diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,9134 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1300,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0015384615384615385,
+      "grad_norm": 60.89158644985447,
+      "learning_rate": 2e-05,
+      "loss": 1.2107,
+      "step": 1
+    },
+    {
+      "epoch": 0.003076923076923077,
+      "grad_norm": 31.499919740759644,
+      "learning_rate": 1.9999992699997636e-05,
+      "loss": 0.7837,
+      "step": 2
+    },
+    {
+      "epoch": 0.004615384615384616,
+      "grad_norm": 6.103692995379824,
+      "learning_rate": 1.999997080000119e-05,
+      "loss": 0.3218,
+      "step": 3
+    },
+    {
+      "epoch": 0.006153846153846154,
+      "grad_norm": 2.6468306725602764,
+      "learning_rate": 1.9999934300042646e-05,
+      "loss": 0.2287,
+      "step": 4
+    },
+    {
+      "epoch": 0.007692307692307693,
+      "grad_norm": 3.4867739541054443,
+      "learning_rate": 1.9999883200175286e-05,
+      "loss": 0.2423,
+      "step": 5
+    },
+    {
+      "epoch": 0.009230769230769232,
+      "grad_norm": 2.641228316079499,
+      "learning_rate": 1.9999817500473724e-05,
+      "loss": 0.2253,
+      "step": 6
+    },
+    {
+      "epoch": 0.010769230769230769,
+      "grad_norm": 1.205202032246029,
+      "learning_rate": 1.9999737201033877e-05,
+      "loss": 0.1668,
+      "step": 7
+    },
+    {
+      "epoch": 0.012307692307692308,
+      "grad_norm": 1.9244405042319512,
+      "learning_rate": 1.999964230197298e-05,
+      "loss": 0.1847,
+      "step": 8
+    },
+    {
+      "epoch": 0.013846153846153847,
+      "grad_norm": 1.4464951947932236,
+      "learning_rate": 1.999953280342959e-05,
+      "loss": 0.1606,
+      "step": 9
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 1.4255296128290331,
+      "learning_rate": 1.999940870556357e-05,
+      "loss": 0.1471,
+      "step": 10
+    },
+    {
+      "epoch": 0.016923076923076923,
+      "grad_norm": 1.3096222246733347,
+      "learning_rate": 1.9999270008556108e-05,
+      "loss": 0.1463,
+      "step": 11
+    },
+    {
+      "epoch": 0.018461538461538463,
+      "grad_norm": 1.018551951164187,
+      "learning_rate": 1.99991167126097e-05,
+      "loss": 0.1405,
+      "step": 12
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1849805458454175,
+      "learning_rate": 1.9998948817948157e-05,
+      "loss": 0.1381,
+      "step": 13
+    },
+    {
+      "epoch": 0.021538461538461538,
+      "grad_norm": 1.01325599101221,
+      "learning_rate": 1.9998766324816606e-05,
+      "loss": 0.1453,
+      "step": 14
+    },
+    {
+      "epoch": 0.023076923076923078,
+      "grad_norm": 1.0760344920602294,
+      "learning_rate": 1.999856923348149e-05,
+      "loss": 0.119,
+      "step": 15
+    },
+    {
+      "epoch": 0.024615384615384615,
+      "grad_norm": 1.2012958039069457,
+      "learning_rate": 1.9998357544230558e-05,
+      "loss": 0.1507,
+      "step": 16
+    },
+    {
+      "epoch": 0.026153846153846153,
+      "grad_norm": 0.8734600178036379,
+      "learning_rate": 1.9998131257372878e-05,
+      "loss": 0.1334,
+      "step": 17
+    },
+    {
+      "epoch": 0.027692307692307693,
+      "grad_norm": 0.7321244232592351,
+      "learning_rate": 1.9997890373238827e-05,
+      "loss": 0.1259,
+      "step": 18
+    },
+    {
+      "epoch": 0.02923076923076923,
+      "grad_norm": 0.8736141545923927,
+      "learning_rate": 1.99976348921801e-05,
+      "loss": 0.1328,
+      "step": 19
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 0.7433251995484863,
+      "learning_rate": 1.9997364814569696e-05,
+      "loss": 0.1155,
+      "step": 20
+    },
+    {
+      "epoch": 0.03230769230769231,
+      "grad_norm": 0.7421987879217785,
+      "learning_rate": 1.9997080140801932e-05,
+      "loss": 0.1321,
+      "step": 21
+    },
+    {
+      "epoch": 0.033846153846153845,
+      "grad_norm": 0.7417777234873172,
+      "learning_rate": 1.9996780871292428e-05,
+      "loss": 0.1082,
+      "step": 22
+    },
+    {
+      "epoch": 0.03538461538461538,
+      "grad_norm": 1.2517479737195332,
+      "learning_rate": 1.999646700647812e-05,
+      "loss": 0.1219,
+      "step": 23
+    },
+    {
+      "epoch": 0.036923076923076927,
+      "grad_norm": 0.9094077693848758,
+      "learning_rate": 1.9996138546817248e-05,
+      "loss": 0.1314,
+      "step": 24
+    },
+    {
+      "epoch": 0.038461538461538464,
+      "grad_norm": 0.7797984417608248,
+      "learning_rate": 1.9995795492789368e-05,
+      "loss": 0.1331,
+      "step": 25
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7976816314521917,
+      "learning_rate": 1.9995437844895337e-05,
+      "loss": 0.114,
+      "step": 26
+    },
+    {
+      "epoch": 0.04153846153846154,
+      "grad_norm": 0.8888193123533839,
+      "learning_rate": 1.9995065603657317e-05,
+      "loss": 0.1399,
+      "step": 27
+    },
+    {
+      "epoch": 0.043076923076923075,
+      "grad_norm": 0.8169350017845466,
+      "learning_rate": 1.9994678769618784e-05,
+      "loss": 0.1148,
+      "step": 28
+    },
+    {
+      "epoch": 0.04461538461538461,
+      "grad_norm": 0.6804970735056192,
+      "learning_rate": 1.999427734334452e-05,
+      "loss": 0.1223,
+      "step": 29
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 0.7358036673244082,
+      "learning_rate": 1.99938613254206e-05,
+      "loss": 0.1281,
+      "step": 30
+    },
+    {
+      "epoch": 0.047692307692307694,
+      "grad_norm": 0.726915700239593,
+      "learning_rate": 1.9993430716454415e-05,
+      "loss": 0.123,
+      "step": 31
+    },
+    {
+      "epoch": 0.04923076923076923,
+      "grad_norm": 0.8370086699996789,
+      "learning_rate": 1.9992985517074653e-05,
+      "loss": 0.1121,
+      "step": 32
+    },
+    {
+      "epoch": 0.05076923076923077,
+      "grad_norm": 0.870233844578574,
+      "learning_rate": 1.9992525727931303e-05,
+      "loss": 0.1242,
+      "step": 33
+    },
+    {
+      "epoch": 0.052307692307692305,
+      "grad_norm": 0.6572319680505859,
+      "learning_rate": 1.9992051349695662e-05,
+      "loss": 0.115,
+      "step": 34
+    },
+    {
+      "epoch": 0.05384615384615385,
+      "grad_norm": 0.8992858978866627,
+      "learning_rate": 1.9991562383060316e-05,
+      "loss": 0.1173,
+      "step": 35
+    },
+    {
+      "epoch": 0.055384615384615386,
+      "grad_norm": 0.6238895374463513,
+      "learning_rate": 1.9991058828739164e-05,
+      "loss": 0.1188,
+      "step": 36
+    },
+    {
+      "epoch": 0.05692307692307692,
+      "grad_norm": 0.5905879008819883,
+      "learning_rate": 1.9990540687467394e-05,
+      "loss": 0.1162,
+      "step": 37
+    },
+    {
+      "epoch": 0.05846153846153846,
+      "grad_norm": 0.7475485922157608,
+      "learning_rate": 1.9990007960001487e-05,
+      "loss": 0.1341,
+      "step": 38
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7062185012543412,
+      "learning_rate": 1.9989460647119232e-05,
+      "loss": 0.1276,
+      "step": 39
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 0.6940436820752224,
+      "learning_rate": 1.9988898749619702e-05,
+      "loss": 0.1173,
+      "step": 40
+    },
+    {
+      "epoch": 0.06307692307692307,
+      "grad_norm": 0.6162083864347133,
+      "learning_rate": 1.998832226832327e-05,
+      "loss": 0.1148,
+      "step": 41
+    },
+    {
+      "epoch": 0.06461538461538462,
+      "grad_norm": 0.5010819273602969,
+      "learning_rate": 1.9987731204071596e-05,
+      "loss": 0.0971,
+      "step": 42
+    },
+    {
+      "epoch": 0.06615384615384616,
+      "grad_norm": 0.6715601368175806,
+      "learning_rate": 1.9987125557727633e-05,
+      "loss": 0.1279,
+      "step": 43
+    },
+    {
+      "epoch": 0.06769230769230769,
+      "grad_norm": 0.6360471415422658,
+      "learning_rate": 1.998650533017563e-05,
+      "loss": 0.1106,
+      "step": 44
+    },
+    {
+      "epoch": 0.06923076923076923,
+      "grad_norm": 0.6293809806468533,
+      "learning_rate": 1.9985870522321118e-05,
+      "loss": 0.1367,
+      "step": 45
+    },
+    {
+      "epoch": 0.07076923076923076,
+      "grad_norm": 0.6592095131341471,
+      "learning_rate": 1.9985221135090917e-05,
+      "loss": 0.1149,
+      "step": 46
+    },
+    {
+      "epoch": 0.07230769230769231,
+      "grad_norm": 0.7078516914096971,
+      "learning_rate": 1.9984557169433126e-05,
+      "loss": 0.1115,
+      "step": 47
+    },
+    {
+      "epoch": 0.07384615384615385,
+      "grad_norm": 0.5877725742836197,
+      "learning_rate": 1.9983878626317147e-05,
+      "loss": 0.1236,
+      "step": 48
+    },
+    {
+      "epoch": 0.07538461538461538,
+      "grad_norm": 0.5884104828878249,
+      "learning_rate": 1.9983185506733643e-05,
+      "loss": 0.12,
+      "step": 49
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.549670934870528,
+      "learning_rate": 1.9982477811694578e-05,
+      "loss": 0.1104,
+      "step": 50
+    },
+    {
+      "epoch": 0.07846153846153846,
+      "grad_norm": 0.8220615421808383,
+      "learning_rate": 1.9981755542233175e-05,
+      "loss": 0.1219,
+      "step": 51
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6798315000266852,
+      "learning_rate": 1.998101869940396e-05,
+      "loss": 0.1182,
+      "step": 52
+    },
+    {
+      "epoch": 0.08153846153846153,
+      "grad_norm": 0.5366855615637781,
+      "learning_rate": 1.9980267284282718e-05,
+      "loss": 0.1017,
+      "step": 53
+    },
+    {
+      "epoch": 0.08307692307692308,
+      "grad_norm": 0.7397417525631895,
+      "learning_rate": 1.9979501297966516e-05,
+      "loss": 0.114,
+      "step": 54
+    },
+    {
+      "epoch": 0.08461538461538462,
+      "grad_norm": 0.7086921001674501,
+      "learning_rate": 1.9978720741573693e-05,
+      "loss": 0.1179,
+      "step": 55
+    },
+    {
+      "epoch": 0.08615384615384615,
+      "grad_norm": 0.7075396570589779,
+      "learning_rate": 1.9977925616243865e-05,
+      "loss": 0.1143,
+      "step": 56
+    },
+    {
+      "epoch": 0.0876923076923077,
+      "grad_norm": 0.5477986445198524,
+      "learning_rate": 1.9977115923137912e-05,
+      "loss": 0.1115,
+      "step": 57
+    },
+    {
+      "epoch": 0.08923076923076922,
+      "grad_norm": 0.5949900627411874,
+      "learning_rate": 1.997629166343799e-05,
+      "loss": 0.1056,
+      "step": 58
+    },
+    {
+      "epoch": 0.09076923076923077,
+      "grad_norm": 0.60111264237703,
+      "learning_rate": 1.9975452838347513e-05,
+      "loss": 0.1097,
+      "step": 59
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 0.6683497211941376,
+      "learning_rate": 1.997459944909117e-05,
+      "loss": 0.1164,
+      "step": 60
+    },
+    {
+      "epoch": 0.09384615384615384,
+      "grad_norm": 0.7704030456569986,
+      "learning_rate": 1.9973731496914914e-05,
+      "loss": 0.1164,
+      "step": 61
+    },
+    {
+      "epoch": 0.09538461538461539,
+      "grad_norm": 0.707210267374331,
+      "learning_rate": 1.9972848983085945e-05,
+      "loss": 0.1192,
+      "step": 62
+    },
+    {
+      "epoch": 0.09692307692307692,
+      "grad_norm": 0.6016618867618871,
+      "learning_rate": 1.9971951908892743e-05,
+      "loss": 0.111,
+      "step": 63
+    },
+    {
+      "epoch": 0.09846153846153846,
+      "grad_norm": 0.662240507889871,
+      "learning_rate": 1.997104027564503e-05,
+      "loss": 0.1163,
+      "step": 64
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5798279422985586,
+      "learning_rate": 1.9970114084673796e-05,
+      "loss": 0.1102,
+      "step": 65
+    },
+    {
+      "epoch": 0.10153846153846154,
+      "grad_norm": 0.5611805453593508,
+      "learning_rate": 1.9969173337331283e-05,
+      "loss": 0.1104,
+      "step": 66
+    },
+    {
+      "epoch": 0.10307692307692308,
+      "grad_norm": 0.7514312817329308,
+      "learning_rate": 1.996821803499097e-05,
+      "loss": 0.1239,
+      "step": 67
+    },
+    {
+      "epoch": 0.10461538461538461,
+      "grad_norm": 0.5863699065118245,
+      "learning_rate": 1.9967248179047612e-05,
+      "loss": 0.1132,
+      "step": 68
+    },
+    {
+      "epoch": 0.10615384615384615,
+      "grad_norm": 0.5464510643573779,
+      "learning_rate": 1.9966263770917192e-05,
+      "loss": 0.1211,
+      "step": 69
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 0.566891174856317,
+      "learning_rate": 1.996526481203695e-05,
+      "loss": 0.1077,
+      "step": 70
+    },
+    {
+      "epoch": 0.10923076923076923,
+      "grad_norm": 0.6246113449127777,
+      "learning_rate": 1.9964251303865362e-05,
+      "loss": 0.1074,
+      "step": 71
+    },
+    {
+      "epoch": 0.11076923076923077,
+      "grad_norm": 0.6001393125416704,
+      "learning_rate": 1.9963223247882154e-05,
+      "loss": 0.1086,
+      "step": 72
+    },
+    {
+      "epoch": 0.1123076923076923,
+      "grad_norm": 0.5021027088495942,
+      "learning_rate": 1.996218064558829e-05,
+      "loss": 0.0983,
+      "step": 73
+    },
+    {
+      "epoch": 0.11384615384615385,
+      "grad_norm": 0.5707671002037145,
+      "learning_rate": 1.9961123498505965e-05,
+      "loss": 0.1076,
+      "step": 74
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.6742358069569517,
+      "learning_rate": 1.9960051808178616e-05,
+      "loss": 0.1148,
+      "step": 75
+    },
+    {
+      "epoch": 0.11692307692307692,
+      "grad_norm": 0.48574494522940415,
+      "learning_rate": 1.995896557617091e-05,
+      "loss": 0.1104,
+      "step": 76
+    },
+    {
+      "epoch": 0.11846153846153847,
+      "grad_norm": 0.5943862604998493,
+      "learning_rate": 1.9957864804068752e-05,
+      "loss": 0.1076,
+      "step": 77
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6097303023134425,
+      "learning_rate": 1.9956749493479263e-05,
+      "loss": 0.1115,
+      "step": 78
+    },
+    {
+      "epoch": 0.12153846153846154,
+      "grad_norm": 0.6195963924681878,
+      "learning_rate": 1.99556196460308e-05,
+      "loss": 0.1212,
+      "step": 79
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 0.5930349140433203,
+      "learning_rate": 1.9954475263372943e-05,
+      "loss": 0.1074,
+      "step": 80
+    },
+    {
+      "epoch": 0.12461538461538461,
+      "grad_norm": 0.5556371192720652,
+      "learning_rate": 1.995331634717649e-05,
+      "loss": 0.1064,
+      "step": 81
+    },
+    {
+      "epoch": 0.12615384615384614,
+      "grad_norm": 0.6530827388442122,
+      "learning_rate": 1.9952142899133452e-05,
+      "loss": 0.1121,
+      "step": 82
+    },
+    {
+      "epoch": 0.1276923076923077,
+      "grad_norm": 0.8859927355734215,
+      "learning_rate": 1.9950954920957074e-05,
+      "loss": 0.1152,
+      "step": 83
+    },
+    {
+      "epoch": 0.12923076923076923,
+      "grad_norm": 0.5935612351835159,
+      "learning_rate": 1.99497524143818e-05,
+      "loss": 0.1179,
+      "step": 84
+    },
+    {
+      "epoch": 0.13076923076923078,
+      "grad_norm": 0.4991766795445498,
+      "learning_rate": 1.994853538116329e-05,
+      "loss": 0.1149,
+      "step": 85
+    },
+    {
+      "epoch": 0.13230769230769232,
+      "grad_norm": 0.5555671407736567,
+      "learning_rate": 1.9947303823078418e-05,
+      "loss": 0.1012,
+      "step": 86
+    },
+    {
+      "epoch": 0.13384615384615384,
+      "grad_norm": 0.8909613249571459,
+      "learning_rate": 1.994605774192525e-05,
+      "loss": 0.0979,
+      "step": 87
+    },
+    {
+      "epoch": 0.13538461538461538,
+      "grad_norm": 0.5310150106650096,
+      "learning_rate": 1.9944797139523068e-05,
+      "loss": 0.1088,
+      "step": 88
+    },
+    {
+      "epoch": 0.13692307692307693,
+      "grad_norm": 0.749673935234374,
+      "learning_rate": 1.994352201771236e-05,
+      "loss": 0.1034,
+      "step": 89
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 0.5378188928120664,
+      "learning_rate": 1.99422323783548e-05,
+      "loss": 0.1057,
+      "step": 90
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.511920881530378,
+      "learning_rate": 1.9940928223333254e-05,
+      "loss": 0.1121,
+      "step": 91
+    },
+    {
+      "epoch": 0.14153846153846153,
+      "grad_norm": 0.5822824054135747,
+      "learning_rate": 1.99396095545518e-05,
+      "loss": 0.1148,
+      "step": 92
+    },
+    {
+      "epoch": 0.14307692307692307,
+      "grad_norm": 0.6672810329410515,
+      "learning_rate": 1.9938276373935688e-05,
+      "loss": 0.1158,
+      "step": 93
+    },
+    {
+      "epoch": 0.14461538461538462,
+      "grad_norm": 0.4952193512759965,
+      "learning_rate": 1.9936928683431368e-05,
+      "loss": 0.1076,
+      "step": 94
+    },
+    {
+      "epoch": 0.14615384615384616,
+      "grad_norm": 0.625543647549158,
+      "learning_rate": 1.9935566485006464e-05,
+      "loss": 0.1125,
+      "step": 95
+    },
+    {
+      "epoch": 0.1476923076923077,
+      "grad_norm": 0.6035686191037584,
+      "learning_rate": 1.993418978064979e-05,
+      "loss": 0.109,
+      "step": 96
+    },
+    {
+      "epoch": 0.14923076923076922,
+      "grad_norm": 0.5236314046610593,
+      "learning_rate": 1.993279857237133e-05,
+      "loss": 0.1152,
+      "step": 97
+    },
+    {
+      "epoch": 0.15076923076923077,
+      "grad_norm": 0.6169915505799958,
+      "learning_rate": 1.9931392862202255e-05,
+      "loss": 0.1187,
+      "step": 98
+    },
+    {
+      "epoch": 0.1523076923076923,
+      "grad_norm": 0.5079224872644268,
+      "learning_rate": 1.99299726521949e-05,
+      "loss": 0.1066,
+      "step": 99
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.6279665388236755,
+      "learning_rate": 1.992853794442277e-05,
+      "loss": 0.1077,
+      "step": 100
+    },
+    {
+      "epoch": 0.15538461538461537,
+      "grad_norm": 0.5259964466933738,
+      "learning_rate": 1.992708874098054e-05,
+      "loss": 0.1107,
+      "step": 101
+    },
+    {
+      "epoch": 0.15692307692307692,
+      "grad_norm": 0.5221587453615723,
+      "learning_rate": 1.9925625043984052e-05,
+      "loss": 0.103,
+      "step": 102
+    },
+    {
+      "epoch": 0.15846153846153846,
+      "grad_norm": 0.5299978643337,
+      "learning_rate": 1.9924146855570298e-05,
+      "loss": 0.1083,
+      "step": 103
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5822846737863101,
+      "learning_rate": 1.9922654177897436e-05,
+      "loss": 0.1069,
+      "step": 104
+    },
+    {
+      "epoch": 0.16153846153846155,
+      "grad_norm": 0.5319364893865596,
+      "learning_rate": 1.9921147013144782e-05,
+      "loss": 0.0997,
+      "step": 105
+    },
+    {
+      "epoch": 0.16307692307692306,
+      "grad_norm": 0.6977409875795292,
+      "learning_rate": 1.9919625363512788e-05,
+      "loss": 0.118,
+      "step": 106
+    },
+    {
+      "epoch": 0.1646153846153846,
+      "grad_norm": 0.48032168376624,
+      "learning_rate": 1.9918089231223066e-05,
+      "loss": 0.1048,
+      "step": 107
+    },
+    {
+      "epoch": 0.16615384615384615,
+      "grad_norm": 0.5258925153786688,
+      "learning_rate": 1.991653861851837e-05,
+      "loss": 0.1125,
+      "step": 108
+    },
+    {
+      "epoch": 0.1676923076923077,
+      "grad_norm": 0.7141043684744723,
+      "learning_rate": 1.99149735276626e-05,
+      "loss": 0.1181,
+      "step": 109
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 0.5038196623101785,
+      "learning_rate": 1.9913393960940785e-05,
+      "loss": 0.0971,
+      "step": 110
+    },
+    {
+      "epoch": 0.17076923076923076,
+      "grad_norm": 0.48472163979718835,
+      "learning_rate": 1.9911799920659093e-05,
+      "loss": 0.1039,
+      "step": 111
+    },
+    {
+      "epoch": 0.1723076923076923,
+      "grad_norm": 0.669381566002829,
+      "learning_rate": 1.9910191409144825e-05,
+      "loss": 0.1073,
+      "step": 112
+    },
+    {
+      "epoch": 0.17384615384615384,
+      "grad_norm": 0.57115456078115,
+      "learning_rate": 1.9908568428746408e-05,
+      "loss": 0.0954,
+      "step": 113
+    },
+    {
+      "epoch": 0.1753846153846154,
+      "grad_norm": 0.5395647646350689,
+      "learning_rate": 1.9906930981833392e-05,
+      "loss": 0.1088,
+      "step": 114
+    },
+    {
+      "epoch": 0.17692307692307693,
+      "grad_norm": 0.5912868508214638,
+      "learning_rate": 1.9905279070796454e-05,
+      "loss": 0.1035,
+      "step": 115
+    },
+    {
+      "epoch": 0.17846153846153845,
+      "grad_norm": 0.56400423327266,
+      "learning_rate": 1.9903612698047387e-05,
+      "loss": 0.1223,
+      "step": 116
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6509772627591756,
+      "learning_rate": 1.9901931866019087e-05,
+      "loss": 0.1165,
+      "step": 117
+    },
+    {
+      "epoch": 0.18153846153846154,
+      "grad_norm": 0.4920607096781014,
+      "learning_rate": 1.990023657716558e-05,
+      "loss": 0.1036,
+      "step": 118
+    },
+    {
+      "epoch": 0.18307692307692308,
+      "grad_norm": 1.003150053450431,
+      "learning_rate": 1.989852683396198e-05,
+      "loss": 0.109,
+      "step": 119
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 0.5598317796088923,
+      "learning_rate": 1.9896802638904512e-05,
+      "loss": 0.1099,
+      "step": 120
+    },
+    {
+      "epoch": 0.18615384615384614,
+      "grad_norm": 1.5739160303036723,
+      "learning_rate": 1.9895063994510512e-05,
+      "loss": 0.1149,
+      "step": 121
+    },
+    {
+      "epoch": 0.18769230769230769,
+      "grad_norm": 0.6575202377156634,
+      "learning_rate": 1.9893310903318394e-05,
+      "loss": 0.1115,
+      "step": 122
+    },
+    {
+      "epoch": 0.18923076923076923,
+      "grad_norm": 0.6474947347376268,
+      "learning_rate": 1.9891543367887675e-05,
+      "loss": 0.1105,
+      "step": 123
+    },
+    {
+      "epoch": 0.19076923076923077,
+      "grad_norm": 0.6273901725857466,
+      "learning_rate": 1.9889761390798952e-05,
+      "loss": 0.1137,
+      "step": 124
+    },
+    {
+      "epoch": 0.19230769230769232,
+      "grad_norm": 0.5821482969182745,
+      "learning_rate": 1.988796497465392e-05,
+      "loss": 0.0932,
+      "step": 125
+    },
+    {
+      "epoch": 0.19384615384615383,
+      "grad_norm": 0.6213235073882689,
+      "learning_rate": 1.9886154122075344e-05,
+      "loss": 0.1111,
+      "step": 126
+    },
+    {
+      "epoch": 0.19538461538461538,
+      "grad_norm": 0.6063602698529751,
+      "learning_rate": 1.988432883570707e-05,
+      "loss": 0.1102,
+      "step": 127
+    },
+    {
+      "epoch": 0.19692307692307692,
+      "grad_norm": 0.5820333042475602,
+      "learning_rate": 1.9882489118214013e-05,
+      "loss": 0.0904,
+      "step": 128
+    },
+    {
+      "epoch": 0.19846153846153847,
+      "grad_norm": 0.5911684289948976,
+      "learning_rate": 1.9880634972282168e-05,
+      "loss": 0.1168,
+      "step": 129
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6156260168090836,
+      "learning_rate": 1.987876640061858e-05,
+      "loss": 0.1037,
+      "step": 130
+    },
+    {
+      "epoch": 0.20153846153846153,
+      "grad_norm": 0.6339706412660758,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.1023,
+      "step": 131
+    },
+    {
+      "epoch": 0.20307692307692307,
+      "grad_norm": 0.5925980576093635,
+      "learning_rate": 1.9874985991029725e-05,
+      "loss": 0.1058,
+      "step": 132
+    },
+    {
+      "epoch": 0.20461538461538462,
+      "grad_norm": 0.44542298631854954,
+      "learning_rate": 1.987307415862385e-05,
+      "loss": 0.0947,
+      "step": 133
+    },
+    {
+      "epoch": 0.20615384615384616,
+      "grad_norm": 0.6583404135837835,
+      "learning_rate": 1.987114791152503e-05,
+      "loss": 0.1079,
+      "step": 134
+    },
+    {
+      "epoch": 0.2076923076923077,
+      "grad_norm": 0.6888868393971377,
+      "learning_rate": 1.9869207252545582e-05,
+      "loss": 0.1046,
+      "step": 135
+    },
+    {
+      "epoch": 0.20923076923076922,
+      "grad_norm": 0.5864137998381262,
+      "learning_rate": 1.9867252184518878e-05,
+      "loss": 0.1128,
+      "step": 136
+    },
+    {
+      "epoch": 0.21076923076923076,
+      "grad_norm": 0.6371765472826588,
+      "learning_rate": 1.986528271029931e-05,
+      "loss": 0.1108,
+      "step": 137
+    },
+    {
+      "epoch": 0.2123076923076923,
+      "grad_norm": 0.6990532523375615,
+      "learning_rate": 1.9863298832762317e-05,
+      "loss": 0.1121,
+      "step": 138
+    },
+    {
+      "epoch": 0.21384615384615385,
+      "grad_norm": 0.5777908223862595,
+      "learning_rate": 1.9861300554804357e-05,
+      "loss": 0.1041,
+      "step": 139
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 0.5437019574342667,
+      "learning_rate": 1.985928787934292e-05,
+      "loss": 0.1041,
+      "step": 140
+    },
+    {
+      "epoch": 0.2169230769230769,
+      "grad_norm": 0.5652145969922582,
+      "learning_rate": 1.985726080931651e-05,
+      "loss": 0.0977,
+      "step": 141
+    },
+    {
+      "epoch": 0.21846153846153846,
+      "grad_norm": 0.6831475109006377,
+      "learning_rate": 1.9855219347684654e-05,
+      "loss": 0.1136,
+      "step": 142
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5819561900456608,
+      "learning_rate": 1.9853163497427885e-05,
+      "loss": 0.1022,
+      "step": 143
+    },
+    {
+      "epoch": 0.22153846153846155,
+      "grad_norm": 0.5243548380003794,
+      "learning_rate": 1.985109326154774e-05,
+      "loss": 0.1068,
+      "step": 144
+    },
+    {
+      "epoch": 0.2230769230769231,
+      "grad_norm": 0.6756018205387149,
+      "learning_rate": 1.9849008643066774e-05,
+      "loss": 0.1116,
+      "step": 145
+    },
+    {
+      "epoch": 0.2246153846153846,
+      "grad_norm": 0.5271647520035011,
+      "learning_rate": 1.9846909645028524e-05,
+      "loss": 0.1004,
+      "step": 146
+    },
+    {
+      "epoch": 0.22615384615384615,
+      "grad_norm": 0.49376009484771377,
+      "learning_rate": 1.984479627049753e-05,
+      "loss": 0.0986,
+      "step": 147
+    },
+    {
+      "epoch": 0.2276923076923077,
+      "grad_norm": 0.5009074295953662,
+      "learning_rate": 1.9842668522559326e-05,
+      "loss": 0.1083,
+      "step": 148
+    },
+    {
+      "epoch": 0.22923076923076924,
+      "grad_norm": 0.5120110822425591,
+      "learning_rate": 1.9840526404320415e-05,
+      "loss": 0.1035,
+      "step": 149
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.4393813675110799,
+      "learning_rate": 1.9838369918908295e-05,
+      "loss": 0.1116,
+      "step": 150
+    },
+    {
+      "epoch": 0.2323076923076923,
+      "grad_norm": 0.6224825279714632,
+      "learning_rate": 1.983619906947144e-05,
+      "loss": 0.1067,
+      "step": 151
+    },
+    {
+      "epoch": 0.23384615384615384,
+      "grad_norm": 0.5038143439477663,
+      "learning_rate": 1.9834013859179284e-05,
+      "loss": 0.1104,
+      "step": 152
+    },
+    {
+      "epoch": 0.2353846153846154,
+      "grad_norm": 0.600725559118982,
+      "learning_rate": 1.9831814291222233e-05,
+      "loss": 0.1114,
+      "step": 153
+    },
+    {
+      "epoch": 0.23692307692307693,
+      "grad_norm": 0.561148124391653,
+      "learning_rate": 1.982960036881167e-05,
+      "loss": 0.1129,
+      "step": 154
+    },
+    {
+      "epoch": 0.23846153846153847,
+      "grad_norm": 0.4254379063861859,
+      "learning_rate": 1.982737209517991e-05,
+      "loss": 0.1018,
+      "step": 155
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5302552159073871,
+      "learning_rate": 1.982512947358024e-05,
+      "loss": 0.1103,
+      "step": 156
+    },
+    {
+      "epoch": 0.24153846153846154,
+      "grad_norm": 0.4768982697935648,
+      "learning_rate": 1.982287250728689e-05,
+      "loss": 0.0947,
+      "step": 157
+    },
+    {
+      "epoch": 0.24307692307692308,
+      "grad_norm": 0.4859749472919824,
+      "learning_rate": 1.9820601199595027e-05,
+      "loss": 0.1049,
+      "step": 158
+    },
+    {
+      "epoch": 0.24461538461538462,
+      "grad_norm": 0.49517524710561983,
+      "learning_rate": 1.981831555382076e-05,
+      "loss": 0.104,
+      "step": 159
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 0.462632720362608,
+      "learning_rate": 1.981601557330114e-05,
+      "loss": 0.101,
+      "step": 160
+    },
+    {
+      "epoch": 0.24769230769230768,
+      "grad_norm": 0.46154906358075987,
+      "learning_rate": 1.9813701261394136e-05,
+      "loss": 0.0971,
+      "step": 161
+    },
+    {
+      "epoch": 0.24923076923076923,
+      "grad_norm": 0.47520236421542805,
+      "learning_rate": 1.9811372621478643e-05,
+      "loss": 0.0887,
+      "step": 162
+    },
+    {
+      "epoch": 0.25076923076923074,
+      "grad_norm": 0.4386838659657528,
+      "learning_rate": 1.980902965695448e-05,
+      "loss": 0.1054,
+      "step": 163
+    },
+    {
+      "epoch": 0.2523076923076923,
+      "grad_norm": 0.5157191022451466,
+      "learning_rate": 1.9806672371242372e-05,
+      "loss": 0.1031,
+      "step": 164
+    },
+    {
+      "epoch": 0.25384615384615383,
+      "grad_norm": 0.49794944103500205,
+      "learning_rate": 1.9804300767783958e-05,
+      "loss": 0.1058,
+      "step": 165
+    },
+    {
+      "epoch": 0.2553846153846154,
+      "grad_norm": 0.5125213345791613,
+      "learning_rate": 1.9801914850041787e-05,
+      "loss": 0.1012,
+      "step": 166
+    },
+    {
+      "epoch": 0.2569230769230769,
+      "grad_norm": 0.5380352018124325,
+      "learning_rate": 1.979951462149929e-05,
+      "loss": 0.1096,
+      "step": 167
+    },
+    {
+      "epoch": 0.25846153846153846,
+      "grad_norm": 0.5283064485685516,
+      "learning_rate": 1.979710008566081e-05,
+      "loss": 0.0985,
+      "step": 168
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5187932463326723,
+      "learning_rate": 1.979467124605156e-05,
+      "loss": 0.0908,
+      "step": 169
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 0.5436736115984068,
+      "learning_rate": 1.979222810621766e-05,
+      "loss": 0.1033,
+      "step": 170
+    },
+    {
+      "epoch": 0.2630769230769231,
+      "grad_norm": 0.4474415033552116,
+      "learning_rate": 1.9789770669726088e-05,
+      "loss": 0.1053,
+      "step": 171
+    },
+    {
+      "epoch": 0.26461538461538464,
+      "grad_norm": 0.544303710798621,
+      "learning_rate": 1.97872989401647e-05,
+      "loss": 0.1159,
+      "step": 172
+    },
+    {
+      "epoch": 0.26615384615384613,
+      "grad_norm": 0.6201173307395067,
+      "learning_rate": 1.9784812921142232e-05,
+      "loss": 0.098,
+      "step": 173
+    },
+    {
+      "epoch": 0.2676923076923077,
+      "grad_norm": 0.447026396377833,
+      "learning_rate": 1.9782312616288262e-05,
+      "loss": 0.106,
+      "step": 174
+    },
+    {
+      "epoch": 0.2692307692307692,
+      "grad_norm": 0.5016997942574009,
+      "learning_rate": 1.977979802925324e-05,
+      "loss": 0.1059,
+      "step": 175
+    },
+    {
+      "epoch": 0.27076923076923076,
+      "grad_norm": 0.5883579853747641,
+      "learning_rate": 1.977726916370847e-05,
+      "loss": 0.1049,
+      "step": 176
+    },
+    {
+      "epoch": 0.2723076923076923,
+      "grad_norm": 0.43307778470945424,
+      "learning_rate": 1.977472602334609e-05,
+      "loss": 0.0917,
+      "step": 177
+    },
+    {
+      "epoch": 0.27384615384615385,
+      "grad_norm": 0.4108595764973751,
+      "learning_rate": 1.977216861187909e-05,
+      "loss": 0.0892,
+      "step": 178
+    },
+    {
+      "epoch": 0.2753846153846154,
+      "grad_norm": 0.5320809885623358,
+      "learning_rate": 1.976959693304129e-05,
+      "loss": 0.0994,
+      "step": 179
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 0.46480787256802003,
+      "learning_rate": 1.9767010990587342e-05,
+      "loss": 0.1013,
+      "step": 180
+    },
+    {
+      "epoch": 0.2784615384615385,
+      "grad_norm": 0.545604832179799,
+      "learning_rate": 1.9764410788292724e-05,
+      "loss": 0.0989,
+      "step": 181
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.43112552357630096,
+      "learning_rate": 1.976179632995373e-05,
+      "loss": 0.1045,
+      "step": 182
+    },
+    {
+      "epoch": 0.2815384615384615,
+      "grad_norm": 0.6658366935016972,
+      "learning_rate": 1.9759167619387474e-05,
+      "loss": 0.1233,
+      "step": 183
+    },
+    {
+      "epoch": 0.28307692307692306,
+      "grad_norm": 0.46835535422995606,
+      "learning_rate": 1.9756524660431876e-05,
+      "loss": 0.0956,
+      "step": 184
+    },
+    {
+      "epoch": 0.2846153846153846,
+      "grad_norm": 0.4607296339259721,
+      "learning_rate": 1.9753867456945653e-05,
+      "loss": 0.1161,
+      "step": 185
+    },
+    {
+      "epoch": 0.28615384615384615,
+      "grad_norm": 0.41625713094163824,
+      "learning_rate": 1.9751196012808328e-05,
+      "loss": 0.0896,
+      "step": 186
+    },
+    {
+      "epoch": 0.2876923076923077,
+      "grad_norm": 0.46438501415946737,
+      "learning_rate": 1.9748510331920204e-05,
+      "loss": 0.1016,
+      "step": 187
+    },
+    {
+      "epoch": 0.28923076923076924,
+      "grad_norm": 0.5188858027802394,
+      "learning_rate": 1.9745810418202383e-05,
+      "loss": 0.1199,
+      "step": 188
+    },
+    {
+      "epoch": 0.2907692307692308,
+      "grad_norm": 0.4563595166293062,
+      "learning_rate": 1.9743096275596735e-05,
+      "loss": 0.1028,
+      "step": 189
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 0.48359296387625755,
+      "learning_rate": 1.9740367908065914e-05,
+      "loss": 0.0925,
+      "step": 190
+    },
+    {
+      "epoch": 0.29384615384615387,
+      "grad_norm": 0.5276131684607785,
+      "learning_rate": 1.9737625319593338e-05,
+      "loss": 0.1083,
+      "step": 191
+    },
+    {
+      "epoch": 0.2953846153846154,
+      "grad_norm": 0.38771222370278047,
+      "learning_rate": 1.973486851418318e-05,
+      "loss": 0.0904,
+      "step": 192
+    },
+    {
+      "epoch": 0.2969230769230769,
+      "grad_norm": 0.5429417039030707,
+      "learning_rate": 1.9732097495860388e-05,
+      "loss": 0.0963,
+      "step": 193
+    },
+    {
+      "epoch": 0.29846153846153844,
+      "grad_norm": 0.5257696459782908,
+      "learning_rate": 1.9729312268670642e-05,
+      "loss": 0.0976,
+      "step": 194
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4550685781931343,
+      "learning_rate": 1.972651283668038e-05,
+      "loss": 0.0992,
+      "step": 195
+    },
+    {
+      "epoch": 0.30153846153846153,
+      "grad_norm": 0.49086953834293273,
+      "learning_rate": 1.9723699203976768e-05,
+      "loss": 0.0994,
+      "step": 196
+    },
+    {
+      "epoch": 0.3030769230769231,
+      "grad_norm": 0.5264210943330065,
+      "learning_rate": 1.9720871374667714e-05,
+      "loss": 0.1021,
+      "step": 197
+    },
+    {
+      "epoch": 0.3046153846153846,
+      "grad_norm": 0.507327301576836,
+      "learning_rate": 1.9718029352881857e-05,
+      "loss": 0.0972,
+      "step": 198
+    },
+    {
+      "epoch": 0.30615384615384617,
+      "grad_norm": 0.4732432615412584,
+      "learning_rate": 1.971517314276854e-05,
+      "loss": 0.0918,
+      "step": 199
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.6042723204316705,
+      "learning_rate": 1.9712302748497838e-05,
+      "loss": 0.1064,
+      "step": 200
+    },
+    {
+      "epoch": 0.30923076923076925,
+      "grad_norm": 0.46540271774907827,
+      "learning_rate": 1.9709418174260523e-05,
+      "loss": 0.0906,
+      "step": 201
+    },
+    {
+      "epoch": 0.31076923076923074,
+      "grad_norm": 0.5214921994645452,
+      "learning_rate": 1.9706519424268077e-05,
+      "loss": 0.093,
+      "step": 202
+    },
+    {
+      "epoch": 0.3123076923076923,
+      "grad_norm": 0.4249410321063521,
+      "learning_rate": 1.9703606502752674e-05,
+      "loss": 0.1011,
+      "step": 203
+    },
+    {
+      "epoch": 0.31384615384615383,
+      "grad_norm": 0.5120931365008091,
+      "learning_rate": 1.970067941396719e-05,
+      "loss": 0.1093,
+      "step": 204
+    },
+    {
+      "epoch": 0.3153846153846154,
+      "grad_norm": 0.46019038450225697,
+      "learning_rate": 1.9697738162185163e-05,
+      "loss": 0.1094,
+      "step": 205
+    },
+    {
+      "epoch": 0.3169230769230769,
+      "grad_norm": 0.43539468765926526,
+      "learning_rate": 1.969478275170083e-05,
+      "loss": 0.0867,
+      "step": 206
+    },
+    {
+      "epoch": 0.31846153846153846,
+      "grad_norm": 0.499590646424274,
+      "learning_rate": 1.969181318682909e-05,
+      "loss": 0.0908,
+      "step": 207
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.540818973654555,
+      "learning_rate": 1.9688829471905507e-05,
+      "loss": 0.1023,
+      "step": 208
+    },
+    {
+      "epoch": 0.32153846153846155,
+      "grad_norm": 0.4730876526778854,
+      "learning_rate": 1.9685831611286312e-05,
+      "loss": 0.1026,
+      "step": 209
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 0.49776463757579537,
+      "learning_rate": 1.968281960934838e-05,
+      "loss": 0.0996,
+      "step": 210
+    },
+    {
+      "epoch": 0.32461538461538464,
+      "grad_norm": 0.4311409270544766,
+      "learning_rate": 1.967979347048923e-05,
+      "loss": 0.099,
+      "step": 211
+    },
+    {
+      "epoch": 0.3261538461538461,
+      "grad_norm": 0.437076731735488,
+      "learning_rate": 1.9676753199127033e-05,
+      "loss": 0.0947,
+      "step": 212
+    },
+    {
+      "epoch": 0.32769230769230767,
+      "grad_norm": 0.4670077669146603,
+      "learning_rate": 1.9673698799700582e-05,
+      "loss": 0.0933,
+      "step": 213
+    },
+    {
+      "epoch": 0.3292307692307692,
+      "grad_norm": 0.4919569174836587,
+      "learning_rate": 1.9670630276669305e-05,
+      "loss": 0.1001,
+      "step": 214
+    },
+    {
+      "epoch": 0.33076923076923076,
+      "grad_norm": 0.47210284604663544,
+      "learning_rate": 1.9667547634513248e-05,
+      "loss": 0.1009,
+      "step": 215
+    },
+    {
+      "epoch": 0.3323076923076923,
+      "grad_norm": 0.37651995607019445,
+      "learning_rate": 1.9664450877733065e-05,
+      "loss": 0.0898,
+      "step": 216
+    },
+    {
+      "epoch": 0.33384615384615385,
+      "grad_norm": 0.534190751602756,
+      "learning_rate": 1.9661340010850025e-05,
+      "loss": 0.1104,
+      "step": 217
+    },
+    {
+      "epoch": 0.3353846153846154,
+      "grad_norm": 0.4951807767134038,
+      "learning_rate": 1.9658215038405997e-05,
+      "loss": 0.1068,
+      "step": 218
+    },
+    {
+      "epoch": 0.33692307692307694,
+      "grad_norm": 0.5728927862246131,
+      "learning_rate": 1.9655075964963443e-05,
+      "loss": 0.1055,
+      "step": 219
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 0.47984244005432625,
+      "learning_rate": 1.9651922795105404e-05,
+      "loss": 0.1021,
+      "step": 220
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5413827050893782,
+      "learning_rate": 1.9648755533435517e-05,
+      "loss": 0.1075,
+      "step": 221
+    },
+    {
+      "epoch": 0.3415384615384615,
+      "grad_norm": 0.4794837943323609,
+      "learning_rate": 1.9645574184577982e-05,
+      "loss": 0.1042,
+      "step": 222
+    },
+    {
+      "epoch": 0.34307692307692306,
+      "grad_norm": 0.47022722324174265,
+      "learning_rate": 1.9642378753177573e-05,
+      "loss": 0.0931,
+      "step": 223
+    },
+    {
+      "epoch": 0.3446153846153846,
+      "grad_norm": 0.43452426712372355,
+      "learning_rate": 1.963916924389962e-05,
+      "loss": 0.0918,
+      "step": 224
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.4951463295701747,
+      "learning_rate": 1.9635945661430006e-05,
+      "loss": 0.0958,
+      "step": 225
+    },
+    {
+      "epoch": 0.3476923076923077,
+      "grad_norm": 0.5231519816188233,
+      "learning_rate": 1.9632708010475166e-05,
+      "loss": 0.0985,
+      "step": 226
+    },
+    {
+      "epoch": 0.34923076923076923,
+      "grad_norm": 0.4125636678302975,
+      "learning_rate": 1.9629456295762067e-05,
+      "loss": 0.0922,
+      "step": 227
+    },
+    {
+      "epoch": 0.3507692307692308,
+      "grad_norm": 0.45904306816999607,
+      "learning_rate": 1.962619052203822e-05,
+      "loss": 0.0992,
+      "step": 228
+    },
+    {
+      "epoch": 0.3523076923076923,
+      "grad_norm": 0.4970133845701183,
+      "learning_rate": 1.9622910694071654e-05,
+      "loss": 0.0932,
+      "step": 229
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 0.37664569358821015,
+      "learning_rate": 1.961961681665092e-05,
+      "loss": 0.0906,
+      "step": 230
+    },
+    {
+      "epoch": 0.3553846153846154,
+      "grad_norm": 0.4876745281374299,
+      "learning_rate": 1.9616308894585078e-05,
+      "loss": 0.0994,
+      "step": 231
+    },
+    {
+      "epoch": 0.3569230769230769,
+      "grad_norm": 0.4680368478168109,
+      "learning_rate": 1.9612986932703698e-05,
+      "loss": 0.1076,
+      "step": 232
+    },
+    {
+      "epoch": 0.35846153846153844,
+      "grad_norm": 0.4857848795913984,
+      "learning_rate": 1.9609650935856847e-05,
+      "loss": 0.1032,
+      "step": 233
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7291002862027431,
+      "learning_rate": 1.9606300908915076e-05,
+      "loss": 0.1071,
+      "step": 234
+    },
+    {
+      "epoch": 0.36153846153846153,
+      "grad_norm": 0.6588220126790021,
+      "learning_rate": 1.9602936856769432e-05,
+      "loss": 0.0943,
+      "step": 235
+    },
+    {
+      "epoch": 0.3630769230769231,
+      "grad_norm": 0.5848274657760204,
+      "learning_rate": 1.959955878433143e-05,
+      "loss": 0.1132,
+      "step": 236
+    },
+    {
+      "epoch": 0.3646153846153846,
+      "grad_norm": 0.49893024052580587,
+      "learning_rate": 1.9596166696533062e-05,
+      "loss": 0.1029,
+      "step": 237
+    },
+    {
+      "epoch": 0.36615384615384616,
+      "grad_norm": 0.5662642891333689,
+      "learning_rate": 1.959276059832677e-05,
+      "loss": 0.1079,
+      "step": 238
+    },
+    {
+      "epoch": 0.3676923076923077,
+      "grad_norm": 0.5411680818094676,
+      "learning_rate": 1.9589340494685464e-05,
+      "loss": 0.0942,
+      "step": 239
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 0.49716800756716995,
+      "learning_rate": 1.95859063906025e-05,
+      "loss": 0.1028,
+      "step": 240
+    },
+    {
+      "epoch": 0.3707692307692308,
+      "grad_norm": 0.5605318150208776,
+      "learning_rate": 1.9582458291091664e-05,
+      "loss": 0.0976,
+      "step": 241
+    },
+    {
+      "epoch": 0.3723076923076923,
+      "grad_norm": 0.6030924939164457,
+      "learning_rate": 1.9578996201187187e-05,
+      "loss": 0.1022,
+      "step": 242
+    },
+    {
+      "epoch": 0.37384615384615383,
+      "grad_norm": 0.4416558734944546,
+      "learning_rate": 1.957552012594372e-05,
+      "loss": 0.0903,
+      "step": 243
+    },
+    {
+      "epoch": 0.37538461538461537,
+      "grad_norm": 0.5329784217079914,
+      "learning_rate": 1.957203007043634e-05,
+      "loss": 0.1164,
+      "step": 244
+    },
+    {
+      "epoch": 0.3769230769230769,
+      "grad_norm": 0.4644145144804026,
+      "learning_rate": 1.956852603976052e-05,
+      "loss": 0.0995,
+      "step": 245
+    },
+    {
+      "epoch": 0.37846153846153846,
+      "grad_norm": 0.43308288576075765,
+      "learning_rate": 1.9565008039032158e-05,
+      "loss": 0.0866,
+      "step": 246
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.556662892440361,
+      "learning_rate": 1.9561476073387527e-05,
+      "loss": 0.0977,
+      "step": 247
+    },
+    {
+      "epoch": 0.38153846153846155,
+      "grad_norm": 0.5153442689897639,
+      "learning_rate": 1.9557930147983303e-05,
+      "loss": 0.0979,
+      "step": 248
+    },
+    {
+      "epoch": 0.3830769230769231,
+      "grad_norm": 0.4731621488928602,
+      "learning_rate": 1.9554370267996537e-05,
+      "loss": 0.1052,
+      "step": 249
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.4461201241494072,
+      "learning_rate": 1.9550796438624656e-05,
+      "loss": 0.0947,
+      "step": 250
+    },
+    {
+      "epoch": 0.3861538461538462,
+      "grad_norm": 0.47389146156943096,
+      "learning_rate": 1.954720866508546e-05,
+      "loss": 0.0995,
+      "step": 251
+    },
+    {
+      "epoch": 0.38769230769230767,
+      "grad_norm": 0.54566585639894,
+      "learning_rate": 1.9543606952617088e-05,
+      "loss": 0.1016,
+      "step": 252
+    },
+    {
+      "epoch": 0.3892307692307692,
+      "grad_norm": 0.582991146196512,
+      "learning_rate": 1.9539991306478046e-05,
+      "loss": 0.1188,
+      "step": 253
+    },
+    {
+      "epoch": 0.39076923076923076,
+      "grad_norm": 0.4578903975000862,
+      "learning_rate": 1.9536361731947182e-05,
+      "loss": 0.0928,
+      "step": 254
+    },
+    {
+      "epoch": 0.3923076923076923,
+      "grad_norm": 0.590776299920465,
+      "learning_rate": 1.953271823432367e-05,
+      "loss": 0.1132,
+      "step": 255
+    },
+    {
+      "epoch": 0.39384615384615385,
+      "grad_norm": 0.6253009080591825,
+      "learning_rate": 1.9529060818927032e-05,
+      "loss": 0.0979,
+      "step": 256
+    },
+    {
+      "epoch": 0.3953846153846154,
+      "grad_norm": 0.531397871124118,
+      "learning_rate": 1.952538949109708e-05,
+      "loss": 0.0959,
+      "step": 257
+    },
+    {
+      "epoch": 0.39692307692307693,
+      "grad_norm": 0.46705935199941867,
+      "learning_rate": 1.9521704256193962e-05,
+      "loss": 0.104,
+      "step": 258
+    },
+    {
+      "epoch": 0.3984615384615385,
+      "grad_norm": 0.5705361981014778,
+      "learning_rate": 1.9518005119598124e-05,
+      "loss": 0.1083,
+      "step": 259
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5695832340146562,
+      "learning_rate": 1.9514292086710307e-05,
+      "loss": 0.0992,
+      "step": 260
+    },
+    {
+      "epoch": 0.4015384615384615,
+      "grad_norm": 0.4464529106040974,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.0987,
+      "step": 261
+    },
+    {
+      "epoch": 0.40307692307692305,
+      "grad_norm": 0.43620342439803284,
+      "learning_rate": 1.9506824353763127e-05,
+      "loss": 0.103,
+      "step": 262
+    },
+    {
+      "epoch": 0.4046153846153846,
+      "grad_norm": 0.48936436282657725,
+      "learning_rate": 1.9503069664606663e-05,
+      "loss": 0.1125,
+      "step": 263
+    },
+    {
+      "epoch": 0.40615384615384614,
+      "grad_norm": 0.5935399529172605,
+      "learning_rate": 1.9499301100963987e-05,
+      "loss": 0.1065,
+      "step": 264
+    },
+    {
+      "epoch": 0.4076923076923077,
+      "grad_norm": 0.5489632101646817,
+      "learning_rate": 1.9495518668337204e-05,
+      "loss": 0.0874,
+      "step": 265
+    },
+    {
+      "epoch": 0.40923076923076923,
+      "grad_norm": 0.4872571558922733,
+      "learning_rate": 1.949172237224867e-05,
+      "loss": 0.1067,
+      "step": 266
+    },
+    {
+      "epoch": 0.4107692307692308,
+      "grad_norm": 0.495735492589365,
+      "learning_rate": 1.9487912218240983e-05,
+      "loss": 0.1025,
+      "step": 267
+    },
+    {
+      "epoch": 0.4123076923076923,
+      "grad_norm": 0.520835006964255,
+      "learning_rate": 1.9484088211876963e-05,
+      "loss": 0.0926,
+      "step": 268
+    },
+    {
+      "epoch": 0.41384615384615386,
+      "grad_norm": 0.4979596342187618,
+      "learning_rate": 1.9480250358739667e-05,
+      "loss": 0.1001,
+      "step": 269
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 0.5398277717258808,
+      "learning_rate": 1.9476398664432356e-05,
+      "loss": 0.1038,
+      "step": 270
+    },
+    {
+      "epoch": 0.4169230769230769,
+      "grad_norm": 0.4620486050085766,
+      "learning_rate": 1.947253313457851e-05,
+      "loss": 0.0909,
+      "step": 271
+    },
+    {
+      "epoch": 0.41846153846153844,
+      "grad_norm": 0.5088637188704777,
+      "learning_rate": 1.9468653774821803e-05,
+      "loss": 0.1038,
+      "step": 272
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5312467309003265,
+      "learning_rate": 1.94647605908261e-05,
+      "loss": 0.1003,
+      "step": 273
+    },
+    {
+      "epoch": 0.42153846153846153,
+      "grad_norm": 0.5353118330387302,
+      "learning_rate": 1.9460853588275454e-05,
+      "loss": 0.0915,
+      "step": 274
+    },
+    {
+      "epoch": 0.4230769230769231,
+      "grad_norm": 0.5883396050759481,
+      "learning_rate": 1.9456932772874092e-05,
+      "loss": 0.0964,
+      "step": 275
+    },
+    {
+      "epoch": 0.4246153846153846,
+      "grad_norm": 0.3782557300574429,
+      "learning_rate": 1.9452998150346403e-05,
+      "loss": 0.0941,
+      "step": 276
+    },
+    {
+      "epoch": 0.42615384615384616,
+      "grad_norm": 0.5038727416567336,
+      "learning_rate": 1.944904972643694e-05,
+      "loss": 0.0966,
+      "step": 277
+    },
+    {
+      "epoch": 0.4276923076923077,
+      "grad_norm": 0.6044080637491093,
+      "learning_rate": 1.9445087506910403e-05,
+      "loss": 0.0945,
+      "step": 278
+    },
+    {
+      "epoch": 0.42923076923076925,
+      "grad_norm": 0.4283489498603955,
+      "learning_rate": 1.944111149755164e-05,
+      "loss": 0.0923,
+      "step": 279
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 0.4755286664345765,
+      "learning_rate": 1.9437121704165612e-05,
+      "loss": 0.0981,
+      "step": 280
+    },
+    {
+      "epoch": 0.4323076923076923,
+      "grad_norm": 0.4034836375835124,
+      "learning_rate": 1.9433118132577432e-05,
+      "loss": 0.0956,
+      "step": 281
+    },
+    {
+      "epoch": 0.4338461538461538,
+      "grad_norm": 0.5243724118676012,
+      "learning_rate": 1.9429100788632313e-05,
+      "loss": 0.0968,
+      "step": 282
+    },
+    {
+      "epoch": 0.43538461538461537,
+      "grad_norm": 0.4909950952163068,
+      "learning_rate": 1.9425069678195577e-05,
+      "loss": 0.0836,
+      "step": 283
+    },
+    {
+      "epoch": 0.4369230769230769,
+      "grad_norm": 0.6197661444062966,
+      "learning_rate": 1.9421024807152652e-05,
+      "loss": 0.1062,
+      "step": 284
+    },
+    {
+      "epoch": 0.43846153846153846,
+      "grad_norm": 0.4641188532073514,
+      "learning_rate": 1.9416966181409047e-05,
+      "loss": 0.0907,
+      "step": 285
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.581115105716189,
+      "learning_rate": 1.9412893806890358e-05,
+      "loss": 0.1136,
+      "step": 286
+    },
+    {
+      "epoch": 0.44153846153846155,
+      "grad_norm": 0.5532740369160736,
+      "learning_rate": 1.9408807689542257e-05,
+      "loss": 0.095,
+      "step": 287
+    },
+    {
+      "epoch": 0.4430769230769231,
+      "grad_norm": 0.610460188491936,
+      "learning_rate": 1.9404707835330475e-05,
+      "loss": 0.102,
+      "step": 288
+    },
+    {
+      "epoch": 0.44461538461538463,
+      "grad_norm": 0.48489275085594646,
+      "learning_rate": 1.94005942502408e-05,
+      "loss": 0.1086,
+      "step": 289
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 0.4350745012942025,
+      "learning_rate": 1.9396466940279067e-05,
+      "loss": 0.1027,
+      "step": 290
+    },
+    {
+      "epoch": 0.44769230769230767,
+      "grad_norm": 0.4950020730573976,
+      "learning_rate": 1.9392325911471154e-05,
+      "loss": 0.1055,
+      "step": 291
+    },
+    {
+      "epoch": 0.4492307692307692,
+      "grad_norm": 0.4788608653907359,
+      "learning_rate": 1.9388171169862967e-05,
+      "loss": 0.1056,
+      "step": 292
+    },
+    {
+      "epoch": 0.45076923076923076,
+      "grad_norm": 0.451391245157289,
+      "learning_rate": 1.9384002721520423e-05,
+      "loss": 0.1034,
+      "step": 293
+    },
+    {
+      "epoch": 0.4523076923076923,
+      "grad_norm": 0.46706566593370247,
+      "learning_rate": 1.9379820572529463e-05,
+      "loss": 0.1032,
+      "step": 294
+    },
+    {
+      "epoch": 0.45384615384615384,
+      "grad_norm": 0.5208831601136724,
+      "learning_rate": 1.937562472899603e-05,
+      "loss": 0.1048,
+      "step": 295
+    },
+    {
+      "epoch": 0.4553846153846154,
+      "grad_norm": 0.4746886544547485,
+      "learning_rate": 1.9371415197046054e-05,
+      "loss": 0.0962,
+      "step": 296
+    },
+    {
+      "epoch": 0.45692307692307693,
+      "grad_norm": 0.5099090103562837,
+      "learning_rate": 1.936719198282545e-05,
+      "loss": 0.0975,
+      "step": 297
+    },
+    {
+      "epoch": 0.4584615384615385,
+      "grad_norm": 0.5371114216801979,
+      "learning_rate": 1.936295509250012e-05,
+      "loss": 0.0995,
+      "step": 298
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4583000967152501,
+      "learning_rate": 1.935870453225592e-05,
+      "loss": 0.0867,
+      "step": 299
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.4824466157164947,
+      "learning_rate": 1.9354440308298676e-05,
+      "loss": 0.1027,
+      "step": 300
+    },
+    {
+      "epoch": 0.46307692307692305,
+      "grad_norm": 0.3731332197286788,
+      "learning_rate": 1.9350162426854152e-05,
+      "loss": 0.0937,
+      "step": 301
+    },
+    {
+      "epoch": 0.4646153846153846,
+      "grad_norm": 0.5904943219369095,
+      "learning_rate": 1.9345870894168056e-05,
+      "loss": 0.1018,
+      "step": 302
+    },
+    {
+      "epoch": 0.46615384615384614,
+      "grad_norm": 0.4937106647701358,
+      "learning_rate": 1.934156571650603e-05,
+      "loss": 0.099,
+      "step": 303
+    },
+    {
+      "epoch": 0.4676923076923077,
+      "grad_norm": 0.3616397861267969,
+      "learning_rate": 1.9337246900153637e-05,
+      "loss": 0.0899,
+      "step": 304
+    },
+    {
+      "epoch": 0.46923076923076923,
+      "grad_norm": 0.4337547379197528,
+      "learning_rate": 1.933291445141635e-05,
+      "loss": 0.1094,
+      "step": 305
+    },
+    {
+      "epoch": 0.4707692307692308,
+      "grad_norm": 0.47132758397572894,
+      "learning_rate": 1.932856837661954e-05,
+      "loss": 0.0932,
+      "step": 306
+    },
+    {
+      "epoch": 0.4723076923076923,
+      "grad_norm": 0.5249116456685259,
+      "learning_rate": 1.9324208682108493e-05,
+      "loss": 0.0955,
+      "step": 307
+    },
+    {
+      "epoch": 0.47384615384615386,
+      "grad_norm": 0.4099876886395493,
+      "learning_rate": 1.931983537424835e-05,
+      "loss": 0.0925,
+      "step": 308
+    },
+    {
+      "epoch": 0.4753846153846154,
+      "grad_norm": 0.4571620412260989,
+      "learning_rate": 1.931544845942415e-05,
+      "loss": 0.0974,
+      "step": 309
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 0.5347160775026656,
+      "learning_rate": 1.9311047944040792e-05,
+      "loss": 0.1021,
+      "step": 310
+    },
+    {
+      "epoch": 0.47846153846153844,
+      "grad_norm": 0.45731698473766913,
+      "learning_rate": 1.9306633834523022e-05,
+      "loss": 0.0995,
+      "step": 311
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4463396489272035,
+      "learning_rate": 1.930220613731545e-05,
+      "loss": 0.0936,
+      "step": 312
+    },
+    {
+      "epoch": 0.4815384615384615,
+      "grad_norm": 0.449418485015138,
+      "learning_rate": 1.9297764858882516e-05,
+      "loss": 0.0883,
+      "step": 313
+    },
+    {
+      "epoch": 0.48307692307692307,
+      "grad_norm": 0.5562810223949127,
+      "learning_rate": 1.9293310005708485e-05,
+      "loss": 0.1089,
+      "step": 314
+    },
+    {
+      "epoch": 0.4846153846153846,
+      "grad_norm": 0.40867491037166837,
+      "learning_rate": 1.9288841584297445e-05,
+      "loss": 0.087,
+      "step": 315
+    },
+    {
+      "epoch": 0.48615384615384616,
+      "grad_norm": 0.6068014941268159,
+      "learning_rate": 1.9284359601173295e-05,
+      "loss": 0.1002,
+      "step": 316
+    },
+    {
+      "epoch": 0.4876923076923077,
+      "grad_norm": 0.5485840263407294,
+      "learning_rate": 1.927986406287973e-05,
+      "loss": 0.1071,
+      "step": 317
+    },
+    {
+      "epoch": 0.48923076923076925,
+      "grad_norm": 0.4326599983402119,
+      "learning_rate": 1.9275354975980245e-05,
+      "loss": 0.0972,
+      "step": 318
+    },
+    {
+      "epoch": 0.4907692307692308,
+      "grad_norm": 0.48038555828580165,
+      "learning_rate": 1.92708323470581e-05,
+      "loss": 0.0962,
+      "step": 319
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 0.5096604190301004,
+      "learning_rate": 1.926629618271634e-05,
+      "loss": 0.1089,
+      "step": 320
+    },
+    {
+      "epoch": 0.4938461538461538,
+      "grad_norm": 0.5074786000084778,
+      "learning_rate": 1.9261746489577767e-05,
+      "loss": 0.0953,
+      "step": 321
+    },
+    {
+      "epoch": 0.49538461538461537,
+      "grad_norm": 0.3551898568112085,
+      "learning_rate": 1.9257183274284934e-05,
+      "loss": 0.0812,
+      "step": 322
+    },
+    {
+      "epoch": 0.4969230769230769,
+      "grad_norm": 0.4154869588023741,
+      "learning_rate": 1.925260654350014e-05,
+      "loss": 0.0878,
+      "step": 323
+    },
+    {
+      "epoch": 0.49846153846153846,
+      "grad_norm": 0.46068163561485903,
+      "learning_rate": 1.924801630390541e-05,
+      "loss": 0.0994,
+      "step": 324
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49072829349035685,
+      "learning_rate": 1.92434125622025e-05,
+      "loss": 0.1022,
+      "step": 325
+    },
+    {
+      "epoch": 0.5015384615384615,
+      "grad_norm": 0.43272682508224936,
+      "learning_rate": 1.9238795325112867e-05,
+      "loss": 0.1084,
+      "step": 326
+    },
+    {
+      "epoch": 0.5030769230769231,
+      "grad_norm": 0.4497283552516036,
+      "learning_rate": 1.9234164599377692e-05,
+      "loss": 0.0967,
+      "step": 327
+    },
+    {
+      "epoch": 0.5046153846153846,
+      "grad_norm": 0.4492106413880127,
+      "learning_rate": 1.9229520391757828e-05,
+      "loss": 0.1007,
+      "step": 328
+    },
+    {
+      "epoch": 0.5061538461538462,
+      "grad_norm": 0.4890385196278267,
+      "learning_rate": 1.9224862709033823e-05,
+      "loss": 0.09,
+      "step": 329
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 0.4508644606981008,
+      "learning_rate": 1.9220191558005897e-05,
+      "loss": 0.0969,
+      "step": 330
+    },
+    {
+      "epoch": 0.5092307692307693,
+      "grad_norm": 0.5761631636154984,
+      "learning_rate": 1.9215506945493933e-05,
+      "loss": 0.1051,
+      "step": 331
+    },
+    {
+      "epoch": 0.5107692307692308,
+      "grad_norm": 0.4285970063917521,
+      "learning_rate": 1.921080887833746e-05,
+      "loss": 0.0952,
+      "step": 332
+    },
+    {
+      "epoch": 0.5123076923076924,
+      "grad_norm": 0.46226415222748873,
+      "learning_rate": 1.9206097363395668e-05,
+      "loss": 0.099,
+      "step": 333
+    },
+    {
+      "epoch": 0.5138461538461538,
+      "grad_norm": 0.44278457207986294,
+      "learning_rate": 1.9201372407547367e-05,
+      "loss": 0.1025,
+      "step": 334
+    },
+    {
+      "epoch": 0.5153846153846153,
+      "grad_norm": 0.5672748490771847,
+      "learning_rate": 1.9196634017690993e-05,
+      "loss": 0.1013,
+      "step": 335
+    },
+    {
+      "epoch": 0.5169230769230769,
+      "grad_norm": 0.4566472958118853,
+      "learning_rate": 1.9191882200744602e-05,
+      "loss": 0.1158,
+      "step": 336
+    },
+    {
+      "epoch": 0.5184615384615384,
+      "grad_norm": 0.5248322529451815,
+      "learning_rate": 1.9187116963645845e-05,
+      "loss": 0.0957,
+      "step": 337
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.44219344780447695,
+      "learning_rate": 1.918233831335197e-05,
+      "loss": 0.0975,
+      "step": 338
+    },
+    {
+      "epoch": 0.5215384615384615,
+      "grad_norm": 0.378145742243331,
+      "learning_rate": 1.9177546256839814e-05,
+      "loss": 0.0855,
+      "step": 339
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 0.4543314495704193,
+      "learning_rate": 1.9172740801105777e-05,
+      "loss": 0.101,
+      "step": 340
+    },
+    {
+      "epoch": 0.5246153846153846,
+      "grad_norm": 0.5388495866373283,
+      "learning_rate": 1.9167921953165827e-05,
+      "loss": 0.1077,
+      "step": 341
+    },
+    {
+      "epoch": 0.5261538461538462,
+      "grad_norm": 0.4931391711463696,
+      "learning_rate": 1.9163089720055484e-05,
+      "loss": 0.0937,
+      "step": 342
+    },
+    {
+      "epoch": 0.5276923076923077,
+      "grad_norm": 0.5226419648010249,
+      "learning_rate": 1.9158244108829815e-05,
+      "loss": 0.0978,
+      "step": 343
+    },
+    {
+      "epoch": 0.5292307692307693,
+      "grad_norm": 0.514073044628901,
+      "learning_rate": 1.915338512656341e-05,
+      "loss": 0.1116,
+      "step": 344
+    },
+    {
+      "epoch": 0.5307692307692308,
+      "grad_norm": 0.536463160406073,
+      "learning_rate": 1.9148512780350384e-05,
+      "loss": 0.09,
+      "step": 345
+    },
+    {
+      "epoch": 0.5323076923076923,
+      "grad_norm": 0.40472290941801015,
+      "learning_rate": 1.914362707730437e-05,
+      "loss": 0.096,
+      "step": 346
+    },
+    {
+      "epoch": 0.5338461538461539,
+      "grad_norm": 0.4419558495038164,
+      "learning_rate": 1.9138728024558494e-05,
+      "loss": 0.1043,
+      "step": 347
+    },
+    {
+      "epoch": 0.5353846153846153,
+      "grad_norm": 0.45250393937417616,
+      "learning_rate": 1.913381562926538e-05,
+      "loss": 0.1044,
+      "step": 348
+    },
+    {
+      "epoch": 0.536923076923077,
+      "grad_norm": 0.4970429188190532,
+      "learning_rate": 1.9128889898597117e-05,
+      "loss": 0.1002,
+      "step": 349
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.5382027157131241,
+      "learning_rate": 1.912395083974528e-05,
+      "loss": 0.0992,
+      "step": 350
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4173235180068536,
+      "learning_rate": 1.91189984599209e-05,
+      "loss": 0.0873,
+      "step": 351
+    },
+    {
+      "epoch": 0.5415384615384615,
+      "grad_norm": 0.42760297701096356,
+      "learning_rate": 1.9114032766354453e-05,
+      "loss": 0.096,
+      "step": 352
+    },
+    {
+      "epoch": 0.5430769230769231,
+      "grad_norm": 0.524358521462744,
+      "learning_rate": 1.910905376629585e-05,
+      "loss": 0.0959,
+      "step": 353
+    },
+    {
+      "epoch": 0.5446153846153846,
+      "grad_norm": 0.47250449647238696,
+      "learning_rate": 1.910406146701444e-05,
+      "loss": 0.0849,
+      "step": 354
+    },
+    {
+      "epoch": 0.5461538461538461,
+      "grad_norm": 0.44398596927938827,
+      "learning_rate": 1.9099055875798974e-05,
+      "loss": 0.1024,
+      "step": 355
+    },
+    {
+      "epoch": 0.5476923076923077,
+      "grad_norm": 0.46499606497240575,
+      "learning_rate": 1.9094036999957623e-05,
+      "loss": 0.0925,
+      "step": 356
+    },
+    {
+      "epoch": 0.5492307692307692,
+      "grad_norm": 0.3886634015877156,
+      "learning_rate": 1.9089004846817947e-05,
+      "loss": 0.0807,
+      "step": 357
+    },
+    {
+      "epoch": 0.5507692307692308,
+      "grad_norm": 0.4118047391602366,
+      "learning_rate": 1.908395942372689e-05,
+      "loss": 0.0988,
+      "step": 358
+    },
+    {
+      "epoch": 0.5523076923076923,
+      "grad_norm": 0.434600405160762,
+      "learning_rate": 1.9078900738050776e-05,
+      "loss": 0.1069,
+      "step": 359
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 0.4148807319841889,
+      "learning_rate": 1.9073828797175284e-05,
+      "loss": 0.1028,
+      "step": 360
+    },
+    {
+      "epoch": 0.5553846153846154,
+      "grad_norm": 0.42028520435641853,
+      "learning_rate": 1.9068743608505454e-05,
+      "loss": 0.1056,
+      "step": 361
+    },
+    {
+      "epoch": 0.556923076923077,
+      "grad_norm": 0.43510420212696854,
+      "learning_rate": 1.9063645179465663e-05,
+      "loss": 0.1009,
+      "step": 362
+    },
+    {
+      "epoch": 0.5584615384615385,
+      "grad_norm": 0.4936074681803371,
+      "learning_rate": 1.905853351749962e-05,
+      "loss": 0.1,
+      "step": 363
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.3632144105495107,
+      "learning_rate": 1.9053408630070352e-05,
+      "loss": 0.0916,
+      "step": 364
+    },
+    {
+      "epoch": 0.5615384615384615,
+      "grad_norm": 0.3536771592078361,
+      "learning_rate": 1.9048270524660197e-05,
+      "loss": 0.0913,
+      "step": 365
+    },
+    {
+      "epoch": 0.563076923076923,
+      "grad_norm": 0.483138657389253,
+      "learning_rate": 1.904311920877079e-05,
+      "loss": 0.0992,
+      "step": 366
+    },
+    {
+      "epoch": 0.5646153846153846,
+      "grad_norm": 0.4057661988297129,
+      "learning_rate": 1.903795468992306e-05,
+      "loss": 0.1027,
+      "step": 367
+    },
+    {
+      "epoch": 0.5661538461538461,
+      "grad_norm": 0.40666719198061757,
+      "learning_rate": 1.9032776975657207e-05,
+      "loss": 0.0981,
+      "step": 368
+    },
+    {
+      "epoch": 0.5676923076923077,
+      "grad_norm": 0.44070081776133335,
+      "learning_rate": 1.902758607353269e-05,
+      "loss": 0.1104,
+      "step": 369
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 0.5047821437081584,
+      "learning_rate": 1.9022381991128235e-05,
+      "loss": 0.0963,
+      "step": 370
+    },
+    {
+      "epoch": 0.5707692307692308,
+      "grad_norm": 0.42260385596144445,
+      "learning_rate": 1.9017164736041795e-05,
+      "loss": 0.1029,
+      "step": 371
+    },
+    {
+      "epoch": 0.5723076923076923,
+      "grad_norm": 0.4244718881899699,
+      "learning_rate": 1.9011934315890576e-05,
+      "loss": 0.0966,
+      "step": 372
+    },
+    {
+      "epoch": 0.5738461538461539,
+      "grad_norm": 0.4953235016748773,
+      "learning_rate": 1.9006690738310988e-05,
+      "loss": 0.1037,
+      "step": 373
+    },
+    {
+      "epoch": 0.5753846153846154,
+      "grad_norm": 0.3770254505903871,
+      "learning_rate": 1.900143401095866e-05,
+      "loss": 0.086,
+      "step": 374
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.41771963874965723,
+      "learning_rate": 1.8996164141508412e-05,
+      "loss": 0.0858,
+      "step": 375
+    },
+    {
+      "epoch": 0.5784615384615385,
+      "grad_norm": 0.4061141404696647,
+      "learning_rate": 1.899088113765426e-05,
+      "loss": 0.1021,
+      "step": 376
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.45946326931814974,
+      "learning_rate": 1.898558500710939e-05,
+      "loss": 0.0979,
+      "step": 377
+    },
+    {
+      "epoch": 0.5815384615384616,
+      "grad_norm": 0.4143960427519077,
+      "learning_rate": 1.8980275757606157e-05,
+      "loss": 0.0965,
+      "step": 378
+    },
+    {
+      "epoch": 0.583076923076923,
+      "grad_norm": 0.40587432220084,
+      "learning_rate": 1.8974953396896066e-05,
+      "loss": 0.0953,
+      "step": 379
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 0.4036824003812063,
+      "learning_rate": 1.8969617932749766e-05,
+      "loss": 0.0947,
+      "step": 380
+    },
+    {
+      "epoch": 0.5861538461538461,
+      "grad_norm": 0.431070678494207,
+      "learning_rate": 1.896426937295704e-05,
+      "loss": 0.103,
+      "step": 381
+    },
+    {
+      "epoch": 0.5876923076923077,
+      "grad_norm": 0.3955536485889764,
+      "learning_rate": 1.8958907725326783e-05,
+      "loss": 0.0928,
+      "step": 382
+    },
+    {
+      "epoch": 0.5892307692307692,
+      "grad_norm": 0.4357741837489556,
+      "learning_rate": 1.8953532997687008e-05,
+      "loss": 0.0997,
+      "step": 383
+    },
+    {
+      "epoch": 0.5907692307692308,
+      "grad_norm": 0.39754441486925746,
+      "learning_rate": 1.8948145197884815e-05,
+      "loss": 0.0859,
+      "step": 384
+    },
+    {
+      "epoch": 0.5923076923076923,
+      "grad_norm": 0.43549014378990397,
+      "learning_rate": 1.89427443337864e-05,
+      "loss": 0.0934,
+      "step": 385
+    },
+    {
+      "epoch": 0.5938461538461538,
+      "grad_norm": 0.42300645280983395,
+      "learning_rate": 1.893733041327702e-05,
+      "loss": 0.0951,
+      "step": 386
+    },
+    {
+      "epoch": 0.5953846153846154,
+      "grad_norm": 0.41282699722561766,
+      "learning_rate": 1.8931903444261007e-05,
+      "loss": 0.0941,
+      "step": 387
+    },
+    {
+      "epoch": 0.5969230769230769,
+      "grad_norm": 0.4480032982295742,
+      "learning_rate": 1.8926463434661738e-05,
+      "loss": 0.0905,
+      "step": 388
+    },
+    {
+      "epoch": 0.5984615384615385,
+      "grad_norm": 0.44297004392731376,
+      "learning_rate": 1.8921010392421628e-05,
+      "loss": 0.1016,
+      "step": 389
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5433110545398998,
+      "learning_rate": 1.8915544325502123e-05,
+      "loss": 0.0935,
+      "step": 390
+    },
+    {
+      "epoch": 0.6015384615384616,
+      "grad_norm": 0.40571632988210427,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.0937,
+      "step": 391
+    },
+    {
+      "epoch": 0.6030769230769231,
+      "grad_norm": 0.43732620714659554,
+      "learning_rate": 1.8904573149565766e-05,
+      "loss": 0.0947,
+      "step": 392
+    },
+    {
+      "epoch": 0.6046153846153847,
+      "grad_norm": 0.4600862732120384,
+      "learning_rate": 1.889906805656684e-05,
+      "loss": 0.095,
+      "step": 393
+    },
+    {
+      "epoch": 0.6061538461538462,
+      "grad_norm": 0.44446566484906674,
+      "learning_rate": 1.8893549970924335e-05,
+      "loss": 0.1005,
+      "step": 394
+    },
+    {
+      "epoch": 0.6076923076923076,
+      "grad_norm": 0.4372133849372649,
+      "learning_rate": 1.888801890069467e-05,
+      "loss": 0.0926,
+      "step": 395
+    },
+    {
+      "epoch": 0.6092307692307692,
+      "grad_norm": 0.4496197566055402,
+      "learning_rate": 1.8882474853953193e-05,
+      "loss": 0.0889,
+      "step": 396
+    },
+    {
+      "epoch": 0.6107692307692307,
+      "grad_norm": 0.43381954647371906,
+      "learning_rate": 1.8876917838794226e-05,
+      "loss": 0.0892,
+      "step": 397
+    },
+    {
+      "epoch": 0.6123076923076923,
+      "grad_norm": 0.5656036739453132,
+      "learning_rate": 1.8871347863331015e-05,
+      "loss": 0.087,
+      "step": 398
+    },
+    {
+      "epoch": 0.6138461538461538,
+      "grad_norm": 0.42869081240488516,
+      "learning_rate": 1.886576493569572e-05,
+      "loss": 0.0999,
+      "step": 399
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.9645884517575197,
+      "learning_rate": 1.8860169064039422e-05,
+      "loss": 0.0877,
+      "step": 400
+    },
+    {
+      "epoch": 0.6169230769230769,
+      "grad_norm": 0.3909376825178073,
+      "learning_rate": 1.8854560256532098e-05,
+      "loss": 0.09,
+      "step": 401
+    },
+    {
+      "epoch": 0.6184615384615385,
+      "grad_norm": 0.41607837043434165,
+      "learning_rate": 1.884893852136261e-05,
+      "loss": 0.0962,
+      "step": 402
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.4118435975798478,
+      "learning_rate": 1.884330386673869e-05,
+      "loss": 0.0914,
+      "step": 403
+    },
+    {
+      "epoch": 0.6215384615384615,
+      "grad_norm": 0.41468438143981523,
+      "learning_rate": 1.8837656300886937e-05,
+      "loss": 0.0932,
+      "step": 404
+    },
+    {
+      "epoch": 0.6230769230769231,
+      "grad_norm": 0.42739734031323684,
+      "learning_rate": 1.8831995832052802e-05,
+      "loss": 0.1006,
+      "step": 405
+    },
+    {
+      "epoch": 0.6246153846153846,
+      "grad_norm": 0.40676045777564296,
+      "learning_rate": 1.8826322468500567e-05,
+      "loss": 0.0911,
+      "step": 406
+    },
+    {
+      "epoch": 0.6261538461538462,
+      "grad_norm": 0.4400751065379498,
+      "learning_rate": 1.8820636218513354e-05,
+      "loss": 0.099,
+      "step": 407
+    },
+    {
+      "epoch": 0.6276923076923077,
+      "grad_norm": 0.42437410068559034,
+      "learning_rate": 1.8814937090393082e-05,
+      "loss": 0.0968,
+      "step": 408
+    },
+    {
+      "epoch": 0.6292307692307693,
+      "grad_norm": 0.3919701002838343,
+      "learning_rate": 1.8809225092460488e-05,
+      "loss": 0.0936,
+      "step": 409
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 0.5028949538298951,
+      "learning_rate": 1.880350023305509e-05,
+      "loss": 0.0948,
+      "step": 410
+    },
+    {
+      "epoch": 0.6323076923076923,
+      "grad_norm": 0.4075578032238594,
+      "learning_rate": 1.8797762520535178e-05,
+      "loss": 0.0981,
+      "step": 411
+    },
+    {
+      "epoch": 0.6338461538461538,
+      "grad_norm": 0.34153544828671295,
+      "learning_rate": 1.8792011963277827e-05,
+      "loss": 0.0887,
+      "step": 412
+    },
+    {
+      "epoch": 0.6353846153846154,
+      "grad_norm": 0.49338561568395417,
+      "learning_rate": 1.8786248569678847e-05,
+      "loss": 0.1008,
+      "step": 413
+    },
+    {
+      "epoch": 0.6369230769230769,
+      "grad_norm": 0.3911057155983341,
+      "learning_rate": 1.8780472348152792e-05,
+      "loss": 0.0957,
+      "step": 414
+    },
+    {
+      "epoch": 0.6384615384615384,
+      "grad_norm": 0.4863828952857313,
+      "learning_rate": 1.8774683307132956e-05,
+      "loss": 0.0962,
+      "step": 415
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4959721632631627,
+      "learning_rate": 1.876888145507133e-05,
+      "loss": 0.0974,
+      "step": 416
+    },
+    {
+      "epoch": 0.6415384615384615,
+      "grad_norm": 0.44751359386684436,
+      "learning_rate": 1.8763066800438638e-05,
+      "loss": 0.1072,
+      "step": 417
+    },
+    {
+      "epoch": 0.6430769230769231,
+      "grad_norm": 0.4168309844234487,
+      "learning_rate": 1.8757239351724262e-05,
+      "loss": 0.0986,
+      "step": 418
+    },
+    {
+      "epoch": 0.6446153846153846,
+      "grad_norm": 0.48992479490511887,
+      "learning_rate": 1.8751399117436292e-05,
+      "loss": 0.1142,
+      "step": 419
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 0.5250869322124576,
+      "learning_rate": 1.8745546106101466e-05,
+      "loss": 0.0966,
+      "step": 420
+    },
+    {
+      "epoch": 0.6476923076923077,
+      "grad_norm": 0.41428994308159467,
+      "learning_rate": 1.873968032626518e-05,
+      "loss": 0.0964,
+      "step": 421
+    },
+    {
+      "epoch": 0.6492307692307693,
+      "grad_norm": 0.40863286499944806,
+      "learning_rate": 1.8733801786491487e-05,
+      "loss": 0.0982,
+      "step": 422
+    },
+    {
+      "epoch": 0.6507692307692308,
+      "grad_norm": 0.40530569813116685,
+      "learning_rate": 1.8727910495363043e-05,
+      "loss": 0.0893,
+      "step": 423
+    },
+    {
+      "epoch": 0.6523076923076923,
+      "grad_norm": 0.3994309755298727,
+      "learning_rate": 1.872200646148115e-05,
+      "loss": 0.0917,
+      "step": 424
+    },
+    {
+      "epoch": 0.6538461538461539,
+      "grad_norm": 0.41310852897533573,
+      "learning_rate": 1.8716089693465696e-05,
+      "loss": 0.0869,
+      "step": 425
+    },
+    {
+      "epoch": 0.6553846153846153,
+      "grad_norm": 0.4477653763330846,
+      "learning_rate": 1.8710160199955158e-05,
+      "loss": 0.1001,
+      "step": 426
+    },
+    {
+      "epoch": 0.6569230769230769,
+      "grad_norm": 0.4749180294295519,
+      "learning_rate": 1.8704217989606606e-05,
+      "loss": 0.0989,
+      "step": 427
+    },
+    {
+      "epoch": 0.6584615384615384,
+      "grad_norm": 0.4374196609501067,
+      "learning_rate": 1.869826307109567e-05,
+      "loss": 0.0955,
+      "step": 428
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.39438024250010495,
+      "learning_rate": 1.869229545311653e-05,
+      "loss": 0.0939,
+      "step": 429
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 0.4203742463904514,
+      "learning_rate": 1.8686315144381914e-05,
+      "loss": 0.0964,
+      "step": 430
+    },
+    {
+      "epoch": 0.6630769230769231,
+      "grad_norm": 0.4761695840193704,
+      "learning_rate": 1.8680322153623077e-05,
+      "loss": 0.1037,
+      "step": 431
+    },
+    {
+      "epoch": 0.6646153846153846,
+      "grad_norm": 0.3677131847676501,
+      "learning_rate": 1.8674316489589782e-05,
+      "loss": 0.09,
+      "step": 432
+    },
+    {
+      "epoch": 0.6661538461538462,
+      "grad_norm": 0.4356656851311741,
+      "learning_rate": 1.8668298161050308e-05,
+      "loss": 0.1005,
+      "step": 433
+    },
+    {
+      "epoch": 0.6676923076923077,
+      "grad_norm": 0.5173595615830835,
+      "learning_rate": 1.8662267176791418e-05,
+      "loss": 0.104,
+      "step": 434
+    },
+    {
+      "epoch": 0.6692307692307692,
+      "grad_norm": 0.565640191679134,
+      "learning_rate": 1.8656223545618345e-05,
+      "loss": 0.1059,
+      "step": 435
+    },
+    {
+      "epoch": 0.6707692307692308,
+      "grad_norm": 0.5304970770470983,
+      "learning_rate": 1.8650167276354802e-05,
+      "loss": 0.0969,
+      "step": 436
+    },
+    {
+      "epoch": 0.6723076923076923,
+      "grad_norm": 0.3582385963482857,
+      "learning_rate": 1.8644098377842934e-05,
+      "loss": 0.0901,
+      "step": 437
+    },
+    {
+      "epoch": 0.6738461538461539,
+      "grad_norm": 0.4778942975736773,
+      "learning_rate": 1.863801685894335e-05,
+      "loss": 0.0997,
+      "step": 438
+    },
+    {
+      "epoch": 0.6753846153846154,
+      "grad_norm": 0.6029016365712476,
+      "learning_rate": 1.8631922728535054e-05,
+      "loss": 0.0938,
+      "step": 439
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 0.6376765327998692,
+      "learning_rate": 1.8625815995515493e-05,
+      "loss": 0.1069,
+      "step": 440
+    },
+    {
+      "epoch": 0.6784615384615384,
+      "grad_norm": 0.4435527425755955,
+      "learning_rate": 1.8619696668800494e-05,
+      "loss": 0.0934,
+      "step": 441
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5412282043268205,
+      "learning_rate": 1.8613564757324276e-05,
+      "loss": 0.099,
+      "step": 442
+    },
+    {
+      "epoch": 0.6815384615384615,
+      "grad_norm": 0.5181149152675278,
+      "learning_rate": 1.860742027003944e-05,
+      "loss": 0.1045,
+      "step": 443
+    },
+    {
+      "epoch": 0.683076923076923,
+      "grad_norm": 0.40925583148990097,
+      "learning_rate": 1.8601263215916928e-05,
+      "loss": 0.0893,
+      "step": 444
+    },
+    {
+      "epoch": 0.6846153846153846,
+      "grad_norm": 0.42471412248636775,
+      "learning_rate": 1.8595093603946053e-05,
+      "loss": 0.0986,
+      "step": 445
+    },
+    {
+      "epoch": 0.6861538461538461,
+      "grad_norm": 0.35148548657598855,
+      "learning_rate": 1.858891144313445e-05,
+      "loss": 0.0911,
+      "step": 446
+    },
+    {
+      "epoch": 0.6876923076923077,
+      "grad_norm": 0.40331605360650485,
+      "learning_rate": 1.8582716742508066e-05,
+      "loss": 0.0967,
+      "step": 447
+    },
+    {
+      "epoch": 0.6892307692307692,
+      "grad_norm": 0.44814576432850384,
+      "learning_rate": 1.8576509511111182e-05,
+      "loss": 0.0917,
+      "step": 448
+    },
+    {
+      "epoch": 0.6907692307692308,
+      "grad_norm": 0.4489452721920411,
+      "learning_rate": 1.8570289758006346e-05,
+      "loss": 0.0989,
+      "step": 449
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.3822084964419905,
+      "learning_rate": 1.8564057492274407e-05,
+      "loss": 0.0923,
+      "step": 450
+    },
+    {
+      "epoch": 0.6938461538461539,
+      "grad_norm": 0.47514026675128,
+      "learning_rate": 1.8557812723014476e-05,
+      "loss": 0.1076,
+      "step": 451
+    },
+    {
+      "epoch": 0.6953846153846154,
+      "grad_norm": 0.5418245481869945,
+      "learning_rate": 1.8551555459343918e-05,
+      "loss": 0.1116,
+      "step": 452
+    },
+    {
+      "epoch": 0.696923076923077,
+      "grad_norm": 0.5023521612267582,
+      "learning_rate": 1.8545285710398343e-05,
+      "loss": 0.1078,
+      "step": 453
+    },
+    {
+      "epoch": 0.6984615384615385,
+      "grad_norm": 0.5662436210170848,
+      "learning_rate": 1.8539003485331584e-05,
+      "loss": 0.0951,
+      "step": 454
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.42546907920042626,
+      "learning_rate": 1.853270879331569e-05,
+      "loss": 0.0942,
+      "step": 455
+    },
+    {
+      "epoch": 0.7015384615384616,
+      "grad_norm": 0.41449390666943065,
+      "learning_rate": 1.8526401643540924e-05,
+      "loss": 0.0994,
+      "step": 456
+    },
+    {
+      "epoch": 0.703076923076923,
+      "grad_norm": 0.48839987100466653,
+      "learning_rate": 1.852008204521572e-05,
+      "loss": 0.1008,
+      "step": 457
+    },
+    {
+      "epoch": 0.7046153846153846,
+      "grad_norm": 0.43738519055543357,
+      "learning_rate": 1.8513750007566696e-05,
+      "loss": 0.0954,
+      "step": 458
+    },
+    {
+      "epoch": 0.7061538461538461,
+      "grad_norm": 0.41775219787174245,
+      "learning_rate": 1.850740553983863e-05,
+      "loss": 0.0875,
+      "step": 459
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 0.7358518540594707,
+      "learning_rate": 1.8501048651294447e-05,
+      "loss": 0.0933,
+      "step": 460
+    },
+    {
+      "epoch": 0.7092307692307692,
+      "grad_norm": 0.4241607505141305,
+      "learning_rate": 1.8494679351215212e-05,
+      "loss": 0.1001,
+      "step": 461
+    },
+    {
+      "epoch": 0.7107692307692308,
+      "grad_norm": 0.4683808796780716,
+      "learning_rate": 1.84882976489001e-05,
+      "loss": 0.0853,
+      "step": 462
+    },
+    {
+      "epoch": 0.7123076923076923,
+      "grad_norm": 0.44346576858606745,
+      "learning_rate": 1.8481903553666405e-05,
+      "loss": 0.0995,
+      "step": 463
+    },
+    {
+      "epoch": 0.7138461538461538,
+      "grad_norm": 0.45180799988411885,
+      "learning_rate": 1.84754970748495e-05,
+      "loss": 0.0848,
+      "step": 464
+    },
+    {
+      "epoch": 0.7153846153846154,
+      "grad_norm": 0.4173368978655163,
+      "learning_rate": 1.846907822180286e-05,
+      "loss": 0.0954,
+      "step": 465
+    },
+    {
+      "epoch": 0.7169230769230769,
+      "grad_norm": 0.42068772817462785,
+      "learning_rate": 1.8462647003898005e-05,
+      "loss": 0.0951,
+      "step": 466
+    },
+    {
+      "epoch": 0.7184615384615385,
+      "grad_norm": 0.49661800554688595,
+      "learning_rate": 1.845620343052452e-05,
+      "loss": 0.0949,
+      "step": 467
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48646267743412663,
+      "learning_rate": 1.844974751109002e-05,
+      "loss": 0.0964,
+      "step": 468
+    },
+    {
+      "epoch": 0.7215384615384616,
+      "grad_norm": 0.4106987772360316,
+      "learning_rate": 1.8443279255020153e-05,
+      "loss": 0.0974,
+      "step": 469
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 0.3921781146014997,
+      "learning_rate": 1.843679867175858e-05,
+      "loss": 0.0848,
+      "step": 470
+    },
+    {
+      "epoch": 0.7246153846153847,
+      "grad_norm": 0.44000195303506684,
+      "learning_rate": 1.8430305770766947e-05,
+      "loss": 0.1023,
+      "step": 471
+    },
+    {
+      "epoch": 0.7261538461538461,
+      "grad_norm": 0.5186352660566815,
+      "learning_rate": 1.84238005615249e-05,
+      "loss": 0.0943,
+      "step": 472
+    },
+    {
+      "epoch": 0.7276923076923076,
+      "grad_norm": 0.4684858223587338,
+      "learning_rate": 1.8417283053530047e-05,
+      "loss": 0.1027,
+      "step": 473
+    },
+    {
+      "epoch": 0.7292307692307692,
+      "grad_norm": 0.39874288545350156,
+      "learning_rate": 1.8410753256297948e-05,
+      "loss": 0.1023,
+      "step": 474
+    },
+    {
+      "epoch": 0.7307692307692307,
+      "grad_norm": 0.3687327920324425,
+      "learning_rate": 1.8404211179362116e-05,
+      "loss": 0.102,
+      "step": 475
+    },
+    {
+      "epoch": 0.7323076923076923,
+      "grad_norm": 0.37907564191738674,
+      "learning_rate": 1.8397656832273982e-05,
+      "loss": 0.0902,
+      "step": 476
+    },
+    {
+      "epoch": 0.7338461538461538,
+      "grad_norm": 0.5844119747750028,
+      "learning_rate": 1.8391090224602895e-05,
+      "loss": 0.1024,
+      "step": 477
+    },
+    {
+      "epoch": 0.7353846153846154,
+      "grad_norm": 0.4973700871932812,
+      "learning_rate": 1.8384511365936112e-05,
+      "loss": 0.0986,
+      "step": 478
+    },
+    {
+      "epoch": 0.7369230769230769,
+      "grad_norm": 0.4220278832402647,
+      "learning_rate": 1.837792026587876e-05,
+      "loss": 0.0947,
+      "step": 479
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 0.48741372675693995,
+      "learning_rate": 1.837131693405386e-05,
+      "loss": 0.1039,
+      "step": 480
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5095101140536407,
+      "learning_rate": 1.8364701380102267e-05,
+      "loss": 0.1014,
+      "step": 481
+    },
+    {
+      "epoch": 0.7415384615384616,
+      "grad_norm": 0.3686730605957879,
+      "learning_rate": 1.8358073613682705e-05,
+      "loss": 0.0879,
+      "step": 482
+    },
+    {
+      "epoch": 0.7430769230769231,
+      "grad_norm": 0.4018358216960023,
+      "learning_rate": 1.8351433644471708e-05,
+      "loss": 0.0951,
+      "step": 483
+    },
+    {
+      "epoch": 0.7446153846153846,
+      "grad_norm": 0.42272315942515715,
+      "learning_rate": 1.8344781482163635e-05,
+      "loss": 0.1006,
+      "step": 484
+    },
+    {
+      "epoch": 0.7461538461538462,
+      "grad_norm": 0.36114348383132194,
+      "learning_rate": 1.8338117136470648e-05,
+      "loss": 0.0876,
+      "step": 485
+    },
+    {
+      "epoch": 0.7476923076923077,
+      "grad_norm": 0.44109486857223745,
+      "learning_rate": 1.8331440617122694e-05,
+      "loss": 0.1014,
+      "step": 486
+    },
+    {
+      "epoch": 0.7492307692307693,
+      "grad_norm": 0.4735198416376007,
+      "learning_rate": 1.8324751933867496e-05,
+      "loss": 0.101,
+      "step": 487
+    },
+    {
+      "epoch": 0.7507692307692307,
+      "grad_norm": 0.5212542112337067,
+      "learning_rate": 1.831805109647053e-05,
+      "loss": 0.0965,
+      "step": 488
+    },
+    {
+      "epoch": 0.7523076923076923,
+      "grad_norm": 0.3867194456131576,
+      "learning_rate": 1.831133811471503e-05,
+      "loss": 0.0923,
+      "step": 489
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 0.4232158553279915,
+      "learning_rate": 1.8304612998401947e-05,
+      "loss": 0.086,
+      "step": 490
+    },
+    {
+      "epoch": 0.7553846153846154,
+      "grad_norm": 0.4882561711422476,
+      "learning_rate": 1.829787575734995e-05,
+      "loss": 0.084,
+      "step": 491
+    },
+    {
+      "epoch": 0.7569230769230769,
+      "grad_norm": 0.49255164116648453,
+      "learning_rate": 1.8291126401395425e-05,
+      "loss": 0.1011,
+      "step": 492
+    },
+    {
+      "epoch": 0.7584615384615384,
+      "grad_norm": 0.39462697940845287,
+      "learning_rate": 1.8284364940392426e-05,
+      "loss": 0.0981,
+      "step": 493
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.42965433064236386,
+      "learning_rate": 1.8277591384212693e-05,
+      "loss": 0.1049,
+      "step": 494
+    },
+    {
+      "epoch": 0.7615384615384615,
+      "grad_norm": 0.42153666897549313,
+      "learning_rate": 1.827080574274562e-05,
+      "loss": 0.1015,
+      "step": 495
+    },
+    {
+      "epoch": 0.7630769230769231,
+      "grad_norm": 0.5244873385591531,
+      "learning_rate": 1.8264008025898248e-05,
+      "loss": 0.1021,
+      "step": 496
+    },
+    {
+      "epoch": 0.7646153846153846,
+      "grad_norm": 0.48573346836145637,
+      "learning_rate": 1.825719824359524e-05,
+      "loss": 0.1012,
+      "step": 497
+    },
+    {
+      "epoch": 0.7661538461538462,
+      "grad_norm": 0.4201750596778976,
+      "learning_rate": 1.8250376405778897e-05,
+      "loss": 0.0914,
+      "step": 498
+    },
+    {
+      "epoch": 0.7676923076923077,
+      "grad_norm": 0.3689138791311251,
+      "learning_rate": 1.824354252240909e-05,
+      "loss": 0.0971,
+      "step": 499
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.446354377870851,
+      "learning_rate": 1.8236696603463297e-05,
+      "loss": 0.1104,
+      "step": 500
+    },
+    {
+      "epoch": 0.7707692307692308,
+      "grad_norm": 0.5302794302913245,
+      "learning_rate": 1.8229838658936566e-05,
+      "loss": 0.0913,
+      "step": 501
+    },
+    {
+      "epoch": 0.7723076923076924,
+      "grad_norm": 0.6151683521942346,
+      "learning_rate": 1.8222968698841495e-05,
+      "loss": 0.0842,
+      "step": 502
+    },
+    {
+      "epoch": 0.7738461538461539,
+      "grad_norm": 0.4018248722446391,
+      "learning_rate": 1.821608673320823e-05,
+      "loss": 0.0896,
+      "step": 503
+    },
+    {
+      "epoch": 0.7753846153846153,
+      "grad_norm": 0.36582381181570395,
+      "learning_rate": 1.8209192772084446e-05,
+      "loss": 0.0899,
+      "step": 504
+    },
+    {
+      "epoch": 0.7769230769230769,
+      "grad_norm": 0.42007534297254007,
+      "learning_rate": 1.820228682553533e-05,
+      "loss": 0.1012,
+      "step": 505
+    },
+    {
+      "epoch": 0.7784615384615384,
+      "grad_norm": 0.3706728601304087,
+      "learning_rate": 1.8195368903643565e-05,
+      "loss": 0.0937,
+      "step": 506
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3386897127268892,
+      "learning_rate": 1.818843901650932e-05,
+      "loss": 0.0864,
+      "step": 507
+    },
+    {
+      "epoch": 0.7815384615384615,
+      "grad_norm": 0.4604244232600737,
+      "learning_rate": 1.8181497174250236e-05,
+      "loss": 0.0886,
+      "step": 508
+    },
+    {
+      "epoch": 0.7830769230769231,
+      "grad_norm": 0.46651975178638583,
+      "learning_rate": 1.8174543387001403e-05,
+      "loss": 0.0942,
+      "step": 509
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 0.37252137712706535,
+      "learning_rate": 1.8167577664915354e-05,
+      "loss": 0.098,
+      "step": 510
+    },
+    {
+      "epoch": 0.7861538461538462,
+      "grad_norm": 0.4549144644337013,
+      "learning_rate": 1.816060001816205e-05,
+      "loss": 0.1043,
+      "step": 511
+    },
+    {
+      "epoch": 0.7876923076923077,
+      "grad_norm": 0.49452287761067387,
+      "learning_rate": 1.8153610456928853e-05,
+      "loss": 0.0976,
+      "step": 512
+    },
+    {
+      "epoch": 0.7892307692307692,
+      "grad_norm": 0.5123177646707533,
+      "learning_rate": 1.8146608991420533e-05,
+      "loss": 0.0974,
+      "step": 513
+    },
+    {
+      "epoch": 0.7907692307692308,
+      "grad_norm": 0.39618956334101985,
+      "learning_rate": 1.8139595631859228e-05,
+      "loss": 0.0891,
+      "step": 514
+    },
+    {
+      "epoch": 0.7923076923076923,
+      "grad_norm": 0.3864877721629785,
+      "learning_rate": 1.8132570388484442e-05,
+      "loss": 0.0849,
+      "step": 515
+    },
+    {
+      "epoch": 0.7938461538461539,
+      "grad_norm": 0.48621924311976095,
+      "learning_rate": 1.8125533271553045e-05,
+      "loss": 0.1014,
+      "step": 516
+    },
+    {
+      "epoch": 0.7953846153846154,
+      "grad_norm": 0.46058902465123464,
+      "learning_rate": 1.811848429133922e-05,
+      "loss": 0.1114,
+      "step": 517
+    },
+    {
+      "epoch": 0.796923076923077,
+      "grad_norm": 0.40700166379805536,
+      "learning_rate": 1.811142345813449e-05,
+      "loss": 0.0903,
+      "step": 518
+    },
+    {
+      "epoch": 0.7984615384615384,
+      "grad_norm": 0.38144448483836707,
+      "learning_rate": 1.810435078224767e-05,
+      "loss": 0.0999,
+      "step": 519
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4087914237321906,
+      "learning_rate": 1.809726627400487e-05,
+      "loss": 0.0905,
+      "step": 520
+    },
+    {
+      "epoch": 0.8015384615384615,
+      "grad_norm": 0.454530116045963,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.0981,
+      "step": 521
+    },
+    {
+      "epoch": 0.803076923076923,
+      "grad_norm": 0.387856092368395,
+      "learning_rate": 1.8083061801842133e-05,
+      "loss": 0.0914,
+      "step": 522
+    },
+    {
+      "epoch": 0.8046153846153846,
+      "grad_norm": 0.4175456777206446,
+      "learning_rate": 1.8075941858660737e-05,
+      "loss": 0.0996,
+      "step": 523
+    },
+    {
+      "epoch": 0.8061538461538461,
+      "grad_norm": 0.7864656011521015,
+      "learning_rate": 1.8068810124600403e-05,
+      "loss": 0.0899,
+      "step": 524
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.48921911883769525,
+      "learning_rate": 1.8061666610073465e-05,
+      "loss": 0.0979,
+      "step": 525
+    },
+    {
+      "epoch": 0.8092307692307692,
+      "grad_norm": 0.48154421315499524,
+      "learning_rate": 1.805451132550946e-05,
+      "loss": 0.1026,
+      "step": 526
+    },
+    {
+      "epoch": 0.8107692307692308,
+      "grad_norm": 0.45822965917426073,
+      "learning_rate": 1.8047344281355112e-05,
+      "loss": 0.1056,
+      "step": 527
+    },
+    {
+      "epoch": 0.8123076923076923,
+      "grad_norm": 0.4699180830359611,
+      "learning_rate": 1.8040165488074294e-05,
+      "loss": 0.0978,
+      "step": 528
+    },
+    {
+      "epoch": 0.8138461538461539,
+      "grad_norm": 0.4386694887627527,
+      "learning_rate": 1.8032974956148064e-05,
+      "loss": 0.0852,
+      "step": 529
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 0.3982089506558907,
+      "learning_rate": 1.8025772696074593e-05,
+      "loss": 0.0847,
+      "step": 530
+    },
+    {
+      "epoch": 0.816923076923077,
+      "grad_norm": 0.4451526496740161,
+      "learning_rate": 1.8018558718369187e-05,
+      "loss": 0.0949,
+      "step": 531
+    },
+    {
+      "epoch": 0.8184615384615385,
+      "grad_norm": 0.42797389669933644,
+      "learning_rate": 1.8011333033564255e-05,
+      "loss": 0.096,
+      "step": 532
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5816907168348651,
+      "learning_rate": 1.8004095652209304e-05,
+      "loss": 0.0979,
+      "step": 533
+    },
+    {
+      "epoch": 0.8215384615384616,
+      "grad_norm": 0.4209426200782058,
+      "learning_rate": 1.799684658487091e-05,
+      "loss": 0.0912,
+      "step": 534
+    },
+    {
+      "epoch": 0.823076923076923,
+      "grad_norm": 0.4459278908996159,
+      "learning_rate": 1.7989585842132713e-05,
+      "loss": 0.0932,
+      "step": 535
+    },
+    {
+      "epoch": 0.8246153846153846,
+      "grad_norm": 0.3996692934376165,
+      "learning_rate": 1.7982313434595405e-05,
+      "loss": 0.0846,
+      "step": 536
+    },
+    {
+      "epoch": 0.8261538461538461,
+      "grad_norm": 0.46838655788136374,
+      "learning_rate": 1.7975029372876706e-05,
+      "loss": 0.0931,
+      "step": 537
+    },
+    {
+      "epoch": 0.8276923076923077,
+      "grad_norm": 0.44272759560569114,
+      "learning_rate": 1.7967733667611346e-05,
+      "loss": 0.0991,
+      "step": 538
+    },
+    {
+      "epoch": 0.8292307692307692,
+      "grad_norm": 0.6005247882009117,
+      "learning_rate": 1.7960426329451062e-05,
+      "loss": 0.0947,
+      "step": 539
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 0.43973649591876807,
+      "learning_rate": 1.7953107369064563e-05,
+      "loss": 0.0946,
+      "step": 540
+    },
+    {
+      "epoch": 0.8323076923076923,
+      "grad_norm": 0.36356195409947895,
+      "learning_rate": 1.7945776797137544e-05,
+      "loss": 0.0829,
+      "step": 541
+    },
+    {
+      "epoch": 0.8338461538461538,
+      "grad_norm": 0.4199497642090339,
+      "learning_rate": 1.7938434624372638e-05,
+      "loss": 0.0925,
+      "step": 542
+    },
+    {
+      "epoch": 0.8353846153846154,
+      "grad_norm": 0.5142377507204998,
+      "learning_rate": 1.7931080861489425e-05,
+      "loss": 0.0973,
+      "step": 543
+    },
+    {
+      "epoch": 0.8369230769230769,
+      "grad_norm": 0.5226363888586346,
+      "learning_rate": 1.7923715519224397e-05,
+      "loss": 0.1065,
+      "step": 544
+    },
+    {
+      "epoch": 0.8384615384615385,
+      "grad_norm": 0.4729813679434869,
+      "learning_rate": 1.791633860833096e-05,
+      "loss": 0.0895,
+      "step": 545
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3606695909284096,
+      "learning_rate": 1.7908950139579406e-05,
+      "loss": 0.0832,
+      "step": 546
+    },
+    {
+      "epoch": 0.8415384615384616,
+      "grad_norm": 0.570980416724036,
+      "learning_rate": 1.7901550123756906e-05,
+      "loss": 0.1002,
+      "step": 547
+    },
+    {
+      "epoch": 0.8430769230769231,
+      "grad_norm": 0.5594313800490134,
+      "learning_rate": 1.7894138571667482e-05,
+      "loss": 0.1008,
+      "step": 548
+    },
+    {
+      "epoch": 0.8446153846153847,
+      "grad_norm": 0.4315817673384466,
+      "learning_rate": 1.7886715494132008e-05,
+      "loss": 0.0888,
+      "step": 549
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.3916992320727455,
+      "learning_rate": 1.787928090198818e-05,
+      "loss": 0.0858,
+      "step": 550
+    },
+    {
+      "epoch": 0.8476923076923077,
+      "grad_norm": 0.5103149586075123,
+      "learning_rate": 1.7871834806090502e-05,
+      "loss": 0.1078,
+      "step": 551
+    },
+    {
+      "epoch": 0.8492307692307692,
+      "grad_norm": 0.5179913026426116,
+      "learning_rate": 1.7864377217310282e-05,
+      "loss": 0.0941,
+      "step": 552
+    },
+    {
+      "epoch": 0.8507692307692307,
+      "grad_norm": 0.5462964124972205,
+      "learning_rate": 1.7856908146535602e-05,
+      "loss": 0.0844,
+      "step": 553
+    },
+    {
+      "epoch": 0.8523076923076923,
+      "grad_norm": 0.40593162957333356,
+      "learning_rate": 1.784942760467131e-05,
+      "loss": 0.0943,
+      "step": 554
+    },
+    {
+      "epoch": 0.8538461538461538,
+      "grad_norm": 0.38375960879766213,
+      "learning_rate": 1.7841935602638997e-05,
+      "loss": 0.0804,
+      "step": 555
+    },
+    {
+      "epoch": 0.8553846153846154,
+      "grad_norm": 0.38136544658194893,
+      "learning_rate": 1.7834432151376992e-05,
+      "loss": 0.0945,
+      "step": 556
+    },
+    {
+      "epoch": 0.8569230769230769,
+      "grad_norm": 0.4194877597337682,
+      "learning_rate": 1.7826917261840337e-05,
+      "loss": 0.092,
+      "step": 557
+    },
+    {
+      "epoch": 0.8584615384615385,
+      "grad_norm": 0.4923272058939714,
+      "learning_rate": 1.7819390945000775e-05,
+      "loss": 0.0986,
+      "step": 558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4875318696461418,
+      "learning_rate": 1.781185321184673e-05,
+      "loss": 0.1017,
+      "step": 559
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 0.39052926422478096,
+      "learning_rate": 1.7804304073383298e-05,
+      "loss": 0.1091,
+      "step": 560
+    },
+    {
+      "epoch": 0.8630769230769231,
+      "grad_norm": 0.38763635529234175,
+      "learning_rate": 1.7796743540632226e-05,
+      "loss": 0.0825,
+      "step": 561
+    },
+    {
+      "epoch": 0.8646153846153846,
+      "grad_norm": 0.3905770386460791,
+      "learning_rate": 1.778917162463189e-05,
+      "loss": 0.0875,
+      "step": 562
+    },
+    {
+      "epoch": 0.8661538461538462,
+      "grad_norm": 0.5703432509432513,
+      "learning_rate": 1.77815883364373e-05,
+      "loss": 0.1001,
+      "step": 563
+    },
+    {
+      "epoch": 0.8676923076923077,
+      "grad_norm": 0.4597959384906591,
+      "learning_rate": 1.777399368712005e-05,
+      "loss": 0.0882,
+      "step": 564
+    },
+    {
+      "epoch": 0.8692307692307693,
+      "grad_norm": 0.45607761156419957,
+      "learning_rate": 1.7766387687768338e-05,
+      "loss": 0.0991,
+      "step": 565
+    },
+    {
+      "epoch": 0.8707692307692307,
+      "grad_norm": 0.39506065545533703,
+      "learning_rate": 1.7758770349486924e-05,
+      "loss": 0.0943,
+      "step": 566
+    },
+    {
+      "epoch": 0.8723076923076923,
+      "grad_norm": 0.4913700102929278,
+      "learning_rate": 1.7751141683397128e-05,
+      "loss": 0.1026,
+      "step": 567
+    },
+    {
+      "epoch": 0.8738461538461538,
+      "grad_norm": 0.4481916097965894,
+      "learning_rate": 1.7743501700636804e-05,
+      "loss": 0.1035,
+      "step": 568
+    },
+    {
+      "epoch": 0.8753846153846154,
+      "grad_norm": 0.40743252425944065,
+      "learning_rate": 1.7735850412360332e-05,
+      "loss": 0.0915,
+      "step": 569
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 0.36201263572487186,
+      "learning_rate": 1.7728187829738596e-05,
+      "loss": 0.096,
+      "step": 570
+    },
+    {
+      "epoch": 0.8784615384615385,
+      "grad_norm": 0.41877789233929563,
+      "learning_rate": 1.772051396395897e-05,
+      "loss": 0.0921,
+      "step": 571
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.43388638709693894,
+      "learning_rate": 1.7712828826225303e-05,
+      "loss": 0.1016,
+      "step": 572
+    },
+    {
+      "epoch": 0.8815384615384615,
+      "grad_norm": 0.4621448068884329,
+      "learning_rate": 1.7705132427757895e-05,
+      "loss": 0.0957,
+      "step": 573
+    },
+    {
+      "epoch": 0.8830769230769231,
+      "grad_norm": 0.384208019183389,
+      "learning_rate": 1.7697424779793497e-05,
+      "loss": 0.0812,
+      "step": 574
+    },
+    {
+      "epoch": 0.8846153846153846,
+      "grad_norm": 0.35252484159436787,
+      "learning_rate": 1.7689705893585273e-05,
+      "loss": 0.0918,
+      "step": 575
+    },
+    {
+      "epoch": 0.8861538461538462,
+      "grad_norm": 0.40143829209617715,
+      "learning_rate": 1.7681975780402807e-05,
+      "loss": 0.0843,
+      "step": 576
+    },
+    {
+      "epoch": 0.8876923076923077,
+      "grad_norm": 0.4307373370195823,
+      "learning_rate": 1.7674234451532065e-05,
+      "loss": 0.0961,
+      "step": 577
+    },
+    {
+      "epoch": 0.8892307692307693,
+      "grad_norm": 0.46711054218809045,
+      "learning_rate": 1.766648191827539e-05,
+      "loss": 0.0918,
+      "step": 578
+    },
+    {
+      "epoch": 0.8907692307692308,
+      "grad_norm": 0.43239124932710876,
+      "learning_rate": 1.7658718191951483e-05,
+      "loss": 0.0865,
+      "step": 579
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 0.36077221637299095,
+      "learning_rate": 1.7650943283895393e-05,
+      "loss": 0.0791,
+      "step": 580
+    },
+    {
+      "epoch": 0.8938461538461538,
+      "grad_norm": 0.452062987622938,
+      "learning_rate": 1.7643157205458483e-05,
+      "loss": 0.0977,
+      "step": 581
+    },
+    {
+      "epoch": 0.8953846153846153,
+      "grad_norm": 0.3729118607027113,
+      "learning_rate": 1.7635359968008438e-05,
+      "loss": 0.0964,
+      "step": 582
+    },
+    {
+      "epoch": 0.8969230769230769,
+      "grad_norm": 0.31864024958361536,
+      "learning_rate": 1.7627551582929223e-05,
+      "loss": 0.0706,
+      "step": 583
+    },
+    {
+      "epoch": 0.8984615384615384,
+      "grad_norm": 0.4897692560311474,
+      "learning_rate": 1.761973206162109e-05,
+      "loss": 0.095,
+      "step": 584
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.43634370121716065,
+      "learning_rate": 1.7611901415500536e-05,
+      "loss": 0.1012,
+      "step": 585
+    },
+    {
+      "epoch": 0.9015384615384615,
+      "grad_norm": 0.4303303314789484,
+      "learning_rate": 1.7604059656000313e-05,
+      "loss": 0.0991,
+      "step": 586
+    },
+    {
+      "epoch": 0.9030769230769231,
+      "grad_norm": 0.3965278873786633,
+      "learning_rate": 1.759620679456939e-05,
+      "loss": 0.0889,
+      "step": 587
+    },
+    {
+      "epoch": 0.9046153846153846,
+      "grad_norm": 0.3471384513881275,
+      "learning_rate": 1.758834284267295e-05,
+      "loss": 0.0901,
+      "step": 588
+    },
+    {
+      "epoch": 0.9061538461538462,
+      "grad_norm": 0.35615891292502644,
+      "learning_rate": 1.7580467811792374e-05,
+      "loss": 0.084,
+      "step": 589
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 0.43993009413917056,
+      "learning_rate": 1.7572581713425195e-05,
+      "loss": 0.0986,
+      "step": 590
+    },
+    {
+      "epoch": 0.9092307692307692,
+      "grad_norm": 0.41369164253274765,
+      "learning_rate": 1.7564684559085138e-05,
+      "loss": 0.0888,
+      "step": 591
+    },
+    {
+      "epoch": 0.9107692307692308,
+      "grad_norm": 0.3683031999775181,
+      "learning_rate": 1.7556776360302038e-05,
+      "loss": 0.0912,
+      "step": 592
+    },
+    {
+      "epoch": 0.9123076923076923,
+      "grad_norm": 0.4537934276100818,
+      "learning_rate": 1.7548857128621878e-05,
+      "loss": 0.0921,
+      "step": 593
+    },
+    {
+      "epoch": 0.9138461538461539,
+      "grad_norm": 0.43435742031494934,
+      "learning_rate": 1.7540926875606734e-05,
+      "loss": 0.0869,
+      "step": 594
+    },
+    {
+      "epoch": 0.9153846153846154,
+      "grad_norm": 0.38341724227303553,
+      "learning_rate": 1.753298561283478e-05,
+      "loss": 0.0903,
+      "step": 595
+    },
+    {
+      "epoch": 0.916923076923077,
+      "grad_norm": 0.41746460815399494,
+      "learning_rate": 1.7525033351900268e-05,
+      "loss": 0.1006,
+      "step": 596
+    },
+    {
+      "epoch": 0.9184615384615384,
+      "grad_norm": 0.3777485553214645,
+      "learning_rate": 1.7517070104413497e-05,
+      "loss": 0.0947,
+      "step": 597
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.41830568154844494,
+      "learning_rate": 1.7509095882000823e-05,
+      "loss": 0.1039,
+      "step": 598
+    },
+    {
+      "epoch": 0.9215384615384615,
+      "grad_norm": 0.36147731651312526,
+      "learning_rate": 1.7501110696304598e-05,
+      "loss": 0.0875,
+      "step": 599
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.3427586374547824,
+      "learning_rate": 1.7493114558983207e-05,
+      "loss": 0.0816,
+      "step": 600
+    },
+    {
+      "epoch": 0.9246153846153846,
+      "grad_norm": 0.3939804203050907,
+      "learning_rate": 1.7485107481711014e-05,
+      "loss": 0.0946,
+      "step": 601
+    },
+    {
+      "epoch": 0.9261538461538461,
+      "grad_norm": 0.381608432957127,
+      "learning_rate": 1.7477089476178354e-05,
+      "loss": 0.0929,
+      "step": 602
+    },
+    {
+      "epoch": 0.9276923076923077,
+      "grad_norm": 0.4292122557448878,
+      "learning_rate": 1.7469060554091518e-05,
+      "loss": 0.0965,
+      "step": 603
+    },
+    {
+      "epoch": 0.9292307692307692,
+      "grad_norm": 0.3762116025408573,
+      "learning_rate": 1.7461020727172736e-05,
+      "loss": 0.0938,
+      "step": 604
+    },
+    {
+      "epoch": 0.9307692307692308,
+      "grad_norm": 0.3455952506447102,
+      "learning_rate": 1.745297000716016e-05,
+      "loss": 0.0741,
+      "step": 605
+    },
+    {
+      "epoch": 0.9323076923076923,
+      "grad_norm": 0.38631458311410155,
+      "learning_rate": 1.7444908405807845e-05,
+      "loss": 0.0865,
+      "step": 606
+    },
+    {
+      "epoch": 0.9338461538461539,
+      "grad_norm": 0.37275711289343844,
+      "learning_rate": 1.7436835934885735e-05,
+      "loss": 0.0886,
+      "step": 607
+    },
+    {
+      "epoch": 0.9353846153846154,
+      "grad_norm": 0.4401140100719851,
+      "learning_rate": 1.742875260617964e-05,
+      "loss": 0.0976,
+      "step": 608
+    },
+    {
+      "epoch": 0.936923076923077,
+      "grad_norm": 0.3921855404257059,
+      "learning_rate": 1.7420658431491224e-05,
+      "loss": 0.0928,
+      "step": 609
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 0.4113951217410347,
+      "learning_rate": 1.741255342263798e-05,
+      "loss": 0.0885,
+      "step": 610
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3262775097841378,
+      "learning_rate": 1.7404437591453237e-05,
+      "loss": 0.0873,
+      "step": 611
+    },
+    {
+      "epoch": 0.9415384615384615,
+      "grad_norm": 0.42350913769649273,
+      "learning_rate": 1.73963109497861e-05,
+      "loss": 0.0958,
+      "step": 612
+    },
+    {
+      "epoch": 0.943076923076923,
+      "grad_norm": 0.36988134430557396,
+      "learning_rate": 1.7388173509501475e-05,
+      "loss": 0.0928,
+      "step": 613
+    },
+    {
+      "epoch": 0.9446153846153846,
+      "grad_norm": 0.39410658459241166,
+      "learning_rate": 1.7380025282480028e-05,
+      "loss": 0.0868,
+      "step": 614
+    },
+    {
+      "epoch": 0.9461538461538461,
+      "grad_norm": 0.41180225877084126,
+      "learning_rate": 1.7371866280618176e-05,
+      "loss": 0.0911,
+      "step": 615
+    },
+    {
+      "epoch": 0.9476923076923077,
+      "grad_norm": 0.35890123689534553,
+      "learning_rate": 1.7363696515828062e-05,
+      "loss": 0.0899,
+      "step": 616
+    },
+    {
+      "epoch": 0.9492307692307692,
+      "grad_norm": 0.2987459340450158,
+      "learning_rate": 1.7355516000037555e-05,
+      "loss": 0.0769,
+      "step": 617
+    },
+    {
+      "epoch": 0.9507692307692308,
+      "grad_norm": 0.43301969400641127,
+      "learning_rate": 1.73473247451902e-05,
+      "loss": 0.0892,
+      "step": 618
+    },
+    {
+      "epoch": 0.9523076923076923,
+      "grad_norm": 0.379899779524929,
+      "learning_rate": 1.733912276324524e-05,
+      "loss": 0.0751,
+      "step": 619
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 0.35284361586327184,
+      "learning_rate": 1.7330910066177574e-05,
+      "loss": 0.0941,
+      "step": 620
+    },
+    {
+      "epoch": 0.9553846153846154,
+      "grad_norm": 0.4459109096619344,
+      "learning_rate": 1.7322686665977738e-05,
+      "loss": 0.086,
+      "step": 621
+    },
+    {
+      "epoch": 0.9569230769230769,
+      "grad_norm": 0.4302144299712948,
+      "learning_rate": 1.7314452574651902e-05,
+      "loss": 0.0865,
+      "step": 622
+    },
+    {
+      "epoch": 0.9584615384615385,
+      "grad_norm": 0.4053685427552524,
+      "learning_rate": 1.7306207804221845e-05,
+      "loss": 0.092,
+      "step": 623
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.46831862180021,
+      "learning_rate": 1.7297952366724935e-05,
+      "loss": 0.1046,
+      "step": 624
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 0.4538044661350598,
+      "learning_rate": 1.7289686274214116e-05,
+      "loss": 0.1076,
+      "step": 625
+    },
+    {
+      "epoch": 0.963076923076923,
+      "grad_norm": 0.48389404664226576,
+      "learning_rate": 1.7281409538757886e-05,
+      "loss": 0.0923,
+      "step": 626
+    },
+    {
+      "epoch": 0.9646153846153847,
+      "grad_norm": 0.3865282157362884,
+      "learning_rate": 1.727312217244028e-05,
+      "loss": 0.0881,
+      "step": 627
+    },
+    {
+      "epoch": 0.9661538461538461,
+      "grad_norm": 0.41052873518942035,
+      "learning_rate": 1.726482418736086e-05,
+      "loss": 0.0982,
+      "step": 628
+    },
+    {
+      "epoch": 0.9676923076923077,
+      "grad_norm": 0.4633845480863643,
+      "learning_rate": 1.7256515595634688e-05,
+      "loss": 0.0963,
+      "step": 629
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 0.49146329005893413,
+      "learning_rate": 1.7248196409392312e-05,
+      "loss": 0.0924,
+      "step": 630
+    },
+    {
+      "epoch": 0.9707692307692307,
+      "grad_norm": 0.35983268721595085,
+      "learning_rate": 1.7239866640779745e-05,
+      "loss": 0.0884,
+      "step": 631
+    },
+    {
+      "epoch": 0.9723076923076923,
+      "grad_norm": 0.490761583695885,
+      "learning_rate": 1.7231526301958454e-05,
+      "loss": 0.0973,
+      "step": 632
+    },
+    {
+      "epoch": 0.9738461538461538,
+      "grad_norm": 0.3594410933659223,
+      "learning_rate": 1.722317540510534e-05,
+      "loss": 0.0929,
+      "step": 633
+    },
+    {
+      "epoch": 0.9753846153846154,
+      "grad_norm": 0.32824532953007957,
+      "learning_rate": 1.7214813962412715e-05,
+      "loss": 0.0886,
+      "step": 634
+    },
+    {
+      "epoch": 0.9769230769230769,
+      "grad_norm": 0.3557861935512496,
+      "learning_rate": 1.720644198608829e-05,
+      "loss": 0.0885,
+      "step": 635
+    },
+    {
+      "epoch": 0.9784615384615385,
+      "grad_norm": 0.45266556902976496,
+      "learning_rate": 1.7198059488355153e-05,
+      "loss": 0.0933,
+      "step": 636
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.40557942924254287,
+      "learning_rate": 1.7189666481451755e-05,
+      "loss": 0.0951,
+      "step": 637
+    },
+    {
+      "epoch": 0.9815384615384616,
+      "grad_norm": 0.35736570345844687,
+      "learning_rate": 1.718126297763189e-05,
+      "loss": 0.0888,
+      "step": 638
+    },
+    {
+      "epoch": 0.9830769230769231,
+      "grad_norm": 0.4631511259456779,
+      "learning_rate": 1.717284898916468e-05,
+      "loss": 0.091,
+      "step": 639
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 0.43267790292943314,
+      "learning_rate": 1.7164424528334548e-05,
+      "loss": 0.0961,
+      "step": 640
+    },
+    {
+      "epoch": 0.9861538461538462,
+      "grad_norm": 0.46968347009435635,
+      "learning_rate": 1.715598960744121e-05,
+      "loss": 0.0887,
+      "step": 641
+    },
+    {
+      "epoch": 0.9876923076923076,
+      "grad_norm": 0.447356115085785,
+      "learning_rate": 1.7147544238799664e-05,
+      "loss": 0.0908,
+      "step": 642
+    },
+    {
+      "epoch": 0.9892307692307692,
+      "grad_norm": 0.32693771260343074,
+      "learning_rate": 1.7139088434740142e-05,
+      "loss": 0.0872,
+      "step": 643
+    },
+    {
+      "epoch": 0.9907692307692307,
+      "grad_norm": 0.422322021584036,
+      "learning_rate": 1.7130622207608126e-05,
+      "loss": 0.0975,
+      "step": 644
+    },
+    {
+      "epoch": 0.9923076923076923,
+      "grad_norm": 0.34862552621490467,
+      "learning_rate": 1.712214556976431e-05,
+      "loss": 0.0944,
+      "step": 645
+    },
+    {
+      "epoch": 0.9938461538461538,
+      "grad_norm": 0.3800495940989224,
+      "learning_rate": 1.7113658533584594e-05,
+      "loss": 0.0793,
+      "step": 646
+    },
+    {
+      "epoch": 0.9953846153846154,
+      "grad_norm": 0.5118689124842419,
+      "learning_rate": 1.7105161111460046e-05,
+      "loss": 0.0949,
+      "step": 647
+    },
+    {
+      "epoch": 0.9969230769230769,
+      "grad_norm": 0.412157009432569,
+      "learning_rate": 1.7096653315796915e-05,
+      "loss": 0.1011,
+      "step": 648
+    },
+    {
+      "epoch": 0.9984615384615385,
+      "grad_norm": 0.3780510510068702,
+      "learning_rate": 1.7088135159016584e-05,
+      "loss": 0.0982,
+      "step": 649
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.41945932684457304,
+      "learning_rate": 1.7079606653555563e-05,
+      "loss": 0.0953,
+      "step": 650
+    },
+    {
+      "epoch": 1.0015384615384615,
+      "grad_norm": 0.43622049064952106,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.0749,
+      "step": 651
+    },
+    {
+      "epoch": 1.003076923076923,
+      "grad_norm": 0.38132209199435835,
+      "learning_rate": 1.706251864641304e-05,
+      "loss": 0.0726,
+      "step": 652
+    },
+    {
+      "epoch": 1.0046153846153847,
+      "grad_norm": 0.30052022007987506,
+      "learning_rate": 1.7053959169680033e-05,
+      "loss": 0.0715,
+      "step": 653
+    },
+    {
+      "epoch": 1.0061538461538462,
+      "grad_norm": 0.3185644660405221,
+      "learning_rate": 1.7045389394163297e-05,
+      "loss": 0.0688,
+      "step": 654
+    },
+    {
+      "epoch": 1.0076923076923077,
+      "grad_norm": 0.35208492988636686,
+      "learning_rate": 1.7036809332374713e-05,
+      "loss": 0.0689,
+      "step": 655
+    },
+    {
+      "epoch": 1.0092307692307692,
+      "grad_norm": 0.44091806848570947,
+      "learning_rate": 1.7028218996841173e-05,
+      "loss": 0.0709,
+      "step": 656
+    },
+    {
+      "epoch": 1.0107692307692309,
+      "grad_norm": 0.37173547317449596,
+      "learning_rate": 1.7019618400104572e-05,
+      "loss": 0.0663,
+      "step": 657
+    },
+    {
+      "epoch": 1.0123076923076924,
+      "grad_norm": 0.3752761957439137,
+      "learning_rate": 1.7011007554721778e-05,
+      "loss": 0.0696,
+      "step": 658
+    },
+    {
+      "epoch": 1.0138461538461538,
+      "grad_norm": 0.3413943290234514,
+      "learning_rate": 1.700238647326464e-05,
+      "loss": 0.0675,
+      "step": 659
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 0.37509528395315783,
+      "learning_rate": 1.6993755168319934e-05,
+      "loss": 0.0718,
+      "step": 660
+    },
+    {
+      "epoch": 1.0169230769230768,
+      "grad_norm": 0.39194820102366207,
+      "learning_rate": 1.6985113652489374e-05,
+      "loss": 0.062,
+      "step": 661
+    },
+    {
+      "epoch": 1.0184615384615385,
+      "grad_norm": 0.296028112394497,
+      "learning_rate": 1.697646193838957e-05,
+      "loss": 0.0658,
+      "step": 662
+    },
+    {
+      "epoch": 1.02,
+      "grad_norm": 0.3737919319488901,
+      "learning_rate": 1.6967800038652035e-05,
+      "loss": 0.0615,
+      "step": 663
+    },
+    {
+      "epoch": 1.0215384615384615,
+      "grad_norm": 0.36005445736837366,
+      "learning_rate": 1.6959127965923144e-05,
+      "loss": 0.0691,
+      "step": 664
+    },
+    {
+      "epoch": 1.023076923076923,
+      "grad_norm": 0.3514084769551997,
+      "learning_rate": 1.695044573286413e-05,
+      "loss": 0.0724,
+      "step": 665
+    },
+    {
+      "epoch": 1.0246153846153847,
+      "grad_norm": 0.37783560046213566,
+      "learning_rate": 1.6941753352151057e-05,
+      "loss": 0.063,
+      "step": 666
+    },
+    {
+      "epoch": 1.0261538461538462,
+      "grad_norm": 0.3639679221324666,
+      "learning_rate": 1.69330508364748e-05,
+      "loss": 0.0682,
+      "step": 667
+    },
+    {
+      "epoch": 1.0276923076923077,
+      "grad_norm": 0.42252833886280206,
+      "learning_rate": 1.692433819854104e-05,
+      "loss": 0.0694,
+      "step": 668
+    },
+    {
+      "epoch": 1.0292307692307692,
+      "grad_norm": 0.44692488489777815,
+      "learning_rate": 1.6915615451070234e-05,
+      "loss": 0.0705,
+      "step": 669
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 0.3678986564514892,
+      "learning_rate": 1.6906882606797595e-05,
+      "loss": 0.0657,
+      "step": 670
+    },
+    {
+      "epoch": 1.0323076923076924,
+      "grad_norm": 0.43913282105160406,
+      "learning_rate": 1.689813967847308e-05,
+      "loss": 0.071,
+      "step": 671
+    },
+    {
+      "epoch": 1.0338461538461539,
+      "grad_norm": 0.32313030736032045,
+      "learning_rate": 1.6889386678861365e-05,
+      "loss": 0.0665,
+      "step": 672
+    },
+    {
+      "epoch": 1.0353846153846153,
+      "grad_norm": 0.3913357186511293,
+      "learning_rate": 1.6880623620741843e-05,
+      "loss": 0.0677,
+      "step": 673
+    },
+    {
+      "epoch": 1.0369230769230768,
+      "grad_norm": 0.3502824876065054,
+      "learning_rate": 1.6871850516908575e-05,
+      "loss": 0.0651,
+      "step": 674
+    },
+    {
+      "epoch": 1.0384615384615385,
+      "grad_norm": 0.5615618864576786,
+      "learning_rate": 1.68630673801703e-05,
+      "loss": 0.0669,
+      "step": 675
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 0.45343396226547117,
+      "learning_rate": 1.68542742233504e-05,
+      "loss": 0.0817,
+      "step": 676
+    },
+    {
+      "epoch": 1.0415384615384615,
+      "grad_norm": 0.40223580906946954,
+      "learning_rate": 1.684547105928689e-05,
+      "loss": 0.0691,
+      "step": 677
+    },
+    {
+      "epoch": 1.043076923076923,
+      "grad_norm": 0.39525415008396314,
+      "learning_rate": 1.683665790083239e-05,
+      "loss": 0.0709,
+      "step": 678
+    },
+    {
+      "epoch": 1.0446153846153847,
+      "grad_norm": 0.29528931632742,
+      "learning_rate": 1.682783476085412e-05,
+      "loss": 0.0562,
+      "step": 679
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 0.4540404912733847,
+      "learning_rate": 1.6819001652233867e-05,
+      "loss": 0.0668,
+      "step": 680
+    },
+    {
+      "epoch": 1.0476923076923077,
+      "grad_norm": 0.4901824401666557,
+      "learning_rate": 1.6810158587867973e-05,
+      "loss": 0.0735,
+      "step": 681
+    },
+    {
+      "epoch": 1.0492307692307692,
+      "grad_norm": 0.35657036998216646,
+      "learning_rate": 1.6801305580667318e-05,
+      "loss": 0.0659,
+      "step": 682
+    },
+    {
+      "epoch": 1.0507692307692307,
+      "grad_norm": 0.34213142704883004,
+      "learning_rate": 1.679244264355729e-05,
+      "loss": 0.0631,
+      "step": 683
+    },
+    {
+      "epoch": 1.0523076923076924,
+      "grad_norm": 0.39311143967753803,
+      "learning_rate": 1.6783569789477795e-05,
+      "loss": 0.0652,
+      "step": 684
+    },
+    {
+      "epoch": 1.0538461538461539,
+      "grad_norm": 0.3501379606305712,
+      "learning_rate": 1.677468703138319e-05,
+      "loss": 0.0699,
+      "step": 685
+    },
+    {
+      "epoch": 1.0553846153846154,
+      "grad_norm": 0.359427195373609,
+      "learning_rate": 1.6765794382242315e-05,
+      "loss": 0.0657,
+      "step": 686
+    },
+    {
+      "epoch": 1.0569230769230769,
+      "grad_norm": 0.3900165610051957,
+      "learning_rate": 1.6756891855038436e-05,
+      "loss": 0.0648,
+      "step": 687
+    },
+    {
+      "epoch": 1.0584615384615386,
+      "grad_norm": 0.3768191856115309,
+      "learning_rate": 1.6747979462769253e-05,
+      "loss": 0.0661,
+      "step": 688
+    },
+    {
+      "epoch": 1.06,
+      "grad_norm": 0.3788900739237935,
+      "learning_rate": 1.673905721844686e-05,
+      "loss": 0.0689,
+      "step": 689
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 0.38294691852219787,
+      "learning_rate": 1.6730125135097736e-05,
+      "loss": 0.0733,
+      "step": 690
+    },
+    {
+      "epoch": 1.063076923076923,
+      "grad_norm": 0.35955788229196967,
+      "learning_rate": 1.6721183225762726e-05,
+      "loss": 0.0653,
+      "step": 691
+    },
+    {
+      "epoch": 1.0646153846153845,
+      "grad_norm": 0.4314633145507115,
+      "learning_rate": 1.6712231503497028e-05,
+      "loss": 0.0693,
+      "step": 692
+    },
+    {
+      "epoch": 1.0661538461538462,
+      "grad_norm": 0.3819072156771492,
+      "learning_rate": 1.670326998137016e-05,
+      "loss": 0.0686,
+      "step": 693
+    },
+    {
+      "epoch": 1.0676923076923077,
+      "grad_norm": 0.32895823745294955,
+      "learning_rate": 1.669429867246594e-05,
+      "loss": 0.0696,
+      "step": 694
+    },
+    {
+      "epoch": 1.0692307692307692,
+      "grad_norm": 0.4310560863117649,
+      "learning_rate": 1.668531758988249e-05,
+      "loss": 0.0715,
+      "step": 695
+    },
+    {
+      "epoch": 1.0707692307692307,
+      "grad_norm": 0.49530944777177144,
+      "learning_rate": 1.6676326746732197e-05,
+      "loss": 0.0684,
+      "step": 696
+    },
+    {
+      "epoch": 1.0723076923076924,
+      "grad_norm": 0.3399836957762591,
+      "learning_rate": 1.666732615614169e-05,
+      "loss": 0.0772,
+      "step": 697
+    },
+    {
+      "epoch": 1.073846153846154,
+      "grad_norm": 0.3717915125573905,
+      "learning_rate": 1.665831583125184e-05,
+      "loss": 0.0735,
+      "step": 698
+    },
+    {
+      "epoch": 1.0753846153846154,
+      "grad_norm": 0.4045082908654806,
+      "learning_rate": 1.6649295785217722e-05,
+      "loss": 0.0712,
+      "step": 699
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.38922594233503904,
+      "learning_rate": 1.664026603120861e-05,
+      "loss": 0.0709,
+      "step": 700
+    },
+    {
+      "epoch": 1.0784615384615384,
+      "grad_norm": 0.5013858686476608,
+      "learning_rate": 1.6631226582407954e-05,
+      "loss": 0.0764,
+      "step": 701
+    },
+    {
+      "epoch": 1.08,
+      "grad_norm": 0.3717981315840534,
+      "learning_rate": 1.6622177452013347e-05,
+      "loss": 0.0705,
+      "step": 702
+    },
+    {
+      "epoch": 1.0815384615384616,
+      "grad_norm": 0.38788483529072765,
+      "learning_rate": 1.661311865323652e-05,
+      "loss": 0.0719,
+      "step": 703
+    },
+    {
+      "epoch": 1.083076923076923,
+      "grad_norm": 1.053330169816148,
+      "learning_rate": 1.660405019930333e-05,
+      "loss": 0.0735,
+      "step": 704
+    },
+    {
+      "epoch": 1.0846153846153845,
+      "grad_norm": 0.6972695498767989,
+      "learning_rate": 1.6594972103453727e-05,
+      "loss": 0.0718,
+      "step": 705
+    },
+    {
+      "epoch": 1.0861538461538462,
+      "grad_norm": 0.37342568285958827,
+      "learning_rate": 1.6585884378941727e-05,
+      "loss": 0.0753,
+      "step": 706
+    },
+    {
+      "epoch": 1.0876923076923077,
+      "grad_norm": 0.4267000057052577,
+      "learning_rate": 1.6576787039035417e-05,
+      "loss": 0.0649,
+      "step": 707
+    },
+    {
+      "epoch": 1.0892307692307692,
+      "grad_norm": 0.36310651208043226,
+      "learning_rate": 1.6567680097016917e-05,
+      "loss": 0.0686,
+      "step": 708
+    },
+    {
+      "epoch": 1.0907692307692307,
+      "grad_norm": 0.42763138085861285,
+      "learning_rate": 1.6558563566182365e-05,
+      "loss": 0.0657,
+      "step": 709
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 0.4251635596635269,
+      "learning_rate": 1.65494374598419e-05,
+      "loss": 0.077,
+      "step": 710
+    },
+    {
+      "epoch": 1.093846153846154,
+      "grad_norm": 0.6823950567290125,
+      "learning_rate": 1.6540301791319647e-05,
+      "loss": 0.0642,
+      "step": 711
+    },
+    {
+      "epoch": 1.0953846153846154,
+      "grad_norm": 0.30086732326600235,
+      "learning_rate": 1.6531156573953677e-05,
+      "loss": 0.0635,
+      "step": 712
+    },
+    {
+      "epoch": 1.096923076923077,
+      "grad_norm": 0.3272124590429498,
+      "learning_rate": 1.652200182109602e-05,
+      "loss": 0.0711,
+      "step": 713
+    },
+    {
+      "epoch": 1.0984615384615384,
+      "grad_norm": 0.3783056835805405,
+      "learning_rate": 1.6512837546112617e-05,
+      "loss": 0.0702,
+      "step": 714
+    },
+    {
+      "epoch": 1.1,
+      "grad_norm": 0.34618660310859284,
+      "learning_rate": 1.6503663762383312e-05,
+      "loss": 0.066,
+      "step": 715
+    },
+    {
+      "epoch": 1.1015384615384616,
+      "grad_norm": 0.3509817308378501,
+      "learning_rate": 1.6494480483301836e-05,
+      "loss": 0.0681,
+      "step": 716
+    },
+    {
+      "epoch": 1.103076923076923,
+      "grad_norm": 0.435238684258864,
+      "learning_rate": 1.6485287722275783e-05,
+      "loss": 0.0654,
+      "step": 717
+    },
+    {
+      "epoch": 1.1046153846153846,
+      "grad_norm": 0.4858221064894468,
+      "learning_rate": 1.6476085492726582e-05,
+      "loss": 0.0621,
+      "step": 718
+    },
+    {
+      "epoch": 1.106153846153846,
+      "grad_norm": 0.5547851396481795,
+      "learning_rate": 1.6466873808089496e-05,
+      "loss": 0.065,
+      "step": 719
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 0.3551740998729382,
+      "learning_rate": 1.645765268181359e-05,
+      "loss": 0.0695,
+      "step": 720
+    },
+    {
+      "epoch": 1.1092307692307692,
+      "grad_norm": 0.5426959849480466,
+      "learning_rate": 1.6448422127361707e-05,
+      "loss": 0.0679,
+      "step": 721
+    },
+    {
+      "epoch": 1.1107692307692307,
+      "grad_norm": 0.4014572952029352,
+      "learning_rate": 1.6439182158210468e-05,
+      "loss": 0.0682,
+      "step": 722
+    },
+    {
+      "epoch": 1.1123076923076922,
+      "grad_norm": 0.37264714876106975,
+      "learning_rate": 1.642993278785023e-05,
+      "loss": 0.0622,
+      "step": 723
+    },
+    {
+      "epoch": 1.113846153846154,
+      "grad_norm": 0.4085985468005736,
+      "learning_rate": 1.642067402978508e-05,
+      "loss": 0.0708,
+      "step": 724
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.3741294371749341,
+      "learning_rate": 1.64114058975328e-05,
+      "loss": 0.0631,
+      "step": 725
+    },
+    {
+      "epoch": 1.116923076923077,
+      "grad_norm": 0.45876871469434627,
+      "learning_rate": 1.640212840462488e-05,
+      "loss": 0.0663,
+      "step": 726
+    },
+    {
+      "epoch": 1.1184615384615384,
+      "grad_norm": 0.3422472176749258,
+      "learning_rate": 1.639284156460646e-05,
+      "loss": 0.0613,
+      "step": 727
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 0.39030560028876443,
+      "learning_rate": 1.6383545391036327e-05,
+      "loss": 0.0659,
+      "step": 728
+    },
+    {
+      "epoch": 1.1215384615384616,
+      "grad_norm": 0.32978927481973175,
+      "learning_rate": 1.63742398974869e-05,
+      "loss": 0.0583,
+      "step": 729
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 0.3669713631154274,
+      "learning_rate": 1.63649250975442e-05,
+      "loss": 0.0723,
+      "step": 730
+    },
+    {
+      "epoch": 1.1246153846153846,
+      "grad_norm": 0.35663226905770323,
+      "learning_rate": 1.6355601004807856e-05,
+      "loss": 0.0687,
+      "step": 731
+    },
+    {
+      "epoch": 1.126153846153846,
+      "grad_norm": 0.39492116903405033,
+      "learning_rate": 1.6346267632891027e-05,
+      "loss": 0.0705,
+      "step": 732
+    },
+    {
+      "epoch": 1.1276923076923078,
+      "grad_norm": 0.36069535113212814,
+      "learning_rate": 1.6336924995420453e-05,
+      "loss": 0.0634,
+      "step": 733
+    },
+    {
+      "epoch": 1.1292307692307693,
+      "grad_norm": 0.4776466189937346,
+      "learning_rate": 1.6327573106036384e-05,
+      "loss": 0.0772,
+      "step": 734
+    },
+    {
+      "epoch": 1.1307692307692307,
+      "grad_norm": 0.3872373026437133,
+      "learning_rate": 1.6318211978392588e-05,
+      "loss": 0.0729,
+      "step": 735
+    },
+    {
+      "epoch": 1.1323076923076922,
+      "grad_norm": 0.3782947210886841,
+      "learning_rate": 1.630884162615631e-05,
+      "loss": 0.0698,
+      "step": 736
+    },
+    {
+      "epoch": 1.1338461538461537,
+      "grad_norm": 0.38313732578124937,
+      "learning_rate": 1.6299462063008272e-05,
+      "loss": 0.0659,
+      "step": 737
+    },
+    {
+      "epoch": 1.1353846153846154,
+      "grad_norm": 0.32705882716559326,
+      "learning_rate": 1.6290073302642637e-05,
+      "loss": 0.0649,
+      "step": 738
+    },
+    {
+      "epoch": 1.136923076923077,
+      "grad_norm": 0.3322452475653334,
+      "learning_rate": 1.6280675358767005e-05,
+      "loss": 0.0675,
+      "step": 739
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 0.3711248581770431,
+      "learning_rate": 1.6271268245102377e-05,
+      "loss": 0.0737,
+      "step": 740
+    },
+    {
+      "epoch": 1.1400000000000001,
+      "grad_norm": 0.39494242888459563,
+      "learning_rate": 1.626185197538314e-05,
+      "loss": 0.0706,
+      "step": 741
+    },
+    {
+      "epoch": 1.1415384615384616,
+      "grad_norm": 0.3116998586738848,
+      "learning_rate": 1.6252426563357054e-05,
+      "loss": 0.0641,
+      "step": 742
+    },
+    {
+      "epoch": 1.143076923076923,
+      "grad_norm": 0.42963979037379485,
+      "learning_rate": 1.6242992022785225e-05,
+      "loss": 0.0683,
+      "step": 743
+    },
+    {
+      "epoch": 1.1446153846153846,
+      "grad_norm": 0.36479367771942905,
+      "learning_rate": 1.623354836744209e-05,
+      "loss": 0.0713,
+      "step": 744
+    },
+    {
+      "epoch": 1.146153846153846,
+      "grad_norm": 0.6468927692309728,
+      "learning_rate": 1.6224095611115385e-05,
+      "loss": 0.0652,
+      "step": 745
+    },
+    {
+      "epoch": 1.1476923076923078,
+      "grad_norm": 0.6063113124725473,
+      "learning_rate": 1.6214633767606142e-05,
+      "loss": 0.0746,
+      "step": 746
+    },
+    {
+      "epoch": 1.1492307692307693,
+      "grad_norm": 0.4304453405042257,
+      "learning_rate": 1.620516285072866e-05,
+      "loss": 0.0698,
+      "step": 747
+    },
+    {
+      "epoch": 1.1507692307692308,
+      "grad_norm": 0.317379420605374,
+      "learning_rate": 1.6195682874310473e-05,
+      "loss": 0.0647,
+      "step": 748
+    },
+    {
+      "epoch": 1.1523076923076923,
+      "grad_norm": 0.39131831324165883,
+      "learning_rate": 1.6186193852192356e-05,
+      "loss": 0.0755,
+      "step": 749
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.35798460402065585,
+      "learning_rate": 1.617669579822829e-05,
+      "loss": 0.0746,
+      "step": 750
+    },
+    {
+      "epoch": 1.1553846153846155,
+      "grad_norm": 0.3502631169305045,
+      "learning_rate": 1.6167188726285433e-05,
+      "loss": 0.0697,
+      "step": 751
+    },
+    {
+      "epoch": 1.156923076923077,
+      "grad_norm": 0.40987521557285755,
+      "learning_rate": 1.6157672650244113e-05,
+      "loss": 0.0798,
+      "step": 752
+    },
+    {
+      "epoch": 1.1584615384615384,
+      "grad_norm": 0.352033824405812,
+      "learning_rate": 1.6148147583997813e-05,
+      "loss": 0.0678,
+      "step": 753
+    },
+    {
+      "epoch": 1.16,
+      "grad_norm": 0.4345817539730366,
+      "learning_rate": 1.6138613541453127e-05,
+      "loss": 0.0728,
+      "step": 754
+    },
+    {
+      "epoch": 1.1615384615384616,
+      "grad_norm": 0.35074866147767675,
+      "learning_rate": 1.6129070536529767e-05,
+      "loss": 0.0687,
+      "step": 755
+    },
+    {
+      "epoch": 1.1630769230769231,
+      "grad_norm": 0.36887909746623887,
+      "learning_rate": 1.611951858316052e-05,
+      "loss": 0.072,
+      "step": 756
+    },
+    {
+      "epoch": 1.1646153846153846,
+      "grad_norm": 0.2983907917380261,
+      "learning_rate": 1.6109957695291246e-05,
+      "loss": 0.0685,
+      "step": 757
+    },
+    {
+      "epoch": 1.166153846153846,
+      "grad_norm": 0.3934269192468973,
+      "learning_rate": 1.610038788688084e-05,
+      "loss": 0.0677,
+      "step": 758
+    },
+    {
+      "epoch": 1.1676923076923078,
+      "grad_norm": 0.37318389733402674,
+      "learning_rate": 1.6090809171901237e-05,
+      "loss": 0.0718,
+      "step": 759
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 0.38645652176143314,
+      "learning_rate": 1.6081221564337356e-05,
+      "loss": 0.0687,
+      "step": 760
+    },
+    {
+      "epoch": 1.1707692307692308,
+      "grad_norm": 0.38329410093067623,
+      "learning_rate": 1.6071625078187113e-05,
+      "loss": 0.0666,
+      "step": 761
+    },
+    {
+      "epoch": 1.1723076923076923,
+      "grad_norm": 0.30301582024440277,
+      "learning_rate": 1.6062019727461384e-05,
+      "loss": 0.0671,
+      "step": 762
+    },
+    {
+      "epoch": 1.1738461538461538,
+      "grad_norm": 0.32424463849184526,
+      "learning_rate": 1.605240552618398e-05,
+      "loss": 0.0715,
+      "step": 763
+    },
+    {
+      "epoch": 1.1753846153846155,
+      "grad_norm": 0.35922202384394764,
+      "learning_rate": 1.6042782488391644e-05,
+      "loss": 0.0732,
+      "step": 764
+    },
+    {
+      "epoch": 1.176923076923077,
+      "grad_norm": 0.3647659403794086,
+      "learning_rate": 1.603315062813401e-05,
+      "loss": 0.0707,
+      "step": 765
+    },
+    {
+      "epoch": 1.1784615384615384,
+      "grad_norm": 0.34982427128139343,
+      "learning_rate": 1.6023509959473608e-05,
+      "loss": 0.0725,
+      "step": 766
+    },
+    {
+      "epoch": 1.18,
+      "grad_norm": 0.427718797984176,
+      "learning_rate": 1.601386049648581e-05,
+      "loss": 0.0737,
+      "step": 767
+    },
+    {
+      "epoch": 1.1815384615384614,
+      "grad_norm": 0.3924968355298374,
+      "learning_rate": 1.6004202253258844e-05,
+      "loss": 0.0649,
+      "step": 768
+    },
+    {
+      "epoch": 1.1830769230769231,
+      "grad_norm": 0.30260539928386654,
+      "learning_rate": 1.5994535243893742e-05,
+      "loss": 0.0695,
+      "step": 769
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 0.35596304966031656,
+      "learning_rate": 1.5984859482504347e-05,
+      "loss": 0.067,
+      "step": 770
+    },
+    {
+      "epoch": 1.1861538461538461,
+      "grad_norm": 0.3129839541191524,
+      "learning_rate": 1.5975174983217273e-05,
+      "loss": 0.069,
+      "step": 771
+    },
+    {
+      "epoch": 1.1876923076923076,
+      "grad_norm": 0.3674547487152307,
+      "learning_rate": 1.5965481760171897e-05,
+      "loss": 0.0631,
+      "step": 772
+    },
+    {
+      "epoch": 1.1892307692307693,
+      "grad_norm": 0.3335296888041543,
+      "learning_rate": 1.5955779827520327e-05,
+      "loss": 0.0641,
+      "step": 773
+    },
+    {
+      "epoch": 1.1907692307692308,
+      "grad_norm": 0.39407159089071714,
+      "learning_rate": 1.5946069199427387e-05,
+      "loss": 0.071,
+      "step": 774
+    },
+    {
+      "epoch": 1.1923076923076923,
+      "grad_norm": 0.37085530721330084,
+      "learning_rate": 1.5936349890070602e-05,
+      "loss": 0.0733,
+      "step": 775
+    },
+    {
+      "epoch": 1.1938461538461538,
+      "grad_norm": 0.3065421660271998,
+      "learning_rate": 1.592662191364017e-05,
+      "loss": 0.0646,
+      "step": 776
+    },
+    {
+      "epoch": 1.1953846153846155,
+      "grad_norm": 0.4157113450091342,
+      "learning_rate": 1.5916885284338937e-05,
+      "loss": 0.0719,
+      "step": 777
+    },
+    {
+      "epoch": 1.196923076923077,
+      "grad_norm": 0.37188603187355035,
+      "learning_rate": 1.5907140016382385e-05,
+      "loss": 0.0682,
+      "step": 778
+    },
+    {
+      "epoch": 1.1984615384615385,
+      "grad_norm": 0.34699200375559375,
+      "learning_rate": 1.5897386123998613e-05,
+      "loss": 0.0745,
+      "step": 779
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.31961705854482836,
+      "learning_rate": 1.588762362142831e-05,
+      "loss": 0.0647,
+      "step": 780
+    },
+    {
+      "epoch": 1.2015384615384614,
+      "grad_norm": 0.35665468708944836,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.0657,
+      "step": 781
+    },
+    {
+      "epoch": 1.2030769230769232,
+      "grad_norm": 0.3379980896855412,
+      "learning_rate": 1.586807284275369e-05,
+      "loss": 0.0631,
+      "step": 782
+    },
+    {
+      "epoch": 1.2046153846153846,
+      "grad_norm": 0.398246969459009,
+      "learning_rate": 1.5858284595193514e-05,
+      "loss": 0.0732,
+      "step": 783
+    },
+    {
+      "epoch": 1.2061538461538461,
+      "grad_norm": 0.34921906662522734,
+      "learning_rate": 1.584848779453506e-05,
+      "loss": 0.0627,
+      "step": 784
+    },
+    {
+      "epoch": 1.2076923076923076,
+      "grad_norm": 0.37101620781077305,
+      "learning_rate": 1.5838682455081657e-05,
+      "loss": 0.0621,
+      "step": 785
+    },
+    {
+      "epoch": 1.209230769230769,
+      "grad_norm": 0.341332526505076,
+      "learning_rate": 1.5828868591149104e-05,
+      "loss": 0.063,
+      "step": 786
+    },
+    {
+      "epoch": 1.2107692307692308,
+      "grad_norm": 0.3041915924214054,
+      "learning_rate": 1.581904621706565e-05,
+      "loss": 0.0678,
+      "step": 787
+    },
+    {
+      "epoch": 1.2123076923076923,
+      "grad_norm": 0.36665714928540116,
+      "learning_rate": 1.580921534717196e-05,
+      "loss": 0.0743,
+      "step": 788
+    },
+    {
+      "epoch": 1.2138461538461538,
+      "grad_norm": 0.3377775940167995,
+      "learning_rate": 1.5799375995821116e-05,
+      "loss": 0.064,
+      "step": 789
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 0.35353779842042976,
+      "learning_rate": 1.5789528177378574e-05,
+      "loss": 0.079,
+      "step": 790
+    },
+    {
+      "epoch": 1.216923076923077,
+      "grad_norm": 0.3232778165659884,
+      "learning_rate": 1.577967190622215e-05,
+      "loss": 0.066,
+      "step": 791
+    },
+    {
+      "epoch": 1.2184615384615385,
+      "grad_norm": 0.3577677354201108,
+      "learning_rate": 1.5769807196742008e-05,
+      "loss": 0.0653,
+      "step": 792
+    },
+    {
+      "epoch": 1.22,
+      "grad_norm": 0.31783041296834197,
+      "learning_rate": 1.5759934063340627e-05,
+      "loss": 0.0706,
+      "step": 793
+    },
+    {
+      "epoch": 1.2215384615384615,
+      "grad_norm": 0.40205477740857387,
+      "learning_rate": 1.575005252043279e-05,
+      "loss": 0.0725,
+      "step": 794
+    },
+    {
+      "epoch": 1.2230769230769232,
+      "grad_norm": 0.3925627479595986,
+      "learning_rate": 1.5740162582445545e-05,
+      "loss": 0.07,
+      "step": 795
+    },
+    {
+      "epoch": 1.2246153846153847,
+      "grad_norm": 0.3311986658898681,
+      "learning_rate": 1.5730264263818212e-05,
+      "loss": 0.0762,
+      "step": 796
+    },
+    {
+      "epoch": 1.2261538461538461,
+      "grad_norm": 0.35012400117324377,
+      "learning_rate": 1.5720357579002346e-05,
+      "loss": 0.0621,
+      "step": 797
+    },
+    {
+      "epoch": 1.2276923076923076,
+      "grad_norm": 0.351604049632638,
+      "learning_rate": 1.5710442542461705e-05,
+      "loss": 0.0638,
+      "step": 798
+    },
+    {
+      "epoch": 1.2292307692307691,
+      "grad_norm": 0.356583134009754,
+      "learning_rate": 1.5700519168672248e-05,
+      "loss": 0.0713,
+      "step": 799
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.3285074874868023,
+      "learning_rate": 1.5690587472122104e-05,
+      "loss": 0.0643,
+      "step": 800
+    },
+    {
+      "epoch": 1.2323076923076923,
+      "grad_norm": 0.3450768181316114,
+      "learning_rate": 1.568064746731156e-05,
+      "loss": 0.0633,
+      "step": 801
+    },
+    {
+      "epoch": 1.2338461538461538,
+      "grad_norm": 0.3539947251115169,
+      "learning_rate": 1.5670699168753022e-05,
+      "loss": 0.0662,
+      "step": 802
+    },
+    {
+      "epoch": 1.2353846153846153,
+      "grad_norm": 0.32288987773337546,
+      "learning_rate": 1.5660742590971014e-05,
+      "loss": 0.0679,
+      "step": 803
+    },
+    {
+      "epoch": 1.236923076923077,
+      "grad_norm": 0.5408106149836094,
+      "learning_rate": 1.5650777748502144e-05,
+      "loss": 0.0739,
+      "step": 804
+    },
+    {
+      "epoch": 1.2384615384615385,
+      "grad_norm": 0.3333851374522989,
+      "learning_rate": 1.5640804655895086e-05,
+      "loss": 0.0666,
+      "step": 805
+    },
+    {
+      "epoch": 1.24,
+      "grad_norm": 0.3514156125809535,
+      "learning_rate": 1.5630823327710558e-05,
+      "loss": 0.0609,
+      "step": 806
+    },
+    {
+      "epoch": 1.2415384615384615,
+      "grad_norm": 0.4355584652770312,
+      "learning_rate": 1.5620833778521306e-05,
+      "loss": 0.0698,
+      "step": 807
+    },
+    {
+      "epoch": 1.2430769230769232,
+      "grad_norm": 0.3863399662303979,
+      "learning_rate": 1.561083602291208e-05,
+      "loss": 0.0747,
+      "step": 808
+    },
+    {
+      "epoch": 1.2446153846153847,
+      "grad_norm": 0.4272070354943623,
+      "learning_rate": 1.5600830075479604e-05,
+      "loss": 0.0625,
+      "step": 809
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 0.34034969408428656,
+      "learning_rate": 1.559081595083256e-05,
+      "loss": 0.0611,
+      "step": 810
+    },
+    {
+      "epoch": 1.2476923076923077,
+      "grad_norm": 0.42893862551923334,
+      "learning_rate": 1.5580793663591583e-05,
+      "loss": 0.0669,
+      "step": 811
+    },
+    {
+      "epoch": 1.2492307692307691,
+      "grad_norm": 0.37930257242842513,
+      "learning_rate": 1.557076322838922e-05,
+      "loss": 0.0699,
+      "step": 812
+    },
+    {
+      "epoch": 1.2507692307692309,
+      "grad_norm": 0.33145455304026866,
+      "learning_rate": 1.5560724659869905e-05,
+      "loss": 0.0681,
+      "step": 813
+    },
+    {
+      "epoch": 1.2523076923076923,
+      "grad_norm": 0.4059667858998905,
+      "learning_rate": 1.555067797268995e-05,
+      "loss": 0.068,
+      "step": 814
+    },
+    {
+      "epoch": 1.2538461538461538,
+      "grad_norm": 0.3467956370682722,
+      "learning_rate": 1.5540623181517532e-05,
+      "loss": 0.0659,
+      "step": 815
+    },
+    {
+      "epoch": 1.2553846153846153,
+      "grad_norm": 0.5088686209429053,
+      "learning_rate": 1.5530560301032644e-05,
+      "loss": 0.0795,
+      "step": 816
+    },
+    {
+      "epoch": 1.2569230769230768,
+      "grad_norm": 0.3799873270111144,
+      "learning_rate": 1.5520489345927095e-05,
+      "loss": 0.0691,
+      "step": 817
+    },
+    {
+      "epoch": 1.2584615384615385,
+      "grad_norm": 0.41391636034513485,
+      "learning_rate": 1.551041033090449e-05,
+      "loss": 0.0721,
+      "step": 818
+    },
+    {
+      "epoch": 1.26,
+      "grad_norm": 0.40772591130685176,
+      "learning_rate": 1.5500323270680194e-05,
+      "loss": 0.0681,
+      "step": 819
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 0.3349423110799659,
+      "learning_rate": 1.549022817998132e-05,
+      "loss": 0.0623,
+      "step": 820
+    },
+    {
+      "epoch": 1.2630769230769232,
+      "grad_norm": 0.5203080319006015,
+      "learning_rate": 1.5480125073546705e-05,
+      "loss": 0.0659,
+      "step": 821
+    },
+    {
+      "epoch": 1.2646153846153847,
+      "grad_norm": 0.42351756936892165,
+      "learning_rate": 1.5470013966126886e-05,
+      "loss": 0.0702,
+      "step": 822
+    },
+    {
+      "epoch": 1.2661538461538462,
+      "grad_norm": 0.3483945108051493,
+      "learning_rate": 1.5459894872484083e-05,
+      "loss": 0.0636,
+      "step": 823
+    },
+    {
+      "epoch": 1.2676923076923077,
+      "grad_norm": 0.37839615755712913,
+      "learning_rate": 1.5449767807392184e-05,
+      "loss": 0.0676,
+      "step": 824
+    },
+    {
+      "epoch": 1.2692307692307692,
+      "grad_norm": 0.4465419577943629,
+      "learning_rate": 1.5439632785636707e-05,
+      "loss": 0.0631,
+      "step": 825
+    },
+    {
+      "epoch": 1.2707692307692309,
+      "grad_norm": 0.39462290024075386,
+      "learning_rate": 1.542948982201479e-05,
+      "loss": 0.0634,
+      "step": 826
+    },
+    {
+      "epoch": 1.2723076923076924,
+      "grad_norm": 0.3797179172041511,
+      "learning_rate": 1.5419338931335155e-05,
+      "loss": 0.0699,
+      "step": 827
+    },
+    {
+      "epoch": 1.2738461538461539,
+      "grad_norm": 0.3416577637686262,
+      "learning_rate": 1.5409180128418123e-05,
+      "loss": 0.0636,
+      "step": 828
+    },
+    {
+      "epoch": 1.2753846153846153,
+      "grad_norm": 0.4613004722112054,
+      "learning_rate": 1.539901342809554e-05,
+      "loss": 0.0684,
+      "step": 829
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 0.42702108586314186,
+      "learning_rate": 1.5388838845210798e-05,
+      "loss": 0.0792,
+      "step": 830
+    },
+    {
+      "epoch": 1.2784615384615385,
+      "grad_norm": 0.3530542594699597,
+      "learning_rate": 1.5378656394618788e-05,
+      "loss": 0.0647,
+      "step": 831
+    },
+    {
+      "epoch": 1.28,
+      "grad_norm": 0.43241017204879106,
+      "learning_rate": 1.5368466091185893e-05,
+      "loss": 0.0781,
+      "step": 832
+    },
+    {
+      "epoch": 1.2815384615384615,
+      "grad_norm": 0.5132567987415939,
+      "learning_rate": 1.5358267949789968e-05,
+      "loss": 0.0728,
+      "step": 833
+    },
+    {
+      "epoch": 1.283076923076923,
+      "grad_norm": 0.36683064622627476,
+      "learning_rate": 1.5348061985320298e-05,
+      "loss": 0.0725,
+      "step": 834
+    },
+    {
+      "epoch": 1.2846153846153845,
+      "grad_norm": 0.33492481470642155,
+      "learning_rate": 1.53378482126776e-05,
+      "loss": 0.0644,
+      "step": 835
+    },
+    {
+      "epoch": 1.2861538461538462,
+      "grad_norm": 0.3210179795426996,
+      "learning_rate": 1.5327626646773975e-05,
+      "loss": 0.0694,
+      "step": 836
+    },
+    {
+      "epoch": 1.2876923076923077,
+      "grad_norm": 0.3454605703445639,
+      "learning_rate": 1.5317397302532933e-05,
+      "loss": 0.063,
+      "step": 837
+    },
+    {
+      "epoch": 1.2892307692307692,
+      "grad_norm": 0.36111805478759756,
+      "learning_rate": 1.530716019488931e-05,
+      "loss": 0.0729,
+      "step": 838
+    },
+    {
+      "epoch": 1.290769230769231,
+      "grad_norm": 0.4150741361759226,
+      "learning_rate": 1.529691533878929e-05,
+      "loss": 0.0692,
+      "step": 839
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 0.44151832030522525,
+      "learning_rate": 1.528666274919037e-05,
+      "loss": 0.0668,
+      "step": 840
+    },
+    {
+      "epoch": 1.2938461538461539,
+      "grad_norm": 0.39281808968582055,
+      "learning_rate": 1.527640244106133e-05,
+      "loss": 0.0614,
+      "step": 841
+    },
+    {
+      "epoch": 1.2953846153846154,
+      "grad_norm": 0.3355371699856118,
+      "learning_rate": 1.526613442938223e-05,
+      "loss": 0.066,
+      "step": 842
+    },
+    {
+      "epoch": 1.2969230769230768,
+      "grad_norm": 0.3343752965517015,
+      "learning_rate": 1.5255858729144368e-05,
+      "loss": 0.0618,
+      "step": 843
+    },
+    {
+      "epoch": 1.2984615384615386,
+      "grad_norm": 0.4219851770072625,
+      "learning_rate": 1.5245575355350273e-05,
+      "loss": 0.0706,
+      "step": 844
+    },
+    {
+      "epoch": 1.3,
+      "grad_norm": 0.47236267011925603,
+      "learning_rate": 1.5235284323013674e-05,
+      "loss": 0.0768,
+      "step": 845
+    },
+    {
+      "epoch": 1.3015384615384615,
+      "grad_norm": 0.3160781099772521,
+      "learning_rate": 1.5224985647159489e-05,
+      "loss": 0.0677,
+      "step": 846
+    },
+    {
+      "epoch": 1.303076923076923,
+      "grad_norm": 0.4211840324694599,
+      "learning_rate": 1.5214679342823786e-05,
+      "loss": 0.0703,
+      "step": 847
+    },
+    {
+      "epoch": 1.3046153846153845,
+      "grad_norm": 0.4382171224080807,
+      "learning_rate": 1.5204365425053773e-05,
+      "loss": 0.0728,
+      "step": 848
+    },
+    {
+      "epoch": 1.3061538461538462,
+      "grad_norm": 0.3730723749029246,
+      "learning_rate": 1.5194043908907774e-05,
+      "loss": 0.0644,
+      "step": 849
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.43669340037201293,
+      "learning_rate": 1.518371480945521e-05,
+      "loss": 0.0725,
+      "step": 850
+    },
+    {
+      "epoch": 1.3092307692307692,
+      "grad_norm": 0.29956287148615507,
+      "learning_rate": 1.5173378141776569e-05,
+      "loss": 0.0595,
+      "step": 851
+    },
+    {
+      "epoch": 1.3107692307692307,
+      "grad_norm": 0.4010201519686623,
+      "learning_rate": 1.5163033920963393e-05,
+      "loss": 0.0715,
+      "step": 852
+    },
+    {
+      "epoch": 1.3123076923076922,
+      "grad_norm": 0.3284401287146176,
+      "learning_rate": 1.515268216211825e-05,
+      "loss": 0.0649,
+      "step": 853
+    },
+    {
+      "epoch": 1.3138461538461539,
+      "grad_norm": 0.36834718137576217,
+      "learning_rate": 1.5142322880354706e-05,
+      "loss": 0.0761,
+      "step": 854
+    },
+    {
+      "epoch": 1.3153846153846154,
+      "grad_norm": 0.4038185340311018,
+      "learning_rate": 1.5131956090797326e-05,
+      "loss": 0.0713,
+      "step": 855
+    },
+    {
+      "epoch": 1.3169230769230769,
+      "grad_norm": 0.3485378538805922,
+      "learning_rate": 1.5121581808581623e-05,
+      "loss": 0.0721,
+      "step": 856
+    },
+    {
+      "epoch": 1.3184615384615386,
+      "grad_norm": 0.324099392540781,
+      "learning_rate": 1.5111200048854055e-05,
+      "loss": 0.0704,
+      "step": 857
+    },
+    {
+      "epoch": 1.32,
+      "grad_norm": 0.3958228511004757,
+      "learning_rate": 1.5100810826771997e-05,
+      "loss": 0.0665,
+      "step": 858
+    },
+    {
+      "epoch": 1.3215384615384616,
+      "grad_norm": 0.4382455185601089,
+      "learning_rate": 1.5090414157503715e-05,
+      "loss": 0.076,
+      "step": 859
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 0.3260556756343216,
+      "learning_rate": 1.5080010056228353e-05,
+      "loss": 0.0671,
+      "step": 860
+    },
+    {
+      "epoch": 1.3246153846153845,
+      "grad_norm": 0.3038142480938741,
+      "learning_rate": 1.5069598538135905e-05,
+      "loss": 0.0681,
+      "step": 861
+    },
+    {
+      "epoch": 1.3261538461538462,
+      "grad_norm": 0.3817393937929787,
+      "learning_rate": 1.505917961842719e-05,
+      "loss": 0.07,
+      "step": 862
+    },
+    {
+      "epoch": 1.3276923076923077,
+      "grad_norm": 0.43653178960329847,
+      "learning_rate": 1.504875331231384e-05,
+      "loss": 0.0682,
+      "step": 863
+    },
+    {
+      "epoch": 1.3292307692307692,
+      "grad_norm": 0.4191601359367802,
+      "learning_rate": 1.5038319635018264e-05,
+      "loss": 0.0709,
+      "step": 864
+    },
+    {
+      "epoch": 1.3307692307692307,
+      "grad_norm": 0.40482174127743453,
+      "learning_rate": 1.5027878601773633e-05,
+      "loss": 0.0681,
+      "step": 865
+    },
+    {
+      "epoch": 1.3323076923076922,
+      "grad_norm": 0.39987356115193523,
+      "learning_rate": 1.5017430227823867e-05,
+      "loss": 0.0653,
+      "step": 866
+    },
+    {
+      "epoch": 1.333846153846154,
+      "grad_norm": 0.36706667375847013,
+      "learning_rate": 1.5006974528423585e-05,
+      "loss": 0.0631,
+      "step": 867
+    },
+    {
+      "epoch": 1.3353846153846154,
+      "grad_norm": 0.3710865070928732,
+      "learning_rate": 1.4996511518838129e-05,
+      "loss": 0.0608,
+      "step": 868
+    },
+    {
+      "epoch": 1.3369230769230769,
+      "grad_norm": 0.5238263977692204,
+      "learning_rate": 1.4986041214343487e-05,
+      "loss": 0.0739,
+      "step": 869
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 0.38666104678955204,
+      "learning_rate": 1.4975563630226311e-05,
+      "loss": 0.0515,
+      "step": 870
+    },
+    {
+      "epoch": 1.34,
+      "grad_norm": 0.31886781420723,
+      "learning_rate": 1.4965078781783882e-05,
+      "loss": 0.0649,
+      "step": 871
+    },
+    {
+      "epoch": 1.3415384615384616,
+      "grad_norm": 0.32582039756760706,
+      "learning_rate": 1.4954586684324077e-05,
+      "loss": 0.0682,
+      "step": 872
+    },
+    {
+      "epoch": 1.343076923076923,
+      "grad_norm": 0.4323139162984884,
+      "learning_rate": 1.494408735316537e-05,
+      "loss": 0.0722,
+      "step": 873
+    },
+    {
+      "epoch": 1.3446153846153845,
+      "grad_norm": 0.4522628240441855,
+      "learning_rate": 1.4933580803636787e-05,
+      "loss": 0.0641,
+      "step": 874
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.38114193229980947,
+      "learning_rate": 1.4923067051077893e-05,
+      "loss": 0.0724,
+      "step": 875
+    },
+    {
+      "epoch": 1.3476923076923077,
+      "grad_norm": 0.4247141871033444,
+      "learning_rate": 1.4912546110838775e-05,
+      "loss": 0.0713,
+      "step": 876
+    },
+    {
+      "epoch": 1.3492307692307692,
+      "grad_norm": 0.4058780791287694,
+      "learning_rate": 1.490201799828001e-05,
+      "loss": 0.0741,
+      "step": 877
+    },
+    {
+      "epoch": 1.3507692307692307,
+      "grad_norm": 0.43398667126598395,
+      "learning_rate": 1.4891482728772645e-05,
+      "loss": 0.0694,
+      "step": 878
+    },
+    {
+      "epoch": 1.3523076923076922,
+      "grad_norm": 0.4177065232740013,
+      "learning_rate": 1.4880940317698182e-05,
+      "loss": 0.0731,
+      "step": 879
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 0.3284355368343791,
+      "learning_rate": 1.4870390780448545e-05,
+      "loss": 0.0747,
+      "step": 880
+    },
+    {
+      "epoch": 1.3553846153846154,
+      "grad_norm": 0.44954781903955254,
+      "learning_rate": 1.485983413242606e-05,
+      "loss": 0.0628,
+      "step": 881
+    },
+    {
+      "epoch": 1.356923076923077,
+      "grad_norm": 0.5129727422741225,
+      "learning_rate": 1.4849270389043444e-05,
+      "loss": 0.0687,
+      "step": 882
+    },
+    {
+      "epoch": 1.3584615384615384,
+      "grad_norm": 0.4467443171166444,
+      "learning_rate": 1.4838699565723764e-05,
+      "loss": 0.0652,
+      "step": 883
+    },
+    {
+      "epoch": 1.3599999999999999,
+      "grad_norm": 0.4496929006760978,
+      "learning_rate": 1.4828121677900427e-05,
+      "loss": 0.0744,
+      "step": 884
+    },
+    {
+      "epoch": 1.3615384615384616,
+      "grad_norm": 0.3414293713875311,
+      "learning_rate": 1.4817536741017153e-05,
+      "loss": 0.0728,
+      "step": 885
+    },
+    {
+      "epoch": 1.363076923076923,
+      "grad_norm": 0.316120367911126,
+      "learning_rate": 1.4806944770527958e-05,
+      "loss": 0.0666,
+      "step": 886
+    },
+    {
+      "epoch": 1.3646153846153846,
+      "grad_norm": 0.40460300362699714,
+      "learning_rate": 1.479634578189712e-05,
+      "loss": 0.0727,
+      "step": 887
+    },
+    {
+      "epoch": 1.3661538461538463,
+      "grad_norm": 0.45659447717946505,
+      "learning_rate": 1.4785739790599174e-05,
+      "loss": 0.0782,
+      "step": 888
+    },
+    {
+      "epoch": 1.3676923076923078,
+      "grad_norm": 0.40368455430062866,
+      "learning_rate": 1.4775126812118865e-05,
+      "loss": 0.0704,
+      "step": 889
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 0.35547903664705943,
+      "learning_rate": 1.4764506861951151e-05,
+      "loss": 0.0774,
+      "step": 890
+    },
+    {
+      "epoch": 1.3707692307692307,
+      "grad_norm": 0.3259048929714102,
+      "learning_rate": 1.4753879955601162e-05,
+      "loss": 0.0568,
+      "step": 891
+    },
+    {
+      "epoch": 1.3723076923076922,
+      "grad_norm": 0.33398368713521703,
+      "learning_rate": 1.474324610858419e-05,
+      "loss": 0.0696,
+      "step": 892
+    },
+    {
+      "epoch": 1.373846153846154,
+      "grad_norm": 0.3975414141870634,
+      "learning_rate": 1.4732605336425651e-05,
+      "loss": 0.0653,
+      "step": 893
+    },
+    {
+      "epoch": 1.3753846153846154,
+      "grad_norm": 0.3695667545172178,
+      "learning_rate": 1.472195765466108e-05,
+      "loss": 0.0657,
+      "step": 894
+    },
+    {
+      "epoch": 1.376923076923077,
+      "grad_norm": 0.3070295149212851,
+      "learning_rate": 1.4711303078836098e-05,
+      "loss": 0.0607,
+      "step": 895
+    },
+    {
+      "epoch": 1.3784615384615384,
+      "grad_norm": 0.43373009062750495,
+      "learning_rate": 1.4700641624506392e-05,
+      "loss": 0.077,
+      "step": 896
+    },
+    {
+      "epoch": 1.38,
+      "grad_norm": 0.3207986109640768,
+      "learning_rate": 1.4689973307237687e-05,
+      "loss": 0.0663,
+      "step": 897
+    },
+    {
+      "epoch": 1.3815384615384616,
+      "grad_norm": 0.33029490975452497,
+      "learning_rate": 1.4679298142605735e-05,
+      "loss": 0.0661,
+      "step": 898
+    },
+    {
+      "epoch": 1.383076923076923,
+      "grad_norm": 0.39797201272606875,
+      "learning_rate": 1.466861614619628e-05,
+      "loss": 0.0683,
+      "step": 899
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.3569032293188448,
+      "learning_rate": 1.465792733360504e-05,
+      "loss": 0.0712,
+      "step": 900
+    },
+    {
+      "epoch": 1.3861538461538463,
+      "grad_norm": 0.33179043133017183,
+      "learning_rate": 1.4647231720437687e-05,
+      "loss": 0.076,
+      "step": 901
+    },
+    {
+      "epoch": 1.3876923076923076,
+      "grad_norm": 0.3185134355526598,
+      "learning_rate": 1.4636529322309825e-05,
+      "loss": 0.0716,
+      "step": 902
+    },
+    {
+      "epoch": 1.3892307692307693,
+      "grad_norm": 0.38927875322344885,
+      "learning_rate": 1.4625820154846953e-05,
+      "loss": 0.0607,
+      "step": 903
+    },
+    {
+      "epoch": 1.3907692307692308,
+      "grad_norm": 0.3592258244516518,
+      "learning_rate": 1.4615104233684467e-05,
+      "loss": 0.0636,
+      "step": 904
+    },
+    {
+      "epoch": 1.3923076923076922,
+      "grad_norm": 0.440027348689472,
+      "learning_rate": 1.4604381574467616e-05,
+      "loss": 0.0812,
+      "step": 905
+    },
+    {
+      "epoch": 1.393846153846154,
+      "grad_norm": 0.31797170088925414,
+      "learning_rate": 1.4593652192851487e-05,
+      "loss": 0.0624,
+      "step": 906
+    },
+    {
+      "epoch": 1.3953846153846154,
+      "grad_norm": 0.44417520359272944,
+      "learning_rate": 1.4582916104500977e-05,
+      "loss": 0.0734,
+      "step": 907
+    },
+    {
+      "epoch": 1.396923076923077,
+      "grad_norm": 0.34291352445561796,
+      "learning_rate": 1.457217332509079e-05,
+      "loss": 0.0684,
+      "step": 908
+    },
+    {
+      "epoch": 1.3984615384615384,
+      "grad_norm": 0.3078336428229946,
+      "learning_rate": 1.4561423870305383e-05,
+      "loss": 0.0594,
+      "step": 909
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 0.45773936133687876,
+      "learning_rate": 1.4550667755838965e-05,
+      "loss": 0.0735,
+      "step": 910
+    },
+    {
+      "epoch": 1.4015384615384616,
+      "grad_norm": 0.3093789239621972,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.0665,
+      "step": 911
+    },
+    {
+      "epoch": 1.403076923076923,
+      "grad_norm": 0.33556510387782423,
+      "learning_rate": 1.4529135610688529e-05,
+      "loss": 0.067,
+      "step": 912
+    },
+    {
+      "epoch": 1.4046153846153846,
+      "grad_norm": 0.3821254219317531,
+      "learning_rate": 1.4518359611441452e-05,
+      "loss": 0.0795,
+      "step": 913
+    },
+    {
+      "epoch": 1.406153846153846,
+      "grad_norm": 0.3968625825470183,
+      "learning_rate": 1.4507577015387204e-05,
+      "loss": 0.0694,
+      "step": 914
+    },
+    {
+      "epoch": 1.4076923076923076,
+      "grad_norm": 0.40920318095123365,
+      "learning_rate": 1.4496787838268378e-05,
+      "loss": 0.0751,
+      "step": 915
+    },
+    {
+      "epoch": 1.4092307692307693,
+      "grad_norm": 0.333518429842856,
+      "learning_rate": 1.4485992095837178e-05,
+      "loss": 0.0678,
+      "step": 916
+    },
+    {
+      "epoch": 1.4107692307692308,
+      "grad_norm": 0.4180262330660976,
+      "learning_rate": 1.4475189803855399e-05,
+      "loss": 0.0669,
+      "step": 917
+    },
+    {
+      "epoch": 1.4123076923076923,
+      "grad_norm": 0.3270671609911376,
+      "learning_rate": 1.4464380978094386e-05,
+      "loss": 0.0628,
+      "step": 918
+    },
+    {
+      "epoch": 1.413846153846154,
+      "grad_norm": 0.3321568177478066,
+      "learning_rate": 1.445356563433503e-05,
+      "loss": 0.0675,
+      "step": 919
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 0.4278414076070176,
+      "learning_rate": 1.4442743788367741e-05,
+      "loss": 0.0756,
+      "step": 920
+    },
+    {
+      "epoch": 1.416923076923077,
+      "grad_norm": 0.3573839526102382,
+      "learning_rate": 1.4431915455992416e-05,
+      "loss": 0.0743,
+      "step": 921
+    },
+    {
+      "epoch": 1.4184615384615384,
+      "grad_norm": 0.3965761887911091,
+      "learning_rate": 1.4421080653018426e-05,
+      "loss": 0.0686,
+      "step": 922
+    },
+    {
+      "epoch": 1.42,
+      "grad_norm": 0.3277382293498253,
+      "learning_rate": 1.4410239395264594e-05,
+      "loss": 0.0593,
+      "step": 923
+    },
+    {
+      "epoch": 1.4215384615384616,
+      "grad_norm": 0.413916370460171,
+      "learning_rate": 1.4399391698559153e-05,
+      "loss": 0.057,
+      "step": 924
+    },
+    {
+      "epoch": 1.4230769230769231,
+      "grad_norm": 0.40814754638707723,
+      "learning_rate": 1.438853757873975e-05,
+      "loss": 0.0673,
+      "step": 925
+    },
+    {
+      "epoch": 1.4246153846153846,
+      "grad_norm": 0.37821667270068476,
+      "learning_rate": 1.4377677051653404e-05,
+      "loss": 0.0703,
+      "step": 926
+    },
+    {
+      "epoch": 1.426153846153846,
+      "grad_norm": 0.5102274833966777,
+      "learning_rate": 1.4366810133156495e-05,
+      "loss": 0.076,
+      "step": 927
+    },
+    {
+      "epoch": 1.4276923076923076,
+      "grad_norm": 0.44547602217095333,
+      "learning_rate": 1.4355936839114718e-05,
+      "loss": 0.0747,
+      "step": 928
+    },
+    {
+      "epoch": 1.4292307692307693,
+      "grad_norm": 0.3911040675941954,
+      "learning_rate": 1.43450571854031e-05,
+      "loss": 0.0715,
+      "step": 929
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 0.3796479456399161,
+      "learning_rate": 1.4334171187905928e-05,
+      "loss": 0.0732,
+      "step": 930
+    },
+    {
+      "epoch": 1.4323076923076923,
+      "grad_norm": 0.33940059068204437,
+      "learning_rate": 1.4323278862516774e-05,
+      "loss": 0.0702,
+      "step": 931
+    },
+    {
+      "epoch": 1.4338461538461538,
+      "grad_norm": 0.3509046699774803,
+      "learning_rate": 1.431238022513843e-05,
+      "loss": 0.0657,
+      "step": 932
+    },
+    {
+      "epoch": 1.4353846153846153,
+      "grad_norm": 0.36668044712623865,
+      "learning_rate": 1.430147529168292e-05,
+      "loss": 0.0625,
+      "step": 933
+    },
+    {
+      "epoch": 1.436923076923077,
+      "grad_norm": 0.4281432928035457,
+      "learning_rate": 1.4290564078071445e-05,
+      "loss": 0.0813,
+      "step": 934
+    },
+    {
+      "epoch": 1.4384615384615385,
+      "grad_norm": 0.33710258376419955,
+      "learning_rate": 1.4279646600234388e-05,
+      "loss": 0.068,
+      "step": 935
+    },
+    {
+      "epoch": 1.44,
+      "grad_norm": 0.32354838701141914,
+      "learning_rate": 1.4268722874111265e-05,
+      "loss": 0.0648,
+      "step": 936
+    },
+    {
+      "epoch": 1.4415384615384617,
+      "grad_norm": 0.3021595123301058,
+      "learning_rate": 1.4257792915650728e-05,
+      "loss": 0.0653,
+      "step": 937
+    },
+    {
+      "epoch": 1.4430769230769231,
+      "grad_norm": 0.44802198690971745,
+      "learning_rate": 1.4246856740810517e-05,
+      "loss": 0.0768,
+      "step": 938
+    },
+    {
+      "epoch": 1.4446153846153846,
+      "grad_norm": 0.3593940216420963,
+      "learning_rate": 1.4235914365557455e-05,
+      "loss": 0.0638,
+      "step": 939
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 0.34159209475332464,
+      "learning_rate": 1.4224965805867413e-05,
+      "loss": 0.0643,
+      "step": 940
+    },
+    {
+      "epoch": 1.4476923076923076,
+      "grad_norm": 0.35984267550149907,
+      "learning_rate": 1.4214011077725293e-05,
+      "loss": 0.0723,
+      "step": 941
+    },
+    {
+      "epoch": 1.4492307692307693,
+      "grad_norm": 0.39317799185440994,
+      "learning_rate": 1.4203050197125005e-05,
+      "loss": 0.0643,
+      "step": 942
+    },
+    {
+      "epoch": 1.4507692307692308,
+      "grad_norm": 0.336857468826351,
+      "learning_rate": 1.4192083180069441e-05,
+      "loss": 0.0618,
+      "step": 943
+    },
+    {
+      "epoch": 1.4523076923076923,
+      "grad_norm": 0.35186803870389044,
+      "learning_rate": 1.4181110042570447e-05,
+      "loss": 0.0683,
+      "step": 944
+    },
+    {
+      "epoch": 1.4538461538461538,
+      "grad_norm": 0.29942739165233795,
+      "learning_rate": 1.4170130800648814e-05,
+      "loss": 0.0652,
+      "step": 945
+    },
+    {
+      "epoch": 1.4553846153846153,
+      "grad_norm": 0.34891798063795276,
+      "learning_rate": 1.4159145470334237e-05,
+      "loss": 0.0696,
+      "step": 946
+    },
+    {
+      "epoch": 1.456923076923077,
+      "grad_norm": 0.4077733821140671,
+      "learning_rate": 1.4148154067665305e-05,
+      "loss": 0.0699,
+      "step": 947
+    },
+    {
+      "epoch": 1.4584615384615385,
+      "grad_norm": 0.40284125050375413,
+      "learning_rate": 1.4137156608689469e-05,
+      "loss": 0.0742,
+      "step": 948
+    },
+    {
+      "epoch": 1.46,
+      "grad_norm": 0.42566126627046824,
+      "learning_rate": 1.4126153109463025e-05,
+      "loss": 0.0642,
+      "step": 949
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.32133228937421,
+      "learning_rate": 1.411514358605109e-05,
+      "loss": 0.068,
+      "step": 950
+    },
+    {
+      "epoch": 1.463076923076923,
+      "grad_norm": 0.36895430822571307,
+      "learning_rate": 1.410412805452757e-05,
+      "loss": 0.0667,
+      "step": 951
+    },
+    {
+      "epoch": 1.4646153846153847,
+      "grad_norm": 0.3166854845654635,
+      "learning_rate": 1.4093106530975146e-05,
+      "loss": 0.061,
+      "step": 952
+    },
+    {
+      "epoch": 1.4661538461538461,
+      "grad_norm": 0.5003043211716166,
+      "learning_rate": 1.4082079031485253e-05,
+      "loss": 0.0656,
+      "step": 953
+    },
+    {
+      "epoch": 1.4676923076923076,
+      "grad_norm": 0.39875601093346025,
+      "learning_rate": 1.4071045572158038e-05,
+      "loss": 0.0681,
+      "step": 954
+    },
+    {
+      "epoch": 1.4692307692307693,
+      "grad_norm": 0.3993223750486499,
+      "learning_rate": 1.4060006169102363e-05,
+      "loss": 0.0692,
+      "step": 955
+    },
+    {
+      "epoch": 1.4707692307692308,
+      "grad_norm": 0.3477397845990775,
+      "learning_rate": 1.4048960838435755e-05,
+      "loss": 0.0702,
+      "step": 956
+    },
+    {
+      "epoch": 1.4723076923076923,
+      "grad_norm": 0.4126048706500608,
+      "learning_rate": 1.403790959628441e-05,
+      "loss": 0.0703,
+      "step": 957
+    },
+    {
+      "epoch": 1.4738461538461538,
+      "grad_norm": 0.3012393164802371,
+      "learning_rate": 1.4026852458783141e-05,
+      "loss": 0.0631,
+      "step": 958
+    },
+    {
+      "epoch": 1.4753846153846153,
+      "grad_norm": 0.41825312268062614,
+      "learning_rate": 1.4015789442075376e-05,
+      "loss": 0.0639,
+      "step": 959
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 0.35965621239571177,
+      "learning_rate": 1.4004720562313125e-05,
+      "loss": 0.0631,
+      "step": 960
+    },
+    {
+      "epoch": 1.4784615384615385,
+      "grad_norm": 0.329531529153342,
+      "learning_rate": 1.3993645835656955e-05,
+      "loss": 0.0794,
+      "step": 961
+    },
+    {
+      "epoch": 1.48,
+      "grad_norm": 0.34849163011002293,
+      "learning_rate": 1.3982565278275976e-05,
+      "loss": 0.067,
+      "step": 962
+    },
+    {
+      "epoch": 1.4815384615384615,
+      "grad_norm": 0.3857507607852618,
+      "learning_rate": 1.3971478906347806e-05,
+      "loss": 0.0786,
+      "step": 963
+    },
+    {
+      "epoch": 1.483076923076923,
+      "grad_norm": 0.3354179087450714,
+      "learning_rate": 1.3960386736058552e-05,
+      "loss": 0.0653,
+      "step": 964
+    },
+    {
+      "epoch": 1.4846153846153847,
+      "grad_norm": 0.4517425643119976,
+      "learning_rate": 1.394928878360279e-05,
+      "loss": 0.0696,
+      "step": 965
+    },
+    {
+      "epoch": 1.4861538461538462,
+      "grad_norm": 0.3726999175701167,
+      "learning_rate": 1.3938185065183534e-05,
+      "loss": 0.0703,
+      "step": 966
+    },
+    {
+      "epoch": 1.4876923076923076,
+      "grad_norm": 0.32753752193184116,
+      "learning_rate": 1.3927075597012215e-05,
+      "loss": 0.0688,
+      "step": 967
+    },
+    {
+      "epoch": 1.4892307692307694,
+      "grad_norm": 0.3625776900702585,
+      "learning_rate": 1.391596039530867e-05,
+      "loss": 0.0699,
+      "step": 968
+    },
+    {
+      "epoch": 1.4907692307692308,
+      "grad_norm": 0.4221103719027582,
+      "learning_rate": 1.3904839476301091e-05,
+      "loss": 0.0745,
+      "step": 969
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 0.3300247767704618,
+      "learning_rate": 1.3893712856226028e-05,
+      "loss": 0.063,
+      "step": 970
+    },
+    {
+      "epoch": 1.4938461538461538,
+      "grad_norm": 0.35465084506366884,
+      "learning_rate": 1.388258055132835e-05,
+      "loss": 0.0628,
+      "step": 971
+    },
+    {
+      "epoch": 1.4953846153846153,
+      "grad_norm": 0.3956402484871412,
+      "learning_rate": 1.3871442577861234e-05,
+      "loss": 0.0718,
+      "step": 972
+    },
+    {
+      "epoch": 1.496923076923077,
+      "grad_norm": 0.32259127869403026,
+      "learning_rate": 1.3860298952086118e-05,
+      "loss": 0.0673,
+      "step": 973
+    },
+    {
+      "epoch": 1.4984615384615385,
+      "grad_norm": 0.30587983895963383,
+      "learning_rate": 1.3849149690272704e-05,
+      "loss": 0.0642,
+      "step": 974
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.3268162425630392,
+      "learning_rate": 1.383799480869892e-05,
+      "loss": 0.0655,
+      "step": 975
+    },
+    {
+      "epoch": 1.5015384615384615,
+      "grad_norm": 0.35344360561508237,
+      "learning_rate": 1.3826834323650899e-05,
+      "loss": 0.0717,
+      "step": 976
+    },
+    {
+      "epoch": 1.503076923076923,
+      "grad_norm": 0.4240517174731372,
+      "learning_rate": 1.3815668251422953e-05,
+      "loss": 0.0767,
+      "step": 977
+    },
+    {
+      "epoch": 1.5046153846153847,
+      "grad_norm": 0.3890020659351386,
+      "learning_rate": 1.3804496608317557e-05,
+      "loss": 0.0707,
+      "step": 978
+    },
+    {
+      "epoch": 1.5061538461538462,
+      "grad_norm": 0.34522079737200434,
+      "learning_rate": 1.3793319410645307e-05,
+      "loss": 0.0704,
+      "step": 979
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 0.35470349339672885,
+      "learning_rate": 1.3782136674724924e-05,
+      "loss": 0.0681,
+      "step": 980
+    },
+    {
+      "epoch": 1.5092307692307694,
+      "grad_norm": 0.37327575501099824,
+      "learning_rate": 1.3770948416883205e-05,
+      "loss": 0.0609,
+      "step": 981
+    },
+    {
+      "epoch": 1.5107692307692306,
+      "grad_norm": 0.3727506860019033,
+      "learning_rate": 1.3759754653455013e-05,
+      "loss": 0.0624,
+      "step": 982
+    },
+    {
+      "epoch": 1.5123076923076924,
+      "grad_norm": 0.3169662482724166,
+      "learning_rate": 1.3748555400783245e-05,
+      "loss": 0.0679,
+      "step": 983
+    },
+    {
+      "epoch": 1.5138461538461538,
+      "grad_norm": 0.318203785112587,
+      "learning_rate": 1.3737350675218819e-05,
+      "loss": 0.0721,
+      "step": 984
+    },
+    {
+      "epoch": 1.5153846153846153,
+      "grad_norm": 0.3459767757333529,
+      "learning_rate": 1.3726140493120639e-05,
+      "loss": 0.0692,
+      "step": 985
+    },
+    {
+      "epoch": 1.516923076923077,
+      "grad_norm": 0.35213169704344977,
+      "learning_rate": 1.3714924870855573e-05,
+      "loss": 0.059,
+      "step": 986
+    },
+    {
+      "epoch": 1.5184615384615383,
+      "grad_norm": 0.444276142733395,
+      "learning_rate": 1.3703703824798438e-05,
+      "loss": 0.0684,
+      "step": 987
+    },
+    {
+      "epoch": 1.52,
+      "grad_norm": 0.35207361668294573,
+      "learning_rate": 1.3692477371331965e-05,
+      "loss": 0.0744,
+      "step": 988
+    },
+    {
+      "epoch": 1.5215384615384615,
+      "grad_norm": 0.35999767828776713,
+      "learning_rate": 1.3681245526846782e-05,
+      "loss": 0.0713,
+      "step": 989
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 0.39013944310312315,
+      "learning_rate": 1.3670008307741388e-05,
+      "loss": 0.0725,
+      "step": 990
+    },
+    {
+      "epoch": 1.5246153846153847,
+      "grad_norm": 0.38236299835311816,
+      "learning_rate": 1.3658765730422126e-05,
+      "loss": 0.0679,
+      "step": 991
+    },
+    {
+      "epoch": 1.5261538461538462,
+      "grad_norm": 0.3391508447409303,
+      "learning_rate": 1.3647517811303164e-05,
+      "loss": 0.0674,
+      "step": 992
+    },
+    {
+      "epoch": 1.5276923076923077,
+      "grad_norm": 0.6412950373687982,
+      "learning_rate": 1.3636264566806473e-05,
+      "loss": 0.0741,
+      "step": 993
+    },
+    {
+      "epoch": 1.5292307692307694,
+      "grad_norm": 0.3350577163257067,
+      "learning_rate": 1.362500601336179e-05,
+      "loss": 0.0614,
+      "step": 994
+    },
+    {
+      "epoch": 1.5307692307692307,
+      "grad_norm": 0.3626403986182635,
+      "learning_rate": 1.3613742167406614e-05,
+      "loss": 0.0691,
+      "step": 995
+    },
+    {
+      "epoch": 1.5323076923076924,
+      "grad_norm": 0.3470732304493993,
+      "learning_rate": 1.3602473045386165e-05,
+      "loss": 0.0742,
+      "step": 996
+    },
+    {
+      "epoch": 1.5338461538461539,
+      "grad_norm": 0.33260189078997204,
+      "learning_rate": 1.3591198663753358e-05,
+      "loss": 0.0686,
+      "step": 997
+    },
+    {
+      "epoch": 1.5353846153846153,
+      "grad_norm": 0.28916096300010014,
+      "learning_rate": 1.3579919038968805e-05,
+      "loss": 0.0638,
+      "step": 998
+    },
+    {
+      "epoch": 1.536923076923077,
+      "grad_norm": 0.43372086950966615,
+      "learning_rate": 1.3568634187500762e-05,
+      "loss": 0.07,
+      "step": 999
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.40435446045850537,
+      "learning_rate": 1.3557344125825113e-05,
+      "loss": 0.0686,
+      "step": 1000
+    },
+    {
+      "epoch": 1.54,
+      "grad_norm": 0.3600272169695901,
+      "learning_rate": 1.3546048870425356e-05,
+      "loss": 0.075,
+      "step": 1001
+    },
+    {
+      "epoch": 1.5415384615384615,
+      "grad_norm": 0.31816083037803755,
+      "learning_rate": 1.3534748437792573e-05,
+      "loss": 0.0737,
+      "step": 1002
+    },
+    {
+      "epoch": 1.543076923076923,
+      "grad_norm": 0.33273148854011647,
+      "learning_rate": 1.3523442844425393e-05,
+      "loss": 0.0665,
+      "step": 1003
+    },
+    {
+      "epoch": 1.5446153846153847,
+      "grad_norm": 0.37169937500283534,
+      "learning_rate": 1.3512132106829996e-05,
+      "loss": 0.0654,
+      "step": 1004
+    },
+    {
+      "epoch": 1.546153846153846,
+      "grad_norm": 0.329535717539556,
+      "learning_rate": 1.3500816241520059e-05,
+      "loss": 0.0648,
+      "step": 1005
+    },
+    {
+      "epoch": 1.5476923076923077,
+      "grad_norm": 0.25023308868192906,
+      "learning_rate": 1.3489495265016753e-05,
+      "loss": 0.0558,
+      "step": 1006
+    },
+    {
+      "epoch": 1.5492307692307692,
+      "grad_norm": 0.3555280320936877,
+      "learning_rate": 1.3478169193848705e-05,
+      "loss": 0.0695,
+      "step": 1007
+    },
+    {
+      "epoch": 1.5507692307692307,
+      "grad_norm": 0.3533955661877233,
+      "learning_rate": 1.346683804455199e-05,
+      "loss": 0.0651,
+      "step": 1008
+    },
+    {
+      "epoch": 1.5523076923076924,
+      "grad_norm": 0.3802507308411157,
+      "learning_rate": 1.3455501833670089e-05,
+      "loss": 0.0737,
+      "step": 1009
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 0.3088088179326214,
+      "learning_rate": 1.3444160577753872e-05,
+      "loss": 0.0681,
+      "step": 1010
+    },
+    {
+      "epoch": 1.5553846153846154,
+      "grad_norm": 0.2929009087213979,
+      "learning_rate": 1.3432814293361585e-05,
+      "loss": 0.062,
+      "step": 1011
+    },
+    {
+      "epoch": 1.556923076923077,
+      "grad_norm": 0.37467092752746656,
+      "learning_rate": 1.34214629970588e-05,
+      "loss": 0.0709,
+      "step": 1012
+    },
+    {
+      "epoch": 1.5584615384615383,
+      "grad_norm": 0.33157137643925366,
+      "learning_rate": 1.3410106705418424e-05,
+      "loss": 0.0655,
+      "step": 1013
+    },
+    {
+      "epoch": 1.56,
+      "grad_norm": 0.36425079800863963,
+      "learning_rate": 1.3398745435020642e-05,
+      "loss": 0.0717,
+      "step": 1014
+    },
+    {
+      "epoch": 1.5615384615384615,
+      "grad_norm": 0.3794215971913442,
+      "learning_rate": 1.3387379202452917e-05,
+      "loss": 0.07,
+      "step": 1015
+    },
+    {
+      "epoch": 1.563076923076923,
+      "grad_norm": 0.35292364976240903,
+      "learning_rate": 1.337600802430995e-05,
+      "loss": 0.0737,
+      "step": 1016
+    },
+    {
+      "epoch": 1.5646153846153847,
+      "grad_norm": 0.36309299670186634,
+      "learning_rate": 1.3364631917193671e-05,
+      "loss": 0.0678,
+      "step": 1017
+    },
+    {
+      "epoch": 1.566153846153846,
+      "grad_norm": 0.43174462874417424,
+      "learning_rate": 1.33532508977132e-05,
+      "loss": 0.0664,
+      "step": 1018
+    },
+    {
+      "epoch": 1.5676923076923077,
+      "grad_norm": 0.2980059236096669,
+      "learning_rate": 1.3341864982484828e-05,
+      "loss": 0.0592,
+      "step": 1019
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 0.3505262612950089,
+      "learning_rate": 1.3330474188132004e-05,
+      "loss": 0.0735,
+      "step": 1020
+    },
+    {
+      "epoch": 1.5707692307692307,
+      "grad_norm": 0.3640243702743533,
+      "learning_rate": 1.3319078531285286e-05,
+      "loss": 0.0684,
+      "step": 1021
+    },
+    {
+      "epoch": 1.5723076923076924,
+      "grad_norm": 0.37347682099255064,
+      "learning_rate": 1.3307678028582342e-05,
+      "loss": 0.0631,
+      "step": 1022
+    },
+    {
+      "epoch": 1.573846153846154,
+      "grad_norm": 0.30683658621574744,
+      "learning_rate": 1.329627269666791e-05,
+      "loss": 0.064,
+      "step": 1023
+    },
+    {
+      "epoch": 1.5753846153846154,
+      "grad_norm": 0.38325694612265304,
+      "learning_rate": 1.328486255219378e-05,
+      "loss": 0.0748,
+      "step": 1024
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.34834422809752846,
+      "learning_rate": 1.3273447611818768e-05,
+      "loss": 0.0706,
+      "step": 1025
+    },
+    {
+      "epoch": 1.5784615384615384,
+      "grad_norm": 0.33431857255788205,
+      "learning_rate": 1.3262027892208696e-05,
+      "loss": 0.0607,
+      "step": 1026
+    },
+    {
+      "epoch": 1.58,
+      "grad_norm": 0.36935448940068555,
+      "learning_rate": 1.3250603410036356e-05,
+      "loss": 0.0657,
+      "step": 1027
+    },
+    {
+      "epoch": 1.5815384615384616,
+      "grad_norm": 0.5029604354345404,
+      "learning_rate": 1.3239174181981496e-05,
+      "loss": 0.0709,
+      "step": 1028
+    },
+    {
+      "epoch": 1.583076923076923,
+      "grad_norm": 0.34010742851810977,
+      "learning_rate": 1.3227740224730799e-05,
+      "loss": 0.0639,
+      "step": 1029
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 0.35326235063290296,
+      "learning_rate": 1.3216301554977844e-05,
+      "loss": 0.07,
+      "step": 1030
+    },
+    {
+      "epoch": 1.586153846153846,
+      "grad_norm": 0.371088178596748,
+      "learning_rate": 1.3204858189423097e-05,
+      "loss": 0.0689,
+      "step": 1031
+    },
+    {
+      "epoch": 1.5876923076923077,
+      "grad_norm": 0.3525413313256349,
+      "learning_rate": 1.3193410144773876e-05,
+      "loss": 0.0646,
+      "step": 1032
+    },
+    {
+      "epoch": 1.5892307692307692,
+      "grad_norm": 0.47439467360396914,
+      "learning_rate": 1.3181957437744333e-05,
+      "loss": 0.0665,
+      "step": 1033
+    },
+    {
+      "epoch": 1.5907692307692307,
+      "grad_norm": 0.3389914679073154,
+      "learning_rate": 1.3170500085055424e-05,
+      "loss": 0.0663,
+      "step": 1034
+    },
+    {
+      "epoch": 1.5923076923076924,
+      "grad_norm": 0.34343403261475547,
+      "learning_rate": 1.3159038103434889e-05,
+      "loss": 0.0693,
+      "step": 1035
+    },
+    {
+      "epoch": 1.5938461538461537,
+      "grad_norm": 0.30611366359410186,
+      "learning_rate": 1.314757150961723e-05,
+      "loss": 0.0676,
+      "step": 1036
+    },
+    {
+      "epoch": 1.5953846153846154,
+      "grad_norm": 0.44257355506663903,
+      "learning_rate": 1.3136100320343674e-05,
+      "loss": 0.0701,
+      "step": 1037
+    },
+    {
+      "epoch": 1.596923076923077,
+      "grad_norm": 0.32909363170626144,
+      "learning_rate": 1.3124624552362166e-05,
+      "loss": 0.0627,
+      "step": 1038
+    },
+    {
+      "epoch": 1.5984615384615384,
+      "grad_norm": 0.38341925530098436,
+      "learning_rate": 1.3113144222427334e-05,
+      "loss": 0.0718,
+      "step": 1039
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 0.3390749420830194,
+      "learning_rate": 1.3101659347300462e-05,
+      "loss": 0.0613,
+      "step": 1040
+    },
+    {
+      "epoch": 1.6015384615384616,
+      "grad_norm": 0.3462273945212573,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.0735,
+      "step": 1041
+    },
+    {
+      "epoch": 1.603076923076923,
+      "grad_norm": 0.3695043310551725,
+      "learning_rate": 1.3078676028548908e-05,
+      "loss": 0.0694,
+      "step": 1042
+    },
+    {
+      "epoch": 1.6046153846153848,
+      "grad_norm": 0.3530125489309211,
+      "learning_rate": 1.3067177618479883e-05,
+      "loss": 0.0631,
+      "step": 1043
+    },
+    {
+      "epoch": 1.606153846153846,
+      "grad_norm": 0.32964686006837474,
+      "learning_rate": 1.305567473033008e-05,
+      "loss": 0.0679,
+      "step": 1044
+    },
+    {
+      "epoch": 1.6076923076923078,
+      "grad_norm": 0.3713682256507068,
+      "learning_rate": 1.3044167380893726e-05,
+      "loss": 0.0701,
+      "step": 1045
+    },
+    {
+      "epoch": 1.6092307692307692,
+      "grad_norm": 0.3762742954199106,
+      "learning_rate": 1.3032655586971552e-05,
+      "loss": 0.0628,
+      "step": 1046
+    },
+    {
+      "epoch": 1.6107692307692307,
+      "grad_norm": 0.3265101319762403,
+      "learning_rate": 1.3021139365370787e-05,
+      "loss": 0.0656,
+      "step": 1047
+    },
+    {
+      "epoch": 1.6123076923076924,
+      "grad_norm": 0.28711481275165773,
+      "learning_rate": 1.300961873290512e-05,
+      "loss": 0.068,
+      "step": 1048
+    },
+    {
+      "epoch": 1.6138461538461537,
+      "grad_norm": 0.4010476086338245,
+      "learning_rate": 1.2998093706394676e-05,
+      "loss": 0.0717,
+      "step": 1049
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.3145419470926991,
+      "learning_rate": 1.2986564302666e-05,
+      "loss": 0.0694,
+      "step": 1050
+    },
+    {
+      "epoch": 1.616923076923077,
+      "grad_norm": 0.3926012570647249,
+      "learning_rate": 1.297503053855203e-05,
+      "loss": 0.0713,
+      "step": 1051
+    },
+    {
+      "epoch": 1.6184615384615384,
+      "grad_norm": 0.3304899518184055,
+      "learning_rate": 1.2963492430892066e-05,
+      "loss": 0.0665,
+      "step": 1052
+    },
+    {
+      "epoch": 1.62,
+      "grad_norm": 0.3249058555050507,
+      "learning_rate": 1.295194999653175e-05,
+      "loss": 0.0716,
+      "step": 1053
+    },
+    {
+      "epoch": 1.6215384615384614,
+      "grad_norm": 0.34563682161792825,
+      "learning_rate": 1.294040325232304e-05,
+      "loss": 0.0689,
+      "step": 1054
+    },
+    {
+      "epoch": 1.623076923076923,
+      "grad_norm": 0.3623157841595741,
+      "learning_rate": 1.292885221512419e-05,
+      "loss": 0.0686,
+      "step": 1055
+    },
+    {
+      "epoch": 1.6246153846153846,
+      "grad_norm": 0.4032343951986735,
+      "learning_rate": 1.291729690179972e-05,
+      "loss": 0.0709,
+      "step": 1056
+    },
+    {
+      "epoch": 1.626153846153846,
+      "grad_norm": 0.42548798300770907,
+      "learning_rate": 1.2905737329220394e-05,
+      "loss": 0.0769,
+      "step": 1057
+    },
+    {
+      "epoch": 1.6276923076923078,
+      "grad_norm": 0.38063725968362255,
+      "learning_rate": 1.2894173514263191e-05,
+      "loss": 0.0718,
+      "step": 1058
+    },
+    {
+      "epoch": 1.6292307692307693,
+      "grad_norm": 0.4260166630050922,
+      "learning_rate": 1.2882605473811282e-05,
+      "loss": 0.0755,
+      "step": 1059
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 0.3183113830256884,
+      "learning_rate": 1.2871033224754022e-05,
+      "loss": 0.0659,
+      "step": 1060
+    },
+    {
+      "epoch": 1.6323076923076925,
+      "grad_norm": 0.33423253446578904,
+      "learning_rate": 1.2859456783986892e-05,
+      "loss": 0.0675,
+      "step": 1061
+    },
+    {
+      "epoch": 1.6338461538461537,
+      "grad_norm": 0.38425643174539403,
+      "learning_rate": 1.2847876168411506e-05,
+      "loss": 0.0668,
+      "step": 1062
+    },
+    {
+      "epoch": 1.6353846153846154,
+      "grad_norm": 0.39834995039582705,
+      "learning_rate": 1.2836291394935568e-05,
+      "loss": 0.0702,
+      "step": 1063
+    },
+    {
+      "epoch": 1.636923076923077,
+      "grad_norm": 0.3592062233324518,
+      "learning_rate": 1.2824702480472846e-05,
+      "loss": 0.0698,
+      "step": 1064
+    },
+    {
+      "epoch": 1.6384615384615384,
+      "grad_norm": 0.30141364654191644,
+      "learning_rate": 1.2813109441943166e-05,
+      "loss": 0.0608,
+      "step": 1065
+    },
+    {
+      "epoch": 1.6400000000000001,
+      "grad_norm": 0.31650470078418513,
+      "learning_rate": 1.280151229627237e-05,
+      "loss": 0.0727,
+      "step": 1066
+    },
+    {
+      "epoch": 1.6415384615384614,
+      "grad_norm": 0.3073328562892288,
+      "learning_rate": 1.2789911060392295e-05,
+      "loss": 0.0597,
+      "step": 1067
+    },
+    {
+      "epoch": 1.643076923076923,
+      "grad_norm": 0.352934869488259,
+      "learning_rate": 1.2778305751240749e-05,
+      "loss": 0.0646,
+      "step": 1068
+    },
+    {
+      "epoch": 1.6446153846153846,
+      "grad_norm": 0.3477210583426811,
+      "learning_rate": 1.2766696385761494e-05,
+      "loss": 0.0694,
+      "step": 1069
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 0.31714834966112554,
+      "learning_rate": 1.2755082980904206e-05,
+      "loss": 0.0693,
+      "step": 1070
+    },
+    {
+      "epoch": 1.6476923076923078,
+      "grad_norm": 0.3328830298996729,
+      "learning_rate": 1.274346555362446e-05,
+      "loss": 0.0651,
+      "step": 1071
+    },
+    {
+      "epoch": 1.6492307692307693,
+      "grad_norm": 0.4081142662510038,
+      "learning_rate": 1.2731844120883705e-05,
+      "loss": 0.0677,
+      "step": 1072
+    },
+    {
+      "epoch": 1.6507692307692308,
+      "grad_norm": 0.3634578478803434,
+      "learning_rate": 1.2720218699649243e-05,
+      "loss": 0.0674,
+      "step": 1073
+    },
+    {
+      "epoch": 1.6523076923076923,
+      "grad_norm": 0.38192663091543777,
+      "learning_rate": 1.270858930689419e-05,
+      "loss": 0.0701,
+      "step": 1074
+    },
+    {
+      "epoch": 1.6538461538461537,
+      "grad_norm": 0.3172214274621381,
+      "learning_rate": 1.269695595959747e-05,
+      "loss": 0.0704,
+      "step": 1075
+    },
+    {
+      "epoch": 1.6553846153846155,
+      "grad_norm": 0.38108705084014816,
+      "learning_rate": 1.2685318674743769e-05,
+      "loss": 0.0773,
+      "step": 1076
+    },
+    {
+      "epoch": 1.656923076923077,
+      "grad_norm": 0.32308540109864475,
+      "learning_rate": 1.2673677469323532e-05,
+      "loss": 0.0648,
+      "step": 1077
+    },
+    {
+      "epoch": 1.6584615384615384,
+      "grad_norm": 0.397581930467522,
+      "learning_rate": 1.2662032360332926e-05,
+      "loss": 0.0639,
+      "step": 1078
+    },
+    {
+      "epoch": 1.6600000000000001,
+      "grad_norm": 0.428597873391097,
+      "learning_rate": 1.2650383364773812e-05,
+      "loss": 0.0649,
+      "step": 1079
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 0.3627584918043698,
+      "learning_rate": 1.2638730499653731e-05,
+      "loss": 0.0685,
+      "step": 1080
+    },
+    {
+      "epoch": 1.6630769230769231,
+      "grad_norm": 0.39434710297076264,
+      "learning_rate": 1.262707378198587e-05,
+      "loss": 0.0623,
+      "step": 1081
+    },
+    {
+      "epoch": 1.6646153846153846,
+      "grad_norm": 0.4398747205110735,
+      "learning_rate": 1.2615413228789044e-05,
+      "loss": 0.0731,
+      "step": 1082
+    },
+    {
+      "epoch": 1.666153846153846,
+      "grad_norm": 0.42581951735488976,
+      "learning_rate": 1.2603748857087668e-05,
+      "loss": 0.0722,
+      "step": 1083
+    },
+    {
+      "epoch": 1.6676923076923078,
+      "grad_norm": 0.33276187980893557,
+      "learning_rate": 1.2592080683911726e-05,
+      "loss": 0.0696,
+      "step": 1084
+    },
+    {
+      "epoch": 1.669230769230769,
+      "grad_norm": 0.41489629589539934,
+      "learning_rate": 1.258040872629676e-05,
+      "loss": 0.0719,
+      "step": 1085
+    },
+    {
+      "epoch": 1.6707692307692308,
+      "grad_norm": 0.3446331527576127,
+      "learning_rate": 1.2568733001283828e-05,
+      "loss": 0.0559,
+      "step": 1086
+    },
+    {
+      "epoch": 1.6723076923076923,
+      "grad_norm": 0.35834532855166323,
+      "learning_rate": 1.2557053525919503e-05,
+      "loss": 0.0746,
+      "step": 1087
+    },
+    {
+      "epoch": 1.6738461538461538,
+      "grad_norm": 0.4154468285899175,
+      "learning_rate": 1.2545370317255817e-05,
+      "loss": 0.0685,
+      "step": 1088
+    },
+    {
+      "epoch": 1.6753846153846155,
+      "grad_norm": 0.3870928793180616,
+      "learning_rate": 1.2533683392350264e-05,
+      "loss": 0.081,
+      "step": 1089
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 0.32370828199921,
+      "learning_rate": 1.252199276826576e-05,
+      "loss": 0.0535,
+      "step": 1090
+    },
+    {
+      "epoch": 1.6784615384615384,
+      "grad_norm": 0.36882979778488023,
+      "learning_rate": 1.2510298462070619e-05,
+      "loss": 0.0671,
+      "step": 1091
+    },
+    {
+      "epoch": 1.6800000000000002,
+      "grad_norm": 0.32092949735914467,
+      "learning_rate": 1.2498600490838535e-05,
+      "loss": 0.0626,
+      "step": 1092
+    },
+    {
+      "epoch": 1.6815384615384614,
+      "grad_norm": 0.5727815255595513,
+      "learning_rate": 1.2486898871648552e-05,
+      "loss": 0.0734,
+      "step": 1093
+    },
+    {
+      "epoch": 1.6830769230769231,
+      "grad_norm": 0.3458495040870996,
+      "learning_rate": 1.2475193621585036e-05,
+      "loss": 0.0653,
+      "step": 1094
+    },
+    {
+      "epoch": 1.6846153846153846,
+      "grad_norm": 0.35967546735381495,
+      "learning_rate": 1.2463484757737663e-05,
+      "loss": 0.0705,
+      "step": 1095
+    },
+    {
+      "epoch": 1.6861538461538461,
+      "grad_norm": 0.38237562063590214,
+      "learning_rate": 1.2451772297201376e-05,
+      "loss": 0.0661,
+      "step": 1096
+    },
+    {
+      "epoch": 1.6876923076923078,
+      "grad_norm": 0.47686706186257616,
+      "learning_rate": 1.2440056257076376e-05,
+      "loss": 0.0613,
+      "step": 1097
+    },
+    {
+      "epoch": 1.689230769230769,
+      "grad_norm": 0.46719251208554496,
+      "learning_rate": 1.2428336654468085e-05,
+      "loss": 0.0662,
+      "step": 1098
+    },
+    {
+      "epoch": 1.6907692307692308,
+      "grad_norm": 0.3686979455788657,
+      "learning_rate": 1.241661350648713e-05,
+      "loss": 0.0735,
+      "step": 1099
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.33821254997323613,
+      "learning_rate": 1.240488683024931e-05,
+      "loss": 0.0671,
+      "step": 1100
+    },
+    {
+      "epoch": 1.6938461538461538,
+      "grad_norm": 0.314525906392349,
+      "learning_rate": 1.2393156642875579e-05,
+      "loss": 0.0593,
+      "step": 1101
+    },
+    {
+      "epoch": 1.6953846153846155,
+      "grad_norm": 0.37318354708691176,
+      "learning_rate": 1.2381422961492018e-05,
+      "loss": 0.0612,
+      "step": 1102
+    },
+    {
+      "epoch": 1.696923076923077,
+      "grad_norm": 0.38841429558968976,
+      "learning_rate": 1.2369685803229802e-05,
+      "loss": 0.0618,
+      "step": 1103
+    },
+    {
+      "epoch": 1.6984615384615385,
+      "grad_norm": 0.4236775137197504,
+      "learning_rate": 1.2357945185225194e-05,
+      "loss": 0.0657,
+      "step": 1104
+    },
+    {
+      "epoch": 1.7,
+      "grad_norm": 0.33764294486901314,
+      "learning_rate": 1.2346201124619502e-05,
+      "loss": 0.0649,
+      "step": 1105
+    },
+    {
+      "epoch": 1.7015384615384614,
+      "grad_norm": 0.4418410618698052,
+      "learning_rate": 1.2334453638559057e-05,
+      "loss": 0.0687,
+      "step": 1106
+    },
+    {
+      "epoch": 1.7030769230769232,
+      "grad_norm": 0.3121204113437058,
+      "learning_rate": 1.2322702744195192e-05,
+      "loss": 0.0611,
+      "step": 1107
+    },
+    {
+      "epoch": 1.7046153846153846,
+      "grad_norm": 0.3445242526989668,
+      "learning_rate": 1.231094845868422e-05,
+      "loss": 0.0702,
+      "step": 1108
+    },
+    {
+      "epoch": 1.7061538461538461,
+      "grad_norm": 0.3181762559945284,
+      "learning_rate": 1.2299190799187405e-05,
+      "loss": 0.0617,
+      "step": 1109
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 0.35189126228058387,
+      "learning_rate": 1.2287429782870936e-05,
+      "loss": 0.0654,
+      "step": 1110
+    },
+    {
+      "epoch": 1.709230769230769,
+      "grad_norm": 0.3893933792288036,
+      "learning_rate": 1.22756654269059e-05,
+      "loss": 0.0731,
+      "step": 1111
+    },
+    {
+      "epoch": 1.7107692307692308,
+      "grad_norm": 0.36384662807264734,
+      "learning_rate": 1.2263897748468265e-05,
+      "loss": 0.0748,
+      "step": 1112
+    },
+    {
+      "epoch": 1.7123076923076923,
+      "grad_norm": 0.35150416540561524,
+      "learning_rate": 1.2252126764738845e-05,
+      "loss": 0.063,
+      "step": 1113
+    },
+    {
+      "epoch": 1.7138461538461538,
+      "grad_norm": 0.37272242307664083,
+      "learning_rate": 1.2240352492903282e-05,
+      "loss": 0.0658,
+      "step": 1114
+    },
+    {
+      "epoch": 1.7153846153846155,
+      "grad_norm": 0.34210913074854066,
+      "learning_rate": 1.222857495015202e-05,
+      "loss": 0.0631,
+      "step": 1115
+    },
+    {
+      "epoch": 1.7169230769230768,
+      "grad_norm": 0.3261822986248889,
+      "learning_rate": 1.2216794153680274e-05,
+      "loss": 0.0682,
+      "step": 1116
+    },
+    {
+      "epoch": 1.7184615384615385,
+      "grad_norm": 0.38673754458534254,
+      "learning_rate": 1.2205010120688012e-05,
+      "loss": 0.0744,
+      "step": 1117
+    },
+    {
+      "epoch": 1.72,
+      "grad_norm": 0.322372777644759,
+      "learning_rate": 1.2193222868379933e-05,
+      "loss": 0.0695,
+      "step": 1118
+    },
+    {
+      "epoch": 1.7215384615384615,
+      "grad_norm": 0.3627083236569345,
+      "learning_rate": 1.2181432413965428e-05,
+      "loss": 0.0742,
+      "step": 1119
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 0.3603766452102839,
+      "learning_rate": 1.2169638774658566e-05,
+      "loss": 0.0701,
+      "step": 1120
+    },
+    {
+      "epoch": 1.7246153846153847,
+      "grad_norm": 0.400801215452547,
+      "learning_rate": 1.2157841967678064e-05,
+      "loss": 0.073,
+      "step": 1121
+    },
+    {
+      "epoch": 1.7261538461538461,
+      "grad_norm": 0.35897450361326094,
+      "learning_rate": 1.2146042010247268e-05,
+      "loss": 0.0713,
+      "step": 1122
+    },
+    {
+      "epoch": 1.7276923076923076,
+      "grad_norm": 0.32908439200162665,
+      "learning_rate": 1.2134238919594122e-05,
+      "loss": 0.073,
+      "step": 1123
+    },
+    {
+      "epoch": 1.7292307692307691,
+      "grad_norm": 0.3195562766482599,
+      "learning_rate": 1.2122432712951142e-05,
+      "loss": 0.0639,
+      "step": 1124
+    },
+    {
+      "epoch": 1.7307692307692308,
+      "grad_norm": 0.3845258420163186,
+      "learning_rate": 1.2110623407555398e-05,
+      "loss": 0.061,
+      "step": 1125
+    },
+    {
+      "epoch": 1.7323076923076923,
+      "grad_norm": 0.3101783646667082,
+      "learning_rate": 1.2098811020648475e-05,
+      "loss": 0.0612,
+      "step": 1126
+    },
+    {
+      "epoch": 1.7338461538461538,
+      "grad_norm": 0.40958171970575463,
+      "learning_rate": 1.2086995569476474e-05,
+      "loss": 0.0602,
+      "step": 1127
+    },
+    {
+      "epoch": 1.7353846153846155,
+      "grad_norm": 0.36131350860502126,
+      "learning_rate": 1.2075177071289952e-05,
+      "loss": 0.0716,
+      "step": 1128
+    },
+    {
+      "epoch": 1.7369230769230768,
+      "grad_norm": 0.3245396638728491,
+      "learning_rate": 1.2063355543343925e-05,
+      "loss": 0.0664,
+      "step": 1129
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 0.34948238997533076,
+      "learning_rate": 1.2051531002897823e-05,
+      "loss": 0.0743,
+      "step": 1130
+    },
+    {
+      "epoch": 1.74,
+      "grad_norm": 0.33482788469371083,
+      "learning_rate": 1.2039703467215489e-05,
+      "loss": 0.07,
+      "step": 1131
+    },
+    {
+      "epoch": 1.7415384615384615,
+      "grad_norm": 0.3106255436550464,
+      "learning_rate": 1.2027872953565125e-05,
+      "loss": 0.0656,
+      "step": 1132
+    },
+    {
+      "epoch": 1.7430769230769232,
+      "grad_norm": 0.3230656932594039,
+      "learning_rate": 1.2016039479219293e-05,
+      "loss": 0.0673,
+      "step": 1133
+    },
+    {
+      "epoch": 1.7446153846153845,
+      "grad_norm": 0.41382601752001624,
+      "learning_rate": 1.2004203061454864e-05,
+      "loss": 0.0665,
+      "step": 1134
+    },
+    {
+      "epoch": 1.7461538461538462,
+      "grad_norm": 0.3742034275244198,
+      "learning_rate": 1.1992363717553015e-05,
+      "loss": 0.0752,
+      "step": 1135
+    },
+    {
+      "epoch": 1.7476923076923077,
+      "grad_norm": 0.36413647568781654,
+      "learning_rate": 1.1980521464799197e-05,
+      "loss": 0.071,
+      "step": 1136
+    },
+    {
+      "epoch": 1.7492307692307691,
+      "grad_norm": 0.373128673273757,
+      "learning_rate": 1.1968676320483103e-05,
+      "loss": 0.0702,
+      "step": 1137
+    },
+    {
+      "epoch": 1.7507692307692309,
+      "grad_norm": 0.3372236802921592,
+      "learning_rate": 1.1956828301898648e-05,
+      "loss": 0.0678,
+      "step": 1138
+    },
+    {
+      "epoch": 1.7523076923076923,
+      "grad_norm": 0.40138229132959674,
+      "learning_rate": 1.194497742634395e-05,
+      "loss": 0.0717,
+      "step": 1139
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 0.33981746405629404,
+      "learning_rate": 1.1933123711121284e-05,
+      "loss": 0.0707,
+      "step": 1140
+    },
+    {
+      "epoch": 1.7553846153846155,
+      "grad_norm": 0.31343455180354113,
+      "learning_rate": 1.1921267173537085e-05,
+      "loss": 0.0649,
+      "step": 1141
+    },
+    {
+      "epoch": 1.7569230769230768,
+      "grad_norm": 0.40773047383140626,
+      "learning_rate": 1.1909407830901905e-05,
+      "loss": 0.0643,
+      "step": 1142
+    },
+    {
+      "epoch": 1.7584615384615385,
+      "grad_norm": 0.3528139994471679,
+      "learning_rate": 1.1897545700530387e-05,
+      "loss": 0.0611,
+      "step": 1143
+    },
+    {
+      "epoch": 1.76,
+      "grad_norm": 0.4034945403026721,
+      "learning_rate": 1.1885680799741249e-05,
+      "loss": 0.0743,
+      "step": 1144
+    },
+    {
+      "epoch": 1.7615384615384615,
+      "grad_norm": 0.41781100877516925,
+      "learning_rate": 1.187381314585725e-05,
+      "loss": 0.0678,
+      "step": 1145
+    },
+    {
+      "epoch": 1.7630769230769232,
+      "grad_norm": 0.36789269013488135,
+      "learning_rate": 1.186194275620517e-05,
+      "loss": 0.0662,
+      "step": 1146
+    },
+    {
+      "epoch": 1.7646153846153845,
+      "grad_norm": 0.4280148153142728,
+      "learning_rate": 1.1850069648115785e-05,
+      "loss": 0.0703,
+      "step": 1147
+    },
+    {
+      "epoch": 1.7661538461538462,
+      "grad_norm": 0.34082568199061963,
+      "learning_rate": 1.1838193838923835e-05,
+      "loss": 0.0664,
+      "step": 1148
+    },
+    {
+      "epoch": 1.7676923076923077,
+      "grad_norm": 0.40801522956984687,
+      "learning_rate": 1.1826315345968014e-05,
+      "loss": 0.0684,
+      "step": 1149
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.5130512133374202,
+      "learning_rate": 1.1814434186590922e-05,
+      "loss": 0.0673,
+      "step": 1150
+    },
+    {
+      "epoch": 1.7707692307692309,
+      "grad_norm": 0.41911565805686907,
+      "learning_rate": 1.180255037813906e-05,
+      "loss": 0.0721,
+      "step": 1151
+    },
+    {
+      "epoch": 1.7723076923076924,
+      "grad_norm": 0.4585867938819892,
+      "learning_rate": 1.1790663937962789e-05,
+      "loss": 0.07,
+      "step": 1152
+    },
+    {
+      "epoch": 1.7738461538461539,
+      "grad_norm": 0.4179899966673112,
+      "learning_rate": 1.1778774883416325e-05,
+      "loss": 0.0742,
+      "step": 1153
+    },
+    {
+      "epoch": 1.7753846153846153,
+      "grad_norm": 0.38592221008164235,
+      "learning_rate": 1.1766883231857686e-05,
+      "loss": 0.0744,
+      "step": 1154
+    },
+    {
+      "epoch": 1.7769230769230768,
+      "grad_norm": 0.35229801433260116,
+      "learning_rate": 1.1754989000648693e-05,
+      "loss": 0.0726,
+      "step": 1155
+    },
+    {
+      "epoch": 1.7784615384615385,
+      "grad_norm": 0.340673871886494,
+      "learning_rate": 1.1743092207154929e-05,
+      "loss": 0.0636,
+      "step": 1156
+    },
+    {
+      "epoch": 1.78,
+      "grad_norm": 0.3495561290874491,
+      "learning_rate": 1.1731192868745716e-05,
+      "loss": 0.056,
+      "step": 1157
+    },
+    {
+      "epoch": 1.7815384615384615,
+      "grad_norm": 0.3734216760946435,
+      "learning_rate": 1.1719291002794096e-05,
+      "loss": 0.0681,
+      "step": 1158
+    },
+    {
+      "epoch": 1.7830769230769232,
+      "grad_norm": 0.3417807928208906,
+      "learning_rate": 1.1707386626676798e-05,
+      "loss": 0.0729,
+      "step": 1159
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 0.3437047768651414,
+      "learning_rate": 1.1695479757774217e-05,
+      "loss": 0.0694,
+      "step": 1160
+    },
+    {
+      "epoch": 1.7861538461538462,
+      "grad_norm": 0.47420240578398315,
+      "learning_rate": 1.1683570413470384e-05,
+      "loss": 0.0676,
+      "step": 1161
+    },
+    {
+      "epoch": 1.7876923076923077,
+      "grad_norm": 0.3250761492618577,
+      "learning_rate": 1.1671658611152954e-05,
+      "loss": 0.0733,
+      "step": 1162
+    },
+    {
+      "epoch": 1.7892307692307692,
+      "grad_norm": 0.38754162428676897,
+      "learning_rate": 1.1659744368213159e-05,
+      "loss": 0.0621,
+      "step": 1163
+    },
+    {
+      "epoch": 1.790769230769231,
+      "grad_norm": 0.33675515675870277,
+      "learning_rate": 1.1647827702045802e-05,
+      "loss": 0.0672,
+      "step": 1164
+    },
+    {
+      "epoch": 1.7923076923076922,
+      "grad_norm": 0.44334916416488984,
+      "learning_rate": 1.163590863004922e-05,
+      "loss": 0.0767,
+      "step": 1165
+    },
+    {
+      "epoch": 1.7938461538461539,
+      "grad_norm": 0.36810606943536545,
+      "learning_rate": 1.1623987169625261e-05,
+      "loss": 0.0757,
+      "step": 1166
+    },
+    {
+      "epoch": 1.7953846153846154,
+      "grad_norm": 0.3878325616113022,
+      "learning_rate": 1.1612063338179269e-05,
+      "loss": 0.0674,
+      "step": 1167
+    },
+    {
+      "epoch": 1.7969230769230768,
+      "grad_norm": 0.36763716766560983,
+      "learning_rate": 1.1600137153120039e-05,
+      "loss": 0.0694,
+      "step": 1168
+    },
+    {
+      "epoch": 1.7984615384615386,
+      "grad_norm": 0.294038432326372,
+      "learning_rate": 1.1588208631859808e-05,
+      "loss": 0.0624,
+      "step": 1169
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 0.3066013994606816,
+      "learning_rate": 1.1576277791814219e-05,
+      "loss": 0.064,
+      "step": 1170
+    },
+    {
+      "epoch": 1.8015384615384615,
+      "grad_norm": 0.3791728156018938,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.0685,
+      "step": 1171
+    },
+    {
+      "epoch": 1.803076923076923,
+      "grad_norm": 0.34991093402366064,
+      "learning_rate": 1.1552409225046472e-05,
+      "loss": 0.0702,
+      "step": 1172
+    },
+    {
+      "epoch": 1.8046153846153845,
+      "grad_norm": 0.35156981872263665,
+      "learning_rate": 1.154047153317243e-05,
+      "loss": 0.0648,
+      "step": 1173
+    },
+    {
+      "epoch": 1.8061538461538462,
+      "grad_norm": 0.47105508744307845,
+      "learning_rate": 1.152853159220922e-05,
+      "loss": 0.0662,
+      "step": 1174
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.34193953754457573,
+      "learning_rate": 1.1516589419589159e-05,
+      "loss": 0.0662,
+      "step": 1175
+    },
+    {
+      "epoch": 1.8092307692307692,
+      "grad_norm": 0.3080144424855269,
+      "learning_rate": 1.1504645032747832e-05,
+      "loss": 0.0718,
+      "step": 1176
+    },
+    {
+      "epoch": 1.810769230769231,
+      "grad_norm": 0.3166715432520699,
+      "learning_rate": 1.1492698449124042e-05,
+      "loss": 0.0632,
+      "step": 1177
+    },
+    {
+      "epoch": 1.8123076923076922,
+      "grad_norm": 0.3116243301245059,
+      "learning_rate": 1.148074968615981e-05,
+      "loss": 0.0668,
+      "step": 1178
+    },
+    {
+      "epoch": 1.8138461538461539,
+      "grad_norm": 0.3923187036355073,
+      "learning_rate": 1.1468798761300335e-05,
+      "loss": 0.0721,
+      "step": 1179
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 0.3293677766442383,
+      "learning_rate": 1.1456845691993975e-05,
+      "loss": 0.0636,
+      "step": 1180
+    },
+    {
+      "epoch": 1.8169230769230769,
+      "grad_norm": 0.33075818730038914,
+      "learning_rate": 1.1444890495692214e-05,
+      "loss": 0.0688,
+      "step": 1181
+    },
+    {
+      "epoch": 1.8184615384615386,
+      "grad_norm": 0.3270293740539091,
+      "learning_rate": 1.1432933189849647e-05,
+      "loss": 0.0675,
+      "step": 1182
+    },
+    {
+      "epoch": 1.8199999999999998,
+      "grad_norm": 0.2923335748923596,
+      "learning_rate": 1.1420973791923941e-05,
+      "loss": 0.0701,
+      "step": 1183
+    },
+    {
+      "epoch": 1.8215384615384616,
+      "grad_norm": 0.3093982567613493,
+      "learning_rate": 1.1409012319375828e-05,
+      "loss": 0.0659,
+      "step": 1184
+    },
+    {
+      "epoch": 1.823076923076923,
+      "grad_norm": 0.3799881014599721,
+      "learning_rate": 1.1397048789669061e-05,
+      "loss": 0.0714,
+      "step": 1185
+    },
+    {
+      "epoch": 1.8246153846153845,
+      "grad_norm": 0.36261575151573644,
+      "learning_rate": 1.13850832202704e-05,
+      "loss": 0.0693,
+      "step": 1186
+    },
+    {
+      "epoch": 1.8261538461538462,
+      "grad_norm": 0.3132085768992873,
+      "learning_rate": 1.1373115628649582e-05,
+      "loss": 0.0682,
+      "step": 1187
+    },
+    {
+      "epoch": 1.8276923076923077,
+      "grad_norm": 0.3461207920147732,
+      "learning_rate": 1.1361146032279295e-05,
+      "loss": 0.0711,
+      "step": 1188
+    },
+    {
+      "epoch": 1.8292307692307692,
+      "grad_norm": 0.3150515227081572,
+      "learning_rate": 1.1349174448635158e-05,
+      "loss": 0.0592,
+      "step": 1189
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 0.33512808523002696,
+      "learning_rate": 1.1337200895195688e-05,
+      "loss": 0.0697,
+      "step": 1190
+    },
+    {
+      "epoch": 1.8323076923076922,
+      "grad_norm": 0.35007766877677027,
+      "learning_rate": 1.1325225389442278e-05,
+      "loss": 0.0667,
+      "step": 1191
+    },
+    {
+      "epoch": 1.833846153846154,
+      "grad_norm": 0.34216015472816874,
+      "learning_rate": 1.1313247948859168e-05,
+      "loss": 0.0769,
+      "step": 1192
+    },
+    {
+      "epoch": 1.8353846153846154,
+      "grad_norm": 0.34775267785403957,
+      "learning_rate": 1.1301268590933434e-05,
+      "loss": 0.0663,
+      "step": 1193
+    },
+    {
+      "epoch": 1.8369230769230769,
+      "grad_norm": 0.33480692927199907,
+      "learning_rate": 1.1289287333154941e-05,
+      "loss": 0.0762,
+      "step": 1194
+    },
+    {
+      "epoch": 1.8384615384615386,
+      "grad_norm": 0.32637056156167593,
+      "learning_rate": 1.1277304193016332e-05,
+      "loss": 0.0713,
+      "step": 1195
+    },
+    {
+      "epoch": 1.8399999999999999,
+      "grad_norm": 0.36829791104284954,
+      "learning_rate": 1.1265319188012995e-05,
+      "loss": 0.0737,
+      "step": 1196
+    },
+    {
+      "epoch": 1.8415384615384616,
+      "grad_norm": 0.36062035876290466,
+      "learning_rate": 1.1253332335643043e-05,
+      "loss": 0.0689,
+      "step": 1197
+    },
+    {
+      "epoch": 1.843076923076923,
+      "grad_norm": 0.3711995102972711,
+      "learning_rate": 1.124134365340729e-05,
+      "loss": 0.0642,
+      "step": 1198
+    },
+    {
+      "epoch": 1.8446153846153845,
+      "grad_norm": 0.30208317420183395,
+      "learning_rate": 1.1229353158809216e-05,
+      "loss": 0.0643,
+      "step": 1199
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.33621334713281376,
+      "learning_rate": 1.1217360869354948e-05,
+      "loss": 0.0716,
+      "step": 1200
+    },
+    {
+      "epoch": 1.8476923076923077,
+      "grad_norm": 0.4109500586936904,
+      "learning_rate": 1.1205366802553231e-05,
+      "loss": 0.0627,
+      "step": 1201
+    },
+    {
+      "epoch": 1.8492307692307692,
+      "grad_norm": 0.3355086005987176,
+      "learning_rate": 1.1193370975915414e-05,
+      "loss": 0.0651,
+      "step": 1202
+    },
+    {
+      "epoch": 1.8507692307692307,
+      "grad_norm": 0.33311821347747733,
+      "learning_rate": 1.118137340695541e-05,
+      "loss": 0.0707,
+      "step": 1203
+    },
+    {
+      "epoch": 1.8523076923076922,
+      "grad_norm": 0.3783217276792658,
+      "learning_rate": 1.1169374113189669e-05,
+      "loss": 0.0688,
+      "step": 1204
+    },
+    {
+      "epoch": 1.853846153846154,
+      "grad_norm": 0.43132614741663744,
+      "learning_rate": 1.1157373112137171e-05,
+      "loss": 0.0799,
+      "step": 1205
+    },
+    {
+      "epoch": 1.8553846153846154,
+      "grad_norm": 0.2922605690189237,
+      "learning_rate": 1.1145370421319377e-05,
+      "loss": 0.0592,
+      "step": 1206
+    },
+    {
+      "epoch": 1.856923076923077,
+      "grad_norm": 0.3223022612523878,
+      "learning_rate": 1.1133366058260232e-05,
+      "loss": 0.0647,
+      "step": 1207
+    },
+    {
+      "epoch": 1.8584615384615386,
+      "grad_norm": 0.3018670306666712,
+      "learning_rate": 1.11213600404861e-05,
+      "loss": 0.0685,
+      "step": 1208
+    },
+    {
+      "epoch": 1.8599999999999999,
+      "grad_norm": 0.3889446684181748,
+      "learning_rate": 1.1109352385525782e-05,
+      "loss": 0.0766,
+      "step": 1209
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 0.4477268822990713,
+      "learning_rate": 1.1097343110910452e-05,
+      "loss": 0.0645,
+      "step": 1210
+    },
+    {
+      "epoch": 1.863076923076923,
+      "grad_norm": 0.33752078106966593,
+      "learning_rate": 1.1085332234173664e-05,
+      "loss": 0.0724,
+      "step": 1211
+    },
+    {
+      "epoch": 1.8646153846153846,
+      "grad_norm": 0.33949138455564243,
+      "learning_rate": 1.1073319772851299e-05,
+      "loss": 0.0635,
+      "step": 1212
+    },
+    {
+      "epoch": 1.8661538461538463,
+      "grad_norm": 0.339715631924927,
+      "learning_rate": 1.106130574448156e-05,
+      "loss": 0.0708,
+      "step": 1213
+    },
+    {
+      "epoch": 1.8676923076923075,
+      "grad_norm": 0.38187248950026703,
+      "learning_rate": 1.1049290166604928e-05,
+      "loss": 0.0648,
+      "step": 1214
+    },
+    {
+      "epoch": 1.8692307692307693,
+      "grad_norm": 0.4755273814617505,
+      "learning_rate": 1.1037273056764157e-05,
+      "loss": 0.0676,
+      "step": 1215
+    },
+    {
+      "epoch": 1.8707692307692307,
+      "grad_norm": 0.35341205325615144,
+      "learning_rate": 1.1025254432504234e-05,
+      "loss": 0.0742,
+      "step": 1216
+    },
+    {
+      "epoch": 1.8723076923076922,
+      "grad_norm": 0.3254062177799031,
+      "learning_rate": 1.1013234311372353e-05,
+      "loss": 0.0649,
+      "step": 1217
+    },
+    {
+      "epoch": 1.873846153846154,
+      "grad_norm": 0.31246579893201815,
+      "learning_rate": 1.1001212710917897e-05,
+      "loss": 0.0649,
+      "step": 1218
+    },
+    {
+      "epoch": 1.8753846153846154,
+      "grad_norm": 0.33928603890752973,
+      "learning_rate": 1.0989189648692408e-05,
+      "loss": 0.07,
+      "step": 1219
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 0.3450403297222314,
+      "learning_rate": 1.0977165142249566e-05,
+      "loss": 0.067,
+      "step": 1220
+    },
+    {
+      "epoch": 1.8784615384615386,
+      "grad_norm": 0.3423853230449218,
+      "learning_rate": 1.0965139209145153e-05,
+      "loss": 0.0702,
+      "step": 1221
+    },
+    {
+      "epoch": 1.88,
+      "grad_norm": 0.32174442781609414,
+      "learning_rate": 1.0953111866937038e-05,
+      "loss": 0.0701,
+      "step": 1222
+    },
+    {
+      "epoch": 1.8815384615384616,
+      "grad_norm": 0.3676532027177272,
+      "learning_rate": 1.0941083133185146e-05,
+      "loss": 0.0705,
+      "step": 1223
+    },
+    {
+      "epoch": 1.883076923076923,
+      "grad_norm": 0.41997872234324835,
+      "learning_rate": 1.0929053025451432e-05,
+      "loss": 0.0659,
+      "step": 1224
+    },
+    {
+      "epoch": 1.8846153846153846,
+      "grad_norm": 0.4110637997348029,
+      "learning_rate": 1.0917021561299864e-05,
+      "loss": 0.0729,
+      "step": 1225
+    },
+    {
+      "epoch": 1.8861538461538463,
+      "grad_norm": 0.38396879751089963,
+      "learning_rate": 1.090498875829638e-05,
+      "loss": 0.0786,
+      "step": 1226
+    },
+    {
+      "epoch": 1.8876923076923076,
+      "grad_norm": 0.3530760565665137,
+      "learning_rate": 1.089295463400888e-05,
+      "loss": 0.0736,
+      "step": 1227
+    },
+    {
+      "epoch": 1.8892307692307693,
+      "grad_norm": 0.3828535267508954,
+      "learning_rate": 1.0880919206007193e-05,
+      "loss": 0.0756,
+      "step": 1228
+    },
+    {
+      "epoch": 1.8907692307692308,
+      "grad_norm": 0.4620093651236917,
+      "learning_rate": 1.0868882491863048e-05,
+      "loss": 0.0672,
+      "step": 1229
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 0.34394347660813923,
+      "learning_rate": 1.0856844509150056e-05,
+      "loss": 0.0628,
+      "step": 1230
+    },
+    {
+      "epoch": 1.893846153846154,
+      "grad_norm": 0.34240923368033127,
+      "learning_rate": 1.0844805275443673e-05,
+      "loss": 0.0663,
+      "step": 1231
+    },
+    {
+      "epoch": 1.8953846153846152,
+      "grad_norm": 0.4216830435498585,
+      "learning_rate": 1.0832764808321186e-05,
+      "loss": 0.0731,
+      "step": 1232
+    },
+    {
+      "epoch": 1.896923076923077,
+      "grad_norm": 0.34817819208217915,
+      "learning_rate": 1.0820723125361685e-05,
+      "loss": 0.0696,
+      "step": 1233
+    },
+    {
+      "epoch": 1.8984615384615384,
+      "grad_norm": 0.3361061093793128,
+      "learning_rate": 1.0808680244146035e-05,
+      "loss": 0.0608,
+      "step": 1234
+    },
+    {
+      "epoch": 1.9,
+      "grad_norm": 0.3127731515392646,
+      "learning_rate": 1.0796636182256846e-05,
+      "loss": 0.0654,
+      "step": 1235
+    },
+    {
+      "epoch": 1.9015384615384616,
+      "grad_norm": 0.34974984864646785,
+      "learning_rate": 1.0784590957278452e-05,
+      "loss": 0.0696,
+      "step": 1236
+    },
+    {
+      "epoch": 1.903076923076923,
+      "grad_norm": 0.3241892888461775,
+      "learning_rate": 1.077254458679689e-05,
+      "loss": 0.0615,
+      "step": 1237
+    },
+    {
+      "epoch": 1.9046153846153846,
+      "grad_norm": 0.37471636624302423,
+      "learning_rate": 1.0760497088399863e-05,
+      "loss": 0.0657,
+      "step": 1238
+    },
+    {
+      "epoch": 1.9061538461538463,
+      "grad_norm": 0.4476088794896166,
+      "learning_rate": 1.074844847967673e-05,
+      "loss": 0.0696,
+      "step": 1239
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 0.35752628031586536,
+      "learning_rate": 1.0736398778218458e-05,
+      "loss": 0.0662,
+      "step": 1240
+    },
+    {
+      "epoch": 1.9092307692307693,
+      "grad_norm": 0.40745054566528704,
+      "learning_rate": 1.0724348001617626e-05,
+      "loss": 0.0688,
+      "step": 1241
+    },
+    {
+      "epoch": 1.9107692307692308,
+      "grad_norm": 0.39673775592008276,
+      "learning_rate": 1.0712296167468366e-05,
+      "loss": 0.0636,
+      "step": 1242
+    },
+    {
+      "epoch": 1.9123076923076923,
+      "grad_norm": 0.36255910467507374,
+      "learning_rate": 1.0700243293366365e-05,
+      "loss": 0.0606,
+      "step": 1243
+    },
+    {
+      "epoch": 1.913846153846154,
+      "grad_norm": 0.5136376481460118,
+      "learning_rate": 1.0688189396908826e-05,
+      "loss": 0.0645,
+      "step": 1244
+    },
+    {
+      "epoch": 1.9153846153846152,
+      "grad_norm": 0.3099567930694205,
+      "learning_rate": 1.0676134495694439e-05,
+      "loss": 0.0644,
+      "step": 1245
+    },
+    {
+      "epoch": 1.916923076923077,
+      "grad_norm": 0.37000890844695966,
+      "learning_rate": 1.0664078607323367e-05,
+      "loss": 0.0687,
+      "step": 1246
+    },
+    {
+      "epoch": 1.9184615384615384,
+      "grad_norm": 0.34210135022540517,
+      "learning_rate": 1.0652021749397216e-05,
+      "loss": 0.0699,
+      "step": 1247
+    },
+    {
+      "epoch": 1.92,
+      "grad_norm": 0.3609562310788875,
+      "learning_rate": 1.0639963939519005e-05,
+      "loss": 0.0738,
+      "step": 1248
+    },
+    {
+      "epoch": 1.9215384615384616,
+      "grad_norm": 0.3700975852964543,
+      "learning_rate": 1.0627905195293135e-05,
+      "loss": 0.0664,
+      "step": 1249
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.3859166868285145,
+      "learning_rate": 1.0615845534325384e-05,
+      "loss": 0.0713,
+      "step": 1250
+    },
+    {
+      "epoch": 1.9246153846153846,
+      "grad_norm": 0.4115264368388931,
+      "learning_rate": 1.0603784974222862e-05,
+      "loss": 0.0705,
+      "step": 1251
+    },
+    {
+      "epoch": 1.926153846153846,
+      "grad_norm": 0.3145083279396702,
+      "learning_rate": 1.0591723532593992e-05,
+      "loss": 0.0604,
+      "step": 1252
+    },
+    {
+      "epoch": 1.9276923076923076,
+      "grad_norm": 0.3197633328292509,
+      "learning_rate": 1.0579661227048484e-05,
+      "loss": 0.0604,
+      "step": 1253
+    },
+    {
+      "epoch": 1.9292307692307693,
+      "grad_norm": 0.3622577046638927,
+      "learning_rate": 1.056759807519731e-05,
+      "loss": 0.0716,
+      "step": 1254
+    },
+    {
+      "epoch": 1.9307692307692308,
+      "grad_norm": 0.3497451203917846,
+      "learning_rate": 1.0555534094652675e-05,
+      "loss": 0.0673,
+      "step": 1255
+    },
+    {
+      "epoch": 1.9323076923076923,
+      "grad_norm": 0.3320313053856611,
+      "learning_rate": 1.0543469303028002e-05,
+      "loss": 0.0635,
+      "step": 1256
+    },
+    {
+      "epoch": 1.933846153846154,
+      "grad_norm": 0.3233116362666333,
+      "learning_rate": 1.0531403717937888e-05,
+      "loss": 0.062,
+      "step": 1257
+    },
+    {
+      "epoch": 1.9353846153846153,
+      "grad_norm": 0.3685802421725291,
+      "learning_rate": 1.0519337356998094e-05,
+      "loss": 0.0692,
+      "step": 1258
+    },
+    {
+      "epoch": 1.936923076923077,
+      "grad_norm": 0.336210753592891,
+      "learning_rate": 1.0507270237825513e-05,
+      "loss": 0.0675,
+      "step": 1259
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 0.3289919747166384,
+      "learning_rate": 1.0495202378038144e-05,
+      "loss": 0.0657,
+      "step": 1260
+    },
+    {
+      "epoch": 1.94,
+      "grad_norm": 0.33651073591370484,
+      "learning_rate": 1.0483133795255072e-05,
+      "loss": 0.0717,
+      "step": 1261
+    },
+    {
+      "epoch": 1.9415384615384617,
+      "grad_norm": 0.3648322054454426,
+      "learning_rate": 1.0471064507096427e-05,
+      "loss": 0.0793,
+      "step": 1262
+    },
+    {
+      "epoch": 1.943076923076923,
+      "grad_norm": 0.33381701009336806,
+      "learning_rate": 1.045899453118338e-05,
+      "loss": 0.0617,
+      "step": 1263
+    },
+    {
+      "epoch": 1.9446153846153846,
+      "grad_norm": 0.3641612409279995,
+      "learning_rate": 1.0446923885138101e-05,
+      "loss": 0.0597,
+      "step": 1264
+    },
+    {
+      "epoch": 1.9461538461538461,
+      "grad_norm": 0.34585249190313433,
+      "learning_rate": 1.0434852586583737e-05,
+      "loss": 0.0674,
+      "step": 1265
+    },
+    {
+      "epoch": 1.9476923076923076,
+      "grad_norm": 0.3289388167113336,
+      "learning_rate": 1.0422780653144392e-05,
+      "loss": 0.0735,
+      "step": 1266
+    },
+    {
+      "epoch": 1.9492307692307693,
+      "grad_norm": 0.3887566915602357,
+      "learning_rate": 1.0410708102445091e-05,
+      "loss": 0.0766,
+      "step": 1267
+    },
+    {
+      "epoch": 1.9507692307692308,
+      "grad_norm": 0.30555149229418166,
+      "learning_rate": 1.0398634952111766e-05,
+      "loss": 0.0718,
+      "step": 1268
+    },
+    {
+      "epoch": 1.9523076923076923,
+      "grad_norm": 0.32317250467343567,
+      "learning_rate": 1.0386561219771222e-05,
+      "loss": 0.0652,
+      "step": 1269
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 0.3006439298211834,
+      "learning_rate": 1.0374486923051117e-05,
+      "loss": 0.0623,
+      "step": 1270
+    },
+    {
+      "epoch": 1.9553846153846153,
+      "grad_norm": 0.33386303419709507,
+      "learning_rate": 1.0362412079579925e-05,
+      "loss": 0.0701,
+      "step": 1271
+    },
+    {
+      "epoch": 1.956923076923077,
+      "grad_norm": 0.33099092128522206,
+      "learning_rate": 1.0350336706986925e-05,
+      "loss": 0.068,
+      "step": 1272
+    },
+    {
+      "epoch": 1.9584615384615385,
+      "grad_norm": 0.3262986897907234,
+      "learning_rate": 1.0338260822902166e-05,
+      "loss": 0.0614,
+      "step": 1273
+    },
+    {
+      "epoch": 1.96,
+      "grad_norm": 0.34485599175028897,
+      "learning_rate": 1.0326184444956449e-05,
+      "loss": 0.0658,
+      "step": 1274
+    },
+    {
+      "epoch": 1.9615384615384617,
+      "grad_norm": 0.31792277168038097,
+      "learning_rate": 1.0314107590781284e-05,
+      "loss": 0.0574,
+      "step": 1275
+    },
+    {
+      "epoch": 1.963076923076923,
+      "grad_norm": 0.3208911821070251,
+      "learning_rate": 1.030203027800889e-05,
+      "loss": 0.0633,
+      "step": 1276
+    },
+    {
+      "epoch": 1.9646153846153847,
+      "grad_norm": 0.33055045597617766,
+      "learning_rate": 1.0289952524272147e-05,
+      "loss": 0.0706,
+      "step": 1277
+    },
+    {
+      "epoch": 1.9661538461538461,
+      "grad_norm": 0.34088210301412697,
+      "learning_rate": 1.027787434720458e-05,
+      "loss": 0.0739,
+      "step": 1278
+    },
+    {
+      "epoch": 1.9676923076923076,
+      "grad_norm": 0.32073888049730687,
+      "learning_rate": 1.0265795764440335e-05,
+      "loss": 0.0651,
+      "step": 1279
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 0.3076473967656385,
+      "learning_rate": 1.025371679361415e-05,
+      "loss": 0.0683,
+      "step": 1280
+    },
+    {
+      "epoch": 1.9707692307692306,
+      "grad_norm": 0.3621459788069735,
+      "learning_rate": 1.0241637452361323e-05,
+      "loss": 0.0677,
+      "step": 1281
+    },
+    {
+      "epoch": 1.9723076923076923,
+      "grad_norm": 0.3573542495717642,
+      "learning_rate": 1.0229557758317703e-05,
+      "loss": 0.0679,
+      "step": 1282
+    },
+    {
+      "epoch": 1.9738461538461538,
+      "grad_norm": 0.3483850527238064,
+      "learning_rate": 1.0217477729119648e-05,
+      "loss": 0.074,
+      "step": 1283
+    },
+    {
+      "epoch": 1.9753846153846153,
+      "grad_norm": 0.3063790927358412,
+      "learning_rate": 1.0205397382404006e-05,
+      "loss": 0.0652,
+      "step": 1284
+    },
+    {
+      "epoch": 1.976923076923077,
+      "grad_norm": 0.34619566294720616,
+      "learning_rate": 1.0193316735808085e-05,
+      "loss": 0.0676,
+      "step": 1285
+    },
+    {
+      "epoch": 1.9784615384615385,
+      "grad_norm": 0.46164644987132686,
+      "learning_rate": 1.018123580696964e-05,
+      "loss": 0.0711,
+      "step": 1286
+    },
+    {
+      "epoch": 1.98,
+      "grad_norm": 0.33352482622772156,
+      "learning_rate": 1.0169154613526831e-05,
+      "loss": 0.0696,
+      "step": 1287
+    },
+    {
+      "epoch": 1.9815384615384617,
+      "grad_norm": 0.3101921292214142,
+      "learning_rate": 1.0157073173118207e-05,
+      "loss": 0.0647,
+      "step": 1288
+    },
+    {
+      "epoch": 1.983076923076923,
+      "grad_norm": 0.31015657112545686,
+      "learning_rate": 1.0144991503382676e-05,
+      "loss": 0.0614,
+      "step": 1289
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 0.3503219066829283,
+      "learning_rate": 1.0132909621959482e-05,
+      "loss": 0.077,
+      "step": 1290
+    },
+    {
+      "epoch": 1.9861538461538462,
+      "grad_norm": 0.2992678727460164,
+      "learning_rate": 1.0120827546488175e-05,
+      "loss": 0.065,
+      "step": 1291
+    },
+    {
+      "epoch": 1.9876923076923076,
+      "grad_norm": 0.3232588558361219,
+      "learning_rate": 1.0108745294608595e-05,
+      "loss": 0.0596,
+      "step": 1292
+    },
+    {
+      "epoch": 1.9892307692307694,
+      "grad_norm": 0.3219005317029137,
+      "learning_rate": 1.0096662883960833e-05,
+      "loss": 0.0693,
+      "step": 1293
+    },
+    {
+      "epoch": 1.9907692307692306,
+      "grad_norm": 0.36479544578271617,
+      "learning_rate": 1.0084580332185214e-05,
+      "loss": 0.071,
+      "step": 1294
+    },
+    {
+      "epoch": 1.9923076923076923,
+      "grad_norm": 0.36208947868270336,
+      "learning_rate": 1.0072497656922266e-05,
+      "loss": 0.0697,
+      "step": 1295
+    },
+    {
+      "epoch": 1.9938461538461538,
+      "grad_norm": 0.374662600377983,
+      "learning_rate": 1.0060414875812709e-05,
+      "loss": 0.0724,
+      "step": 1296
+    },
+    {
+      "epoch": 1.9953846153846153,
+      "grad_norm": 0.4368778771016829,
+      "learning_rate": 1.0048332006497406e-05,
+      "loss": 0.0777,
+      "step": 1297
+    },
+    {
+      "epoch": 1.996923076923077,
+      "grad_norm": 0.40482597277564725,
+      "learning_rate": 1.003624906661735e-05,
+      "loss": 0.0672,
+      "step": 1298
+    },
+    {
+      "epoch": 1.9984615384615385,
+      "grad_norm": 0.3944268530980909,
+      "learning_rate": 1.0024166073813634e-05,
+      "loss": 0.0619,
+      "step": 1299
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.3312157049217859,
+      "learning_rate": 1.0012083045727445e-05,
+      "loss": 0.0633,
+      "step": 1300
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2600,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 4,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 263636178075648.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}