
+      "loss": 1.846,
+      "step": 8378
+    },
+    {
+      "epoch": 0.8815360336664914,
+      "grad_norm": 1.2944754362106323,
+      "learning_rate": 7.2718307422871445e-06,
+      "loss": 1.2121,
+      "step": 8379
+    },
+    {
+      "epoch": 0.8816412414518674,
+      "grad_norm": 1.729756474494934,
+      "learning_rate": 7.259078774704198e-06,
+      "loss": 2.0314,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8817464492372435,
+      "grad_norm": 1.872869610786438,
+      "learning_rate": 7.24633757681904e-06,
+      "loss": 1.9162,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8818516570226197,
+      "grad_norm": 1.577900767326355,
+      "learning_rate": 7.233607150111255e-06,
+      "loss": 1.697,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8819568648079958,
+      "grad_norm": 1.8250850439071655,
+      "learning_rate": 7.2208874960592145e-06,
+      "loss": 1.1092,
+      "step": 8383
+    },
+    {
+      "epoch": 0.8820620725933719,
+      "grad_norm": 1.5504481792449951,
+      "learning_rate": 7.208178616139994e-06,
+      "loss": 1.265,
+      "step": 8384
+    },
+    {
+      "epoch": 0.882167280378748,
+      "grad_norm": 1.681287407875061,
+      "learning_rate": 7.195480511829411e-06,
+      "loss": 1.366,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8822724881641242,
+      "grad_norm": 1.5954519510269165,
+      "learning_rate": 7.18279318460211e-06,
+      "loss": 1.4569,
+      "step": 8386
+    },
+    {
+      "epoch": 0.8823776959495002,
+      "grad_norm": 1.6172798871994019,
+      "learning_rate": 7.1701166359313894e-06,
+      "loss": 1.6745,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8824829037348764,
+      "grad_norm": 1.5901143550872803,
+      "learning_rate": 7.157450867289317e-06,
+      "loss": 1.6646,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8825881115202525,
+      "grad_norm": 1.5728827714920044,
+      "learning_rate": 7.1447958801467816e-06,
+      "loss": 1.5367,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8826933193056287,
+      "grad_norm": 1.7115237712860107,
+      "learning_rate": 7.132151675973331e-06,
+      "loss": 1.3662,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8827985270910047,
+      "grad_norm": 1.2988559007644653,
+      "learning_rate": 7.119518256237279e-06,
+      "loss": 1.5863,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8829037348763809,
+      "grad_norm": 1.656714916229248,
+      "learning_rate": 7.106895622405752e-06,
+      "loss": 1.6779,
+      "step": 8392
+    },
+    {
+      "epoch": 0.883008942661757,
+      "grad_norm": 1.3429776430130005,
+      "learning_rate": 7.0942837759445325e-06,
+      "loss": 1.6569,
+      "step": 8393
+    },
+    {
+      "epoch": 0.883114150447133,
+      "grad_norm": 1.6043792963027954,
+      "learning_rate": 7.081682718318194e-06,
+      "loss": 1.4003,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8832193582325092,
+      "grad_norm": 1.8559693098068237,
+      "learning_rate": 7.069092450990089e-06,
+      "loss": 1.6701,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8833245660178853,
+      "grad_norm": 1.5783679485321045,
+      "learning_rate": 7.056512975422269e-06,
+      "loss": 1.5224,
+      "step": 8396
+    },
+    {
+      "epoch": 0.8834297738032615,
+      "grad_norm": 1.5466448068618774,
+      "learning_rate": 7.0439442930755105e-06,
+      "loss": 1.5427,
+      "step": 8397
+    },
+    {
+      "epoch": 0.8835349815886375,
+      "grad_norm": 1.4492648839950562,
+      "learning_rate": 7.031386405409434e-06,
+      "loss": 1.3369,
+      "step": 8398
+    },
+    {
+      "epoch": 0.8836401893740137,
+      "grad_norm": 1.4293040037155151,
+      "learning_rate": 7.018839313882286e-06,
+      "loss": 1.3867,
+      "step": 8399
+    },
+    {
+      "epoch": 0.8837453971593898,
+      "grad_norm": 2.3548078536987305,
+      "learning_rate": 7.006303019951177e-06,
+      "loss": 1.9804,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8838506049447659,
+      "grad_norm": 1.3848503828048706,
+      "learning_rate": 6.993777525071887e-06,
+      "loss": 1.1785,
+      "step": 8401
+    },
+    {
+      "epoch": 0.883955812730142,
+      "grad_norm": 1.6691620349884033,
+      "learning_rate": 6.98126283069892e-06,
+      "loss": 1.5292,
+      "step": 8402
+    },
+    {
+      "epoch": 0.8840610205155182,
+      "grad_norm": 1.5758638381958008,
+      "learning_rate": 6.968758938285614e-06,
+      "loss": 1.3405,
+      "step": 8403
+    },
+    {
+      "epoch": 0.8841662283008943,
+      "grad_norm": 1.6990686655044556,
+      "learning_rate": 6.956265849283994e-06,
+      "loss": 1.3229,
+      "step": 8404
+    },
+    {
+      "epoch": 0.8842714360862703,
+      "grad_norm": 2.8560450077056885,
+      "learning_rate": 6.943783565144812e-06,
+      "loss": 1.6618,
+      "step": 8405
+    },
+    {
+      "epoch": 0.8843766438716465,
+      "grad_norm": 1.6072392463684082,
+      "learning_rate": 6.931312087317632e-06,
+      "loss": 1.7917,
+      "step": 8406
+    },
+    {
+      "epoch": 0.8844818516570226,
+      "grad_norm": 1.5466362237930298,
+      "learning_rate": 6.918851417250693e-06,
+      "loss": 1.321,
+      "step": 8407
+    },
+    {
+      "epoch": 0.8845870594423987,
+      "grad_norm": 1.8458850383758545,
+      "learning_rate": 6.906401556391051e-06,
+      "loss": 1.6448,
+      "step": 8408
+    },
+    {
+      "epoch": 0.8846922672277748,
+      "grad_norm": 2.208996295928955,
+      "learning_rate": 6.893962506184448e-06,
+      "loss": 1.8648,
+      "step": 8409
+    },
+    {
+      "epoch": 0.884797475013151,
+      "grad_norm": 1.3810334205627441,
+      "learning_rate": 6.8815342680753735e-06,
+      "loss": 1.7885,
+      "step": 8410
+    },
+    {
+      "epoch": 0.8849026827985271,
+      "grad_norm": 1.5404844284057617,
+      "learning_rate": 6.869116843507106e-06,
+      "loss": 1.5237,
+      "step": 8411
+    },
+    {
+      "epoch": 0.8850078905839032,
+      "grad_norm": 1.3252919912338257,
+      "learning_rate": 6.856710233921626e-06,
+      "loss": 1.9188,
+      "step": 8412
+    },
+    {
+      "epoch": 0.8851130983692793,
+      "grad_norm": 1.3163280487060547,
+      "learning_rate": 6.844314440759647e-06,
+      "loss": 1.8506,
+      "step": 8413
+    },
+    {
+      "epoch": 0.8852183061546555,
+      "grad_norm": 1.7833784818649292,
+      "learning_rate": 6.8319294654607065e-06,
+      "loss": 1.2899,
+      "step": 8414
+    },
+    {
+      "epoch": 0.8853235139400316,
+      "grad_norm": 2.266003370285034,
+      "learning_rate": 6.8195553094629995e-06,
+      "loss": 1.1743,
+      "step": 8415
+    },
+    {
+      "epoch": 0.8854287217254077,
+      "grad_norm": 2.029226303100586,
+      "learning_rate": 6.807191974203486e-06,
+      "loss": 1.04,
+      "step": 8416
+    },
+    {
+      "epoch": 0.8855339295107838,
+      "grad_norm": 1.7757220268249512,
+      "learning_rate": 6.7948394611178964e-06,
+      "loss": 1.2583,
+      "step": 8417
+    },
+    {
+      "epoch": 0.88563913729616,
+      "grad_norm": 2.46942138671875,
+      "learning_rate": 6.782497771640694e-06,
+      "loss": 1.5153,
+      "step": 8418
+    },
+    {
+      "epoch": 0.885744345081536,
+      "grad_norm": 2.080439805984497,
+      "learning_rate": 6.770166907205044e-06,
+      "loss": 0.9513,
+      "step": 8419
+    },
+    {
+      "epoch": 0.8858495528669121,
+      "grad_norm": 1.3978705406188965,
+      "learning_rate": 6.7578468692429345e-06,
+      "loss": 1.2028,
+      "step": 8420
+    },
+    {
+      "epoch": 0.8859547606522883,
+      "grad_norm": 1.2934406995773315,
+      "learning_rate": 6.7455376591850195e-06,
+      "loss": 1.6844,
+      "step": 8421
+    },
+    {
+      "epoch": 0.8860599684376644,
+      "grad_norm": 1.9341928958892822,
+      "learning_rate": 6.733239278460735e-06,
+      "loss": 1.5859,
+      "step": 8422
+    },
+    {
+      "epoch": 0.8861651762230405,
+      "grad_norm": 1.2040367126464844,
+      "learning_rate": 6.7209517284982704e-06,
+      "loss": 1.3829,
+      "step": 8423
+    },
+    {
+      "epoch": 0.8862703840084166,
+      "grad_norm": 1.6010814905166626,
+      "learning_rate": 6.7086750107244965e-06,
+      "loss": 1.3713,
+      "step": 8424
+    },
+    {
+      "epoch": 0.8863755917937928,
+      "grad_norm": 2.1789660453796387,
+      "learning_rate": 6.696409126565107e-06,
+      "loss": 1.9039,
+      "step": 8425
+    },
+    {
+      "epoch": 0.8864807995791688,
+      "grad_norm": 1.1743876934051514,
+      "learning_rate": 6.684154077444482e-06,
+      "loss": 1.6147,
+      "step": 8426
+    },
+    {
+      "epoch": 0.886586007364545,
+      "grad_norm": 1.855284571647644,
+      "learning_rate": 6.6719098647857525e-06,
+      "loss": 1.5138,
+      "step": 8427
+    },
+    {
+      "epoch": 0.8866912151499211,
+      "grad_norm": 1.862816333770752,
+      "learning_rate": 6.659676490010824e-06,
+      "loss": 1.5119,
+      "step": 8428
+    },
+    {
+      "epoch": 0.8867964229352973,
+      "grad_norm": 1.4365988969802856,
+      "learning_rate": 6.647453954540295e-06,
+      "loss": 1.6655,
+      "step": 8429
+    },
+    {
+      "epoch": 0.8869016307206733,
+      "grad_norm": 1.137057900428772,
+      "learning_rate": 6.635242259793528e-06,
+      "loss": 1.6778,
+      "step": 8430
+    },
+    {
+      "epoch": 0.8870068385060494,
+      "grad_norm": 1.6340605020523071,
+      "learning_rate": 6.623041407188646e-06,
+      "loss": 1.7023,
+      "step": 8431
+    },
+    {
+      "epoch": 0.8871120462914256,
+      "grad_norm": 1.682706356048584,
+      "learning_rate": 6.610851398142482e-06,
+      "loss": 1.6385,
+      "step": 8432
+    },
+    {
+      "epoch": 0.8872172540768016,
+      "grad_norm": 2.859862804412842,
+      "learning_rate": 6.598672234070602e-06,
+      "loss": 1.4929,
+      "step": 8433
+    },
+    {
+      "epoch": 0.8873224618621778,
+      "grad_norm": 1.687212347984314,
+      "learning_rate": 6.586503916387366e-06,
+      "loss": 1.3471,
+      "step": 8434
+    },
+    {
+      "epoch": 0.8874276696475539,
+      "grad_norm": 1.9600337743759155,
+      "learning_rate": 6.574346446505841e-06,
+      "loss": 1.8195,
+      "step": 8435
+    },
+    {
+      "epoch": 0.8875328774329301,
+      "grad_norm": 1.6050618886947632,
+      "learning_rate": 6.562199825837789e-06,
+      "loss": 1.4023,
+      "step": 8436
+    },
+    {
+      "epoch": 0.8876380852183061,
+      "grad_norm": 1.2669776678085327,
+      "learning_rate": 6.550064055793815e-06,
+      "loss": 1.6314,
+      "step": 8437
+    },
+    {
+      "epoch": 0.8877432930036823,
+      "grad_norm": 1.4774951934814453,
+      "learning_rate": 6.537939137783166e-06,
+      "loss": 1.5599,
+      "step": 8438
+    },
+    {
+      "epoch": 0.8878485007890584,
+      "grad_norm": 2.499821424484253,
+      "learning_rate": 6.525825073213876e-06,
+      "loss": 1.9315,
+      "step": 8439
+    },
+    {
+      "epoch": 0.8879537085744345,
+      "grad_norm": 1.3363654613494873,
+      "learning_rate": 6.513721863492739e-06,
+      "loss": 1.8184,
+      "step": 8440
+    },
+    {
+      "epoch": 0.8880589163598106,
+      "grad_norm": 1.872019648551941,
+      "learning_rate": 6.501629510025231e-06,
+      "loss": 1.8917,
+      "step": 8441
+    },
+    {
+      "epoch": 0.8881641241451868,
+      "grad_norm": 1.3895902633666992,
+      "learning_rate": 6.489548014215585e-06,
+      "loss": 1.296,
+      "step": 8442
+    },
+    {
+      "epoch": 0.8882693319305629,
+      "grad_norm": 1.1768591403961182,
+      "learning_rate": 6.4774773774668225e-06,
+      "loss": 1.3739,
+      "step": 8443
+    },
+    {
+      "epoch": 0.8883745397159389,
+      "grad_norm": 1.620846152305603,
+      "learning_rate": 6.465417601180657e-06,
+      "loss": 1.3514,
+      "step": 8444
+    },
+    {
+      "epoch": 0.8884797475013151,
+      "grad_norm": 1.3851511478424072,
+      "learning_rate": 6.453368686757533e-06,
+      "loss": 1.2824,
+      "step": 8445
+    },
+    {
+      "epoch": 0.8885849552866912,
+      "grad_norm": 1.9092847108840942,
+      "learning_rate": 6.441330635596665e-06,
+      "loss": 1.6623,
+      "step": 8446
+    },
+    {
+      "epoch": 0.8886901630720674,
+      "grad_norm": 1.3836404085159302,
+      "learning_rate": 6.42930344909598e-06,
+      "loss": 1.932,
+      "step": 8447
+    },
+    {
+      "epoch": 0.8887953708574434,
+      "grad_norm": 1.2607841491699219,
+      "learning_rate": 6.417287128652172e-06,
+      "loss": 1.5857,
+      "step": 8448
+    },
+    {
+      "epoch": 0.8889005786428196,
+      "grad_norm": 1.6962370872497559,
+      "learning_rate": 6.405281675660657e-06,
+      "loss": 1.7428,
+      "step": 8449
+    },
+    {
+      "epoch": 0.8890057864281957,
+      "grad_norm": 1.8756303787231445,
+      "learning_rate": 6.393287091515565e-06,
+      "loss": 1.3188,
+      "step": 8450
+    },
+    {
+      "epoch": 0.8891109942135718,
+      "grad_norm": 1.370875358581543,
+      "learning_rate": 6.3813033776098045e-06,
+      "loss": 1.9644,
+      "step": 8451
+    },
+    {
+      "epoch": 0.8892162019989479,
+      "grad_norm": 1.3783890008926392,
+      "learning_rate": 6.369330535335016e-06,
+      "loss": 1.3712,
+      "step": 8452
+    },
+    {
+      "epoch": 0.8893214097843241,
+      "grad_norm": 1.7575373649597168,
+      "learning_rate": 6.357368566081534e-06,
+      "loss": 1.7293,
+      "step": 8453
+    },
+    {
+      "epoch": 0.8894266175697002,
+      "grad_norm": 1.3849029541015625,
+      "learning_rate": 6.345417471238501e-06,
+      "loss": 1.6498,
+      "step": 8454
+    },
+    {
+      "epoch": 0.8895318253550762,
+      "grad_norm": 2.384744167327881,
+      "learning_rate": 6.333477252193731e-06,
+      "loss": 2.3091,
+      "step": 8455
+    },
+    {
+      "epoch": 0.8896370331404524,
+      "grad_norm": 1.8546522855758667,
+      "learning_rate": 6.321547910333814e-06,
+      "loss": 1.5911,
+      "step": 8456
+    },
+    {
+      "epoch": 0.8897422409258285,
+      "grad_norm": 1.3433914184570312,
+      "learning_rate": 6.309629447044074e-06,
+      "loss": 2.1943,
+      "step": 8457
+    },
+    {
+      "epoch": 0.8898474487112046,
+      "grad_norm": 1.997424840927124,
+      "learning_rate": 6.297721863708528e-06,
+      "loss": 1.8741,
+      "step": 8458
+    },
+    {
+      "epoch": 0.8899526564965807,
+      "grad_norm": 1.5434318780899048,
+      "learning_rate": 6.285825161710002e-06,
+      "loss": 0.8767,
+      "step": 8459
+    },
+    {
+      "epoch": 0.8900578642819569,
+      "grad_norm": 1.7368338108062744,
+      "learning_rate": 6.273939342430013e-06,
+      "loss": 1.6961,
+      "step": 8460
+    },
+    {
+      "epoch": 0.890163072067333,
+      "grad_norm": 1.936927318572998,
+      "learning_rate": 6.262064407248791e-06,
+      "loss": 0.8594,
+      "step": 8461
+    },
+    {
+      "epoch": 0.8902682798527091,
+      "grad_norm": 1.7954018115997314,
+      "learning_rate": 6.250200357545377e-06,
+      "loss": 1.6956,
+      "step": 8462
+    },
+    {
+      "epoch": 0.8903734876380852,
+      "grad_norm": 1.5254756212234497,
+      "learning_rate": 6.238347194697492e-06,
+      "loss": 1.6629,
+      "step": 8463
+    },
+    {
+      "epoch": 0.8904786954234614,
+      "grad_norm": 1.8098357915878296,
+      "learning_rate": 6.226504920081566e-06,
+      "loss": 1.17,
+      "step": 8464
+    },
+    {
+      "epoch": 0.8905839032088374,
+      "grad_norm": 1.3831403255462646,
+      "learning_rate": 6.214673535072868e-06,
+      "loss": 1.6587,
+      "step": 8465
+    },
+    {
+      "epoch": 0.8906891109942136,
+      "grad_norm": 1.8217341899871826,
+      "learning_rate": 6.202853041045298e-06,
+      "loss": 1.7295,
+      "step": 8466
+    },
+    {
+      "epoch": 0.8907943187795897,
+      "grad_norm": 0.8784952759742737,
+      "learning_rate": 6.191043439371535e-06,
+      "loss": 2.3105,
+      "step": 8467
+    },
+    {
+      "epoch": 0.8908995265649658,
+      "grad_norm": 1.1799579858779907,
+      "learning_rate": 6.179244731422984e-06,
+      "loss": 1.1108,
+      "step": 8468
+    },
+    {
+      "epoch": 0.8910047343503419,
+      "grad_norm": 1.5560115575790405,
+      "learning_rate": 6.167456918569792e-06,
+      "loss": 1.9163,
+      "step": 8469
+    },
+    {
+      "epoch": 0.891109942135718,
+      "grad_norm": 1.6790612936019897,
+      "learning_rate": 6.155680002180864e-06,
+      "loss": 1.5781,
+      "step": 8470
+    },
+    {
+      "epoch": 0.8912151499210942,
+      "grad_norm": 2.0735700130462646,
+      "learning_rate": 6.143913983623795e-06,
+      "loss": 1.601,
+      "step": 8471
+    },
+    {
+      "epoch": 0.8913203577064703,
+      "grad_norm": 1.3503835201263428,
+      "learning_rate": 6.132158864264914e-06,
+      "loss": 1.8698,
+      "step": 8472
+    },
+    {
+      "epoch": 0.8914255654918464,
+      "grad_norm": 1.1508896350860596,
+      "learning_rate": 6.120414645469341e-06,
+      "loss": 1.8827,
+      "step": 8473
+    },
+    {
+      "epoch": 0.8915307732772225,
+      "grad_norm": 2.075125217437744,
+      "learning_rate": 6.108681328600874e-06,
+      "loss": 1.8445,
+      "step": 8474
+    },
+    {
+      "epoch": 0.8916359810625987,
+      "grad_norm": 1.5181657075881958,
+      "learning_rate": 6.0969589150220554e-06,
+      "loss": 1.8365,
+      "step": 8475
+    },
+    {
+      "epoch": 0.8917411888479747,
+      "grad_norm": 1.7302663326263428,
+      "learning_rate": 6.085247406094197e-06,
+      "loss": 1.0664,
+      "step": 8476
+    },
+    {
+      "epoch": 0.8918463966333509,
+      "grad_norm": 1.5871552228927612,
+      "learning_rate": 6.073546803177299e-06,
+      "loss": 1.7855,
+      "step": 8477
+    },
+    {
+      "epoch": 0.891951604418727,
+      "grad_norm": 1.1431766748428345,
+      "learning_rate": 6.0618571076301085e-06,
+      "loss": 2.0511,
+      "step": 8478
+    },
+    {
+      "epoch": 0.8920568122041032,
+      "grad_norm": 1.5150169134140015,
+      "learning_rate": 6.050178320810141e-06,
+      "loss": 1.4898,
+      "step": 8479
+    },
+    {
+      "epoch": 0.8921620199894792,
+      "grad_norm": 2.173424005508423,
+      "learning_rate": 6.038510444073586e-06,
+      "loss": 1.1078,
+      "step": 8480
+    },
+    {
+      "epoch": 0.8922672277748553,
+      "grad_norm": 2.229217529296875,
+      "learning_rate": 6.026853478775396e-06,
+      "loss": 1.9185,
+      "step": 8481
+    },
+    {
+      "epoch": 0.8923724355602315,
+      "grad_norm": 2.0256011486053467,
+      "learning_rate": 6.015207426269276e-06,
+      "loss": 2.1693,
+      "step": 8482
+    },
+    {
+      "epoch": 0.8924776433456075,
+      "grad_norm": 1.8043358325958252,
+      "learning_rate": 6.003572287907633e-06,
+      "loss": 1.7443,
+      "step": 8483
+    },
+    {
+      "epoch": 0.8925828511309837,
+      "grad_norm": 1.3562211990356445,
+      "learning_rate": 5.991948065041608e-06,
+      "loss": 1.4683,
+      "step": 8484
+    },
+    {
+      "epoch": 0.8926880589163598,
+      "grad_norm": 1.540953278541565,
+      "learning_rate": 5.9803347590211e-06,
+      "loss": 1.6193,
+      "step": 8485
+    },
+    {
+      "epoch": 0.892793266701736,
+      "grad_norm": 1.7050999402999878,
+      "learning_rate": 5.968732371194729e-06,
+      "loss": 1.4715,
+      "step": 8486
+    },
+    {
+      "epoch": 0.892898474487112,
+      "grad_norm": 2.0793585777282715,
+      "learning_rate": 5.957140902909819e-06,
+      "loss": 1.22,
+      "step": 8487
+    },
+    {
+      "epoch": 0.8930036822724882,
+      "grad_norm": 2.1180810928344727,
+      "learning_rate": 5.945560355512458e-06,
+      "loss": 1.2444,
+      "step": 8488
+    },
+    {
+      "epoch": 0.8931088900578643,
+      "grad_norm": 1.1381313800811768,
+      "learning_rate": 5.93399073034745e-06,
+      "loss": 1.5193,
+      "step": 8489
+    },
+    {
+      "epoch": 0.8932140978432404,
+      "grad_norm": 1.6418750286102295,
+      "learning_rate": 5.922432028758362e-06,
+      "loss": 1.7449,
+      "step": 8490
+    },
+    {
+      "epoch": 0.8933193056286165,
+      "grad_norm": 1.5433701276779175,
+      "learning_rate": 5.910884252087457e-06,
+      "loss": 1.4825,
+      "step": 8491
+    },
+    {
+      "epoch": 0.8934245134139926,
+      "grad_norm": 1.8106061220169067,
+      "learning_rate": 5.8993474016757145e-06,
+      "loss": 1.542,
+      "step": 8492
+    },
+    {
+      "epoch": 0.8935297211993688,
+      "grad_norm": 1.619828701019287,
+      "learning_rate": 5.88782147886291e-06,
+      "loss": 1.8641,
+      "step": 8493
+    },
+    {
+      "epoch": 0.8936349289847448,
+      "grad_norm": 1.3758325576782227,
+      "learning_rate": 5.876306484987481e-06,
+      "loss": 1.3756,
+      "step": 8494
+    },
+    {
+      "epoch": 0.893740136770121,
+      "grad_norm": 1.4138890504837036,
+      "learning_rate": 5.8648024213866396e-06,
+      "loss": 1.3945,
+      "step": 8495
+    },
+    {
+      "epoch": 0.8938453445554971,
+      "grad_norm": 2.485138416290283,
+      "learning_rate": 5.853309289396314e-06,
+      "loss": 1.6306,
+      "step": 8496
+    },
+    {
+      "epoch": 0.8939505523408732,
+      "grad_norm": 1.9388067722320557,
+      "learning_rate": 5.841827090351171e-06,
+      "loss": 1.4469,
+      "step": 8497
+    },
+    {
+      "epoch": 0.8940557601262493,
+      "grad_norm": 1.8217087984085083,
+      "learning_rate": 5.830355825584577e-06,
+      "loss": 1.3221,
+      "step": 8498
+    },
+    {
+      "epoch": 0.8941609679116255,
+      "grad_norm": 1.5511425733566284,
+      "learning_rate": 5.818895496428689e-06,
+      "loss": 1.3081,
+      "step": 8499
+    },
+    {
+      "epoch": 0.8942661756970016,
+      "grad_norm": 2.5423057079315186,
+      "learning_rate": 5.8074461042143095e-06,
+      "loss": 1.1463,
+      "step": 8500
+    },
+    {
+      "epoch": 0.8943713834823777,
+      "grad_norm": 1.830324411392212,
+      "learning_rate": 5.796007650271063e-06,
+      "loss": 1.4945,
+      "step": 8501
+    },
+    {
+      "epoch": 0.8944765912677538,
+      "grad_norm": 2.4133145809173584,
+      "learning_rate": 5.784580135927242e-06,
+      "loss": 1.5212,
+      "step": 8502
+    },
+    {
+      "epoch": 0.89458179905313,
+      "grad_norm": 1.9608807563781738,
+      "learning_rate": 5.7731635625098755e-06,
+      "loss": 1.3602,
+      "step": 8503
+    },
+    {
+      "epoch": 0.8946870068385061,
+      "grad_norm": 2.1056807041168213,
+      "learning_rate": 5.761757931344758e-06,
+      "loss": 1.5038,
+      "step": 8504
+    },
+    {
+      "epoch": 0.8947922146238821,
+      "grad_norm": 2.061222553253174,
+      "learning_rate": 5.750363243756363e-06,
+      "loss": 1.9814,
+      "step": 8505
+    },
+    {
+      "epoch": 0.8948974224092583,
+      "grad_norm": 2.296455144882202,
+      "learning_rate": 5.738979501067921e-06,
+      "loss": 2.0531,
+      "step": 8506
+    },
+    {
+      "epoch": 0.8950026301946344,
+      "grad_norm": 2.311011552810669,
+      "learning_rate": 5.727606704601407e-06,
+      "loss": 1.719,
+      "step": 8507
+    },
+    {
+      "epoch": 0.8951078379800105,
+      "grad_norm": 1.3462337255477905,
+      "learning_rate": 5.7162448556774995e-06,
+      "loss": 1.7112,
+      "step": 8508
+    },
+    {
+      "epoch": 0.8952130457653866,
+      "grad_norm": 1.3177099227905273,
+      "learning_rate": 5.704893955615598e-06,
+      "loss": 1.8481,
+      "step": 8509
+    },
+    {
+      "epoch": 0.8953182535507628,
+      "grad_norm": 1.1072567701339722,
+      "learning_rate": 5.693554005733859e-06,
+      "loss": 1.6795,
+      "step": 8510
+    },
+    {
+      "epoch": 0.8954234613361389,
+      "grad_norm": 1.5659089088439941,
+      "learning_rate": 5.68222500734914e-06,
+      "loss": 1.4823,
+      "step": 8511
+    },
+    {
+      "epoch": 0.895528669121515,
+      "grad_norm": 1.4302914142608643,
+      "learning_rate": 5.6709069617770675e-06,
+      "loss": 1.2352,
+      "step": 8512
+    },
+    {
+      "epoch": 0.8956338769068911,
+      "grad_norm": 1.4420124292373657,
+      "learning_rate": 5.659599870331944e-06,
+      "loss": 1.5293,
+      "step": 8513
+    },
+    {
+      "epoch": 0.8957390846922673,
+      "grad_norm": 2.497387170791626,
+      "learning_rate": 5.64830373432681e-06,
+      "loss": 1.2493,
+      "step": 8514
+    },
+    {
+      "epoch": 0.8958442924776433,
+      "grad_norm": 1.808634877204895,
+      "learning_rate": 5.637018555073492e-06,
+      "loss": 1.8987,
+      "step": 8515
+    },
+    {
+      "epoch": 0.8959495002630194,
+      "grad_norm": 2.0679233074188232,
+      "learning_rate": 5.625744333882488e-06,
+      "loss": 1.3279,
+      "step": 8516
+    },
+    {
+      "epoch": 0.8960547080483956,
+      "grad_norm": 1.3654800653457642,
+      "learning_rate": 5.614481072063005e-06,
+      "loss": 1.6085,
+      "step": 8517
+    },
+    {
+      "epoch": 0.8961599158337717,
+      "grad_norm": 1.6575655937194824,
+      "learning_rate": 5.603228770923041e-06,
+      "loss": 1.1494,
+      "step": 8518
+    },
+    {
+      "epoch": 0.8962651236191478,
+      "grad_norm": 1.3457351922988892,
+      "learning_rate": 5.591987431769285e-06,
+      "loss": 1.7025,
+      "step": 8519
+    },
+    {
+      "epoch": 0.8963703314045239,
+      "grad_norm": 1.5918161869049072,
+      "learning_rate": 5.580757055907137e-06,
+      "loss": 1.3449,
+      "step": 8520
+    },
+    {
+      "epoch": 0.8964755391899001,
+      "grad_norm": 1.6226881742477417,
+      "learning_rate": 5.5695376446407656e-06,
+      "loss": 1.6743,
+      "step": 8521
+    },
+    {
+      "epoch": 0.8965807469752761,
+      "grad_norm": 1.7749323844909668,
+      "learning_rate": 5.558329199273038e-06,
+      "loss": 1.8004,
+      "step": 8522
+    },
+    {
+      "epoch": 0.8966859547606523,
+      "grad_norm": 1.319456696510315,
+      "learning_rate": 5.547131721105536e-06,
+      "loss": 1.5449,
+      "step": 8523
+    },
+    {
+      "epoch": 0.8967911625460284,
+      "grad_norm": 1.598297119140625,
+      "learning_rate": 5.53594521143862e-06,
+      "loss": 1.7227,
+      "step": 8524
+    },
+    {
+      "epoch": 0.8968963703314046,
+      "grad_norm": 2.359658718109131,
+      "learning_rate": 5.524769671571317e-06,
+      "loss": 1.5778,
+      "step": 8525
+    },
+    {
+      "epoch": 0.8970015781167806,
+      "grad_norm": 1.2139389514923096,
+      "learning_rate": 5.5136051028014e-06,
+      "loss": 2.0547,
+      "step": 8526
+    },
+    {
+      "epoch": 0.8971067859021568,
+      "grad_norm": 1.9385275840759277,
+      "learning_rate": 5.50245150642541e-06,
+      "loss": 1.374,
+      "step": 8527
+    },
+    {
+      "epoch": 0.8972119936875329,
+      "grad_norm": 1.8721158504486084,
+      "learning_rate": 5.491308883738544e-06,
+      "loss": 1.3719,
+      "step": 8528
+    },
+    {
+      "epoch": 0.8973172014729089,
+      "grad_norm": 1.8050134181976318,
+      "learning_rate": 5.480177236034756e-06,
+      "loss": 1.4891,
+      "step": 8529
+    },
+    {
+      "epoch": 0.8974224092582851,
+      "grad_norm": 1.385988712310791,
+      "learning_rate": 5.469056564606767e-06,
+      "loss": 1.7196,
+      "step": 8530
+    },
+    {
+      "epoch": 0.8975276170436612,
+      "grad_norm": 1.7569900751113892,
+      "learning_rate": 5.4579468707459225e-06,
+      "loss": 1.7562,
+      "step": 8531
+    },
+    {
+      "epoch": 0.8976328248290374,
+      "grad_norm": 1.610825538635254,
+      "learning_rate": 5.446848155742401e-06,
+      "loss": 1.6719,
+      "step": 8532
+    },
+    {
+      "epoch": 0.8977380326144134,
+      "grad_norm": 2.0842642784118652,
+      "learning_rate": 5.435760420885061e-06,
+      "loss": 1.8032,
+      "step": 8533
+    },
+    {
+      "epoch": 0.8978432403997896,
+      "grad_norm": 1.5236550569534302,
+      "learning_rate": 5.42468366746145e-06,
+      "loss": 1.4211,
+      "step": 8534
+    },
+    {
+      "epoch": 0.8979484481851657,
+      "grad_norm": 1.6460191011428833,
+      "learning_rate": 5.4136178967579054e-06,
+      "loss": 1.5113,
+      "step": 8535
+    },
+    {
+      "epoch": 0.8980536559705419,
+      "grad_norm": 2.305039405822754,
+      "learning_rate": 5.402563110059456e-06,
+      "loss": 1.5592,
+      "step": 8536
+    },
+    {
+      "epoch": 0.8981588637559179,
+      "grad_norm": 1.2543623447418213,
+      "learning_rate": 5.3915193086498286e-06,
+      "loss": 1.4207,
+      "step": 8537
+    },
+    {
+      "epoch": 0.8982640715412941,
+      "grad_norm": 1.8181365728378296,
+      "learning_rate": 5.380486493811543e-06,
+      "loss": 1.5938,
+      "step": 8538
+    },
+    {
+      "epoch": 0.8983692793266702,
+      "grad_norm": 1.4921931028366089,
+      "learning_rate": 5.3694646668257855e-06,
+      "loss": 1.7251,
+      "step": 8539
+    },
+    {
+      "epoch": 0.8984744871120462,
+      "grad_norm": 1.6796791553497314,
+      "learning_rate": 5.358453828972465e-06,
+      "loss": 1.7007,
+      "step": 8540
+    },
+    {
+      "epoch": 0.8985796948974224,
+      "grad_norm": 1.4309808015823364,
+      "learning_rate": 5.3474539815302815e-06,
+      "loss": 1.6625,
+      "step": 8541
+    },
+    {
+      "epoch": 0.8986849026827985,
+      "grad_norm": 1.6457716226577759,
+      "learning_rate": 5.336465125776579e-06,
+      "loss": 1.4362,
+      "step": 8542
+    },
+    {
+      "epoch": 0.8987901104681747,
+      "grad_norm": 1.1779471635818481,
+      "learning_rate": 5.325487262987439e-06,
+      "loss": 1.5655,
+      "step": 8543
+    },
+    {
+      "epoch": 0.8988953182535507,
+      "grad_norm": 1.411882996559143,
+      "learning_rate": 5.314520394437728e-06,
+      "loss": 1.5253,
+      "step": 8544
+    },
+    {
+      "epoch": 0.8990005260389269,
+      "grad_norm": 1.2842921018600464,
+      "learning_rate": 5.303564521400961e-06,
+      "loss": 1.5975,
+      "step": 8545
+    },
+    {
+      "epoch": 0.899105733824303,
+      "grad_norm": 2.102381467819214,
+      "learning_rate": 5.292619645149433e-06,
+      "loss": 2.1535,
+      "step": 8546
+    },
+    {
+      "epoch": 0.8992109416096791,
+      "grad_norm": 2.0416359901428223,
+      "learning_rate": 5.281685766954114e-06,
+      "loss": 1.4009,
+      "step": 8547
+    },
+    {
+      "epoch": 0.8993161493950552,
+      "grad_norm": 1.6468279361724854,
+      "learning_rate": 5.270762888084712e-06,
+      "loss": 1.8449,
+      "step": 8548
+    },
+    {
+      "epoch": 0.8994213571804314,
+      "grad_norm": 1.4324162006378174,
+      "learning_rate": 5.259851009809702e-06,
+      "loss": 1.5525,
+      "step": 8549
+    },
+    {
+      "epoch": 0.8995265649658075,
+      "grad_norm": 1.8984599113464355,
+      "learning_rate": 5.2489501333962135e-06,
+      "loss": 2.262,
+      "step": 8550
+    },
+    {
+      "epoch": 0.8996317727511836,
+      "grad_norm": 1.3665709495544434,
+      "learning_rate": 5.238060260110145e-06,
+      "loss": 1.2597,
+      "step": 8551
+    },
+    {
+      "epoch": 0.8997369805365597,
+      "grad_norm": 1.834430456161499,
+      "learning_rate": 5.227181391216096e-06,
+      "loss": 1.4875,
+      "step": 8552
+    },
+    {
+      "epoch": 0.8998421883219359,
+      "grad_norm": 1.4777787923812866,
+      "learning_rate": 5.2163135279773904e-06,
+      "loss": 1.5871,
+      "step": 8553
+    },
+    {
+      "epoch": 0.8999473961073119,
+      "grad_norm": 1.73262619972229,
+      "learning_rate": 5.205456671656061e-06,
+      "loss": 1.4334,
+      "step": 8554
+    },
+    {
+      "epoch": 0.900052603892688,
+      "grad_norm": 2.0025711059570312,
+      "learning_rate": 5.194610823512913e-06,
+      "loss": 1.5515,
+      "step": 8555
+    },
+    {
+      "epoch": 0.9001578116780642,
+      "grad_norm": 2.1046454906463623,
+      "learning_rate": 5.183775984807415e-06,
+      "loss": 1.7606,
+      "step": 8556
+    },
+    {
+      "epoch": 0.9002630194634403,
+      "grad_norm": 2.093031167984009,
+      "learning_rate": 5.172952156797795e-06,
+      "loss": 1.0529,
+      "step": 8557
+    },
+    {
+      "epoch": 0.9003682272488164,
+      "grad_norm": 3.1299211978912354,
+      "learning_rate": 5.1621393407409904e-06,
+      "loss": 1.3046,
+      "step": 8558
+    },
+    {
+      "epoch": 0.9004734350341925,
+      "grad_norm": 1.444140076637268,
+      "learning_rate": 5.151337537892631e-06,
+      "loss": 1.6833,
+      "step": 8559
+    },
+    {
+      "epoch": 0.9005786428195687,
+      "grad_norm": 1.5545951128005981,
+      "learning_rate": 5.140546749507136e-06,
+      "loss": 1.5637,
+      "step": 8560
+    },
+    {
+      "epoch": 0.9006838506049447,
+      "grad_norm": 1.6997040510177612,
+      "learning_rate": 5.129766976837569e-06,
+      "loss": 1.3593,
+      "step": 8561
+    },
+    {
+      "epoch": 0.9007890583903209,
+      "grad_norm": 2.1797568798065186,
+      "learning_rate": 5.118998221135762e-06,
+      "loss": 1.5224,
+      "step": 8562
+    },
+    {
+      "epoch": 0.900894266175697,
+      "grad_norm": 1.9926127195358276,
+      "learning_rate": 5.10824048365226e-06,
+      "loss": 1.8785,
+      "step": 8563
+    },
+    {
+      "epoch": 0.9009994739610732,
+      "grad_norm": 1.50703763961792,
+      "learning_rate": 5.097493765636318e-06,
+      "loss": 1.1829,
+      "step": 8564
+    },
+    {
+      "epoch": 0.9011046817464492,
+      "grad_norm": 2.1419105529785156,
+      "learning_rate": 5.086758068335917e-06,
+      "loss": 1.2841,
+      "step": 8565
+    },
+    {
+      "epoch": 0.9012098895318253,
+      "grad_norm": 2.1330740451812744,
+      "learning_rate": 5.076033392997758e-06,
+      "loss": 1.5291,
+      "step": 8566
+    },
+    {
+      "epoch": 0.9013150973172015,
+      "grad_norm": 1.655969262123108,
+      "learning_rate": 5.06531974086728e-06,
+      "loss": 0.8968,
+      "step": 8567
+    },
+    {
+      "epoch": 0.9014203051025776,
+      "grad_norm": 1.8075084686279297,
+      "learning_rate": 5.054617113188586e-06,
+      "loss": 2.1496,
+      "step": 8568
+    },
+    {
+      "epoch": 0.9015255128879537,
+      "grad_norm": 1.51304292678833,
+      "learning_rate": 5.043925511204573e-06,
+      "loss": 1.8161,
+      "step": 8569
+    },
+    {
+      "epoch": 0.9016307206733298,
+      "grad_norm": 1.4924466609954834,
+      "learning_rate": 5.0332449361568e-06,
+      "loss": 1.731,
+      "step": 8570
+    },
+    {
+      "epoch": 0.901735928458706,
+      "grad_norm": 1.9998409748077393,
+      "learning_rate": 5.0225753892855776e-06,
+      "loss": 1.6692,
+      "step": 8571
+    },
+    {
+      "epoch": 0.901841136244082,
+      "grad_norm": 1.3246253728866577,
+      "learning_rate": 5.011916871829925e-06,
+      "loss": 1.6233,
+      "step": 8572
+    },
+    {
+      "epoch": 0.9019463440294582,
+      "grad_norm": 1.8206862211227417,
+      "learning_rate": 5.0012693850275736e-06,
+      "loss": 1.7423,
+      "step": 8573
+    },
+    {
+      "epoch": 0.9020515518148343,
+      "grad_norm": 2.139857292175293,
+      "learning_rate": 4.9906329301149914e-06,
+      "loss": 1.4421,
+      "step": 8574
+    },
+    {
+      "epoch": 0.9021567596002105,
+      "grad_norm": 2.2966580390930176,
+      "learning_rate": 4.980007508327345e-06,
+      "loss": 1.6362,
+      "step": 8575
+    },
+    {
+      "epoch": 0.9022619673855865,
+      "grad_norm": 1.746172308921814,
+      "learning_rate": 4.969393120898525e-06,
+      "loss": 1.2432,
+      "step": 8576
+    },
+    {
+      "epoch": 0.9023671751709627,
+      "grad_norm": 1.2494480609893799,
+      "learning_rate": 4.958789769061156e-06,
+      "loss": 1.7419,
+      "step": 8577
+    },
+    {
+      "epoch": 0.9024723829563388,
+      "grad_norm": 1.3508919477462769,
+      "learning_rate": 4.948197454046577e-06,
+      "loss": 1.7533,
+      "step": 8578
+    },
+    {
+      "epoch": 0.9025775907417148,
+      "grad_norm": 1.8284884691238403,
+      "learning_rate": 4.937616177084814e-06,
+      "loss": 1.5497,
+      "step": 8579
+    },
+    {
+      "epoch": 0.902682798527091,
+      "grad_norm": 1.5557414293289185,
+      "learning_rate": 4.927045939404673e-06,
+      "loss": 1.8222,
+      "step": 8580
+    },
+    {
+      "epoch": 0.9027880063124671,
+      "grad_norm": 1.6787934303283691,
+      "learning_rate": 4.916486742233606e-06,
+      "loss": 1.385,
+      "step": 8581
+    },
+    {
+      "epoch": 0.9028932140978433,
+      "grad_norm": 1.3682160377502441,
+      "learning_rate": 4.90593858679782e-06,
+      "loss": 1.768,
+      "step": 8582
+    },
+    {
+      "epoch": 0.9029984218832193,
+      "grad_norm": 2.39587140083313,
+      "learning_rate": 4.89540147432227e-06,
+      "loss": 1.8781,
+      "step": 8583
+    },
+    {
+      "epoch": 0.9031036296685955,
+      "grad_norm": 1.224987268447876,
+      "learning_rate": 4.884875406030565e-06,
+      "loss": 1.6377,
+      "step": 8584
+    },
+    {
+      "epoch": 0.9032088374539716,
+      "grad_norm": 1.3309333324432373,
+      "learning_rate": 4.874360383145072e-06,
+      "loss": 1.7443,
+      "step": 8585
+    },
+    {
+      "epoch": 0.9033140452393477,
+      "grad_norm": 3.1084442138671875,
+      "learning_rate": 4.863856406886869e-06,
+      "loss": 2.1872,
+      "step": 8586
+    },
+    {
+      "epoch": 0.9034192530247238,
+      "grad_norm": 1.980631709098816,
+      "learning_rate": 4.853363478475748e-06,
+      "loss": 1.273,
+      "step": 8587
+    },
+    {
+      "epoch": 0.9035244608101,
+      "grad_norm": 1.4033839702606201,
+      "learning_rate": 4.8428815991302005e-06,
+      "loss": 1.3123,
+      "step": 8588
+    },
+    {
+      "epoch": 0.9036296685954761,
+      "grad_norm": 1.6415311098098755,
+      "learning_rate": 4.832410770067486e-06,
+      "loss": 1.5425,
+      "step": 8589
+    },
+    {
+      "epoch": 0.9037348763808521,
+      "grad_norm": 1.605977177619934,
+      "learning_rate": 4.821950992503521e-06,
+      "loss": 1.6757,
+      "step": 8590
+    },
+    {
+      "epoch": 0.9038400841662283,
+      "grad_norm": 2.0766685009002686,
+      "learning_rate": 4.811502267652968e-06,
+      "loss": 1.7121,
+      "step": 8591
+    },
+    {
+      "epoch": 0.9039452919516044,
+      "grad_norm": 1.8514540195465088,
+      "learning_rate": 4.801064596729221e-06,
+      "loss": 1.3121,
+      "step": 8592
+    },
+    {
+      "epoch": 0.9040504997369805,
+      "grad_norm": 1.8307629823684692,
+      "learning_rate": 4.790637980944346e-06,
+      "loss": 1.871,
+      "step": 8593
+    },
+    {
+      "epoch": 0.9041557075223566,
+      "grad_norm": 1.437143325805664,
+      "learning_rate": 4.780222421509184e-06,
+      "loss": 1.8685,
+      "step": 8594
+    },
+    {
+      "epoch": 0.9042609153077328,
+      "grad_norm": 1.956531047821045,
+      "learning_rate": 4.769817919633235e-06,
+      "loss": 1.5374,
+      "step": 8595
+    },
+    {
+      "epoch": 0.9043661230931089,
+      "grad_norm": 1.2559046745300293,
+      "learning_rate": 4.759424476524732e-06,
+      "loss": 2.0446,
+      "step": 8596
+    },
+    {
+      "epoch": 0.904471330878485,
+      "grad_norm": 1.7142302989959717,
+      "learning_rate": 4.749042093390654e-06,
+      "loss": 1.7011,
+      "step": 8597
+    },
+    {
+      "epoch": 0.9045765386638611,
+      "grad_norm": 2.1067309379577637,
+      "learning_rate": 4.738670771436671e-06,
+      "loss": 1.5651,
+      "step": 8598
+    },
+    {
+      "epoch": 0.9046817464492373,
+      "grad_norm": 1.2750693559646606,
+      "learning_rate": 4.72831051186714e-06,
+      "loss": 1.9008,
+      "step": 8599
+    },
+    {
+      "epoch": 0.9047869542346134,
+      "grad_norm": 1.5802183151245117,
+      "learning_rate": 4.717961315885211e-06,
+      "loss": 2.3383,
+      "step": 8600
+    },
+    {
+      "epoch": 0.9048921620199895,
+      "grad_norm": 1.7453382015228271,
+      "learning_rate": 4.707623184692655e-06,
+      "loss": 1.7165,
+      "step": 8601
+    },
+    {
+      "epoch": 0.9049973698053656,
+      "grad_norm": 1.6323179006576538,
+      "learning_rate": 4.697296119490047e-06,
+      "loss": 1.3792,
+      "step": 8602
+    },
+    {
+      "epoch": 0.9051025775907418,
+      "grad_norm": 1.3392667770385742,
+      "learning_rate": 4.686980121476614e-06,
+      "loss": 1.369,
+      "step": 8603
+    },
+    {
+      "epoch": 0.9052077853761178,
+      "grad_norm": 2.2435269355773926,
+      "learning_rate": 4.67667519185031e-06,
+      "loss": 1.8554,
+      "step": 8604
+    },
+    {
+      "epoch": 0.9053129931614939,
+      "grad_norm": 1.6639388799667358,
+      "learning_rate": 4.666381331807834e-06,
+      "loss": 1.2958,
+      "step": 8605
+    },
+    {
+      "epoch": 0.9054182009468701,
+      "grad_norm": 1.4905084371566772,
+      "learning_rate": 4.656098542544574e-06,
+      "loss": 1.7928,
+      "step": 8606
+    },
+    {
+      "epoch": 0.9055234087322462,
+      "grad_norm": 1.625221848487854,
+      "learning_rate": 4.645826825254607e-06,
+      "loss": 1.9308,
+      "step": 8607
+    },
+    {
+      "epoch": 0.9056286165176223,
+      "grad_norm": 1.763044834136963,
+      "learning_rate": 4.6355661811308015e-06,
+      "loss": 1.7369,
+      "step": 8608
+    },
+    {
+      "epoch": 0.9057338243029984,
+      "grad_norm": 3.1937339305877686,
+      "learning_rate": 4.625316611364661e-06,
+      "loss": 1.1817,
+      "step": 8609
+    },
+    {
+      "epoch": 0.9058390320883746,
+      "grad_norm": 1.484457015991211,
+      "learning_rate": 4.615078117146421e-06,
+      "loss": 1.6794,
+      "step": 8610
+    },
+    {
+      "epoch": 0.9059442398737506,
+      "grad_norm": 1.7743538618087769,
+      "learning_rate": 4.604850699665087e-06,
+      "loss": 2.1803,
+      "step": 8611
+    },
+    {
+      "epoch": 0.9060494476591268,
+      "grad_norm": 2.6126396656036377,
+      "learning_rate": 4.594634360108319e-06,
+      "loss": 1.6458,
+      "step": 8612
+    },
+    {
+      "epoch": 0.9061546554445029,
+      "grad_norm": 2.2468442916870117,
+      "learning_rate": 4.584429099662468e-06,
+      "loss": 1.8922,
+      "step": 8613
+    },
+    {
+      "epoch": 0.9062598632298791,
+      "grad_norm": 2.1800897121429443,
+      "learning_rate": 4.574234919512698e-06,
+      "loss": 1.5657,
+      "step": 8614
+    },
+    {
+      "epoch": 0.9063650710152551,
+      "grad_norm": 2.317814826965332,
+      "learning_rate": 4.564051820842796e-06,
+      "loss": 1.7934,
+      "step": 8615
+    },
+    {
+      "epoch": 0.9064702788006312,
+      "grad_norm": 1.8590099811553955,
+      "learning_rate": 4.553879804835282e-06,
+      "loss": 2.0423,
+      "step": 8616
+    },
+    {
+      "epoch": 0.9065754865860074,
+      "grad_norm": 1.0570123195648193,
+      "learning_rate": 4.543718872671421e-06,
+      "loss": 2.114,
+      "step": 8617
+    },
+    {
+      "epoch": 0.9066806943713834,
+      "grad_norm": 1.6926145553588867,
+      "learning_rate": 4.533569025531137e-06,
+      "loss": 1.5145,
+      "step": 8618
+    },
+    {
+      "epoch": 0.9067859021567596,
+      "grad_norm": 2.080049514770508,
+      "learning_rate": 4.523430264593132e-06,
+      "loss": 2.0468,
+      "step": 8619
+    },
+    {
+      "epoch": 0.9068911099421357,
+      "grad_norm": 1.7230383157730103,
+      "learning_rate": 4.5133025910347845e-06,
+      "loss": 1.407,
+      "step": 8620
+    },
+    {
+      "epoch": 0.9069963177275119,
+      "grad_norm": 1.7170302867889404,
+      "learning_rate": 4.5031860060321455e-06,
+      "loss": 1.0797,
+      "step": 8621
+    },
+    {
+      "epoch": 0.9071015255128879,
+      "grad_norm": 1.9875949621200562,
+      "learning_rate": 4.493080510760083e-06,
+      "loss": 1.2423,
+      "step": 8622
+    },
+    {
+      "epoch": 0.9072067332982641,
+      "grad_norm": 2.1834206581115723,
+      "learning_rate": 4.482986106392073e-06,
+      "loss": 1.2699,
+      "step": 8623
+    },
+    {
+      "epoch": 0.9073119410836402,
+      "grad_norm": 1.1872564554214478,
+      "learning_rate": 4.472902794100342e-06,
+      "loss": 1.7113,
+      "step": 8624
+    },
+    {
+      "epoch": 0.9074171488690163,
+      "grad_norm": 1.889709711074829,
+      "learning_rate": 4.4628305750558656e-06,
+      "loss": 1.9109,
+      "step": 8625
+    },
+    {
+      "epoch": 0.9075223566543924,
+      "grad_norm": 1.6247581243515015,
+      "learning_rate": 4.452769450428273e-06,
+      "loss": 1.1514,
+      "step": 8626
+    },
+    {
+      "epoch": 0.9076275644397686,
+      "grad_norm": 1.985897183418274,
+      "learning_rate": 4.442719421385922e-06,
+      "loss": 1.7738,
+      "step": 8627
+    },
+    {
+      "epoch": 0.9077327722251447,
+      "grad_norm": 1.4728162288665771,
+      "learning_rate": 4.4326804890959195e-06,
+      "loss": 1.8999,
+      "step": 8628
+    },
+    {
+      "epoch": 0.9078379800105207,
+      "grad_norm": 1.9778163433074951,
+      "learning_rate": 4.422652654724036e-06,
+      "loss": 2.0382,
+      "step": 8629
+    },
+    {
+      "epoch": 0.9079431877958969,
+      "grad_norm": 1.3184362649917603,
+      "learning_rate": 4.412635919434749e-06,
+      "loss": 1.6149,
+      "step": 8630
+    },
+    {
+      "epoch": 0.908048395581273,
+      "grad_norm": 1.9387370347976685,
+      "learning_rate": 4.402630284391318e-06,
+      "loss": 1.4208,
+      "step": 8631
+    },
+    {
+      "epoch": 0.9081536033666492,
+      "grad_norm": 1.576529860496521,
+      "learning_rate": 4.392635750755625e-06,
+      "loss": 1.4622,
+      "step": 8632
+    },
+    {
+      "epoch": 0.9082588111520252,
+      "grad_norm": 2.166128396987915,
+      "learning_rate": 4.382652319688307e-06,
+      "loss": 2.017,
+      "step": 8633
+    },
+    {
+      "epoch": 0.9083640189374014,
+      "grad_norm": 1.8880326747894287,
+      "learning_rate": 4.372679992348727e-06,
+      "loss": 1.7115,
+      "step": 8634
+    },
+    {
+      "epoch": 0.9084692267227775,
+      "grad_norm": 1.9160791635513306,
+      "learning_rate": 4.362718769894925e-06,
+      "loss": 2.233,
+      "step": 8635
+    },
+    {
+      "epoch": 0.9085744345081536,
+      "grad_norm": 1.0861622095108032,
+      "learning_rate": 4.352768653483652e-06,
+      "loss": 1.4918,
+      "step": 8636
+    },
+    {
+      "epoch": 0.9086796422935297,
+      "grad_norm": 1.6084988117218018,
+      "learning_rate": 4.342829644270429e-06,
+      "loss": 1.69,
+      "step": 8637
+    },
+    {
+      "epoch": 0.9087848500789059,
+      "grad_norm": 1.7415450811386108,
+      "learning_rate": 4.332901743409379e-06,
+      "loss": 1.4335,
+      "step": 8638
+    },
+    {
+      "epoch": 0.908890057864282,
+      "grad_norm": 1.4303922653198242,
+      "learning_rate": 4.322984952053433e-06,
+      "loss": 1.3406,
+      "step": 8639
+    },
+    {
+      "epoch": 0.908995265649658,
+      "grad_norm": 1.2648975849151611,
+      "learning_rate": 4.313079271354192e-06,
+      "loss": 1.3649,
+      "step": 8640
+    },
+    {
+      "epoch": 0.9091004734350342,
+      "grad_norm": 1.7955864667892456,
+      "learning_rate": 4.303184702461948e-06,
+      "loss": 1.5743,
+      "step": 8641
+    },
+    {
+      "epoch": 0.9092056812204103,
+      "grad_norm": 1.23116934299469,
+      "learning_rate": 4.293301246525761e-06,
+      "loss": 1.4245,
+      "step": 8642
+    },
+    {
+      "epoch": 0.9093108890057864,
+      "grad_norm": 1.7114944458007812,
+      "learning_rate": 4.283428904693343e-06,
+      "loss": 1.8552,
+      "step": 8643
+    },
+    {
+      "epoch": 0.9094160967911625,
+      "grad_norm": 1.141519546508789,
+      "learning_rate": 4.273567678111123e-06,
+      "loss": 1.773,
+      "step": 8644
+    },
+    {
+      "epoch": 0.9095213045765387,
+      "grad_norm": 1.2643916606903076,
+      "learning_rate": 4.263717567924286e-06,
+      "loss": 1.8676,
+      "step": 8645
+    },
+    {
+      "epoch": 0.9096265123619148,
+      "grad_norm": 1.395230770111084,
+      "learning_rate": 4.2538785752766816e-06,
+      "loss": 1.591,
+      "step": 8646
+    },
+    {
+      "epoch": 0.9097317201472909,
+      "grad_norm": 1.6044540405273438,
+      "learning_rate": 4.244050701310853e-06,
+      "loss": 2.1957,
+      "step": 8647
+    },
+    {
+      "epoch": 0.909836927932667,
+      "grad_norm": 1.558889389038086,
+      "learning_rate": 4.234233947168109e-06,
+      "loss": 1.3307,
+      "step": 8648
+    },
+    {
+      "epoch": 0.9099421357180432,
+      "grad_norm": 2.091890811920166,
+      "learning_rate": 4.224428313988416e-06,
+      "loss": 1.092,
+      "step": 8649
+    },
+    {
+      "epoch": 0.9100473435034192,
+      "grad_norm": 2.1650590896606445,
+      "learning_rate": 4.214633802910506e-06,
+      "loss": 1.9047,
+      "step": 8650
+    },
+    {
+      "epoch": 0.9101525512887954,
+      "grad_norm": 1.981006145477295,
+      "learning_rate": 4.204850415071748e-06,
+      "loss": 1.1051,
+      "step": 8651
+    },
+    {
+      "epoch": 0.9102577590741715,
+      "grad_norm": 1.77347993850708,
+      "learning_rate": 4.195078151608256e-06,
+      "loss": 1.6043,
+      "step": 8652
+    },
+    {
+      "epoch": 0.9103629668595477,
+      "grad_norm": 3.0252108573913574,
+      "learning_rate": 4.185317013654866e-06,
+      "loss": 1.4648,
+      "step": 8653
+    },
+    {
+      "epoch": 0.9104681746449237,
+      "grad_norm": 1.7558192014694214,
+      "learning_rate": 4.175567002345104e-06,
+      "loss": 1.2865,
+      "step": 8654
+    },
+    {
+      "epoch": 0.9105733824302998,
+      "grad_norm": 1.6132521629333496,
+      "learning_rate": 4.165828118811199e-06,
+      "loss": 1.5495,
+      "step": 8655
+    },
+    {
+      "epoch": 0.910678590215676,
+      "grad_norm": 1.244947910308838,
+      "learning_rate": 4.156100364184101e-06,
+      "loss": 2.0034,
+      "step": 8656
+    },
+    {
+      "epoch": 0.910783798001052,
+      "grad_norm": 1.8241870403289795,
+      "learning_rate": 4.146383739593474e-06,
+      "loss": 1.6338,
+      "step": 8657
+    },
+    {
+      "epoch": 0.9108890057864282,
+      "grad_norm": 0.9370825290679932,
+      "learning_rate": 4.136678246167636e-06,
+      "loss": 1.6983,
+      "step": 8658
+    },
+    {
+      "epoch": 0.9109942135718043,
+      "grad_norm": 1.7532540559768677,
+      "learning_rate": 4.12698388503372e-06,
+      "loss": 1.5103,
+      "step": 8659
+    },
+    {
+      "epoch": 0.9110994213571805,
+      "grad_norm": 1.93152916431427,
+      "learning_rate": 4.1173006573174354e-06,
+      "loss": 1.9619,
+      "step": 8660
+    },
+    {
+      "epoch": 0.9112046291425565,
+      "grad_norm": 1.5335841178894043,
+      "learning_rate": 4.107628564143306e-06,
+      "loss": 1.519,
+      "step": 8661
+    },
+    {
+      "epoch": 0.9113098369279327,
+      "grad_norm": 1.8598703145980835,
+      "learning_rate": 4.0979676066345005e-06,
+      "loss": 1.3632,
+      "step": 8662
+    },
+    {
+      "epoch": 0.9114150447133088,
+      "grad_norm": 1.7030982971191406,
+      "learning_rate": 4.08831778591291e-06,
+      "loss": 1.0882,
+      "step": 8663
+    },
+    {
+      "epoch": 0.911520252498685,
+      "grad_norm": 1.5589040517807007,
+      "learning_rate": 4.078679103099159e-06,
+      "loss": 2.0111,
+      "step": 8664
+    },
+    {
+      "epoch": 0.911625460284061,
+      "grad_norm": 1.4710445404052734,
+      "learning_rate": 4.069051559312531e-06,
+      "loss": 1.8036,
+      "step": 8665
+    },
+    {
+      "epoch": 0.9117306680694371,
+      "grad_norm": 2.0354957580566406,
+      "learning_rate": 4.0594351556710544e-06,
+      "loss": 1.6073,
+      "step": 8666
+    },
+    {
+      "epoch": 0.9118358758548133,
+      "grad_norm": 1.0362532138824463,
+      "learning_rate": 4.049829893291457e-06,
+      "loss": 1.689,
+      "step": 8667
+    },
+    {
+      "epoch": 0.9119410836401893,
+      "grad_norm": 1.7926151752471924,
+      "learning_rate": 4.040235773289147e-06,
+      "loss": 1.7301,
+      "step": 8668
+    },
+    {
+      "epoch": 0.9120462914255655,
+      "grad_norm": 1.6435024738311768,
+      "learning_rate": 4.030652796778267e-06,
+      "loss": 1.5168,
+      "step": 8669
+    },
+    {
+      "epoch": 0.9121514992109416,
+      "grad_norm": 1.7422984838485718,
+      "learning_rate": 4.021080964871671e-06,
+      "loss": 1.7809,
+      "step": 8670
+    },
+    {
+      "epoch": 0.9122567069963178,
+      "grad_norm": 2.355043411254883,
+      "learning_rate": 4.011520278680891e-06,
+      "loss": 1.9162,
+      "step": 8671
+    },
+    {
+      "epoch": 0.9123619147816938,
+      "grad_norm": 1.7827218770980835,
+      "learning_rate": 4.001970739316163e-06,
+      "loss": 1.322,
+      "step": 8672
+    },
+    {
+      "epoch": 0.91246712256707,
+      "grad_norm": 1.2466920614242554,
+      "learning_rate": 3.9924323478864655e-06,
+      "loss": 1.2959,
+      "step": 8673
+    },
+    {
+      "epoch": 0.9125723303524461,
+      "grad_norm": 1.3089375495910645,
+      "learning_rate": 3.982905105499468e-06,
+      "loss": 1.3969,
+      "step": 8674
+    },
+    {
+      "epoch": 0.9126775381378222,
+      "grad_norm": 1.8237321376800537,
+      "learning_rate": 3.973389013261497e-06,
+      "loss": 1.2769,
+      "step": 8675
+    },
+    {
+      "epoch": 0.9127827459231983,
+      "grad_norm": 1.5145193338394165,
+      "learning_rate": 3.9638840722776685e-06,
+      "loss": 1.4357,
+      "step": 8676
+    },
+    {
+      "epoch": 0.9128879537085745,
+      "grad_norm": 1.9698090553283691,
+      "learning_rate": 3.954390283651754e-06,
+      "loss": 1.4994,
+      "step": 8677
+    },
+    {
+      "epoch": 0.9129931614939506,
+      "grad_norm": 1.725354790687561,
+      "learning_rate": 3.944907648486196e-06,
+      "loss": 1.4207,
+      "step": 8678
+    },
+    {
+      "epoch": 0.9130983692793266,
+      "grad_norm": 2.0058798789978027,
+      "learning_rate": 3.935436167882234e-06,
+      "loss": 1.1381,
+      "step": 8679
+    },
+    {
+      "epoch": 0.9132035770647028,
+      "grad_norm": 1.658754587173462,
+      "learning_rate": 3.925975842939733e-06,
+      "loss": 1.7292,
+      "step": 8680
+    },
+    {
+      "epoch": 0.9133087848500789,
+      "grad_norm": 1.5891391038894653,
+      "learning_rate": 3.916526674757293e-06,
+      "loss": 1.0397,
+      "step": 8681
+    },
+    {
+      "epoch": 0.913413992635455,
+      "grad_norm": 1.2456856966018677,
+      "learning_rate": 3.907088664432224e-06,
+      "loss": 2.1817,
+      "step": 8682
+    },
+    {
+      "epoch": 0.9135192004208311,
+      "grad_norm": 1.409684658050537,
+      "learning_rate": 3.897661813060494e-06,
+      "loss": 1.3057,
+      "step": 8683
+    },
+    {
+      "epoch": 0.9136244082062073,
+      "grad_norm": 1.887300968170166,
+      "learning_rate": 3.8882461217368604e-06,
+      "loss": 1.7419,
+      "step": 8684
+    },
+    {
+      "epoch": 0.9137296159915834,
+      "grad_norm": 1.338424801826477,
+      "learning_rate": 3.878841591554716e-06,
+      "loss": 1.6628,
+      "step": 8685
+    },
+    {
+      "epoch": 0.9138348237769595,
+      "grad_norm": 1.1566272974014282,
+      "learning_rate": 3.869448223606165e-06,
+      "loss": 1.7944,
+      "step": 8686
+    },
+    {
+      "epoch": 0.9139400315623356,
+      "grad_norm": 1.7602925300598145,
+      "learning_rate": 3.860066018982056e-06,
+      "loss": 1.2639,
+      "step": 8687
+    },
+    {
+      "epoch": 0.9140452393477118,
+      "grad_norm": 1.099798560142517,
+      "learning_rate": 3.850694978771896e-06,
+      "loss": 1.5982,
+      "step": 8688
+    },
+    {
+      "epoch": 0.9141504471330878,
+      "grad_norm": 1.6427594423294067,
+      "learning_rate": 3.841335104063904e-06,
+      "loss": 2.0918,
+      "step": 8689
+    },
+    {
+      "epoch": 0.914255654918464,
+      "grad_norm": 1.4541171789169312,
+      "learning_rate": 3.831986395945042e-06,
+      "loss": 1.8839,
+      "step": 8690
+    },
+    {
+      "epoch": 0.9143608627038401,
+      "grad_norm": 1.9176745414733887,
+      "learning_rate": 3.822648855500921e-06,
+      "loss": 1.5762,
+      "step": 8691
+    },
+    {
+      "epoch": 0.9144660704892162,
+      "grad_norm": 1.3941978216171265,
+      "learning_rate": 3.813322483815862e-06,
+      "loss": 1.6778,
+      "step": 8692
+    },
+    {
+      "epoch": 0.9145712782745923,
+      "grad_norm": 2.07476544380188,
+      "learning_rate": 3.8040072819729545e-06,
+      "loss": 1.2133,
+      "step": 8693
+    },
+    {
+      "epoch": 0.9146764860599684,
+      "grad_norm": 1.7748063802719116,
+      "learning_rate": 3.794703251053899e-06,
+      "loss": 1.7154,
+      "step": 8694
+    },
+    {
+      "epoch": 0.9147816938453446,
+      "grad_norm": 1.5125572681427002,
+      "learning_rate": 3.785410392139166e-06,
+      "loss": 1.6641,
+      "step": 8695
+    },
+    {
+      "epoch": 0.9148869016307207,
+      "grad_norm": 2.81387996673584,
+      "learning_rate": 3.776128706307902e-06,
+      "loss": 1.3597,
+      "step": 8696
+    },
+    {
+      "epoch": 0.9149921094160968,
+      "grad_norm": 2.081625461578369,
+      "learning_rate": 3.7668581946379345e-06,
+      "loss": 1.3875,
+      "step": 8697
+    },
+    {
+      "epoch": 0.9150973172014729,
+      "grad_norm": 1.5099921226501465,
+      "learning_rate": 3.7575988582058575e-06,
+      "loss": 2.1729,
+      "step": 8698
+    },
+    {
+      "epoch": 0.9152025249868491,
+      "grad_norm": 1.3036530017852783,
+      "learning_rate": 3.7483506980868997e-06,
+      "loss": 0.8875,
+      "step": 8699
+    },
+    {
+      "epoch": 0.9153077327722251,
+      "grad_norm": 1.566401720046997,
+      "learning_rate": 3.7391137153550137e-06,
+      "loss": 1.5267,
+      "step": 8700
+    },
+    {
+      "epoch": 0.9154129405576013,
+      "grad_norm": 1.6144717931747437,
+      "learning_rate": 3.7298879110828965e-06,
+      "loss": 1.6758,
+      "step": 8701
+    },
+    {
+      "epoch": 0.9155181483429774,
+      "grad_norm": 1.2582496404647827,
+      "learning_rate": 3.7206732863418804e-06,
+      "loss": 1.6363,
+      "step": 8702
+    },
+    {
+      "epoch": 0.9156233561283535,
+      "grad_norm": 1.8955252170562744,
+      "learning_rate": 3.71146984220202e-06,
+      "loss": 1.6015,
+      "step": 8703
+    },
+    {
+      "epoch": 0.9157285639137296,
+      "grad_norm": 1.6370620727539062,
+      "learning_rate": 3.702277579732116e-06,
+      "loss": 1.4619,
+      "step": 8704
+    },
+    {
+      "epoch": 0.9158337716991057,
+      "grad_norm": 2.216202974319458,
+      "learning_rate": 3.6930964999995933e-06,
+      "loss": 1.4897,
+      "step": 8705
+    },
+    {
+      "epoch": 0.9159389794844819,
+      "grad_norm": 1.8498296737670898,
+      "learning_rate": 3.683926604070653e-06,
+      "loss": 1.7555,
+      "step": 8706
+    },
+    {
+      "epoch": 0.9160441872698579,
+      "grad_norm": 2.3383429050445557,
+      "learning_rate": 3.6747678930101558e-06,
+      "loss": 1.4773,
+      "step": 8707
+    },
+    {
+      "epoch": 0.9161493950552341,
+      "grad_norm": 1.3877893686294556,
+      "learning_rate": 3.6656203678816723e-06,
+      "loss": 1.9327,
+      "step": 8708
+    },
+    {
+      "epoch": 0.9162546028406102,
+      "grad_norm": 1.5240907669067383,
+      "learning_rate": 3.6564840297474757e-06,
+      "loss": 2.1459,
+      "step": 8709
+    },
+    {
+      "epoch": 0.9163598106259864,
+      "grad_norm": 1.7055188417434692,
+      "learning_rate": 3.6473588796685386e-06,
+      "loss": 1.5657,
+      "step": 8710
+    },
+    {
+      "epoch": 0.9164650184113624,
+      "grad_norm": 2.1579809188842773,
+      "learning_rate": 3.6382449187045144e-06,
+      "loss": 1.5808,
+      "step": 8711
+    },
+    {
+      "epoch": 0.9165702261967386,
+      "grad_norm": 1.276623010635376,
+      "learning_rate": 3.629142147913811e-06,
+      "loss": 1.9783,
+      "step": 8712
+    },
+    {
+      "epoch": 0.9166754339821147,
+      "grad_norm": 2.286100149154663,
+      "learning_rate": 3.6200505683534945e-06,
+      "loss": 1.714,
+      "step": 8713
+    },
+    {
+      "epoch": 0.9167806417674907,
+      "grad_norm": 2.115018367767334,
+      "learning_rate": 3.6109701810793208e-06,
+      "loss": 1.7412,
+      "step": 8714
+    },
+    {
+      "epoch": 0.9168858495528669,
+      "grad_norm": 1.6776834726333618,
+      "learning_rate": 3.6019009871457897e-06,
+      "loss": 1.715,
+      "step": 8715
+    },
+    {
+      "epoch": 0.916991057338243,
+      "grad_norm": 1.3437702655792236,
+      "learning_rate": 3.59284298760606e-06,
+      "loss": 1.1101,
+      "step": 8716
+    },
+    {
+      "epoch": 0.9170962651236192,
+      "grad_norm": 3.0596518516540527,
+      "learning_rate": 3.5837961835120006e-06,
+      "loss": 1.4793,
+      "step": 8717
+    },
+    {
+      "epoch": 0.9172014729089952,
+      "grad_norm": 1.5906575918197632,
+      "learning_rate": 3.5747605759142157e-06,
+      "loss": 1.6282,
+      "step": 8718
+    },
+    {
+      "epoch": 0.9173066806943714,
+      "grad_norm": 2.017958402633667,
+      "learning_rate": 3.565736165861966e-06,
+      "loss": 1.9276,
+      "step": 8719
+    },
+    {
+      "epoch": 0.9174118884797475,
+      "grad_norm": 2.3949334621429443,
+      "learning_rate": 3.5567229544032133e-06,
+      "loss": 0.857,
+      "step": 8720
+    },
+    {
+      "epoch": 0.9175170962651236,
+      "grad_norm": 3.835064649581909,
+      "learning_rate": 3.5477209425846538e-06,
+      "loss": 1.3035,
+      "step": 8721
+    },
+    {
+      "epoch": 0.9176223040504997,
+      "grad_norm": 1.541480302810669,
+      "learning_rate": 3.538730131451651e-06,
+      "loss": 1.6198,
+      "step": 8722
+    },
+    {
+      "epoch": 0.9177275118358759,
+      "grad_norm": 1.4783751964569092,
+      "learning_rate": 3.529750522048281e-06,
+      "loss": 1.5319,
+      "step": 8723
+    },
+    {
+      "epoch": 0.917832719621252,
+      "grad_norm": 1.9219422340393066,
+      "learning_rate": 3.5207821154173093e-06,
+      "loss": 1.6938,
+      "step": 8724
+    },
+    {
+      "epoch": 0.917937927406628,
+      "grad_norm": 1.5939526557922363,
+      "learning_rate": 3.5118249126002035e-06,
+      "loss": 1.4352,
+      "step": 8725
+    },
+    {
+      "epoch": 0.9180431351920042,
+      "grad_norm": 1.3447009325027466,
+      "learning_rate": 3.5028789146371533e-06,
+      "loss": 1.7133,
+      "step": 8726
+    },
+    {
+      "epoch": 0.9181483429773803,
+      "grad_norm": 1.859089970588684,
+      "learning_rate": 3.4939441225670054e-06,
+      "loss": 1.8013,
+      "step": 8727
+    },
+    {
+      "epoch": 0.9182535507627565,
+      "grad_norm": 1.8480535745620728,
+      "learning_rate": 3.4850205374273416e-06,
+      "loss": 1.5742,
+      "step": 8728
+    },
+    {
+      "epoch": 0.9183587585481325,
+      "grad_norm": 2.163987159729004,
+      "learning_rate": 3.476108160254443e-06,
+      "loss": 1.2029,
+      "step": 8729
+    },
+    {
+      "epoch": 0.9184639663335087,
+      "grad_norm": 1.8570356369018555,
+      "learning_rate": 3.4672069920832493e-06,
+      "loss": 1.558,
+      "step": 8730
+    },
+    {
+      "epoch": 0.9185691741188848,
+      "grad_norm": 1.9687753915786743,
+      "learning_rate": 3.4583170339474223e-06,
+      "loss": 1.6714,
+      "step": 8731
+    },
+    {
+      "epoch": 0.9186743819042609,
+      "grad_norm": 1.7975976467132568,
+      "learning_rate": 3.4494382868793474e-06,
+      "loss": 1.623,
+      "step": 8732
+    },
+    {
+      "epoch": 0.918779589689637,
+      "grad_norm": 1.992436170578003,
+      "learning_rate": 3.440570751910066e-06,
+      "loss": 1.2232,
+      "step": 8733
+    },
+    {
+      "epoch": 0.9188847974750132,
+      "grad_norm": 1.5416083335876465,
+      "learning_rate": 3.4317144300693328e-06,
+      "loss": 1.3667,
+      "step": 8734
+    },
+    {
+      "epoch": 0.9189900052603893,
+      "grad_norm": 2.0780837535858154,
+      "learning_rate": 3.4228693223856136e-06,
+      "loss": 1.6205,
+      "step": 8735
+    },
+    {
+      "epoch": 0.9190952130457654,
+      "grad_norm": 2.3178086280822754,
+      "learning_rate": 3.4140354298860756e-06,
+      "loss": 1.6387,
+      "step": 8736
+    },
+    {
+      "epoch": 0.9192004208311415,
+      "grad_norm": 2.03950834274292,
+      "learning_rate": 3.405212753596532e-06,
+      "loss": 1.2235,
+      "step": 8737
+    },
+    {
+      "epoch": 0.9193056286165177,
+      "grad_norm": 1.5267329216003418,
+      "learning_rate": 3.3964012945415624e-06,
+      "loss": 1.9435,
+      "step": 8738
+    },
+    {
+      "epoch": 0.9194108364018937,
+      "grad_norm": 1.4271742105484009,
+      "learning_rate": 3.3876010537444046e-06,
+      "loss": 1.3572,
+      "step": 8739
+    },
+    {
+      "epoch": 0.9195160441872698,
+      "grad_norm": 1.5185612440109253,
+      "learning_rate": 3.3788120322269855e-06,
+      "loss": 1.1357,
+      "step": 8740
+    },
+    {
+      "epoch": 0.919621251972646,
+      "grad_norm": 2.3991973400115967,
+      "learning_rate": 3.3700342310099773e-06,
+      "loss": 2.0348,
+      "step": 8741
+    },
+    {
+      "epoch": 0.9197264597580221,
+      "grad_norm": 1.1119378805160522,
+      "learning_rate": 3.361267651112676e-06,
+      "loss": 1.4776,
+      "step": 8742
+    },
+    {
+      "epoch": 0.9198316675433982,
+      "grad_norm": 2.3870489597320557,
+      "learning_rate": 3.3525122935531562e-06,
+      "loss": 1.8226,
+      "step": 8743
+    },
+    {
+      "epoch": 0.9199368753287743,
+      "grad_norm": 1.734883189201355,
+      "learning_rate": 3.3437681593481486e-06,
+      "loss": 1.6325,
+      "step": 8744
+    },
+    {
+      "epoch": 0.9200420831141505,
+      "grad_norm": 1.6427862644195557,
+      "learning_rate": 3.3350352495130298e-06,
+      "loss": 1.6529,
+      "step": 8745
+    },
+    {
+      "epoch": 0.9201472908995265,
+      "grad_norm": 1.838613748550415,
+      "learning_rate": 3.326313565061978e-06,
+      "loss": 1.3113,
+      "step": 8746
+    },
+    {
+      "epoch": 0.9202524986849027,
+      "grad_norm": 2.063891887664795,
+      "learning_rate": 3.3176031070077827e-06,
+      "loss": 2.1088,
+      "step": 8747
+    },
+    {
+      "epoch": 0.9203577064702788,
+      "grad_norm": 1.299898624420166,
+      "learning_rate": 3.3089038763619684e-06,
+      "loss": 1.7546,
+      "step": 8748
+    },
+    {
+      "epoch": 0.920462914255655,
+      "grad_norm": 1.4041413068771362,
+      "learning_rate": 3.30021587413476e-06,
+      "loss": 2.0231,
+      "step": 8749
+    },
+    {
+      "epoch": 0.920568122041031,
+      "grad_norm": 2.1937994956970215,
+      "learning_rate": 3.2915391013350392e-06,
+      "loss": 2.0115,
+      "step": 8750
+    },
+    {
+      "epoch": 0.9206733298264071,
+      "grad_norm": 1.2597063779830933,
+      "learning_rate": 3.282873558970445e-06,
+      "loss": 1.7548,
+      "step": 8751
+    },
+    {
+      "epoch": 0.9207785376117833,
+      "grad_norm": 1.6283094882965088,
+      "learning_rate": 3.2742192480472724e-06,
+      "loss": 1.5641,
+      "step": 8752
+    },
+    {
+      "epoch": 0.9208837453971593,
+      "grad_norm": 1.535847783088684,
+      "learning_rate": 3.2655761695704834e-06,
+      "loss": 1.769,
+      "step": 8753
+    },
+    {
+      "epoch": 0.9209889531825355,
+      "grad_norm": 1.572363018989563,
+      "learning_rate": 3.256944324543809e-06,
+      "loss": 1.5744,
+      "step": 8754
+    },
+    {
+      "epoch": 0.9210941609679116,
+      "grad_norm": 1.8032740354537964,
+      "learning_rate": 3.2483237139696255e-06,
+      "loss": 1.1064,
+      "step": 8755
+    },
+    {
+      "epoch": 0.9211993687532878,
+      "grad_norm": 1.5819141864776611,
+      "learning_rate": 3.2397143388489983e-06,
+      "loss": 1.3802,
+      "step": 8756
+    },
+    {
+      "epoch": 0.9213045765386638,
+      "grad_norm": 1.1894673109054565,
+      "learning_rate": 3.2311162001817387e-06,
+      "loss": 1.4237,
+      "step": 8757
+    },
+    {
+      "epoch": 0.92140978432404,
+      "grad_norm": 1.9379583597183228,
+      "learning_rate": 3.2225292989662925e-06,
+      "loss": 1.9112,
+      "step": 8758
+    },
+    {
+      "epoch": 0.9215149921094161,
+      "grad_norm": 1.32359778881073,
+      "learning_rate": 3.21395363619984e-06,
+      "loss": 1.3942,
+      "step": 8759
+    },
+    {
+      "epoch": 0.9216201998947923,
+      "grad_norm": 1.8300716876983643,
+      "learning_rate": 3.2053892128782403e-06,
+      "loss": 1.7214,
+      "step": 8760
+    },
+    {
+      "epoch": 0.9217254076801683,
+      "grad_norm": 1.1215351819992065,
+      "learning_rate": 3.1968360299960643e-06,
+      "loss": 1.3853,
+      "step": 8761
+    },
+    {
+      "epoch": 0.9218306154655445,
+      "grad_norm": 1.482314944267273,
+      "learning_rate": 3.1882940885465397e-06,
+      "loss": 1.6937,
+      "step": 8762
+    },
+    {
+      "epoch": 0.9219358232509206,
+      "grad_norm": 1.5101746320724487,
+      "learning_rate": 3.1797633895216394e-06,
+      "loss": 1.7438,
+      "step": 8763
+    },
+    {
+      "epoch": 0.9220410310362966,
+      "grad_norm": 1.2713983058929443,
+      "learning_rate": 3.171243933911994e-06,
+      "loss": 1.5315,
+      "step": 8764
+    },
+    {
+      "epoch": 0.9221462388216728,
+      "grad_norm": 2.0249693393707275,
+      "learning_rate": 3.1627357227069333e-06,
+      "loss": 1.4189,
+      "step": 8765
+    },
+    {
+      "epoch": 0.9222514466070489,
+      "grad_norm": 1.0391638278961182,
+      "learning_rate": 3.154238756894512e-06,
+      "loss": 1.8894,
+      "step": 8766
+    },
+    {
+      "epoch": 0.9223566543924251,
+      "grad_norm": 1.596531629562378,
+      "learning_rate": 3.1457530374614295e-06,
+      "loss": 2.0511,
+      "step": 8767
+    },
+    {
+      "epoch": 0.9224618621778011,
+      "grad_norm": 1.645261526107788,
+      "learning_rate": 3.1372785653931093e-06,
+      "loss": 1.7624,
+      "step": 8768
+    },
+    {
+      "epoch": 0.9225670699631773,
+      "grad_norm": 1.3146398067474365,
+      "learning_rate": 3.128815341673674e-06,
+      "loss": 1.9661,
+      "step": 8769
+    },
+    {
+      "epoch": 0.9226722777485534,
+      "grad_norm": 1.578999400138855,
+      "learning_rate": 3.120363367285917e-06,
+      "loss": 1.3778,
+      "step": 8770
+    },
+    {
+      "epoch": 0.9227774855339295,
+      "grad_norm": 2.222402811050415,
+      "learning_rate": 3.111922643211351e-06,
+      "loss": 2.074,
+      "step": 8771
+    },
+    {
+      "epoch": 0.9228826933193056,
+      "grad_norm": 1.712213158607483,
+      "learning_rate": 3.1034931704301606e-06,
+      "loss": 1.6251,
+      "step": 8772
+    },
+    {
+      "epoch": 0.9229879011046818,
+      "grad_norm": 1.4497140645980835,
+      "learning_rate": 3.0950749499212283e-06,
+      "loss": 1.5102,
+      "step": 8773
+    },
+    {
+      "epoch": 0.9230931088900579,
+      "grad_norm": 1.7401906251907349,
+      "learning_rate": 3.0866679826621504e-06,
+      "loss": 1.5702,
+      "step": 8774
+    },
+    {
+      "epoch": 0.923198316675434,
+      "grad_norm": 1.3884131908416748,
+      "learning_rate": 3.078272269629201e-06,
+      "loss": 1.6859,
+      "step": 8775
+    },
+    {
+      "epoch": 0.9233035244608101,
+      "grad_norm": 1.8712023496627808,
+      "learning_rate": 3.0698878117973117e-06,
+      "loss": 1.7209,
+      "step": 8776
+    },
+    {
+      "epoch": 0.9234087322461862,
+      "grad_norm": 1.4813895225524902,
+      "learning_rate": 3.0615146101401925e-06,
+      "loss": 1.5664,
+      "step": 8777
+    },
+    {
+      "epoch": 0.9235139400315623,
+      "grad_norm": 1.8341212272644043,
+      "learning_rate": 3.053152665630166e-06,
+      "loss": 1.4758,
+      "step": 8778
+    },
+    {
+      "epoch": 0.9236191478169384,
+      "grad_norm": 2.59242582321167,
+      "learning_rate": 3.0448019792382654e-06,
+      "loss": 1.4894,
+      "step": 8779
+    },
+    {
+      "epoch": 0.9237243556023146,
+      "grad_norm": 1.6962684392929077,
+      "learning_rate": 3.0364625519342603e-06,
+      "loss": 1.1454,
+      "step": 8780
+    },
+    {
+      "epoch": 0.9238295633876907,
+      "grad_norm": 1.346834421157837,
+      "learning_rate": 3.028134384686565e-06,
+      "loss": 1.3534,
+      "step": 8781
+    },
+    {
+      "epoch": 0.9239347711730668,
+      "grad_norm": 1.6630899906158447,
+      "learning_rate": 3.0198174784622944e-06,
+      "loss": 1.7483,
+      "step": 8782
+    },
+    {
+      "epoch": 0.9240399789584429,
+      "grad_norm": 1.2288480997085571,
+      "learning_rate": 3.0115118342272765e-06,
+      "loss": 1.9129,
+      "step": 8783
+    },
+    {
+      "epoch": 0.9241451867438191,
+      "grad_norm": 1.761152982711792,
+      "learning_rate": 3.0032174529460165e-06,
+      "loss": 1.4155,
+      "step": 8784
+    },
+    {
+      "epoch": 0.9242503945291951,
+      "grad_norm": 1.8571745157241821,
+      "learning_rate": 2.9949343355817003e-06,
+      "loss": 1.5797,
+      "step": 8785
+    },
+    {
+      "epoch": 0.9243556023145713,
+      "grad_norm": 1.4947131872177124,
+      "learning_rate": 2.9866624830962366e-06,
+      "loss": 1.5442,
+      "step": 8786
+    },
+    {
+      "epoch": 0.9244608100999474,
+      "grad_norm": 1.9050381183624268,
+      "learning_rate": 2.9784018964502114e-06,
+      "loss": 1.0048,
+      "step": 8787
+    },
+    {
+      "epoch": 0.9245660178853236,
+      "grad_norm": 1.943846583366394,
+      "learning_rate": 2.9701525766028802e-06,
+      "loss": 1.6949,
+      "step": 8788
+    },
+    {
+      "epoch": 0.9246712256706996,
+      "grad_norm": 2.4148600101470947,
+      "learning_rate": 2.9619145245122217e-06,
+      "loss": 0.8479,
+      "step": 8789
+    },
+    {
+      "epoch": 0.9247764334560757,
+      "grad_norm": 1.4801805019378662,
+      "learning_rate": 2.9536877411348808e-06,
+      "loss": 2.1227,
+      "step": 8790
+    },
+    {
+      "epoch": 0.9248816412414519,
+      "grad_norm": 1.3251457214355469,
+      "learning_rate": 2.945472227426227e-06,
+      "loss": 1.5081,
+      "step": 8791
+    },
+    {
+      "epoch": 0.924986849026828,
+      "grad_norm": 2.5353012084960938,
+      "learning_rate": 2.9372679843402863e-06,
+      "loss": 1.4429,
+      "step": 8792
+    },
+    {
+      "epoch": 0.9250920568122041,
+      "grad_norm": 1.5721760988235474,
+      "learning_rate": 2.9290750128297963e-06,
+      "loss": 1.9756,
+      "step": 8793
+    },
+    {
+      "epoch": 0.9251972645975802,
+      "grad_norm": 1.1288405656814575,
+      "learning_rate": 2.9208933138461737e-06,
+      "loss": 2.0431,
+      "step": 8794
+    },
+    {
+      "epoch": 0.9253024723829564,
+      "grad_norm": 1.4272089004516602,
+      "learning_rate": 2.9127228883395472e-06,
+      "loss": 1.9728,
+      "step": 8795
+    },
+    {
+      "epoch": 0.9254076801683324,
+      "grad_norm": 1.9231061935424805,
+      "learning_rate": 2.904563737258692e-06,
+      "loss": 2.0053,
+      "step": 8796
+    },
+    {
+      "epoch": 0.9255128879537086,
+      "grad_norm": 1.3104902505874634,
+      "learning_rate": 2.8964158615511383e-06,
+      "loss": 1.5842,
+      "step": 8797
+    },
+    {
+      "epoch": 0.9256180957390847,
+      "grad_norm": 1.5530825853347778,
+      "learning_rate": 2.8882792621630406e-06,
+      "loss": 1.1644,
+      "step": 8798
+    },
+    {
+      "epoch": 0.9257233035244609,
+      "grad_norm": 1.2885637283325195,
+      "learning_rate": 2.8801539400393097e-06,
+      "loss": 1.7493,
+      "step": 8799
+    },
+    {
+      "epoch": 0.9258285113098369,
+      "grad_norm": 1.3112657070159912,
+      "learning_rate": 2.8720398961234907e-06,
+      "loss": 1.2708,
+      "step": 8800
+    },
+    {
+      "epoch": 0.925933719095213,
+      "grad_norm": 1.7676241397857666,
+      "learning_rate": 2.86393713135783e-06,
+      "loss": 1.7696,
+      "step": 8801
+    },
+    {
+      "epoch": 0.9260389268805892,
+      "grad_norm": 1.611799955368042,
+      "learning_rate": 2.8558456466832973e-06,
+      "loss": 1.4262,
+      "step": 8802
+    },
+    {
+      "epoch": 0.9261441346659652,
+      "grad_norm": 1.5839171409606934,
+      "learning_rate": 2.8477654430395185e-06,
+      "loss": 1.9976,
+      "step": 8803
+    },
+    {
+      "epoch": 0.9262493424513414,
+      "grad_norm": 1.8299293518066406,
+      "learning_rate": 2.839696521364821e-06,
+      "loss": 1.5226,
+      "step": 8804
+    },
+    {
+      "epoch": 0.9263545502367175,
+      "grad_norm": 2.451498508453369,
+      "learning_rate": 2.8316388825962324e-06,
+      "loss": 1.4629,
+      "step": 8805
+    },
+    {
+      "epoch": 0.9264597580220937,
+      "grad_norm": 1.6395529508590698,
+      "learning_rate": 2.823592527669461e-06,
+      "loss": 1.8826,
+      "step": 8806
+    },
+    {
+      "epoch": 0.9265649658074697,
+      "grad_norm": 1.9740159511566162,
+      "learning_rate": 2.8155574575188694e-06,
+      "loss": 1.2348,
+      "step": 8807
+    },
+    {
+      "epoch": 0.9266701735928459,
+      "grad_norm": 2.034790515899658,
+      "learning_rate": 2.8075336730775894e-06,
+      "loss": 1.6714,
+      "step": 8808
+    },
+    {
+      "epoch": 0.926775381378222,
+      "grad_norm": 1.232695460319519,
+      "learning_rate": 2.7995211752773752e-06,
+      "loss": 1.2372,
+      "step": 8809
+    },
+    {
+      "epoch": 0.9268805891635981,
+      "grad_norm": 1.4230536222457886,
+      "learning_rate": 2.7915199650486944e-06,
+      "loss": 1.8896,
+      "step": 8810
+    },
+    {
+      "epoch": 0.9269857969489742,
+      "grad_norm": 1.6737197637557983,
+      "learning_rate": 2.7835300433207035e-06,
+      "loss": 1.6089,
+      "step": 8811
+    },
+    {
+      "epoch": 0.9270910047343504,
+      "grad_norm": 2.7729249000549316,
+      "learning_rate": 2.7755514110212264e-06,
+      "loss": 1.311,
+      "step": 8812
+    },
+    {
+      "epoch": 0.9271962125197265,
+      "grad_norm": 1.7959167957305908,
+      "learning_rate": 2.767584069076823e-06,
+      "loss": 1.3979,
+      "step": 8813
+    },
+    {
+      "epoch": 0.9273014203051025,
+      "grad_norm": 0.992779016494751,
+      "learning_rate": 2.7596280184126965e-06,
+      "loss": 1.7532,
+      "step": 8814
+    },
+    {
+      "epoch": 0.9274066280904787,
+      "grad_norm": 1.6769089698791504,
+      "learning_rate": 2.751683259952764e-06,
+      "loss": 1.6832,
+      "step": 8815
+    },
+    {
+      "epoch": 0.9275118358758548,
+      "grad_norm": 1.5358198881149292,
+      "learning_rate": 2.7437497946196322e-06,
+      "loss": 1.1027,
+      "step": 8816
+    },
+    {
+      "epoch": 0.9276170436612309,
+      "grad_norm": 1.683468222618103,
+      "learning_rate": 2.7358276233345747e-06,
+      "loss": 1.014,
+      "step": 8817
+    },
+    {
+      "epoch": 0.927722251446607,
+      "grad_norm": 1.826439619064331,
+      "learning_rate": 2.727916747017556e-06,
+      "loss": 1.3374,
+      "step": 8818
+    },
+    {
+      "epoch": 0.9278274592319832,
+      "grad_norm": 1.3325717449188232,
+      "learning_rate": 2.7200171665872742e-06,
+      "loss": 1.8112,
+      "step": 8819
+    },
+    {
+      "epoch": 0.9279326670173593,
+      "grad_norm": 1.434065818786621,
+      "learning_rate": 2.7121288829610624e-06,
+      "loss": 1.5424,
+      "step": 8820
+    },
+    {
+      "epoch": 0.9280378748027354,
+      "grad_norm": 1.613349437713623,
+      "learning_rate": 2.7042518970549546e-06,
+      "loss": 1.3637,
+      "step": 8821
+    },
+    {
+      "epoch": 0.9281430825881115,
+      "grad_norm": 1.9332239627838135,
+      "learning_rate": 2.696386209783697e-06,
+      "loss": 1.5076,
+      "step": 8822
+    },
+    {
+      "epoch": 0.9282482903734877,
+      "grad_norm": 1.393325686454773,
+      "learning_rate": 2.6885318220606914e-06,
+      "loss": 1.6009,
+      "step": 8823
+    },
+    {
+      "epoch": 0.9283534981588638,
+      "grad_norm": 1.6824408769607544,
+      "learning_rate": 2.6806887347980427e-06,
+      "loss": 1.6287,
+      "step": 8824
+    },
+    {
+      "epoch": 0.9284587059442398,
+      "grad_norm": 1.7882195711135864,
+      "learning_rate": 2.6728569489065437e-06,
+      "loss": 2.0023,
+      "step": 8825
+    },
+    {
+      "epoch": 0.928563913729616,
+      "grad_norm": 1.559647560119629,
+      "learning_rate": 2.6650364652956894e-06,
+      "loss": 1.3792,
+      "step": 8826
+    },
+    {
+      "epoch": 0.9286691215149921,
+      "grad_norm": 1.2669860124588013,
+      "learning_rate": 2.657227284873609e-06,
+      "loss": 1.6591,
+      "step": 8827
+    },
+    {
+      "epoch": 0.9287743293003682,
+      "grad_norm": 1.9442473649978638,
+      "learning_rate": 2.6494294085472103e-06,
+      "loss": 1.7604,
+      "step": 8828
+    },
+    {
+      "epoch": 0.9288795370857443,
+      "grad_norm": 1.3520489931106567,
+      "learning_rate": 2.6416428372219914e-06,
+      "loss": 1.8279,
+      "step": 8829
+    },
+    {
+      "epoch": 0.9289847448711205,
+      "grad_norm": 1.2436840534210205,
+      "learning_rate": 2.6338675718022064e-06,
+      "loss": 1.9452,
+      "step": 8830
+    },
+    {
+      "epoch": 0.9290899526564966,
+      "grad_norm": 1.687325119972229,
+      "learning_rate": 2.6261036131907557e-06,
+      "loss": 1.1885,
+      "step": 8831
+    },
+    {
+      "epoch": 0.9291951604418727,
+      "grad_norm": 1.489946722984314,
+      "learning_rate": 2.61835096228924e-06,
+      "loss": 1.9571,
+      "step": 8832
+    },
+    {
+      "epoch": 0.9293003682272488,
+      "grad_norm": 1.1309078931808472,
+      "learning_rate": 2.6106096199979614e-06,
+      "loss": 1.4592,
+      "step": 8833
+    },
+    {
+      "epoch": 0.929405576012625,
+      "grad_norm": 1.816552758216858,
+      "learning_rate": 2.6028795872159005e-06,
+      "loss": 1.5374,
+      "step": 8834
+    },
+    {
+      "epoch": 0.929510783798001,
+      "grad_norm": 1.753360390663147,
+      "learning_rate": 2.5951608648406955e-06,
+      "loss": 1.5404,
+      "step": 8835
+    },
+    {
+      "epoch": 0.9296159915833772,
+      "grad_norm": 1.432955265045166,
+      "learning_rate": 2.587453453768729e-06,
+      "loss": 1.7795,
+      "step": 8836
+    },
+    {
+      "epoch": 0.9297211993687533,
+      "grad_norm": 2.5248990058898926,
+      "learning_rate": 2.579757354895018e-06,
+      "loss": 1.5791,
+      "step": 8837
+    },
+    {
+      "epoch": 0.9298264071541295,
+      "grad_norm": 1.3202687501907349,
+      "learning_rate": 2.5720725691132706e-06,
+      "loss": 1.4432,
+      "step": 8838
+    },
+    {
+      "epoch": 0.9299316149395055,
+      "grad_norm": 1.9619760513305664,
+      "learning_rate": 2.564399097315928e-06,
+      "loss": 1.7022,
+      "step": 8839
+    },
+    {
+      "epoch": 0.9300368227248816,
+      "grad_norm": 1.2252382040023804,
+      "learning_rate": 2.5567369403940776e-06,
+      "loss": 1.579,
+      "step": 8840
+    },
+    {
+      "epoch": 0.9301420305102578,
+      "grad_norm": 1.8557909727096558,
+      "learning_rate": 2.5490860992374745e-06,
+      "loss": 1.6796,
+      "step": 8841
+    },
+    {
+      "epoch": 0.9302472382956338,
+      "grad_norm": 1.5600543022155762,
+      "learning_rate": 2.5414465747346182e-06,
+      "loss": 1.5379,
+      "step": 8842
+    },
+    {
+      "epoch": 0.93035244608101,
+      "grad_norm": 1.3198964595794678,
+      "learning_rate": 2.5338183677726334e-06,
+      "loss": 1.8882,
+      "step": 8843
+    },
+    {
+      "epoch": 0.9304576538663861,
+      "grad_norm": 1.934952974319458,
+      "learning_rate": 2.526201479237389e-06,
+      "loss": 0.9782,
+      "step": 8844
+    },
+    {
+      "epoch": 0.9305628616517623,
+      "grad_norm": 1.6024317741394043,
+      "learning_rate": 2.5185959100133883e-06,
+      "loss": 1.7564,
+      "step": 8845
+    },
+    {
+      "epoch": 0.9306680694371383,
+      "grad_norm": 1.7014966011047363,
+      "learning_rate": 2.5110016609838473e-06,
+      "loss": 1.5487,
+      "step": 8846
+    },
+    {
+      "epoch": 0.9307732772225145,
+      "grad_norm": 1.7204549312591553,
+      "learning_rate": 2.50341873303066e-06,
+      "loss": 1.7045,
+      "step": 8847
+    },
+    {
+      "epoch": 0.9308784850078906,
+      "grad_norm": 1.1743104457855225,
+      "learning_rate": 2.495847127034401e-06,
+      "loss": 1.7379,
+      "step": 8848
+    },
+    {
+      "epoch": 0.9309836927932666,
+      "grad_norm": 1.450788140296936,
+      "learning_rate": 2.4882868438743436e-06,
+      "loss": 1.2717,
+      "step": 8849
+    },
+    {
+      "epoch": 0.9310889005786428,
+      "grad_norm": 1.5399311780929565,
+      "learning_rate": 2.480737884428441e-06,
+      "loss": 1.3663,
+      "step": 8850
+    },
+    {
+      "epoch": 0.931194108364019,
+      "grad_norm": 1.6990025043487549,
+      "learning_rate": 2.4732002495733154e-06,
+      "loss": 1.8612,
+      "step": 8851
+    },
+    {
+      "epoch": 0.9312993161493951,
+      "grad_norm": 1.3516472578048706,
+      "learning_rate": 2.46567394018431e-06,
+      "loss": 1.5048,
+      "step": 8852
+    },
+    {
+      "epoch": 0.9314045239347711,
+      "grad_norm": 2.059854030609131,
+      "learning_rate": 2.458158957135415e-06,
+      "loss": 1.6879,
+      "step": 8853
+    },
+    {
+      "epoch": 0.9315097317201473,
+      "grad_norm": 1.1983468532562256,
+      "learning_rate": 2.4506553012993093e-06,
+      "loss": 1.7273,
+      "step": 8854
+    },
+    {
+      "epoch": 0.9316149395055234,
+      "grad_norm": 1.6070380210876465,
+      "learning_rate": 2.443162973547386e-06,
+      "loss": 1.3611,
+      "step": 8855
+    },
+    {
+      "epoch": 0.9317201472908996,
+      "grad_norm": 1.6409918069839478,
+      "learning_rate": 2.435681974749704e-06,
+      "loss": 1.5396,
+      "step": 8856
+    },
+    {
+      "epoch": 0.9318253550762756,
+      "grad_norm": 2.1190476417541504,
+      "learning_rate": 2.4282123057750016e-06,
+      "loss": 1.9413,
+      "step": 8857
+    },
+    {
+      "epoch": 0.9319305628616518,
+      "grad_norm": 1.454859733581543,
+      "learning_rate": 2.4207539674907075e-06,
+      "loss": 1.6396,
+      "step": 8858
+    },
+    {
+      "epoch": 0.9320357706470279,
+      "grad_norm": 1.9500346183776855,
+      "learning_rate": 2.413306960762929e-06,
+      "loss": 1.0076,
+      "step": 8859
+    },
+    {
+      "epoch": 0.932140978432404,
+      "grad_norm": 2.006927013397217,
+      "learning_rate": 2.4058712864564736e-06,
+      "loss": 1.8154,
+      "step": 8860
+    },
+    {
+      "epoch": 0.9322461862177801,
+      "grad_norm": 1.5914230346679688,
+      "learning_rate": 2.398446945434818e-06,
+      "loss": 1.5964,
+      "step": 8861
+    },
+    {
+      "epoch": 0.9323513940031563,
+      "grad_norm": 1.940948724746704,
+      "learning_rate": 2.3910339385601168e-06,
+      "loss": 1.8728,
+      "step": 8862
+    },
+    {
+      "epoch": 0.9324566017885324,
+      "grad_norm": 1.6678147315979004,
+      "learning_rate": 2.383632266693225e-06,
+      "loss": 2.0333,
+      "step": 8863
+    },
+    {
+      "epoch": 0.9325618095739084,
+      "grad_norm": 1.5024465322494507,
+      "learning_rate": 2.376241930693679e-06,
+      "loss": 1.7904,
+      "step": 8864
+    },
+    {
+      "epoch": 0.9326670173592846,
+      "grad_norm": 1.273901343345642,
+      "learning_rate": 2.368862931419702e-06,
+      "loss": 1.3617,
+      "step": 8865
+    },
+    {
+      "epoch": 0.9327722251446607,
+      "grad_norm": 1.9819680452346802,
+      "learning_rate": 2.3614952697281534e-06,
+      "loss": 1.5756,
+      "step": 8866
+    },
+    {
+      "epoch": 0.9328774329300368,
+      "grad_norm": 1.4044946432113647,
+      "learning_rate": 2.354138946474671e-06,
+      "loss": 1.7253,
+      "step": 8867
+    },
+    {
+      "epoch": 0.9329826407154129,
+      "grad_norm": 1.6213405132293701,
+      "learning_rate": 2.346793962513483e-06,
+      "loss": 1.9362,
+      "step": 8868
+    },
+    {
+      "epoch": 0.9330878485007891,
+      "grad_norm": 2.6082890033721924,
+      "learning_rate": 2.3394603186975393e-06,
+      "loss": 1.5571,
+      "step": 8869
+    },
+    {
+      "epoch": 0.9331930562861652,
+      "grad_norm": 1.7648690938949585,
+      "learning_rate": 2.332138015878482e-06,
+      "loss": 1.5838,
+      "step": 8870
+    },
+    {
+      "epoch": 0.9332982640715413,
+      "grad_norm": 1.2517858743667603,
+      "learning_rate": 2.324827054906631e-06,
+      "loss": 1.4428,
+      "step": 8871
+    },
+    {
+      "epoch": 0.9334034718569174,
+      "grad_norm": 1.9067564010620117,
+      "learning_rate": 2.317527436630973e-06,
+      "loss": 2.0886,
+      "step": 8872
+    },
+    {
+      "epoch": 0.9335086796422936,
+      "grad_norm": 1.4557033777236938,
+      "learning_rate": 2.310239161899208e-06,
+      "loss": 1.4884,
+      "step": 8873
+    },
+    {
+      "epoch": 0.9336138874276696,
+      "grad_norm": 1.6103370189666748,
+      "learning_rate": 2.3029622315576595e-06,
+      "loss": 1.6499,
+      "step": 8874
+    },
+    {
+      "epoch": 0.9337190952130457,
+      "grad_norm": 1.6522938013076782,
+      "learning_rate": 2.2956966464514175e-06,
+      "loss": 1.7791,
+      "step": 8875
+    },
+    {
+      "epoch": 0.9338243029984219,
+      "grad_norm": 1.3017587661743164,
+      "learning_rate": 2.288442407424185e-06,
+      "loss": 1.9411,
+      "step": 8876
+    },
+    {
+      "epoch": 0.933929510783798,
+      "grad_norm": 1.478482961654663,
+      "learning_rate": 2.2811995153183776e-06,
+      "loss": 1.6778,
+      "step": 8877
+    },
+    {
+      "epoch": 0.9340347185691741,
+      "grad_norm": 2.009734630584717,
+      "learning_rate": 2.2739679709750885e-06,
+      "loss": 1.4429,
+      "step": 8878
+    },
+    {
+      "epoch": 0.9341399263545502,
+      "grad_norm": 1.2722523212432861,
+      "learning_rate": 2.2667477752341017e-06,
+      "loss": 1.4058,
+      "step": 8879
+    },
+    {
+      "epoch": 0.9342451341399264,
+      "grad_norm": 1.124986171722412,
+      "learning_rate": 2.2595389289338575e-06,
+      "loss": 1.3739,
+      "step": 8880
+    },
+    {
+      "epoch": 0.9343503419253024,
+      "grad_norm": 1.039442777633667,
+      "learning_rate": 2.252341432911509e-06,
+      "loss": 1.7472,
+      "step": 8881
+    },
+    {
+      "epoch": 0.9344555497106786,
+      "grad_norm": 2.12754487991333,
+      "learning_rate": 2.245155288002876e-06,
+      "loss": 1.2261,
+      "step": 8882
+    },
+    {
+      "epoch": 0.9345607574960547,
+      "grad_norm": 1.5434430837631226,
+      "learning_rate": 2.2379804950424576e-06,
+      "loss": 1.5896,
+      "step": 8883
+    },
+    {
+      "epoch": 0.9346659652814309,
+      "grad_norm": 1.6226022243499756,
+      "learning_rate": 2.2308170548634435e-06,
+      "loss": 1.3338,
+      "step": 8884
+    },
+    {
+      "epoch": 0.9347711730668069,
+      "grad_norm": 1.7707509994506836,
+      "learning_rate": 2.2236649682977117e-06,
+      "loss": 1.332,
+      "step": 8885
+    },
+    {
+      "epoch": 0.934876380852183,
+      "grad_norm": 1.5284020900726318,
+      "learning_rate": 2.2165242361757764e-06,
+      "loss": 1.669,
+      "step": 8886
+    },
+    {
+      "epoch": 0.9349815886375592,
+      "grad_norm": 1.250614881515503,
+      "learning_rate": 2.2093948593268963e-06,
+      "loss": 1.7689,
+      "step": 8887
+    },
+    {
+      "epoch": 0.9350867964229354,
+      "grad_norm": 1.4621397256851196,
+      "learning_rate": 2.202276838578976e-06,
+      "loss": 1.5322,
+      "step": 8888
+    },
+    {
+      "epoch": 0.9351920042083114,
+      "grad_norm": 1.8375461101531982,
+      "learning_rate": 2.1951701747585982e-06,
+      "loss": 1.6119,
+      "step": 8889
+    },
+    {
+      "epoch": 0.9352972119936875,
+      "grad_norm": 1.5318689346313477,
+      "learning_rate": 2.188074868691059e-06,
+      "loss": 1.2107,
+      "step": 8890
+    },
+    {
+      "epoch": 0.9354024197790637,
+      "grad_norm": 2.4939355850219727,
+      "learning_rate": 2.1809909212002767e-06,
+      "loss": 1.7044,
+      "step": 8891
+    },
+    {
+      "epoch": 0.9355076275644397,
+      "grad_norm": 1.93202543258667,
+      "learning_rate": 2.1739183331089263e-06,
+      "loss": 1.3337,
+      "step": 8892
+    },
+    {
+      "epoch": 0.9356128353498159,
+      "grad_norm": 1.680690884590149,
+      "learning_rate": 2.166857105238307e-06,
+      "loss": 1.9268,
+      "step": 8893
+    },
+    {
+      "epoch": 0.935718043135192,
+      "grad_norm": 1.772756576538086,
+      "learning_rate": 2.159807238408418e-06,
+      "loss": 1.7988,
+      "step": 8894
+    },
+    {
+      "epoch": 0.9358232509205682,
+      "grad_norm": 1.440395474433899,
+      "learning_rate": 2.1527687334379266e-06,
+      "loss": 1.2087,
+      "step": 8895
+    },
+    {
+      "epoch": 0.9359284587059442,
+      "grad_norm": 1.6496853828430176,
+      "learning_rate": 2.1457415911442013e-06,
+      "loss": 1.5179,
+      "step": 8896
+    },
+    {
+      "epoch": 0.9360336664913204,
+      "grad_norm": 1.3402870893478394,
+      "learning_rate": 2.1387258123432673e-06,
+      "loss": 1.7395,
+      "step": 8897
+    },
+    {
+      "epoch": 0.9361388742766965,
+      "grad_norm": 1.9559837579727173,
+      "learning_rate": 2.131721397849862e-06,
+      "loss": 1.5187,
+      "step": 8898
+    },
+    {
+      "epoch": 0.9362440820620725,
+      "grad_norm": 2.084958076477051,
+      "learning_rate": 2.1247283484773785e-06,
+      "loss": 1.8087,
+      "step": 8899
+    },
+    {
+      "epoch": 0.9363492898474487,
+      "grad_norm": 2.734379291534424,
+      "learning_rate": 2.1177466650379007e-06,
+      "loss": 1.286,
+      "step": 8900
+    },
+    {
+      "epoch": 0.9364544976328248,
+      "grad_norm": 1.1719292402267456,
+      "learning_rate": 2.1107763483421805e-06,
+      "loss": 1.8817,
+      "step": 8901
+    },
+    {
+      "epoch": 0.936559705418201,
+      "grad_norm": 1.631633996963501,
+      "learning_rate": 2.103817399199659e-06,
+      "loss": 1.5544,
+      "step": 8902
+    },
+    {
+      "epoch": 0.936664913203577,
+      "grad_norm": 2.258840799331665,
+      "learning_rate": 2.0968698184184565e-06,
+      "loss": 1.4009,
+      "step": 8903
+    },
+    {
+      "epoch": 0.9367701209889532,
+      "grad_norm": 1.6519112586975098,
+      "learning_rate": 2.0899336068053833e-06,
+      "loss": 1.5555,
+      "step": 8904
+    },
+    {
+      "epoch": 0.9368753287743293,
+      "grad_norm": 1.7025299072265625,
+      "learning_rate": 2.0830087651658945e-06,
+      "loss": 2.2402,
+      "step": 8905
+    },
+    {
+      "epoch": 0.9369805365597054,
+      "grad_norm": 1.8736423254013062,
+      "learning_rate": 2.076095294304181e-06,
+      "loss": 1.5652,
+      "step": 8906
+    },
+    {
+      "epoch": 0.9370857443450815,
+      "grad_norm": 1.4368400573730469,
+      "learning_rate": 2.069193195023067e-06,
+      "loss": 2.0581,
+      "step": 8907
+    },
+    {
+      "epoch": 0.9371909521304577,
+      "grad_norm": 1.2674177885055542,
+      "learning_rate": 2.0623024681240554e-06,
+      "loss": 1.8723,
+      "step": 8908
+    },
+    {
+      "epoch": 0.9372961599158338,
+      "grad_norm": 1.2145195007324219,
+      "learning_rate": 2.0554231144073623e-06,
+      "loss": 1.8348,
+      "step": 8909
+    },
+    {
+      "epoch": 0.9374013677012099,
+      "grad_norm": 2.1379621028900146,
+      "learning_rate": 2.04855513467187e-06,
+      "loss": 1.5549,
+      "step": 8910
+    },
+    {
+      "epoch": 0.937506575486586,
+      "grad_norm": 2.2935433387756348,
+      "learning_rate": 2.041698529715097e-06,
+      "loss": 0.869,
+      "step": 8911
+    },
+    {
+      "epoch": 0.9376117832719622,
+      "grad_norm": 2.0212948322296143,
+      "learning_rate": 2.034853300333328e-06,
+      "loss": 2.2311,
+      "step": 8912
+    },
+    {
+      "epoch": 0.9377169910573382,
+      "grad_norm": 1.7877557277679443,
+      "learning_rate": 2.0280194473214497e-06,
+      "loss": 1.214,
+      "step": 8913
+    },
+    {
+      "epoch": 0.9378221988427143,
+      "grad_norm": 1.538549780845642,
+      "learning_rate": 2.0211969714730496e-06,
+      "loss": 1.2042,
+      "step": 8914
+    },
+    {
+      "epoch": 0.9379274066280905,
+      "grad_norm": 1.2701212167739868,
+      "learning_rate": 2.0143858735804154e-06,
+      "loss": 1.6479,
+      "step": 8915
+    },
+    {
+      "epoch": 0.9380326144134666,
+      "grad_norm": 2.006523609161377,
+      "learning_rate": 2.007586154434493e-06,
+      "loss": 1.9177,
+      "step": 8916
+    },
+    {
+      "epoch": 0.9381378221988427,
+      "grad_norm": 1.5271292924880981,
+      "learning_rate": 2.000797814824906e-06,
+      "loss": 1.2413,
+      "step": 8917
+    },
+    {
+      "epoch": 0.9382430299842188,
+      "grad_norm": 1.2923787832260132,
+      "learning_rate": 1.9940208555399685e-06,
+      "loss": 1.626,
+      "step": 8918
+    },
+    {
+      "epoch": 0.938348237769595,
+      "grad_norm": 2.304880380630493,
+      "learning_rate": 1.987255277366662e-06,
+      "loss": 1.7644,
+      "step": 8919
+    },
+    {
+      "epoch": 0.9384534455549711,
+      "grad_norm": 1.4698824882507324,
+      "learning_rate": 1.9805010810906464e-06,
+      "loss": 1.9222,
+      "step": 8920
+    },
+    {
+      "epoch": 0.9385586533403472,
+      "grad_norm": 1.2957773208618164,
+      "learning_rate": 1.9737582674962728e-06,
+      "loss": 1.5484,
+      "step": 8921
+    },
+    {
+      "epoch": 0.9386638611257233,
+      "grad_norm": 1.2710483074188232,
+      "learning_rate": 1.967026837366559e-06,
+      "loss": 1.4719,
+      "step": 8922
+    },
+    {
+      "epoch": 0.9387690689110995,
+      "grad_norm": 1.4735010862350464,
+      "learning_rate": 1.9603067914832017e-06,
+      "loss": 1.6447,
+      "step": 8923
+    },
+    {
+      "epoch": 0.9388742766964755,
+      "grad_norm": 1.5090333223342896,
+      "learning_rate": 1.9535981306265884e-06,
+      "loss": 1.9183,
+      "step": 8924
+    },
+    {
+      "epoch": 0.9389794844818516,
+      "grad_norm": 1.4228880405426025,
+      "learning_rate": 1.9469008555757505e-06,
+      "loss": 1.8077,
+      "step": 8925
+    },
+    {
+      "epoch": 0.9390846922672278,
+      "grad_norm": 1.4772889614105225,
+      "learning_rate": 1.9402149671084446e-06,
+      "loss": 1.4166,
+      "step": 8926
+    },
+    {
+      "epoch": 0.9391899000526039,
+      "grad_norm": 2.280817985534668,
+      "learning_rate": 1.9335404660010713e-06,
+      "loss": 1.458,
+      "step": 8927
+    },
+    {
+      "epoch": 0.93929510783798,
+      "grad_norm": 2.147027015686035,
+      "learning_rate": 1.926877353028711e-06,
+      "loss": 0.6933,
+      "step": 8928
+    },
+    {
+      "epoch": 0.9394003156233561,
+      "grad_norm": 1.2121834754943848,
+      "learning_rate": 1.9202256289651446e-06,
+      "loss": 1.8354,
+      "step": 8929
+    },
+    {
+      "epoch": 0.9395055234087323,
+      "grad_norm": 1.4258288145065308,
+      "learning_rate": 1.913585294582798e-06,
+      "loss": 1.588,
+      "step": 8930
+    },
+    {
+      "epoch": 0.9396107311941083,
+      "grad_norm": 1.710619330406189,
+      "learning_rate": 1.9069563506527998e-06,
+      "loss": 1.7881,
+      "step": 8931
+    },
+    {
+      "epoch": 0.9397159389794845,
+      "grad_norm": 1.781956434249878,
+      "learning_rate": 1.9003387979449562e-06,
+      "loss": 1.416,
+      "step": 8932
+    },
+    {
+      "epoch": 0.9398211467648606,
+      "grad_norm": 2.1186389923095703,
+      "learning_rate": 1.8937326372277408e-06,
+      "loss": 1.2819,
+      "step": 8933
+    },
+    {
+      "epoch": 0.9399263545502368,
+      "grad_norm": 1.4919960498809814,
+      "learning_rate": 1.8871378692682851e-06,
+      "loss": 1.6042,
+      "step": 8934
+    },
+    {
+      "epoch": 0.9400315623356128,
+      "grad_norm": 1.996641755104065,
+      "learning_rate": 1.8805544948324317e-06,
+      "loss": 1.4017,
+      "step": 8935
+    },
+    {
+      "epoch": 0.940136770120989,
+      "grad_norm": 1.87164306640625,
+      "learning_rate": 1.87398251468468e-06,
+      "loss": 1.7661,
+      "step": 8936
+    },
+    {
+      "epoch": 0.9402419779063651,
+      "grad_norm": 1.9063903093338013,
+      "learning_rate": 1.867421929588231e-06,
+      "loss": 1.7494,
+      "step": 8937
+    },
+    {
+      "epoch": 0.9403471856917411,
+      "grad_norm": 3.1897265911102295,
+      "learning_rate": 1.8608727403049309e-06,
+      "loss": 1.7943,
+      "step": 8938
+    },
+    {
+      "epoch": 0.9404523934771173,
+      "grad_norm": 6.021173000335693,
+      "learning_rate": 1.8543349475953043e-06,
+      "loss": 1.4585,
+      "step": 8939
+    },
+    {
+      "epoch": 0.9405576012624934,
+      "grad_norm": 2.0118443965911865,
+      "learning_rate": 1.8478085522185773e-06,
+      "loss": 1.4911,
+      "step": 8940
+    },
+    {
+      "epoch": 0.9406628090478696,
+      "grad_norm": 1.1734709739685059,
+      "learning_rate": 1.8412935549326438e-06,
+      "loss": 1.9156,
+      "step": 8941
+    },
+    {
+      "epoch": 0.9407680168332456,
+      "grad_norm": 1.911544919013977,
+      "learning_rate": 1.834789956494043e-06,
+      "loss": 1.7905,
+      "step": 8942
+    },
+    {
+      "epoch": 0.9408732246186218,
+      "grad_norm": 1.821887731552124,
+      "learning_rate": 1.8282977576580484e-06,
+      "loss": 1.6208,
+      "step": 8943
+    },
+    {
+      "epoch": 0.9409784324039979,
+      "grad_norm": 1.2936660051345825,
+      "learning_rate": 1.8218169591785572e-06,
+      "loss": 1.9121,
+      "step": 8944
+    },
+    {
+      "epoch": 0.941083640189374,
+      "grad_norm": 1.6752891540527344,
+      "learning_rate": 1.8153475618081673e-06,
+      "loss": 1.3991,
+      "step": 8945
+    },
+    {
+      "epoch": 0.9411888479747501,
+      "grad_norm": 1.817868709564209,
+      "learning_rate": 1.8088895662981554e-06,
+      "loss": 1.518,
+      "step": 8946
+    },
+    {
+      "epoch": 0.9412940557601263,
+      "grad_norm": 1.5891584157943726,
+      "learning_rate": 1.802442973398455e-06,
+      "loss": 1.763,
+      "step": 8947
+    },
+    {
+      "epoch": 0.9413992635455024,
+      "grad_norm": 1.1031978130340576,
+      "learning_rate": 1.7960077838577117e-06,
+      "loss": 1.5801,
+      "step": 8948
+    },
+    {
+      "epoch": 0.9415044713308784,
+      "grad_norm": 1.214778184890747,
+      "learning_rate": 1.7895839984231943e-06,
+      "loss": 1.9143,
+      "step": 8949
+    },
+    {
+      "epoch": 0.9416096791162546,
+      "grad_norm": 1.8901784420013428,
+      "learning_rate": 1.7831716178408841e-06,
+      "loss": 1.397,
+      "step": 8950
+    },
+    {
+      "epoch": 0.9417148869016307,
+      "grad_norm": 1.900726079940796,
+      "learning_rate": 1.776770642855441e-06,
+      "loss": 1.8984,
+      "step": 8951
+    },
+    {
+      "epoch": 0.9418200946870069,
+      "grad_norm": 1.5284584760665894,
+      "learning_rate": 1.7703810742101813e-06,
+      "loss": 1.8202,
+      "step": 8952
+    },
+    {
+      "epoch": 0.9419253024723829,
+      "grad_norm": 1.6591954231262207,
+      "learning_rate": 1.7640029126471003e-06,
+      "loss": 1.2281,
+      "step": 8953
+    },
+    {
+      "epoch": 0.9420305102577591,
+      "grad_norm": 1.8267433643341064,
+      "learning_rate": 1.7576361589068834e-06,
+      "loss": 1.3939,
+      "step": 8954
+    },
+    {
+      "epoch": 0.9421357180431352,
+      "grad_norm": 1.8574069738388062,
+      "learning_rate": 1.7512808137288727e-06,
+      "loss": 1.6839,
+      "step": 8955
+    },
+    {
+      "epoch": 0.9422409258285113,
+      "grad_norm": 2.1914479732513428,
+      "learning_rate": 1.7449368778511e-06,
+      "loss": 1.6874,
+      "step": 8956
+    },
+    {
+      "epoch": 0.9423461336138874,
+      "grad_norm": 1.5957982540130615,
+      "learning_rate": 1.7386043520102534e-06,
+      "loss": 1.8577,
+      "step": 8957
+    },
+    {
+      "epoch": 0.9424513413992636,
+      "grad_norm": 1.8510699272155762,
+      "learning_rate": 1.732283236941723e-06,
+      "loss": 1.8177,
+      "step": 8958
+    },
+    {
+      "epoch": 0.9425565491846397,
+      "grad_norm": 1.7489219903945923,
+      "learning_rate": 1.7259735333795545e-06,
+      "loss": 1.9759,
+      "step": 8959
+    },
+    {
+      "epoch": 0.9426617569700158,
+      "grad_norm": 1.5425418615341187,
+      "learning_rate": 1.7196752420564622e-06,
+      "loss": 1.5018,
+      "step": 8960
+    },
+    {
+      "epoch": 0.9427669647553919,
+      "grad_norm": 1.9875203371047974,
+      "learning_rate": 1.7133883637038494e-06,
+      "loss": 1.3086,
+      "step": 8961
+    },
+    {
+      "epoch": 0.942872172540768,
+      "grad_norm": 1.7433385848999023,
+      "learning_rate": 1.7071128990518103e-06,
+      "loss": 1.4756,
+      "step": 8962
+    },
+    {
+      "epoch": 0.9429773803261441,
+      "grad_norm": 1.6027960777282715,
+      "learning_rate": 1.700848848829073e-06,
+      "loss": 2.1611,
+      "step": 8963
+    },
+    {
+      "epoch": 0.9430825881115202,
+      "grad_norm": 1.8459538221359253,
+      "learning_rate": 1.6945962137630668e-06,
+      "loss": 1.7465,
+      "step": 8964
+    },
+    {
+      "epoch": 0.9431877958968964,
+      "grad_norm": 1.179093599319458,
+      "learning_rate": 1.6883549945798883e-06,
+      "loss": 1.8516,
+      "step": 8965
+    },
+    {
+      "epoch": 0.9432930036822725,
+      "grad_norm": 1.7608375549316406,
+      "learning_rate": 1.6821251920043246e-06,
+      "loss": 1.5069,
+      "step": 8966
+    },
+    {
+      "epoch": 0.9433982114676486,
+      "grad_norm": 1.7923959493637085,
+      "learning_rate": 1.675906806759786e-06,
+      "loss": 1.3353,
+      "step": 8967
+    },
+    {
+      "epoch": 0.9435034192530247,
+      "grad_norm": 1.7323315143585205,
+      "learning_rate": 1.6696998395684394e-06,
+      "loss": 1.7977,
+      "step": 8968
+    },
+    {
+      "epoch": 0.9436086270384009,
+      "grad_norm": 1.0940277576446533,
+      "learning_rate": 1.6635042911510413e-06,
+      "loss": 1.8435,
+      "step": 8969
+    },
+    {
+      "epoch": 0.9437138348237769,
+      "grad_norm": 1.5782438516616821,
+      "learning_rate": 1.6573201622270828e-06,
+      "loss": 1.428,
+      "step": 8970
+    },
+    {
+      "epoch": 0.9438190426091531,
+      "grad_norm": 1.616591453552246,
+      "learning_rate": 1.6511474535146899e-06,
+      "loss": 1.3093,
+      "step": 8971
+    },
+    {
+      "epoch": 0.9439242503945292,
+      "grad_norm": 1.872626781463623,
+      "learning_rate": 1.6449861657306998e-06,
+      "loss": 1.1612,
+      "step": 8972
+    },
+    {
+      "epoch": 0.9440294581799054,
+      "grad_norm": 2.045409679412842,
+      "learning_rate": 1.6388362995905848e-06,
+      "loss": 1.7662,
+      "step": 8973
+    },
+    {
+      "epoch": 0.9441346659652814,
+      "grad_norm": 1.8982219696044922,
+      "learning_rate": 1.6326978558085182e-06,
+      "loss": 1.5276,
+      "step": 8974
+    },
+    {
+      "epoch": 0.9442398737506575,
+      "grad_norm": 2.2144486904144287,
+      "learning_rate": 1.6265708350973296e-06,
+      "loss": 1.3016,
+      "step": 8975
+    },
+    {
+      "epoch": 0.9443450815360337,
+      "grad_norm": 1.7731724977493286,
+      "learning_rate": 1.6204552381685278e-06,
+      "loss": 1.884,
+      "step": 8976
+    },
+    {
+      "epoch": 0.9444502893214097,
+      "grad_norm": 1.628603219985962,
+      "learning_rate": 1.6143510657323112e-06,
+      "loss": 1.5011,
+      "step": 8977
+    },
+    {
+      "epoch": 0.9445554971067859,
+      "grad_norm": 1.250891089439392,
+      "learning_rate": 1.6082583184975352e-06,
+      "loss": 1.5904,
+      "step": 8978
+    },
+    {
+      "epoch": 0.944660704892162,
+      "grad_norm": 1.6449693441390991,
+      "learning_rate": 1.6021769971717116e-06,
+      "loss": 1.9314,
+      "step": 8979
+    },
+    {
+      "epoch": 0.9447659126775382,
+      "grad_norm": 1.9107930660247803,
+      "learning_rate": 1.5961071024610752e-06,
+      "loss": 1.5378,
+      "step": 8980
+    },
+    {
+      "epoch": 0.9448711204629142,
+      "grad_norm": 1.551930546760559,
+      "learning_rate": 1.5900486350704625e-06,
+      "loss": 1.4893,
+      "step": 8981
+    },
+    {
+      "epoch": 0.9449763282482904,
+      "grad_norm": 1.4413615465164185,
+      "learning_rate": 1.584001595703466e-06,
+      "loss": 1.8078,
+      "step": 8982
+    },
+    {
+      "epoch": 0.9450815360336665,
+      "grad_norm": 1.8441150188446045,
+      "learning_rate": 1.5779659850622797e-06,
+      "loss": 1.6129,
+      "step": 8983
+    },
+    {
+      "epoch": 0.9451867438190427,
+      "grad_norm": 1.5965230464935303,
+      "learning_rate": 1.5719418038477985e-06,
+      "loss": 2.0439,
+      "step": 8984
+    },
+    {
+      "epoch": 0.9452919516044187,
+      "grad_norm": 2.328061819076538,
+      "learning_rate": 1.5659290527596071e-06,
+      "loss": 1.4441,
+      "step": 8985
+    },
+    {
+      "epoch": 0.9453971593897948,
+      "grad_norm": 2.004664182662964,
+      "learning_rate": 1.559927732495936e-06,
+      "loss": 1.7041,
+      "step": 8986
+    },
+    {
+      "epoch": 0.945502367175171,
+      "grad_norm": 1.5392452478408813,
+      "learning_rate": 1.5539378437536944e-06,
+      "loss": 1.8988,
+      "step": 8987
+    },
+    {
+      "epoch": 0.945607574960547,
+      "grad_norm": 1.6413873434066772,
+      "learning_rate": 1.5479593872284926e-06,
+      "loss": 1.6044,
+      "step": 8988
+    },
+    {
+      "epoch": 0.9457127827459232,
+      "grad_norm": 2.1332480907440186,
+      "learning_rate": 1.5419923636145639e-06,
+      "loss": 1.8659,
+      "step": 8989
+    },
+    {
+      "epoch": 0.9458179905312993,
+      "grad_norm": 1.1487113237380981,
+      "learning_rate": 1.536036773604843e-06,
+      "loss": 1.6138,
+      "step": 8990
+    },
+    {
+      "epoch": 0.9459231983166755,
+      "grad_norm": 1.1979764699935913,
+      "learning_rate": 1.5300926178909435e-06,
+      "loss": 1.684,
+      "step": 8991
+    },
+    {
+      "epoch": 0.9460284061020515,
+      "grad_norm": 1.9243789911270142,
+      "learning_rate": 1.5241598971631354e-06,
+      "loss": 1.5863,
+      "step": 8992
+    },
+    {
+      "epoch": 0.9461336138874277,
+      "grad_norm": 1.8303254842758179,
+      "learning_rate": 1.5182386121103676e-06,
+      "loss": 1.7,
+      "step": 8993
+    },
+    {
+      "epoch": 0.9462388216728038,
+      "grad_norm": 1.7681361436843872,
+      "learning_rate": 1.5123287634202454e-06,
+      "loss": 1.521,
+      "step": 8994
+    },
+    {
+      "epoch": 0.9463440294581799,
+      "grad_norm": 1.3203057050704956,
+      "learning_rate": 1.5064303517790757e-06,
+      "loss": 1.8113,
+      "step": 8995
+    },
+    {
+      "epoch": 0.946449237243556,
+      "grad_norm": 1.5649878978729248,
+      "learning_rate": 1.5005433778718213e-06,
+      "loss": 1.4567,
+      "step": 8996
+    },
+    {
+      "epoch": 0.9465544450289322,
+      "grad_norm": 1.9078657627105713,
+      "learning_rate": 1.4946678423821248e-06,
+      "loss": 1.6237,
+      "step": 8997
+    },
+    {
+      "epoch": 0.9466596528143083,
+      "grad_norm": 1.2230523824691772,
+      "learning_rate": 1.4888037459922622e-06,
+      "loss": 1.6624,
+      "step": 8998
+    },
+    {
+      "epoch": 0.9467648605996843,
+      "grad_norm": 2.302837371826172,
+      "learning_rate": 1.4829510893832332e-06,
+      "loss": 1.7364,
+      "step": 8999
+    },
+    {
+      "epoch": 0.9468700683850605,
+      "grad_norm": 1.3794721364974976,
+      "learning_rate": 1.4771098732346943e-06,
+      "loss": 1.9772,
+      "step": 9000
+    },
+    {
+      "epoch": 0.9469752761704366,
+      "grad_norm": 1.796860694885254,
+      "learning_rate": 1.4712800982249474e-06,
+      "loss": 1.4391,
+      "step": 9001
+    },
+    {
+      "epoch": 0.9470804839558127,
+      "grad_norm": 1.5692015886306763,
+      "learning_rate": 1.4654617650310176e-06,
+      "loss": 1.5924,
+      "step": 9002
+    },
+    {
+      "epoch": 0.9471856917411888,
+      "grad_norm": 1.5295088291168213,
+      "learning_rate": 1.4596548743285198e-06,
+      "loss": 1.4613,
+      "step": 9003
+    },
+    {
+      "epoch": 0.947290899526565,
+      "grad_norm": 2.021333932876587,
+      "learning_rate": 1.453859426791826e-06,
+      "loss": 1.9719,
+      "step": 9004
+    },
+    {
+      "epoch": 0.9473961073119411,
+      "grad_norm": 1.1852548122406006,
+      "learning_rate": 1.448075423093942e-06,
+      "loss": 1.3462,
+      "step": 9005
+    },
+    {
+      "epoch": 0.9475013150973172,
+      "grad_norm": 1.628966212272644,
+      "learning_rate": 1.4423028639065195e-06,
+      "loss": 2.1885,
+      "step": 9006
+    },
+    {
+      "epoch": 0.9476065228826933,
+      "grad_norm": 1.4825149774551392,
+      "learning_rate": 1.4365417498999334e-06,
+      "loss": 1.4696,
+      "step": 9007
+    },
+    {
+      "epoch": 0.9477117306680695,
+      "grad_norm": 1.7557880878448486,
+      "learning_rate": 1.430792081743182e-06,
+      "loss": 1.3468,
+      "step": 9008
+    },
+    {
+      "epoch": 0.9478169384534455,
+      "grad_norm": 1.8458571434020996,
+      "learning_rate": 1.4250538601039642e-06,
+      "loss": 1.804,
+      "step": 9009
+    },
+    {
+      "epoch": 0.9479221462388217,
+      "grad_norm": 1.6594163179397583,
+      "learning_rate": 1.4193270856486585e-06,
+      "loss": 1.3729,
+      "step": 9010
+    },
+    {
+      "epoch": 0.9480273540241978,
+      "grad_norm": 1.11782968044281,
+      "learning_rate": 1.413611759042266e-06,
+      "loss": 1.8055,
+      "step": 9011
+    },
+    {
+      "epoch": 0.948132561809574,
+      "grad_norm": 1.8020600080490112,
+      "learning_rate": 1.4079078809485002e-06,
+      "loss": 1.6278,
+      "step": 9012
+    },
+    {
+      "epoch": 0.94823776959495,
+      "grad_norm": 1.9331527948379517,
+      "learning_rate": 1.4022154520297425e-06,
+      "loss": 1.819,
+      "step": 9013
+    },
+    {
+      "epoch": 0.9483429773803261,
+      "grad_norm": 1.2520698308944702,
+      "learning_rate": 1.396534472947031e-06,
+      "loss": 1.757,
+      "step": 9014
+    },
+    {
+      "epoch": 0.9484481851657023,
+      "grad_norm": 1.4876186847686768,
+      "learning_rate": 1.3908649443600707e-06,
+      "loss": 1.5025,
+      "step": 9015
+    },
+    {
+      "epoch": 0.9485533929510784,
+      "grad_norm": 1.5193793773651123,
+      "learning_rate": 1.3852068669272688e-06,
+      "loss": 1.8764,
+      "step": 9016
+    },
+    {
+      "epoch": 0.9486586007364545,
+      "grad_norm": 1.3971233367919922,
+      "learning_rate": 1.3795602413056442e-06,
+      "loss": 1.3161,
+      "step": 9017
+    },
+    {
+      "epoch": 0.9487638085218306,
+      "grad_norm": 1.4068975448608398,
+      "learning_rate": 1.3739250681509497e-06,
+      "loss": 1.2116,
+      "step": 9018
+    },
+    {
+      "epoch": 0.9488690163072068,
+      "grad_norm": 1.3958295583724976,
+      "learning_rate": 1.3683013481175621e-06,
+      "loss": 1.4589,
+      "step": 9019
+    },
+    {
+      "epoch": 0.9489742240925828,
+      "grad_norm": 1.2469528913497925,
+      "learning_rate": 1.3626890818585591e-06,
+      "loss": 1.808,
+      "step": 9020
+    },
+    {
+      "epoch": 0.949079431877959,
+      "grad_norm": 1.113645076751709,
+      "learning_rate": 1.3570882700256637e-06,
+      "loss": 1.4229,
+      "step": 9021
+    },
+    {
+      "epoch": 0.9491846396633351,
+      "grad_norm": 1.3229107856750488,
+      "learning_rate": 1.351498913269289e-06,
+      "loss": 1.8561,
+      "step": 9022
+    },
+    {
+      "epoch": 0.9492898474487113,
+      "grad_norm": 1.7333238124847412,
+      "learning_rate": 1.3459210122384934e-06,
+      "loss": 1.3316,
+      "step": 9023
+    },
+    {
+      "epoch": 0.9493950552340873,
+      "grad_norm": 1.8732205629348755,
+      "learning_rate": 1.340354567581037e-06,
+      "loss": 1.7108,
+      "step": 9024
+    },
+    {
+      "epoch": 0.9495002630194634,
+      "grad_norm": 1.6077425479888916,
+      "learning_rate": 1.3347995799433355e-06,
+      "loss": 1.6701,
+      "step": 9025
+    },
+    {
+      "epoch": 0.9496054708048396,
+      "grad_norm": 2.317784070968628,
+      "learning_rate": 1.3292560499704398e-06,
+      "loss": 1.5069,
+      "step": 9026
+    },
+    {
+      "epoch": 0.9497106785902156,
+      "grad_norm": 2.149060010910034,
+      "learning_rate": 1.3237239783061462e-06,
+      "loss": 0.9591,
+      "step": 9027
+    },
+    {
+      "epoch": 0.9498158863755918,
+      "grad_norm": 1.2673544883728027,
+      "learning_rate": 1.3182033655928404e-06,
+      "loss": 1.7899,
+      "step": 9028
+    },
+    {
+      "epoch": 0.9499210941609679,
+      "grad_norm": 1.8225202560424805,
+      "learning_rate": 1.3126942124716213e-06,
+      "loss": 1.5151,
+      "step": 9029
+    },
+    {
+      "epoch": 0.9500263019463441,
+      "grad_norm": 1.9195647239685059,
+      "learning_rate": 1.3071965195822656e-06,
+      "loss": 1.5733,
+      "step": 9030
+    },
+    {
+      "epoch": 0.9501315097317201,
+      "grad_norm": 2.377620220184326,
+      "learning_rate": 1.301710287563196e-06,
+      "loss": 1.1683,
+      "step": 9031
+    },
+    {
+      "epoch": 0.9502367175170963,
+      "grad_norm": 1.6789700984954834,
+      "learning_rate": 1.2962355170514917e-06,
+      "loss": 1.5605,
+      "step": 9032
+    },
+    {
+      "epoch": 0.9503419253024724,
+      "grad_norm": 1.4375206232070923,
+      "learning_rate": 1.2907722086829332e-06,
+      "loss": 1.3032,
+      "step": 9033
+    },
+    {
+      "epoch": 0.9504471330878485,
+      "grad_norm": 1.3653327226638794,
+      "learning_rate": 1.2853203630919686e-06,
+      "loss": 1.4516,
+      "step": 9034
+    },
+    {
+      "epoch": 0.9505523408732246,
+      "grad_norm": 1.7745347023010254,
+      "learning_rate": 1.279879980911669e-06,
+      "loss": 1.5384,
+      "step": 9035
+    },
+    {
+      "epoch": 0.9506575486586007,
+      "grad_norm": 1.7795803546905518,
+      "learning_rate": 1.2744510627738516e-06,
+      "loss": 1.8052,
+      "step": 9036
+    },
+    {
+      "epoch": 0.9507627564439769,
+      "grad_norm": 1.5467740297317505,
+      "learning_rate": 1.2690336093089228e-06,
+      "loss": 1.7491,
+      "step": 9037
+    },
+    {
+      "epoch": 0.9508679642293529,
+      "grad_norm": 1.2983478307724,
+      "learning_rate": 1.263627621146013e-06,
+      "loss": 1.6505,
+      "step": 9038
+    },
+    {
+      "epoch": 0.9509731720147291,
+      "grad_norm": 2.1444427967071533,
+      "learning_rate": 1.2582330989128977e-06,
+      "loss": 1.3857,
+      "step": 9039
+    },
+    {
+      "epoch": 0.9510783798001052,
+      "grad_norm": 1.9414992332458496,
+      "learning_rate": 1.252850043236009e-06,
+      "loss": 1.2304,
+      "step": 9040
+    },
+    {
+      "epoch": 0.9511835875854813,
+      "grad_norm": 1.6080853939056396,
+      "learning_rate": 1.2474784547404916e-06,
+      "loss": 1.8033,
+      "step": 9041
+    },
+    {
+      "epoch": 0.9512887953708574,
+      "grad_norm": 1.4745537042617798,
+      "learning_rate": 1.2421183340501242e-06,
+      "loss": 2.1338,
+      "step": 9042
+    },
+    {
+      "epoch": 0.9513940031562336,
+      "grad_norm": 1.2472604513168335,
+      "learning_rate": 1.2367696817873419e-06,
+      "loss": 2.3056,
+      "step": 9043
+    },
+    {
+      "epoch": 0.9514992109416097,
+      "grad_norm": 1.7281334400177002,
+      "learning_rate": 1.2314324985732818e-06,
+      "loss": 1.5415,
+      "step": 9044
+    },
+    {
+      "epoch": 0.9516044187269858,
+      "grad_norm": 1.563148856163025,
+      "learning_rate": 1.2261067850277252e-06,
+      "loss": 1.6814,
+      "step": 9045
+    },
+    {
+      "epoch": 0.9517096265123619,
+      "grad_norm": 1.6207623481750488,
+      "learning_rate": 1.2207925417691334e-06,
+      "loss": 1.566,
+      "step": 9046
+    },
+    {
+      "epoch": 0.951814834297738,
+      "grad_norm": 1.622622013092041,
+      "learning_rate": 1.215489769414635e-06,
+      "loss": 1.4526,
+      "step": 9047
+    },
+    {
+      "epoch": 0.9519200420831142,
+      "grad_norm": 1.5105351209640503,
+      "learning_rate": 1.210198468580015e-06,
+      "loss": 1.8531,
+      "step": 9048
+    },
+    {
+      "epoch": 0.9520252498684902,
+      "grad_norm": 1.3942266702651978,
+      "learning_rate": 1.2049186398797374e-06,
+      "loss": 1.8842,
+      "step": 9049
+    },
+    {
+      "epoch": 0.9521304576538664,
+      "grad_norm": 1.670791745185852,
+      "learning_rate": 1.1996502839269453e-06,
+      "loss": 1.9948,
+      "step": 9050
+    },
+    {
+      "epoch": 0.9522356654392425,
+      "grad_norm": 1.3902188539505005,
+      "learning_rate": 1.1943934013334047e-06,
+      "loss": 1.091,
+      "step": 9051
+    },
+    {
+      "epoch": 0.9523408732246186,
+      "grad_norm": 1.8161249160766602,
+      "learning_rate": 1.189147992709616e-06,
+      "loss": 2.1526,
+      "step": 9052
+    },
+    {
+      "epoch": 0.9524460810099947,
+      "grad_norm": 1.8096421957015991,
+      "learning_rate": 1.1839140586646923e-06,
+      "loss": 0.9295,
+      "step": 9053
+    },
+    {
+      "epoch": 0.9525512887953709,
+      "grad_norm": 1.7646183967590332,
+      "learning_rate": 1.178691599806425e-06,
+      "loss": 1.7518,
+      "step": 9054
+    },
+    {
+      "epoch": 0.952656496580747,
+      "grad_norm": 1.164070963859558,
+      "learning_rate": 1.173480616741296e-06,
+      "loss": 1.5824,
+      "step": 9055
+    },
+    {
+      "epoch": 0.9527617043661231,
+      "grad_norm": 1.876624584197998,
+      "learning_rate": 1.168281110074443e-06,
+      "loss": 1.4504,
+      "step": 9056
+    },
+    {
+      "epoch": 0.9528669121514992,
+      "grad_norm": 1.5865232944488525,
+      "learning_rate": 1.1630930804096495e-06,
+      "loss": 1.5255,
+      "step": 9057
+    },
+    {
+      "epoch": 0.9529721199368754,
+      "grad_norm": 1.653921127319336,
+      "learning_rate": 1.1579165283494009e-06,
+      "loss": 1.5343,
+      "step": 9058
+    },
+    {
+      "epoch": 0.9530773277222514,
+      "grad_norm": 3.066737651824951,
+      "learning_rate": 1.152751454494827e-06,
+      "loss": 1.3524,
+      "step": 9059
+    },
+    {
+      "epoch": 0.9531825355076275,
+      "grad_norm": 1.9732846021652222,
+      "learning_rate": 1.1475978594457149e-06,
+      "loss": 1.6233,
+      "step": 9060
+    },
+    {
+      "epoch": 0.9532877432930037,
+      "grad_norm": 1.4492088556289673,
+      "learning_rate": 1.1424557438005634e-06,
+      "loss": 1.6444,
+      "step": 9061
+    },
+    {
+      "epoch": 0.9533929510783798,
+      "grad_norm": 1.6626065969467163,
+      "learning_rate": 1.1373251081565062e-06,
+      "loss": 1.6292,
+      "step": 9062
+    },
+    {
+      "epoch": 0.9534981588637559,
+      "grad_norm": 1.2074167728424072,
+      "learning_rate": 1.132205953109311e-06,
+      "loss": 1.9072,
+      "step": 9063
+    },
+    {
+      "epoch": 0.953603366649132,
+      "grad_norm": 1.8209060430526733,
+      "learning_rate": 1.127098279253491e-06,
+      "loss": 1.4265,
+      "step": 9064
+    },
+    {
+      "epoch": 0.9537085744345082,
+      "grad_norm": 1.6529898643493652,
+      "learning_rate": 1.1220020871821723e-06,
+      "loss": 1.3912,
+      "step": 9065
+    },
+    {
+      "epoch": 0.9538137822198842,
+      "grad_norm": 1.3081694841384888,
+      "learning_rate": 1.1169173774871478e-06,
+      "loss": 1.8764,
+      "step": 9066
+    },
+    {
+      "epoch": 0.9539189900052604,
+      "grad_norm": 1.5082989931106567,
+      "learning_rate": 1.1118441507589006e-06,
+      "loss": 1.6502,
+      "step": 9067
+    },
+    {
+      "epoch": 0.9540241977906365,
+      "grad_norm": 1.7318092584609985,
+      "learning_rate": 1.1067824075865485e-06,
+      "loss": 1.3874,
+      "step": 9068
+    },
+    {
+      "epoch": 0.9541294055760127,
+      "grad_norm": 1.9931410551071167,
+      "learning_rate": 1.1017321485579102e-06,
+      "loss": 2.344,
+      "step": 9069
+    },
+    {
+      "epoch": 0.9542346133613887,
+      "grad_norm": 1.397017240524292,
+      "learning_rate": 1.096693374259461e-06,
+      "loss": 1.9131,
+      "step": 9070
+    },
+    {
+      "epoch": 0.9543398211467649,
+      "grad_norm": 0.9994755983352661,
+      "learning_rate": 1.0916660852763216e-06,
+      "loss": 1.4483,
+      "step": 9071
+    },
+    {
+      "epoch": 0.954445028932141,
+      "grad_norm": 1.567875623703003,
+      "learning_rate": 1.0866502821923031e-06,
+      "loss": 1.4536,
+      "step": 9072
+    },
+    {
+      "epoch": 0.954550236717517,
+      "grad_norm": 1.6587287187576294,
+      "learning_rate": 1.0816459655898726e-06,
+      "loss": 1.8396,
+      "step": 9073
+    },
+    {
+      "epoch": 0.9546554445028932,
+      "grad_norm": 1.5894458293914795,
+      "learning_rate": 1.0766531360501654e-06,
+      "loss": 1.3923,
+      "step": 9074
+    },
+    {
+      "epoch": 0.9547606522882693,
+      "grad_norm": 1.1351405382156372,
+      "learning_rate": 1.0716717941529841e-06,
+      "loss": 1.543,
+      "step": 9075
+    },
+    {
+      "epoch": 0.9548658600736455,
+      "grad_norm": 1.4005601406097412,
+      "learning_rate": 1.0667019404767996e-06,
+      "loss": 1.7229,
+      "step": 9076
+    },
+    {
+      "epoch": 0.9549710678590215,
+      "grad_norm": 2.1269547939300537,
+      "learning_rate": 1.0617435755987281e-06,
+      "loss": 1.5129,
+      "step": 9077
+    },
+    {
+      "epoch": 0.9550762756443977,
+      "grad_norm": 1.9350544214248657,
+      "learning_rate": 1.0567967000945866e-06,
+      "loss": 1.4607,
+      "step": 9078
+    },
+    {
+      "epoch": 0.9551814834297738,
+      "grad_norm": 1.4004015922546387,
+      "learning_rate": 1.0518613145388378e-06,
+      "loss": 1.907,
+      "step": 9079
+    },
+    {
+      "epoch": 0.95528669121515,
+      "grad_norm": 1.2579517364501953,
+      "learning_rate": 1.04693741950459e-06,
+      "loss": 1.4664,
+      "step": 9080
+    },
+    {
+      "epoch": 0.955391899000526,
+      "grad_norm": 1.7487457990646362,
+      "learning_rate": 1.0420250155636745e-06,
+      "loss": 1.6682,
+      "step": 9081
+    },
+    {
+      "epoch": 0.9554971067859022,
+      "grad_norm": 2.7791035175323486,
+      "learning_rate": 1.0371241032865242e-06,
+      "loss": 1.2912,
+      "step": 9082
+    },
+    {
+      "epoch": 0.9556023145712783,
+      "grad_norm": 1.7350082397460938,
+      "learning_rate": 1.0322346832422613e-06,
+      "loss": 1.7033,
+      "step": 9083
+    },
+    {
+      "epoch": 0.9557075223566543,
+      "grad_norm": 2.800816535949707,
+      "learning_rate": 1.0273567559987097e-06,
+      "loss": 1.7198,
+      "step": 9084
+    },
+    {
+      "epoch": 0.9558127301420305,
+      "grad_norm": 1.4789823293685913,
+      "learning_rate": 1.0224903221222938e-06,
+      "loss": 1.9354,
+      "step": 9085
+    },
+    {
+      "epoch": 0.9559179379274066,
+      "grad_norm": 2.029716968536377,
+      "learning_rate": 1.0176353821781616e-06,
+      "loss": 1.2883,
+      "step": 9086
+    },
+    {
+      "epoch": 0.9560231457127828,
+      "grad_norm": 1.2375006675720215,
+      "learning_rate": 1.0127919367301064e-06,
+      "loss": 1.7844,
+      "step": 9087
+    },
+    {
+      "epoch": 0.9561283534981588,
+      "grad_norm": 1.4403471946716309,
+      "learning_rate": 1.0079599863405454e-06,
+      "loss": 1.4304,
+      "step": 9088
+    },
+    {
+      "epoch": 0.956233561283535,
+      "grad_norm": 1.9305444955825806,
+      "learning_rate": 1.0031395315706183e-06,
+      "loss": 1.041,
+      "step": 9089
+    },
+    {
+      "epoch": 0.9563387690689111,
+      "grad_norm": 1.9020962715148926,
+      "learning_rate": 9.983305729801107e-07,
+      "loss": 1.609,
+      "step": 9090
+    },
+    {
+      "epoch": 0.9564439768542872,
+      "grad_norm": 2.1255598068237305,
+      "learning_rate": 9.93533111127476e-07,
+      "loss": 1.4914,
+      "step": 9091
+    },
+    {
+      "epoch": 0.9565491846396633,
+      "grad_norm": 2.1392078399658203,
+      "learning_rate": 9.88747146569813e-07,
+      "loss": 1.6591,
+      "step": 9092
+    },
+    {
+      "epoch": 0.9566543924250395,
+      "grad_norm": 1.4976539611816406,
+      "learning_rate": 9.8397267986291e-07,
+      "loss": 1.5315,
+      "step": 9093
+    },
+    {
+      "epoch": 0.9567596002104156,
+      "grad_norm": 1.4068844318389893,
+      "learning_rate": 9.79209711561202e-07,
+      "loss": 1.2471,
+      "step": 9094
+    },
+    {
+      "epoch": 0.9568648079957917,
+      "grad_norm": 1.8085061311721802,
+      "learning_rate": 9.744582422178127e-07,
+      "loss": 1.3131,
+      "step": 9095
+    },
+    {
+      "epoch": 0.9569700157811678,
+      "grad_norm": 1.4883860349655151,
+      "learning_rate": 9.6971827238449e-07,
+      "loss": 1.8711,
+      "step": 9096
+    },
+    {
+      "epoch": 0.957075223566544,
+      "grad_norm": 1.6276978254318237,
+      "learning_rate": 9.649898026117043e-07,
+      "loss": 1.5425,
+      "step": 9097
+    },
+    {
+      "epoch": 0.95718043135192,
+      "grad_norm": 1.1478794813156128,
+      "learning_rate": 9.602728334485278e-07,
+      "loss": 1.746,
+      "step": 9098
+    },
+    {
+      "epoch": 0.9572856391372961,
+      "grad_norm": 1.3659400939941406,
+      "learning_rate": 9.555673654427332e-07,
+      "loss": 1.3815,
+      "step": 9099
+    },
+    {
+      "epoch": 0.9573908469226723,
+      "grad_norm": 1.6303960084915161,
+      "learning_rate": 9.508733991407615e-07,
+      "loss": 1.6709,
+      "step": 9100
+    },
+    {
+      "epoch": 0.9574960547080484,
+      "grad_norm": 1.6804249286651611,
+      "learning_rate": 9.461909350876985e-07,
+      "loss": 1.5344,
+      "step": 9101
+    },
+    {
+      "epoch": 0.9576012624934245,
+      "grad_norm": 1.8790122270584106,
+      "learning_rate": 9.415199738272984e-07,
+      "loss": 1.4744,
+      "step": 9102
+    },
+    {
+      "epoch": 0.9577064702788006,
+      "grad_norm": 1.4778565168380737,
+      "learning_rate": 9.36860515902005e-07,
+      "loss": 1.7999,
+      "step": 9103
+    },
+    {
+      "epoch": 0.9578116780641768,
+      "grad_norm": 1.2392534017562866,
+      "learning_rate": 9.322125618528854e-07,
+      "loss": 1.7232,
+      "step": 9104
+    },
+    {
+      "epoch": 0.9579168858495528,
+      "grad_norm": 1.5124746561050415,
+      "learning_rate": 9.275761122196968e-07,
+      "loss": 2.0347,
+      "step": 9105
+    },
+    {
+      "epoch": 0.958022093634929,
+      "grad_norm": 1.9507633447647095,
+      "learning_rate": 9.229511675408642e-07,
+      "loss": 1.6567,
+      "step": 9106
+    },
+    {
+      "epoch": 0.9581273014203051,
+      "grad_norm": 1.7960712909698486,
+      "learning_rate": 9.183377283534578e-07,
+      "loss": 1.7705,
+      "step": 9107
+    },
+    {
+      "epoch": 0.9582325092056813,
+      "grad_norm": 2.1413114070892334,
+      "learning_rate": 9.137357951932157e-07,
+      "loss": 1.7895,
+      "step": 9108
+    },
+    {
+      "epoch": 0.9583377169910573,
+      "grad_norm": 1.709755301475525,
+      "learning_rate": 9.091453685945661e-07,
+      "loss": 1.7887,
+      "step": 9109
+    },
+    {
+      "epoch": 0.9584429247764334,
+      "grad_norm": 1.2714898586273193,
+      "learning_rate": 9.045664490905492e-07,
+      "loss": 1.8393,
+      "step": 9110
+    },
+    {
+      "epoch": 0.9585481325618096,
+      "grad_norm": 1.3688026666641235,
+      "learning_rate": 8.999990372129286e-07,
+      "loss": 1.6004,
+      "step": 9111
+    },
+    {
+      "epoch": 0.9586533403471857,
+      "grad_norm": 1.3194514513015747,
+      "learning_rate": 8.95443133492091e-07,
+      "loss": 1.1129,
+      "step": 9112
+    },
+    {
+      "epoch": 0.9587585481325618,
+      "grad_norm": 1.4183218479156494,
+      "learning_rate": 8.90898738457091e-07,
+      "loss": 1.8257,
+      "step": 9113
+    },
+    {
+      "epoch": 0.9588637559179379,
+      "grad_norm": 1.4666954278945923,
+      "learning_rate": 8.863658526356622e-07,
+      "loss": 1.6854,
+      "step": 9114
+    },
+    {
+      "epoch": 0.9589689637033141,
+      "grad_norm": 1.620832085609436,
+      "learning_rate": 8.818444765541944e-07,
+      "loss": 1.9613,
+      "step": 9115
+    },
+    {
+      "epoch": 0.9590741714886901,
+      "grad_norm": 1.858279824256897,
+      "learning_rate": 8.773346107377456e-07,
+      "loss": 1.7767,
+      "step": 9116
+    },
+    {
+      "epoch": 0.9591793792740663,
+      "grad_norm": 1.720293402671814,
+      "learning_rate": 8.72836255710019e-07,
+      "loss": 1.8888,
+      "step": 9117
+    },
+    {
+      "epoch": 0.9592845870594424,
+      "grad_norm": 1.7517911195755005,
+      "learning_rate": 8.683494119934076e-07,
+      "loss": 1.449,
+      "step": 9118
+    },
+    {
+      "epoch": 0.9593897948448186,
+      "grad_norm": 0.9157357215881348,
+      "learning_rate": 8.638740801089396e-07,
+      "loss": 1.5477,
+      "step": 9119
+    },
+    {
+      "epoch": 0.9594950026301946,
+      "grad_norm": 1.0953036546707153,
+      "learning_rate": 8.59410260576321e-07,
+      "loss": 1.6595,
+      "step": 9120
+    },
+    {
+      "epoch": 0.9596002104155708,
+      "grad_norm": 1.7858799695968628,
+      "learning_rate": 8.549579539139374e-07,
+      "loss": 1.6835,
+      "step": 9121
+    },
+    {
+      "epoch": 0.9597054182009469,
+      "grad_norm": 1.9369888305664062,
+      "learning_rate": 8.505171606388085e-07,
+      "loss": 2.1841,
+      "step": 9122
+    },
+    {
+      "epoch": 0.9598106259863229,
+      "grad_norm": 1.8354744911193848,
+      "learning_rate": 8.460878812666217e-07,
+      "loss": 1.823,
+      "step": 9123
+    },
+    {
+      "epoch": 0.9599158337716991,
+      "grad_norm": 1.8804636001586914,
+      "learning_rate": 8.416701163117546e-07,
+      "loss": 2.0077,
+      "step": 9124
+    },
+    {
+      "epoch": 0.9600210415570752,
+      "grad_norm": 1.981581449508667,
+      "learning_rate": 8.372638662872079e-07,
+      "loss": 1.5237,
+      "step": 9125
+    },
+    {
+      "epoch": 0.9601262493424514,
+      "grad_norm": 1.4924275875091553,
+      "learning_rate": 8.328691317046722e-07,
+      "loss": 1.5502,
+      "step": 9126
+    },
+    {
+      "epoch": 0.9602314571278274,
+      "grad_norm": 1.8991860151290894,
+      "learning_rate": 8.28485913074506e-07,
+      "loss": 1.5965,
+      "step": 9127
+    },
+    {
+      "epoch": 0.9603366649132036,
+      "grad_norm": 0.9699298739433289,
+      "learning_rate": 8.24114210905691e-07,
+      "loss": 2.2511,
+      "step": 9128
+    },
+    {
+      "epoch": 0.9604418726985797,
+      "grad_norm": 1.3279988765716553,
+      "learning_rate": 8.197540257059321e-07,
+      "loss": 1.6017,
+      "step": 9129
+    },
+    {
+      "epoch": 0.9605470804839558,
+      "grad_norm": 1.4322397708892822,
+      "learning_rate": 8.154053579815358e-07,
+      "loss": 1.5948,
+      "step": 9130
+    },
+    {
+      "epoch": 0.9606522882693319,
+      "grad_norm": 1.4159713983535767,
+      "learning_rate": 8.110682082375087e-07,
+      "loss": 1.4411,
+      "step": 9131
+    },
+    {
+      "epoch": 0.9607574960547081,
+      "grad_norm": 1.7567311525344849,
+      "learning_rate": 8.067425769775039e-07,
+      "loss": 1.3948,
+      "step": 9132
+    },
+    {
+      "epoch": 0.9608627038400842,
+      "grad_norm": 2.1508755683898926,
+      "learning_rate": 8.024284647038527e-07,
+      "loss": 1.8868,
+      "step": 9133
+    },
+    {
+      "epoch": 0.9609679116254602,
+      "grad_norm": 2.318276882171631,
+      "learning_rate": 7.981258719175322e-07,
+      "loss": 1.4337,
+      "step": 9134
+    },
+    {
+      "epoch": 0.9610731194108364,
+      "grad_norm": 1.1780331134796143,
+      "learning_rate": 7.938347991181982e-07,
+      "loss": 1.6706,
+      "step": 9135
+    },
+    {
+      "epoch": 0.9611783271962125,
+      "grad_norm": 2.116865634918213,
+      "learning_rate": 7.895552468041412e-07,
+      "loss": 1.8336,
+      "step": 9136
+    },
+    {
+      "epoch": 0.9612835349815886,
+      "grad_norm": 1.3692445755004883,
+      "learning_rate": 7.852872154723412e-07,
+      "loss": 1.9935,
+      "step": 9137
+    },
+    {
+      "epoch": 0.9613887427669647,
+      "grad_norm": 1.171465277671814,
+      "learning_rate": 7.810307056184352e-07,
+      "loss": 1.6881,
+      "step": 9138
+    },
+    {
+      "epoch": 0.9614939505523409,
+      "grad_norm": 1.1161651611328125,
+      "learning_rate": 7.767857177367055e-07,
+      "loss": 1.6728,
+      "step": 9139
+    },
+    {
+      "epoch": 0.961599158337717,
+      "grad_norm": 1.869456171989441,
+      "learning_rate": 7.725522523201245e-07,
+      "loss": 1.1257,
+      "step": 9140
+    },
+    {
+      "epoch": 0.9617043661230931,
+      "grad_norm": 1.2645540237426758,
+      "learning_rate": 7.683303098602989e-07,
+      "loss": 1.4703,
+      "step": 9141
+    },
+    {
+      "epoch": 0.9618095739084692,
+      "grad_norm": 1.4299430847167969,
+      "learning_rate": 7.641198908475144e-07,
+      "loss": 1.6185,
+      "step": 9142
+    },
+    {
+      "epoch": 0.9619147816938454,
+      "grad_norm": 2.149198532104492,
+      "learning_rate": 7.599209957707021e-07,
+      "loss": 1.4881,
+      "step": 9143
+    },
+    {
+      "epoch": 0.9620199894792215,
+      "grad_norm": 1.9210573434829712,
+      "learning_rate": 7.55733625117483e-07,
+      "loss": 2.0232,
+      "step": 9144
+    },
+    {
+      "epoch": 0.9621251972645976,
+      "grad_norm": 2.284515380859375,
+      "learning_rate": 7.515577793741124e-07,
+      "loss": 1.4036,
+      "step": 9145
+    },
+    {
+      "epoch": 0.9622304050499737,
+      "grad_norm": 1.6469871997833252,
+      "learning_rate": 7.473934590255249e-07,
+      "loss": 1.4179,
+      "step": 9146
+    },
+    {
+      "epoch": 0.9623356128353499,
+      "grad_norm": 1.3524980545043945,
+      "learning_rate": 7.432406645552891e-07,
+      "loss": 1.4155,
+      "step": 9147
+    },
+    {
+      "epoch": 0.9624408206207259,
+      "grad_norm": 1.9270532131195068,
+      "learning_rate": 7.39099396445686e-07,
+      "loss": 1.7492,
+      "step": 9148
+    },
+    {
+      "epoch": 0.962546028406102,
+      "grad_norm": 1.6175739765167236,
+      "learning_rate": 7.349696551776086e-07,
+      "loss": 2.1136,
+      "step": 9149
+    },
+    {
+      "epoch": 0.9626512361914782,
+      "grad_norm": 1.5902010202407837,
+      "learning_rate": 7.308514412306289e-07,
+      "loss": 1.8893,
+      "step": 9150
+    },
+    {
+      "epoch": 0.9627564439768543,
+      "grad_norm": 1.2558472156524658,
+      "learning_rate": 7.267447550829865e-07,
+      "loss": 1.5318,
+      "step": 9151
+    },
+    {
+      "epoch": 0.9628616517622304,
+      "grad_norm": 1.7991538047790527,
+      "learning_rate": 7.226495972115776e-07,
+      "loss": 1.2432,
+      "step": 9152
+    },
+    {
+      "epoch": 0.9629668595476065,
+      "grad_norm": 1.6260236501693726,
+      "learning_rate": 7.185659680919554e-07,
+      "loss": 1.7901,
+      "step": 9153
+    },
+    {
+      "epoch": 0.9630720673329827,
+      "grad_norm": 1.319873571395874,
+      "learning_rate": 7.144938681983515e-07,
+      "loss": 1.4871,
+      "step": 9154
+    },
+    {
+      "epoch": 0.9631772751183587,
+      "grad_norm": 1.6436864137649536,
+      "learning_rate": 7.104332980036211e-07,
+      "loss": 1.2697,
+      "step": 9155
+    },
+    {
+      "epoch": 0.9632824829037349,
+      "grad_norm": 1.0608686208724976,
+      "learning_rate": 7.063842579793311e-07,
+      "loss": 1.8403,
+      "step": 9156
+    },
+    {
+      "epoch": 0.963387690689111,
+      "grad_norm": 1.5618516206741333,
+      "learning_rate": 7.023467485956614e-07,
+      "loss": 1.4696,
+      "step": 9157
+    },
+    {
+      "epoch": 0.9634928984744872,
+      "grad_norm": 1.3680403232574463,
+      "learning_rate": 6.983207703214811e-07,
+      "loss": 1.6647,
+      "step": 9158
+    },
+    {
+      "epoch": 0.9635981062598632,
+      "grad_norm": 2.8899571895599365,
+      "learning_rate": 6.943063236243275e-07,
+      "loss": 1.7808,
+      "step": 9159
+    },
+    {
+      "epoch": 0.9637033140452393,
+      "grad_norm": 1.5044270753860474,
+      "learning_rate": 6.903034089703719e-07,
+      "loss": 1.8185,
+      "step": 9160
+    },
+    {
+      "epoch": 0.9638085218306155,
+      "grad_norm": 1.38004469871521,
+      "learning_rate": 6.863120268244649e-07,
+      "loss": 1.6207,
+      "step": 9161
+    },
+    {
+      "epoch": 0.9639137296159915,
+      "grad_norm": 2.003220558166504,
+      "learning_rate": 6.823321776501024e-07,
+      "loss": 1.5458,
+      "step": 9162
+    },
+    {
+      "epoch": 0.9640189374013677,
+      "grad_norm": 1.4244567155838013,
+      "learning_rate": 6.783638619094701e-07,
+      "loss": 1.2926,
+      "step": 9163
+    },
+    {
+      "epoch": 0.9641241451867438,
+      "grad_norm": 1.428270697593689,
+      "learning_rate": 6.744070800633661e-07,
+      "loss": 1.5582,
+      "step": 9164
+    },
+    {
+      "epoch": 0.96422935297212,
+      "grad_norm": 1.64469313621521,
+      "learning_rate": 6.704618325713119e-07,
+      "loss": 1.7808,
+      "step": 9165
+    },
+    {
+      "epoch": 0.964334560757496,
+      "grad_norm": 1.6442152261734009,
+      "learning_rate": 6.665281198914408e-07,
+      "loss": 1.4003,
+      "step": 9166
+    },
+    {
+      "epoch": 0.9644397685428722,
+      "grad_norm": 1.406569480895996,
+      "learning_rate": 6.626059424805542e-07,
+      "loss": 1.4821,
+      "step": 9167
+    },
+    {
+      "epoch": 0.9645449763282483,
+      "grad_norm": 1.100145936012268,
+      "learning_rate": 6.586953007941321e-07,
+      "loss": 1.6582,
+      "step": 9168
+    },
+    {
+      "epoch": 0.9646501841136244,
+      "grad_norm": 1.651857614517212,
+      "learning_rate": 6.547961952863002e-07,
+      "loss": 1.7471,
+      "step": 9169
+    },
+    {
+      "epoch": 0.9647553918990005,
+      "grad_norm": 1.3238495588302612,
+      "learning_rate": 6.509086264098407e-07,
+      "loss": 1.7772,
+      "step": 9170
+    },
+    {
+      "epoch": 0.9648605996843767,
+      "grad_norm": 2.2787668704986572,
+      "learning_rate": 6.470325946162259e-07,
+      "loss": 1.3509,
+      "step": 9171
+    },
+    {
+      "epoch": 0.9649658074697528,
+      "grad_norm": 1.7403054237365723,
+      "learning_rate": 6.431681003555622e-07,
+      "loss": 1.4681,
+      "step": 9172
+    },
+    {
+      "epoch": 0.9650710152551288,
+      "grad_norm": 1.6274385452270508,
+      "learning_rate": 6.393151440766021e-07,
+      "loss": 1.6136,
+      "step": 9173
+    },
+    {
+      "epoch": 0.965176223040505,
+      "grad_norm": 1.5703790187835693,
+      "learning_rate": 6.354737262267873e-07,
+      "loss": 1.4191,
+      "step": 9174
+    },
+    {
+      "epoch": 0.9652814308258811,
+      "grad_norm": 1.5442698001861572,
+      "learning_rate": 6.316438472522057e-07,
+      "loss": 1.1351,
+      "step": 9175
+    },
+    {
+      "epoch": 0.9653866386112573,
+      "grad_norm": 1.7261892557144165,
+      "learning_rate": 6.278255075976125e-07,
+      "loss": 1.1987,
+      "step": 9176
+    },
+    {
+      "epoch": 0.9654918463966333,
+      "grad_norm": 1.5503300428390503,
+      "learning_rate": 6.240187077064307e-07,
+      "loss": 1.8204,
+      "step": 9177
+    },
+    {
+      "epoch": 0.9655970541820095,
+      "grad_norm": 1.6254150867462158,
+      "learning_rate": 6.202234480207069e-07,
+      "loss": 1.6718,
+      "step": 9178
+    },
+    {
+      "epoch": 0.9657022619673856,
+      "grad_norm": 1.0417579412460327,
+      "learning_rate": 6.164397289811885e-07,
+      "loss": 1.3387,
+      "step": 9179
+    },
+    {
+      "epoch": 0.9658074697527617,
+      "grad_norm": 1.2805875539779663,
+      "learning_rate": 6.126675510272572e-07,
+      "loss": 1.2551,
+      "step": 9180
+    },
+    {
+      "epoch": 0.9659126775381378,
+      "grad_norm": 1.3211545944213867,
+      "learning_rate": 6.089069145969739e-07,
+      "loss": 1.65,
+      "step": 9181
+    },
+    {
+      "epoch": 0.966017885323514,
+      "grad_norm": 4.055902481079102,
+      "learning_rate": 6.051578201270336e-07,
+      "loss": 1.776,
+      "step": 9182
+    },
+    {
+      "epoch": 0.9661230931088901,
+      "grad_norm": 1.3434686660766602,
+      "learning_rate": 6.014202680528324e-07,
+      "loss": 1.4995,
+      "step": 9183
+    },
+    {
+      "epoch": 0.9662283008942661,
+      "grad_norm": 1.8617429733276367,
+      "learning_rate": 5.976942588083678e-07,
+      "loss": 1.7132,
+      "step": 9184
+    },
+    {
+      "epoch": 0.9663335086796423,
+      "grad_norm": 1.6335538625717163,
+      "learning_rate": 5.939797928263602e-07,
+      "loss": 1.5556,
+      "step": 9185
+    },
+    {
+      "epoch": 0.9664387164650184,
+      "grad_norm": 1.5544825792312622,
+      "learning_rate": 5.902768705381312e-07,
+      "loss": 1.736,
+      "step": 9186
+    },
+    {
+      "epoch": 0.9665439242503945,
+      "grad_norm": 1.460477590560913,
+      "learning_rate": 5.865854923737035e-07,
+      "loss": 1.4154,
+      "step": 9187
+    },
+    {
+      "epoch": 0.9666491320357706,
+      "grad_norm": 1.5231237411499023,
+      "learning_rate": 5.829056587617455e-07,
+      "loss": 1.3271,
+      "step": 9188
+    },
+    {
+      "epoch": 0.9667543398211468,
+      "grad_norm": 1.2489863634109497,
+      "learning_rate": 5.792373701295706e-07,
+      "loss": 1.9272,
+      "step": 9189
+    },
+    {
+      "epoch": 0.9668595476065229,
+      "grad_norm": 1.2746580839157104,
+      "learning_rate": 5.755806269031827e-07,
+      "loss": 1.4012,
+      "step": 9190
+    },
+    {
+      "epoch": 0.966964755391899,
+      "grad_norm": 1.310562252998352,
+      "learning_rate": 5.719354295072199e-07,
+      "loss": 1.9932,
+      "step": 9191
+    },
+    {
+      "epoch": 0.9670699631772751,
+      "grad_norm": 1.565401554107666,
+      "learning_rate": 5.68301778364988e-07,
+      "loss": 1.5477,
+      "step": 9192
+    },
+    {
+      "epoch": 0.9671751709626513,
+      "grad_norm": 1.3643187284469604,
+      "learning_rate": 5.646796738984495e-07,
+      "loss": 1.7313,
+      "step": 9193
+    },
+    {
+      "epoch": 0.9672803787480273,
+      "grad_norm": 2.277108907699585,
+      "learning_rate": 5.610691165282234e-07,
+      "loss": 1.4127,
+      "step": 9194
+    },
+    {
+      "epoch": 0.9673855865334035,
+      "grad_norm": 1.1328575611114502,
+      "learning_rate": 5.574701066735965e-07,
+      "loss": 2.0227,
+      "step": 9195
+    },
+    {
+      "epoch": 0.9674907943187796,
+      "grad_norm": 1.7198408842086792,
+      "learning_rate": 5.538826447525125e-07,
+      "loss": 1.7175,
+      "step": 9196
+    },
+    {
+      "epoch": 0.9675960021041558,
+      "grad_norm": 1.514518141746521,
+      "learning_rate": 5.503067311815713e-07,
+      "loss": 1.512,
+      "step": 9197
+    },
+    {
+      "epoch": 0.9677012098895318,
+      "grad_norm": 1.6654568910598755,
+      "learning_rate": 5.467423663760296e-07,
+      "loss": 1.337,
+      "step": 9198
+    },
+    {
+      "epoch": 0.9678064176749079,
+      "grad_norm": 1.9775021076202393,
+      "learning_rate": 5.431895507498008e-07,
+      "loss": 1.1514,
+      "step": 9199
+    },
+    {
+      "epoch": 0.9679116254602841,
+      "grad_norm": 1.0919164419174194,
+      "learning_rate": 5.39648284715466e-07,
+      "loss": 1.7102,
+      "step": 9200
+    },
+    {
+      "epoch": 0.9680168332456601,
+      "grad_norm": 1.3616853952407837,
+      "learning_rate": 5.361185686842629e-07,
+      "loss": 1.7125,
+      "step": 9201
+    },
+    {
+      "epoch": 0.9681220410310363,
+      "grad_norm": 1.843559741973877,
+      "learning_rate": 5.326004030660747e-07,
+      "loss": 1.1107,
+      "step": 9202
+    },
+    {
+      "epoch": 0.9682272488164124,
+      "grad_norm": 1.5905053615570068,
+      "learning_rate": 5.290937882694746e-07,
+      "loss": 1.4774,
+      "step": 9203
+    },
+    {
+      "epoch": 0.9683324566017886,
+      "grad_norm": 1.699774146080017,
+      "learning_rate": 5.255987247016591e-07,
+      "loss": 1.6351,
+      "step": 9204
+    },
+    {
+      "epoch": 0.9684376643871646,
+      "grad_norm": 2.0290257930755615,
+      "learning_rate": 5.221152127685036e-07,
+      "loss": 2.2553,
+      "step": 9205
+    },
+    {
+      "epoch": 0.9685428721725408,
+      "grad_norm": 2.343132972717285,
+      "learning_rate": 5.186432528745289e-07,
+      "loss": 0.9011,
+      "step": 9206
+    },
+    {
+      "epoch": 0.9686480799579169,
+      "grad_norm": 2.2507107257843018,
+      "learning_rate": 5.151828454229346e-07,
+      "loss": 1.7659,
+      "step": 9207
+    },
+    {
+      "epoch": 0.9687532877432931,
+      "grad_norm": 1.9188868999481201,
+      "learning_rate": 5.117339908155549e-07,
+      "loss": 1.9411,
+      "step": 9208
+    },
+    {
+      "epoch": 0.9688584955286691,
+      "grad_norm": 1.9085512161254883,
+      "learning_rate": 5.082966894529028e-07,
+      "loss": 1.3046,
+      "step": 9209
+    },
+    {
+      "epoch": 0.9689637033140452,
+      "grad_norm": 1.6628177165985107,
+      "learning_rate": 5.04870941734148e-07,
+      "loss": 1.3953,
+      "step": 9210
+    },
+    {
+      "epoch": 0.9690689110994214,
+      "grad_norm": 1.5158159732818604,
+      "learning_rate": 5.014567480570831e-07,
+      "loss": 1.9355,
+      "step": 9211
+    },
+    {
+      "epoch": 0.9691741188847974,
+      "grad_norm": 1.3943663835525513,
+      "learning_rate": 4.980541088182133e-07,
+      "loss": 1.5488,
+      "step": 9212
+    },
+    {
+      "epoch": 0.9692793266701736,
+      "grad_norm": 1.2974984645843506,
+      "learning_rate": 4.946630244126782e-07,
+      "loss": 1.4926,
+      "step": 9213
+    },
+    {
+      "epoch": 0.9693845344555497,
+      "grad_norm": 1.5148580074310303,
+      "learning_rate": 4.912834952342515e-07,
+      "loss": 1.7226,
+      "step": 9214
+    },
+    {
+      "epoch": 0.9694897422409259,
+      "grad_norm": 1.8382201194763184,
+      "learning_rate": 4.879155216753972e-07,
+      "loss": 1.6377,
+      "step": 9215
+    },
+    {
+      "epoch": 0.9695949500263019,
+      "grad_norm": 1.510947823524475,
+      "learning_rate": 4.845591041272358e-07,
+      "loss": 2.0674,
+      "step": 9216
+    },
+    {
+      "epoch": 0.9697001578116781,
+      "grad_norm": 2.992677927017212,
+      "learning_rate": 4.812142429795219e-07,
+      "loss": 1.6575,
+      "step": 9217
+    },
+    {
+      "epoch": 0.9698053655970542,
+      "grad_norm": 2.957728147506714,
+      "learning_rate": 4.778809386206895e-07,
+      "loss": 1.6637,
+      "step": 9218
+    },
+    {
+      "epoch": 0.9699105733824303,
+      "grad_norm": 1.4280604124069214,
+      "learning_rate": 4.745591914378289e-07,
+      "loss": 1.9542,
+      "step": 9219
+    },
+    {
+      "epoch": 0.9700157811678064,
+      "grad_norm": 2.1100828647613525,
+      "learning_rate": 4.71249001816676e-07,
+      "loss": 1.6198,
+      "step": 9220
+    },
+    {
+      "epoch": 0.9701209889531826,
+      "grad_norm": 1.218092918395996,
+      "learning_rate": 4.6795037014163436e-07,
+      "loss": 1.2455,
+      "step": 9221
+    },
+    {
+      "epoch": 0.9702261967385587,
+      "grad_norm": 1.3231292963027954,
+      "learning_rate": 4.6466329679577536e-07,
+      "loss": 1.4456,
+      "step": 9222
+    },
+    {
+      "epoch": 0.9703314045239347,
+      "grad_norm": 1.482961893081665,
+      "learning_rate": 4.613877821607937e-07,
+      "loss": 2.1301,
+      "step": 9223
+    },
+    {
+      "epoch": 0.9704366123093109,
+      "grad_norm": 1.3625333309173584,
+      "learning_rate": 4.581238266170851e-07,
+      "loss": 1.631,
+      "step": 9224
+    },
+    {
+      "epoch": 0.970541820094687,
+      "grad_norm": 1.9850952625274658,
+      "learning_rate": 4.548714305436685e-07,
+      "loss": 1.8183,
+      "step": 9225
+    },
+    {
+      "epoch": 0.9706470278800631,
+      "grad_norm": 2.0296475887298584,
+      "learning_rate": 4.5163059431824194e-07,
+      "loss": 1.5821,
+      "step": 9226
+    },
+    {
+      "epoch": 0.9707522356654392,
+      "grad_norm": 1.2871662378311157,
+      "learning_rate": 4.484013183171376e-07,
+      "loss": 1.5999,
+      "step": 9227
+    },
+    {
+      "epoch": 0.9708574434508154,
+      "grad_norm": 1.5916684865951538,
+      "learning_rate": 4.4518360291538883e-07,
+      "loss": 1.8337,
+      "step": 9228
+    },
+    {
+      "epoch": 0.9709626512361915,
+      "grad_norm": 1.2991007566452026,
+      "learning_rate": 4.4197744848663017e-07,
+      "loss": 1.2721,
+      "step": 9229
+    },
+    {
+      "epoch": 0.9710678590215676,
+      "grad_norm": 1.4421882629394531,
+      "learning_rate": 4.3878285540319694e-07,
+      "loss": 1.684,
+      "step": 9230
+    },
+    {
+      "epoch": 0.9711730668069437,
+      "grad_norm": 1.5107342004776,
+      "learning_rate": 4.355998240360704e-07,
+      "loss": 1.1753,
+      "step": 9231
+    },
+    {
+      "epoch": 0.9712782745923199,
+      "grad_norm": 1.8420056104660034,
+      "learning_rate": 4.324283547548658e-07,
+      "loss": 1.5455,
+      "step": 9232
+    },
+    {
+      "epoch": 0.9713834823776959,
+      "grad_norm": 2.1014416217803955,
+      "learning_rate": 4.2926844792789967e-07,
+      "loss": 1.3949,
+      "step": 9233
+    },
+    {
+      "epoch": 0.971488690163072,
+      "grad_norm": 1.452222466468811,
+      "learning_rate": 4.261201039221008e-07,
+      "loss": 1.5272,
+      "step": 9234
+    },
+    {
+      "epoch": 0.9715938979484482,
+      "grad_norm": 1.7641346454620361,
+      "learning_rate": 4.2298332310308775e-07,
+      "loss": 1.3401,
+      "step": 9235
+    },
+    {
+      "epoch": 0.9716991057338243,
+      "grad_norm": 1.8564509153366089,
+      "learning_rate": 4.1985810583512473e-07,
+      "loss": 1.4316,
+      "step": 9236
+    },
+    {
+      "epoch": 0.9718043135192004,
+      "grad_norm": 1.6556718349456787,
+      "learning_rate": 4.167444524811215e-07,
+      "loss": 1.5263,
+      "step": 9237
+    },
+    {
+      "epoch": 0.9719095213045765,
+      "grad_norm": 1.864583969116211,
+      "learning_rate": 4.136423634026776e-07,
+      "loss": 1.4365,
+      "step": 9238
+    },
+    {
+      "epoch": 0.9720147290899527,
+      "grad_norm": 1.7581238746643066,
+      "learning_rate": 4.1055183896001606e-07,
+      "loss": 1.5028,
+      "step": 9239
+    },
+    {
+      "epoch": 0.9721199368753288,
+      "grad_norm": 1.2646266222000122,
+      "learning_rate": 4.074728795120275e-07,
+      "loss": 1.6811,
+      "step": 9240
+    },
+    {
+      "epoch": 0.9722251446607049,
+      "grad_norm": 1.6444590091705322,
+      "learning_rate": 4.0440548541625935e-07,
+      "loss": 1.6224,
+      "step": 9241
+    },
+    {
+      "epoch": 0.972330352446081,
+      "grad_norm": 1.8099961280822754,
+      "learning_rate": 4.013496570289155e-07,
+      "loss": 1.5547,
+      "step": 9242
+    },
+    {
+      "epoch": 0.9724355602314572,
+      "grad_norm": 2.0473053455352783,
+      "learning_rate": 3.983053947048676e-07,
+      "loss": 1.4913,
+      "step": 9243
+    },
+    {
+      "epoch": 0.9725407680168332,
+      "grad_norm": 1.557085394859314,
+      "learning_rate": 3.9527269879764406e-07,
+      "loss": 1.415,
+      "step": 9244
+    },
+    {
+      "epoch": 0.9726459758022094,
+      "grad_norm": 1.8768304586410522,
+      "learning_rate": 3.9225156965939647e-07,
+      "loss": 1.5624,
+      "step": 9245
+    },
+    {
+      "epoch": 0.9727511835875855,
+      "grad_norm": 1.8402485847473145,
+      "learning_rate": 3.892420076409886e-07,
+      "loss": 1.712,
+      "step": 9246
+    },
+    {
+      "epoch": 0.9728563913729616,
+      "grad_norm": 1.5501086711883545,
+      "learning_rate": 3.862440130918854e-07,
+      "loss": 1.5368,
+      "step": 9247
+    },
+    {
+      "epoch": 0.9729615991583377,
+      "grad_norm": 2.000622272491455,
+      "learning_rate": 3.832575863602528e-07,
+      "loss": 2.0911,
+      "step": 9248
+    },
+    {
+      "epoch": 0.9730668069437138,
+      "grad_norm": 1.5791133642196655,
+      "learning_rate": 3.8028272779287997e-07,
+      "loss": 1.7847,
+      "step": 9249
+    },
+    {
+      "epoch": 0.97317201472909,
+      "grad_norm": 1.8884774446487427,
+      "learning_rate": 3.7731943773523515e-07,
+      "loss": 1.5126,
+      "step": 9250
+    },
+    {
+      "epoch": 0.973277222514466,
+      "grad_norm": 1.8291395902633667,
+      "learning_rate": 3.7436771653143187e-07,
+      "loss": 1.6039,
+      "step": 9251
+    },
+    {
+      "epoch": 0.9733824302998422,
+      "grad_norm": 1.446852684020996,
+      "learning_rate": 3.7142756452425155e-07,
+      "loss": 1.3811,
+      "step": 9252
+    },
+    {
+      "epoch": 0.9734876380852183,
+      "grad_norm": 1.450920820236206,
+      "learning_rate": 3.68498982055121e-07,
+      "loss": 1.5193,
+      "step": 9253
+    },
+    {
+      "epoch": 0.9735928458705945,
+      "grad_norm": 1.6216645240783691,
+      "learning_rate": 3.655819694641127e-07,
+      "loss": 1.7471,
+      "step": 9254
+    },
+    {
+      "epoch": 0.9736980536559705,
+      "grad_norm": 1.7787151336669922,
+      "learning_rate": 3.62676527089989e-07,
+      "loss": 1.8712,
+      "step": 9255
+    },
+    {
+      "epoch": 0.9738032614413467,
+      "grad_norm": 1.63424813747406,
+      "learning_rate": 3.5978265527014666e-07,
+      "loss": 1.8227,
+      "step": 9256
+    },
+    {
+      "epoch": 0.9739084692267228,
+      "grad_norm": 2.0026700496673584,
+      "learning_rate": 3.56900354340628e-07,
+      "loss": 1.7547,
+      "step": 9257
+    },
+    {
+      "epoch": 0.9740136770120988,
+      "grad_norm": 1.6562840938568115,
+      "learning_rate": 3.5402962463616517e-07,
+      "loss": 2.0525,
+      "step": 9258
+    },
+    {
+      "epoch": 0.974118884797475,
+      "grad_norm": 2.0857789516448975,
+      "learning_rate": 3.511704664901139e-07,
+      "loss": 1.4605,
+      "step": 9259
+    },
+    {
+      "epoch": 0.9742240925828511,
+      "grad_norm": 1.9922007322311401,
+      "learning_rate": 3.483228802344973e-07,
+      "loss": 1.4324,
+      "step": 9260
+    },
+    {
+      "epoch": 0.9743293003682273,
+      "grad_norm": 2.007927417755127,
+      "learning_rate": 3.454868661999955e-07,
+      "loss": 1.6704,
+      "step": 9261
+    },
+    {
+      "epoch": 0.9744345081536033,
+      "grad_norm": 1.3496975898742676,
+      "learning_rate": 3.42662424715956e-07,
+      "loss": 1.6075,
+      "step": 9262
+    },
+    {
+      "epoch": 0.9745397159389795,
+      "grad_norm": 1.6474722623825073,
+      "learning_rate": 3.39849556110361e-07,
+      "loss": 1.8737,
+      "step": 9263
+    },
+    {
+      "epoch": 0.9746449237243556,
+      "grad_norm": 1.7069952487945557,
+      "learning_rate": 3.370482607098602e-07,
+      "loss": 1.2898,
+      "step": 9264
+    },
+    {
+      "epoch": 0.9747501315097317,
+      "grad_norm": 1.7524888515472412,
+      "learning_rate": 3.342585388397712e-07,
+      "loss": 1.5437,
+      "step": 9265
+    },
+    {
+      "epoch": 0.9748553392951078,
+      "grad_norm": 1.8465501070022583,
+      "learning_rate": 3.3148039082404605e-07,
+      "loss": 1.4052,
+      "step": 9266
+    },
+    {
+      "epoch": 0.974960547080484,
+      "grad_norm": 1.5890626907348633,
+      "learning_rate": 3.2871381698529324e-07,
+      "loss": 1.616,
+      "step": 9267
+    },
+    {
+      "epoch": 0.9750657548658601,
+      "grad_norm": 1.5656459331512451,
+      "learning_rate": 3.259588176448003e-07,
+      "loss": 1.3757,
+      "step": 9268
+    },
+    {
+      "epoch": 0.9751709626512362,
+      "grad_norm": 1.868996262550354,
+      "learning_rate": 3.2321539312248903e-07,
+      "loss": 1.3586,
+      "step": 9269
+    },
+    {
+      "epoch": 0.9752761704366123,
+      "grad_norm": 1.7891117334365845,
+      "learning_rate": 3.204835437369491e-07,
+      "loss": 1.2176,
+      "step": 9270
+    },
+    {
+      "epoch": 0.9753813782219884,
+      "grad_norm": 1.7648265361785889,
+      "learning_rate": 3.177632698054156e-07,
+      "loss": 1.7038,
+      "step": 9271
+    },
+    {
+      "epoch": 0.9754865860073646,
+      "grad_norm": 1.6850988864898682,
+      "learning_rate": 3.150545716437914e-07,
+      "loss": 1.4181,
+      "step": 9272
+    },
+    {
+      "epoch": 0.9755917937927406,
+      "grad_norm": 1.7368006706237793,
+      "learning_rate": 3.1235744956662484e-07,
+      "loss": 1.3408,
+      "step": 9273
+    },
+    {
+      "epoch": 0.9756970015781168,
+      "grad_norm": 2.081385374069214,
+      "learning_rate": 3.0967190388712097e-07,
+      "loss": 1.3886,
+      "step": 9274
+    },
+    {
+      "epoch": 0.9758022093634929,
+      "grad_norm": 2.454695224761963,
+      "learning_rate": 3.0699793491715256e-07,
+      "loss": 1.1868,
+      "step": 9275
+    },
+    {
+      "epoch": 0.975907417148869,
+      "grad_norm": 1.7129137516021729,
+      "learning_rate": 3.043355429672268e-07,
+      "loss": 1.7567,
+      "step": 9276
+    },
+    {
+      "epoch": 0.9760126249342451,
+      "grad_norm": 1.3494802713394165,
+      "learning_rate": 3.016847283465185e-07,
+      "loss": 1.4914,
+      "step": 9277
+    },
+    {
+      "epoch": 0.9761178327196213,
+      "grad_norm": 1.3890875577926636,
+      "learning_rate": 2.990454913628704e-07,
+      "loss": 1.3622,
+      "step": 9278
+    },
+    {
+      "epoch": 0.9762230405049974,
+      "grad_norm": 2.182384967803955,
+      "learning_rate": 2.9641783232275955e-07,
+      "loss": 1.4551,
+      "step": 9279
+    },
+    {
+      "epoch": 0.9763282482903735,
+      "grad_norm": 1.537185788154602,
+      "learning_rate": 2.938017515313418e-07,
+      "loss": 1.7531,
+      "step": 9280
+    },
+    {
+      "epoch": 0.9764334560757496,
+      "grad_norm": 1.466887354850769,
+      "learning_rate": 2.9119724929239645e-07,
+      "loss": 1.567,
+      "step": 9281
+    },
+    {
+      "epoch": 0.9765386638611258,
+      "grad_norm": 1.78278648853302,
+      "learning_rate": 2.886043259083704e-07,
+      "loss": 1.3838,
+      "step": 9282
+    },
+    {
+      "epoch": 0.9766438716465018,
+      "grad_norm": 1.6056292057037354,
+      "learning_rate": 2.860229816803894e-07,
+      "loss": 1.1808,
+      "step": 9283
+    },
+    {
+      "epoch": 0.9767490794318779,
+      "grad_norm": 1.9775320291519165,
+      "learning_rate": 2.834532169082138e-07,
+      "loss": 1.6403,
+      "step": 9284
+    },
+    {
+      "epoch": 0.9768542872172541,
+      "grad_norm": 1.4951421022415161,
+      "learning_rate": 2.8089503189024926e-07,
+      "loss": 1.6209,
+      "step": 9285
+    },
+    {
+      "epoch": 0.9769594950026302,
+      "grad_norm": 2.1086246967315674,
+      "learning_rate": 2.7834842692358033e-07,
+      "loss": 1.6395,
+      "step": 9286
+    },
+    {
+      "epoch": 0.9770647027880063,
+      "grad_norm": 1.4690982103347778,
+      "learning_rate": 2.7581340230393717e-07,
+      "loss": 1.6837,
+      "step": 9287
+    },
+    {
+      "epoch": 0.9771699105733824,
+      "grad_norm": 1.447407603263855,
+      "learning_rate": 2.7328995832568426e-07,
+      "loss": 1.6867,
+      "step": 9288
+    },
+    {
+      "epoch": 0.9772751183587586,
+      "grad_norm": 1.7926383018493652,
+      "learning_rate": 2.7077809528188724e-07,
+      "loss": 1.4406,
+      "step": 9289
+    },
+    {
+      "epoch": 0.9773803261441346,
+      "grad_norm": 2.195378541946411,
+      "learning_rate": 2.6827781346423496e-07,
+      "loss": 1.808,
+      "step": 9290
+    },
+    {
+      "epoch": 0.9774855339295108,
+      "grad_norm": 1.9407440423965454,
+      "learning_rate": 2.657891131630619e-07,
+      "loss": 1.362,
+      "step": 9291
+    },
+    {
+      "epoch": 0.9775907417148869,
+      "grad_norm": 1.7981421947479248,
+      "learning_rate": 2.633119946673923e-07,
+      "loss": 1.4832,
+      "step": 9292
+    },
+    {
+      "epoch": 0.9776959495002631,
+      "grad_norm": 1.5545809268951416,
+      "learning_rate": 2.608464582648629e-07,
+      "loss": 1.8864,
+      "step": 9293
+    },
+    {
+      "epoch": 0.9778011572856391,
+      "grad_norm": 1.306902289390564,
+      "learning_rate": 2.583925042418112e-07,
+      "loss": 1.5875,
+      "step": 9294
+    },
+    {
+      "epoch": 0.9779063650710152,
+      "grad_norm": 2.3127174377441406,
+      "learning_rate": 2.5595013288318703e-07,
+      "loss": 1.6263,
+      "step": 9295
+    },
+    {
+      "epoch": 0.9780115728563914,
+      "grad_norm": 2.087738275527954,
+      "learning_rate": 2.5351934447263026e-07,
+      "loss": 1.7355,
+      "step": 9296
+    },
+    {
+      "epoch": 0.9781167806417674,
+      "grad_norm": 1.412264108657837,
+      "learning_rate": 2.5110013929241504e-07,
+      "loss": 1.5116,
+      "step": 9297
+    },
+    {
+      "epoch": 0.9782219884271436,
+      "grad_norm": 1.165013313293457,
+      "learning_rate": 2.4869251762348333e-07,
+      "loss": 1.421,
+      "step": 9298
+    },
+    {
+      "epoch": 0.9783271962125197,
+      "grad_norm": 2.194392442703247,
+      "learning_rate": 2.462964797454004e-07,
+      "loss": 1.9825,
+      "step": 9299
+    },
+    {
+      "epoch": 0.9784324039978959,
+      "grad_norm": 1.6436448097229004,
+      "learning_rate": 2.4391202593643246e-07,
+      "loss": 1.6435,
+      "step": 9300
+    },
+    {
+      "epoch": 0.9785376117832719,
+      "grad_norm": 0.9663065075874329,
+      "learning_rate": 2.4153915647348034e-07,
+      "loss": 1.5033,
+      "step": 9301
+    },
+    {
+      "epoch": 0.9786428195686481,
+      "grad_norm": 1.650208592414856,
+      "learning_rate": 2.391778716320792e-07,
+      "loss": 1.4037,
+      "step": 9302
+    },
+    {
+      "epoch": 0.9787480273540242,
+      "grad_norm": 1.5926425457000732,
+      "learning_rate": 2.3682817168644288e-07,
+      "loss": 1.9914,
+      "step": 9303
+    },
+    {
+      "epoch": 0.9788532351394004,
+      "grad_norm": 1.1928669214248657,
+      "learning_rate": 2.3449005690945324e-07,
+      "loss": 1.5487,
+      "step": 9304
+    },
+    {
+      "epoch": 0.9789584429247764,
+      "grad_norm": 1.331669569015503,
+      "learning_rate": 2.321635275726042e-07,
+      "loss": 1.9395,
+      "step": 9305
+    },
+    {
+      "epoch": 0.9790636507101526,
+      "grad_norm": 1.4218162298202515,
+      "learning_rate": 2.2984858394607956e-07,
+      "loss": 1.4188,
+      "step": 9306
+    },
+    {
+      "epoch": 0.9791688584955287,
+      "grad_norm": 1.6478614807128906,
+      "learning_rate": 2.275452262986977e-07,
+      "loss": 1.7231,
+      "step": 9307
+    },
+    {
+      "epoch": 0.9792740662809047,
+      "grad_norm": 1.5847270488739014,
+      "learning_rate": 2.252534548979446e-07,
+      "loss": 1.3288,
+      "step": 9308
+    },
+    {
+      "epoch": 0.9793792740662809,
+      "grad_norm": 1.5312392711639404,
+      "learning_rate": 2.2297327000996293e-07,
+      "loss": 1.4311,
+      "step": 9309
+    },
+    {
+      "epoch": 0.979484481851657,
+      "grad_norm": 1.6331710815429688,
+      "learning_rate": 2.207046718995409e-07,
+      "loss": 1.8195,
+      "step": 9310
+    },
+    {
+      "epoch": 0.9795896896370332,
+      "grad_norm": 1.5755873918533325,
+      "learning_rate": 2.1844766083011226e-07,
+      "loss": 1.5262,
+      "step": 9311
+    },
+    {
+      "epoch": 0.9796948974224092,
+      "grad_norm": 1.929260492324829,
+      "learning_rate": 2.162022370637895e-07,
+      "loss": 1.5236,
+      "step": 9312
+    },
+    {
+      "epoch": 0.9798001052077854,
+      "grad_norm": 1.6517144441604614,
+      "learning_rate": 2.1396840086131964e-07,
+      "loss": 1.9677,
+      "step": 9313
+    },
+    {
+      "epoch": 0.9799053129931615,
+      "grad_norm": 1.5350128412246704,
+      "learning_rate": 2.1174615248210626e-07,
+      "loss": 1.2628,
+      "step": 9314
+    },
+    {
+      "epoch": 0.9800105207785376,
+      "grad_norm": 1.6820600032806396,
+      "learning_rate": 2.0953549218423185e-07,
+      "loss": 2.0707,
+      "step": 9315
+    },
+    {
+      "epoch": 0.9801157285639137,
+      "grad_norm": 2.348896026611328,
+      "learning_rate": 2.0733642022437994e-07,
+      "loss": 1.3693,
+      "step": 9316
+    },
+    {
+      "epoch": 0.9802209363492899,
+      "grad_norm": 3.21044659614563,
+      "learning_rate": 2.0514893685795733e-07,
+      "loss": 1.4473,
+      "step": 9317
+    },
+    {
+      "epoch": 0.980326144134666,
+      "grad_norm": 1.5892785787582397,
+      "learning_rate": 2.0297304233896087e-07,
+      "loss": 1.5056,
+      "step": 9318
+    },
+    {
+      "epoch": 0.980431351920042,
+      "grad_norm": 1.6441622972488403,
+      "learning_rate": 2.008087369200773e-07,
+      "loss": 1.6748,
+      "step": 9319
+    },
+    {
+      "epoch": 0.9805365597054182,
+      "grad_norm": 1.501565933227539,
+      "learning_rate": 1.9865602085265002e-07,
+      "loss": 1.3629,
+      "step": 9320
+    },
+    {
+      "epoch": 0.9806417674907943,
+      "grad_norm": 1.4179154634475708,
+      "learning_rate": 1.9651489438666792e-07,
+      "loss": 1.7201,
+      "step": 9321
+    },
+    {
+      "epoch": 0.9807469752761704,
+      "grad_norm": 1.6609681844711304,
+      "learning_rate": 1.943853577707544e-07,
+      "loss": 1.7947,
+      "step": 9322
+    },
+    {
+      "epoch": 0.9808521830615465,
+      "grad_norm": 1.9472432136535645,
+      "learning_rate": 1.922674112522227e-07,
+      "loss": 1.6767,
+      "step": 9323
+    },
+    {
+      "epoch": 0.9809573908469227,
+      "grad_norm": 1.3472323417663574,
+      "learning_rate": 1.9016105507702054e-07,
+      "loss": 1.2528,
+      "step": 9324
+    },
+    {
+      "epoch": 0.9810625986322988,
+      "grad_norm": 1.9179092645645142,
+      "learning_rate": 1.8806628948974114e-07,
+      "loss": 1.5153,
+      "step": 9325
+    },
+    {
+      "epoch": 0.9811678064176749,
+      "grad_norm": 2.2002909183502197,
+      "learning_rate": 1.859831147336566e-07,
+      "loss": 1.584,
+      "step": 9326
+    },
+    {
+      "epoch": 0.981273014203051,
+      "grad_norm": 2.532457113265991,
+      "learning_rate": 1.8391153105067338e-07,
+      "loss": 2.4041,
+      "step": 9327
+    },
+    {
+      "epoch": 0.9813782219884272,
+      "grad_norm": 1.7120616436004639,
+      "learning_rate": 1.8185153868135462e-07,
+      "loss": 1.6406,
+      "step": 9328
+    },
+    {
+      "epoch": 0.9814834297738032,
+      "grad_norm": 1.6035572290420532,
+      "learning_rate": 1.798031378649201e-07,
+      "loss": 1.5754,
+      "step": 9329
+    },
+    {
+      "epoch": 0.9815886375591794,
+      "grad_norm": 1.2862604856491089,
+      "learning_rate": 1.7776632883924615e-07,
+      "loss": 1.7776,
+      "step": 9330
+    },
+    {
+      "epoch": 0.9816938453445555,
+      "grad_norm": 2.551787853240967,
+      "learning_rate": 1.7574111184086582e-07,
+      "loss": 1.962,
+      "step": 9331
+    },
+    {
+      "epoch": 0.9817990531299317,
+      "grad_norm": 1.6095662117004395,
+      "learning_rate": 1.7372748710495768e-07,
+      "loss": 1.6107,
+      "step": 9332
+    },
+    {
+      "epoch": 0.9819042609153077,
+      "grad_norm": 2.01579213142395,
+      "learning_rate": 1.7172545486535685e-07,
+      "loss": 1.1388,
+      "step": 9333
+    },
+    {
+      "epoch": 0.9820094687006838,
+      "grad_norm": 2.388930559158325,
+      "learning_rate": 1.6973501535455516e-07,
+      "loss": 1.7735,
+      "step": 9334
+    },
+    {
+      "epoch": 0.98211467648606,
+      "grad_norm": 1.5731823444366455,
+      "learning_rate": 1.6775616880368994e-07,
+      "loss": 1.653,
+      "step": 9335
+    },
+    {
+      "epoch": 0.9822198842714361,
+      "grad_norm": 1.1225214004516602,
+      "learning_rate": 1.6578891544255514e-07,
+      "loss": 1.8642,
+      "step": 9336
+    },
+    {
+      "epoch": 0.9823250920568122,
+      "grad_norm": 1.4525314569473267,
+      "learning_rate": 1.638332554996125e-07,
+      "loss": 1.2603,
+      "step": 9337
+    },
+    {
+      "epoch": 0.9824302998421883,
+      "grad_norm": 2.0482780933380127,
+      "learning_rate": 1.6188918920195806e-07,
+      "loss": 1.4663,
+      "step": 9338
+    },
+    {
+      "epoch": 0.9825355076275645,
+      "grad_norm": 1.067059874534607,
+      "learning_rate": 1.5995671677535573e-07,
+      "loss": 1.371,
+      "step": 9339
+    },
+    {
+      "epoch": 0.9826407154129405,
+      "grad_norm": 2.4665608406066895,
+      "learning_rate": 1.5803583844421488e-07,
+      "loss": 1.533,
+      "step": 9340
+    },
+    {
+      "epoch": 0.9827459231983167,
+      "grad_norm": 1.2421523332595825,
+      "learning_rate": 1.561265544316015e-07,
+      "loss": 1.3759,
+      "step": 9341
+    },
+    {
+      "epoch": 0.9828511309836928,
+      "grad_norm": 1.3852927684783936,
+      "learning_rate": 1.5422886495922718e-07,
+      "loss": 1.5905,
+      "step": 9342
+    },
+    {
+      "epoch": 0.982956338769069,
+      "grad_norm": 1.9068608283996582,
+      "learning_rate": 1.5234277024747112e-07,
+      "loss": 1.7971,
+      "step": 9343
+    },
+    {
+      "epoch": 0.983061546554445,
+      "grad_norm": 1.5108956098556519,
+      "learning_rate": 1.5046827051536928e-07,
+      "loss": 1.8727,
+      "step": 9344
+    },
+    {
+      "epoch": 0.9831667543398211,
+      "grad_norm": 1.5862867832183838,
+      "learning_rate": 1.4860536598058085e-07,
+      "loss": 1.9341,
+      "step": 9345
+    },
+    {
+      "epoch": 0.9832719621251973,
+      "grad_norm": 1.2081376314163208,
+      "learning_rate": 1.4675405685944387e-07,
+      "loss": 1.7027,
+      "step": 9346
+    },
+    {
+      "epoch": 0.9833771699105733,
+      "grad_norm": 1.2375844717025757,
+      "learning_rate": 1.4491434336696418e-07,
+      "loss": 1.5949,
+      "step": 9347
+    },
+    {
+      "epoch": 0.9834823776959495,
+      "grad_norm": 1.2393487691879272,
+      "learning_rate": 1.430862257167598e-07,
+      "loss": 1.7839,
+      "step": 9348
+    },
+    {
+      "epoch": 0.9835875854813256,
+      "grad_norm": 2.2204091548919678,
+      "learning_rate": 1.412697041211275e-07,
+      "loss": 1.5249,
+      "step": 9349
+    },
+    {
+      "epoch": 0.9836927932667018,
+      "grad_norm": 1.9110403060913086,
+      "learning_rate": 1.3946477879102083e-07,
+      "loss": 1.1106,
+      "step": 9350
+    },
+    {
+      "epoch": 0.9837980010520778,
+      "grad_norm": 1.352942705154419,
+      "learning_rate": 1.3767144993602766e-07,
+      "loss": 1.3338,
+      "step": 9351
+    },
+    {
+      "epoch": 0.983903208837454,
+      "grad_norm": 1.511385440826416,
+      "learning_rate": 1.3588971776441472e-07,
+      "loss": 1.8067,
+      "step": 9352
+    },
+    {
+      "epoch": 0.9840084166228301,
+      "grad_norm": 1.3924455642700195,
+      "learning_rate": 1.3411958248309431e-07,
+      "loss": 1.8532,
+      "step": 9353
+    },
+    {
+      "epoch": 0.9841136244082062,
+      "grad_norm": 1.5775933265686035,
+      "learning_rate": 1.3236104429760199e-07,
+      "loss": 1.602,
+      "step": 9354
+    },
+    {
+      "epoch": 0.9842188321935823,
+      "grad_norm": 1.5030940771102905,
+      "learning_rate": 1.306141034121744e-07,
+      "loss": 1.8972,
+      "step": 9355
+    },
+    {
+      "epoch": 0.9843240399789585,
+      "grad_norm": 2.3042783737182617,
+      "learning_rate": 1.2887876002967149e-07,
+      "loss": 1.3577,
+      "step": 9356
+    },
+    {
+      "epoch": 0.9844292477643346,
+      "grad_norm": 2.0904388427734375,
+      "learning_rate": 1.2715501435159872e-07,
+      "loss": 1.3343,
+      "step": 9357
+    },
+    {
+      "epoch": 0.9845344555497106,
+      "grad_norm": 1.2769339084625244,
+      "learning_rate": 1.254428665781515e-07,
+      "loss": 1.8264,
+      "step": 9358
+    },
+    {
+      "epoch": 0.9846396633350868,
+      "grad_norm": 1.4355113506317139,
+      "learning_rate": 1.2374231690813754e-07,
+      "loss": 1.8146,
+      "step": 9359
+    },
+    {
+      "epoch": 0.9847448711204629,
+      "grad_norm": 1.9717060327529907,
+      "learning_rate": 1.2205336553904323e-07,
+      "loss": 1.3479,
+      "step": 9360
+    },
+    {
+      "epoch": 0.984850078905839,
+      "grad_norm": 1.9314119815826416,
+      "learning_rate": 1.203760126670117e-07,
+      "loss": 1.4207,
+      "step": 9361
+    },
+    {
+      "epoch": 0.9849552866912151,
+      "grad_norm": 1.9035725593566895,
+      "learning_rate": 1.1871025848680939e-07,
+      "loss": 1.2061,
+      "step": 9362
+    },
+    {
+      "epoch": 0.9850604944765913,
+      "grad_norm": 1.5623804330825806,
+      "learning_rate": 1.1705610319188154e-07,
+      "loss": 1.5218,
+      "step": 9363
+    },
+    {
+      "epoch": 0.9851657022619674,
+      "grad_norm": 1.5503050088882446,
+      "learning_rate": 1.1541354697431894e-07,
+      "loss": 1.8319,
+      "step": 9364
+    },
+    {
+      "epoch": 0.9852709100473435,
+      "grad_norm": 1.8744232654571533,
+      "learning_rate": 1.1378259002488013e-07,
+      "loss": 1.2055,
+      "step": 9365
+    },
+    {
+      "epoch": 0.9853761178327196,
+      "grad_norm": 1.7430399656295776,
+      "learning_rate": 1.1216323253294691e-07,
+      "loss": 1.44,
+      "step": 9366
+    },
+    {
+      "epoch": 0.9854813256180958,
+      "grad_norm": 1.298393964767456,
+      "learning_rate": 1.1055547468658001e-07,
+      "loss": 1.7648,
+      "step": 9367
+    },
+    {
+      "epoch": 0.9855865334034719,
+      "grad_norm": 1.0370997190475464,
+      "learning_rate": 1.089593166724634e-07,
+      "loss": 1.4384,
+      "step": 9368
+    },
+    {
+      "epoch": 0.985691741188848,
+      "grad_norm": 1.9467689990997314,
+      "learning_rate": 1.0737475867598212e-07,
+      "loss": 1.8683,
+      "step": 9369
+    },
+    {
+      "epoch": 0.9857969489742241,
+      "grad_norm": 1.4504022598266602,
+      "learning_rate": 1.0580180088112234e-07,
+      "loss": 1.2465,
+      "step": 9370
+    },
+    {
+      "epoch": 0.9859021567596002,
+      "grad_norm": 1.4652175903320312,
+      "learning_rate": 1.0424044347056017e-07,
+      "loss": 1.6443,
+      "step": 9371
+    },
+    {
+      "epoch": 0.9860073645449763,
+      "grad_norm": 1.7554495334625244,
+      "learning_rate": 1.0269068662560611e-07,
+      "loss": 1.0831,
+      "step": 9372
+    },
+    {
+      "epoch": 0.9861125723303524,
+      "grad_norm": 1.1378939151763916,
+      "learning_rate": 1.0115253052622731e-07,
+      "loss": 1.7837,
+      "step": 9373
+    },
+    {
+      "epoch": 0.9862177801157286,
+      "grad_norm": 1.8769643306732178,
+      "learning_rate": 9.962597535104756e-08,
+      "loss": 1.515,
+      "step": 9374
+    },
+    {
+      "epoch": 0.9863229879011047,
+      "grad_norm": 1.4734489917755127,
+      "learning_rate": 9.811102127733618e-08,
+      "loss": 2.1066,
+      "step": 9375
+    },
+    {
+      "epoch": 0.9864281956864808,
+      "grad_norm": 1.2402914762496948,
+      "learning_rate": 9.660766848101909e-08,
+      "loss": 1.6076,
+      "step": 9376
+    },
+    {
+      "epoch": 0.9865334034718569,
+      "grad_norm": 1.601291537284851,
+      "learning_rate": 9.511591713668999e-08,
+      "loss": 1.5006,
+      "step": 9377
+    },
+    {
+      "epoch": 0.9866386112572331,
+      "grad_norm": 1.4172643423080444,
+      "learning_rate": 9.363576741755476e-08,
+      "loss": 1.2291,
+      "step": 9378
+    },
+    {
+      "epoch": 0.9867438190426091,
+      "grad_norm": 1.5718390941619873,
+      "learning_rate": 9.216721949553142e-08,
+      "loss": 1.8789,
+      "step": 9379
+    },
+    {
+      "epoch": 0.9868490268279853,
+      "grad_norm": 1.7157665491104126,
+      "learning_rate": 9.071027354112804e-08,
+      "loss": 1.6226,
+      "step": 9380
+    },
+    {
+      "epoch": 0.9869542346133614,
+      "grad_norm": 1.4585630893707275,
+      "learning_rate": 8.926492972355371e-08,
+      "loss": 1.5842,
+      "step": 9381
+    },
+    {
+      "epoch": 0.9870594423987376,
+      "grad_norm": 1.299195647239685,
+      "learning_rate": 8.783118821064085e-08,
+      "loss": 1.4625,
+      "step": 9382
+    },
+    {
+      "epoch": 0.9871646501841136,
+      "grad_norm": 1.305501937866211,
+      "learning_rate": 8.640904916888959e-08,
+      "loss": 1.4309,
+      "step": 9383
+    },
+    {
+      "epoch": 0.9872698579694897,
+      "grad_norm": 1.4798643589019775,
+      "learning_rate": 8.499851276344561e-08,
+      "loss": 1.4667,
+      "step": 9384
+    },
+    {
+      "epoch": 0.9873750657548659,
+      "grad_norm": 1.9000688791275024,
+      "learning_rate": 8.359957915812233e-08,
+      "loss": 1.4393,
+      "step": 9385
+    },
+    {
+      "epoch": 0.9874802735402419,
+      "grad_norm": 1.4523059129714966,
+      "learning_rate": 8.221224851535647e-08,
+      "loss": 1.5859,
+      "step": 9386
+    },
+    {
+      "epoch": 0.9875854813256181,
+      "grad_norm": 1.9257595539093018,
+      "learning_rate": 8.08365209962525e-08,
+      "loss": 1.7144,
+      "step": 9387
+    },
+    {
+      "epoch": 0.9876906891109942,
+      "grad_norm": 1.2338892221450806,
+      "learning_rate": 7.94723967605937e-08,
+      "loss": 1.5707,
+      "step": 9388
+    },
+    {
+      "epoch": 0.9877958968963704,
+      "grad_norm": 1.9463906288146973,
+      "learning_rate": 7.81198759667645e-08,
+      "loss": 1.7662,
+      "step": 9389
+    },
+    {
+      "epoch": 0.9879011046817464,
+      "grad_norm": 1.2963135242462158,
+      "learning_rate": 7.677895877183927e-08,
+      "loss": 1.6959,
+      "step": 9390
+    },
+    {
+      "epoch": 0.9880063124671226,
+      "grad_norm": 1.3970755338668823,
+      "learning_rate": 7.544964533153787e-08,
+      "loss": 1.334,
+      "step": 9391
+    },
+    {
+      "epoch": 0.9881115202524987,
+      "grad_norm": 1.601676344871521,
+      "learning_rate": 7.41319358002146e-08,
+      "loss": 1.6125,
+      "step": 9392
+    },
+    {
+      "epoch": 0.9882167280378747,
+      "grad_norm": 1.3263901472091675,
+      "learning_rate": 7.282583033091372e-08,
+      "loss": 1.9106,
+      "step": 9393
+    },
+    {
+      "epoch": 0.9883219358232509,
+      "grad_norm": 1.8151953220367432,
+      "learning_rate": 7.15313290752917e-08,
+      "loss": 1.8621,
+      "step": 9394
+    },
+    {
+      "epoch": 0.988427143608627,
+      "grad_norm": 1.4019142389297485,
+      "learning_rate": 7.024843218368382e-08,
+      "loss": 1.565,
+      "step": 9395
+    },
+    {
+      "epoch": 0.9885323513940032,
+      "grad_norm": 2.185393810272217,
+      "learning_rate": 6.897713980505982e-08,
+      "loss": 1.4799,
+      "step": 9396
+    },
+    {
+      "epoch": 0.9886375591793792,
+      "grad_norm": 1.357749104499817,
+      "learning_rate": 6.771745208705715e-08,
+      "loss": 1.5501,
+      "step": 9397
+    },
+    {
+      "epoch": 0.9887427669647554,
+      "grad_norm": 1.7001276016235352,
+      "learning_rate": 6.646936917595881e-08,
+      "loss": 1.5436,
+      "step": 9398
+    },
+    {
+      "epoch": 0.9888479747501315,
+      "grad_norm": 2.9914774894714355,
+      "learning_rate": 6.52328912167044e-08,
+      "loss": 1.9068,
+      "step": 9399
+    },
+    {
+      "epoch": 0.9889531825355077,
+      "grad_norm": 1.8778984546661377,
+      "learning_rate": 6.400801835286796e-08,
+      "loss": 1.6909,
+      "step": 9400
+    },
+    {
+      "epoch": 0.9890583903208837,
+      "grad_norm": 1.316900372505188,
+      "learning_rate": 6.279475072670238e-08,
+      "loss": 1.4041,
+      "step": 9401
+    },
+    {
+      "epoch": 0.9891635981062599,
+      "grad_norm": 2.04650616645813,
+      "learning_rate": 6.159308847909495e-08,
+      "loss": 1.7704,
+      "step": 9402
+    },
+    {
+      "epoch": 0.989268805891636,
+      "grad_norm": 2.6405978202819824,
+      "learning_rate": 6.040303174958961e-08,
+      "loss": 1.8183,
+      "step": 9403
+    },
+    {
+      "epoch": 0.989374013677012,
+      "grad_norm": 1.2661974430084229,
+      "learning_rate": 5.922458067639802e-08,
+      "loss": 2.0856,
+      "step": 9404
+    },
+    {
+      "epoch": 0.9894792214623882,
+      "grad_norm": 1.692745566368103,
+      "learning_rate": 5.805773539634407e-08,
+      "loss": 1.5816,
+      "step": 9405
+    },
+    {
+      "epoch": 0.9895844292477644,
+      "grad_norm": 1.9099050760269165,
+      "learning_rate": 5.690249604495268e-08,
+      "loss": 1.4334,
+      "step": 9406
+    },
+    {
+      "epoch": 0.9896896370331405,
+      "grad_norm": 1.8465721607208252,
+      "learning_rate": 5.57588627563721e-08,
+      "loss": 1.7839,
+      "step": 9407
+    },
+    {
+      "epoch": 0.9897948448185165,
+      "grad_norm": 1.444919228553772,
+      "learning_rate": 5.46268356634072e-08,
+      "loss": 1.5675,
+      "step": 9408
+    },
+    {
+      "epoch": 0.9899000526038927,
+      "grad_norm": 1.517018437385559,
+      "learning_rate": 5.3506414897508404e-08,
+      "loss": 1.4653,
+      "step": 9409
+    },
+    {
+      "epoch": 0.9900052603892688,
+      "grad_norm": 1.6585253477096558,
+      "learning_rate": 5.239760058879384e-08,
+      "loss": 2.2469,
+      "step": 9410
+    },
+    {
+      "epoch": 0.9901104681746449,
+      "grad_norm": 1.9332027435302734,
+      "learning_rate": 5.130039286602717e-08,
+      "loss": 1.6433,
+      "step": 9411
+    },
+    {
+      "epoch": 0.990215675960021,
+      "grad_norm": 1.3716708421707153,
+      "learning_rate": 5.02147918566287e-08,
+      "loss": 1.4928,
+      "step": 9412
+    },
+    {
+      "epoch": 0.9903208837453972,
+      "grad_norm": 2.6388068199157715,
+      "learning_rate": 4.9140797686653136e-08,
+      "loss": 1.4978,
+      "step": 9413
+    },
+    {
+      "epoch": 0.9904260915307733,
+      "grad_norm": 1.895566463470459,
+      "learning_rate": 4.807841048082296e-08,
+      "loss": 1.6034,
+      "step": 9414
+    },
+    {
+      "epoch": 0.9905312993161494,
+      "grad_norm": 1.3241360187530518,
+      "learning_rate": 4.702763036252833e-08,
+      "loss": 1.7537,
+      "step": 9415
+    },
+    {
+      "epoch": 0.9906365071015255,
+      "grad_norm": 2.1230876445770264,
+      "learning_rate": 4.598845745376057e-08,
+      "loss": 1.9085,
+      "step": 9416
+    },
+    {
+      "epoch": 0.9907417148869017,
+      "grad_norm": 2.1953372955322266,
+      "learning_rate": 4.496089187522312e-08,
+      "loss": 1.5051,
+      "step": 9417
+    },
+    {
+      "epoch": 0.9908469226722777,
+      "grad_norm": 1.0922819375991821,
+      "learning_rate": 4.3944933746231655e-08,
+      "loss": 1.938,
+      "step": 9418
+    },
+    {
+      "epoch": 0.9909521304576538,
+      "grad_norm": 1.2908408641815186,
+      "learning_rate": 4.294058318475846e-08,
+      "loss": 1.4622,
+      "step": 9419
+    },
+    {
+      "epoch": 0.99105733824303,
+      "grad_norm": 1.8100134134292603,
+      "learning_rate": 4.194784030745469e-08,
+      "loss": 1.8774,
+      "step": 9420
+    },
+    {
+      "epoch": 0.9911625460284061,
+      "grad_norm": 1.8238705396652222,
+      "learning_rate": 4.096670522959478e-08,
+      "loss": 1.5957,
+      "step": 9421
+    },
+    {
+      "epoch": 0.9912677538137822,
+      "grad_norm": 1.4968173503875732,
+      "learning_rate": 3.999717806510983e-08,
+      "loss": 1.6067,
+      "step": 9422
+    },
+    {
+      "epoch": 0.9913729615991583,
+      "grad_norm": 1.1250231266021729,
+      "learning_rate": 3.903925892658755e-08,
+      "loss": 1.5794,
+      "step": 9423
+    },
+    {
+      "epoch": 0.9914781693845345,
+      "grad_norm": 1.5623326301574707,
+      "learning_rate": 3.809294792527229e-08,
+      "loss": 1.7024,
+      "step": 9424
+    },
+    {
+      "epoch": 0.9915833771699105,
+      "grad_norm": 2.0509960651397705,
+      "learning_rate": 3.715824517106503e-08,
+      "loss": 1.489,
+      "step": 9425
+    },
+    {
+      "epoch": 0.9916885849552867,
+      "grad_norm": 1.0217515230178833,
+      "learning_rate": 3.623515077250117e-08,
+      "loss": 1.5209,
+      "step": 9426
+    },
+    {
+      "epoch": 0.9917937927406628,
+      "grad_norm": 1.395032525062561,
+      "learning_rate": 3.532366483677274e-08,
+      "loss": 1.8385,
+      "step": 9427
+    },
+    {
+      "epoch": 0.991899000526039,
+      "grad_norm": 1.5118807554244995,
+      "learning_rate": 3.442378746972841e-08,
+      "loss": 1.3888,
+      "step": 9428
+    },
+    {
+      "epoch": 0.992004208311415,
+      "grad_norm": 1.96005117893219,
+      "learning_rate": 3.3535518775873466e-08,
+      "loss": 1.5879,
+      "step": 9429
+    },
+    {
+      "epoch": 0.9921094160967912,
+      "grad_norm": 1.319014072418213,
+      "learning_rate": 3.265885885835873e-08,
+      "loss": 1.8016,
+      "step": 9430
+    },
+    {
+      "epoch": 0.9922146238821673,
+      "grad_norm": 1.2182525396347046,
+      "learning_rate": 3.179380781898056e-08,
+      "loss": 1.9213,
+      "step": 9431
+    },
+    {
+      "epoch": 0.9923198316675435,
+      "grad_norm": 1.6044673919677734,
+      "learning_rate": 3.0940365758203025e-08,
+      "loss": 1.421,
+      "step": 9432
+    },
+    {
+      "epoch": 0.9924250394529195,
+      "grad_norm": 1.7028839588165283,
+      "learning_rate": 3.009853277512464e-08,
+      "loss": 1.4184,
+      "step": 9433
+    },
+    {
+      "epoch": 0.9925302472382956,
+      "grad_norm": 1.4499717950820923,
+      "learning_rate": 2.9268308967522752e-08,
+      "loss": 1.8255,
+      "step": 9434
+    },
+    {
+      "epoch": 0.9926354550236718,
+      "grad_norm": 2.1826236248016357,
+      "learning_rate": 2.844969443178691e-08,
+      "loss": 2.1099,
+      "step": 9435
+    },
+    {
+      "epoch": 0.9927406628090478,
+      "grad_norm": 1.6331896781921387,
+      "learning_rate": 2.7642689262996625e-08,
+      "loss": 2.0769,
+      "step": 9436
+    },
+    {
+      "epoch": 0.992845870594424,
+      "grad_norm": 1.0556201934814453,
+      "learning_rate": 2.6847293554854712e-08,
+      "loss": 1.2877,
+      "step": 9437
+    },
+    {
+      "epoch": 0.9929510783798001,
+      "grad_norm": 1.582120656967163,
+      "learning_rate": 2.6063507399731735e-08,
+      "loss": 1.6723,
+      "step": 9438
+    },
+    {
+      "epoch": 0.9930562861651763,
+      "grad_norm": 1.865082859992981,
+      "learning_rate": 2.5291330888643772e-08,
+      "loss": 1.612,
+      "step": 9439
+    },
+    {
+      "epoch": 0.9931614939505523,
+      "grad_norm": 1.324381709098816,
+      "learning_rate": 2.453076411127464e-08,
+      "loss": 2.0149,
+      "step": 9440
+    },
+    {
+      "epoch": 0.9932667017359285,
+      "grad_norm": 1.3037350177764893,
+      "learning_rate": 2.378180715593148e-08,
+      "loss": 1.6584,
+      "step": 9441
+    },
+    {
+      "epoch": 0.9933719095213046,
+      "grad_norm": 2.0570385456085205,
+      "learning_rate": 2.304446010958916e-08,
+      "loss": 1.0094,
+      "step": 9442
+    },
+    {
+      "epoch": 0.9934771173066806,
+      "grad_norm": 2.5256316661834717,
+      "learning_rate": 2.2318723057879188e-08,
+      "loss": 1.3119,
+      "step": 9443
+    },
+    {
+      "epoch": 0.9935823250920568,
+      "grad_norm": 1.4097765684127808,
+      "learning_rate": 2.1604596085078587e-08,
+      "loss": 1.5828,
+      "step": 9444
+    },
+    {
+      "epoch": 0.9936875328774329,
+      "grad_norm": 1.6299117803573608,
+      "learning_rate": 2.0902079274121024e-08,
+      "loss": 1.523,
+      "step": 9445
+    },
+    {
+      "epoch": 0.9937927406628091,
+      "grad_norm": 1.7055869102478027,
+      "learning_rate": 2.0211172706574576e-08,
+      "loss": 1.7484,
+      "step": 9446
+    },
+    {
+      "epoch": 0.9938979484481851,
+      "grad_norm": 1.4784151315689087,
+      "learning_rate": 1.953187646268617e-08,
+      "loss": 1.4377,
+      "step": 9447
+    },
+    {
+      "epoch": 0.9940031562335613,
+      "grad_norm": 1.5295408964157104,
+      "learning_rate": 1.886419062132605e-08,
+      "loss": 1.6043,
+      "step": 9448
+    },
+    {
+      "epoch": 0.9941083640189374,
+      "grad_norm": 1.7209928035736084,
+      "learning_rate": 1.8208115260032187e-08,
+      "loss": 1.236,
+      "step": 9449
+    },
+    {
+      "epoch": 0.9942135718043135,
+      "grad_norm": 1.6919333934783936,
+      "learning_rate": 1.7563650455010295e-08,
+      "loss": 1.4099,
+      "step": 9450
+    },
+    {
+      "epoch": 0.9943187795896896,
+      "grad_norm": 1.644214391708374,
+      "learning_rate": 1.6930796281078297e-08,
+      "loss": 1.8462,
+      "step": 9451
+    },
+    {
+      "epoch": 0.9944239873750658,
+      "grad_norm": 1.6857322454452515,
+      "learning_rate": 1.6309552811744067e-08,
+      "loss": 1.568,
+      "step": 9452
+    },
+    {
+      "epoch": 0.9945291951604419,
+      "grad_norm": 1.1969326734542847,
+      "learning_rate": 1.569992011913879e-08,
+      "loss": 2.0799,
+      "step": 9453
+    },
+    {
+      "epoch": 0.994634402945818,
+      "grad_norm": 1.4197190999984741,
+      "learning_rate": 1.5101898274050286e-08,
+      "loss": 1.6347,
+      "step": 9454
+    },
+    {
+      "epoch": 0.9947396107311941,
+      "grad_norm": 1.7226569652557373,
+      "learning_rate": 1.4515487345956313e-08,
+      "loss": 1.5217,
+      "step": 9455
+    },
+    {
+      "epoch": 0.9948448185165703,
+      "grad_norm": 1.4008227586746216,
+      "learning_rate": 1.3940687402924646e-08,
+      "loss": 1.7268,
+      "step": 9456
+    },
+    {
+      "epoch": 0.9949500263019463,
+      "grad_norm": 1.268113613128662,
+      "learning_rate": 1.3377498511712993e-08,
+      "loss": 1.6932,
+      "step": 9457
+    },
+    {
+      "epoch": 0.9950552340873224,
+      "grad_norm": 1.930999994277954,
+      "learning_rate": 1.2825920737724596e-08,
+      "loss": 1.7915,
+      "step": 9458
+    },
+    {
+      "epoch": 0.9951604418726986,
+      "grad_norm": 2.0582447052001953,
+      "learning_rate": 1.2285954145008216e-08,
+      "loss": 1.3104,
+      "step": 9459
+    },
+    {
+      "epoch": 0.9952656496580747,
+      "grad_norm": 1.6370129585266113,
+      "learning_rate": 1.1757598796280355e-08,
+      "loss": 1.6174,
+      "step": 9460
+    },
+    {
+      "epoch": 0.9953708574434508,
+      "grad_norm": 1.1519914865493774,
+      "learning_rate": 1.1240854752880836e-08,
+      "loss": 1.6631,
+      "step": 9461
+    },
+    {
+      "epoch": 0.9954760652288269,
+      "grad_norm": 1.6220288276672363,
+      "learning_rate": 1.0735722074828313e-08,
+      "loss": 2.0331,
+      "step": 9462
+    },
+    {
+      "epoch": 0.9955812730142031,
+      "grad_norm": 1.75868558883667,
+      "learning_rate": 1.0242200820786974e-08,
+      "loss": 1.8015,
+      "step": 9463
+    },
+    {
+      "epoch": 0.9956864807995792,
+      "grad_norm": 1.8576023578643799,
+      "learning_rate": 9.760291048055426e-09,
+      "loss": 1.8266,
+      "step": 9464
+    },
+    {
+      "epoch": 0.9957916885849553,
+      "grad_norm": 1.2487952709197998,
+      "learning_rate": 9.289992812600012e-09,
+      "loss": 2.035,
+      "step": 9465
+    },
+    {
+      "epoch": 0.9958968963703314,
+      "grad_norm": 2.236783027648926,
+      "learning_rate": 8.831306169032604e-09,
+      "loss": 1.9039,
+      "step": 9466
+    },
+    {
+      "epoch": 0.9960021041557076,
+      "grad_norm": 1.4378468990325928,
+      "learning_rate": 8.384231170632805e-09,
+      "loss": 1.9437,
+      "step": 9467
+    },
+    {
+      "epoch": 0.9961073119410836,
+      "grad_norm": 1.5383528470993042,
+      "learning_rate": 7.94876786929244e-09,
+      "loss": 1.2174,
+      "step": 9468
+    },
+    {
+      "epoch": 0.9962125197264597,
+      "grad_norm": 3.4968624114990234,
+      "learning_rate": 7.52491631560437e-09,
+      "loss": 1.4205,
+      "step": 9469
+    },
+    {
+      "epoch": 0.9963177275118359,
+      "grad_norm": 1.9415160417556763,
+      "learning_rate": 7.112676558784781e-09,
+      "loss": 1.4909,
+      "step": 9470
+    },
+    {
+      "epoch": 0.996422935297212,
+      "grad_norm": 1.2908811569213867,
+      "learning_rate": 6.71204864669539e-09,
+      "loss": 1.5163,
+      "step": 9471
+    },
+    {
+      "epoch": 0.9965281430825881,
+      "grad_norm": 1.1630277633666992,
+      "learning_rate": 6.3230326258656435e-09,
+      "loss": 1.5793,
+      "step": 9472
+    },
+    {
+      "epoch": 0.9966333508679642,
+      "grad_norm": 1.6427892446517944,
+      "learning_rate": 5.945628541481619e-09,
+      "loss": 1.774,
+      "step": 9473
+    },
+    {
+      "epoch": 0.9967385586533404,
+      "grad_norm": 1.8046869039535522,
+      "learning_rate": 5.579836437341612e-09,
+      "loss": 1.6256,
+      "step": 9474
+    },
+    {
+      "epoch": 0.9968437664387164,
+      "grad_norm": 2.09666109085083,
+      "learning_rate": 5.225656355956066e-09,
+      "loss": 1.5055,
+      "step": 9475
+    },
+    {
+      "epoch": 0.9969489742240926,
+      "grad_norm": 1.8133103847503662,
+      "learning_rate": 4.883088338425434e-09,
+      "loss": 1.7931,
+      "step": 9476
+    },
+    {
+      "epoch": 0.9970541820094687,
+      "grad_norm": 1.8678925037384033,
+      "learning_rate": 4.552132424562317e-09,
+      "loss": 1.9548,
+      "step": 9477
+    },
+    {
+      "epoch": 0.9971593897948449,
+      "grad_norm": 1.6927310228347778,
+      "learning_rate": 4.2327886527693265e-09,
+      "loss": 1.4532,
+      "step": 9478
+    },
+    {
+      "epoch": 0.9972645975802209,
+      "grad_norm": 1.3339251279830933,
+      "learning_rate": 3.925057060150117e-09,
+      "loss": 1.4882,
+      "step": 9479
+    },
+    {
+      "epoch": 0.997369805365597,
+      "grad_norm": 2.199472188949585,
+      "learning_rate": 3.628937682431666e-09,
+      "loss": 1.782,
+      "step": 9480
+    },
+    {
+      "epoch": 0.9974750131509732,
+      "grad_norm": 1.5956065654754639,
+      "learning_rate": 3.344430554008682e-09,
+      "loss": 1.7815,
+      "step": 9481
+    },
+    {
+      "epoch": 0.9975802209363492,
+      "grad_norm": 1.884323000907898,
+      "learning_rate": 3.0715357079103003e-09,
+      "loss": 1.3084,
+      "step": 9482
+    },
+    {
+      "epoch": 0.9976854287217254,
+      "grad_norm": 1.6075438261032104,
+      "learning_rate": 2.8102531758333883e-09,
+      "loss": 1.4152,
+      "step": 9483
+    },
+    {
+      "epoch": 0.9977906365071015,
+      "grad_norm": 1.7321268320083618,
+      "learning_rate": 2.5605829881203414e-09,
+      "loss": 1.776,
+      "step": 9484
+    },
+    {
+      "epoch": 0.9978958442924777,
+      "grad_norm": 1.7372652292251587,
+      "learning_rate": 2.3225251737701846e-09,
+      "loss": 1.2503,
+      "step": 9485
+    },
+    {
+      "epoch": 0.9980010520778537,
+      "grad_norm": 1.7862646579742432,
+      "learning_rate": 2.0960797604052675e-09,
+      "loss": 1.4679,
+      "step": 9486
+    },
+    {
+      "epoch": 0.9981062598632299,
+      "grad_norm": 1.7311826944351196,
+      "learning_rate": 1.881246774348977e-09,
+      "loss": 0.6319,
+      "step": 9487
+    },
+    {
+      "epoch": 0.998211467648606,
+      "grad_norm": 2.2221319675445557,
+      "learning_rate": 1.678026240536923e-09,
+      "loss": 1.4301,
+      "step": 9488
+    },
+    {
+      "epoch": 0.9983166754339821,
+      "grad_norm": 1.4736319780349731,
+      "learning_rate": 1.4864181825613444e-09,
+      "loss": 1.6957,
+      "step": 9489
+    },
+    {
+      "epoch": 0.9984218832193582,
+      "grad_norm": 1.5542229413986206,
+      "learning_rate": 1.3064226226933152e-09,
+      "loss": 1.4633,
+      "step": 9490
+    },
+    {
+      "epoch": 0.9985270910047344,
+      "grad_norm": 1.928504467010498,
+      "learning_rate": 1.1380395818050282e-09,
+      "loss": 1.9097,
+      "step": 9491
+    },
+    {
+      "epoch": 0.9986322987901105,
+      "grad_norm": 1.4703706502914429,
+      "learning_rate": 9.812690794808177e-10,
+      "loss": 1.5353,
+      "step": 9492
+    },
+    {
+      "epoch": 0.9987375065754865,
+      "grad_norm": 1.7381463050842285,
+      "learning_rate": 8.361111339061367e-10,
+      "loss": 1.562,
+      "step": 9493
+    },
+    {
+      "epoch": 0.9988427143608627,
+      "grad_norm": 2.3469841480255127,
+      "learning_rate": 7.025657619563752e-10,
+      "loss": 1.756,
+      "step": 9494
+    },
+    {
+      "epoch": 0.9989479221462388,
+      "grad_norm": 1.3420002460479736,
+      "learning_rate": 5.806329791191445e-10,
+      "loss": 1.4611,
+      "step": 9495
+    },
+    {
+      "epoch": 0.999053129931615,
+      "grad_norm": 1.709895372390747,
+      "learning_rate": 4.703127995608902e-10,
+      "loss": 1.4044,
+      "step": 9496
+    },
+    {
+      "epoch": 0.999158337716991,
+      "grad_norm": 1.3718494176864624,
+      "learning_rate": 3.716052360935862e-10,
+      "loss": 1.7666,
+      "step": 9497
+    },
+    {
+      "epoch": 0.9992635455023672,
+      "grad_norm": 1.1952706575393677,
+      "learning_rate": 2.8451030018583623e-10,
+      "loss": 1.8885,
+      "step": 9498
+    },
+    {
+      "epoch": 0.9993687532877433,
+      "grad_norm": 1.6405409574508667,
+      "learning_rate": 2.0902800194066985e-10,
+      "loss": 1.5163,
+      "step": 9499
+    },
+    {
+      "epoch": 0.9994739610731194,
+      "grad_norm": 1.4296813011169434,
+      "learning_rate": 1.4515835012884893e-10,
+      "loss": 1.4754,
+      "step": 9500
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 9505,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.969403554197504e+17,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}