diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22091 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998958772343011,
+  "eval_steps": 500,
+  "global_step": 3151,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003173265240350051,
+      "grad_norm": 10.848901748657227,
+      "learning_rate": 0.0,
+      "loss": 1.1168,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006346530480700102,
+      "grad_norm": 10.646440505981445,
+      "learning_rate": 3.1645569620253166e-08,
+      "loss": 1.0713,
+      "step": 2
+    },
+    {
+      "epoch": 0.0009519795721050152,
+      "grad_norm": 18.729393005371094,
+      "learning_rate": 6.329113924050633e-08,
+      "loss": 1.1188,
+      "step": 3
+    },
+    {
+      "epoch": 0.0012693060961400204,
+      "grad_norm": 11.392064094543457,
+      "learning_rate": 9.49367088607595e-08,
+      "loss": 1.14,
+      "step": 4
+    },
+    {
+      "epoch": 0.0015866326201750254,
+      "grad_norm": 10.984261512756348,
+      "learning_rate": 1.2658227848101266e-07,
+      "loss": 1.1239,
+      "step": 5
+    },
+    {
+      "epoch": 0.0019039591442100304,
+      "grad_norm": 10.374354362487793,
+      "learning_rate": 1.5822784810126584e-07,
+      "loss": 1.0937,
+      "step": 6
+    },
+    {
+      "epoch": 0.0022212856682450356,
+      "grad_norm": 10.874767303466797,
+      "learning_rate": 1.89873417721519e-07,
+      "loss": 1.0982,
+      "step": 7
+    },
+    {
+      "epoch": 0.002538612192280041,
+      "grad_norm": 11.027266502380371,
+      "learning_rate": 2.2151898734177218e-07,
+      "loss": 1.1157,
+      "step": 8
+    },
+    {
+      "epoch": 0.0028559387163150456,
+      "grad_norm": 10.416919708251953,
+      "learning_rate": 2.5316455696202533e-07,
+      "loss": 1.0905,
+      "step": 9
+    },
+    {
+      "epoch": 0.003173265240350051,
+      "grad_norm": 10.194792747497559,
+      "learning_rate": 2.848101265822785e-07,
+      "loss": 1.0472,
+      "step": 10
+    },
+    {
+      "epoch": 0.003490591764385056,
+      "grad_norm": 10.86874771118164,
+      "learning_rate": 3.164556962025317e-07,
+      "loss": 1.1327,
+      "step": 11
+    },
+    {
+      "epoch": 0.003807918288420061,
+      "grad_norm": 10.13455581665039,
+      "learning_rate": 3.481012658227848e-07,
+      "loss": 1.0522,
+      "step": 12
+    },
+    {
+      "epoch": 0.004125244812455066,
+      "grad_norm": 10.391048431396484,
+      "learning_rate": 3.79746835443038e-07,
+      "loss": 1.0625,
+      "step": 13
+    },
+    {
+      "epoch": 0.004442571336490071,
+      "grad_norm": 10.643158912658691,
+      "learning_rate": 4.1139240506329117e-07,
+      "loss": 1.0626,
+      "step": 14
+    },
+    {
+      "epoch": 0.0047598978605250765,
+      "grad_norm": 10.64275074005127,
+      "learning_rate": 4.4303797468354435e-07,
+      "loss": 1.1172,
+      "step": 15
+    },
+    {
+      "epoch": 0.005077224384560082,
+      "grad_norm": 10.97158145904541,
+      "learning_rate": 4.7468354430379753e-07,
+      "loss": 1.1438,
+      "step": 16
+    },
+    {
+      "epoch": 0.005394550908595086,
+      "grad_norm": 10.86425495147705,
+      "learning_rate": 5.063291139240507e-07,
+      "loss": 1.0899,
+      "step": 17
+    },
+    {
+      "epoch": 0.005711877432630091,
+      "grad_norm": 10.342445373535156,
+      "learning_rate": 5.379746835443038e-07,
+      "loss": 1.0642,
+      "step": 18
+    },
+    {
+      "epoch": 0.0060292039566650964,
+      "grad_norm": 10.646989822387695,
+      "learning_rate": 5.69620253164557e-07,
+      "loss": 1.0734,
+      "step": 19
+    },
+    {
+      "epoch": 0.006346530480700102,
+      "grad_norm": 10.075757026672363,
+      "learning_rate": 6.012658227848102e-07,
+      "loss": 1.0498,
+      "step": 20
+    },
+    {
+      "epoch": 0.006663857004735107,
+      "grad_norm": 10.03166675567627,
+      "learning_rate": 6.329113924050634e-07,
+      "loss": 1.0402,
+      "step": 21
+    },
+    {
+      "epoch": 0.006981183528770112,
+      "grad_norm": 10.57165241241455,
+      "learning_rate": 6.645569620253164e-07,
+      "loss": 1.0836,
+      "step": 22
+    },
+    {
+      "epoch": 0.007298510052805117,
+      "grad_norm": 10.821656227111816,
+      "learning_rate": 6.962025316455696e-07,
+      "loss": 1.0758,
+      "step": 23
+    },
+    {
+      "epoch": 0.007615836576840122,
+      "grad_norm": 8.897995948791504,
+      "learning_rate": 7.278481012658228e-07,
+      "loss": 0.9763,
+      "step": 24
+    },
+    {
+      "epoch": 0.007933163100875128,
+      "grad_norm": 8.885331153869629,
+      "learning_rate": 7.59493670886076e-07,
+      "loss": 0.9519,
+      "step": 25
+    },
+    {
+      "epoch": 0.008250489624910132,
+      "grad_norm": 8.776424407958984,
+      "learning_rate": 7.911392405063293e-07,
+      "loss": 0.9289,
+      "step": 26
+    },
+    {
+      "epoch": 0.008567816148945136,
+      "grad_norm": 8.42698860168457,
+      "learning_rate": 8.227848101265823e-07,
+      "loss": 0.9104,
+      "step": 27
+    },
+    {
+      "epoch": 0.008885142672980142,
+      "grad_norm": 8.485051155090332,
+      "learning_rate": 8.544303797468355e-07,
+      "loss": 0.9074,
+      "step": 28
+    },
+    {
+      "epoch": 0.009202469197015147,
+      "grad_norm": 8.01269817352295,
+      "learning_rate": 8.860759493670887e-07,
+      "loss": 0.8682,
+      "step": 29
+    },
+    {
+      "epoch": 0.009519795721050153,
+      "grad_norm": 8.731021881103516,
+      "learning_rate": 9.177215189873419e-07,
+      "loss": 0.9172,
+      "step": 30
+    },
+    {
+      "epoch": 0.009837122245085157,
+      "grad_norm": 7.760385036468506,
+      "learning_rate": 9.493670886075951e-07,
+      "loss": 0.8467,
+      "step": 31
+    },
+    {
+      "epoch": 0.010154448769120163,
+      "grad_norm": 8.05570125579834,
+      "learning_rate": 9.810126582278482e-07,
+      "loss": 0.8499,
+      "step": 32
+    },
+    {
+      "epoch": 0.010471775293155168,
+      "grad_norm": 8.80942440032959,
+      "learning_rate": 1.0126582278481013e-06,
+      "loss": 0.7083,
+      "step": 33
+    },
+    {
+      "epoch": 0.010789101817190172,
+      "grad_norm": 5.450026035308838,
+      "learning_rate": 1.0443037974683546e-06,
+      "loss": 0.6748,
+      "step": 34
+    },
+    {
+      "epoch": 0.011106428341225178,
+      "grad_norm": 5.143744468688965,
+      "learning_rate": 1.0759493670886077e-06,
+      "loss": 0.6462,
+      "step": 35
+    },
+    {
+      "epoch": 0.011423754865260182,
+      "grad_norm": 5.061795711517334,
+      "learning_rate": 1.1075949367088607e-06,
+      "loss": 0.6293,
+      "step": 36
+    },
+    {
+      "epoch": 0.011741081389295189,
+      "grad_norm": 4.609454154968262,
+      "learning_rate": 1.139240506329114e-06,
+      "loss": 0.5921,
+      "step": 37
+    },
+    {
+      "epoch": 0.012058407913330193,
+      "grad_norm": 4.469468116760254,
+      "learning_rate": 1.170886075949367e-06,
+      "loss": 0.6028,
+      "step": 38
+    },
+    {
+      "epoch": 0.012375734437365199,
+      "grad_norm": 4.329920768737793,
+      "learning_rate": 1.2025316455696204e-06,
+      "loss": 0.5507,
+      "step": 39
+    },
+    {
+      "epoch": 0.012693060961400203,
+      "grad_norm": 4.549306869506836,
+      "learning_rate": 1.2341772151898737e-06,
+      "loss": 0.5558,
+      "step": 40
+    },
+    {
+      "epoch": 0.013010387485435208,
+      "grad_norm": 4.089702129364014,
+      "learning_rate": 1.2658227848101267e-06,
+      "loss": 0.5482,
+      "step": 41
+    },
+    {
+      "epoch": 0.013327714009470214,
+      "grad_norm": 3.972285270690918,
+      "learning_rate": 1.2974683544303798e-06,
+      "loss": 0.5172,
+      "step": 42
+    },
+    {
+      "epoch": 0.013645040533505218,
+      "grad_norm": 4.000632286071777,
+      "learning_rate": 1.3291139240506329e-06,
+      "loss": 0.5095,
+      "step": 43
+    },
+    {
+      "epoch": 0.013962367057540224,
+      "grad_norm": 3.4380977153778076,
+      "learning_rate": 1.3607594936708862e-06,
+      "loss": 0.4316,
+      "step": 44
+    },
+    {
+      "epoch": 0.014279693581575229,
+      "grad_norm": 3.15617036819458,
+      "learning_rate": 1.3924050632911392e-06,
+      "loss": 0.4038,
+      "step": 45
+    },
+    {
+      "epoch": 0.014597020105610235,
+      "grad_norm": 2.7378439903259277,
+      "learning_rate": 1.4240506329113925e-06,
+      "loss": 0.3703,
+      "step": 46
+    },
+    {
+      "epoch": 0.014914346629645239,
+      "grad_norm": 2.536205291748047,
+      "learning_rate": 1.4556962025316456e-06,
+      "loss": 0.3511,
+      "step": 47
+    },
+    {
+      "epoch": 0.015231673153680243,
+      "grad_norm": 2.604930877685547,
+      "learning_rate": 1.4873417721518987e-06,
+      "loss": 0.3224,
+      "step": 48
+    },
+    {
+      "epoch": 0.01554899967771525,
+      "grad_norm": 2.1545419692993164,
+      "learning_rate": 1.518987341772152e-06,
+      "loss": 0.3038,
+      "step": 49
+    },
+    {
+      "epoch": 0.015866326201750255,
+      "grad_norm": 2.231151819229126,
+      "learning_rate": 1.550632911392405e-06,
+      "loss": 0.2707,
+      "step": 50
+    },
+    {
+      "epoch": 0.016183652725785258,
+      "grad_norm": 2.079399585723877,
+      "learning_rate": 1.5822784810126585e-06,
+      "loss": 0.277,
+      "step": 51
+    },
+    {
+      "epoch": 0.016500979249820264,
+      "grad_norm": 2.2018184661865234,
+      "learning_rate": 1.6139240506329116e-06,
+      "loss": 0.2528,
+      "step": 52
+    },
+    {
+      "epoch": 0.01681830577385527,
+      "grad_norm": 1.8787950277328491,
+      "learning_rate": 1.6455696202531647e-06,
+      "loss": 0.2448,
+      "step": 53
+    },
+    {
+      "epoch": 0.017135632297890273,
+      "grad_norm": 1.8153671026229858,
+      "learning_rate": 1.677215189873418e-06,
+      "loss": 0.241,
+      "step": 54
+    },
+    {
+      "epoch": 0.01745295882192528,
+      "grad_norm": 1.7260993719100952,
+      "learning_rate": 1.708860759493671e-06,
+      "loss": 0.2316,
+      "step": 55
+    },
+    {
+      "epoch": 0.017770285345960285,
+      "grad_norm": 2.256204128265381,
+      "learning_rate": 1.7405063291139243e-06,
+      "loss": 0.2245,
+      "step": 56
+    },
+    {
+      "epoch": 0.01808761186999529,
+      "grad_norm": 1.5609785318374634,
+      "learning_rate": 1.7721518987341774e-06,
+      "loss": 0.2088,
+      "step": 57
+    },
+    {
+      "epoch": 0.018404938394030294,
+      "grad_norm": 1.4666439294815063,
+      "learning_rate": 1.8037974683544305e-06,
+      "loss": 0.1954,
+      "step": 58
+    },
+    {
+      "epoch": 0.0187222649180653,
+      "grad_norm": 1.4496690034866333,
+      "learning_rate": 1.8354430379746838e-06,
+      "loss": 0.1939,
+      "step": 59
+    },
+    {
+      "epoch": 0.019039591442100306,
+      "grad_norm": 1.525081992149353,
+      "learning_rate": 1.8670886075949368e-06,
+      "loss": 0.1927,
+      "step": 60
+    },
+    {
+      "epoch": 0.01935691796613531,
+      "grad_norm": 1.402138590812683,
+      "learning_rate": 1.8987341772151901e-06,
+      "loss": 0.1764,
+      "step": 61
+    },
+    {
+      "epoch": 0.019674244490170315,
+      "grad_norm": 1.702570915222168,
+      "learning_rate": 1.930379746835443e-06,
+      "loss": 0.1757,
+      "step": 62
+    },
+    {
+      "epoch": 0.01999157101420532,
+      "grad_norm": 1.2852612733840942,
+      "learning_rate": 1.9620253164556965e-06,
+      "loss": 0.1669,
+      "step": 63
+    },
+    {
+      "epoch": 0.020308897538240327,
+      "grad_norm": 1.3921505212783813,
+      "learning_rate": 1.9936708860759498e-06,
+      "loss": 0.1587,
+      "step": 64
+    },
+    {
+      "epoch": 0.02062622406227533,
+      "grad_norm": 1.2438849210739136,
+      "learning_rate": 2.0253164556962026e-06,
+      "loss": 0.1636,
+      "step": 65
+    },
+    {
+      "epoch": 0.020943550586310335,
+      "grad_norm": 1.7521835565567017,
+      "learning_rate": 2.056962025316456e-06,
+      "loss": 0.1473,
+      "step": 66
+    },
+    {
+      "epoch": 0.02126087711034534,
+      "grad_norm": 1.8567091226577759,
+      "learning_rate": 2.088607594936709e-06,
+      "loss": 0.1542,
+      "step": 67
+    },
+    {
+      "epoch": 0.021578203634380344,
+      "grad_norm": 1.4712021350860596,
+      "learning_rate": 2.120253164556962e-06,
+      "loss": 0.1594,
+      "step": 68
+    },
+    {
+      "epoch": 0.02189553015841535,
+      "grad_norm": 1.8508087396621704,
+      "learning_rate": 2.1518987341772153e-06,
+      "loss": 0.1337,
+      "step": 69
+    },
+    {
+      "epoch": 0.022212856682450356,
+      "grad_norm": 3.110574722290039,
+      "learning_rate": 2.1835443037974686e-06,
+      "loss": 0.1472,
+      "step": 70
+    },
+    {
+      "epoch": 0.022530183206485362,
+      "grad_norm": 1.2160632610321045,
+      "learning_rate": 2.2151898734177215e-06,
+      "loss": 0.143,
+      "step": 71
+    },
+    {
+      "epoch": 0.022847509730520365,
+      "grad_norm": 1.1466506719589233,
+      "learning_rate": 2.2468354430379748e-06,
+      "loss": 0.1243,
+      "step": 72
+    },
+    {
+      "epoch": 0.02316483625455537,
+      "grad_norm": 5.218608856201172,
+      "learning_rate": 2.278481012658228e-06,
+      "loss": 0.1309,
+      "step": 73
+    },
+    {
+      "epoch": 0.023482162778590377,
+      "grad_norm": 1.1374642848968506,
+      "learning_rate": 2.3101265822784813e-06,
+      "loss": 0.1449,
+      "step": 74
+    },
+    {
+      "epoch": 0.02379948930262538,
+      "grad_norm": 1.4869107007980347,
+      "learning_rate": 2.341772151898734e-06,
+      "loss": 0.1191,
+      "step": 75
+    },
+    {
+      "epoch": 0.024116815826660386,
+      "grad_norm": 1.077081561088562,
+      "learning_rate": 2.3734177215189875e-06,
+      "loss": 0.1342,
+      "step": 76
+    },
+    {
+      "epoch": 0.024434142350695392,
+      "grad_norm": 0.8993155360221863,
+      "learning_rate": 2.4050632911392408e-06,
+      "loss": 0.1203,
+      "step": 77
+    },
+    {
+      "epoch": 0.024751468874730398,
+      "grad_norm": 0.9718343019485474,
+      "learning_rate": 2.4367088607594936e-06,
+      "loss": 0.1166,
+      "step": 78
+    },
+    {
+      "epoch": 0.0250687953987654,
+      "grad_norm": 1.0919193029403687,
+      "learning_rate": 2.4683544303797473e-06,
+      "loss": 0.1159,
+      "step": 79
+    },
+    {
+      "epoch": 0.025386121922800407,
+      "grad_norm": 0.9666793942451477,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1236,
+      "step": 80
+    },
+    {
+      "epoch": 0.025703448446835413,
+      "grad_norm": 0.9448258280754089,
+      "learning_rate": 2.5316455696202535e-06,
+      "loss": 0.1172,
+      "step": 81
+    },
+    {
+      "epoch": 0.026020774970870415,
+      "grad_norm": 1.0400398969650269,
+      "learning_rate": 2.5632911392405064e-06,
+      "loss": 0.1193,
+      "step": 82
+    },
+    {
+      "epoch": 0.02633810149490542,
+      "grad_norm": 0.8915061354637146,
+      "learning_rate": 2.5949367088607596e-06,
+      "loss": 0.1081,
+      "step": 83
+    },
+    {
+      "epoch": 0.026655428018940427,
+      "grad_norm": 0.8930220007896423,
+      "learning_rate": 2.626582278481013e-06,
+      "loss": 0.1135,
+      "step": 84
+    },
+    {
+      "epoch": 0.026972754542975434,
+      "grad_norm": 1.110547661781311,
+      "learning_rate": 2.6582278481012658e-06,
+      "loss": 0.1222,
+      "step": 85
+    },
+    {
+      "epoch": 0.027290081067010436,
+      "grad_norm": 1.0935736894607544,
+      "learning_rate": 2.689873417721519e-06,
+      "loss": 0.1054,
+      "step": 86
+    },
+    {
+      "epoch": 0.027607407591045442,
+      "grad_norm": 0.9467206001281738,
+      "learning_rate": 2.7215189873417724e-06,
+      "loss": 0.1059,
+      "step": 87
+    },
+    {
+      "epoch": 0.02792473411508045,
+      "grad_norm": 0.791374683380127,
+      "learning_rate": 2.7531645569620256e-06,
+      "loss": 0.1019,
+      "step": 88
+    },
+    {
+      "epoch": 0.02824206063911545,
+      "grad_norm": 1.3509459495544434,
+      "learning_rate": 2.7848101265822785e-06,
+      "loss": 0.1054,
+      "step": 89
+    },
+    {
+      "epoch": 0.028559387163150457,
+      "grad_norm": 0.8940509557723999,
+      "learning_rate": 2.8164556962025318e-06,
+      "loss": 0.1022,
+      "step": 90
+    },
+    {
+      "epoch": 0.028876713687185463,
+      "grad_norm": 0.920413613319397,
+      "learning_rate": 2.848101265822785e-06,
+      "loss": 0.1011,
+      "step": 91
+    },
+    {
+      "epoch": 0.02919404021122047,
+      "grad_norm": 1.2662198543548584,
+      "learning_rate": 2.879746835443038e-06,
+      "loss": 0.0961,
+      "step": 92
+    },
+    {
+      "epoch": 0.029511366735255472,
+      "grad_norm": 1.05869722366333,
+      "learning_rate": 2.9113924050632912e-06,
+      "loss": 0.0998,
+      "step": 93
+    },
+    {
+      "epoch": 0.029828693259290478,
+      "grad_norm": 0.9342459440231323,
+      "learning_rate": 2.9430379746835445e-06,
+      "loss": 0.1043,
+      "step": 94
+    },
+    {
+      "epoch": 0.030146019783325484,
+      "grad_norm": 0.9262270331382751,
+      "learning_rate": 2.9746835443037974e-06,
+      "loss": 0.1031,
+      "step": 95
+    },
+    {
+      "epoch": 0.030463346307360487,
+      "grad_norm": 0.8977397084236145,
+      "learning_rate": 3.0063291139240506e-06,
+      "loss": 0.0951,
+      "step": 96
+    },
+    {
+      "epoch": 0.030780672831395493,
+      "grad_norm": 0.9080337285995483,
+      "learning_rate": 3.037974683544304e-06,
+      "loss": 0.0967,
+      "step": 97
+    },
+    {
+      "epoch": 0.0310979993554305,
+      "grad_norm": 2.3133058547973633,
+      "learning_rate": 3.0696202531645576e-06,
+      "loss": 0.0949,
+      "step": 98
+    },
+    {
+      "epoch": 0.0314153258794655,
+      "grad_norm": 0.8277332782745361,
+      "learning_rate": 3.10126582278481e-06,
+      "loss": 0.0963,
+      "step": 99
+    },
+    {
+      "epoch": 0.03173265240350051,
+      "grad_norm": 0.7884067296981812,
+      "learning_rate": 3.132911392405064e-06,
+      "loss": 0.0903,
+      "step": 100
+    },
+    {
+      "epoch": 0.032049978927535513,
+      "grad_norm": 1.0298603773117065,
+      "learning_rate": 3.164556962025317e-06,
+      "loss": 0.0903,
+      "step": 101
+    },
+    {
+      "epoch": 0.032367305451570516,
+      "grad_norm": 0.8253909945487976,
+      "learning_rate": 3.1962025316455695e-06,
+      "loss": 0.0868,
+      "step": 102
+    },
+    {
+      "epoch": 0.032684631975605526,
+      "grad_norm": 1.0553210973739624,
+      "learning_rate": 3.2278481012658232e-06,
+      "loss": 0.103,
+      "step": 103
+    },
+    {
+      "epoch": 0.03300195849964053,
+      "grad_norm": 0.7972857356071472,
+      "learning_rate": 3.2594936708860765e-06,
+      "loss": 0.0989,
+      "step": 104
+    },
+    {
+      "epoch": 0.03331928502367553,
+      "grad_norm": 0.9157866835594177,
+      "learning_rate": 3.2911392405063294e-06,
+      "loss": 0.0855,
+      "step": 105
+    },
+    {
+      "epoch": 0.03363661154771054,
+      "grad_norm": 0.798462986946106,
+      "learning_rate": 3.3227848101265827e-06,
+      "loss": 0.0935,
+      "step": 106
+    },
+    {
+      "epoch": 0.03395393807174554,
+      "grad_norm": 0.7930102944374084,
+      "learning_rate": 3.354430379746836e-06,
+      "loss": 0.0938,
+      "step": 107
+    },
+    {
+      "epoch": 0.034271264595780546,
+      "grad_norm": 0.9606144428253174,
+      "learning_rate": 3.386075949367089e-06,
+      "loss": 0.0945,
+      "step": 108
+    },
+    {
+      "epoch": 0.034588591119815555,
+      "grad_norm": 0.9231522083282471,
+      "learning_rate": 3.417721518987342e-06,
+      "loss": 0.0865,
+      "step": 109
+    },
+    {
+      "epoch": 0.03490591764385056,
+      "grad_norm": 0.9299225807189941,
+      "learning_rate": 3.4493670886075954e-06,
+      "loss": 0.0882,
+      "step": 110
+    },
+    {
+      "epoch": 0.03522324416788557,
+      "grad_norm": 0.9133904576301575,
+      "learning_rate": 3.4810126582278487e-06,
+      "loss": 0.0837,
+      "step": 111
+    },
+    {
+      "epoch": 0.03554057069192057,
+      "grad_norm": 0.9055621027946472,
+      "learning_rate": 3.5126582278481015e-06,
+      "loss": 0.0814,
+      "step": 112
+    },
+    {
+      "epoch": 0.03585789721595557,
+      "grad_norm": 0.7394254803657532,
+      "learning_rate": 3.544303797468355e-06,
+      "loss": 0.0797,
+      "step": 113
+    },
+    {
+      "epoch": 0.03617522373999058,
+      "grad_norm": 0.9343309998512268,
+      "learning_rate": 3.575949367088608e-06,
+      "loss": 0.0794,
+      "step": 114
+    },
+    {
+      "epoch": 0.036492550264025585,
+      "grad_norm": 0.8105984330177307,
+      "learning_rate": 3.607594936708861e-06,
+      "loss": 0.0795,
+      "step": 115
+    },
+    {
+      "epoch": 0.03680987678806059,
+      "grad_norm": 3.030195951461792,
+      "learning_rate": 3.6392405063291142e-06,
+      "loss": 0.0857,
+      "step": 116
+    },
+    {
+      "epoch": 0.0371272033120956,
+      "grad_norm": 1.4652644395828247,
+      "learning_rate": 3.6708860759493675e-06,
+      "loss": 0.0804,
+      "step": 117
+    },
+    {
+      "epoch": 0.0374445298361306,
+      "grad_norm": 1.0011448860168457,
+      "learning_rate": 3.7025316455696204e-06,
+      "loss": 0.0759,
+      "step": 118
+    },
+    {
+      "epoch": 0.0377618563601656,
+      "grad_norm": 0.7789915204048157,
+      "learning_rate": 3.7341772151898737e-06,
+      "loss": 0.077,
+      "step": 119
+    },
+    {
+      "epoch": 0.03807918288420061,
+      "grad_norm": 5.796139240264893,
+      "learning_rate": 3.765822784810127e-06,
+      "loss": 0.0789,
+      "step": 120
+    },
+    {
+      "epoch": 0.038396509408235614,
+      "grad_norm": 1.3558639287948608,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 0.0802,
+      "step": 121
+    },
+    {
+      "epoch": 0.03871383593227062,
+      "grad_norm": 0.9625980854034424,
+      "learning_rate": 3.829113924050633e-06,
+      "loss": 0.0872,
+      "step": 122
+    },
+    {
+      "epoch": 0.039031162456305626,
+      "grad_norm": 0.7643633484840393,
+      "learning_rate": 3.860759493670886e-06,
+      "loss": 0.0721,
+      "step": 123
+    },
+    {
+      "epoch": 0.03934848898034063,
+      "grad_norm": 0.9799569249153137,
+      "learning_rate": 3.89240506329114e-06,
+      "loss": 0.0793,
+      "step": 124
+    },
+    {
+      "epoch": 0.03966581550437564,
+      "grad_norm": 0.8280842900276184,
+      "learning_rate": 3.924050632911393e-06,
+      "loss": 0.0782,
+      "step": 125
+    },
+    {
+      "epoch": 0.03998314202841064,
+      "grad_norm": 1.1285597085952759,
+      "learning_rate": 3.955696202531646e-06,
+      "loss": 0.0793,
+      "step": 126
+    },
+    {
+      "epoch": 0.040300468552445644,
+      "grad_norm": 0.9049203991889954,
+      "learning_rate": 3.9873417721518995e-06,
+      "loss": 0.0718,
+      "step": 127
+    },
+    {
+      "epoch": 0.04061779507648065,
+      "grad_norm": 0.7365930080413818,
+      "learning_rate": 4.018987341772152e-06,
+      "loss": 0.0743,
+      "step": 128
+    },
+    {
+      "epoch": 0.040935121600515656,
+      "grad_norm": 0.7526193857192993,
+      "learning_rate": 4.050632911392405e-06,
+      "loss": 0.0765,
+      "step": 129
+    },
+    {
+      "epoch": 0.04125244812455066,
+      "grad_norm": 0.910602331161499,
+      "learning_rate": 4.082278481012659e-06,
+      "loss": 0.0837,
+      "step": 130
+    },
+    {
+      "epoch": 0.04156977464858567,
+      "grad_norm": 0.7549585700035095,
+      "learning_rate": 4.113924050632912e-06,
+      "loss": 0.0776,
+      "step": 131
+    },
+    {
+      "epoch": 0.04188710117262067,
+      "grad_norm": 0.620275616645813,
+      "learning_rate": 4.145569620253165e-06,
+      "loss": 0.073,
+      "step": 132
+    },
+    {
+      "epoch": 0.04220442769665567,
+      "grad_norm": 0.7038996815681458,
+      "learning_rate": 4.177215189873418e-06,
+      "loss": 0.0765,
+      "step": 133
+    },
+    {
+      "epoch": 0.04252175422069068,
+      "grad_norm": 1.0639665126800537,
+      "learning_rate": 4.208860759493671e-06,
+      "loss": 0.0679,
+      "step": 134
+    },
+    {
+      "epoch": 0.042839080744725686,
+      "grad_norm": 0.872046709060669,
+      "learning_rate": 4.240506329113924e-06,
+      "loss": 0.0764,
+      "step": 135
+    },
+    {
+      "epoch": 0.04315640726876069,
+      "grad_norm": 0.8458957076072693,
+      "learning_rate": 4.272151898734178e-06,
+      "loss": 0.0764,
+      "step": 136
+    },
+    {
+      "epoch": 0.0434737337927957,
+      "grad_norm": 0.6807056665420532,
+      "learning_rate": 4.303797468354431e-06,
+      "loss": 0.0669,
+      "step": 137
+    },
+    {
+      "epoch": 0.0437910603168307,
+      "grad_norm": 1.1797236204147339,
+      "learning_rate": 4.3354430379746835e-06,
+      "loss": 0.0783,
+      "step": 138
+    },
+    {
+      "epoch": 0.04410838684086571,
+      "grad_norm": 0.9501144289970398,
+      "learning_rate": 4.367088607594937e-06,
+      "loss": 0.0751,
+      "step": 139
+    },
+    {
+      "epoch": 0.04442571336490071,
+      "grad_norm": 0.7160568833351135,
+      "learning_rate": 4.39873417721519e-06,
+      "loss": 0.0689,
+      "step": 140
+    },
+    {
+      "epoch": 0.044743039888935715,
+      "grad_norm": 0.7635315656661987,
+      "learning_rate": 4.430379746835443e-06,
+      "loss": 0.073,
+      "step": 141
+    },
+    {
+      "epoch": 0.045060366412970725,
+      "grad_norm": 0.7010704874992371,
+      "learning_rate": 4.462025316455697e-06,
+      "loss": 0.0687,
+      "step": 142
+    },
+    {
+      "epoch": 0.04537769293700573,
+      "grad_norm": 0.6945084929466248,
+      "learning_rate": 4.4936708860759495e-06,
+      "loss": 0.0613,
+      "step": 143
+    },
+    {
+      "epoch": 0.04569501946104073,
+      "grad_norm": 0.844071090221405,
+      "learning_rate": 4.525316455696203e-06,
+      "loss": 0.08,
+      "step": 144
+    },
+    {
+      "epoch": 0.04601234598507574,
+      "grad_norm": 0.8242946267127991,
+      "learning_rate": 4.556962025316456e-06,
+      "loss": 0.0751,
+      "step": 145
+    },
+    {
+      "epoch": 0.04632967250911074,
+      "grad_norm": 0.6641847491264343,
+      "learning_rate": 4.588607594936709e-06,
+      "loss": 0.0621,
+      "step": 146
+    },
+    {
+      "epoch": 0.046646999033145745,
+      "grad_norm": 0.6343713402748108,
+      "learning_rate": 4.620253164556963e-06,
+      "loss": 0.0599,
+      "step": 147
+    },
+    {
+      "epoch": 0.046964325557180754,
+      "grad_norm": 0.7122145891189575,
+      "learning_rate": 4.6518987341772155e-06,
+      "loss": 0.0665,
+      "step": 148
+    },
+    {
+      "epoch": 0.04728165208121576,
+      "grad_norm": 0.9016188383102417,
+      "learning_rate": 4.683544303797468e-06,
+      "loss": 0.0751,
+      "step": 149
+    },
+    {
+      "epoch": 0.04759897860525076,
+      "grad_norm": 0.7317721843719482,
+      "learning_rate": 4.715189873417722e-06,
+      "loss": 0.0613,
+      "step": 150
+    },
+    {
+      "epoch": 0.04791630512928577,
+      "grad_norm": 0.5918257832527161,
+      "learning_rate": 4.746835443037975e-06,
+      "loss": 0.0707,
+      "step": 151
+    },
+    {
+      "epoch": 0.04823363165332077,
+      "grad_norm": 0.8951603770256042,
+      "learning_rate": 4.778481012658228e-06,
+      "loss": 0.0616,
+      "step": 152
+    },
+    {
+      "epoch": 0.048550958177355774,
+      "grad_norm": 0.6995075345039368,
+      "learning_rate": 4.8101265822784815e-06,
+      "loss": 0.0694,
+      "step": 153
+    },
+    {
+      "epoch": 0.048868284701390784,
+      "grad_norm": 0.7023273706436157,
+      "learning_rate": 4.841772151898735e-06,
+      "loss": 0.0717,
+      "step": 154
+    },
+    {
+      "epoch": 0.049185611225425786,
+      "grad_norm": 0.67479407787323,
+      "learning_rate": 4.873417721518987e-06,
+      "loss": 0.0681,
+      "step": 155
+    },
+    {
+      "epoch": 0.049502937749460796,
+      "grad_norm": 0.8265228867530823,
+      "learning_rate": 4.905063291139241e-06,
+      "loss": 0.0637,
+      "step": 156
+    },
+    {
+      "epoch": 0.0498202642734958,
+      "grad_norm": 0.6701403856277466,
+      "learning_rate": 4.936708860759495e-06,
+      "loss": 0.0678,
+      "step": 157
+    },
+    {
+      "epoch": 0.0501375907975308,
+      "grad_norm": 0.7492559552192688,
+      "learning_rate": 4.968354430379747e-06,
+      "loss": 0.0656,
+      "step": 158
+    },
+    {
+      "epoch": 0.05045491732156581,
+      "grad_norm": 0.7431222796440125,
+      "learning_rate": 5e-06,
+      "loss": 0.0698,
+      "step": 159
+    },
+    {
+      "epoch": 0.05077224384560081,
+      "grad_norm": 0.7292019724845886,
+      "learning_rate": 5.031645569620254e-06,
+      "loss": 0.0707,
+      "step": 160
+    },
+    {
+      "epoch": 0.051089570369635816,
+      "grad_norm": 1.1787906885147095,
+      "learning_rate": 5.063291139240507e-06,
+      "loss": 0.0723,
+      "step": 161
+    },
+    {
+      "epoch": 0.051406896893670825,
+      "grad_norm": 1.4319536685943604,
+      "learning_rate": 5.094936708860761e-06,
+      "loss": 0.0724,
+      "step": 162
+    },
+    {
+      "epoch": 0.05172422341770583,
+      "grad_norm": 1.1190286874771118,
+      "learning_rate": 5.126582278481013e-06,
+      "loss": 0.0597,
+      "step": 163
+    },
+    {
+      "epoch": 0.05204154994174083,
+      "grad_norm": 1.2695000171661377,
+      "learning_rate": 5.1582278481012656e-06,
+      "loss": 0.0628,
+      "step": 164
+    },
+    {
+      "epoch": 0.05235887646577584,
+      "grad_norm": 0.7486492991447449,
+      "learning_rate": 5.189873417721519e-06,
+      "loss": 0.0658,
+      "step": 165
+    },
+    {
+      "epoch": 0.05267620298981084,
+      "grad_norm": 0.7061361074447632,
+      "learning_rate": 5.221518987341773e-06,
+      "loss": 0.0642,
+      "step": 166
+    },
+    {
+      "epoch": 0.052993529513845845,
+      "grad_norm": 0.6353309750556946,
+      "learning_rate": 5.253164556962026e-06,
+      "loss": 0.0646,
+      "step": 167
+    },
+    {
+      "epoch": 0.053310856037880855,
+      "grad_norm": 0.7401093244552612,
+      "learning_rate": 5.2848101265822796e-06,
+      "loss": 0.0618,
+      "step": 168
+    },
+    {
+      "epoch": 0.05362818256191586,
+      "grad_norm": 0.781416654586792,
+      "learning_rate": 5.3164556962025316e-06,
+      "loss": 0.0702,
+      "step": 169
+    },
+    {
+      "epoch": 0.05394550908595087,
+      "grad_norm": 0.6498562097549438,
+      "learning_rate": 5.348101265822785e-06,
+      "loss": 0.0645,
+      "step": 170
+    },
+    {
+      "epoch": 0.05426283560998587,
+      "grad_norm": 0.7381700873374939,
+      "learning_rate": 5.379746835443038e-06,
+      "loss": 0.0627,
+      "step": 171
+    },
+    {
+      "epoch": 0.05458016213402087,
+      "grad_norm": 0.7094889283180237,
+      "learning_rate": 5.411392405063292e-06,
+      "loss": 0.0649,
+      "step": 172
+    },
+    {
+      "epoch": 0.05489748865805588,
+      "grad_norm": 0.6543947458267212,
+      "learning_rate": 5.443037974683545e-06,
+      "loss": 0.0634,
+      "step": 173
+    },
+    {
+      "epoch": 0.055214815182090884,
+      "grad_norm": 1.0508382320404053,
+      "learning_rate": 5.474683544303798e-06,
+      "loss": 0.0727,
+      "step": 174
+    },
+    {
+      "epoch": 0.05553214170612589,
+      "grad_norm": 0.8786133527755737,
+      "learning_rate": 5.506329113924051e-06,
+      "loss": 0.0698,
+      "step": 175
+    },
+    {
+      "epoch": 0.0558494682301609,
+      "grad_norm": 0.7369667887687683,
+      "learning_rate": 5.537974683544304e-06,
+      "loss": 0.0666,
+      "step": 176
+    },
+    {
+      "epoch": 0.0561667947541959,
+      "grad_norm": 0.843385636806488,
+      "learning_rate": 5.569620253164557e-06,
+      "loss": 0.0567,
+      "step": 177
+    },
+    {
+      "epoch": 0.0564841212782309,
+      "grad_norm": 0.9363443851470947,
+      "learning_rate": 5.601265822784811e-06,
+      "loss": 0.0599,
+      "step": 178
+    },
+    {
+      "epoch": 0.05680144780226591,
+      "grad_norm": 0.6885288953781128,
+      "learning_rate": 5.6329113924050636e-06,
+      "loss": 0.0624,
+      "step": 179
+    },
+    {
+      "epoch": 0.057118774326300914,
+      "grad_norm": 0.6786429286003113,
+      "learning_rate": 5.664556962025317e-06,
+      "loss": 0.0571,
+      "step": 180
+    },
+    {
+      "epoch": 0.05743610085033592,
+      "grad_norm": 1.0804567337036133,
+      "learning_rate": 5.69620253164557e-06,
+      "loss": 0.0586,
+      "step": 181
+    },
+    {
+      "epoch": 0.057753427374370926,
+      "grad_norm": 0.7370650172233582,
+      "learning_rate": 5.727848101265824e-06,
+      "loss": 0.0645,
+      "step": 182
+    },
+    {
+      "epoch": 0.05807075389840593,
+      "grad_norm": 0.7486072778701782,
+      "learning_rate": 5.759493670886076e-06,
+      "loss": 0.0639,
+      "step": 183
+    },
+    {
+      "epoch": 0.05838808042244094,
+      "grad_norm": 0.7497734427452087,
+      "learning_rate": 5.7911392405063296e-06,
+      "loss": 0.0576,
+      "step": 184
+    },
+    {
+      "epoch": 0.05870540694647594,
+      "grad_norm": 1.2832496166229248,
+      "learning_rate": 5.8227848101265824e-06,
+      "loss": 0.0659,
+      "step": 185
+    },
+    {
+      "epoch": 0.059022733470510944,
+      "grad_norm": 0.7222813367843628,
+      "learning_rate": 5.854430379746836e-06,
+      "loss": 0.0635,
+      "step": 186
+    },
+    {
+      "epoch": 0.05934005999454595,
+      "grad_norm": 0.8825712203979492,
+      "learning_rate": 5.886075949367089e-06,
+      "loss": 0.0658,
+      "step": 187
+    },
+    {
+      "epoch": 0.059657386518580956,
+      "grad_norm": 0.8911665678024292,
+      "learning_rate": 5.917721518987343e-06,
+      "loss": 0.0656,
+      "step": 188
+    },
+    {
+      "epoch": 0.05997471304261596,
+      "grad_norm": 0.7345048189163208,
+      "learning_rate": 5.949367088607595e-06,
+      "loss": 0.0596,
+      "step": 189
+    },
+    {
+      "epoch": 0.06029203956665097,
+      "grad_norm": 0.68448805809021,
+      "learning_rate": 5.9810126582278484e-06,
+      "loss": 0.0594,
+      "step": 190
+    },
+    {
+      "epoch": 0.06060936609068597,
+      "grad_norm": 0.6588714718818665,
+      "learning_rate": 6.012658227848101e-06,
+      "loss": 0.0612,
+      "step": 191
+    },
+    {
+      "epoch": 0.06092669261472097,
+      "grad_norm": 0.7801559567451477,
+      "learning_rate": 6.044303797468355e-06,
+      "loss": 0.0635,
+      "step": 192
+    },
+    {
+      "epoch": 0.06124401913875598,
+      "grad_norm": 0.678334653377533,
+      "learning_rate": 6.075949367088608e-06,
+      "loss": 0.066,
+      "step": 193
+    },
+    {
+      "epoch": 0.061561345662790985,
+      "grad_norm": 0.9699953198432922,
+      "learning_rate": 6.1075949367088616e-06,
+      "loss": 0.0566,
+      "step": 194
+    },
+    {
+      "epoch": 0.06187867218682599,
+      "grad_norm": 0.6815274357795715,
+      "learning_rate": 6.139240506329115e-06,
+      "loss": 0.0553,
+      "step": 195
+    },
+    {
+      "epoch": 0.062195998710861,
+      "grad_norm": 0.8181425929069519,
+      "learning_rate": 6.170886075949367e-06,
+      "loss": 0.0664,
+      "step": 196
+    },
+    {
+      "epoch": 0.062513325234896,
+      "grad_norm": 0.7274137139320374,
+      "learning_rate": 6.20253164556962e-06,
+      "loss": 0.0624,
+      "step": 197
+    },
+    {
+      "epoch": 0.062830651758931,
+      "grad_norm": 0.8435655236244202,
+      "learning_rate": 6.234177215189874e-06,
+      "loss": 0.0552,
+      "step": 198
+    },
+    {
+      "epoch": 0.063147978282966,
+      "grad_norm": 0.5994033217430115,
+      "learning_rate": 6.265822784810128e-06,
+      "loss": 0.0545,
+      "step": 199
+    },
+    {
+      "epoch": 0.06346530480700102,
+      "grad_norm": 0.7202048897743225,
+      "learning_rate": 6.2974683544303804e-06,
+      "loss": 0.0595,
+      "step": 200
+    },
+    {
+      "epoch": 0.06378263133103602,
+      "grad_norm": 0.6187375783920288,
+      "learning_rate": 6.329113924050634e-06,
+      "loss": 0.0503,
+      "step": 201
+    },
+    {
+      "epoch": 0.06409995785507103,
+      "grad_norm": 0.8126381039619446,
+      "learning_rate": 6.360759493670886e-06,
+      "loss": 0.0618,
+      "step": 202
+    },
+    {
+      "epoch": 0.06441728437910603,
+      "grad_norm": 0.803071141242981,
+      "learning_rate": 6.392405063291139e-06,
+      "loss": 0.0613,
+      "step": 203
+    },
+    {
+      "epoch": 0.06473461090314103,
+      "grad_norm": 0.7738210558891296,
+      "learning_rate": 6.424050632911393e-06,
+      "loss": 0.0521,
+      "step": 204
+    },
+    {
+      "epoch": 0.06505193742717605,
+      "grad_norm": 0.7081449627876282,
+      "learning_rate": 6.4556962025316464e-06,
+      "loss": 0.0551,
+      "step": 205
+    },
+    {
+      "epoch": 0.06536926395121105,
+      "grad_norm": 0.8450528979301453,
+      "learning_rate": 6.487341772151899e-06,
+      "loss": 0.0566,
+      "step": 206
+    },
+    {
+      "epoch": 0.06568659047524605,
+      "grad_norm": 0.7632423043251038,
+      "learning_rate": 6.518987341772153e-06,
+      "loss": 0.059,
+      "step": 207
+    },
+    {
+      "epoch": 0.06600391699928106,
+      "grad_norm": 0.7753247618675232,
+      "learning_rate": 6.550632911392406e-06,
+      "loss": 0.0611,
+      "step": 208
+    },
+    {
+      "epoch": 0.06632124352331606,
+      "grad_norm": 0.6930252313613892,
+      "learning_rate": 6.582278481012659e-06,
+      "loss": 0.0643,
+      "step": 209
+    },
+    {
+      "epoch": 0.06663857004735106,
+      "grad_norm": 0.7667645215988159,
+      "learning_rate": 6.613924050632912e-06,
+      "loss": 0.063,
+      "step": 210
+    },
+    {
+      "epoch": 0.06695589657138608,
+      "grad_norm": 0.7646639347076416,
+      "learning_rate": 6.645569620253165e-06,
+      "loss": 0.0609,
+      "step": 211
+    },
+    {
+      "epoch": 0.06727322309542108,
+      "grad_norm": 0.6697304248809814,
+      "learning_rate": 6.677215189873418e-06,
+      "loss": 0.0562,
+      "step": 212
+    },
+    {
+      "epoch": 0.06759054961945608,
+      "grad_norm": 0.6435572504997253,
+      "learning_rate": 6.708860759493672e-06,
+      "loss": 0.0508,
+      "step": 213
+    },
+    {
+      "epoch": 0.06790787614349109,
+      "grad_norm": 0.7295902967453003,
+      "learning_rate": 6.740506329113925e-06,
+      "loss": 0.0575,
+      "step": 214
+    },
+    {
+      "epoch": 0.06822520266752609,
+      "grad_norm": 0.684455156326294,
+      "learning_rate": 6.772151898734178e-06,
+      "loss": 0.0634,
+      "step": 215
+    },
+    {
+      "epoch": 0.06854252919156109,
+      "grad_norm": 0.6214410662651062,
+      "learning_rate": 6.8037974683544305e-06,
+      "loss": 0.0557,
+      "step": 216
+    },
+    {
+      "epoch": 0.06885985571559611,
+      "grad_norm": 0.5960308313369751,
+      "learning_rate": 6.835443037974684e-06,
+      "loss": 0.0555,
+      "step": 217
+    },
+    {
+      "epoch": 0.06917718223963111,
+      "grad_norm": 0.6423066258430481,
+      "learning_rate": 6.867088607594937e-06,
+      "loss": 0.0538,
+      "step": 218
+    },
+    {
+      "epoch": 0.06949450876366611,
+      "grad_norm": 0.6987835168838501,
+      "learning_rate": 6.898734177215191e-06,
+      "loss": 0.055,
+      "step": 219
+    },
+    {
+      "epoch": 0.06981183528770112,
+      "grad_norm": 0.7723947763442993,
+      "learning_rate": 6.930379746835444e-06,
+      "loss": 0.0613,
+      "step": 220
+    },
+    {
+      "epoch": 0.07012916181173612,
+      "grad_norm": 0.5927456021308899,
+      "learning_rate": 6.962025316455697e-06,
+      "loss": 0.0507,
+      "step": 221
+    },
+    {
+      "epoch": 0.07044648833577113,
+      "grad_norm": 0.7041898965835571,
+      "learning_rate": 6.993670886075949e-06,
+      "loss": 0.056,
+      "step": 222
+    },
+    {
+      "epoch": 0.07076381485980614,
+      "grad_norm": 0.9441041350364685,
+      "learning_rate": 7.025316455696203e-06,
+      "loss": 0.0617,
+      "step": 223
+    },
+    {
+      "epoch": 0.07108114138384114,
+      "grad_norm": 0.7225652933120728,
+      "learning_rate": 7.056962025316456e-06,
+      "loss": 0.0573,
+      "step": 224
+    },
+    {
+      "epoch": 0.07139846790787614,
+      "grad_norm": 0.7080259919166565,
+      "learning_rate": 7.08860759493671e-06,
+      "loss": 0.055,
+      "step": 225
+    },
+    {
+      "epoch": 0.07171579443191115,
+      "grad_norm": 0.8079982399940491,
+      "learning_rate": 7.1202531645569625e-06,
+      "loss": 0.0641,
+      "step": 226
+    },
+    {
+      "epoch": 0.07203312095594615,
+      "grad_norm": 0.7268806099891663,
+      "learning_rate": 7.151898734177216e-06,
+      "loss": 0.0571,
+      "step": 227
+    },
+    {
+      "epoch": 0.07235044747998116,
+      "grad_norm": 1.2894824743270874,
+      "learning_rate": 7.18354430379747e-06,
+      "loss": 0.0574,
+      "step": 228
+    },
+    {
+      "epoch": 0.07266777400401617,
+      "grad_norm": 0.8462490439414978,
+      "learning_rate": 7.215189873417722e-06,
+      "loss": 0.054,
+      "step": 229
+    },
+    {
+      "epoch": 0.07298510052805117,
+      "grad_norm": 2.3804852962493896,
+      "learning_rate": 7.246835443037975e-06,
+      "loss": 0.0567,
+      "step": 230
+    },
+    {
+      "epoch": 0.07330242705208617,
+      "grad_norm": 0.7559064030647278,
+      "learning_rate": 7.2784810126582285e-06,
+      "loss": 0.0583,
+      "step": 231
+    },
+    {
+      "epoch": 0.07361975357612117,
+      "grad_norm": 0.5622048377990723,
+      "learning_rate": 7.310126582278481e-06,
+      "loss": 0.0472,
+      "step": 232
+    },
+    {
+      "epoch": 0.07393708010015618,
+      "grad_norm": 0.7366349101066589,
+      "learning_rate": 7.341772151898735e-06,
+      "loss": 0.0527,
+      "step": 233
+    },
+    {
+      "epoch": 0.0742544066241912,
+      "grad_norm": 1.389522671699524,
+      "learning_rate": 7.373417721518989e-06,
+      "loss": 0.0575,
+      "step": 234
+    },
+    {
+      "epoch": 0.0745717331482262,
+      "grad_norm": 1.0432237386703491,
+      "learning_rate": 7.405063291139241e-06,
+      "loss": 0.0565,
+      "step": 235
+    },
+    {
+      "epoch": 0.0748890596722612,
+      "grad_norm": 0.7934345602989197,
+      "learning_rate": 7.436708860759494e-06,
+      "loss": 0.0514,
+      "step": 236
+    },
+    {
+      "epoch": 0.0752063861962962,
+      "grad_norm": 0.9964441657066345,
+      "learning_rate": 7.468354430379747e-06,
+      "loss": 0.0592,
+      "step": 237
+    },
+    {
+      "epoch": 0.0755237127203312,
+      "grad_norm": 0.6250876784324646,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.052,
+      "step": 238
+    },
+    {
+      "epoch": 0.07584103924436622,
+      "grad_norm": 1.0673632621765137,
+      "learning_rate": 7.531645569620254e-06,
+      "loss": 0.0547,
+      "step": 239
+    },
+    {
+      "epoch": 0.07615836576840122,
+      "grad_norm": 0.8667101860046387,
+      "learning_rate": 7.563291139240508e-06,
+      "loss": 0.06,
+      "step": 240
+    },
+    {
+      "epoch": 0.07647569229243623,
+      "grad_norm": 0.6332155466079712,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 0.0551,
+      "step": 241
+    },
+    {
+      "epoch": 0.07679301881647123,
+      "grad_norm": 0.6795554161071777,
+      "learning_rate": 7.6265822784810125e-06,
+      "loss": 0.0531,
+      "step": 242
+    },
+    {
+      "epoch": 0.07711034534050623,
+      "grad_norm": 0.6087434887886047,
+      "learning_rate": 7.658227848101265e-06,
+      "loss": 0.0494,
+      "step": 243
+    },
+    {
+      "epoch": 0.07742767186454123,
+      "grad_norm": 0.6335604786872864,
+      "learning_rate": 7.68987341772152e-06,
+      "loss": 0.0502,
+      "step": 244
+    },
+    {
+      "epoch": 0.07774499838857625,
+      "grad_norm": 0.6387730836868286,
+      "learning_rate": 7.721518987341773e-06,
+      "loss": 0.0561,
+      "step": 245
+    },
+    {
+      "epoch": 0.07806232491261125,
+      "grad_norm": 0.6156599521636963,
+      "learning_rate": 7.753164556962026e-06,
+      "loss": 0.0527,
+      "step": 246
+    },
+    {
+      "epoch": 0.07837965143664626,
+      "grad_norm": 0.7905254364013672,
+      "learning_rate": 7.78481012658228e-06,
+      "loss": 0.0558,
+      "step": 247
+    },
+    {
+      "epoch": 0.07869697796068126,
+      "grad_norm": 0.7257440686225891,
+      "learning_rate": 7.816455696202531e-06,
+      "loss": 0.0534,
+      "step": 248
+    },
+    {
+      "epoch": 0.07901430448471626,
+      "grad_norm": 0.7667980790138245,
+      "learning_rate": 7.848101265822786e-06,
+      "loss": 0.0589,
+      "step": 249
+    },
+    {
+      "epoch": 0.07933163100875128,
+      "grad_norm": 0.6432068943977356,
+      "learning_rate": 7.879746835443039e-06,
+      "loss": 0.0509,
+      "step": 250
+    },
+    {
+      "epoch": 0.07964895753278628,
+      "grad_norm": 0.6008541584014893,
+      "learning_rate": 7.911392405063292e-06,
+      "loss": 0.0498,
+      "step": 251
+    },
+    {
+      "epoch": 0.07996628405682128,
+      "grad_norm": 0.9412493109703064,
+      "learning_rate": 7.943037974683544e-06,
+      "loss": 0.0536,
+      "step": 252
+    },
+    {
+      "epoch": 0.08028361058085628,
+      "grad_norm": 0.6630102396011353,
+      "learning_rate": 7.974683544303799e-06,
+      "loss": 0.0569,
+      "step": 253
+    },
+    {
+      "epoch": 0.08060093710489129,
+      "grad_norm": 0.7627852559089661,
+      "learning_rate": 8.006329113924052e-06,
+      "loss": 0.0472,
+      "step": 254
+    },
+    {
+      "epoch": 0.08091826362892629,
+      "grad_norm": 0.6052554249763489,
+      "learning_rate": 8.037974683544305e-06,
+      "loss": 0.0491,
+      "step": 255
+    },
+    {
+      "epoch": 0.0812355901529613,
+      "grad_norm": 0.6256715059280396,
+      "learning_rate": 8.069620253164558e-06,
+      "loss": 0.0536,
+      "step": 256
+    },
+    {
+      "epoch": 0.08155291667699631,
+      "grad_norm": 0.9521887898445129,
+      "learning_rate": 8.10126582278481e-06,
+      "loss": 0.0491,
+      "step": 257
+    },
+    {
+      "epoch": 0.08187024320103131,
+      "grad_norm": 0.6414262652397156,
+      "learning_rate": 8.132911392405063e-06,
+      "loss": 0.0557,
+      "step": 258
+    },
+    {
+      "epoch": 0.08218756972506631,
+      "grad_norm": 0.5295684337615967,
+      "learning_rate": 8.164556962025318e-06,
+      "loss": 0.0422,
+      "step": 259
+    },
+    {
+      "epoch": 0.08250489624910132,
+      "grad_norm": 0.7351981997489929,
+      "learning_rate": 8.19620253164557e-06,
+      "loss": 0.0562,
+      "step": 260
+    },
+    {
+      "epoch": 0.08282222277313632,
+      "grad_norm": 0.6484821438789368,
+      "learning_rate": 8.227848101265824e-06,
+      "loss": 0.0515,
+      "step": 261
+    },
+    {
+      "epoch": 0.08313954929717134,
+      "grad_norm": 0.6362297534942627,
+      "learning_rate": 8.259493670886076e-06,
+      "loss": 0.0493,
+      "step": 262
+    },
+    {
+      "epoch": 0.08345687582120634,
+      "grad_norm": 0.6204566359519958,
+      "learning_rate": 8.29113924050633e-06,
+      "loss": 0.0505,
+      "step": 263
+    },
+    {
+      "epoch": 0.08377420234524134,
+      "grad_norm": 0.6443067789077759,
+      "learning_rate": 8.322784810126582e-06,
+      "loss": 0.0552,
+      "step": 264
+    },
+    {
+      "epoch": 0.08409152886927634,
+      "grad_norm": 0.6477574110031128,
+      "learning_rate": 8.354430379746837e-06,
+      "loss": 0.055,
+      "step": 265
+    },
+    {
+      "epoch": 0.08440885539331135,
+      "grad_norm": 0.9887301325798035,
+      "learning_rate": 8.38607594936709e-06,
+      "loss": 0.0558,
+      "step": 266
+    },
+    {
+      "epoch": 0.08472618191734636,
+      "grad_norm": 0.5635358095169067,
+      "learning_rate": 8.417721518987342e-06,
+      "loss": 0.0534,
+      "step": 267
+    },
+    {
+      "epoch": 0.08504350844138137,
+      "grad_norm": 0.7857146859169006,
+      "learning_rate": 8.449367088607595e-06,
+      "loss": 0.0554,
+      "step": 268
+    },
+    {
+      "epoch": 0.08536083496541637,
+      "grad_norm": 0.5973455309867859,
+      "learning_rate": 8.481012658227848e-06,
+      "loss": 0.0472,
+      "step": 269
+    },
+    {
+      "epoch": 0.08567816148945137,
+      "grad_norm": 0.6638344526290894,
+      "learning_rate": 8.512658227848101e-06,
+      "loss": 0.0481,
+      "step": 270
+    },
+    {
+      "epoch": 0.08599548801348637,
+      "grad_norm": 0.723565936088562,
+      "learning_rate": 8.544303797468356e-06,
+      "loss": 0.0514,
+      "step": 271
+    },
+    {
+      "epoch": 0.08631281453752138,
+      "grad_norm": 0.5519058108329773,
+      "learning_rate": 8.575949367088608e-06,
+      "loss": 0.0471,
+      "step": 272
+    },
+    {
+      "epoch": 0.08663014106155639,
+      "grad_norm": 0.6533753275871277,
+      "learning_rate": 8.607594936708861e-06,
+      "loss": 0.0544,
+      "step": 273
+    },
+    {
+      "epoch": 0.0869474675855914,
+      "grad_norm": 1.0921151638031006,
+      "learning_rate": 8.639240506329116e-06,
+      "loss": 0.0516,
+      "step": 274
+    },
+    {
+      "epoch": 0.0872647941096264,
+      "grad_norm": 0.5981336236000061,
+      "learning_rate": 8.670886075949367e-06,
+      "loss": 0.051,
+      "step": 275
+    },
+    {
+      "epoch": 0.0875821206336614,
+      "grad_norm": 0.9020387530326843,
+      "learning_rate": 8.70253164556962e-06,
+      "loss": 0.048,
+      "step": 276
+    },
+    {
+      "epoch": 0.0878994471576964,
+      "grad_norm": 0.7535433769226074,
+      "learning_rate": 8.734177215189874e-06,
+      "loss": 0.0492,
+      "step": 277
+    },
+    {
+      "epoch": 0.08821677368173142,
+      "grad_norm": 0.5968385934829712,
+      "learning_rate": 8.765822784810127e-06,
+      "loss": 0.0488,
+      "step": 278
+    },
+    {
+      "epoch": 0.08853410020576642,
+      "grad_norm": 0.7003240585327148,
+      "learning_rate": 8.79746835443038e-06,
+      "loss": 0.0529,
+      "step": 279
+    },
+    {
+      "epoch": 0.08885142672980142,
+      "grad_norm": 0.7311054468154907,
+      "learning_rate": 8.829113924050635e-06,
+      "loss": 0.0587,
+      "step": 280
+    },
+    {
+      "epoch": 0.08916875325383643,
+      "grad_norm": 0.7550223469734192,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 0.0468,
+      "step": 281
+    },
+    {
+      "epoch": 0.08948607977787143,
+      "grad_norm": 0.6127842664718628,
+      "learning_rate": 8.892405063291139e-06,
+      "loss": 0.0539,
+      "step": 282
+    },
+    {
+      "epoch": 0.08980340630190643,
+      "grad_norm": 0.8750295042991638,
+      "learning_rate": 8.924050632911393e-06,
+      "loss": 0.0535,
+      "step": 283
+    },
+    {
+      "epoch": 0.09012073282594145,
+      "grad_norm": 0.5803536176681519,
+      "learning_rate": 8.955696202531646e-06,
+      "loss": 0.055,
+      "step": 284
+    },
+    {
+      "epoch": 0.09043805934997645,
+      "grad_norm": 0.7908932566642761,
+      "learning_rate": 8.987341772151899e-06,
+      "loss": 0.051,
+      "step": 285
+    },
+    {
+      "epoch": 0.09075538587401145,
+      "grad_norm": 0.549761950969696,
+      "learning_rate": 9.018987341772154e-06,
+      "loss": 0.0544,
+      "step": 286
+    },
+    {
+      "epoch": 0.09107271239804646,
+      "grad_norm": 0.6864755153656006,
+      "learning_rate": 9.050632911392407e-06,
+      "loss": 0.0491,
+      "step": 287
+    },
+    {
+      "epoch": 0.09139003892208146,
+      "grad_norm": 0.8759939670562744,
+      "learning_rate": 9.08227848101266e-06,
+      "loss": 0.0452,
+      "step": 288
+    },
+    {
+      "epoch": 0.09170736544611646,
+      "grad_norm": 0.48853012919425964,
+      "learning_rate": 9.113924050632912e-06,
+      "loss": 0.0463,
+      "step": 289
+    },
+    {
+      "epoch": 0.09202469197015148,
+      "grad_norm": 0.6499919295310974,
+      "learning_rate": 9.145569620253165e-06,
+      "loss": 0.0472,
+      "step": 290
+    },
+    {
+      "epoch": 0.09234201849418648,
+      "grad_norm": 0.7799652218818665,
+      "learning_rate": 9.177215189873418e-06,
+      "loss": 0.0514,
+      "step": 291
+    },
+    {
+      "epoch": 0.09265934501822148,
+      "grad_norm": 0.6204949021339417,
+      "learning_rate": 9.208860759493673e-06,
+      "loss": 0.0525,
+      "step": 292
+    },
+    {
+      "epoch": 0.09297667154225649,
+      "grad_norm": 0.7858293056488037,
+      "learning_rate": 9.240506329113925e-06,
+      "loss": 0.0437,
+      "step": 293
+    },
+    {
+      "epoch": 0.09329399806629149,
+      "grad_norm": 0.7692995071411133,
+      "learning_rate": 9.272151898734178e-06,
+      "loss": 0.0502,
+      "step": 294
+    },
+    {
+      "epoch": 0.0936113245903265,
+      "grad_norm": 0.6089361310005188,
+      "learning_rate": 9.303797468354431e-06,
+      "loss": 0.0501,
+      "step": 295
+    },
+    {
+      "epoch": 0.09392865111436151,
+      "grad_norm": 0.5847312808036804,
+      "learning_rate": 9.335443037974684e-06,
+      "loss": 0.0471,
+      "step": 296
+    },
+    {
+      "epoch": 0.09424597763839651,
+      "grad_norm": 5.030068397521973,
+      "learning_rate": 9.367088607594937e-06,
+      "loss": 0.0557,
+      "step": 297
+    },
+    {
+      "epoch": 0.09456330416243151,
+      "grad_norm": 0.6561621427536011,
+      "learning_rate": 9.398734177215191e-06,
+      "loss": 0.05,
+      "step": 298
+    },
+    {
+      "epoch": 0.09488063068646652,
+      "grad_norm": 0.6542370319366455,
+      "learning_rate": 9.430379746835444e-06,
+      "loss": 0.0536,
+      "step": 299
+    },
+    {
+      "epoch": 0.09519795721050152,
+      "grad_norm": 0.9324794411659241,
+      "learning_rate": 9.462025316455697e-06,
+      "loss": 0.0463,
+      "step": 300
+    },
+    {
+      "epoch": 0.09551528373453654,
+      "grad_norm": 0.573015570640564,
+      "learning_rate": 9.49367088607595e-06,
+      "loss": 0.0478,
+      "step": 301
+    },
+    {
+      "epoch": 0.09583261025857154,
+      "grad_norm": 0.6793656945228577,
+      "learning_rate": 9.525316455696203e-06,
+      "loss": 0.0452,
+      "step": 302
+    },
+    {
+      "epoch": 0.09614993678260654,
+      "grad_norm": 0.6263553500175476,
+      "learning_rate": 9.556962025316456e-06,
+      "loss": 0.0442,
+      "step": 303
+    },
+    {
+      "epoch": 0.09646726330664154,
+      "grad_norm": 0.519311785697937,
+      "learning_rate": 9.58860759493671e-06,
+      "loss": 0.045,
+      "step": 304
+    },
+    {
+      "epoch": 0.09678458983067655,
+      "grad_norm": 0.6338351368904114,
+      "learning_rate": 9.620253164556963e-06,
+      "loss": 0.045,
+      "step": 305
+    },
+    {
+      "epoch": 0.09710191635471155,
+      "grad_norm": 0.6338417530059814,
+      "learning_rate": 9.651898734177216e-06,
+      "loss": 0.0474,
+      "step": 306
+    },
+    {
+      "epoch": 0.09741924287874656,
+      "grad_norm": 0.6606894135475159,
+      "learning_rate": 9.68354430379747e-06,
+      "loss": 0.052,
+      "step": 307
+    },
+    {
+      "epoch": 0.09773656940278157,
+      "grad_norm": 0.6472856402397156,
+      "learning_rate": 9.715189873417722e-06,
+      "loss": 0.0486,
+      "step": 308
+    },
+    {
+      "epoch": 0.09805389592681657,
+      "grad_norm": 0.7079848647117615,
+      "learning_rate": 9.746835443037975e-06,
+      "loss": 0.0504,
+      "step": 309
+    },
+    {
+      "epoch": 0.09837122245085157,
+      "grad_norm": 0.6880467534065247,
+      "learning_rate": 9.778481012658229e-06,
+      "loss": 0.0537,
+      "step": 310
+    },
+    {
+      "epoch": 0.09868854897488658,
+      "grad_norm": 1.5069619417190552,
+      "learning_rate": 9.810126582278482e-06,
+      "loss": 0.0584,
+      "step": 311
+    },
+    {
+      "epoch": 0.09900587549892159,
+      "grad_norm": 0.5906437635421753,
+      "learning_rate": 9.841772151898735e-06,
+      "loss": 0.0479,
+      "step": 312
+    },
+    {
+      "epoch": 0.0993232020229566,
+      "grad_norm": 0.9024403095245361,
+      "learning_rate": 9.87341772151899e-06,
+      "loss": 0.0477,
+      "step": 313
+    },
+    {
+      "epoch": 0.0996405285469916,
+      "grad_norm": 0.6381876468658447,
+      "learning_rate": 9.90506329113924e-06,
+      "loss": 0.0554,
+      "step": 314
+    },
+    {
+      "epoch": 0.0999578550710266,
+      "grad_norm": 0.9446009993553162,
+      "learning_rate": 9.936708860759493e-06,
+      "loss": 0.0474,
+      "step": 315
+    },
+    {
+      "epoch": 0.1002751815950616,
+      "grad_norm": 0.6961436867713928,
+      "learning_rate": 9.968354430379748e-06,
+      "loss": 0.0539,
+      "step": 316
+    },
+    {
+      "epoch": 0.1005925081190966,
+      "grad_norm": 0.4731083810329437,
+      "learning_rate": 1e-05,
+      "loss": 0.0451,
+      "step": 317
+    },
+    {
+      "epoch": 0.10090983464313162,
+      "grad_norm": 0.6064469218254089,
+      "learning_rate": 9.999996930033877e-06,
+      "loss": 0.0484,
+      "step": 318
+    },
+    {
+      "epoch": 0.10122716116716662,
+      "grad_norm": 0.5192763805389404,
+      "learning_rate": 9.999987720139277e-06,
+      "loss": 0.0448,
+      "step": 319
+    },
+    {
+      "epoch": 0.10154448769120163,
+      "grad_norm": 0.615347146987915,
+      "learning_rate": 9.999972370327508e-06,
+      "loss": 0.0474,
+      "step": 320
+    },
+    {
+      "epoch": 0.10186181421523663,
+      "grad_norm": 0.5881875157356262,
+      "learning_rate": 9.99995088061742e-06,
+      "loss": 0.0463,
+      "step": 321
+    },
+    {
+      "epoch": 0.10217914073927163,
+      "grad_norm": 0.6094655394554138,
+      "learning_rate": 9.999923251035404e-06,
+      "loss": 0.0486,
+      "step": 322
+    },
+    {
+      "epoch": 0.10249646726330665,
+      "grad_norm": 0.8739184141159058,
+      "learning_rate": 9.999889481615387e-06,
+      "loss": 0.0411,
+      "step": 323
+    },
+    {
+      "epoch": 0.10281379378734165,
+      "grad_norm": 0.5956976413726807,
+      "learning_rate": 9.99984957239884e-06,
+      "loss": 0.0543,
+      "step": 324
+    },
+    {
+      "epoch": 0.10313112031137665,
+      "grad_norm": 0.5529938340187073,
+      "learning_rate": 9.999803523434766e-06,
+      "loss": 0.0467,
+      "step": 325
+    },
+    {
+      "epoch": 0.10344844683541166,
+      "grad_norm": 0.5876514315605164,
+      "learning_rate": 9.999751334779716e-06,
+      "loss": 0.049,
+      "step": 326
+    },
+    {
+      "epoch": 0.10376577335944666,
+      "grad_norm": 0.6124856472015381,
+      "learning_rate": 9.999693006497777e-06,
+      "loss": 0.0536,
+      "step": 327
+    },
+    {
+      "epoch": 0.10408309988348166,
+      "grad_norm": 0.5201764702796936,
+      "learning_rate": 9.999628538660574e-06,
+      "loss": 0.0468,
+      "step": 328
+    },
+    {
+      "epoch": 0.10440042640751668,
+      "grad_norm": 0.5368020534515381,
+      "learning_rate": 9.999557931347273e-06,
+      "loss": 0.0504,
+      "step": 329
+    },
+    {
+      "epoch": 0.10471775293155168,
+      "grad_norm": 0.5911468863487244,
+      "learning_rate": 9.99948118464458e-06,
+      "loss": 0.0514,
+      "step": 330
+    },
+    {
+      "epoch": 0.10503507945558668,
+      "grad_norm": 0.5092113018035889,
+      "learning_rate": 9.999398298646738e-06,
+      "loss": 0.0454,
+      "step": 331
+    },
+    {
+      "epoch": 0.10535240597962169,
+      "grad_norm": 0.5567737817764282,
+      "learning_rate": 9.99930927345553e-06,
+      "loss": 0.048,
+      "step": 332
+    },
+    {
+      "epoch": 0.10566973250365669,
+      "grad_norm": 0.5333291888237,
+      "learning_rate": 9.999214109180278e-06,
+      "loss": 0.0485,
+      "step": 333
+    },
+    {
+      "epoch": 0.10598705902769169,
+      "grad_norm": 2.8673856258392334,
+      "learning_rate": 9.999112805937841e-06,
+      "loss": 0.0454,
+      "step": 334
+    },
+    {
+      "epoch": 0.10630438555172671,
+      "grad_norm": 0.616701066493988,
+      "learning_rate": 9.999005363852619e-06,
+      "loss": 0.0449,
+      "step": 335
+    },
+    {
+      "epoch": 0.10662171207576171,
+      "grad_norm": 0.5830521583557129,
+      "learning_rate": 9.99889178305655e-06,
+      "loss": 0.0479,
+      "step": 336
+    },
+    {
+      "epoch": 0.10693903859979671,
+      "grad_norm": 0.5477436780929565,
+      "learning_rate": 9.998772063689109e-06,
+      "loss": 0.0452,
+      "step": 337
+    },
+    {
+      "epoch": 0.10725636512383172,
+      "grad_norm": 0.6000690460205078,
+      "learning_rate": 9.99864620589731e-06,
+      "loss": 0.0458,
+      "step": 338
+    },
+    {
+      "epoch": 0.10757369164786672,
+      "grad_norm": 0.7424272298812866,
+      "learning_rate": 9.998514209835703e-06,
+      "loss": 0.0448,
+      "step": 339
+    },
+    {
+      "epoch": 0.10789101817190173,
+      "grad_norm": 0.5593220591545105,
+      "learning_rate": 9.998376075666382e-06,
+      "loss": 0.0524,
+      "step": 340
+    },
+    {
+      "epoch": 0.10820834469593674,
+      "grad_norm": 0.5366954803466797,
+      "learning_rate": 9.99823180355897e-06,
+      "loss": 0.046,
+      "step": 341
+    },
+    {
+      "epoch": 0.10852567121997174,
+      "grad_norm": 0.6852601170539856,
+      "learning_rate": 9.99808139369063e-06,
+      "loss": 0.0458,
+      "step": 342
+    },
+    {
+      "epoch": 0.10884299774400674,
+      "grad_norm": 0.6158245205879211,
+      "learning_rate": 9.997924846246067e-06,
+      "loss": 0.0475,
+      "step": 343
+    },
+    {
+      "epoch": 0.10916032426804174,
+      "grad_norm": 0.5623180270195007,
+      "learning_rate": 9.997762161417517e-06,
+      "loss": 0.0473,
+      "step": 344
+    },
+    {
+      "epoch": 0.10947765079207675,
+      "grad_norm": 2.489149808883667,
+      "learning_rate": 9.997593339404757e-06,
+      "loss": 0.0427,
+      "step": 345
+    },
+    {
+      "epoch": 0.10979497731611176,
+      "grad_norm": 1.1507155895233154,
+      "learning_rate": 9.997418380415093e-06,
+      "loss": 0.0447,
+      "step": 346
+    },
+    {
+      "epoch": 0.11011230384014677,
+      "grad_norm": 0.8314547538757324,
+      "learning_rate": 9.99723728466338e-06,
+      "loss": 0.0472,
+      "step": 347
+    },
+    {
+      "epoch": 0.11042963036418177,
+      "grad_norm": 0.6330522894859314,
+      "learning_rate": 9.997050052371992e-06,
+      "loss": 0.0495,
+      "step": 348
+    },
+    {
+      "epoch": 0.11074695688821677,
+      "grad_norm": 0.6322121620178223,
+      "learning_rate": 9.996856683770857e-06,
+      "loss": 0.0407,
+      "step": 349
+    },
+    {
+      "epoch": 0.11106428341225177,
+      "grad_norm": 0.6152644753456116,
+      "learning_rate": 9.996657179097422e-06,
+      "loss": 0.0513,
+      "step": 350
+    },
+    {
+      "epoch": 0.11138160993628679,
+      "grad_norm": 0.5624389052391052,
+      "learning_rate": 9.996451538596678e-06,
+      "loss": 0.0411,
+      "step": 351
+    },
+    {
+      "epoch": 0.1116989364603218,
+      "grad_norm": 0.6775932312011719,
+      "learning_rate": 9.996239762521152e-06,
+      "loss": 0.0472,
+      "step": 352
+    },
+    {
+      "epoch": 0.1120162629843568,
+      "grad_norm": 0.6296716332435608,
+      "learning_rate": 9.996021851130897e-06,
+      "loss": 0.0453,
+      "step": 353
+    },
+    {
+      "epoch": 0.1123335895083918,
+      "grad_norm": 0.5194743275642395,
+      "learning_rate": 9.995797804693508e-06,
+      "loss": 0.0435,
+      "step": 354
+    },
+    {
+      "epoch": 0.1126509160324268,
+      "grad_norm": 0.9628121852874756,
+      "learning_rate": 9.995567623484111e-06,
+      "loss": 0.0519,
+      "step": 355
+    },
+    {
+      "epoch": 0.1129682425564618,
+      "grad_norm": 1.077511191368103,
+      "learning_rate": 9.995331307785365e-06,
+      "loss": 0.0495,
+      "step": 356
+    },
+    {
+      "epoch": 0.11328556908049682,
+      "grad_norm": 0.9500876069068909,
+      "learning_rate": 9.995088857887463e-06,
+      "loss": 0.0446,
+      "step": 357
+    },
+    {
+      "epoch": 0.11360289560453182,
+      "grad_norm": 0.6982964277267456,
+      "learning_rate": 9.994840274088131e-06,
+      "loss": 0.0596,
+      "step": 358
+    },
+    {
+      "epoch": 0.11392022212856683,
+      "grad_norm": 0.9036064743995667,
+      "learning_rate": 9.994585556692624e-06,
+      "loss": 0.0443,
+      "step": 359
+    },
+    {
+      "epoch": 0.11423754865260183,
+      "grad_norm": 0.865353524684906,
+      "learning_rate": 9.994324706013731e-06,
+      "loss": 0.0373,
+      "step": 360
+    },
+    {
+      "epoch": 0.11455487517663683,
+      "grad_norm": 0.6371411085128784,
+      "learning_rate": 9.994057722371777e-06,
+      "loss": 0.0476,
+      "step": 361
+    },
+    {
+      "epoch": 0.11487220170067183,
+      "grad_norm": 0.8190805912017822,
+      "learning_rate": 9.993784606094612e-06,
+      "loss": 0.0441,
+      "step": 362
+    },
+    {
+      "epoch": 0.11518952822470685,
+      "grad_norm": 0.7036651968955994,
+      "learning_rate": 9.99350535751762e-06,
+      "loss": 0.0459,
+      "step": 363
+    },
+    {
+      "epoch": 0.11550685474874185,
+      "grad_norm": 1.523719072341919,
+      "learning_rate": 9.993219976983712e-06,
+      "loss": 0.0424,
+      "step": 364
+    },
+    {
+      "epoch": 0.11582418127277685,
+      "grad_norm": 0.6216347217559814,
+      "learning_rate": 9.992928464843335e-06,
+      "loss": 0.0522,
+      "step": 365
+    },
+    {
+      "epoch": 0.11614150779681186,
+      "grad_norm": 0.8922724723815918,
+      "learning_rate": 9.992630821454458e-06,
+      "loss": 0.0477,
+      "step": 366
+    },
+    {
+      "epoch": 0.11645883432084686,
+      "grad_norm": 0.9970713257789612,
+      "learning_rate": 9.992327047182588e-06,
+      "loss": 0.0407,
+      "step": 367
+    },
+    {
+      "epoch": 0.11677616084488188,
+      "grad_norm": 0.6065555214881897,
+      "learning_rate": 9.99201714240075e-06,
+      "loss": 0.0434,
+      "step": 368
+    },
+    {
+      "epoch": 0.11709348736891688,
+      "grad_norm": 0.6207418441772461,
+      "learning_rate": 9.991701107489508e-06,
+      "loss": 0.0457,
+      "step": 369
+    },
+    {
+      "epoch": 0.11741081389295188,
+      "grad_norm": 0.5973998308181763,
+      "learning_rate": 9.991378942836945e-06,
+      "loss": 0.0445,
+      "step": 370
+    },
+    {
+      "epoch": 0.11772814041698688,
+      "grad_norm": 0.7065964937210083,
+      "learning_rate": 9.991050648838676e-06,
+      "loss": 0.0481,
+      "step": 371
+    },
+    {
+      "epoch": 0.11804546694102189,
+      "grad_norm": 0.6226929426193237,
+      "learning_rate": 9.990716225897844e-06,
+      "loss": 0.0444,
+      "step": 372
+    },
+    {
+      "epoch": 0.11836279346505689,
+      "grad_norm": 0.4240459203720093,
+      "learning_rate": 9.99037567442511e-06,
+      "loss": 0.0468,
+      "step": 373
+    },
+    {
+      "epoch": 0.1186801199890919,
+      "grad_norm": 0.48520708084106445,
+      "learning_rate": 9.990028994838673e-06,
+      "loss": 0.0486,
+      "step": 374
+    },
+    {
+      "epoch": 0.11899744651312691,
+      "grad_norm": 0.5164874792098999,
+      "learning_rate": 9.989676187564246e-06,
+      "loss": 0.0392,
+      "step": 375
+    },
+    {
+      "epoch": 0.11931477303716191,
+      "grad_norm": 0.5058465003967285,
+      "learning_rate": 9.989317253035077e-06,
+      "loss": 0.0435,
+      "step": 376
+    },
+    {
+      "epoch": 0.11963209956119691,
+      "grad_norm": 0.4859192669391632,
+      "learning_rate": 9.988952191691925e-06,
+      "loss": 0.0449,
+      "step": 377
+    },
+    {
+      "epoch": 0.11994942608523192,
+      "grad_norm": 0.459755003452301,
+      "learning_rate": 9.988581003983087e-06,
+      "loss": 0.0409,
+      "step": 378
+    },
+    {
+      "epoch": 0.12026675260926692,
+      "grad_norm": 0.49375367164611816,
+      "learning_rate": 9.988203690364375e-06,
+      "loss": 0.0462,
+      "step": 379
+    },
+    {
+      "epoch": 0.12058407913330194,
+      "grad_norm": 0.49485915899276733,
+      "learning_rate": 9.987820251299121e-06,
+      "loss": 0.0467,
+      "step": 380
+    },
+    {
+      "epoch": 0.12090140565733694,
+      "grad_norm": 0.5387222766876221,
+      "learning_rate": 9.987430687258189e-06,
+      "loss": 0.0469,
+      "step": 381
+    },
+    {
+      "epoch": 0.12121873218137194,
+      "grad_norm": 0.5329259037971497,
+      "learning_rate": 9.987034998719956e-06,
+      "loss": 0.0443,
+      "step": 382
+    },
+    {
+      "epoch": 0.12153605870540694,
+      "grad_norm": 0.6823049783706665,
+      "learning_rate": 9.986633186170319e-06,
+      "loss": 0.0475,
+      "step": 383
+    },
+    {
+      "epoch": 0.12185338522944195,
+      "grad_norm": 0.5331971049308777,
+      "learning_rate": 9.986225250102703e-06,
+      "loss": 0.0425,
+      "step": 384
+    },
+    {
+      "epoch": 0.12217071175347696,
+      "grad_norm": 0.5682464241981506,
+      "learning_rate": 9.985811191018044e-06,
+      "loss": 0.0444,
+      "step": 385
+    },
+    {
+      "epoch": 0.12248803827751197,
+      "grad_norm": 0.48949867486953735,
+      "learning_rate": 9.985391009424805e-06,
+      "loss": 0.0398,
+      "step": 386
+    },
+    {
+      "epoch": 0.12280536480154697,
+      "grad_norm": 0.6570077538490295,
+      "learning_rate": 9.98496470583896e-06,
+      "loss": 0.0471,
+      "step": 387
+    },
+    {
+      "epoch": 0.12312269132558197,
+      "grad_norm": 0.520440936088562,
+      "learning_rate": 9.984532280784007e-06,
+      "loss": 0.0432,
+      "step": 388
+    },
+    {
+      "epoch": 0.12344001784961697,
+      "grad_norm": 0.5352020263671875,
+      "learning_rate": 9.984093734790955e-06,
+      "loss": 0.0443,
+      "step": 389
+    },
+    {
+      "epoch": 0.12375734437365198,
+      "grad_norm": 0.48413532972335815,
+      "learning_rate": 9.983649068398335e-06,
+      "loss": 0.0464,
+      "step": 390
+    },
+    {
+      "epoch": 0.12407467089768699,
+      "grad_norm": 0.5233762860298157,
+      "learning_rate": 9.98319828215219e-06,
+      "loss": 0.0461,
+      "step": 391
+    },
+    {
+      "epoch": 0.124391997421722,
+      "grad_norm": 0.5393548011779785,
+      "learning_rate": 9.982741376606077e-06,
+      "loss": 0.044,
+      "step": 392
+    },
+    {
+      "epoch": 0.124709323945757,
+      "grad_norm": 0.48046964406967163,
+      "learning_rate": 9.982278352321075e-06,
+      "loss": 0.0448,
+      "step": 393
+    },
+    {
+      "epoch": 0.125026650469792,
+      "grad_norm": 0.47574758529663086,
+      "learning_rate": 9.98180920986577e-06,
+      "loss": 0.0396,
+      "step": 394
+    },
+    {
+      "epoch": 0.12534397699382702,
+      "grad_norm": 0.4889158010482788,
+      "learning_rate": 9.98133394981626e-06,
+      "loss": 0.0417,
+      "step": 395
+    },
+    {
+      "epoch": 0.125661303517862,
+      "grad_norm": 0.5132105946540833,
+      "learning_rate": 9.980852572756159e-06,
+      "loss": 0.0471,
+      "step": 396
+    },
+    {
+      "epoch": 0.12597863004189702,
+      "grad_norm": 0.48804590106010437,
+      "learning_rate": 9.980365079276593e-06,
+      "loss": 0.0458,
+      "step": 397
+    },
+    {
+      "epoch": 0.126295956565932,
+      "grad_norm": 0.4683227241039276,
+      "learning_rate": 9.979871469976197e-06,
+      "loss": 0.0362,
+      "step": 398
+    },
+    {
+      "epoch": 0.12661328308996703,
+      "grad_norm": 0.5643588304519653,
+      "learning_rate": 9.979371745461114e-06,
+      "loss": 0.0461,
+      "step": 399
+    },
+    {
+      "epoch": 0.12693060961400204,
+      "grad_norm": 0.5780950784683228,
+      "learning_rate": 9.978865906345e-06,
+      "loss": 0.0435,
+      "step": 400
+    },
+    {
+      "epoch": 0.12724793613803703,
+      "grad_norm": 0.4467637240886688,
+      "learning_rate": 9.978353953249023e-06,
+      "loss": 0.0424,
+      "step": 401
+    },
+    {
+      "epoch": 0.12756526266207205,
+      "grad_norm": 0.4732384979724884,
+      "learning_rate": 9.97783588680185e-06,
+      "loss": 0.0372,
+      "step": 402
+    },
+    {
+      "epoch": 0.12788258918610704,
+      "grad_norm": 0.5423407554626465,
+      "learning_rate": 9.97731170763966e-06,
+      "loss": 0.0445,
+      "step": 403
+    },
+    {
+      "epoch": 0.12819991571014205,
+      "grad_norm": 0.5567305684089661,
+      "learning_rate": 9.976781416406136e-06,
+      "loss": 0.0492,
+      "step": 404
+    },
+    {
+      "epoch": 0.12851724223417707,
+      "grad_norm": 0.5112591981887817,
+      "learning_rate": 9.976245013752472e-06,
+      "loss": 0.0386,
+      "step": 405
+    },
+    {
+      "epoch": 0.12883456875821206,
+      "grad_norm": 0.9472014307975769,
+      "learning_rate": 9.975702500337365e-06,
+      "loss": 0.0497,
+      "step": 406
+    },
+    {
+      "epoch": 0.12915189528224708,
+      "grad_norm": 0.4987757205963135,
+      "learning_rate": 9.975153876827008e-06,
+      "loss": 0.0377,
+      "step": 407
+    },
+    {
+      "epoch": 0.12946922180628206,
+      "grad_norm": 0.5201554298400879,
+      "learning_rate": 9.974599143895107e-06,
+      "loss": 0.0445,
+      "step": 408
+    },
+    {
+      "epoch": 0.12978654833031708,
+      "grad_norm": 0.52880859375,
+      "learning_rate": 9.974038302222865e-06,
+      "loss": 0.0441,
+      "step": 409
+    },
+    {
+      "epoch": 0.1301038748543521,
+      "grad_norm": 0.5272786617279053,
+      "learning_rate": 9.973471352498991e-06,
+      "loss": 0.0388,
+      "step": 410
+    },
+    {
+      "epoch": 0.13042120137838709,
+      "grad_norm": 0.47186312079429626,
+      "learning_rate": 9.972898295419688e-06,
+      "loss": 0.0425,
+      "step": 411
+    },
+    {
+      "epoch": 0.1307385279024221,
+      "grad_norm": 0.8499534726142883,
+      "learning_rate": 9.972319131688666e-06,
+      "loss": 0.0489,
+      "step": 412
+    },
+    {
+      "epoch": 0.1310558544264571,
+      "grad_norm": 0.6231266260147095,
+      "learning_rate": 9.971733862017126e-06,
+      "loss": 0.0421,
+      "step": 413
+    },
+    {
+      "epoch": 0.1313731809504921,
+      "grad_norm": 0.563764214515686,
+      "learning_rate": 9.971142487123772e-06,
+      "loss": 0.0442,
+      "step": 414
+    },
+    {
+      "epoch": 0.1316905074745271,
+      "grad_norm": 0.5588873624801636,
+      "learning_rate": 9.970545007734807e-06,
+      "loss": 0.046,
+      "step": 415
+    },
+    {
+      "epoch": 0.1320078339985621,
+      "grad_norm": 0.5489698648452759,
+      "learning_rate": 9.969941424583926e-06,
+      "loss": 0.0477,
+      "step": 416
+    },
+    {
+      "epoch": 0.13232516052259713,
+      "grad_norm": 0.5242834091186523,
+      "learning_rate": 9.969331738412323e-06,
+      "loss": 0.045,
+      "step": 417
+    },
+    {
+      "epoch": 0.13264248704663212,
+      "grad_norm": 0.4812135398387909,
+      "learning_rate": 9.96871594996868e-06,
+      "loss": 0.0385,
+      "step": 418
+    },
+    {
+      "epoch": 0.13295981357066713,
+      "grad_norm": 3.0158839225769043,
+      "learning_rate": 9.96809406000918e-06,
+      "loss": 0.0491,
+      "step": 419
+    },
+    {
+      "epoch": 0.13327714009470212,
+      "grad_norm": 0.4738430082798004,
+      "learning_rate": 9.967466069297496e-06,
+      "loss": 0.0408,
+      "step": 420
+    },
+    {
+      "epoch": 0.13359446661873714,
+      "grad_norm": 1.6419224739074707,
+      "learning_rate": 9.966831978604788e-06,
+      "loss": 0.046,
+      "step": 421
+    },
+    {
+      "epoch": 0.13391179314277216,
+      "grad_norm": 0.49285703897476196,
+      "learning_rate": 9.966191788709716e-06,
+      "loss": 0.0476,
+      "step": 422
+    },
+    {
+      "epoch": 0.13422911966680715,
+      "grad_norm": 0.5061706304550171,
+      "learning_rate": 9.965545500398421e-06,
+      "loss": 0.0439,
+      "step": 423
+    },
+    {
+      "epoch": 0.13454644619084216,
+      "grad_norm": 0.7146293520927429,
+      "learning_rate": 9.964893114464539e-06,
+      "loss": 0.0372,
+      "step": 424
+    },
+    {
+      "epoch": 0.13486377271487715,
+      "grad_norm": 0.47010061144828796,
+      "learning_rate": 9.964234631709188e-06,
+      "loss": 0.0404,
+      "step": 425
+    },
+    {
+      "epoch": 0.13518109923891217,
+      "grad_norm": 0.5113213658332825,
+      "learning_rate": 9.963570052940977e-06,
+      "loss": 0.0413,
+      "step": 426
+    },
+    {
+      "epoch": 0.13549842576294718,
+      "grad_norm": 0.4507155120372772,
+      "learning_rate": 9.962899378976001e-06,
+      "loss": 0.043,
+      "step": 427
+    },
+    {
+      "epoch": 0.13581575228698217,
+      "grad_norm": 0.49787163734436035,
+      "learning_rate": 9.962222610637837e-06,
+      "loss": 0.0437,
+      "step": 428
+    },
+    {
+      "epoch": 0.1361330788110172,
+      "grad_norm": 0.5881332755088806,
+      "learning_rate": 9.96153974875755e-06,
+      "loss": 0.0372,
+      "step": 429
+    },
+    {
+      "epoch": 0.13645040533505218,
+      "grad_norm": 1.156383991241455,
+      "learning_rate": 9.960850794173681e-06,
+      "loss": 0.0395,
+      "step": 430
+    },
+    {
+      "epoch": 0.1367677318590872,
+      "grad_norm": 0.4214046895503998,
+      "learning_rate": 9.96015574773226e-06,
+      "loss": 0.0409,
+      "step": 431
+    },
+    {
+      "epoch": 0.13708505838312218,
+      "grad_norm": 0.473908931016922,
+      "learning_rate": 9.959454610286794e-06,
+      "loss": 0.0426,
+      "step": 432
+    },
+    {
+      "epoch": 0.1374023849071572,
+      "grad_norm": 0.516278862953186,
+      "learning_rate": 9.958747382698268e-06,
+      "loss": 0.045,
+      "step": 433
+    },
+    {
+      "epoch": 0.13771971143119222,
+      "grad_norm": 0.4690360724925995,
+      "learning_rate": 9.958034065835151e-06,
+      "loss": 0.0454,
+      "step": 434
+    },
+    {
+      "epoch": 0.1380370379552272,
+      "grad_norm": 0.47334879636764526,
+      "learning_rate": 9.957314660573388e-06,
+      "loss": 0.0421,
+      "step": 435
+    },
+    {
+      "epoch": 0.13835436447926222,
+      "grad_norm": 0.5536037087440491,
+      "learning_rate": 9.956589167796392e-06,
+      "loss": 0.0377,
+      "step": 436
+    },
+    {
+      "epoch": 0.1386716910032972,
+      "grad_norm": 0.46338868141174316,
+      "learning_rate": 9.955857588395065e-06,
+      "loss": 0.0389,
+      "step": 437
+    },
+    {
+      "epoch": 0.13898901752733223,
+      "grad_norm": 0.45655298233032227,
+      "learning_rate": 9.955119923267774e-06,
+      "loss": 0.0352,
+      "step": 438
+    },
+    {
+      "epoch": 0.13930634405136724,
+      "grad_norm": 0.4080408215522766,
+      "learning_rate": 9.95437617332036e-06,
+      "loss": 0.0374,
+      "step": 439
+    },
+    {
+      "epoch": 0.13962367057540223,
+      "grad_norm": 0.5346148610115051,
+      "learning_rate": 9.95362633946614e-06,
+      "loss": 0.0421,
+      "step": 440
+    },
+    {
+      "epoch": 0.13994099709943725,
+      "grad_norm": 0.45880448818206787,
+      "learning_rate": 9.952870422625902e-06,
+      "loss": 0.0392,
+      "step": 441
+    },
+    {
+      "epoch": 0.14025832362347224,
+      "grad_norm": 0.4564225375652313,
+      "learning_rate": 9.952108423727897e-06,
+      "loss": 0.0389,
+      "step": 442
+    },
+    {
+      "epoch": 0.14057565014750725,
+      "grad_norm": 0.48790186643600464,
+      "learning_rate": 9.951340343707852e-06,
+      "loss": 0.0385,
+      "step": 443
+    },
+    {
+      "epoch": 0.14089297667154227,
+      "grad_norm": 0.42506635189056396,
+      "learning_rate": 9.950566183508959e-06,
+      "loss": 0.04,
+      "step": 444
+    },
+    {
+      "epoch": 0.14121030319557726,
+      "grad_norm": 0.6440918445587158,
+      "learning_rate": 9.949785944081876e-06,
+      "loss": 0.0414,
+      "step": 445
+    },
+    {
+      "epoch": 0.14152762971961227,
+      "grad_norm": 0.8506025671958923,
+      "learning_rate": 9.948999626384725e-06,
+      "loss": 0.0422,
+      "step": 446
+    },
+    {
+      "epoch": 0.14184495624364726,
+      "grad_norm": 0.4331110417842865,
+      "learning_rate": 9.948207231383094e-06,
+      "loss": 0.0371,
+      "step": 447
+    },
+    {
+      "epoch": 0.14216228276768228,
+      "grad_norm": 0.5053799152374268,
+      "learning_rate": 9.947408760050035e-06,
+      "loss": 0.0396,
+      "step": 448
+    },
+    {
+      "epoch": 0.14247960929171727,
+      "grad_norm": 0.5332431197166443,
+      "learning_rate": 9.946604213366058e-06,
+      "loss": 0.0409,
+      "step": 449
+    },
+    {
+      "epoch": 0.14279693581575229,
+      "grad_norm": 0.8472349047660828,
+      "learning_rate": 9.945793592319137e-06,
+      "loss": 0.0417,
+      "step": 450
+    },
+    {
+      "epoch": 0.1431142623397873,
+      "grad_norm": 0.5701362490653992,
+      "learning_rate": 9.9449768979047e-06,
+      "loss": 0.0453,
+      "step": 451
+    },
+    {
+      "epoch": 0.1434315888638223,
+      "grad_norm": 0.519732654094696,
+      "learning_rate": 9.944154131125643e-06,
+      "loss": 0.0417,
+      "step": 452
+    },
+    {
+      "epoch": 0.1437489153878573,
+      "grad_norm": 0.4713364541530609,
+      "learning_rate": 9.943325292992308e-06,
+      "loss": 0.0437,
+      "step": 453
+    },
+    {
+      "epoch": 0.1440662419118923,
+      "grad_norm": 0.5003937482833862,
+      "learning_rate": 9.942490384522498e-06,
+      "loss": 0.0368,
+      "step": 454
+    },
+    {
+      "epoch": 0.1443835684359273,
+      "grad_norm": 0.4261319935321808,
+      "learning_rate": 9.94164940674147e-06,
+      "loss": 0.0396,
+      "step": 455
+    },
+    {
+      "epoch": 0.14470089495996233,
+      "grad_norm": 0.864355742931366,
+      "learning_rate": 9.940802360681932e-06,
+      "loss": 0.0367,
+      "step": 456
+    },
+    {
+      "epoch": 0.14501822148399732,
+      "grad_norm": 0.4653952121734619,
+      "learning_rate": 9.939949247384046e-06,
+      "loss": 0.0377,
+      "step": 457
+    },
+    {
+      "epoch": 0.14533554800803233,
+      "grad_norm": 0.6019065380096436,
+      "learning_rate": 9.939090067895422e-06,
+      "loss": 0.0341,
+      "step": 458
+    },
+    {
+      "epoch": 0.14565287453206732,
+      "grad_norm": 0.6753557920455933,
+      "learning_rate": 9.938224823271123e-06,
+      "loss": 0.0428,
+      "step": 459
+    },
+    {
+      "epoch": 0.14597020105610234,
+      "grad_norm": 0.48566654324531555,
+      "learning_rate": 9.937353514573658e-06,
+      "loss": 0.0363,
+      "step": 460
+    },
+    {
+      "epoch": 0.14628752758013736,
+      "grad_norm": 0.5096802711486816,
+      "learning_rate": 9.936476142872979e-06,
+      "loss": 0.0383,
+      "step": 461
+    },
+    {
+      "epoch": 0.14660485410417234,
+      "grad_norm": 0.4175027012825012,
+      "learning_rate": 9.935592709246489e-06,
+      "loss": 0.0364,
+      "step": 462
+    },
+    {
+      "epoch": 0.14692218062820736,
+      "grad_norm": 0.5305432677268982,
+      "learning_rate": 9.934703214779034e-06,
+      "loss": 0.0432,
+      "step": 463
+    },
+    {
+      "epoch": 0.14723950715224235,
+      "grad_norm": 0.5047663450241089,
+      "learning_rate": 9.933807660562898e-06,
+      "loss": 0.0429,
+      "step": 464
+    },
+    {
+      "epoch": 0.14755683367627737,
+      "grad_norm": 0.5636289715766907,
+      "learning_rate": 9.932906047697808e-06,
+      "loss": 0.0378,
+      "step": 465
+    },
+    {
+      "epoch": 0.14787416020031235,
+      "grad_norm": 0.5088527798652649,
+      "learning_rate": 9.931998377290938e-06,
+      "loss": 0.0397,
+      "step": 466
+    },
+    {
+      "epoch": 0.14819148672434737,
+      "grad_norm": 0.5290652513504028,
+      "learning_rate": 9.931084650456892e-06,
+      "loss": 0.0386,
+      "step": 467
+    },
+    {
+      "epoch": 0.1485088132483824,
+      "grad_norm": 0.5171421766281128,
+      "learning_rate": 9.930164868317713e-06,
+      "loss": 0.0464,
+      "step": 468
+    },
+    {
+      "epoch": 0.14882613977241738,
+      "grad_norm": 0.42210906744003296,
+      "learning_rate": 9.929239032002885e-06,
+      "loss": 0.036,
+      "step": 469
+    },
+    {
+      "epoch": 0.1491434662964524,
+      "grad_norm": 0.5078014731407166,
+      "learning_rate": 9.928307142649315e-06,
+      "loss": 0.0356,
+      "step": 470
+    },
+    {
+      "epoch": 0.14946079282048738,
+      "grad_norm": 0.5068497061729431,
+      "learning_rate": 9.927369201401358e-06,
+      "loss": 0.0393,
+      "step": 471
+    },
+    {
+      "epoch": 0.1497781193445224,
+      "grad_norm": 0.4841766953468323,
+      "learning_rate": 9.92642520941079e-06,
+      "loss": 0.0367,
+      "step": 472
+    },
+    {
+      "epoch": 0.15009544586855741,
+      "grad_norm": 0.5417282581329346,
+      "learning_rate": 9.92547516783682e-06,
+      "loss": 0.0457,
+      "step": 473
+    },
+    {
+      "epoch": 0.1504127723925924,
+      "grad_norm": 0.4417283833026886,
+      "learning_rate": 9.924519077846089e-06,
+      "loss": 0.0395,
+      "step": 474
+    },
+    {
+      "epoch": 0.15073009891662742,
+      "grad_norm": 0.7089312076568604,
+      "learning_rate": 9.923556940612658e-06,
+      "loss": 0.0344,
+      "step": 475
+    },
+    {
+      "epoch": 0.1510474254406624,
+      "grad_norm": 0.41839781403541565,
+      "learning_rate": 9.92258875731802e-06,
+      "loss": 0.0364,
+      "step": 476
+    },
+    {
+      "epoch": 0.15136475196469742,
+      "grad_norm": 0.7053573727607727,
+      "learning_rate": 9.921614529151096e-06,
+      "loss": 0.0377,
+      "step": 477
+    },
+    {
+      "epoch": 0.15168207848873244,
+      "grad_norm": 0.550714373588562,
+      "learning_rate": 9.920634257308217e-06,
+      "loss": 0.0431,
+      "step": 478
+    },
+    {
+      "epoch": 0.15199940501276743,
+      "grad_norm": 0.47487685084342957,
+      "learning_rate": 9.91964794299315e-06,
+      "loss": 0.0402,
+      "step": 479
+    },
+    {
+      "epoch": 0.15231673153680245,
+      "grad_norm": 0.5477140545845032,
+      "learning_rate": 9.91865558741707e-06,
+      "loss": 0.0379,
+      "step": 480
+    },
+    {
+      "epoch": 0.15263405806083744,
+      "grad_norm": 0.49598705768585205,
+      "learning_rate": 9.917657191798582e-06,
+      "loss": 0.0411,
+      "step": 481
+    },
+    {
+      "epoch": 0.15295138458487245,
+      "grad_norm": 0.6432579159736633,
+      "learning_rate": 9.916652757363698e-06,
+      "loss": 0.0323,
+      "step": 482
+    },
+    {
+      "epoch": 0.15326871110890747,
+      "grad_norm": 0.7311610579490662,
+      "learning_rate": 9.915642285345852e-06,
+      "loss": 0.0428,
+      "step": 483
+    },
+    {
+      "epoch": 0.15358603763294246,
+      "grad_norm": 0.4128270447254181,
+      "learning_rate": 9.914625776985889e-06,
+      "loss": 0.033,
+      "step": 484
+    },
+    {
+      "epoch": 0.15390336415697747,
+      "grad_norm": 0.5466141700744629,
+      "learning_rate": 9.913603233532067e-06,
+      "loss": 0.042,
+      "step": 485
+    },
+    {
+      "epoch": 0.15422069068101246,
+      "grad_norm": 0.4603215456008911,
+      "learning_rate": 9.912574656240058e-06,
+      "loss": 0.0302,
+      "step": 486
+    },
+    {
+      "epoch": 0.15453801720504748,
+      "grad_norm": 0.49099481105804443,
+      "learning_rate": 9.911540046372939e-06,
+      "loss": 0.0422,
+      "step": 487
+    },
+    {
+      "epoch": 0.15485534372908247,
+      "grad_norm": 0.5768367052078247,
+      "learning_rate": 9.910499405201195e-06,
+      "loss": 0.0388,
+      "step": 488
+    },
+    {
+      "epoch": 0.15517267025311748,
+      "grad_norm": 0.44656628370285034,
+      "learning_rate": 9.909452734002722e-06,
+      "loss": 0.0382,
+      "step": 489
+    },
+    {
+      "epoch": 0.1554899967771525,
+      "grad_norm": 0.4093354642391205,
+      "learning_rate": 9.90840003406282e-06,
+      "loss": 0.0364,
+      "step": 490
+    },
+    {
+      "epoch": 0.1558073233011875,
+      "grad_norm": 0.6989220380783081,
+      "learning_rate": 9.907341306674185e-06,
+      "loss": 0.0419,
+      "step": 491
+    },
+    {
+      "epoch": 0.1561246498252225,
+      "grad_norm": 5.947473049163818,
+      "learning_rate": 9.906276553136924e-06,
+      "loss": 0.041,
+      "step": 492
+    },
+    {
+      "epoch": 0.1564419763492575,
+      "grad_norm": 0.7478053569793701,
+      "learning_rate": 9.905205774758538e-06,
+      "loss": 0.0419,
+      "step": 493
+    },
+    {
+      "epoch": 0.1567593028732925,
+      "grad_norm": 0.5153319239616394,
+      "learning_rate": 9.90412897285393e-06,
+      "loss": 0.0343,
+      "step": 494
+    },
+    {
+      "epoch": 0.15707662939732753,
+      "grad_norm": 0.8592816591262817,
+      "learning_rate": 9.903046148745395e-06,
+      "loss": 0.0401,
+      "step": 495
+    },
+    {
+      "epoch": 0.15739395592136252,
+      "grad_norm": 0.6446492075920105,
+      "learning_rate": 9.90195730376263e-06,
+      "loss": 0.0391,
+      "step": 496
+    },
+    {
+      "epoch": 0.15771128244539753,
+      "grad_norm": 0.4828226566314697,
+      "learning_rate": 9.900862439242719e-06,
+      "loss": 0.0356,
+      "step": 497
+    },
+    {
+      "epoch": 0.15802860896943252,
+      "grad_norm": 0.48052075505256653,
+      "learning_rate": 9.899761556530143e-06,
+      "loss": 0.0389,
+      "step": 498
+    },
+    {
+      "epoch": 0.15834593549346754,
+      "grad_norm": 0.4846798777580261,
+      "learning_rate": 9.89865465697677e-06,
+      "loss": 0.0389,
+      "step": 499
+    },
+    {
+      "epoch": 0.15866326201750255,
+      "grad_norm": 0.5561555027961731,
+      "learning_rate": 9.897541741941858e-06,
+      "loss": 0.0439,
+      "step": 500
+    },
+    {
+      "epoch": 0.15898058854153754,
+      "grad_norm": 0.9551202058792114,
+      "learning_rate": 9.896422812792051e-06,
+      "loss": 0.0473,
+      "step": 501
+    },
+    {
+      "epoch": 0.15929791506557256,
+      "grad_norm": 1.4664722681045532,
+      "learning_rate": 9.895297870901379e-06,
+      "loss": 0.035,
+      "step": 502
+    },
+    {
+      "epoch": 0.15961524158960755,
+      "grad_norm": 0.7867361307144165,
+      "learning_rate": 9.894166917651256e-06,
+      "loss": 0.0408,
+      "step": 503
+    },
+    {
+      "epoch": 0.15993256811364256,
+      "grad_norm": 0.6291322112083435,
+      "learning_rate": 9.893029954430477e-06,
+      "loss": 0.0406,
+      "step": 504
+    },
+    {
+      "epoch": 0.16024989463767755,
+      "grad_norm": 0.5173561573028564,
+      "learning_rate": 9.891886982635217e-06,
+      "loss": 0.0427,
+      "step": 505
+    },
+    {
+      "epoch": 0.16056722116171257,
+      "grad_norm": 0.5291993618011475,
+      "learning_rate": 9.890738003669029e-06,
+      "loss": 0.0375,
+      "step": 506
+    },
+    {
+      "epoch": 0.1608845476857476,
+      "grad_norm": 0.5636836290359497,
+      "learning_rate": 9.889583018942846e-06,
+      "loss": 0.0436,
+      "step": 507
+    },
+    {
+      "epoch": 0.16120187420978258,
+      "grad_norm": 0.5550052523612976,
+      "learning_rate": 9.888422029874971e-06,
+      "loss": 0.0386,
+      "step": 508
+    },
+    {
+      "epoch": 0.1615192007338176,
+      "grad_norm": 0.6075907945632935,
+      "learning_rate": 9.887255037891085e-06,
+      "loss": 0.0359,
+      "step": 509
+    },
+    {
+      "epoch": 0.16183652725785258,
+      "grad_norm": 1.6464531421661377,
+      "learning_rate": 9.886082044424238e-06,
+      "loss": 0.0352,
+      "step": 510
+    },
+    {
+      "epoch": 0.1621538537818876,
+      "grad_norm": 0.7066797614097595,
+      "learning_rate": 9.88490305091485e-06,
+      "loss": 0.0401,
+      "step": 511
+    },
+    {
+      "epoch": 0.1624711803059226,
+      "grad_norm": 0.6348215341567993,
+      "learning_rate": 9.883718058810708e-06,
+      "loss": 0.0401,
+      "step": 512
+    },
+    {
+      "epoch": 0.1627885068299576,
+      "grad_norm": 0.5230064988136292,
+      "learning_rate": 9.882527069566965e-06,
+      "loss": 0.0405,
+      "step": 513
+    },
+    {
+      "epoch": 0.16310583335399262,
+      "grad_norm": 0.5367875099182129,
+      "learning_rate": 9.881330084646144e-06,
+      "loss": 0.0386,
+      "step": 514
+    },
+    {
+      "epoch": 0.1634231598780276,
+      "grad_norm": 0.5916309952735901,
+      "learning_rate": 9.880127105518122e-06,
+      "loss": 0.0427,
+      "step": 515
+    },
+    {
+      "epoch": 0.16374048640206262,
+      "grad_norm": 0.48311853408813477,
+      "learning_rate": 9.878918133660144e-06,
+      "loss": 0.0392,
+      "step": 516
+    },
+    {
+      "epoch": 0.16405781292609764,
+      "grad_norm": 0.5291467905044556,
+      "learning_rate": 9.877703170556811e-06,
+      "loss": 0.0368,
+      "step": 517
+    },
+    {
+      "epoch": 0.16437513945013263,
+      "grad_norm": 0.512001097202301,
+      "learning_rate": 9.876482217700078e-06,
+      "loss": 0.0427,
+      "step": 518
+    },
+    {
+      "epoch": 0.16469246597416765,
+      "grad_norm": 0.5678003430366516,
+      "learning_rate": 9.875255276589261e-06,
+      "loss": 0.0431,
+      "step": 519
+    },
+    {
+      "epoch": 0.16500979249820263,
+      "grad_norm": 0.5127106308937073,
+      "learning_rate": 9.87402234873103e-06,
+      "loss": 0.0372,
+      "step": 520
+    },
+    {
+      "epoch": 0.16532711902223765,
+      "grad_norm": 0.6947556138038635,
+      "learning_rate": 9.872783435639397e-06,
+      "loss": 0.0395,
+      "step": 521
+    },
+    {
+      "epoch": 0.16564444554627264,
+      "grad_norm": 0.47612130641937256,
+      "learning_rate": 9.871538538835734e-06,
+      "loss": 0.0392,
+      "step": 522
+    },
+    {
+      "epoch": 0.16596177207030766,
+      "grad_norm": 0.4998747408390045,
+      "learning_rate": 9.870287659848758e-06,
+      "loss": 0.0403,
+      "step": 523
+    },
+    {
+      "epoch": 0.16627909859434267,
+      "grad_norm": 0.47464555501937866,
+      "learning_rate": 9.869030800214531e-06,
+      "loss": 0.0351,
+      "step": 524
+    },
+    {
+      "epoch": 0.16659642511837766,
+      "grad_norm": 0.48123589158058167,
+      "learning_rate": 9.86776796147646e-06,
+      "loss": 0.0329,
+      "step": 525
+    },
+    {
+      "epoch": 0.16691375164241268,
+      "grad_norm": 0.4649268388748169,
+      "learning_rate": 9.866499145185293e-06,
+      "loss": 0.0376,
+      "step": 526
+    },
+    {
+      "epoch": 0.16723107816644767,
+      "grad_norm": 0.5200221538543701,
+      "learning_rate": 9.86522435289912e-06,
+      "loss": 0.0358,
+      "step": 527
+    },
+    {
+      "epoch": 0.16754840469048268,
+      "grad_norm": 6.899172306060791,
+      "learning_rate": 9.863943586183367e-06,
+      "loss": 0.0329,
+      "step": 528
+    },
+    {
+      "epoch": 0.1678657312145177,
+      "grad_norm": 0.4819231629371643,
+      "learning_rate": 9.8626568466108e-06,
+      "loss": 0.0405,
+      "step": 529
+    },
+    {
+      "epoch": 0.1681830577385527,
+      "grad_norm": 0.568011999130249,
+      "learning_rate": 9.861364135761518e-06,
+      "loss": 0.0406,
+      "step": 530
+    },
+    {
+      "epoch": 0.1685003842625877,
+      "grad_norm": 0.5075401067733765,
+      "learning_rate": 9.86006545522295e-06,
+      "loss": 0.0412,
+      "step": 531
+    },
+    {
+      "epoch": 0.1688177107866227,
+      "grad_norm": 0.5815643072128296,
+      "learning_rate": 9.85876080658986e-06,
+      "loss": 0.0418,
+      "step": 532
+    },
+    {
+      "epoch": 0.1691350373106577,
+      "grad_norm": 0.4880857467651367,
+      "learning_rate": 9.857450191464337e-06,
+      "loss": 0.0387,
+      "step": 533
+    },
+    {
+      "epoch": 0.16945236383469273,
+      "grad_norm": 0.4363384544849396,
+      "learning_rate": 9.856133611455802e-06,
+      "loss": 0.039,
+      "step": 534
+    },
+    {
+      "epoch": 0.16976969035872772,
+      "grad_norm": 0.5793890953063965,
+      "learning_rate": 9.854811068180997e-06,
+      "loss": 0.038,
+      "step": 535
+    },
+    {
+      "epoch": 0.17008701688276273,
+      "grad_norm": 0.48941922187805176,
+      "learning_rate": 9.853482563263981e-06,
+      "loss": 0.037,
+      "step": 536
+    },
+    {
+      "epoch": 0.17040434340679772,
+      "grad_norm": 0.4803932309150696,
+      "learning_rate": 9.852148098336149e-06,
+      "loss": 0.0431,
+      "step": 537
+    },
+    {
+      "epoch": 0.17072166993083274,
+      "grad_norm": 0.4297909140586853,
+      "learning_rate": 9.850807675036199e-06,
+      "loss": 0.0369,
+      "step": 538
+    },
+    {
+      "epoch": 0.17103899645486773,
+      "grad_norm": 0.436347633600235,
+      "learning_rate": 9.849461295010157e-06,
+      "loss": 0.0396,
+      "step": 539
+    },
+    {
+      "epoch": 0.17135632297890274,
+      "grad_norm": 0.5017163157463074,
+      "learning_rate": 9.848108959911355e-06,
+      "loss": 0.0416,
+      "step": 540
+    },
+    {
+      "epoch": 0.17167364950293776,
+      "grad_norm": 0.4987901449203491,
+      "learning_rate": 9.846750671400447e-06,
+      "loss": 0.0421,
+      "step": 541
+    },
+    {
+      "epoch": 0.17199097602697275,
+      "grad_norm": 0.36170661449432373,
+      "learning_rate": 9.84538643114539e-06,
+      "loss": 0.0294,
+      "step": 542
+    },
+    {
+      "epoch": 0.17230830255100776,
+      "grad_norm": 0.5026159286499023,
+      "learning_rate": 9.844016240821455e-06,
+      "loss": 0.0432,
+      "step": 543
+    },
+    {
+      "epoch": 0.17262562907504275,
+      "grad_norm": 0.4289485812187195,
+      "learning_rate": 9.842640102111213e-06,
+      "loss": 0.0353,
+      "step": 544
+    },
+    {
+      "epoch": 0.17294295559907777,
+      "grad_norm": 0.4630093276500702,
+      "learning_rate": 9.841258016704547e-06,
+      "loss": 0.0369,
+      "step": 545
+    },
+    {
+      "epoch": 0.17326028212311279,
+      "grad_norm": 0.4598078429698944,
+      "learning_rate": 9.839869986298637e-06,
+      "loss": 0.0371,
+      "step": 546
+    },
+    {
+      "epoch": 0.17357760864714777,
+      "grad_norm": 0.664010763168335,
+      "learning_rate": 9.838476012597969e-06,
+      "loss": 0.0384,
+      "step": 547
+    },
+    {
+      "epoch": 0.1738949351711828,
+      "grad_norm": 0.5720415711402893,
+      "learning_rate": 9.83707609731432e-06,
+      "loss": 0.0333,
+      "step": 548
+    },
+    {
+      "epoch": 0.17421226169521778,
+      "grad_norm": 0.39645203948020935,
+      "learning_rate": 9.83567024216677e-06,
+      "loss": 0.0354,
+      "step": 549
+    },
+    {
+      "epoch": 0.1745295882192528,
+      "grad_norm": 3.5242910385131836,
+      "learning_rate": 9.834258448881687e-06,
+      "loss": 0.0345,
+      "step": 550
+    },
+    {
+      "epoch": 0.1748469147432878,
+      "grad_norm": 0.9625179171562195,
+      "learning_rate": 9.832840719192737e-06,
+      "loss": 0.0347,
+      "step": 551
+    },
+    {
+      "epoch": 0.1751642412673228,
+      "grad_norm": 0.48980042338371277,
+      "learning_rate": 9.83141705484087e-06,
+      "loss": 0.0399,
+      "step": 552
+    },
+    {
+      "epoch": 0.17548156779135782,
+      "grad_norm": 0.7079441547393799,
+      "learning_rate": 9.82998745757433e-06,
+      "loss": 0.0402,
+      "step": 553
+    },
+    {
+      "epoch": 0.1757988943153928,
+      "grad_norm": 0.7008752822875977,
+      "learning_rate": 9.82855192914864e-06,
+      "loss": 0.0437,
+      "step": 554
+    },
+    {
+      "epoch": 0.17611622083942782,
+      "grad_norm": 0.55488121509552,
+      "learning_rate": 9.827110471326612e-06,
+      "loss": 0.0372,
+      "step": 555
+    },
+    {
+      "epoch": 0.17643354736346284,
+      "grad_norm": 0.4979701638221741,
+      "learning_rate": 9.825663085878331e-06,
+      "loss": 0.0358,
+      "step": 556
+    },
+    {
+      "epoch": 0.17675087388749783,
+      "grad_norm": 0.4795512855052948,
+      "learning_rate": 9.824209774581176e-06,
+      "loss": 0.0436,
+      "step": 557
+    },
+    {
+      "epoch": 0.17706820041153284,
+      "grad_norm": 0.3831399381160736,
+      "learning_rate": 9.822750539219785e-06,
+      "loss": 0.0325,
+      "step": 558
+    },
+    {
+      "epoch": 0.17738552693556783,
+      "grad_norm": 0.3826386332511902,
+      "learning_rate": 9.821285381586086e-06,
+      "loss": 0.0316,
+      "step": 559
+    },
+    {
+      "epoch": 0.17770285345960285,
+      "grad_norm": 0.5129420161247253,
+      "learning_rate": 9.819814303479268e-06,
+      "loss": 0.045,
+      "step": 560
+    },
+    {
+      "epoch": 0.17802017998363784,
+      "grad_norm": 0.40777698159217834,
+      "learning_rate": 9.818337306705797e-06,
+      "loss": 0.0371,
+      "step": 561
+    },
+    {
+      "epoch": 0.17833750650767285,
+      "grad_norm": 0.44024044275283813,
+      "learning_rate": 9.816854393079402e-06,
+      "loss": 0.0328,
+      "step": 562
+    },
+    {
+      "epoch": 0.17865483303170787,
+      "grad_norm": 0.4359908103942871,
+      "learning_rate": 9.815365564421086e-06,
+      "loss": 0.0373,
+      "step": 563
+    },
+    {
+      "epoch": 0.17897215955574286,
+      "grad_norm": 0.4505244493484497,
+      "learning_rate": 9.813870822559107e-06,
+      "loss": 0.0334,
+      "step": 564
+    },
+    {
+      "epoch": 0.17928948607977788,
+      "grad_norm": 0.485819548368454,
+      "learning_rate": 9.812370169328988e-06,
+      "loss": 0.0383,
+      "step": 565
+    },
+    {
+      "epoch": 0.17960681260381287,
+      "grad_norm": 0.4061923921108246,
+      "learning_rate": 9.810863606573512e-06,
+      "loss": 0.0348,
+      "step": 566
+    },
+    {
+      "epoch": 0.17992413912784788,
+      "grad_norm": 0.43856677412986755,
+      "learning_rate": 9.809351136142717e-06,
+      "loss": 0.0354,
+      "step": 567
+    },
+    {
+      "epoch": 0.1802414656518829,
+      "grad_norm": 0.43922320008277893,
+      "learning_rate": 9.807832759893896e-06,
+      "loss": 0.0393,
+      "step": 568
+    },
+    {
+      "epoch": 0.1805587921759179,
+      "grad_norm": 0.5089945197105408,
+      "learning_rate": 9.806308479691595e-06,
+      "loss": 0.0373,
+      "step": 569
+    },
+    {
+      "epoch": 0.1808761186999529,
+      "grad_norm": 0.6611643433570862,
+      "learning_rate": 9.80477829740761e-06,
+      "loss": 0.0385,
+      "step": 570
+    },
+    {
+      "epoch": 0.1811934452239879,
+      "grad_norm": 0.5329744815826416,
+      "learning_rate": 9.803242214920981e-06,
+      "loss": 0.0407,
+      "step": 571
+    },
+    {
+      "epoch": 0.1815107717480229,
+      "grad_norm": 0.4627719223499298,
+      "learning_rate": 9.801700234118e-06,
+      "loss": 0.0367,
+      "step": 572
+    },
+    {
+      "epoch": 0.18182809827205793,
+      "grad_norm": 0.4487936198711395,
+      "learning_rate": 9.800152356892196e-06,
+      "loss": 0.036,
+      "step": 573
+    },
+    {
+      "epoch": 0.18214542479609291,
+      "grad_norm": 0.4558846652507782,
+      "learning_rate": 9.798598585144344e-06,
+      "loss": 0.04,
+      "step": 574
+    },
+    {
+      "epoch": 0.18246275132012793,
+      "grad_norm": 0.4488959312438965,
+      "learning_rate": 9.797038920782453e-06,
+      "loss": 0.0409,
+      "step": 575
+    },
+    {
+      "epoch": 0.18278007784416292,
+      "grad_norm": 0.5080794095993042,
+      "learning_rate": 9.79547336572177e-06,
+      "loss": 0.0429,
+      "step": 576
+    },
+    {
+      "epoch": 0.18309740436819794,
+      "grad_norm": 0.5124521851539612,
+      "learning_rate": 9.793901921884773e-06,
+      "loss": 0.0424,
+      "step": 577
+    },
+    {
+      "epoch": 0.18341473089223292,
+      "grad_norm": 0.434887170791626,
+      "learning_rate": 9.792324591201179e-06,
+      "loss": 0.0411,
+      "step": 578
+    },
+    {
+      "epoch": 0.18373205741626794,
+      "grad_norm": 0.44502225518226624,
+      "learning_rate": 9.790741375607924e-06,
+      "loss": 0.0391,
+      "step": 579
+    },
+    {
+      "epoch": 0.18404938394030296,
+      "grad_norm": 0.4703298509120941,
+      "learning_rate": 9.78915227704918e-06,
+      "loss": 0.0367,
+      "step": 580
+    },
+    {
+      "epoch": 0.18436671046433795,
+      "grad_norm": 0.46545666456222534,
+      "learning_rate": 9.787557297476331e-06,
+      "loss": 0.0381,
+      "step": 581
+    },
+    {
+      "epoch": 0.18468403698837296,
+      "grad_norm": 0.4980582594871521,
+      "learning_rate": 9.785956438847997e-06,
+      "loss": 0.0369,
+      "step": 582
+    },
+    {
+      "epoch": 0.18500136351240795,
+      "grad_norm": 0.7089335918426514,
+      "learning_rate": 9.784349703130008e-06,
+      "loss": 0.0402,
+      "step": 583
+    },
+    {
+      "epoch": 0.18531869003644297,
+      "grad_norm": 0.4325447678565979,
+      "learning_rate": 9.782737092295414e-06,
+      "loss": 0.0372,
+      "step": 584
+    },
+    {
+      "epoch": 0.18563601656047798,
+      "grad_norm": 0.6050954461097717,
+      "learning_rate": 9.78111860832448e-06,
+      "loss": 0.0338,
+      "step": 585
+    },
+    {
+      "epoch": 0.18595334308451297,
+      "grad_norm": 0.5050947666168213,
+      "learning_rate": 9.77949425320468e-06,
+      "loss": 0.0349,
+      "step": 586
+    },
+    {
+      "epoch": 0.186270669608548,
+      "grad_norm": 0.46751752495765686,
+      "learning_rate": 9.777864028930705e-06,
+      "loss": 0.035,
+      "step": 587
+    },
+    {
+      "epoch": 0.18658799613258298,
+      "grad_norm": 0.46274620294570923,
+      "learning_rate": 9.776227937504443e-06,
+      "loss": 0.0354,
+      "step": 588
+    },
+    {
+      "epoch": 0.186905322656618,
+      "grad_norm": 0.5234224200248718,
+      "learning_rate": 9.774585980934994e-06,
+      "loss": 0.0371,
+      "step": 589
+    },
+    {
+      "epoch": 0.187222649180653,
+      "grad_norm": 0.6037426590919495,
+      "learning_rate": 9.77293816123866e-06,
+      "loss": 0.0403,
+      "step": 590
+    },
+    {
+      "epoch": 0.187539975704688,
+      "grad_norm": 0.48288002610206604,
+      "learning_rate": 9.771284480438941e-06,
+      "loss": 0.036,
+      "step": 591
+    },
+    {
+      "epoch": 0.18785730222872302,
+      "grad_norm": 0.5748129487037659,
+      "learning_rate": 9.769624940566531e-06,
+      "loss": 0.0378,
+      "step": 592
+    },
+    {
+      "epoch": 0.188174628752758,
+      "grad_norm": 0.5043009519577026,
+      "learning_rate": 9.767959543659327e-06,
+      "loss": 0.034,
+      "step": 593
+    },
+    {
+      "epoch": 0.18849195527679302,
+      "grad_norm": 0.7868360877037048,
+      "learning_rate": 9.766288291762412e-06,
+      "loss": 0.0347,
+      "step": 594
+    },
+    {
+      "epoch": 0.188809281800828,
+      "grad_norm": 0.5767320394515991,
+      "learning_rate": 9.764611186928059e-06,
+      "loss": 0.0356,
+      "step": 595
+    },
+    {
+      "epoch": 0.18912660832486303,
+      "grad_norm": 0.4922977387905121,
+      "learning_rate": 9.762928231215731e-06,
+      "loss": 0.0339,
+      "step": 596
+    },
+    {
+      "epoch": 0.18944393484889804,
+      "grad_norm": 1.573745846748352,
+      "learning_rate": 9.761239426692077e-06,
+      "loss": 0.0429,
+      "step": 597
+    },
+    {
+      "epoch": 0.18976126137293303,
+      "grad_norm": 0.6214785575866699,
+      "learning_rate": 9.759544775430925e-06,
+      "loss": 0.0382,
+      "step": 598
+    },
+    {
+      "epoch": 0.19007858789696805,
+      "grad_norm": 0.5225369930267334,
+      "learning_rate": 9.757844279513282e-06,
+      "loss": 0.0353,
+      "step": 599
+    },
+    {
+      "epoch": 0.19039591442100304,
+      "grad_norm": 0.5668691992759705,
+      "learning_rate": 9.756137941027335e-06,
+      "loss": 0.0456,
+      "step": 600
+    },
+    {
+      "epoch": 0.19071324094503805,
+      "grad_norm": 0.5518437027931213,
+      "learning_rate": 9.754425762068447e-06,
+      "loss": 0.0363,
+      "step": 601
+    },
+    {
+      "epoch": 0.19103056746907307,
+      "grad_norm": 0.4800831377506256,
+      "learning_rate": 9.752707744739146e-06,
+      "loss": 0.0394,
+      "step": 602
+    },
+    {
+      "epoch": 0.19134789399310806,
+      "grad_norm": 0.5535571575164795,
+      "learning_rate": 9.750983891149138e-06,
+      "loss": 0.0403,
+      "step": 603
+    },
+    {
+      "epoch": 0.19166522051714308,
+      "grad_norm": 0.5235954523086548,
+      "learning_rate": 9.749254203415288e-06,
+      "loss": 0.0369,
+      "step": 604
+    },
+    {
+      "epoch": 0.19198254704117806,
+      "grad_norm": 0.5845518708229065,
+      "learning_rate": 9.747518683661632e-06,
+      "loss": 0.0346,
+      "step": 605
+    },
+    {
+      "epoch": 0.19229987356521308,
+      "grad_norm": 1.2771995067596436,
+      "learning_rate": 9.745777334019365e-06,
+      "loss": 0.0381,
+      "step": 606
+    },
+    {
+      "epoch": 0.1926172000892481,
+      "grad_norm": 0.6013443470001221,
+      "learning_rate": 9.74403015662684e-06,
+      "loss": 0.033,
+      "step": 607
+    },
+    {
+      "epoch": 0.19293452661328309,
+      "grad_norm": 0.4407309591770172,
+      "learning_rate": 9.742277153629564e-06,
+      "loss": 0.036,
+      "step": 608
+    },
+    {
+      "epoch": 0.1932518531373181,
+      "grad_norm": 1.0990318059921265,
+      "learning_rate": 9.740518327180207e-06,
+      "loss": 0.0364,
+      "step": 609
+    },
+    {
+      "epoch": 0.1935691796613531,
+      "grad_norm": 0.5374619364738464,
+      "learning_rate": 9.73875367943858e-06,
+      "loss": 0.0383,
+      "step": 610
+    },
+    {
+      "epoch": 0.1938865061853881,
+      "grad_norm": 0.4873441159725189,
+      "learning_rate": 9.736983212571646e-06,
+      "loss": 0.0319,
+      "step": 611
+    },
+    {
+      "epoch": 0.1942038327094231,
+      "grad_norm": 0.5347971320152283,
+      "learning_rate": 9.735206928753518e-06,
+      "loss": 0.0423,
+      "step": 612
+    },
+    {
+      "epoch": 0.1945211592334581,
+      "grad_norm": 0.5009407997131348,
+      "learning_rate": 9.733424830165446e-06,
+      "loss": 0.0364,
+      "step": 613
+    },
+    {
+      "epoch": 0.19483848575749313,
+      "grad_norm": 0.5272148251533508,
+      "learning_rate": 9.731636918995821e-06,
+      "loss": 0.0379,
+      "step": 614
+    },
+    {
+      "epoch": 0.19515581228152812,
+      "grad_norm": 1.4064557552337646,
+      "learning_rate": 9.729843197440179e-06,
+      "loss": 0.0399,
+      "step": 615
+    },
+    {
+      "epoch": 0.19547313880556313,
+      "grad_norm": 0.4689939320087433,
+      "learning_rate": 9.72804366770118e-06,
+      "loss": 0.0332,
+      "step": 616
+    },
+    {
+      "epoch": 0.19579046532959812,
+      "grad_norm": 0.8608937859535217,
+      "learning_rate": 9.726238331988625e-06,
+      "loss": 0.0391,
+      "step": 617
+    },
+    {
+      "epoch": 0.19610779185363314,
+      "grad_norm": 0.6153189539909363,
+      "learning_rate": 9.72442719251944e-06,
+      "loss": 0.0377,
+      "step": 618
+    },
+    {
+      "epoch": 0.19642511837766816,
+      "grad_norm": 0.45634961128234863,
+      "learning_rate": 9.722610251517685e-06,
+      "loss": 0.0353,
+      "step": 619
+    },
+    {
+      "epoch": 0.19674244490170315,
+      "grad_norm": 3.5250935554504395,
+      "learning_rate": 9.720787511214533e-06,
+      "loss": 0.0396,
+      "step": 620
+    },
+    {
+      "epoch": 0.19705977142573816,
+      "grad_norm": 0.531997799873352,
+      "learning_rate": 9.718958973848287e-06,
+      "loss": 0.0319,
+      "step": 621
+    },
+    {
+      "epoch": 0.19737709794977315,
+      "grad_norm": 0.5198058485984802,
+      "learning_rate": 9.717124641664366e-06,
+      "loss": 0.0329,
+      "step": 622
+    },
+    {
+      "epoch": 0.19769442447380817,
+      "grad_norm": 0.5732515454292297,
+      "learning_rate": 9.715284516915303e-06,
+      "loss": 0.0345,
+      "step": 623
+    },
+    {
+      "epoch": 0.19801175099784318,
+      "grad_norm": 0.5469380021095276,
+      "learning_rate": 9.71343860186075e-06,
+      "loss": 0.0387,
+      "step": 624
+    },
+    {
+      "epoch": 0.19832907752187817,
+      "grad_norm": 1.345428705215454,
+      "learning_rate": 9.711586898767462e-06,
+      "loss": 0.0373,
+      "step": 625
+    },
+    {
+      "epoch": 0.1986464040459132,
+      "grad_norm": 0.5007116794586182,
+      "learning_rate": 9.709729409909308e-06,
+      "loss": 0.0308,
+      "step": 626
+    },
+    {
+      "epoch": 0.19896373056994818,
+      "grad_norm": 0.5097343325614929,
+      "learning_rate": 9.707866137567257e-06,
+      "loss": 0.0379,
+      "step": 627
+    },
+    {
+      "epoch": 0.1992810570939832,
+      "grad_norm": 0.48711511492729187,
+      "learning_rate": 9.705997084029383e-06,
+      "loss": 0.0385,
+      "step": 628
+    },
+    {
+      "epoch": 0.1995983836180182,
+      "grad_norm": 0.5661397576332092,
+      "learning_rate": 9.704122251590862e-06,
+      "loss": 0.0342,
+      "step": 629
+    },
+    {
+      "epoch": 0.1999157101420532,
+      "grad_norm": 0.7066240906715393,
+      "learning_rate": 9.702241642553956e-06,
+      "loss": 0.0336,
+      "step": 630
+    },
+    {
+      "epoch": 0.20023303666608822,
+      "grad_norm": 0.61566162109375,
+      "learning_rate": 9.700355259228033e-06,
+      "loss": 0.0314,
+      "step": 631
+    },
+    {
+      "epoch": 0.2005503631901232,
+      "grad_norm": 0.5768104791641235,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.0373,
+      "step": 632
+    },
+    {
+      "epoch": 0.20086768971415822,
+      "grad_norm": 0.6001889705657959,
+      "learning_rate": 9.696565178982028e-06,
+      "loss": 0.0471,
+      "step": 633
+    },
+    {
+      "epoch": 0.2011850162381932,
+      "grad_norm": 0.518028199672699,
+      "learning_rate": 9.694661486716115e-06,
+      "loss": 0.0356,
+      "step": 634
+    },
+    {
+      "epoch": 0.20150234276222823,
+      "grad_norm": 0.5039363503456116,
+      "learning_rate": 9.692752029469511e-06,
+      "loss": 0.0324,
+      "step": 635
+    },
+    {
+      "epoch": 0.20181966928626324,
+      "grad_norm": 0.4602259695529938,
+      "learning_rate": 9.690836809587005e-06,
+      "loss": 0.0345,
+      "step": 636
+    },
+    {
+      "epoch": 0.20213699581029823,
+      "grad_norm": 0.5799177289009094,
+      "learning_rate": 9.68891582942046e-06,
+      "loss": 0.037,
+      "step": 637
+    },
+    {
+      "epoch": 0.20245432233433325,
+      "grad_norm": 0.5667338967323303,
+      "learning_rate": 9.686989091328814e-06,
+      "loss": 0.04,
+      "step": 638
+    },
+    {
+      "epoch": 0.20277164885836824,
+      "grad_norm": 1.3075451850891113,
+      "learning_rate": 9.685056597678075e-06,
+      "loss": 0.0354,
+      "step": 639
+    },
+    {
+      "epoch": 0.20308897538240325,
+      "grad_norm": 0.5199016332626343,
+      "learning_rate": 9.68311835084132e-06,
+      "loss": 0.0344,
+      "step": 640
+    },
+    {
+      "epoch": 0.20340630190643827,
+      "grad_norm": 0.7419707775115967,
+      "learning_rate": 9.681174353198687e-06,
+      "loss": 0.0369,
+      "step": 641
+    },
+    {
+      "epoch": 0.20372362843047326,
+      "grad_norm": 1.1335899829864502,
+      "learning_rate": 9.679224607137382e-06,
+      "loss": 0.0344,
+      "step": 642
+    },
+    {
+      "epoch": 0.20404095495450827,
+      "grad_norm": 0.46918225288391113,
+      "learning_rate": 9.677269115051665e-06,
+      "loss": 0.0328,
+      "step": 643
+    },
+    {
+      "epoch": 0.20435828147854326,
+      "grad_norm": 0.6314353942871094,
+      "learning_rate": 9.675307879342854e-06,
+      "loss": 0.0409,
+      "step": 644
+    },
+    {
+      "epoch": 0.20467560800257828,
+      "grad_norm": 0.5286089181900024,
+      "learning_rate": 9.67334090241932e-06,
+      "loss": 0.0386,
+      "step": 645
+    },
+    {
+      "epoch": 0.2049929345266133,
+      "grad_norm": 0.4890168309211731,
+      "learning_rate": 9.671368186696488e-06,
+      "loss": 0.03,
+      "step": 646
+    },
+    {
+      "epoch": 0.20531026105064829,
+      "grad_norm": 0.5555238723754883,
+      "learning_rate": 9.669389734596819e-06,
+      "loss": 0.0325,
+      "step": 647
+    },
+    {
+      "epoch": 0.2056275875746833,
+      "grad_norm": 0.5869796276092529,
+      "learning_rate": 9.66740554854983e-06,
+      "loss": 0.0278,
+      "step": 648
+    },
+    {
+      "epoch": 0.2059449140987183,
+      "grad_norm": 0.5451986789703369,
+      "learning_rate": 9.665415630992073e-06,
+      "loss": 0.0405,
+      "step": 649
+    },
+    {
+      "epoch": 0.2062622406227533,
+      "grad_norm": 0.5067962408065796,
+      "learning_rate": 9.663419984367139e-06,
+      "loss": 0.0368,
+      "step": 650
+    },
+    {
+      "epoch": 0.2065795671467883,
+      "grad_norm": 0.5210440754890442,
+      "learning_rate": 9.661418611125657e-06,
+      "loss": 0.0359,
+      "step": 651
+    },
+    {
+      "epoch": 0.2068968936708233,
+      "grad_norm": 0.5049238204956055,
+      "learning_rate": 9.659411513725285e-06,
+      "loss": 0.0353,
+      "step": 652
+    },
+    {
+      "epoch": 0.20721422019485833,
+      "grad_norm": 0.5459165573120117,
+      "learning_rate": 9.657398694630713e-06,
+      "loss": 0.0368,
+      "step": 653
+    },
+    {
+      "epoch": 0.20753154671889332,
+      "grad_norm": 0.5596637725830078,
+      "learning_rate": 9.655380156313653e-06,
+      "loss": 0.0371,
+      "step": 654
+    },
+    {
+      "epoch": 0.20784887324292833,
+      "grad_norm": 0.4836740791797638,
+      "learning_rate": 9.653355901252844e-06,
+      "loss": 0.0325,
+      "step": 655
+    },
+    {
+      "epoch": 0.20816619976696332,
+      "grad_norm": 0.6189664006233215,
+      "learning_rate": 9.651325931934046e-06,
+      "loss": 0.0366,
+      "step": 656
+    },
+    {
+      "epoch": 0.20848352629099834,
+      "grad_norm": 0.4946371912956238,
+      "learning_rate": 9.649290250850029e-06,
+      "loss": 0.0361,
+      "step": 657
+    },
+    {
+      "epoch": 0.20880085281503336,
+      "grad_norm": 0.5777744054794312,
+      "learning_rate": 9.647248860500586e-06,
+      "loss": 0.035,
+      "step": 658
+    },
+    {
+      "epoch": 0.20911817933906834,
+      "grad_norm": 0.5777953863143921,
+      "learning_rate": 9.645201763392513e-06,
+      "loss": 0.0365,
+      "step": 659
+    },
+    {
+      "epoch": 0.20943550586310336,
+      "grad_norm": 0.6436187028884888,
+      "learning_rate": 9.643148962039622e-06,
+      "loss": 0.0321,
+      "step": 660
+    },
+    {
+      "epoch": 0.20975283238713835,
+      "grad_norm": 0.6856178641319275,
+      "learning_rate": 9.641090458962723e-06,
+      "loss": 0.033,
+      "step": 661
+    },
+    {
+      "epoch": 0.21007015891117337,
+      "grad_norm": 0.45667991042137146,
+      "learning_rate": 9.639026256689628e-06,
+      "loss": 0.0319,
+      "step": 662
+    },
+    {
+      "epoch": 0.21038748543520838,
+      "grad_norm": 0.48816201090812683,
+      "learning_rate": 9.636956357755152e-06,
+      "loss": 0.0309,
+      "step": 663
+    },
+    {
+      "epoch": 0.21070481195924337,
+      "grad_norm": 0.598262369632721,
+      "learning_rate": 9.634880764701103e-06,
+      "loss": 0.0353,
+      "step": 664
+    },
+    {
+      "epoch": 0.2110221384832784,
+      "grad_norm": 0.5323143005371094,
+      "learning_rate": 9.632799480076278e-06,
+      "loss": 0.037,
+      "step": 665
+    },
+    {
+      "epoch": 0.21133946500731338,
+      "grad_norm": 0.5205509066581726,
+      "learning_rate": 9.630712506436471e-06,
+      "loss": 0.041,
+      "step": 666
+    },
+    {
+      "epoch": 0.2116567915313484,
+      "grad_norm": 0.5344911813735962,
+      "learning_rate": 9.628619846344453e-06,
+      "loss": 0.0346,
+      "step": 667
+    },
+    {
+      "epoch": 0.21197411805538338,
+      "grad_norm": 1.1526832580566406,
+      "learning_rate": 9.626521502369984e-06,
+      "loss": 0.0341,
+      "step": 668
+    },
+    {
+      "epoch": 0.2122914445794184,
+      "grad_norm": 0.40215981006622314,
+      "learning_rate": 9.624417477089802e-06,
+      "loss": 0.0317,
+      "step": 669
+    },
+    {
+      "epoch": 0.21260877110345341,
+      "grad_norm": 0.5034760236740112,
+      "learning_rate": 9.622307773087623e-06,
+      "loss": 0.0409,
+      "step": 670
+    },
+    {
+      "epoch": 0.2129260976274884,
+      "grad_norm": 0.5594977140426636,
+      "learning_rate": 9.620192392954134e-06,
+      "loss": 0.0355,
+      "step": 671
+    },
+    {
+      "epoch": 0.21324342415152342,
+      "grad_norm": 0.4485695958137512,
+      "learning_rate": 9.618071339286991e-06,
+      "loss": 0.0356,
+      "step": 672
+    },
+    {
+      "epoch": 0.2135607506755584,
+      "grad_norm": 0.607118546962738,
+      "learning_rate": 9.615944614690821e-06,
+      "loss": 0.0324,
+      "step": 673
+    },
+    {
+      "epoch": 0.21387807719959342,
+      "grad_norm": 0.5235499739646912,
+      "learning_rate": 9.613812221777212e-06,
+      "loss": 0.0368,
+      "step": 674
+    },
+    {
+      "epoch": 0.21419540372362844,
+      "grad_norm": 0.5809747576713562,
+      "learning_rate": 9.611674163164716e-06,
+      "loss": 0.0373,
+      "step": 675
+    },
+    {
+      "epoch": 0.21451273024766343,
+      "grad_norm": 0.5115391612052917,
+      "learning_rate": 9.609530441478838e-06,
+      "loss": 0.036,
+      "step": 676
+    },
+    {
+      "epoch": 0.21483005677169845,
+      "grad_norm": 0.45732414722442627,
+      "learning_rate": 9.60738105935204e-06,
+      "loss": 0.0312,
+      "step": 677
+    },
+    {
+      "epoch": 0.21514738329573344,
+      "grad_norm": 0.47172895073890686,
+      "learning_rate": 9.605226019423733e-06,
+      "loss": 0.0374,
+      "step": 678
+    },
+    {
+      "epoch": 0.21546470981976845,
+      "grad_norm": 0.44555094838142395,
+      "learning_rate": 9.603065324340276e-06,
+      "loss": 0.0294,
+      "step": 679
+    },
+    {
+      "epoch": 0.21578203634380347,
+      "grad_norm": 0.6839290857315063,
+      "learning_rate": 9.600898976754977e-06,
+      "loss": 0.0361,
+      "step": 680
+    },
+    {
+      "epoch": 0.21609936286783846,
+      "grad_norm": 0.3829723596572876,
+      "learning_rate": 9.598726979328079e-06,
+      "loss": 0.0295,
+      "step": 681
+    },
+    {
+      "epoch": 0.21641668939187347,
+      "grad_norm": 1.058274507522583,
+      "learning_rate": 9.596549334726766e-06,
+      "loss": 0.0337,
+      "step": 682
+    },
+    {
+      "epoch": 0.21673401591590846,
+      "grad_norm": 0.46028947830200195,
+      "learning_rate": 9.594366045625155e-06,
+      "loss": 0.0353,
+      "step": 683
+    },
+    {
+      "epoch": 0.21705134243994348,
+      "grad_norm": 0.3606852889060974,
+      "learning_rate": 9.592177114704295e-06,
+      "loss": 0.0297,
+      "step": 684
+    },
+    {
+      "epoch": 0.21736866896397847,
+      "grad_norm": 0.6824389696121216,
+      "learning_rate": 9.589982544652167e-06,
+      "loss": 0.0372,
+      "step": 685
+    },
+    {
+      "epoch": 0.21768599548801348,
+      "grad_norm": 3.5930075645446777,
+      "learning_rate": 9.58778233816367e-06,
+      "loss": 0.0341,
+      "step": 686
+    },
+    {
+      "epoch": 0.2180033220120485,
+      "grad_norm": 0.6226539611816406,
+      "learning_rate": 9.585576497940628e-06,
+      "loss": 0.038,
+      "step": 687
+    },
+    {
+      "epoch": 0.2183206485360835,
+      "grad_norm": 0.5187017321586609,
+      "learning_rate": 9.583365026691785e-06,
+      "loss": 0.038,
+      "step": 688
+    },
+    {
+      "epoch": 0.2186379750601185,
+      "grad_norm": 0.7778983116149902,
+      "learning_rate": 9.581147927132797e-06,
+      "loss": 0.0405,
+      "step": 689
+    },
+    {
+      "epoch": 0.2189553015841535,
+      "grad_norm": 0.5897039771080017,
+      "learning_rate": 9.578925201986229e-06,
+      "loss": 0.0356,
+      "step": 690
+    },
+    {
+      "epoch": 0.2192726281081885,
+      "grad_norm": 0.7329979538917542,
+      "learning_rate": 9.576696853981562e-06,
+      "loss": 0.0368,
+      "step": 691
+    },
+    {
+      "epoch": 0.21958995463222353,
+      "grad_norm": 0.4420118033885956,
+      "learning_rate": 9.574462885855173e-06,
+      "loss": 0.0337,
+      "step": 692
+    },
+    {
+      "epoch": 0.21990728115625852,
+      "grad_norm": 0.4787854254245758,
+      "learning_rate": 9.572223300350349e-06,
+      "loss": 0.0364,
+      "step": 693
+    },
+    {
+      "epoch": 0.22022460768029353,
+      "grad_norm": 0.6599032282829285,
+      "learning_rate": 9.569978100217267e-06,
+      "loss": 0.0316,
+      "step": 694
+    },
+    {
+      "epoch": 0.22054193420432852,
+      "grad_norm": 2.6733546257019043,
+      "learning_rate": 9.567727288213005e-06,
+      "loss": 0.0364,
+      "step": 695
+    },
+    {
+      "epoch": 0.22085926072836354,
+      "grad_norm": 0.48702892661094666,
+      "learning_rate": 9.565470867101528e-06,
+      "loss": 0.0348,
+      "step": 696
+    },
+    {
+      "epoch": 0.22117658725239855,
+      "grad_norm": 0.5166319608688354,
+      "learning_rate": 9.56320883965369e-06,
+      "loss": 0.0351,
+      "step": 697
+    },
+    {
+      "epoch": 0.22149391377643354,
+      "grad_norm": 0.4663035571575165,
+      "learning_rate": 9.560941208647231e-06,
+      "loss": 0.0335,
+      "step": 698
+    },
+    {
+      "epoch": 0.22181124030046856,
+      "grad_norm": 0.5864226222038269,
+      "learning_rate": 9.558667976866773e-06,
+      "loss": 0.0334,
+      "step": 699
+    },
+    {
+      "epoch": 0.22212856682450355,
+      "grad_norm": 0.5785648822784424,
+      "learning_rate": 9.55638914710381e-06,
+      "loss": 0.0363,
+      "step": 700
+    },
+    {
+      "epoch": 0.22244589334853856,
+      "grad_norm": 0.8943732380867004,
+      "learning_rate": 9.554104722156716e-06,
+      "loss": 0.0324,
+      "step": 701
+    },
+    {
+      "epoch": 0.22276321987257358,
+      "grad_norm": 0.5664440989494324,
+      "learning_rate": 9.551814704830734e-06,
+      "loss": 0.0325,
+      "step": 702
+    },
+    {
+      "epoch": 0.22308054639660857,
+      "grad_norm": 0.5445716381072998,
+      "learning_rate": 9.549519097937974e-06,
+      "loss": 0.0317,
+      "step": 703
+    },
+    {
+      "epoch": 0.2233978729206436,
+      "grad_norm": 0.4895453453063965,
+      "learning_rate": 9.547217904297411e-06,
+      "loss": 0.0328,
+      "step": 704
+    },
+    {
+      "epoch": 0.22371519944467858,
+      "grad_norm": 0.4746696949005127,
+      "learning_rate": 9.544911126734878e-06,
+      "loss": 0.0335,
+      "step": 705
+    },
+    {
+      "epoch": 0.2240325259687136,
+      "grad_norm": 0.5209836363792419,
+      "learning_rate": 9.542598768083068e-06,
+      "loss": 0.0342,
+      "step": 706
+    },
+    {
+      "epoch": 0.22434985249274858,
+      "grad_norm": 0.4738750159740448,
+      "learning_rate": 9.540280831181525e-06,
+      "loss": 0.0324,
+      "step": 707
+    },
+    {
+      "epoch": 0.2246671790167836,
+      "grad_norm": 0.4653942286968231,
+      "learning_rate": 9.537957318876646e-06,
+      "loss": 0.0332,
+      "step": 708
+    },
+    {
+      "epoch": 0.2249845055408186,
+      "grad_norm": 0.43025103211402893,
+      "learning_rate": 9.53562823402167e-06,
+      "loss": 0.0335,
+      "step": 709
+    },
+    {
+      "epoch": 0.2253018320648536,
+      "grad_norm": 0.5600267648696899,
+      "learning_rate": 9.533293579476684e-06,
+      "loss": 0.0413,
+      "step": 710
+    },
+    {
+      "epoch": 0.22561915858888862,
+      "grad_norm": 0.7055615186691284,
+      "learning_rate": 9.53095335810861e-06,
+      "loss": 0.034,
+      "step": 711
+    },
+    {
+      "epoch": 0.2259364851129236,
+      "grad_norm": 0.49561578035354614,
+      "learning_rate": 9.528607572791209e-06,
+      "loss": 0.0339,
+      "step": 712
+    },
+    {
+      "epoch": 0.22625381163695862,
+      "grad_norm": 0.4777248203754425,
+      "learning_rate": 9.526256226405075e-06,
+      "loss": 0.0357,
+      "step": 713
+    },
+    {
+      "epoch": 0.22657113816099364,
+      "grad_norm": 0.36987778544425964,
+      "learning_rate": 9.523899321837627e-06,
+      "loss": 0.0333,
+      "step": 714
+    },
+    {
+      "epoch": 0.22688846468502863,
+      "grad_norm": 0.5368096232414246,
+      "learning_rate": 9.521536861983115e-06,
+      "loss": 0.0346,
+      "step": 715
+    },
+    {
+      "epoch": 0.22720579120906365,
+      "grad_norm": 0.4266177713871002,
+      "learning_rate": 9.519168849742603e-06,
+      "loss": 0.034,
+      "step": 716
+    },
+    {
+      "epoch": 0.22752311773309863,
+      "grad_norm": 0.510219395160675,
+      "learning_rate": 9.516795288023984e-06,
+      "loss": 0.0313,
+      "step": 717
+    },
+    {
+      "epoch": 0.22784044425713365,
+      "grad_norm": 0.5956120491027832,
+      "learning_rate": 9.514416179741954e-06,
+      "loss": 0.0307,
+      "step": 718
+    },
+    {
+      "epoch": 0.22815777078116867,
+      "grad_norm": 0.6387599110603333,
+      "learning_rate": 9.512031527818028e-06,
+      "loss": 0.0337,
+      "step": 719
+    },
+    {
+      "epoch": 0.22847509730520366,
+      "grad_norm": 0.3651878237724304,
+      "learning_rate": 9.509641335180529e-06,
+      "loss": 0.0308,
+      "step": 720
+    },
+    {
+      "epoch": 0.22879242382923867,
+      "grad_norm": 1.0050114393234253,
+      "learning_rate": 9.507245604764576e-06,
+      "loss": 0.0349,
+      "step": 721
+    },
+    {
+      "epoch": 0.22910975035327366,
+      "grad_norm": 0.4613276720046997,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.0336,
+      "step": 722
+    },
+    {
+      "epoch": 0.22942707687730868,
+      "grad_norm": 1.5468878746032715,
+      "learning_rate": 9.502437542371812e-06,
+      "loss": 0.0316,
+      "step": 723
+    },
+    {
+      "epoch": 0.22974440340134367,
+      "grad_norm": 0.581998884677887,
+      "learning_rate": 9.500025216299235e-06,
+      "loss": 0.0321,
+      "step": 724
+    },
+    {
+      "epoch": 0.23006172992537868,
+      "grad_norm": 0.8211390972137451,
+      "learning_rate": 9.49760736425667e-06,
+      "loss": 0.0342,
+      "step": 725
+    },
+    {
+      "epoch": 0.2303790564494137,
+      "grad_norm": 0.6262272000312805,
+      "learning_rate": 9.495183989213208e-06,
+      "loss": 0.0389,
+      "step": 726
+    },
+    {
+      "epoch": 0.2306963829734487,
+      "grad_norm": 0.7935025691986084,
+      "learning_rate": 9.49275509414472e-06,
+      "loss": 0.035,
+      "step": 727
+    },
+    {
+      "epoch": 0.2310137094974837,
+      "grad_norm": 0.7618018984794617,
+      "learning_rate": 9.490320682033854e-06,
+      "loss": 0.0332,
+      "step": 728
+    },
+    {
+      "epoch": 0.2313310360215187,
+      "grad_norm": 1.1134957075119019,
+      "learning_rate": 9.487880755870038e-06,
+      "loss": 0.0348,
+      "step": 729
+    },
+    {
+      "epoch": 0.2316483625455537,
+      "grad_norm": 0.6110552549362183,
+      "learning_rate": 9.485435318649468e-06,
+      "loss": 0.0371,
+      "step": 730
+    },
+    {
+      "epoch": 0.23196568906958873,
+      "grad_norm": 0.4054882824420929,
+      "learning_rate": 9.482984373375105e-06,
+      "loss": 0.0302,
+      "step": 731
+    },
+    {
+      "epoch": 0.23228301559362372,
+      "grad_norm": 0.5592225790023804,
+      "learning_rate": 9.480527923056679e-06,
+      "loss": 0.0326,
+      "step": 732
+    },
+    {
+      "epoch": 0.23260034211765873,
+      "grad_norm": 2.251985788345337,
+      "learning_rate": 9.478065970710677e-06,
+      "loss": 0.0352,
+      "step": 733
+    },
+    {
+      "epoch": 0.23291766864169372,
+      "grad_norm": 0.5407910346984863,
+      "learning_rate": 9.475598519360345e-06,
+      "loss": 0.0345,
+      "step": 734
+    },
+    {
+      "epoch": 0.23323499516572874,
+      "grad_norm": 0.5035356283187866,
+      "learning_rate": 9.473125572035675e-06,
+      "loss": 0.0325,
+      "step": 735
+    },
+    {
+      "epoch": 0.23355232168976375,
+      "grad_norm": 0.4409586489200592,
+      "learning_rate": 9.470647131773418e-06,
+      "loss": 0.0314,
+      "step": 736
+    },
+    {
+      "epoch": 0.23386964821379874,
+      "grad_norm": 0.7080483436584473,
+      "learning_rate": 9.468163201617063e-06,
+      "loss": 0.0329,
+      "step": 737
+    },
+    {
+      "epoch": 0.23418697473783376,
+      "grad_norm": 0.4371355175971985,
+      "learning_rate": 9.46567378461684e-06,
+      "loss": 0.0311,
+      "step": 738
+    },
+    {
+      "epoch": 0.23450430126186875,
+      "grad_norm": 0.683476448059082,
+      "learning_rate": 9.463178883829725e-06,
+      "loss": 0.0393,
+      "step": 739
+    },
+    {
+      "epoch": 0.23482162778590376,
+      "grad_norm": 0.44821420311927795,
+      "learning_rate": 9.460678502319419e-06,
+      "loss": 0.031,
+      "step": 740
+    },
+    {
+      "epoch": 0.23513895430993875,
+      "grad_norm": 0.4718499481678009,
+      "learning_rate": 9.458172643156354e-06,
+      "loss": 0.0325,
+      "step": 741
+    },
+    {
+      "epoch": 0.23545628083397377,
+      "grad_norm": 0.5231450200080872,
+      "learning_rate": 9.455661309417697e-06,
+      "loss": 0.0355,
+      "step": 742
+    },
+    {
+      "epoch": 0.23577360735800879,
+      "grad_norm": 0.5215026140213013,
+      "learning_rate": 9.453144504187327e-06,
+      "loss": 0.0371,
+      "step": 743
+    },
+    {
+      "epoch": 0.23609093388204377,
+      "grad_norm": 0.7727586627006531,
+      "learning_rate": 9.450622230555849e-06,
+      "loss": 0.0323,
+      "step": 744
+    },
+    {
+      "epoch": 0.2364082604060788,
+      "grad_norm": 0.6420536637306213,
+      "learning_rate": 9.448094491620581e-06,
+      "loss": 0.0341,
+      "step": 745
+    },
+    {
+      "epoch": 0.23672558693011378,
+      "grad_norm": 0.7929708957672119,
+      "learning_rate": 9.44556129048555e-06,
+      "loss": 0.0336,
+      "step": 746
+    },
+    {
+      "epoch": 0.2370429134541488,
+      "grad_norm": 0.5932604670524597,
+      "learning_rate": 9.443022630261495e-06,
+      "loss": 0.0373,
+      "step": 747
+    },
+    {
+      "epoch": 0.2373602399781838,
+      "grad_norm": 0.5067250728607178,
+      "learning_rate": 9.440478514065855e-06,
+      "loss": 0.0285,
+      "step": 748
+    },
+    {
+      "epoch": 0.2376775665022188,
+      "grad_norm": 0.6431717872619629,
+      "learning_rate": 9.437928945022772e-06,
+      "loss": 0.0322,
+      "step": 749
+    },
+    {
+      "epoch": 0.23799489302625382,
+      "grad_norm": 0.5668692588806152,
+      "learning_rate": 9.435373926263078e-06,
+      "loss": 0.0314,
+      "step": 750
+    },
+    {
+      "epoch": 0.2383122195502888,
+      "grad_norm": 0.5935972332954407,
+      "learning_rate": 9.432813460924308e-06,
+      "loss": 0.0344,
+      "step": 751
+    },
+    {
+      "epoch": 0.23862954607432382,
+      "grad_norm": 0.5271613001823425,
+      "learning_rate": 9.430247552150673e-06,
+      "loss": 0.0354,
+      "step": 752
+    },
+    {
+      "epoch": 0.23894687259835884,
+      "grad_norm": 0.5951942205429077,
+      "learning_rate": 9.427676203093077e-06,
+      "loss": 0.0307,
+      "step": 753
+    },
+    {
+      "epoch": 0.23926419912239383,
+      "grad_norm": 0.49435317516326904,
+      "learning_rate": 9.425099416909101e-06,
+      "loss": 0.028,
+      "step": 754
+    },
+    {
+      "epoch": 0.23958152564642884,
+      "grad_norm": 0.5195945501327515,
+      "learning_rate": 9.422517196763002e-06,
+      "loss": 0.0352,
+      "step": 755
+    },
+    {
+      "epoch": 0.23989885217046383,
+      "grad_norm": 0.8718374371528625,
+      "learning_rate": 9.419929545825714e-06,
+      "loss": 0.0281,
+      "step": 756
+    },
+    {
+      "epoch": 0.24021617869449885,
+      "grad_norm": 0.6569284200668335,
+      "learning_rate": 9.417336467274838e-06,
+      "loss": 0.0327,
+      "step": 757
+    },
+    {
+      "epoch": 0.24053350521853384,
+      "grad_norm": 0.6640499234199524,
+      "learning_rate": 9.414737964294636e-06,
+      "loss": 0.038,
+      "step": 758
+    },
+    {
+      "epoch": 0.24085083174256885,
+      "grad_norm": 0.6330658197402954,
+      "learning_rate": 9.412134040076035e-06,
+      "loss": 0.0412,
+      "step": 759
+    },
+    {
+      "epoch": 0.24116815826660387,
+      "grad_norm": 0.9485225677490234,
+      "learning_rate": 9.40952469781662e-06,
+      "loss": 0.0344,
+      "step": 760
+    },
+    {
+      "epoch": 0.24148548479063886,
+      "grad_norm": 0.5819979310035706,
+      "learning_rate": 9.40690994072063e-06,
+      "loss": 0.0331,
+      "step": 761
+    },
+    {
+      "epoch": 0.24180281131467388,
+      "grad_norm": 0.5249894857406616,
+      "learning_rate": 9.404289771998946e-06,
+      "loss": 0.0364,
+      "step": 762
+    },
+    {
+      "epoch": 0.24212013783870887,
+      "grad_norm": 0.6229386925697327,
+      "learning_rate": 9.401664194869105e-06,
+      "loss": 0.0284,
+      "step": 763
+    },
+    {
+      "epoch": 0.24243746436274388,
+      "grad_norm": 1.1252851486206055,
+      "learning_rate": 9.399033212555276e-06,
+      "loss": 0.0364,
+      "step": 764
+    },
+    {
+      "epoch": 0.2427547908867789,
+      "grad_norm": 0.6530624032020569,
+      "learning_rate": 9.396396828288272e-06,
+      "loss": 0.0313,
+      "step": 765
+    },
+    {
+      "epoch": 0.2430721174108139,
+      "grad_norm": 0.6041601300239563,
+      "learning_rate": 9.393755045305538e-06,
+      "loss": 0.0355,
+      "step": 766
+    },
+    {
+      "epoch": 0.2433894439348489,
+      "grad_norm": 0.9328456521034241,
+      "learning_rate": 9.391107866851143e-06,
+      "loss": 0.0326,
+      "step": 767
+    },
+    {
+      "epoch": 0.2437067704588839,
+      "grad_norm": 0.4834827780723572,
+      "learning_rate": 9.388455296175792e-06,
+      "loss": 0.033,
+      "step": 768
+    },
+    {
+      "epoch": 0.2440240969829189,
+      "grad_norm": 0.6758521795272827,
+      "learning_rate": 9.3857973365368e-06,
+      "loss": 0.0334,
+      "step": 769
+    },
+    {
+      "epoch": 0.24434142350695393,
+      "grad_norm": 0.5084520578384399,
+      "learning_rate": 9.383133991198113e-06,
+      "loss": 0.0303,
+      "step": 770
+    },
+    {
+      "epoch": 0.24465875003098891,
+      "grad_norm": 0.5185012817382812,
+      "learning_rate": 9.380465263430276e-06,
+      "loss": 0.0413,
+      "step": 771
+    },
+    {
+      "epoch": 0.24497607655502393,
+      "grad_norm": 0.460488885641098,
+      "learning_rate": 9.377791156510456e-06,
+      "loss": 0.0346,
+      "step": 772
+    },
+    {
+      "epoch": 0.24529340307905892,
+      "grad_norm": 2.3728208541870117,
+      "learning_rate": 9.375111673722415e-06,
+      "loss": 0.0307,
+      "step": 773
+    },
+    {
+      "epoch": 0.24561072960309394,
+      "grad_norm": 0.4785768389701843,
+      "learning_rate": 9.372426818356523e-06,
+      "loss": 0.0339,
+      "step": 774
+    },
+    {
+      "epoch": 0.24592805612712895,
+      "grad_norm": 0.45519906282424927,
+      "learning_rate": 9.369736593709748e-06,
+      "loss": 0.0355,
+      "step": 775
+    },
+    {
+      "epoch": 0.24624538265116394,
+      "grad_norm": 1.4345992803573608,
+      "learning_rate": 9.36704100308565e-06,
+      "loss": 0.0293,
+      "step": 776
+    },
+    {
+      "epoch": 0.24656270917519896,
+      "grad_norm": 0.49177953600883484,
+      "learning_rate": 9.364340049794374e-06,
+      "loss": 0.0317,
+      "step": 777
+    },
+    {
+      "epoch": 0.24688003569923395,
+      "grad_norm": 0.5470086932182312,
+      "learning_rate": 9.361633737152656e-06,
+      "loss": 0.0299,
+      "step": 778
+    },
+    {
+      "epoch": 0.24719736222326896,
+      "grad_norm": 0.7837552428245544,
+      "learning_rate": 9.358922068483813e-06,
+      "loss": 0.03,
+      "step": 779
+    },
+    {
+      "epoch": 0.24751468874730395,
+      "grad_norm": 0.5691478252410889,
+      "learning_rate": 9.356205047117735e-06,
+      "loss": 0.0294,
+      "step": 780
+    },
+    {
+      "epoch": 0.24783201527133897,
+      "grad_norm": 0.589525580406189,
+      "learning_rate": 9.353482676390888e-06,
+      "loss": 0.0302,
+      "step": 781
+    },
+    {
+      "epoch": 0.24814934179537398,
+      "grad_norm": 0.7869940996170044,
+      "learning_rate": 9.350754959646306e-06,
+      "loss": 0.035,
+      "step": 782
+    },
+    {
+      "epoch": 0.24846666831940897,
+      "grad_norm": 0.532746434211731,
+      "learning_rate": 9.348021900233587e-06,
+      "loss": 0.0246,
+      "step": 783
+    },
+    {
+      "epoch": 0.248783994843444,
+      "grad_norm": 0.8993604183197021,
+      "learning_rate": 9.345283501508898e-06,
+      "loss": 0.0356,
+      "step": 784
+    },
+    {
+      "epoch": 0.24910132136747898,
+      "grad_norm": 0.517388105392456,
+      "learning_rate": 9.342539766834945e-06,
+      "loss": 0.0311,
+      "step": 785
+    },
+    {
+      "epoch": 0.249418647891514,
+      "grad_norm": 0.5598728656768799,
+      "learning_rate": 9.339790699581004e-06,
+      "loss": 0.0295,
+      "step": 786
+    },
+    {
+      "epoch": 0.249735974415549,
+      "grad_norm": 0.5227993726730347,
+      "learning_rate": 9.33703630312289e-06,
+      "loss": 0.0326,
+      "step": 787
+    },
+    {
+      "epoch": 0.250053300939584,
+      "grad_norm": 0.5973880290985107,
+      "learning_rate": 9.334276580842966e-06,
+      "loss": 0.0311,
+      "step": 788
+    },
+    {
+      "epoch": 0.250370627463619,
+      "grad_norm": 0.9228516221046448,
+      "learning_rate": 9.331511536130133e-06,
+      "loss": 0.032,
+      "step": 789
+    },
+    {
+      "epoch": 0.25068795398765403,
+      "grad_norm": 0.6291566491127014,
+      "learning_rate": 9.328741172379827e-06,
+      "loss": 0.0298,
+      "step": 790
+    },
+    {
+      "epoch": 0.251005280511689,
+      "grad_norm": 0.7358800768852234,
+      "learning_rate": 9.325965492994018e-06,
+      "loss": 0.0348,
+      "step": 791
+    },
+    {
+      "epoch": 0.251322607035724,
+      "grad_norm": 0.5709190368652344,
+      "learning_rate": 9.323184501381204e-06,
+      "loss": 0.0447,
+      "step": 792
+    },
+    {
+      "epoch": 0.25163993355975905,
+      "grad_norm": 0.5870211720466614,
+      "learning_rate": 9.320398200956403e-06,
+      "loss": 0.0331,
+      "step": 793
+    },
+    {
+      "epoch": 0.25195726008379404,
+      "grad_norm": 0.6013067364692688,
+      "learning_rate": 9.317606595141156e-06,
+      "loss": 0.0284,
+      "step": 794
+    },
+    {
+      "epoch": 0.25227458660782903,
+      "grad_norm": 0.6303397417068481,
+      "learning_rate": 9.314809687363515e-06,
+      "loss": 0.0328,
+      "step": 795
+    },
+    {
+      "epoch": 0.252591913131864,
+      "grad_norm": 0.5101282596588135,
+      "learning_rate": 9.312007481058045e-06,
+      "loss": 0.0332,
+      "step": 796
+    },
+    {
+      "epoch": 0.25290923965589907,
+      "grad_norm": 0.47284528613090515,
+      "learning_rate": 9.309199979665821e-06,
+      "loss": 0.029,
+      "step": 797
+    },
+    {
+      "epoch": 0.25322656617993405,
+      "grad_norm": 0.984382688999176,
+      "learning_rate": 9.306387186634412e-06,
+      "loss": 0.0344,
+      "step": 798
+    },
+    {
+      "epoch": 0.25354389270396904,
+      "grad_norm": 0.5734308362007141,
+      "learning_rate": 9.303569105417894e-06,
+      "loss": 0.0351,
+      "step": 799
+    },
+    {
+      "epoch": 0.2538612192280041,
+      "grad_norm": 1.0161045789718628,
+      "learning_rate": 9.30074573947683e-06,
+      "loss": 0.031,
+      "step": 800
+    },
+    {
+      "epoch": 0.2541785457520391,
+      "grad_norm": 0.5779903531074524,
+      "learning_rate": 9.297917092278272e-06,
+      "loss": 0.0328,
+      "step": 801
+    },
+    {
+      "epoch": 0.25449587227607406,
+      "grad_norm": 0.7489080429077148,
+      "learning_rate": 9.295083167295767e-06,
+      "loss": 0.0301,
+      "step": 802
+    },
+    {
+      "epoch": 0.2548131988001091,
+      "grad_norm": 0.6738011837005615,
+      "learning_rate": 9.292243968009332e-06,
+      "loss": 0.0344,
+      "step": 803
+    },
+    {
+      "epoch": 0.2551305253241441,
+      "grad_norm": 0.419712096452713,
+      "learning_rate": 9.289399497905467e-06,
+      "loss": 0.0303,
+      "step": 804
+    },
+    {
+      "epoch": 0.2554478518481791,
+      "grad_norm": 0.5418825149536133,
+      "learning_rate": 9.286549760477141e-06,
+      "loss": 0.0329,
+      "step": 805
+    },
+    {
+      "epoch": 0.2557651783722141,
+      "grad_norm": 0.4912303686141968,
+      "learning_rate": 9.283694759223795e-06,
+      "loss": 0.0331,
+      "step": 806
+    },
+    {
+      "epoch": 0.2560825048962491,
+      "grad_norm": 0.4939482808113098,
+      "learning_rate": 9.280834497651334e-06,
+      "loss": 0.0281,
+      "step": 807
+    },
+    {
+      "epoch": 0.2563998314202841,
+      "grad_norm": 1.1112980842590332,
+      "learning_rate": 9.277968979272115e-06,
+      "loss": 0.0318,
+      "step": 808
+    },
+    {
+      "epoch": 0.2567171579443191,
+      "grad_norm": 0.6621851325035095,
+      "learning_rate": 9.275098207604958e-06,
+      "loss": 0.036,
+      "step": 809
+    },
+    {
+      "epoch": 0.25703448446835414,
+      "grad_norm": 1.9360847473144531,
+      "learning_rate": 9.272222186175132e-06,
+      "loss": 0.0354,
+      "step": 810
+    },
+    {
+      "epoch": 0.25735181099238913,
+      "grad_norm": 0.7800442576408386,
+      "learning_rate": 9.269340918514353e-06,
+      "loss": 0.0321,
+      "step": 811
+    },
+    {
+      "epoch": 0.2576691375164241,
+      "grad_norm": 0.6715080738067627,
+      "learning_rate": 9.266454408160779e-06,
+      "loss": 0.0289,
+      "step": 812
+    },
+    {
+      "epoch": 0.2579864640404591,
+      "grad_norm": 2.909111499786377,
+      "learning_rate": 9.263562658659005e-06,
+      "loss": 0.0319,
+      "step": 813
+    },
+    {
+      "epoch": 0.25830379056449415,
+      "grad_norm": 0.5100283026695251,
+      "learning_rate": 9.260665673560058e-06,
+      "loss": 0.0332,
+      "step": 814
+    },
+    {
+      "epoch": 0.25862111708852914,
+      "grad_norm": 0.46259382367134094,
+      "learning_rate": 9.257763456421398e-06,
+      "loss": 0.0312,
+      "step": 815
+    },
+    {
+      "epoch": 0.25893844361256413,
+      "grad_norm": 0.7316282391548157,
+      "learning_rate": 9.25485601080691e-06,
+      "loss": 0.034,
+      "step": 816
+    },
+    {
+      "epoch": 0.2592557701365992,
+      "grad_norm": 0.801801323890686,
+      "learning_rate": 9.251943340286896e-06,
+      "loss": 0.0281,
+      "step": 817
+    },
+    {
+      "epoch": 0.25957309666063416,
+      "grad_norm": 0.569520890712738,
+      "learning_rate": 9.249025448438076e-06,
+      "loss": 0.0287,
+      "step": 818
+    },
+    {
+      "epoch": 0.25989042318466915,
+      "grad_norm": 0.48731184005737305,
+      "learning_rate": 9.24610233884358e-06,
+      "loss": 0.0299,
+      "step": 819
+    },
+    {
+      "epoch": 0.2602077497087042,
+      "grad_norm": 0.7043786644935608,
+      "learning_rate": 9.243174015092954e-06,
+      "loss": 0.0267,
+      "step": 820
+    },
+    {
+      "epoch": 0.2605250762327392,
+      "grad_norm": 0.4703860282897949,
+      "learning_rate": 9.24024048078213e-06,
+      "loss": 0.0285,
+      "step": 821
+    },
+    {
+      "epoch": 0.26084240275677417,
+      "grad_norm": 0.5326614379882812,
+      "learning_rate": 9.237301739513456e-06,
+      "loss": 0.0277,
+      "step": 822
+    },
+    {
+      "epoch": 0.26115972928080916,
+      "grad_norm": 0.6649437546730042,
+      "learning_rate": 9.234357794895664e-06,
+      "loss": 0.0287,
+      "step": 823
+    },
+    {
+      "epoch": 0.2614770558048442,
+      "grad_norm": 0.9807226061820984,
+      "learning_rate": 9.231408650543875e-06,
+      "loss": 0.028,
+      "step": 824
+    },
+    {
+      "epoch": 0.2617943823288792,
+      "grad_norm": 0.4513666033744812,
+      "learning_rate": 9.228454310079603e-06,
+      "loss": 0.029,
+      "step": 825
+    },
+    {
+      "epoch": 0.2621117088529142,
+      "grad_norm": 0.5832451581954956,
+      "learning_rate": 9.225494777130735e-06,
+      "loss": 0.0284,
+      "step": 826
+    },
+    {
+      "epoch": 0.2624290353769492,
+      "grad_norm": 0.6744049787521362,
+      "learning_rate": 9.22253005533154e-06,
+      "loss": 0.0343,
+      "step": 827
+    },
+    {
+      "epoch": 0.2627463619009842,
+      "grad_norm": 0.7426224946975708,
+      "learning_rate": 9.219560148322655e-06,
+      "loss": 0.0285,
+      "step": 828
+    },
+    {
+      "epoch": 0.2630636884250192,
+      "grad_norm": 0.6556416153907776,
+      "learning_rate": 9.216585059751083e-06,
+      "loss": 0.028,
+      "step": 829
+    },
+    {
+      "epoch": 0.2633810149490542,
+      "grad_norm": 1.5548570156097412,
+      "learning_rate": 9.213604793270196e-06,
+      "loss": 0.0334,
+      "step": 830
+    },
+    {
+      "epoch": 0.26369834147308924,
+      "grad_norm": 0.8730886578559875,
+      "learning_rate": 9.21061935253972e-06,
+      "loss": 0.0288,
+      "step": 831
+    },
+    {
+      "epoch": 0.2640156679971242,
+      "grad_norm": 0.4794308543205261,
+      "learning_rate": 9.207628741225735e-06,
+      "loss": 0.0307,
+      "step": 832
+    },
+    {
+      "epoch": 0.2643329945211592,
+      "grad_norm": 0.47464418411254883,
+      "learning_rate": 9.204632963000671e-06,
+      "loss": 0.0289,
+      "step": 833
+    },
+    {
+      "epoch": 0.26465032104519426,
+      "grad_norm": 0.6282751560211182,
+      "learning_rate": 9.201632021543305e-06,
+      "loss": 0.0257,
+      "step": 834
+    },
+    {
+      "epoch": 0.26496764756922925,
+      "grad_norm": 0.6647489070892334,
+      "learning_rate": 9.19862592053875e-06,
+      "loss": 0.0323,
+      "step": 835
+    },
+    {
+      "epoch": 0.26528497409326424,
+      "grad_norm": 0.45851823687553406,
+      "learning_rate": 9.195614663678458e-06,
+      "loss": 0.0287,
+      "step": 836
+    },
+    {
+      "epoch": 0.2656023006172993,
+      "grad_norm": 0.44865816831588745,
+      "learning_rate": 9.192598254660212e-06,
+      "loss": 0.0313,
+      "step": 837
+    },
+    {
+      "epoch": 0.26591962714133427,
+      "grad_norm": 0.6523016691207886,
+      "learning_rate": 9.18957669718812e-06,
+      "loss": 0.0319,
+      "step": 838
+    },
+    {
+      "epoch": 0.26623695366536926,
+      "grad_norm": 0.5544301271438599,
+      "learning_rate": 9.186549994972618e-06,
+      "loss": 0.0366,
+      "step": 839
+    },
+    {
+      "epoch": 0.26655428018940425,
+      "grad_norm": 0.5257997512817383,
+      "learning_rate": 9.18351815173045e-06,
+      "loss": 0.0283,
+      "step": 840
+    },
+    {
+      "epoch": 0.2668716067134393,
+      "grad_norm": 0.6622651815414429,
+      "learning_rate": 9.18048117118468e-06,
+      "loss": 0.0301,
+      "step": 841
+    },
+    {
+      "epoch": 0.2671889332374743,
+      "grad_norm": 0.5245144367218018,
+      "learning_rate": 9.177439057064684e-06,
+      "loss": 0.0292,
+      "step": 842
+    },
+    {
+      "epoch": 0.26750625976150927,
+      "grad_norm": 0.5707793235778809,
+      "learning_rate": 9.174391813106129e-06,
+      "loss": 0.0315,
+      "step": 843
+    },
+    {
+      "epoch": 0.2678235862855443,
+      "grad_norm": 0.4293830990791321,
+      "learning_rate": 9.171339443050993e-06,
+      "loss": 0.0276,
+      "step": 844
+    },
+    {
+      "epoch": 0.2681409128095793,
+      "grad_norm": 0.52970951795578,
+      "learning_rate": 9.168281950647545e-06,
+      "loss": 0.0292,
+      "step": 845
+    },
+    {
+      "epoch": 0.2684582393336143,
+      "grad_norm": 0.6536352634429932,
+      "learning_rate": 9.165219339650344e-06,
+      "loss": 0.0349,
+      "step": 846
+    },
+    {
+      "epoch": 0.2687755658576493,
+      "grad_norm": 0.5803107619285583,
+      "learning_rate": 9.162151613820236e-06,
+      "loss": 0.0311,
+      "step": 847
+    },
+    {
+      "epoch": 0.2690928923816843,
+      "grad_norm": 0.8978036642074585,
+      "learning_rate": 9.159078776924347e-06,
+      "loss": 0.0253,
+      "step": 848
+    },
+    {
+      "epoch": 0.2694102189057193,
+      "grad_norm": 1.1981711387634277,
+      "learning_rate": 9.156000832736073e-06,
+      "loss": 0.0301,
+      "step": 849
+    },
+    {
+      "epoch": 0.2697275454297543,
+      "grad_norm": 0.589787483215332,
+      "learning_rate": 9.152917785035097e-06,
+      "loss": 0.0288,
+      "step": 850
+    },
+    {
+      "epoch": 0.27004487195378934,
+      "grad_norm": 0.5028876662254333,
+      "learning_rate": 9.149829637607354e-06,
+      "loss": 0.0309,
+      "step": 851
+    },
+    {
+      "epoch": 0.27036219847782433,
+      "grad_norm": 1.6075416803359985,
+      "learning_rate": 9.146736394245048e-06,
+      "loss": 0.0287,
+      "step": 852
+    },
+    {
+      "epoch": 0.2706795250018593,
+      "grad_norm": 0.5125949382781982,
+      "learning_rate": 9.143638058746642e-06,
+      "loss": 0.0334,
+      "step": 853
+    },
+    {
+      "epoch": 0.27099685152589437,
+      "grad_norm": 2.059253692626953,
+      "learning_rate": 9.140534634916847e-06,
+      "loss": 0.0301,
+      "step": 854
+    },
+    {
+      "epoch": 0.27131417804992936,
+      "grad_norm": 0.5214054584503174,
+      "learning_rate": 9.137426126566627e-06,
+      "loss": 0.0298,
+      "step": 855
+    },
+    {
+      "epoch": 0.27163150457396434,
+      "grad_norm": 1.735802173614502,
+      "learning_rate": 9.134312537513188e-06,
+      "loss": 0.0291,
+      "step": 856
+    },
+    {
+      "epoch": 0.27194883109799933,
+      "grad_norm": 1.03411865234375,
+      "learning_rate": 9.131193871579975e-06,
+      "loss": 0.0322,
+      "step": 857
+    },
+    {
+      "epoch": 0.2722661576220344,
+      "grad_norm": 0.4103931784629822,
+      "learning_rate": 9.128070132596668e-06,
+      "loss": 0.0291,
+      "step": 858
+    },
+    {
+      "epoch": 0.27258348414606937,
+      "grad_norm": 0.6772278547286987,
+      "learning_rate": 9.124941324399175e-06,
+      "loss": 0.0306,
+      "step": 859
+    },
+    {
+      "epoch": 0.27290081067010435,
+      "grad_norm": 0.5598214864730835,
+      "learning_rate": 9.121807450829633e-06,
+      "loss": 0.0327,
+      "step": 860
+    },
+    {
+      "epoch": 0.2732181371941394,
+      "grad_norm": 0.48415231704711914,
+      "learning_rate": 9.118668515736393e-06,
+      "loss": 0.0293,
+      "step": 861
+    },
+    {
+      "epoch": 0.2735354637181744,
+      "grad_norm": 0.5242176651954651,
+      "learning_rate": 9.115524522974027e-06,
+      "loss": 0.0281,
+      "step": 862
+    },
+    {
+      "epoch": 0.2738527902422094,
+      "grad_norm": 0.46247944235801697,
+      "learning_rate": 9.112375476403313e-06,
+      "loss": 0.0268,
+      "step": 863
+    },
+    {
+      "epoch": 0.27417011676624437,
+      "grad_norm": 0.47741997241973877,
+      "learning_rate": 9.109221379891238e-06,
+      "loss": 0.0292,
+      "step": 864
+    },
+    {
+      "epoch": 0.2744874432902794,
+      "grad_norm": 1.0339504480361938,
+      "learning_rate": 9.106062237310993e-06,
+      "loss": 0.0325,
+      "step": 865
+    },
+    {
+      "epoch": 0.2748047698143144,
+      "grad_norm": 1.2948225736618042,
+      "learning_rate": 9.102898052541959e-06,
+      "loss": 0.0295,
+      "step": 866
+    },
+    {
+      "epoch": 0.2751220963383494,
+      "grad_norm": 0.5770873427391052,
+      "learning_rate": 9.099728829469713e-06,
+      "loss": 0.0353,
+      "step": 867
+    },
+    {
+      "epoch": 0.27543942286238443,
+      "grad_norm": 0.6643593907356262,
+      "learning_rate": 9.096554571986017e-06,
+      "loss": 0.0318,
+      "step": 868
+    },
+    {
+      "epoch": 0.2757567493864194,
+      "grad_norm": 0.4977007806301117,
+      "learning_rate": 9.093375283988819e-06,
+      "loss": 0.0349,
+      "step": 869
+    },
+    {
+      "epoch": 0.2760740759104544,
+      "grad_norm": 0.4194231927394867,
+      "learning_rate": 9.09019096938224e-06,
+      "loss": 0.0298,
+      "step": 870
+    },
+    {
+      "epoch": 0.27639140243448945,
+      "grad_norm": 0.4401779770851135,
+      "learning_rate": 9.087001632076574e-06,
+      "loss": 0.0328,
+      "step": 871
+    },
+    {
+      "epoch": 0.27670872895852444,
+      "grad_norm": 0.4803525507450104,
+      "learning_rate": 9.083807275988285e-06,
+      "loss": 0.0285,
+      "step": 872
+    },
+    {
+      "epoch": 0.27702605548255943,
+      "grad_norm": 0.8381597399711609,
+      "learning_rate": 9.080607905039998e-06,
+      "loss": 0.032,
+      "step": 873
+    },
+    {
+      "epoch": 0.2773433820065944,
+      "grad_norm": 0.4429357945919037,
+      "learning_rate": 9.0774035231605e-06,
+      "loss": 0.029,
+      "step": 874
+    },
+    {
+      "epoch": 0.27766070853062946,
+      "grad_norm": 0.7169542908668518,
+      "learning_rate": 9.074194134284726e-06,
+      "loss": 0.0371,
+      "step": 875
+    },
+    {
+      "epoch": 0.27797803505466445,
+      "grad_norm": 0.6945329904556274,
+      "learning_rate": 9.070979742353762e-06,
+      "loss": 0.0302,
+      "step": 876
+    },
+    {
+      "epoch": 0.27829536157869944,
+      "grad_norm": 0.7859143614768982,
+      "learning_rate": 9.067760351314838e-06,
+      "loss": 0.0355,
+      "step": 877
+    },
+    {
+      "epoch": 0.2786126881027345,
+      "grad_norm": 0.862166702747345,
+      "learning_rate": 9.064535965121324e-06,
+      "loss": 0.0297,
+      "step": 878
+    },
+    {
+      "epoch": 0.2789300146267695,
+      "grad_norm": 0.48135456442832947,
+      "learning_rate": 9.061306587732722e-06,
+      "loss": 0.0277,
+      "step": 879
+    },
+    {
+      "epoch": 0.27924734115080446,
+      "grad_norm": 0.44525182247161865,
+      "learning_rate": 9.058072223114662e-06,
+      "loss": 0.0313,
+      "step": 880
+    },
+    {
+      "epoch": 0.27956466767483945,
+      "grad_norm": 0.5589117407798767,
+      "learning_rate": 9.054832875238903e-06,
+      "loss": 0.0338,
+      "step": 881
+    },
+    {
+      "epoch": 0.2798819941988745,
+      "grad_norm": 0.4769975543022156,
+      "learning_rate": 9.051588548083316e-06,
+      "loss": 0.0289,
+      "step": 882
+    },
+    {
+      "epoch": 0.2801993207229095,
+      "grad_norm": 0.5153721570968628,
+      "learning_rate": 9.048339245631897e-06,
+      "loss": 0.0339,
+      "step": 883
+    },
+    {
+      "epoch": 0.2805166472469445,
+      "grad_norm": 0.6283595561981201,
+      "learning_rate": 9.045084971874738e-06,
+      "loss": 0.0276,
+      "step": 884
+    },
+    {
+      "epoch": 0.2808339737709795,
+      "grad_norm": 0.7754523754119873,
+      "learning_rate": 9.041825730808047e-06,
+      "loss": 0.0327,
+      "step": 885
+    },
+    {
+      "epoch": 0.2811513002950145,
+      "grad_norm": 0.46336594223976135,
+      "learning_rate": 9.038561526434128e-06,
+      "loss": 0.0307,
+      "step": 886
+    },
+    {
+      "epoch": 0.2814686268190495,
+      "grad_norm": 0.6135295033454895,
+      "learning_rate": 9.035292362761382e-06,
+      "loss": 0.0306,
+      "step": 887
+    },
+    {
+      "epoch": 0.28178595334308454,
+      "grad_norm": 0.644515872001648,
+      "learning_rate": 9.032018243804292e-06,
+      "loss": 0.0334,
+      "step": 888
+    },
+    {
+      "epoch": 0.2821032798671195,
+      "grad_norm": 0.5998374223709106,
+      "learning_rate": 9.028739173583436e-06,
+      "loss": 0.0283,
+      "step": 889
+    },
+    {
+      "epoch": 0.2824206063911545,
+      "grad_norm": 0.5495613217353821,
+      "learning_rate": 9.025455156125466e-06,
+      "loss": 0.0288,
+      "step": 890
+    },
+    {
+      "epoch": 0.2827379329151895,
+      "grad_norm": 0.7235378623008728,
+      "learning_rate": 9.022166195463112e-06,
+      "loss": 0.0325,
+      "step": 891
+    },
+    {
+      "epoch": 0.28305525943922455,
+      "grad_norm": 0.5239971876144409,
+      "learning_rate": 9.01887229563517e-06,
+      "loss": 0.0329,
+      "step": 892
+    },
+    {
+      "epoch": 0.28337258596325954,
+      "grad_norm": 0.4739692211151123,
+      "learning_rate": 9.01557346068651e-06,
+      "loss": 0.0309,
+      "step": 893
+    },
+    {
+      "epoch": 0.2836899124872945,
+      "grad_norm": 0.5890957117080688,
+      "learning_rate": 9.012269694668053e-06,
+      "loss": 0.0297,
+      "step": 894
+    },
+    {
+      "epoch": 0.28400723901132957,
+      "grad_norm": 0.5504534244537354,
+      "learning_rate": 9.008961001636777e-06,
+      "loss": 0.0289,
+      "step": 895
+    },
+    {
+      "epoch": 0.28432456553536456,
+      "grad_norm": 0.4708574712276459,
+      "learning_rate": 9.005647385655718e-06,
+      "loss": 0.0304,
+      "step": 896
+    },
+    {
+      "epoch": 0.28464189205939955,
+      "grad_norm": 0.5596920847892761,
+      "learning_rate": 9.002328850793946e-06,
+      "loss": 0.034,
+      "step": 897
+    },
+    {
+      "epoch": 0.28495921858343454,
+      "grad_norm": 0.45385676622390747,
+      "learning_rate": 8.99900540112658e-06,
+      "loss": 0.029,
+      "step": 898
+    },
+    {
+      "epoch": 0.2852765451074696,
+      "grad_norm": 0.6831643581390381,
+      "learning_rate": 8.99567704073477e-06,
+      "loss": 0.0299,
+      "step": 899
+    },
+    {
+      "epoch": 0.28559387163150457,
+      "grad_norm": 0.926430881023407,
+      "learning_rate": 8.992343773705698e-06,
+      "loss": 0.0284,
+      "step": 900
+    },
+    {
+      "epoch": 0.28591119815553956,
+      "grad_norm": 0.5740981698036194,
+      "learning_rate": 8.98900560413257e-06,
+      "loss": 0.031,
+      "step": 901
+    },
+    {
+      "epoch": 0.2862285246795746,
+      "grad_norm": 0.6546286344528198,
+      "learning_rate": 8.985662536114614e-06,
+      "loss": 0.0289,
+      "step": 902
+    },
+    {
+      "epoch": 0.2865458512036096,
+      "grad_norm": 0.5189806818962097,
+      "learning_rate": 8.98231457375707e-06,
+      "loss": 0.0338,
+      "step": 903
+    },
+    {
+      "epoch": 0.2868631777276446,
+      "grad_norm": 0.5386642217636108,
+      "learning_rate": 8.978961721171193e-06,
+      "loss": 0.0288,
+      "step": 904
+    },
+    {
+      "epoch": 0.2871805042516796,
+      "grad_norm": 0.3801208436489105,
+      "learning_rate": 8.97560398247424e-06,
+      "loss": 0.0312,
+      "step": 905
+    },
+    {
+      "epoch": 0.2874978307757146,
+      "grad_norm": 0.40569886565208435,
+      "learning_rate": 8.972241361789467e-06,
+      "loss": 0.0292,
+      "step": 906
+    },
+    {
+      "epoch": 0.2878151572997496,
+      "grad_norm": 0.7458774447441101,
+      "learning_rate": 8.968873863246128e-06,
+      "loss": 0.0368,
+      "step": 907
+    },
+    {
+      "epoch": 0.2881324838237846,
+      "grad_norm": 0.9061804413795471,
+      "learning_rate": 8.965501490979467e-06,
+      "loss": 0.0282,
+      "step": 908
+    },
+    {
+      "epoch": 0.28844981034781964,
+      "grad_norm": 1.3380414247512817,
+      "learning_rate": 8.962124249130709e-06,
+      "loss": 0.0341,
+      "step": 909
+    },
+    {
+      "epoch": 0.2887671368718546,
+      "grad_norm": 0.4201684892177582,
+      "learning_rate": 8.95874214184706e-06,
+      "loss": 0.0318,
+      "step": 910
+    },
+    {
+      "epoch": 0.2890844633958896,
+      "grad_norm": 0.39467689394950867,
+      "learning_rate": 8.955355173281709e-06,
+      "loss": 0.0301,
+      "step": 911
+    },
+    {
+      "epoch": 0.28940178991992466,
+      "grad_norm": 0.49978744983673096,
+      "learning_rate": 8.951963347593797e-06,
+      "loss": 0.0354,
+      "step": 912
+    },
+    {
+      "epoch": 0.28971911644395965,
+      "grad_norm": 0.3985583484172821,
+      "learning_rate": 8.948566668948449e-06,
+      "loss": 0.0292,
+      "step": 913
+    },
+    {
+      "epoch": 0.29003644296799463,
+      "grad_norm": 0.4919849932193756,
+      "learning_rate": 8.945165141516733e-06,
+      "loss": 0.0316,
+      "step": 914
+    },
+    {
+      "epoch": 0.2903537694920296,
+      "grad_norm": 0.5306928157806396,
+      "learning_rate": 8.941758769475686e-06,
+      "loss": 0.0304,
+      "step": 915
+    },
+    {
+      "epoch": 0.29067109601606467,
+      "grad_norm": 0.494953453540802,
+      "learning_rate": 8.938347557008282e-06,
+      "loss": 0.0273,
+      "step": 916
+    },
+    {
+      "epoch": 0.29098842254009966,
+      "grad_norm": 0.4031258523464203,
+      "learning_rate": 8.934931508303446e-06,
+      "loss": 0.0302,
+      "step": 917
+    },
+    {
+      "epoch": 0.29130574906413464,
+      "grad_norm": 0.37998414039611816,
+      "learning_rate": 8.931510627556038e-06,
+      "loss": 0.0272,
+      "step": 918
+    },
+    {
+      "epoch": 0.2916230755881697,
+      "grad_norm": 0.5286462306976318,
+      "learning_rate": 8.92808491896685e-06,
+      "loss": 0.0337,
+      "step": 919
+    },
+    {
+      "epoch": 0.2919404021122047,
+      "grad_norm": 0.46686577796936035,
+      "learning_rate": 8.924654386742613e-06,
+      "loss": 0.0295,
+      "step": 920
+    },
+    {
+      "epoch": 0.29225772863623967,
+      "grad_norm": 0.535905122756958,
+      "learning_rate": 8.921219035095969e-06,
+      "loss": 0.0328,
+      "step": 921
+    },
+    {
+      "epoch": 0.2925750551602747,
+      "grad_norm": 0.4379343092441559,
+      "learning_rate": 8.917778868245486e-06,
+      "loss": 0.0286,
+      "step": 922
+    },
+    {
+      "epoch": 0.2928923816843097,
+      "grad_norm": 0.8484111428260803,
+      "learning_rate": 8.91433389041564e-06,
+      "loss": 0.0294,
+      "step": 923
+    },
+    {
+      "epoch": 0.2932097082083447,
+      "grad_norm": 0.4197337031364441,
+      "learning_rate": 8.910884105836817e-06,
+      "loss": 0.0295,
+      "step": 924
+    },
+    {
+      "epoch": 0.2935270347323797,
+      "grad_norm": 0.36424049735069275,
+      "learning_rate": 8.907429518745308e-06,
+      "loss": 0.0272,
+      "step": 925
+    },
+    {
+      "epoch": 0.2938443612564147,
+      "grad_norm": 0.5069578289985657,
+      "learning_rate": 8.903970133383297e-06,
+      "loss": 0.0276,
+      "step": 926
+    },
+    {
+      "epoch": 0.2941616877804497,
+      "grad_norm": 0.43945327401161194,
+      "learning_rate": 8.900505953998864e-06,
+      "loss": 0.0294,
+      "step": 927
+    },
+    {
+      "epoch": 0.2944790143044847,
+      "grad_norm": 1.010656714439392,
+      "learning_rate": 8.897036984845973e-06,
+      "loss": 0.025,
+      "step": 928
+    },
+    {
+      "epoch": 0.29479634082851974,
+      "grad_norm": 0.41510307788848877,
+      "learning_rate": 8.89356323018447e-06,
+      "loss": 0.0288,
+      "step": 929
+    },
+    {
+      "epoch": 0.29511366735255473,
+      "grad_norm": 0.371634304523468,
+      "learning_rate": 8.890084694280084e-06,
+      "loss": 0.0257,
+      "step": 930
+    },
+    {
+      "epoch": 0.2954309938765897,
+      "grad_norm": 0.44098961353302,
+      "learning_rate": 8.886601381404403e-06,
+      "loss": 0.0305,
+      "step": 931
+    },
+    {
+      "epoch": 0.2957483204006247,
+      "grad_norm": 0.44313350319862366,
+      "learning_rate": 8.883113295834893e-06,
+      "loss": 0.0316,
+      "step": 932
+    },
+    {
+      "epoch": 0.29606564692465975,
+      "grad_norm": 0.4331350326538086,
+      "learning_rate": 8.879620441854873e-06,
+      "loss": 0.0245,
+      "step": 933
+    },
+    {
+      "epoch": 0.29638297344869474,
+      "grad_norm": 1.0557185411453247,
+      "learning_rate": 8.876122823753522e-06,
+      "loss": 0.0286,
+      "step": 934
+    },
+    {
+      "epoch": 0.29670029997272973,
+      "grad_norm": 2.303699016571045,
+      "learning_rate": 8.872620445825868e-06,
+      "loss": 0.0281,
+      "step": 935
+    },
+    {
+      "epoch": 0.2970176264967648,
+      "grad_norm": 0.48086345195770264,
+      "learning_rate": 8.869113312372781e-06,
+      "loss": 0.0319,
+      "step": 936
+    },
+    {
+      "epoch": 0.29733495302079976,
+      "grad_norm": 0.4232422113418579,
+      "learning_rate": 8.865601427700975e-06,
+      "loss": 0.026,
+      "step": 937
+    },
+    {
+      "epoch": 0.29765227954483475,
+      "grad_norm": 0.5579431653022766,
+      "learning_rate": 8.862084796122998e-06,
+      "loss": 0.0308,
+      "step": 938
+    },
+    {
+      "epoch": 0.2979696060688698,
+      "grad_norm": 0.7712277173995972,
+      "learning_rate": 8.858563421957226e-06,
+      "loss": 0.0278,
+      "step": 939
+    },
+    {
+      "epoch": 0.2982869325929048,
+      "grad_norm": 0.5227288007736206,
+      "learning_rate": 8.855037309527854e-06,
+      "loss": 0.0318,
+      "step": 940
+    },
+    {
+      "epoch": 0.2986042591169398,
+      "grad_norm": 0.4694115221500397,
+      "learning_rate": 8.851506463164906e-06,
+      "loss": 0.0276,
+      "step": 941
+    },
+    {
+      "epoch": 0.29892158564097476,
+      "grad_norm": 0.5424268245697021,
+      "learning_rate": 8.847970887204212e-06,
+      "loss": 0.0282,
+      "step": 942
+    },
+    {
+      "epoch": 0.2992389121650098,
+      "grad_norm": 0.4243178367614746,
+      "learning_rate": 8.84443058598741e-06,
+      "loss": 0.0302,
+      "step": 943
+    },
+    {
+      "epoch": 0.2995562386890448,
+      "grad_norm": 0.7143673896789551,
+      "learning_rate": 8.840885563861941e-06,
+      "loss": 0.029,
+      "step": 944
+    },
+    {
+      "epoch": 0.2998735652130798,
+      "grad_norm": 0.4195621609687805,
+      "learning_rate": 8.837335825181049e-06,
+      "loss": 0.0272,
+      "step": 945
+    },
+    {
+      "epoch": 0.30019089173711483,
+      "grad_norm": 0.478171169757843,
+      "learning_rate": 8.833781374303757e-06,
+      "loss": 0.0268,
+      "step": 946
+    },
+    {
+      "epoch": 0.3005082182611498,
+      "grad_norm": 0.48147091269493103,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.028,
+      "step": 947
+    },
+    {
+      "epoch": 0.3008255447851848,
+      "grad_norm": 0.48147860169410706,
+      "learning_rate": 8.826658353425044e-06,
+      "loss": 0.0313,
+      "step": 948
+    },
+    {
+      "epoch": 0.30114287130921985,
+      "grad_norm": 1.4652529954910278,
+      "learning_rate": 8.823089792170595e-06,
+      "loss": 0.0306,
+      "step": 949
+    },
+    {
+      "epoch": 0.30146019783325484,
+      "grad_norm": 0.5120874643325806,
+      "learning_rate": 8.819516536213683e-06,
+      "loss": 0.0262,
+      "step": 950
+    },
+    {
+      "epoch": 0.30177752435728983,
+      "grad_norm": 0.5322923064231873,
+      "learning_rate": 8.815938589942224e-06,
+      "loss": 0.0289,
+      "step": 951
+    },
+    {
+      "epoch": 0.3020948508813248,
+      "grad_norm": 0.9768365025520325,
+      "learning_rate": 8.812355957749883e-06,
+      "loss": 0.0292,
+      "step": 952
+    },
+    {
+      "epoch": 0.30241217740535986,
+      "grad_norm": 0.43848299980163574,
+      "learning_rate": 8.808768644036086e-06,
+      "loss": 0.0293,
+      "step": 953
+    },
+    {
+      "epoch": 0.30272950392939485,
+      "grad_norm": 0.40072569251060486,
+      "learning_rate": 8.805176653206004e-06,
+      "loss": 0.0245,
+      "step": 954
+    },
+    {
+      "epoch": 0.30304683045342984,
+      "grad_norm": 1.1225285530090332,
+      "learning_rate": 8.801579989670554e-06,
+      "loss": 0.0314,
+      "step": 955
+    },
+    {
+      "epoch": 0.3033641569774649,
+      "grad_norm": 0.5232235789299011,
+      "learning_rate": 8.797978657846391e-06,
+      "loss": 0.027,
+      "step": 956
+    },
+    {
+      "epoch": 0.30368148350149987,
+      "grad_norm": 0.3462230861186981,
+      "learning_rate": 8.7943726621559e-06,
+      "loss": 0.0213,
+      "step": 957
+    },
+    {
+      "epoch": 0.30399881002553486,
+      "grad_norm": 0.3628953695297241,
+      "learning_rate": 8.790762007027199e-06,
+      "loss": 0.0305,
+      "step": 958
+    },
+    {
+      "epoch": 0.30431613654956985,
+      "grad_norm": 0.7281985282897949,
+      "learning_rate": 8.787146696894118e-06,
+      "loss": 0.029,
+      "step": 959
+    },
+    {
+      "epoch": 0.3046334630736049,
+      "grad_norm": 0.45109686255455017,
+      "learning_rate": 8.78352673619621e-06,
+      "loss": 0.0282,
+      "step": 960
+    },
+    {
+      "epoch": 0.3049507895976399,
+      "grad_norm": 0.4393883943557739,
+      "learning_rate": 8.77990212937874e-06,
+      "loss": 0.0296,
+      "step": 961
+    },
+    {
+      "epoch": 0.30526811612167487,
+      "grad_norm": 0.5017929673194885,
+      "learning_rate": 8.776272880892675e-06,
+      "loss": 0.0289,
+      "step": 962
+    },
+    {
+      "epoch": 0.3055854426457099,
+      "grad_norm": 2.048901319503784,
+      "learning_rate": 8.772638995194682e-06,
+      "loss": 0.0289,
+      "step": 963
+    },
+    {
+      "epoch": 0.3059027691697449,
+      "grad_norm": 0.41658857464790344,
+      "learning_rate": 8.769000476747126e-06,
+      "loss": 0.0269,
+      "step": 964
+    },
+    {
+      "epoch": 0.3062200956937799,
+      "grad_norm": 0.44559794664382935,
+      "learning_rate": 8.765357330018056e-06,
+      "loss": 0.0334,
+      "step": 965
+    },
+    {
+      "epoch": 0.30653742221781494,
+      "grad_norm": 0.41546422243118286,
+      "learning_rate": 8.761709559481207e-06,
+      "loss": 0.0246,
+      "step": 966
+    },
+    {
+      "epoch": 0.3068547487418499,
+      "grad_norm": 0.4419938027858734,
+      "learning_rate": 8.758057169615992e-06,
+      "loss": 0.0319,
+      "step": 967
+    },
+    {
+      "epoch": 0.3071720752658849,
+      "grad_norm": 1.003542184829712,
+      "learning_rate": 8.754400164907496e-06,
+      "loss": 0.0309,
+      "step": 968
+    },
+    {
+      "epoch": 0.3074894017899199,
+      "grad_norm": 0.45393243432044983,
+      "learning_rate": 8.750738549846474e-06,
+      "loss": 0.0263,
+      "step": 969
+    },
+    {
+      "epoch": 0.30780672831395495,
+      "grad_norm": 0.41879016160964966,
+      "learning_rate": 8.747072328929336e-06,
+      "loss": 0.0308,
+      "step": 970
+    },
+    {
+      "epoch": 0.30812405483798994,
+      "grad_norm": 0.4686222970485687,
+      "learning_rate": 8.74340150665815e-06,
+      "loss": 0.0302,
+      "step": 971
+    },
+    {
+      "epoch": 0.3084413813620249,
+      "grad_norm": 0.4682992696762085,
+      "learning_rate": 8.739726087540645e-06,
+      "loss": 0.034,
+      "step": 972
+    },
+    {
+      "epoch": 0.30875870788605997,
+      "grad_norm": 0.5936564803123474,
+      "learning_rate": 8.736046076090176e-06,
+      "loss": 0.0279,
+      "step": 973
+    },
+    {
+      "epoch": 0.30907603441009496,
+      "grad_norm": 0.40318334102630615,
+      "learning_rate": 8.732361476825752e-06,
+      "loss": 0.0281,
+      "step": 974
+    },
+    {
+      "epoch": 0.30939336093412995,
+      "grad_norm": 0.46161365509033203,
+      "learning_rate": 8.728672294272009e-06,
+      "loss": 0.0298,
+      "step": 975
+    },
+    {
+      "epoch": 0.30971068745816493,
+      "grad_norm": 0.3896993398666382,
+      "learning_rate": 8.724978532959216e-06,
+      "loss": 0.0276,
+      "step": 976
+    },
+    {
+      "epoch": 0.3100280139822,
+      "grad_norm": 0.37892207503318787,
+      "learning_rate": 8.721280197423259e-06,
+      "loss": 0.029,
+      "step": 977
+    },
+    {
+      "epoch": 0.31034534050623497,
+      "grad_norm": 0.395465612411499,
+      "learning_rate": 8.717577292205646e-06,
+      "loss": 0.0266,
+      "step": 978
+    },
+    {
+      "epoch": 0.31066266703026996,
+      "grad_norm": 0.47348758578300476,
+      "learning_rate": 8.713869821853492e-06,
+      "loss": 0.0306,
+      "step": 979
+    },
+    {
+      "epoch": 0.310979993554305,
+      "grad_norm": 0.34926560521125793,
+      "learning_rate": 8.710157790919522e-06,
+      "loss": 0.0237,
+      "step": 980
+    },
+    {
+      "epoch": 0.31129732007834,
+      "grad_norm": 0.8807845711708069,
+      "learning_rate": 8.70644120396206e-06,
+      "loss": 0.0301,
+      "step": 981
+    },
+    {
+      "epoch": 0.311614646602375,
+      "grad_norm": 0.5081422924995422,
+      "learning_rate": 8.702720065545024e-06,
+      "loss": 0.0314,
+      "step": 982
+    },
+    {
+      "epoch": 0.31193197312641,
+      "grad_norm": 0.514532208442688,
+      "learning_rate": 8.698994380237921e-06,
+      "loss": 0.0265,
+      "step": 983
+    },
+    {
+      "epoch": 0.312249299650445,
+      "grad_norm": 0.3232363164424896,
+      "learning_rate": 8.695264152615843e-06,
+      "loss": 0.0245,
+      "step": 984
+    },
+    {
+      "epoch": 0.31256662617448,
+      "grad_norm": 0.37816470861434937,
+      "learning_rate": 8.69152938725946e-06,
+      "loss": 0.0272,
+      "step": 985
+    },
+    {
+      "epoch": 0.312883952698515,
+      "grad_norm": 0.44968974590301514,
+      "learning_rate": 8.687790088755008e-06,
+      "loss": 0.0253,
+      "step": 986
+    },
+    {
+      "epoch": 0.31320127922255003,
+      "grad_norm": 0.4260583817958832,
+      "learning_rate": 8.6840462616943e-06,
+      "loss": 0.0306,
+      "step": 987
+    },
+    {
+      "epoch": 0.313518605746585,
+      "grad_norm": 0.4838992953300476,
+      "learning_rate": 8.680297910674704e-06,
+      "loss": 0.028,
+      "step": 988
+    },
+    {
+      "epoch": 0.31383593227062,
+      "grad_norm": 0.44544926285743713,
+      "learning_rate": 8.676545040299145e-06,
+      "loss": 0.0306,
+      "step": 989
+    },
+    {
+      "epoch": 0.31415325879465505,
+      "grad_norm": 0.4807493984699249,
+      "learning_rate": 8.672787655176094e-06,
+      "loss": 0.0295,
+      "step": 990
+    },
+    {
+      "epoch": 0.31447058531869004,
+      "grad_norm": 0.7959880232810974,
+      "learning_rate": 8.669025759919572e-06,
+      "loss": 0.0263,
+      "step": 991
+    },
+    {
+      "epoch": 0.31478791184272503,
+      "grad_norm": 1.1461362838745117,
+      "learning_rate": 8.665259359149132e-06,
+      "loss": 0.0278,
+      "step": 992
+    },
+    {
+      "epoch": 0.31510523836676,
+      "grad_norm": 0.43732452392578125,
+      "learning_rate": 8.661488457489868e-06,
+      "loss": 0.0275,
+      "step": 993
+    },
+    {
+      "epoch": 0.31542256489079507,
+      "grad_norm": 0.5605583190917969,
+      "learning_rate": 8.657713059572392e-06,
+      "loss": 0.0284,
+      "step": 994
+    },
+    {
+      "epoch": 0.31573989141483005,
+      "grad_norm": 0.4159834086894989,
+      "learning_rate": 8.653933170032842e-06,
+      "loss": 0.0288,
+      "step": 995
+    },
+    {
+      "epoch": 0.31605721793886504,
+      "grad_norm": 0.5480255484580994,
+      "learning_rate": 8.650148793512874e-06,
+      "loss": 0.0279,
+      "step": 996
+    },
+    {
+      "epoch": 0.3163745444629001,
+      "grad_norm": 0.3648799657821655,
+      "learning_rate": 8.646359934659648e-06,
+      "loss": 0.0248,
+      "step": 997
+    },
+    {
+      "epoch": 0.3166918709869351,
+      "grad_norm": 0.4637749195098877,
+      "learning_rate": 8.642566598125832e-06,
+      "loss": 0.0269,
+      "step": 998
+    },
+    {
+      "epoch": 0.31700919751097006,
+      "grad_norm": 0.4913865029811859,
+      "learning_rate": 8.638768788569593e-06,
+      "loss": 0.0269,
+      "step": 999
+    },
+    {
+      "epoch": 0.3173265240350051,
+      "grad_norm": 0.489591509103775,
+      "learning_rate": 8.63496651065459e-06,
+      "loss": 0.03,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3176438505590401,
+      "grad_norm": 0.3750690221786499,
+      "learning_rate": 8.631159769049965e-06,
+      "loss": 0.024,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3179611770830751,
+      "grad_norm": 0.41776952147483826,
+      "learning_rate": 8.62734856843035e-06,
+      "loss": 0.029,
+      "step": 1002
+    },
+    {
+      "epoch": 0.3182785036071101,
+      "grad_norm": 0.4538407623767853,
+      "learning_rate": 8.623532913475847e-06,
+      "loss": 0.0334,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3185958301311451,
+      "grad_norm": 0.39228740334510803,
+      "learning_rate": 8.619712808872025e-06,
+      "loss": 0.0285,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3189131566551801,
+      "grad_norm": 0.636155903339386,
+      "learning_rate": 8.615888259309923e-06,
+      "loss": 0.032,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3192304831792151,
+      "grad_norm": 0.4503800570964813,
+      "learning_rate": 8.612059269486036e-06,
+      "loss": 0.0249,
+      "step": 1006
+    },
+    {
+      "epoch": 0.31954780970325014,
+      "grad_norm": 0.4255145192146301,
+      "learning_rate": 8.608225844102312e-06,
+      "loss": 0.0306,
+      "step": 1007
+    },
+    {
+      "epoch": 0.31986513622728513,
+      "grad_norm": 0.42445096373558044,
+      "learning_rate": 8.604387987866144e-06,
+      "loss": 0.0263,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3201824627513201,
+      "grad_norm": 0.3951943516731262,
+      "learning_rate": 8.600545705490369e-06,
+      "loss": 0.0335,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3204997892753551,
+      "grad_norm": 0.3890993595123291,
+      "learning_rate": 8.596699001693257e-06,
+      "loss": 0.0277,
+      "step": 1010
+    },
+    {
+      "epoch": 0.32081711579939015,
+      "grad_norm": 1.4389110803604126,
+      "learning_rate": 8.592847881198508e-06,
+      "loss": 0.0263,
+      "step": 1011
+    },
+    {
+      "epoch": 0.32113444232342514,
+      "grad_norm": 0.562517523765564,
+      "learning_rate": 8.588992348735245e-06,
+      "loss": 0.0248,
+      "step": 1012
+    },
+    {
+      "epoch": 0.32145176884746013,
+      "grad_norm": 0.5117233395576477,
+      "learning_rate": 8.585132409038013e-06,
+      "loss": 0.0303,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3217690953714952,
+      "grad_norm": 0.47211968898773193,
+      "learning_rate": 8.581268066846762e-06,
+      "loss": 0.0338,
+      "step": 1014
+    },
+    {
+      "epoch": 0.32208642189553016,
+      "grad_norm": 0.4438319802284241,
+      "learning_rate": 8.577399326906853e-06,
+      "loss": 0.0312,
+      "step": 1015
+    },
+    {
+      "epoch": 0.32240374841956515,
+      "grad_norm": 0.8474715352058411,
+      "learning_rate": 8.573526193969047e-06,
+      "loss": 0.0302,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3227210749436002,
+      "grad_norm": 0.4471942186355591,
+      "learning_rate": 8.569648672789496e-06,
+      "loss": 0.0277,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3230384014676352,
+      "grad_norm": 0.3686079680919647,
+      "learning_rate": 8.565766768129747e-06,
+      "loss": 0.0241,
+      "step": 1018
+    },
+    {
+      "epoch": 0.32335572799167017,
+      "grad_norm": 0.5485999584197998,
+      "learning_rate": 8.561880484756726e-06,
+      "loss": 0.0303,
+      "step": 1019
+    },
+    {
+      "epoch": 0.32367305451570516,
+      "grad_norm": 0.43232259154319763,
+      "learning_rate": 8.557989827442734e-06,
+      "loss": 0.0283,
+      "step": 1020
+    },
+    {
+      "epoch": 0.3239903810397402,
+      "grad_norm": 0.44287875294685364,
+      "learning_rate": 8.554094800965446e-06,
+      "loss": 0.0283,
+      "step": 1021
+    },
+    {
+      "epoch": 0.3243077075637752,
+      "grad_norm": 0.4500407576560974,
+      "learning_rate": 8.550195410107903e-06,
+      "loss": 0.026,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3246250340878102,
+      "grad_norm": 0.5405159592628479,
+      "learning_rate": 8.546291659658504e-06,
+      "loss": 0.0343,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3249423606118452,
+      "grad_norm": 0.535243570804596,
+      "learning_rate": 8.542383554411e-06,
+      "loss": 0.0286,
+      "step": 1024
+    },
+    {
+      "epoch": 0.3252596871358802,
+      "grad_norm": 0.5369758009910583,
+      "learning_rate": 8.538471099164494e-06,
+      "loss": 0.0311,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3255770136599152,
+      "grad_norm": 0.4696981608867645,
+      "learning_rate": 8.534554298723425e-06,
+      "loss": 0.0305,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3258943401839502,
+      "grad_norm": 0.41171807050704956,
+      "learning_rate": 8.530633157897573e-06,
+      "loss": 0.0278,
+      "step": 1027
+    },
+    {
+      "epoch": 0.32621166670798524,
+      "grad_norm": 0.4900743067264557,
+      "learning_rate": 8.526707681502045e-06,
+      "loss": 0.0303,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3265289932320202,
+      "grad_norm": 0.41155490279197693,
+      "learning_rate": 8.522777874357273e-06,
+      "loss": 0.0263,
+      "step": 1029
+    },
+    {
+      "epoch": 0.3268463197560552,
+      "grad_norm": 0.3709670305252075,
+      "learning_rate": 8.518843741289007e-06,
+      "loss": 0.0255,
+      "step": 1030
+    },
+    {
+      "epoch": 0.32716364628009026,
+      "grad_norm": 0.564120352268219,
+      "learning_rate": 8.51490528712831e-06,
+      "loss": 0.0245,
+      "step": 1031
+    },
+    {
+      "epoch": 0.32748097280412525,
+      "grad_norm": 0.47435125708580017,
+      "learning_rate": 8.510962516711548e-06,
+      "loss": 0.0229,
+      "step": 1032
+    },
+    {
+      "epoch": 0.32779829932816024,
+      "grad_norm": 0.4333440959453583,
+      "learning_rate": 8.50701543488039e-06,
+      "loss": 0.028,
+      "step": 1033
+    },
+    {
+      "epoch": 0.3281156258521953,
+      "grad_norm": 0.41801002621650696,
+      "learning_rate": 8.503064046481802e-06,
+      "loss": 0.0261,
+      "step": 1034
+    },
+    {
+      "epoch": 0.32843295237623027,
+      "grad_norm": 0.46156740188598633,
+      "learning_rate": 8.499108356368033e-06,
+      "loss": 0.0281,
+      "step": 1035
+    },
+    {
+      "epoch": 0.32875027890026526,
+      "grad_norm": 0.4081984758377075,
+      "learning_rate": 8.495148369396617e-06,
+      "loss": 0.0273,
+      "step": 1036
+    },
+    {
+      "epoch": 0.32906760542430025,
+      "grad_norm": 0.44961923360824585,
+      "learning_rate": 8.491184090430365e-06,
+      "loss": 0.0245,
+      "step": 1037
+    },
+    {
+      "epoch": 0.3293849319483353,
+      "grad_norm": 0.4167129397392273,
+      "learning_rate": 8.487215524337357e-06,
+      "loss": 0.0276,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3297022584723703,
+      "grad_norm": 0.5280731916427612,
+      "learning_rate": 8.483242675990939e-06,
+      "loss": 0.0291,
+      "step": 1039
+    },
+    {
+      "epoch": 0.33001958499640527,
+      "grad_norm": 0.4450177252292633,
+      "learning_rate": 8.479265550269714e-06,
+      "loss": 0.0253,
+      "step": 1040
+    },
+    {
+      "epoch": 0.3303369115204403,
+      "grad_norm": 0.3854939043521881,
+      "learning_rate": 8.475284152057542e-06,
+      "loss": 0.0258,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3306542380444753,
+      "grad_norm": 0.6903836727142334,
+      "learning_rate": 8.47129848624352e-06,
+      "loss": 0.0308,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3309715645685103,
+      "grad_norm": 0.387279212474823,
+      "learning_rate": 8.467308557721996e-06,
+      "loss": 0.0263,
+      "step": 1043
+    },
+    {
+      "epoch": 0.3312888910925453,
+      "grad_norm": 0.6145289540290833,
+      "learning_rate": 8.463314371392547e-06,
+      "loss": 0.0277,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3316062176165803,
+      "grad_norm": 0.5422887802124023,
+      "learning_rate": 8.45931593215998e-06,
+      "loss": 0.0262,
+      "step": 1045
+    },
+    {
+      "epoch": 0.3319235441406153,
+      "grad_norm": 0.4166538119316101,
+      "learning_rate": 8.455313244934324e-06,
+      "loss": 0.0259,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3322408706646503,
+      "grad_norm": 0.5057457685470581,
+      "learning_rate": 8.451306314630825e-06,
+      "loss": 0.0271,
+      "step": 1047
+    },
+    {
+      "epoch": 0.33255819718868534,
+      "grad_norm": 0.45026496052742004,
+      "learning_rate": 8.447295146169939e-06,
+      "loss": 0.0289,
+      "step": 1048
+    },
+    {
+      "epoch": 0.33287552371272033,
+      "grad_norm": 0.5630201101303101,
+      "learning_rate": 8.443279744477324e-06,
+      "loss": 0.0265,
+      "step": 1049
+    },
+    {
+      "epoch": 0.3331928502367553,
+      "grad_norm": 0.4210197329521179,
+      "learning_rate": 8.439260114483843e-06,
+      "loss": 0.0252,
+      "step": 1050
+    },
+    {
+      "epoch": 0.33351017676079037,
+      "grad_norm": 0.41802313923835754,
+      "learning_rate": 8.435236261125545e-06,
+      "loss": 0.0242,
+      "step": 1051
+    },
+    {
+      "epoch": 0.33382750328482536,
+      "grad_norm": 0.506890594959259,
+      "learning_rate": 8.43120818934367e-06,
+      "loss": 0.026,
+      "step": 1052
+    },
+    {
+      "epoch": 0.33414482980886034,
+      "grad_norm": 0.5125287175178528,
+      "learning_rate": 8.42717590408463e-06,
+      "loss": 0.028,
+      "step": 1053
+    },
+    {
+      "epoch": 0.33446215633289533,
+      "grad_norm": 0.4136829972267151,
+      "learning_rate": 8.423139410300021e-06,
+      "loss": 0.0266,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3347794828569304,
+      "grad_norm": 0.9653528332710266,
+      "learning_rate": 8.4190987129466e-06,
+      "loss": 0.024,
+      "step": 1055
+    },
+    {
+      "epoch": 0.33509680938096537,
+      "grad_norm": 0.34526655077934265,
+      "learning_rate": 8.415053816986293e-06,
+      "loss": 0.0258,
+      "step": 1056
+    },
+    {
+      "epoch": 0.33541413590500035,
+      "grad_norm": 0.6338776350021362,
+      "learning_rate": 8.411004727386175e-06,
+      "loss": 0.0242,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3357314624290354,
+      "grad_norm": 0.6697638630867004,
+      "learning_rate": 8.40695144911847e-06,
+      "loss": 0.0238,
+      "step": 1058
+    },
+    {
+      "epoch": 0.3360487889530704,
+      "grad_norm": 0.3901837468147278,
+      "learning_rate": 8.402893987160553e-06,
+      "loss": 0.0243,
+      "step": 1059
+    },
+    {
+      "epoch": 0.3363661154771054,
+      "grad_norm": 0.44057050347328186,
+      "learning_rate": 8.39883234649493e-06,
+      "loss": 0.0276,
+      "step": 1060
+    },
+    {
+      "epoch": 0.33668344200114037,
+      "grad_norm": 0.4518550932407379,
+      "learning_rate": 8.394766532109243e-06,
+      "loss": 0.0309,
+      "step": 1061
+    },
+    {
+      "epoch": 0.3370007685251754,
+      "grad_norm": 0.48535799980163574,
+      "learning_rate": 8.390696548996256e-06,
+      "loss": 0.0249,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3373180950492104,
+      "grad_norm": 0.4695781171321869,
+      "learning_rate": 8.386622402153851e-06,
+      "loss": 0.0234,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3376354215732454,
+      "grad_norm": 0.3866165578365326,
+      "learning_rate": 8.382544096585028e-06,
+      "loss": 0.0282,
+      "step": 1064
+    },
+    {
+      "epoch": 0.33795274809728043,
+      "grad_norm": 0.4235221743583679,
+      "learning_rate": 8.378461637297887e-06,
+      "loss": 0.0226,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3382700746213154,
+      "grad_norm": 0.458048015832901,
+      "learning_rate": 8.37437502930564e-06,
+      "loss": 0.0306,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3385874011453504,
+      "grad_norm": 0.46734440326690674,
+      "learning_rate": 8.370284277626576e-06,
+      "loss": 0.0222,
+      "step": 1067
+    },
+    {
+      "epoch": 0.33890472766938545,
+      "grad_norm": 0.6317221522331238,
+      "learning_rate": 8.366189387284093e-06,
+      "loss": 0.0251,
+      "step": 1068
+    },
+    {
+      "epoch": 0.33922205419342044,
+      "grad_norm": 0.36008667945861816,
+      "learning_rate": 8.362090363306655e-06,
+      "loss": 0.0241,
+      "step": 1069
+    },
+    {
+      "epoch": 0.33953938071745543,
+      "grad_norm": 0.47830745577812195,
+      "learning_rate": 8.357987210727807e-06,
+      "loss": 0.0288,
+      "step": 1070
+    },
+    {
+      "epoch": 0.3398567072414904,
+      "grad_norm": 0.5054096579551697,
+      "learning_rate": 8.353879934586171e-06,
+      "loss": 0.0234,
+      "step": 1071
+    },
+    {
+      "epoch": 0.34017403376552546,
+      "grad_norm": 0.44308191537857056,
+      "learning_rate": 8.34976853992542e-06,
+      "loss": 0.0294,
+      "step": 1072
+    },
+    {
+      "epoch": 0.34049136028956045,
+      "grad_norm": 0.3868498206138611,
+      "learning_rate": 8.345653031794292e-06,
+      "loss": 0.0251,
+      "step": 1073
+    },
+    {
+      "epoch": 0.34080868681359544,
+      "grad_norm": 0.4330494701862335,
+      "learning_rate": 8.341533415246578e-06,
+      "loss": 0.0283,
+      "step": 1074
+    },
+    {
+      "epoch": 0.3411260133376305,
+      "grad_norm": 0.4157262444496155,
+      "learning_rate": 8.337409695341107e-06,
+      "loss": 0.0308,
+      "step": 1075
+    },
+    {
+      "epoch": 0.3414433398616655,
+      "grad_norm": 0.38565635681152344,
+      "learning_rate": 8.333281877141758e-06,
+      "loss": 0.0243,
+      "step": 1076
+    },
+    {
+      "epoch": 0.34176066638570046,
+      "grad_norm": 0.48059508204460144,
+      "learning_rate": 8.329149965717429e-06,
+      "loss": 0.0251,
+      "step": 1077
+    },
+    {
+      "epoch": 0.34207799290973545,
+      "grad_norm": 0.4065714478492737,
+      "learning_rate": 8.325013966142054e-06,
+      "loss": 0.0231,
+      "step": 1078
+    },
+    {
+      "epoch": 0.3423953194337705,
+      "grad_norm": 0.36884450912475586,
+      "learning_rate": 8.320873883494586e-06,
+      "loss": 0.0248,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3427126459578055,
+      "grad_norm": 0.44685840606689453,
+      "learning_rate": 8.316729722858987e-06,
+      "loss": 0.0249,
+      "step": 1080
+    },
+    {
+      "epoch": 0.3430299724818405,
+      "grad_norm": 0.483712375164032,
+      "learning_rate": 8.312581489324234e-06,
+      "loss": 0.0257,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3433472990058755,
+      "grad_norm": 0.42289575934410095,
+      "learning_rate": 8.308429187984298e-06,
+      "loss": 0.0296,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3436646255299105,
+      "grad_norm": 0.44183018803596497,
+      "learning_rate": 8.304272823938151e-06,
+      "loss": 0.0268,
+      "step": 1083
+    },
+    {
+      "epoch": 0.3439819520539455,
+      "grad_norm": 0.4222985804080963,
+      "learning_rate": 8.30011240228975e-06,
+      "loss": 0.0287,
+      "step": 1084
+    },
+    {
+      "epoch": 0.34429927857798054,
+      "grad_norm": 2.3519210815429688,
+      "learning_rate": 8.295947928148037e-06,
+      "loss": 0.0252,
+      "step": 1085
+    },
+    {
+      "epoch": 0.3446166051020155,
+      "grad_norm": 0.4917280972003937,
+      "learning_rate": 8.291779406626931e-06,
+      "loss": 0.0275,
+      "step": 1086
+    },
+    {
+      "epoch": 0.3449339316260505,
+      "grad_norm": 0.3875178098678589,
+      "learning_rate": 8.28760684284532e-06,
+      "loss": 0.0231,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3452512581500855,
+      "grad_norm": 0.5463328957557678,
+      "learning_rate": 8.283430241927053e-06,
+      "loss": 0.028,
+      "step": 1088
+    },
+    {
+      "epoch": 0.34556858467412055,
+      "grad_norm": 0.4385169446468353,
+      "learning_rate": 8.27924960900094e-06,
+      "loss": 0.0267,
+      "step": 1089
+    },
+    {
+      "epoch": 0.34588591119815554,
+      "grad_norm": 0.4448505640029907,
+      "learning_rate": 8.275064949200746e-06,
+      "loss": 0.0308,
+      "step": 1090
+    },
+    {
+      "epoch": 0.3462032377221905,
+      "grad_norm": 0.3706183135509491,
+      "learning_rate": 8.270876267665173e-06,
+      "loss": 0.0237,
+      "step": 1091
+    },
+    {
+      "epoch": 0.34652056424622557,
+      "grad_norm": 1.1387298107147217,
+      "learning_rate": 8.266683569537866e-06,
+      "loss": 0.0275,
+      "step": 1092
+    },
+    {
+      "epoch": 0.34683789077026056,
+      "grad_norm": 0.43619418144226074,
+      "learning_rate": 8.262486859967397e-06,
+      "loss": 0.0283,
+      "step": 1093
+    },
+    {
+      "epoch": 0.34715521729429555,
+      "grad_norm": 0.39969122409820557,
+      "learning_rate": 8.258286144107277e-06,
+      "loss": 0.0287,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3474725438183306,
+      "grad_norm": 0.532938539981842,
+      "learning_rate": 8.25408142711592e-06,
+      "loss": 0.0295,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3477898703423656,
+      "grad_norm": 0.392141193151474,
+      "learning_rate": 8.249872714156668e-06,
+      "loss": 0.0245,
+      "step": 1096
+    },
+    {
+      "epoch": 0.34810719686640057,
+      "grad_norm": 0.5717901587486267,
+      "learning_rate": 8.24566001039776e-06,
+      "loss": 0.0274,
+      "step": 1097
+    },
+    {
+      "epoch": 0.34842452339043556,
+      "grad_norm": 0.4331674873828888,
+      "learning_rate": 8.241443321012339e-06,
+      "loss": 0.0266,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3487418499144706,
+      "grad_norm": 0.4639444649219513,
+      "learning_rate": 8.237222651178445e-06,
+      "loss": 0.0252,
+      "step": 1099
+    },
+    {
+      "epoch": 0.3490591764385056,
+      "grad_norm": 0.4274793565273285,
+      "learning_rate": 8.232998006078998e-06,
+      "loss": 0.0267,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3493765029625406,
+      "grad_norm": 0.36444589495658875,
+      "learning_rate": 8.228769390901812e-06,
+      "loss": 0.0264,
+      "step": 1101
+    },
+    {
+      "epoch": 0.3496938294865756,
+      "grad_norm": 0.45189785957336426,
+      "learning_rate": 8.224536810839564e-06,
+      "loss": 0.0231,
+      "step": 1102
+    },
+    {
+      "epoch": 0.3500111560106106,
+      "grad_norm": 0.5885544419288635,
+      "learning_rate": 8.220300271089806e-06,
+      "loss": 0.0264,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3503284825346456,
+      "grad_norm": 0.41699036955833435,
+      "learning_rate": 8.216059776854952e-06,
+      "loss": 0.0264,
+      "step": 1104
+    },
+    {
+      "epoch": 0.3506458090586806,
+      "grad_norm": 0.41668203473091125,
+      "learning_rate": 8.211815333342272e-06,
+      "loss": 0.0251,
+      "step": 1105
+    },
+    {
+      "epoch": 0.35096313558271564,
+      "grad_norm": 0.41037651896476746,
+      "learning_rate": 8.207566945763885e-06,
+      "loss": 0.0261,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3512804621067506,
+      "grad_norm": 0.6969994902610779,
+      "learning_rate": 8.203314619336754e-06,
+      "loss": 0.0243,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3515977886307856,
+      "grad_norm": 0.761831521987915,
+      "learning_rate": 8.199058359282675e-06,
+      "loss": 0.0237,
+      "step": 1108
+    },
+    {
+      "epoch": 0.35191511515482066,
+      "grad_norm": 0.4980373978614807,
+      "learning_rate": 8.19479817082828e-06,
+      "loss": 0.0319,
+      "step": 1109
+    },
+    {
+      "epoch": 0.35223244167885565,
+      "grad_norm": 0.49130550026893616,
+      "learning_rate": 8.190534059205023e-06,
+      "loss": 0.0233,
+      "step": 1110
+    },
+    {
+      "epoch": 0.35254976820289063,
+      "grad_norm": 0.5645861625671387,
+      "learning_rate": 8.186266029649174e-06,
+      "loss": 0.0257,
+      "step": 1111
+    },
+    {
+      "epoch": 0.3528670947269257,
+      "grad_norm": 0.5322470664978027,
+      "learning_rate": 8.181994087401819e-06,
+      "loss": 0.023,
+      "step": 1112
+    },
+    {
+      "epoch": 0.35318442125096067,
+      "grad_norm": 0.809970498085022,
+      "learning_rate": 8.17771823770884e-06,
+      "loss": 0.0269,
+      "step": 1113
+    },
+    {
+      "epoch": 0.35350174777499566,
+      "grad_norm": 0.7466074824333191,
+      "learning_rate": 8.173438485820927e-06,
+      "loss": 0.0314,
+      "step": 1114
+    },
+    {
+      "epoch": 0.35381907429903064,
+      "grad_norm": 0.5680863857269287,
+      "learning_rate": 8.16915483699355e-06,
+      "loss": 0.0237,
+      "step": 1115
+    },
+    {
+      "epoch": 0.3541364008230657,
+      "grad_norm": 0.6937032341957092,
+      "learning_rate": 8.164867296486983e-06,
+      "loss": 0.0257,
+      "step": 1116
+    },
+    {
+      "epoch": 0.3544537273471007,
+      "grad_norm": 0.48837175965309143,
+      "learning_rate": 8.16057586956626e-06,
+      "loss": 0.0278,
+      "step": 1117
+    },
+    {
+      "epoch": 0.35477105387113567,
+      "grad_norm": 0.6901347041130066,
+      "learning_rate": 8.156280561501196e-06,
+      "loss": 0.0251,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3550883803951707,
+      "grad_norm": 0.4559510350227356,
+      "learning_rate": 8.151981377566372e-06,
+      "loss": 0.0261,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3554057069192057,
+      "grad_norm": 0.4369254410266876,
+      "learning_rate": 8.147678323041125e-06,
+      "loss": 0.0215,
+      "step": 1120
+    },
+    {
+      "epoch": 0.3557230334432407,
+      "grad_norm": 0.44330379366874695,
+      "learning_rate": 8.143371403209555e-06,
+      "loss": 0.0223,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3560403599672757,
+      "grad_norm": 0.5089208483695984,
+      "learning_rate": 8.139060623360494e-06,
+      "loss": 0.0292,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3563576864913107,
+      "grad_norm": 0.42668241262435913,
+      "learning_rate": 8.134745988787523e-06,
+      "loss": 0.024,
+      "step": 1123
+    },
+    {
+      "epoch": 0.3566750130153457,
+      "grad_norm": 0.572441816329956,
+      "learning_rate": 8.130427504788956e-06,
+      "loss": 0.0318,
+      "step": 1124
+    },
+    {
+      "epoch": 0.3569923395393807,
+      "grad_norm": 0.48021039366722107,
+      "learning_rate": 8.12610517666783e-06,
+      "loss": 0.0218,
+      "step": 1125
+    },
+    {
+      "epoch": 0.35730966606341574,
+      "grad_norm": 0.8262785077095032,
+      "learning_rate": 8.121779009731912e-06,
+      "loss": 0.0291,
+      "step": 1126
+    },
+    {
+      "epoch": 0.35762699258745073,
+      "grad_norm": 0.6068892478942871,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.0295,
+      "step": 1127
+    },
+    {
+      "epoch": 0.3579443191114857,
+      "grad_norm": 0.7416961193084717,
+      "learning_rate": 8.113115180670286e-06,
+      "loss": 0.0327,
+      "step": 1128
+    },
+    {
+      "epoch": 0.35826164563552076,
+      "grad_norm": 0.6804921627044678,
+      "learning_rate": 8.108777529183644e-06,
+      "loss": 0.0233,
+      "step": 1129
+    },
+    {
+      "epoch": 0.35857897215955575,
+      "grad_norm": 0.520760178565979,
+      "learning_rate": 8.104436060160324e-06,
+      "loss": 0.0322,
+      "step": 1130
+    },
+    {
+      "epoch": 0.35889629868359074,
+      "grad_norm": 0.7374339699745178,
+      "learning_rate": 8.100090778931588e-06,
+      "loss": 0.0249,
+      "step": 1131
+    },
+    {
+      "epoch": 0.35921362520762573,
+      "grad_norm": 0.5813891291618347,
+      "learning_rate": 8.095741690833385e-06,
+      "loss": 0.0274,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3595309517316608,
+      "grad_norm": 0.9124199748039246,
+      "learning_rate": 8.091388801206334e-06,
+      "loss": 0.0261,
+      "step": 1133
+    },
+    {
+      "epoch": 0.35984827825569576,
+      "grad_norm": 0.5454839468002319,
+      "learning_rate": 8.087032115395725e-06,
+      "loss": 0.0275,
+      "step": 1134
+    },
+    {
+      "epoch": 0.36016560477973075,
+      "grad_norm": 0.5077856779098511,
+      "learning_rate": 8.082671638751508e-06,
+      "loss": 0.0245,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3604829313037658,
+      "grad_norm": 1.2194993495941162,
+      "learning_rate": 8.078307376628292e-06,
+      "loss": 0.0283,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3608002578278008,
+      "grad_norm": 0.755836546421051,
+      "learning_rate": 8.07393933438533e-06,
+      "loss": 0.028,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3611175843518358,
+      "grad_norm": 0.5165392160415649,
+      "learning_rate": 8.069567517386518e-06,
+      "loss": 0.0285,
+      "step": 1138
+    },
+    {
+      "epoch": 0.36143491087587076,
+      "grad_norm": 0.6336295008659363,
+      "learning_rate": 8.065191931000391e-06,
+      "loss": 0.0295,
+      "step": 1139
+    },
+    {
+      "epoch": 0.3617522373999058,
+      "grad_norm": 0.9173374176025391,
+      "learning_rate": 8.060812580600106e-06,
+      "loss": 0.031,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3620695639239408,
+      "grad_norm": 0.7403201460838318,
+      "learning_rate": 8.05642947156345e-06,
+      "loss": 0.0277,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3623868904479758,
+      "grad_norm": 0.4352768361568451,
+      "learning_rate": 8.052042609272817e-06,
+      "loss": 0.0293,
+      "step": 1142
+    },
+    {
+      "epoch": 0.36270421697201083,
+      "grad_norm": 0.7653275728225708,
+      "learning_rate": 8.047651999115216e-06,
+      "loss": 0.0267,
+      "step": 1143
+    },
+    {
+      "epoch": 0.3630215434960458,
+      "grad_norm": 0.7468372583389282,
+      "learning_rate": 8.04325764648226e-06,
+      "loss": 0.0246,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3633388700200808,
+      "grad_norm": 0.34332841634750366,
+      "learning_rate": 8.038859556770152e-06,
+      "loss": 0.0249,
+      "step": 1145
+    },
+    {
+      "epoch": 0.36365619654411585,
+      "grad_norm": 0.3993615210056305,
+      "learning_rate": 8.034457735379683e-06,
+      "loss": 0.0281,
+      "step": 1146
+    },
+    {
+      "epoch": 0.36397352306815084,
+      "grad_norm": 0.6524525284767151,
+      "learning_rate": 8.030052187716238e-06,
+      "loss": 0.0215,
+      "step": 1147
+    },
+    {
+      "epoch": 0.36429084959218583,
+      "grad_norm": 0.6711657643318176,
+      "learning_rate": 8.025642919189763e-06,
+      "loss": 0.0284,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3646081761162208,
+      "grad_norm": 0.8700569272041321,
+      "learning_rate": 8.021229935214782e-06,
+      "loss": 0.0256,
+      "step": 1149
+    },
+    {
+      "epoch": 0.36492550264025586,
+      "grad_norm": 0.4704226851463318,
+      "learning_rate": 8.01681324121038e-06,
+      "loss": 0.0278,
+      "step": 1150
+    },
+    {
+      "epoch": 0.36524282916429085,
+      "grad_norm": 0.7297829985618591,
+      "learning_rate": 8.012392842600198e-06,
+      "loss": 0.0269,
+      "step": 1151
+    },
+    {
+      "epoch": 0.36556015568832584,
+      "grad_norm": 1.3700752258300781,
+      "learning_rate": 8.007968744812424e-06,
+      "loss": 0.0261,
+      "step": 1152
+    },
+    {
+      "epoch": 0.3658774822123609,
+      "grad_norm": 0.7244130969047546,
+      "learning_rate": 8.003540953279793e-06,
+      "loss": 0.0276,
+      "step": 1153
+    },
+    {
+      "epoch": 0.36619480873639587,
+      "grad_norm": 0.5642964243888855,
+      "learning_rate": 7.99910947343957e-06,
+      "loss": 0.0255,
+      "step": 1154
+    },
+    {
+      "epoch": 0.36651213526043086,
+      "grad_norm": 0.620916485786438,
+      "learning_rate": 7.994674310733553e-06,
+      "loss": 0.0254,
+      "step": 1155
+    },
+    {
+      "epoch": 0.36682946178446585,
+      "grad_norm": 0.8156549334526062,
+      "learning_rate": 7.99023547060806e-06,
+      "loss": 0.0254,
+      "step": 1156
+    },
+    {
+      "epoch": 0.3671467883085009,
+      "grad_norm": 0.6374293565750122,
+      "learning_rate": 7.985792958513932e-06,
+      "loss": 0.0301,
+      "step": 1157
+    },
+    {
+      "epoch": 0.3674641148325359,
+      "grad_norm": 0.4271999001502991,
+      "learning_rate": 7.98134677990651e-06,
+      "loss": 0.0274,
+      "step": 1158
+    },
+    {
+      "epoch": 0.36778144135657087,
+      "grad_norm": 0.46824926137924194,
+      "learning_rate": 7.976896940245637e-06,
+      "loss": 0.0302,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3680987678806059,
+      "grad_norm": 0.4343389570713043,
+      "learning_rate": 7.972443444995663e-06,
+      "loss": 0.0265,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3684160944046409,
+      "grad_norm": 0.7338358163833618,
+      "learning_rate": 7.967986299625417e-06,
+      "loss": 0.0254,
+      "step": 1161
+    },
+    {
+      "epoch": 0.3687334209286759,
+      "grad_norm": 0.39128246903419495,
+      "learning_rate": 7.963525509608214e-06,
+      "loss": 0.0286,
+      "step": 1162
+    },
+    {
+      "epoch": 0.36905074745271094,
+      "grad_norm": 0.4054844081401825,
+      "learning_rate": 7.95906108042184e-06,
+      "loss": 0.0243,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3693680739767459,
+      "grad_norm": 0.37072888016700745,
+      "learning_rate": 7.954593017548557e-06,
+      "loss": 0.0255,
+      "step": 1164
+    },
+    {
+      "epoch": 0.3696854005007809,
+      "grad_norm": 0.5412118434906006,
+      "learning_rate": 7.950121326475085e-06,
+      "loss": 0.0313,
+      "step": 1165
+    },
+    {
+      "epoch": 0.3700027270248159,
+      "grad_norm": 0.3992641568183899,
+      "learning_rate": 7.9456460126926e-06,
+      "loss": 0.025,
+      "step": 1166
+    },
+    {
+      "epoch": 0.37032005354885095,
+      "grad_norm": 0.441742479801178,
+      "learning_rate": 7.941167081696725e-06,
+      "loss": 0.0293,
+      "step": 1167
+    },
+    {
+      "epoch": 0.37063738007288594,
+      "grad_norm": 0.3998163938522339,
+      "learning_rate": 7.936684538987531e-06,
+      "loss": 0.0262,
+      "step": 1168
+    },
+    {
+      "epoch": 0.3709547065969209,
+      "grad_norm": 0.4399847388267517,
+      "learning_rate": 7.932198390069515e-06,
+      "loss": 0.0275,
+      "step": 1169
+    },
+    {
+      "epoch": 0.37127203312095597,
+      "grad_norm": 0.4310084879398346,
+      "learning_rate": 7.92770864045161e-06,
+      "loss": 0.0276,
+      "step": 1170
+    },
+    {
+      "epoch": 0.37158935964499096,
+      "grad_norm": 0.4524058997631073,
+      "learning_rate": 7.923215295647167e-06,
+      "loss": 0.0253,
+      "step": 1171
+    },
+    {
+      "epoch": 0.37190668616902595,
+      "grad_norm": 0.46805015206336975,
+      "learning_rate": 7.918718361173951e-06,
+      "loss": 0.0253,
+      "step": 1172
+    },
+    {
+      "epoch": 0.37222401269306093,
+      "grad_norm": 0.4731823205947876,
+      "learning_rate": 7.914217842554138e-06,
+      "loss": 0.0296,
+      "step": 1173
+    },
+    {
+      "epoch": 0.372541339217096,
+      "grad_norm": 0.48658451437950134,
+      "learning_rate": 7.909713745314303e-06,
+      "loss": 0.0294,
+      "step": 1174
+    },
+    {
+      "epoch": 0.37285866574113097,
+      "grad_norm": 0.49759331345558167,
+      "learning_rate": 7.905206074985416e-06,
+      "loss": 0.0272,
+      "step": 1175
+    },
+    {
+      "epoch": 0.37317599226516596,
+      "grad_norm": 0.3912714123725891,
+      "learning_rate": 7.900694837102839e-06,
+      "loss": 0.0243,
+      "step": 1176
+    },
+    {
+      "epoch": 0.373493318789201,
+      "grad_norm": 0.603121280670166,
+      "learning_rate": 7.896180037206306e-06,
+      "loss": 0.0302,
+      "step": 1177
+    },
+    {
+      "epoch": 0.373810645313236,
+      "grad_norm": 0.7376512289047241,
+      "learning_rate": 7.891661680839932e-06,
+      "loss": 0.0254,
+      "step": 1178
+    },
+    {
+      "epoch": 0.374127971837271,
+      "grad_norm": 0.40546655654907227,
+      "learning_rate": 7.8871397735522e-06,
+      "loss": 0.0257,
+      "step": 1179
+    },
+    {
+      "epoch": 0.374445298361306,
+      "grad_norm": 0.4415283799171448,
+      "learning_rate": 7.882614320895947e-06,
+      "loss": 0.027,
+      "step": 1180
+    },
+    {
+      "epoch": 0.374762624885341,
+      "grad_norm": 0.39566221833229065,
+      "learning_rate": 7.87808532842837e-06,
+      "loss": 0.0245,
+      "step": 1181
+    },
+    {
+      "epoch": 0.375079951409376,
+      "grad_norm": 0.4054136872291565,
+      "learning_rate": 7.873552801711007e-06,
+      "loss": 0.0286,
+      "step": 1182
+    },
+    {
+      "epoch": 0.375397277933411,
+      "grad_norm": 0.75135737657547,
+      "learning_rate": 7.869016746309743e-06,
+      "loss": 0.0246,
+      "step": 1183
+    },
+    {
+      "epoch": 0.37571460445744603,
+      "grad_norm": 0.49160444736480713,
+      "learning_rate": 7.86447716779479e-06,
+      "loss": 0.0252,
+      "step": 1184
+    },
+    {
+      "epoch": 0.376031930981481,
+      "grad_norm": 0.484737366437912,
+      "learning_rate": 7.859934071740693e-06,
+      "loss": 0.0275,
+      "step": 1185
+    },
+    {
+      "epoch": 0.376349257505516,
+      "grad_norm": 0.41686415672302246,
+      "learning_rate": 7.855387463726307e-06,
+      "loss": 0.0215,
+      "step": 1186
+    },
+    {
+      "epoch": 0.37666658402955105,
+      "grad_norm": 0.4131918251514435,
+      "learning_rate": 7.85083734933481e-06,
+      "loss": 0.027,
+      "step": 1187
+    },
+    {
+      "epoch": 0.37698391055358604,
+      "grad_norm": 0.48149415850639343,
+      "learning_rate": 7.846283734153677e-06,
+      "loss": 0.0278,
+      "step": 1188
+    },
+    {
+      "epoch": 0.37730123707762103,
+      "grad_norm": 0.5752674341201782,
+      "learning_rate": 7.841726623774687e-06,
+      "loss": 0.0324,
+      "step": 1189
+    },
+    {
+      "epoch": 0.377618563601656,
+      "grad_norm": 0.3986679017543793,
+      "learning_rate": 7.83716602379391e-06,
+      "loss": 0.0289,
+      "step": 1190
+    },
+    {
+      "epoch": 0.37793589012569107,
+      "grad_norm": 0.39355382323265076,
+      "learning_rate": 7.832601939811701e-06,
+      "loss": 0.026,
+      "step": 1191
+    },
+    {
+      "epoch": 0.37825321664972605,
+      "grad_norm": 0.5375329852104187,
+      "learning_rate": 7.828034377432694e-06,
+      "loss": 0.0266,
+      "step": 1192
+    },
+    {
+      "epoch": 0.37857054317376104,
+      "grad_norm": 1.5827182531356812,
+      "learning_rate": 7.823463342265793e-06,
+      "loss": 0.0277,
+      "step": 1193
+    },
+    {
+      "epoch": 0.3788878696977961,
+      "grad_norm": 0.44673463702201843,
+      "learning_rate": 7.818888839924165e-06,
+      "loss": 0.0269,
+      "step": 1194
+    },
+    {
+      "epoch": 0.3792051962218311,
+      "grad_norm": 0.46579429507255554,
+      "learning_rate": 7.814310876025242e-06,
+      "loss": 0.0264,
+      "step": 1195
+    },
+    {
+      "epoch": 0.37952252274586606,
+      "grad_norm": 0.7898886799812317,
+      "learning_rate": 7.809729456190699e-06,
+      "loss": 0.0244,
+      "step": 1196
+    },
+    {
+      "epoch": 0.3798398492699011,
+      "grad_norm": 0.649233341217041,
+      "learning_rate": 7.805144586046454e-06,
+      "loss": 0.0213,
+      "step": 1197
+    },
+    {
+      "epoch": 0.3801571757939361,
+      "grad_norm": 0.41746070981025696,
+      "learning_rate": 7.80055627122267e-06,
+      "loss": 0.0242,
+      "step": 1198
+    },
+    {
+      "epoch": 0.3804745023179711,
+      "grad_norm": 0.5043621063232422,
+      "learning_rate": 7.795964517353734e-06,
+      "loss": 0.0253,
+      "step": 1199
+    },
+    {
+      "epoch": 0.3807918288420061,
+      "grad_norm": 0.48764780163764954,
+      "learning_rate": 7.791369330078258e-06,
+      "loss": 0.0256,
+      "step": 1200
+    },
+    {
+      "epoch": 0.3811091553660411,
+      "grad_norm": 0.40910306572914124,
+      "learning_rate": 7.786770715039069e-06,
+      "loss": 0.0241,
+      "step": 1201
+    },
+    {
+      "epoch": 0.3814264818900761,
+      "grad_norm": 0.7150037884712219,
+      "learning_rate": 7.782168677883206e-06,
+      "loss": 0.03,
+      "step": 1202
+    },
+    {
+      "epoch": 0.3817438084141111,
+      "grad_norm": 0.4457259178161621,
+      "learning_rate": 7.777563224261905e-06,
+      "loss": 0.0275,
+      "step": 1203
+    },
+    {
+      "epoch": 0.38206113493814614,
+      "grad_norm": 0.5687321424484253,
+      "learning_rate": 7.772954359830603e-06,
+      "loss": 0.0265,
+      "step": 1204
+    },
+    {
+      "epoch": 0.38237846146218113,
+      "grad_norm": 0.37201011180877686,
+      "learning_rate": 7.768342090248919e-06,
+      "loss": 0.0247,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3826957879862161,
+      "grad_norm": 0.9412732720375061,
+      "learning_rate": 7.763726421180664e-06,
+      "loss": 0.023,
+      "step": 1206
+    },
+    {
+      "epoch": 0.3830131145102511,
+      "grad_norm": 0.4486013352870941,
+      "learning_rate": 7.759107358293814e-06,
+      "loss": 0.0235,
+      "step": 1207
+    },
+    {
+      "epoch": 0.38333044103428615,
+      "grad_norm": 0.6515507698059082,
+      "learning_rate": 7.754484907260513e-06,
+      "loss": 0.0245,
+      "step": 1208
+    },
+    {
+      "epoch": 0.38364776755832114,
+      "grad_norm": 0.39932361245155334,
+      "learning_rate": 7.749859073757072e-06,
+      "loss": 0.0214,
+      "step": 1209
+    },
+    {
+      "epoch": 0.38396509408235613,
+      "grad_norm": 0.42794036865234375,
+      "learning_rate": 7.745229863463951e-06,
+      "loss": 0.0243,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3842824206063912,
+      "grad_norm": 0.6306819915771484,
+      "learning_rate": 7.740597282065756e-06,
+      "loss": 0.0263,
+      "step": 1211
+    },
+    {
+      "epoch": 0.38459974713042616,
+      "grad_norm": 0.5999877452850342,
+      "learning_rate": 7.735961335251236e-06,
+      "loss": 0.0256,
+      "step": 1212
+    },
+    {
+      "epoch": 0.38491707365446115,
+      "grad_norm": 0.49755433201789856,
+      "learning_rate": 7.73132202871327e-06,
+      "loss": 0.0223,
+      "step": 1213
+    },
+    {
+      "epoch": 0.3852344001784962,
+      "grad_norm": 0.4599727690219879,
+      "learning_rate": 7.726679368148863e-06,
+      "loss": 0.0241,
+      "step": 1214
+    },
+    {
+      "epoch": 0.3855517267025312,
+      "grad_norm": 0.4556451737880707,
+      "learning_rate": 7.722033359259142e-06,
+      "loss": 0.0234,
+      "step": 1215
+    },
+    {
+      "epoch": 0.38586905322656617,
+      "grad_norm": 0.5568158626556396,
+      "learning_rate": 7.71738400774934e-06,
+      "loss": 0.0255,
+      "step": 1216
+    },
+    {
+      "epoch": 0.38618637975060116,
+      "grad_norm": 0.4913748502731323,
+      "learning_rate": 7.712731319328798e-06,
+      "loss": 0.0275,
+      "step": 1217
+    },
+    {
+      "epoch": 0.3865037062746362,
+      "grad_norm": 0.4001238942146301,
+      "learning_rate": 7.708075299710955e-06,
+      "loss": 0.0215,
+      "step": 1218
+    },
+    {
+      "epoch": 0.3868210327986712,
+      "grad_norm": 4.385622501373291,
+      "learning_rate": 7.70341595461334e-06,
+      "loss": 0.0271,
+      "step": 1219
+    },
+    {
+      "epoch": 0.3871383593227062,
+      "grad_norm": 0.4039493203163147,
+      "learning_rate": 7.698753289757565e-06,
+      "loss": 0.0237,
+      "step": 1220
+    },
+    {
+      "epoch": 0.3874556858467412,
+      "grad_norm": 0.39531779289245605,
+      "learning_rate": 7.69408731086932e-06,
+      "loss": 0.0269,
+      "step": 1221
+    },
+    {
+      "epoch": 0.3877730123707762,
+      "grad_norm": 0.47068384289741516,
+      "learning_rate": 7.689418023678363e-06,
+      "loss": 0.0312,
+      "step": 1222
+    },
+    {
+      "epoch": 0.3880903388948112,
+      "grad_norm": 0.39732053875923157,
+      "learning_rate": 7.684745433918519e-06,
+      "loss": 0.0224,
+      "step": 1223
+    },
+    {
+      "epoch": 0.3884076654188462,
+      "grad_norm": 0.3397679924964905,
+      "learning_rate": 7.680069547327659e-06,
+      "loss": 0.0241,
+      "step": 1224
+    },
+    {
+      "epoch": 0.38872499194288124,
+      "grad_norm": 0.3972413241863251,
+      "learning_rate": 7.675390369647711e-06,
+      "loss": 0.0248,
+      "step": 1225
+    },
+    {
+      "epoch": 0.3890423184669162,
+      "grad_norm": 0.3952977955341339,
+      "learning_rate": 7.670707906624644e-06,
+      "loss": 0.0234,
+      "step": 1226
+    },
+    {
+      "epoch": 0.3893596449909512,
+      "grad_norm": 0.40310418605804443,
+      "learning_rate": 7.666022164008458e-06,
+      "loss": 0.0252,
+      "step": 1227
+    },
+    {
+      "epoch": 0.38967697151498626,
+      "grad_norm": 0.5499181151390076,
+      "learning_rate": 7.661333147553178e-06,
+      "loss": 0.0282,
+      "step": 1228
+    },
+    {
+      "epoch": 0.38999429803902125,
+      "grad_norm": 0.5339550375938416,
+      "learning_rate": 7.656640863016857e-06,
+      "loss": 0.0253,
+      "step": 1229
+    },
+    {
+      "epoch": 0.39031162456305624,
+      "grad_norm": 0.4359331727027893,
+      "learning_rate": 7.651945316161555e-06,
+      "loss": 0.03,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3906289510870913,
+      "grad_norm": 0.4402971863746643,
+      "learning_rate": 7.647246512753339e-06,
+      "loss": 0.0227,
+      "step": 1231
+    },
+    {
+      "epoch": 0.39094627761112627,
+      "grad_norm": 0.4190260171890259,
+      "learning_rate": 7.642544458562278e-06,
+      "loss": 0.0247,
+      "step": 1232
+    },
+    {
+      "epoch": 0.39126360413516126,
+      "grad_norm": 0.41850367188453674,
+      "learning_rate": 7.63783915936243e-06,
+      "loss": 0.0276,
+      "step": 1233
+    },
+    {
+      "epoch": 0.39158093065919625,
+      "grad_norm": 0.4658275842666626,
+      "learning_rate": 7.633130620931837e-06,
+      "loss": 0.0243,
+      "step": 1234
+    },
+    {
+      "epoch": 0.3918982571832313,
+      "grad_norm": 0.41021329164505005,
+      "learning_rate": 7.628418849052523e-06,
+      "loss": 0.027,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3922155837072663,
+      "grad_norm": 0.36580464243888855,
+      "learning_rate": 7.62370384951048e-06,
+      "loss": 0.0226,
+      "step": 1236
+    },
+    {
+      "epoch": 0.39253291023130127,
+      "grad_norm": 0.39270028471946716,
+      "learning_rate": 7.618985628095662e-06,
+      "loss": 0.0263,
+      "step": 1237
+    },
+    {
+      "epoch": 0.3928502367553363,
+      "grad_norm": 0.40487226843833923,
+      "learning_rate": 7.614264190601981e-06,
+      "loss": 0.0218,
+      "step": 1238
+    },
+    {
+      "epoch": 0.3931675632793713,
+      "grad_norm": 0.8318223357200623,
+      "learning_rate": 7.609539542827298e-06,
+      "loss": 0.0268,
+      "step": 1239
+    },
+    {
+      "epoch": 0.3934848898034063,
+      "grad_norm": 0.5691092610359192,
+      "learning_rate": 7.604811690573418e-06,
+      "loss": 0.0289,
+      "step": 1240
+    },
+    {
+      "epoch": 0.39380221632744133,
+      "grad_norm": 0.5840064287185669,
+      "learning_rate": 7.600080639646077e-06,
+      "loss": 0.0282,
+      "step": 1241
+    },
+    {
+      "epoch": 0.3941195428514763,
+      "grad_norm": 0.43909749388694763,
+      "learning_rate": 7.595346395854946e-06,
+      "loss": 0.023,
+      "step": 1242
+    },
+    {
+      "epoch": 0.3944368693755113,
+      "grad_norm": 0.6908259987831116,
+      "learning_rate": 7.590608965013608e-06,
+      "loss": 0.026,
+      "step": 1243
+    },
+    {
+      "epoch": 0.3947541958995463,
+      "grad_norm": 0.4869570732116699,
+      "learning_rate": 7.585868352939564e-06,
+      "loss": 0.0279,
+      "step": 1244
+    },
+    {
+      "epoch": 0.39507152242358134,
+      "grad_norm": 0.4845676124095917,
+      "learning_rate": 7.581124565454223e-06,
+      "loss": 0.0245,
+      "step": 1245
+    },
+    {
+      "epoch": 0.39538884894761633,
+      "grad_norm": 0.4240732789039612,
+      "learning_rate": 7.576377608382892e-06,
+      "loss": 0.0241,
+      "step": 1246
+    },
+    {
+      "epoch": 0.3957061754716513,
+      "grad_norm": 0.4161009192466736,
+      "learning_rate": 7.571627487554769e-06,
+      "loss": 0.0254,
+      "step": 1247
+    },
+    {
+      "epoch": 0.39602350199568637,
+      "grad_norm": 0.482496440410614,
+      "learning_rate": 7.566874208802939e-06,
+      "loss": 0.0259,
+      "step": 1248
+    },
+    {
+      "epoch": 0.39634082851972136,
+      "grad_norm": 0.361062228679657,
+      "learning_rate": 7.562117777964361e-06,
+      "loss": 0.0225,
+      "step": 1249
+    },
+    {
+      "epoch": 0.39665815504375634,
+      "grad_norm": 0.4240871071815491,
+      "learning_rate": 7.5573582008798706e-06,
+      "loss": 0.0232,
+      "step": 1250
+    },
+    {
+      "epoch": 0.39697548156779133,
+      "grad_norm": 0.3940419554710388,
+      "learning_rate": 7.5525954833941635e-06,
+      "loss": 0.0195,
+      "step": 1251
+    },
+    {
+      "epoch": 0.3972928080918264,
+      "grad_norm": 0.4435242712497711,
+      "learning_rate": 7.547829631355794e-06,
+      "loss": 0.0265,
+      "step": 1252
+    },
+    {
+      "epoch": 0.39761013461586137,
+      "grad_norm": 0.4083918035030365,
+      "learning_rate": 7.543060650617159e-06,
+      "loss": 0.0268,
+      "step": 1253
+    },
+    {
+      "epoch": 0.39792746113989635,
+      "grad_norm": 0.5816600918769836,
+      "learning_rate": 7.538288547034506e-06,
+      "loss": 0.0236,
+      "step": 1254
+    },
+    {
+      "epoch": 0.3982447876639314,
+      "grad_norm": 0.36388111114501953,
+      "learning_rate": 7.533513326467911e-06,
+      "loss": 0.023,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3985621141879664,
+      "grad_norm": 0.4519801735877991,
+      "learning_rate": 7.528734994781284e-06,
+      "loss": 0.024,
+      "step": 1256
+    },
+    {
+      "epoch": 0.3988794407120014,
+      "grad_norm": 0.9629951119422913,
+      "learning_rate": 7.5239535578423474e-06,
+      "loss": 0.0254,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3991967672360364,
+      "grad_norm": 0.3986568748950958,
+      "learning_rate": 7.519169021522643e-06,
+      "loss": 0.0217,
+      "step": 1258
+    },
+    {
+      "epoch": 0.3995140937600714,
+      "grad_norm": 0.3642372488975525,
+      "learning_rate": 7.514381391697518e-06,
+      "loss": 0.0234,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3998314202841064,
+      "grad_norm": 0.6354243159294128,
+      "learning_rate": 7.509590674246112e-06,
+      "loss": 0.0279,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4001487468081414,
+      "grad_norm": 0.5784509181976318,
+      "learning_rate": 7.504796875051368e-06,
+      "loss": 0.0227,
+      "step": 1261
+    },
+    {
+      "epoch": 0.40046607333217643,
+      "grad_norm": 0.4335835874080658,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.0266,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4007833998562114,
+      "grad_norm": 0.5561223030090332,
+      "learning_rate": 7.49520005498251e-06,
+      "loss": 0.0294,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4011007263802464,
+      "grad_norm": 0.6013244390487671,
+      "learning_rate": 7.490397045893164e-06,
+      "loss": 0.0255,
+      "step": 1264
+    },
+    {
+      "epoch": 0.40141805290428145,
+      "grad_norm": 0.3803730607032776,
+      "learning_rate": 7.485590978629991e-06,
+      "loss": 0.0223,
+      "step": 1265
+    },
+    {
+      "epoch": 0.40173537942831644,
+      "grad_norm": 0.3937969505786896,
+      "learning_rate": 7.480781859094778e-06,
+      "loss": 0.023,
+      "step": 1266
+    },
+    {
+      "epoch": 0.40205270595235143,
+      "grad_norm": 0.7828153967857361,
+      "learning_rate": 7.475969693193057e-06,
+      "loss": 0.0241,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4023700324763864,
+      "grad_norm": 0.37793123722076416,
+      "learning_rate": 7.471154486834105e-06,
+      "loss": 0.0279,
+      "step": 1268
+    },
+    {
+      "epoch": 0.40268735900042146,
+      "grad_norm": 0.5311406850814819,
+      "learning_rate": 7.466336245930927e-06,
+      "loss": 0.0253,
+      "step": 1269
+    },
+    {
+      "epoch": 0.40300468552445645,
+      "grad_norm": 0.4039859175682068,
+      "learning_rate": 7.46151497640026e-06,
+      "loss": 0.0241,
+      "step": 1270
+    },
+    {
+      "epoch": 0.40332201204849144,
+      "grad_norm": 0.47483640909194946,
+      "learning_rate": 7.456690684162557e-06,
+      "loss": 0.0199,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4036393385725265,
+      "grad_norm": 0.3460010290145874,
+      "learning_rate": 7.451863375141984e-06,
+      "loss": 0.0215,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4039566650965615,
+      "grad_norm": 0.3834797441959381,
+      "learning_rate": 7.447033055266411e-06,
+      "loss": 0.0198,
+      "step": 1273
+    },
+    {
+      "epoch": 0.40427399162059646,
+      "grad_norm": 0.43220388889312744,
+      "learning_rate": 7.442199730467403e-06,
+      "loss": 0.0207,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4045913181446315,
+      "grad_norm": 0.41027504205703735,
+      "learning_rate": 7.4373634066802206e-06,
+      "loss": 0.0253,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4049086446686665,
+      "grad_norm": 0.9925998449325562,
+      "learning_rate": 7.4325240898438e-06,
+      "loss": 0.0281,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4052259711927015,
+      "grad_norm": 0.48082756996154785,
+      "learning_rate": 7.4276817859007615e-06,
+      "loss": 0.0247,
+      "step": 1277
+    },
+    {
+      "epoch": 0.4055432977167365,
+      "grad_norm": 0.5014410018920898,
+      "learning_rate": 7.422836500797386e-06,
+      "loss": 0.025,
+      "step": 1278
+    },
+    {
+      "epoch": 0.4058606242407715,
+      "grad_norm": 0.35161522030830383,
+      "learning_rate": 7.417988240483618e-06,
+      "loss": 0.0183,
+      "step": 1279
+    },
+    {
+      "epoch": 0.4061779507648065,
+      "grad_norm": 0.31721818447113037,
+      "learning_rate": 7.413137010913055e-06,
+      "loss": 0.021,
+      "step": 1280
+    },
+    {
+      "epoch": 0.4064952772888415,
+      "grad_norm": 0.3986819088459015,
+      "learning_rate": 7.408282818042942e-06,
+      "loss": 0.0242,
+      "step": 1281
+    },
+    {
+      "epoch": 0.40681260381287654,
+      "grad_norm": 0.6125516295433044,
+      "learning_rate": 7.403425667834164e-06,
+      "loss": 0.0185,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4071299303369115,
+      "grad_norm": 0.428974449634552,
+      "learning_rate": 7.398565566251232e-06,
+      "loss": 0.0225,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4074472568609465,
+      "grad_norm": 0.46708357334136963,
+      "learning_rate": 7.393702519262287e-06,
+      "loss": 0.0295,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4077645833849815,
+      "grad_norm": 0.4053926467895508,
+      "learning_rate": 7.388836532839083e-06,
+      "loss": 0.022,
+      "step": 1285
+    },
+    {
+      "epoch": 0.40808190990901655,
+      "grad_norm": 0.45544353127479553,
+      "learning_rate": 7.383967612956988e-06,
+      "loss": 0.0237,
+      "step": 1286
+    },
+    {
+      "epoch": 0.40839923643305154,
+      "grad_norm": 0.4615209102630615,
+      "learning_rate": 7.37909576559497e-06,
+      "loss": 0.0289,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4087165629570865,
+      "grad_norm": 0.42988142371177673,
+      "learning_rate": 7.374220996735588e-06,
+      "loss": 0.0271,
+      "step": 1288
+    },
+    {
+      "epoch": 0.40903388948112157,
+      "grad_norm": 0.3423292934894562,
+      "learning_rate": 7.369343312364994e-06,
+      "loss": 0.0239,
+      "step": 1289
+    },
+    {
+      "epoch": 0.40935121600515656,
+      "grad_norm": 0.4145824909210205,
+      "learning_rate": 7.364462718472919e-06,
+      "loss": 0.023,
+      "step": 1290
+    },
+    {
+      "epoch": 0.40966854252919155,
+      "grad_norm": 0.5536074042320251,
+      "learning_rate": 7.359579221052665e-06,
+      "loss": 0.0252,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4099858690532266,
+      "grad_norm": 0.378606379032135,
+      "learning_rate": 7.354692826101102e-06,
+      "loss": 0.0244,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4103031955772616,
+      "grad_norm": 0.38977566361427307,
+      "learning_rate": 7.349803539618654e-06,
+      "loss": 0.0233,
+      "step": 1293
+    },
+    {
+      "epoch": 0.41062052210129657,
+      "grad_norm": 0.3522484004497528,
+      "learning_rate": 7.344911367609302e-06,
+      "loss": 0.0238,
+      "step": 1294
+    },
+    {
+      "epoch": 0.41093784862533156,
+      "grad_norm": 0.4004107713699341,
+      "learning_rate": 7.340016316080565e-06,
+      "loss": 0.0232,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4112551751493666,
+      "grad_norm": 0.4026682674884796,
+      "learning_rate": 7.335118391043501e-06,
+      "loss": 0.0244,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4115725016734016,
+      "grad_norm": 0.30499836802482605,
+      "learning_rate": 7.330217598512696e-06,
+      "loss": 0.0218,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4118898281974366,
+      "grad_norm": 0.38302183151245117,
+      "learning_rate": 7.3253139445062535e-06,
+      "loss": 0.024,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4122071547214716,
+      "grad_norm": 0.3766491413116455,
+      "learning_rate": 7.320407435045798e-06,
+      "loss": 0.0249,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4125244812455066,
+      "grad_norm": 0.3521221876144409,
+      "learning_rate": 7.315498076156453e-06,
+      "loss": 0.0218,
+      "step": 1300
+    },
+    {
+      "epoch": 0.4128418077695416,
+      "grad_norm": 0.37246978282928467,
+      "learning_rate": 7.310585873866849e-06,
+      "loss": 0.0224,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4131591342935766,
+      "grad_norm": 0.5604162812232971,
+      "learning_rate": 7.3056708342090995e-06,
+      "loss": 0.0174,
+      "step": 1302
+    },
+    {
+      "epoch": 0.41347646081761164,
+      "grad_norm": 0.3549817204475403,
+      "learning_rate": 7.300752963218811e-06,
+      "loss": 0.0238,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4137937873416466,
+      "grad_norm": 0.6549107432365417,
+      "learning_rate": 7.295832266935059e-06,
+      "loss": 0.0252,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4141111138656816,
+      "grad_norm": 0.3568093180656433,
+      "learning_rate": 7.2909087514003925e-06,
+      "loss": 0.0246,
+      "step": 1305
+    },
+    {
+      "epoch": 0.41442844038971666,
+      "grad_norm": 0.41675353050231934,
+      "learning_rate": 7.285982422660823e-06,
+      "loss": 0.0225,
+      "step": 1306
+    },
+    {
+      "epoch": 0.41474576691375165,
+      "grad_norm": 0.37039777636528015,
+      "learning_rate": 7.281053286765816e-06,
+      "loss": 0.0263,
+      "step": 1307
+    },
+    {
+      "epoch": 0.41506309343778663,
+      "grad_norm": 0.46817561984062195,
+      "learning_rate": 7.276121349768282e-06,
+      "loss": 0.0201,
+      "step": 1308
+    },
+    {
+      "epoch": 0.4153804199618217,
+      "grad_norm": 0.38498789072036743,
+      "learning_rate": 7.2711866177245715e-06,
+      "loss": 0.0201,
+      "step": 1309
+    },
+    {
+      "epoch": 0.41569774648585667,
+      "grad_norm": 0.4292822480201721,
+      "learning_rate": 7.266249096694471e-06,
+      "loss": 0.0249,
+      "step": 1310
+    },
+    {
+      "epoch": 0.41601507300989166,
+      "grad_norm": 0.4322212338447571,
+      "learning_rate": 7.2613087927411885e-06,
+      "loss": 0.0266,
+      "step": 1311
+    },
+    {
+      "epoch": 0.41633239953392664,
+      "grad_norm": 0.4073107838630676,
+      "learning_rate": 7.256365711931351e-06,
+      "loss": 0.0207,
+      "step": 1312
+    },
+    {
+      "epoch": 0.4166497260579617,
+      "grad_norm": 0.42184391617774963,
+      "learning_rate": 7.251419860334994e-06,
+      "loss": 0.0224,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4169670525819967,
+      "grad_norm": 0.395214319229126,
+      "learning_rate": 7.246471244025557e-06,
+      "loss": 0.0259,
+      "step": 1314
+    },
+    {
+      "epoch": 0.41728437910603167,
+      "grad_norm": 0.38156789541244507,
+      "learning_rate": 7.241519869079873e-06,
+      "loss": 0.0228,
+      "step": 1315
+    },
+    {
+      "epoch": 0.4176017056300667,
+      "grad_norm": 0.37018758058547974,
+      "learning_rate": 7.236565741578163e-06,
+      "loss": 0.0193,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4179190321541017,
+      "grad_norm": 0.29722678661346436,
+      "learning_rate": 7.231608867604031e-06,
+      "loss": 0.0223,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4182363586781367,
+      "grad_norm": 0.3752012550830841,
+      "learning_rate": 7.226649253244448e-06,
+      "loss": 0.0224,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4185536852021717,
+      "grad_norm": 0.5705034136772156,
+      "learning_rate": 7.221686904589754e-06,
+      "loss": 0.024,
+      "step": 1319
+    },
+    {
+      "epoch": 0.4188710117262067,
+      "grad_norm": 0.35023805499076843,
+      "learning_rate": 7.216721827733646e-06,
+      "loss": 0.0199,
+      "step": 1320
+    },
+    {
+      "epoch": 0.4191883382502417,
+      "grad_norm": 0.3987623155117035,
+      "learning_rate": 7.2117540287731715e-06,
+      "loss": 0.0201,
+      "step": 1321
+    },
+    {
+      "epoch": 0.4195056647742767,
+      "grad_norm": 0.31302085518836975,
+      "learning_rate": 7.206783513808721e-06,
+      "loss": 0.0212,
+      "step": 1322
+    },
+    {
+      "epoch": 0.41982299129831174,
+      "grad_norm": 0.46105459332466125,
+      "learning_rate": 7.201810288944018e-06,
+      "loss": 0.0249,
+      "step": 1323
+    },
+    {
+      "epoch": 0.42014031782234673,
+      "grad_norm": 0.5760583281517029,
+      "learning_rate": 7.1968343602861154e-06,
+      "loss": 0.0205,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4204576443463817,
+      "grad_norm": 0.37246954441070557,
+      "learning_rate": 7.191855733945388e-06,
+      "loss": 0.021,
+      "step": 1325
+    },
+    {
+      "epoch": 0.42077497087041676,
+      "grad_norm": 1.162356972694397,
+      "learning_rate": 7.186874416035519e-06,
+      "loss": 0.0227,
+      "step": 1326
+    },
+    {
+      "epoch": 0.42109229739445175,
+      "grad_norm": 4.277872562408447,
+      "learning_rate": 7.181890412673502e-06,
+      "loss": 0.0226,
+      "step": 1327
+    },
+    {
+      "epoch": 0.42140962391848674,
+      "grad_norm": 0.36676719784736633,
+      "learning_rate": 7.176903729979622e-06,
+      "loss": 0.0216,
+      "step": 1328
+    },
+    {
+      "epoch": 0.42172695044252173,
+      "grad_norm": 0.45080697536468506,
+      "learning_rate": 7.171914374077461e-06,
+      "loss": 0.0243,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4220442769665568,
+      "grad_norm": 0.4264984428882599,
+      "learning_rate": 7.1669223510938794e-06,
+      "loss": 0.0261,
+      "step": 1330
+    },
+    {
+      "epoch": 0.42236160349059176,
+      "grad_norm": 0.42336052656173706,
+      "learning_rate": 7.161927667159013e-06,
+      "loss": 0.0227,
+      "step": 1331
+    },
+    {
+      "epoch": 0.42267893001462675,
+      "grad_norm": 0.5430270433425903,
+      "learning_rate": 7.156930328406268e-06,
+      "loss": 0.0274,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4229962565386618,
+      "grad_norm": 0.5601271390914917,
+      "learning_rate": 7.151930340972306e-06,
+      "loss": 0.0243,
+      "step": 1333
+    },
+    {
+      "epoch": 0.4233135830626968,
+      "grad_norm": 0.34801235795021057,
+      "learning_rate": 7.146927710997047e-06,
+      "loss": 0.0239,
+      "step": 1334
+    },
+    {
+      "epoch": 0.4236309095867318,
+      "grad_norm": 0.9817331433296204,
+      "learning_rate": 7.141922444623649e-06,
+      "loss": 0.026,
+      "step": 1335
+    },
+    {
+      "epoch": 0.42394823611076676,
+      "grad_norm": 0.45608851313591003,
+      "learning_rate": 7.1369145479985135e-06,
+      "loss": 0.0267,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4242655626348018,
+      "grad_norm": 0.38039979338645935,
+      "learning_rate": 7.1319040272712705e-06,
+      "loss": 0.0259,
+      "step": 1337
+    },
+    {
+      "epoch": 0.4245828891588368,
+      "grad_norm": 0.39437422156333923,
+      "learning_rate": 7.126890888594768e-06,
+      "loss": 0.0266,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4249002156828718,
+      "grad_norm": 0.3424970805644989,
+      "learning_rate": 7.121875138125077e-06,
+      "loss": 0.0207,
+      "step": 1339
+    },
+    {
+      "epoch": 0.42521754220690683,
+      "grad_norm": 0.3377830684185028,
+      "learning_rate": 7.116856782021469e-06,
+      "loss": 0.0233,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4255348687309418,
+      "grad_norm": 0.36252957582473755,
+      "learning_rate": 7.111835826446417e-06,
+      "loss": 0.0284,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4258521952549768,
+      "grad_norm": 0.3533332347869873,
+      "learning_rate": 7.106812277565587e-06,
+      "loss": 0.0241,
+      "step": 1342
+    },
+    {
+      "epoch": 0.42616952177901185,
+      "grad_norm": 0.37393268942832947,
+      "learning_rate": 7.101786141547829e-06,
+      "loss": 0.025,
+      "step": 1343
+    },
+    {
+      "epoch": 0.42648684830304684,
+      "grad_norm": 0.4678236246109009,
+      "learning_rate": 7.0967574245651684e-06,
+      "loss": 0.0253,
+      "step": 1344
+    },
+    {
+      "epoch": 0.42680417482708183,
+      "grad_norm": 0.39569568634033203,
+      "learning_rate": 7.091726132792802e-06,
+      "loss": 0.0239,
+      "step": 1345
+    },
+    {
+      "epoch": 0.4271215013511168,
+      "grad_norm": 0.5878258943557739,
+      "learning_rate": 7.08669227240909e-06,
+      "loss": 0.0246,
+      "step": 1346
+    },
+    {
+      "epoch": 0.42743882787515186,
+      "grad_norm": 0.42878949642181396,
+      "learning_rate": 7.0816558495955435e-06,
+      "loss": 0.0231,
+      "step": 1347
+    },
+    {
+      "epoch": 0.42775615439918685,
+      "grad_norm": 0.4272514581680298,
+      "learning_rate": 7.076616870536821e-06,
+      "loss": 0.0198,
+      "step": 1348
+    },
+    {
+      "epoch": 0.42807348092322184,
+      "grad_norm": 0.3287626802921295,
+      "learning_rate": 7.07157534142072e-06,
+      "loss": 0.0233,
+      "step": 1349
+    },
+    {
+      "epoch": 0.4283908074472569,
+      "grad_norm": 0.3793589770793915,
+      "learning_rate": 7.0665312684381705e-06,
+      "loss": 0.0218,
+      "step": 1350
+    },
+    {
+      "epoch": 0.42870813397129187,
+      "grad_norm": 0.7487128376960754,
+      "learning_rate": 7.0614846577832274e-06,
+      "loss": 0.0238,
+      "step": 1351
+    },
+    {
+      "epoch": 0.42902546049532686,
+      "grad_norm": 0.42150741815567017,
+      "learning_rate": 7.056435515653059e-06,
+      "loss": 0.0215,
+      "step": 1352
+    },
+    {
+      "epoch": 0.42934278701936185,
+      "grad_norm": 0.4220375418663025,
+      "learning_rate": 7.0513838482479424e-06,
+      "loss": 0.0247,
+      "step": 1353
+    },
+    {
+      "epoch": 0.4296601135433969,
+      "grad_norm": 0.4421738088130951,
+      "learning_rate": 7.046329661771257e-06,
+      "loss": 0.0258,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4299774400674319,
+      "grad_norm": 0.5215237736701965,
+      "learning_rate": 7.041272962429478e-06,
+      "loss": 0.0194,
+      "step": 1355
+    },
+    {
+      "epoch": 0.43029476659146687,
+      "grad_norm": 0.7795076966285706,
+      "learning_rate": 7.036213756432158e-06,
+      "loss": 0.0216,
+      "step": 1356
+    },
+    {
+      "epoch": 0.4306120931155019,
+      "grad_norm": 0.4417571425437927,
+      "learning_rate": 7.03115204999194e-06,
+      "loss": 0.0265,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4309294196395369,
+      "grad_norm": 0.4062870144844055,
+      "learning_rate": 7.026087849324527e-06,
+      "loss": 0.0214,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4312467461635719,
+      "grad_norm": 0.4482879340648651,
+      "learning_rate": 7.021021160648689e-06,
+      "loss": 0.0211,
+      "step": 1359
+    },
+    {
+      "epoch": 0.43156407268760694,
+      "grad_norm": 0.5520749688148499,
+      "learning_rate": 7.0159519901862515e-06,
+      "loss": 0.0259,
+      "step": 1360
+    },
+    {
+      "epoch": 0.4318813992116419,
+      "grad_norm": 0.3796919584274292,
+      "learning_rate": 7.010880344162087e-06,
+      "loss": 0.0219,
+      "step": 1361
+    },
+    {
+      "epoch": 0.4321987257356769,
+      "grad_norm": 0.4787737727165222,
+      "learning_rate": 7.0058062288041105e-06,
+      "loss": 0.0253,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4325160522597119,
+      "grad_norm": 0.6855406761169434,
+      "learning_rate": 7.000729650343264e-06,
+      "loss": 0.0256,
+      "step": 1363
+    },
+    {
+      "epoch": 0.43283337878374695,
+      "grad_norm": 0.38763228058815,
+      "learning_rate": 6.995650615013516e-06,
+      "loss": 0.0232,
+      "step": 1364
+    },
+    {
+      "epoch": 0.43315070530778194,
+      "grad_norm": 0.3584664463996887,
+      "learning_rate": 6.990569129051858e-06,
+      "loss": 0.0214,
+      "step": 1365
+    },
+    {
+      "epoch": 0.4334680318318169,
+      "grad_norm": 0.39888814091682434,
+      "learning_rate": 6.985485198698281e-06,
+      "loss": 0.0255,
+      "step": 1366
+    },
+    {
+      "epoch": 0.43378535835585197,
+      "grad_norm": 0.43585047125816345,
+      "learning_rate": 6.980398830195785e-06,
+      "loss": 0.0253,
+      "step": 1367
+    },
+    {
+      "epoch": 0.43410268487988696,
+      "grad_norm": 0.3997935354709625,
+      "learning_rate": 6.975310029790361e-06,
+      "loss": 0.0225,
+      "step": 1368
+    },
+    {
+      "epoch": 0.43442001140392195,
+      "grad_norm": 2.98087739944458,
+      "learning_rate": 6.9702188037309855e-06,
+      "loss": 0.0222,
+      "step": 1369
+    },
+    {
+      "epoch": 0.43473733792795693,
+      "grad_norm": 0.5320636630058289,
+      "learning_rate": 6.965125158269619e-06,
+      "loss": 0.0207,
+      "step": 1370
+    },
+    {
+      "epoch": 0.435054664451992,
+      "grad_norm": 0.3632018566131592,
+      "learning_rate": 6.9600290996611855e-06,
+      "loss": 0.0245,
+      "step": 1371
+    },
+    {
+      "epoch": 0.43537199097602697,
+      "grad_norm": 0.4211314916610718,
+      "learning_rate": 6.954930634163577e-06,
+      "loss": 0.0254,
+      "step": 1372
+    },
+    {
+      "epoch": 0.43568931750006196,
+      "grad_norm": 0.39384958148002625,
+      "learning_rate": 6.94982976803764e-06,
+      "loss": 0.0251,
+      "step": 1373
+    },
+    {
+      "epoch": 0.436006644024097,
+      "grad_norm": 0.37354591488838196,
+      "learning_rate": 6.944726507547169e-06,
+      "loss": 0.0242,
+      "step": 1374
+    },
+    {
+      "epoch": 0.436323970548132,
+      "grad_norm": 0.4330283999443054,
+      "learning_rate": 6.939620858958899e-06,
+      "loss": 0.0254,
+      "step": 1375
+    },
+    {
+      "epoch": 0.436641297072167,
+      "grad_norm": 0.3911290466785431,
+      "learning_rate": 6.934512828542498e-06,
+      "loss": 0.0233,
+      "step": 1376
+    },
+    {
+      "epoch": 0.436958623596202,
+      "grad_norm": 0.36045026779174805,
+      "learning_rate": 6.929402422570558e-06,
+      "loss": 0.0243,
+      "step": 1377
+    },
+    {
+      "epoch": 0.437275950120237,
+      "grad_norm": 0.33701351284980774,
+      "learning_rate": 6.924289647318586e-06,
+      "loss": 0.0232,
+      "step": 1378
+    },
+    {
+      "epoch": 0.437593276644272,
+      "grad_norm": 0.40197089314460754,
+      "learning_rate": 6.919174509065003e-06,
+      "loss": 0.0229,
+      "step": 1379
+    },
+    {
+      "epoch": 0.437910603168307,
+      "grad_norm": 0.41575607657432556,
+      "learning_rate": 6.914057014091128e-06,
+      "loss": 0.0259,
+      "step": 1380
+    },
+    {
+      "epoch": 0.43822792969234203,
+      "grad_norm": 0.31211093068122864,
+      "learning_rate": 6.908937168681176e-06,
+      "loss": 0.02,
+      "step": 1381
+    },
+    {
+      "epoch": 0.438545256216377,
+      "grad_norm": 0.345258504152298,
+      "learning_rate": 6.903814979122249e-06,
+      "loss": 0.0214,
+      "step": 1382
+    },
+    {
+      "epoch": 0.438862582740412,
+      "grad_norm": 0.361907958984375,
+      "learning_rate": 6.898690451704324e-06,
+      "loss": 0.0231,
+      "step": 1383
+    },
+    {
+      "epoch": 0.43917990926444705,
+      "grad_norm": 0.41244417428970337,
+      "learning_rate": 6.893563592720254e-06,
+      "loss": 0.0246,
+      "step": 1384
+    },
+    {
+      "epoch": 0.43949723578848204,
+      "grad_norm": 1.380765676498413,
+      "learning_rate": 6.888434408465751e-06,
+      "loss": 0.0226,
+      "step": 1385
+    },
+    {
+      "epoch": 0.43981456231251703,
+      "grad_norm": 0.34231051802635193,
+      "learning_rate": 6.883302905239383e-06,
+      "loss": 0.0224,
+      "step": 1386
+    },
+    {
+      "epoch": 0.4401318888365521,
+      "grad_norm": 0.5770206451416016,
+      "learning_rate": 6.878169089342568e-06,
+      "loss": 0.0258,
+      "step": 1387
+    },
+    {
+      "epoch": 0.44044921536058707,
+      "grad_norm": 0.4780910015106201,
+      "learning_rate": 6.873032967079562e-06,
+      "loss": 0.0241,
+      "step": 1388
+    },
+    {
+      "epoch": 0.44076654188462205,
+      "grad_norm": 0.40991711616516113,
+      "learning_rate": 6.867894544757452e-06,
+      "loss": 0.0239,
+      "step": 1389
+    },
+    {
+      "epoch": 0.44108386840865704,
+      "grad_norm": 0.40073180198669434,
+      "learning_rate": 6.862753828686151e-06,
+      "loss": 0.0212,
+      "step": 1390
+    },
+    {
+      "epoch": 0.4414011949326921,
+      "grad_norm": 0.730179488658905,
+      "learning_rate": 6.85761082517839e-06,
+      "loss": 0.0263,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4417185214567271,
+      "grad_norm": 0.42350083589553833,
+      "learning_rate": 6.85246554054971e-06,
+      "loss": 0.0189,
+      "step": 1392
+    },
+    {
+      "epoch": 0.44203584798076206,
+      "grad_norm": 0.4064870774745941,
+      "learning_rate": 6.847317981118446e-06,
+      "loss": 0.021,
+      "step": 1393
+    },
+    {
+      "epoch": 0.4423531745047971,
+      "grad_norm": 0.39409124851226807,
+      "learning_rate": 6.842168153205734e-06,
+      "loss": 0.0217,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4426705010288321,
+      "grad_norm": 0.31239157915115356,
+      "learning_rate": 6.837016063135491e-06,
+      "loss": 0.0206,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4429878275528671,
+      "grad_norm": 0.6803058981895447,
+      "learning_rate": 6.831861717234416e-06,
+      "loss": 0.026,
+      "step": 1396
+    },
+    {
+      "epoch": 0.4433051540769021,
+      "grad_norm": 0.3440152704715729,
+      "learning_rate": 6.8267051218319766e-06,
+      "loss": 0.0219,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4436224806009371,
+      "grad_norm": 0.42716649174690247,
+      "learning_rate": 6.821546283260398e-06,
+      "loss": 0.021,
+      "step": 1398
+    },
+    {
+      "epoch": 0.4439398071249721,
+      "grad_norm": 0.6273257732391357,
+      "learning_rate": 6.816385207854668e-06,
+      "loss": 0.0263,
+      "step": 1399
+    },
+    {
+      "epoch": 0.4442571336490071,
+      "grad_norm": 0.394522100687027,
+      "learning_rate": 6.811221901952514e-06,
+      "loss": 0.0216,
+      "step": 1400
+    },
+    {
+      "epoch": 0.44457446017304214,
+      "grad_norm": 0.6693580746650696,
+      "learning_rate": 6.806056371894409e-06,
+      "loss": 0.0214,
+      "step": 1401
+    },
+    {
+      "epoch": 0.44489178669707713,
+      "grad_norm": 0.34796902537345886,
+      "learning_rate": 6.800888624023552e-06,
+      "loss": 0.0251,
+      "step": 1402
+    },
+    {
+      "epoch": 0.4452091132211121,
+      "grad_norm": 0.502963125705719,
+      "learning_rate": 6.795718664685869e-06,
+      "loss": 0.0236,
+      "step": 1403
+    },
+    {
+      "epoch": 0.44552643974514716,
+      "grad_norm": 0.3246331512928009,
+      "learning_rate": 6.7905465002299976e-06,
+      "loss": 0.021,
+      "step": 1404
+    },
+    {
+      "epoch": 0.44584376626918215,
+      "grad_norm": 0.33192336559295654,
+      "learning_rate": 6.7853721370072865e-06,
+      "loss": 0.0218,
+      "step": 1405
+    },
+    {
+      "epoch": 0.44616109279321714,
+      "grad_norm": 0.3408469259738922,
+      "learning_rate": 6.780195581371785e-06,
+      "loss": 0.0181,
+      "step": 1406
+    },
+    {
+      "epoch": 0.44647841931725213,
+      "grad_norm": 0.34763434529304504,
+      "learning_rate": 6.775016839680231e-06,
+      "loss": 0.0192,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4467957458412872,
+      "grad_norm": 0.3652612268924713,
+      "learning_rate": 6.76983591829205e-06,
+      "loss": 0.022,
+      "step": 1408
+    },
+    {
+      "epoch": 0.44711307236532216,
+      "grad_norm": 0.3457341492176056,
+      "learning_rate": 6.7646528235693445e-06,
+      "loss": 0.0215,
+      "step": 1409
+    },
+    {
+      "epoch": 0.44743039888935715,
+      "grad_norm": 0.5752133131027222,
+      "learning_rate": 6.759467561876885e-06,
+      "loss": 0.0207,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4477477254133922,
+      "grad_norm": 1.4128748178482056,
+      "learning_rate": 6.754280139582099e-06,
+      "loss": 0.0204,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4480650519374272,
+      "grad_norm": 0.4586191177368164,
+      "learning_rate": 6.749090563055075e-06,
+      "loss": 0.024,
+      "step": 1412
+    },
+    {
+      "epoch": 0.44838237846146217,
+      "grad_norm": 0.3749358654022217,
+      "learning_rate": 6.743898838668539e-06,
+      "loss": 0.0235,
+      "step": 1413
+    },
+    {
+      "epoch": 0.44869970498549716,
+      "grad_norm": 0.5007901191711426,
+      "learning_rate": 6.73870497279786e-06,
+      "loss": 0.0201,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4490170315095322,
+      "grad_norm": 0.3986184895038605,
+      "learning_rate": 6.733508971821037e-06,
+      "loss": 0.0226,
+      "step": 1415
+    },
+    {
+      "epoch": 0.4493343580335672,
+      "grad_norm": 0.42544475197792053,
+      "learning_rate": 6.7283108421186835e-06,
+      "loss": 0.0233,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4496516845576022,
+      "grad_norm": 0.3931899070739746,
+      "learning_rate": 6.723110590074038e-06,
+      "loss": 0.0216,
+      "step": 1417
+    },
+    {
+      "epoch": 0.4499690110816372,
+      "grad_norm": 0.2761775553226471,
+      "learning_rate": 6.7179082220729355e-06,
+      "loss": 0.0173,
+      "step": 1418
+    },
+    {
+      "epoch": 0.4502863376056722,
+      "grad_norm": 0.3823276162147522,
+      "learning_rate": 6.712703744503815e-06,
+      "loss": 0.0203,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4506036641297072,
+      "grad_norm": 0.514716386795044,
+      "learning_rate": 6.707497163757705e-06,
+      "loss": 0.024,
+      "step": 1420
+    },
+    {
+      "epoch": 0.45092099065374225,
+      "grad_norm": 0.3098562955856323,
+      "learning_rate": 6.702288486228215e-06,
+      "loss": 0.0191,
+      "step": 1421
+    },
+    {
+      "epoch": 0.45123831717777724,
+      "grad_norm": 0.4600263237953186,
+      "learning_rate": 6.697077718311533e-06,
+      "loss": 0.0245,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4515556437018122,
+      "grad_norm": 0.4740245044231415,
+      "learning_rate": 6.691864866406407e-06,
+      "loss": 0.0262,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4518729702258472,
+      "grad_norm": 0.400704026222229,
+      "learning_rate": 6.686649936914151e-06,
+      "loss": 0.0256,
+      "step": 1424
+    },
+    {
+      "epoch": 0.45219029674988226,
+      "grad_norm": 0.48187628388404846,
+      "learning_rate": 6.681432936238628e-06,
+      "loss": 0.0232,
+      "step": 1425
+    },
+    {
+      "epoch": 0.45250762327391725,
+      "grad_norm": 0.5575533509254456,
+      "learning_rate": 6.676213870786243e-06,
+      "loss": 0.0264,
+      "step": 1426
+    },
+    {
+      "epoch": 0.45282494979795224,
+      "grad_norm": 0.44394588470458984,
+      "learning_rate": 6.6709927469659385e-06,
+      "loss": 0.0241,
+      "step": 1427
+    },
+    {
+      "epoch": 0.4531422763219873,
+      "grad_norm": 0.7182941436767578,
+      "learning_rate": 6.6657695711891836e-06,
+      "loss": 0.0204,
+      "step": 1428
+    },
+    {
+      "epoch": 0.45345960284602227,
+      "grad_norm": 0.6201386451721191,
+      "learning_rate": 6.660544349869967e-06,
+      "loss": 0.0209,
+      "step": 1429
+    },
+    {
+      "epoch": 0.45377692937005726,
+      "grad_norm": 0.4228150248527527,
+      "learning_rate": 6.655317089424791e-06,
+      "loss": 0.0223,
+      "step": 1430
+    },
+    {
+      "epoch": 0.45409425589409225,
+      "grad_norm": 0.31781598925590515,
+      "learning_rate": 6.650087796272659e-06,
+      "loss": 0.02,
+      "step": 1431
+    },
+    {
+      "epoch": 0.4544115824181273,
+      "grad_norm": 0.4284209907054901,
+      "learning_rate": 6.644856476835074e-06,
+      "loss": 0.0231,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4547289089421623,
+      "grad_norm": 0.46374157071113586,
+      "learning_rate": 6.639623137536023e-06,
+      "loss": 0.0225,
+      "step": 1433
+    },
+    {
+      "epoch": 0.45504623546619727,
+      "grad_norm": 0.44458121061325073,
+      "learning_rate": 6.634387784801977e-06,
+      "loss": 0.0222,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4553635619902323,
+      "grad_norm": 0.40935683250427246,
+      "learning_rate": 6.6291504250618775e-06,
+      "loss": 0.0197,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4556808885142673,
+      "grad_norm": 0.33413878083229065,
+      "learning_rate": 6.623911064747133e-06,
+      "loss": 0.0213,
+      "step": 1436
+    },
+    {
+      "epoch": 0.4559982150383023,
+      "grad_norm": 0.6049994230270386,
+      "learning_rate": 6.618669710291607e-06,
+      "loss": 0.031,
+      "step": 1437
+    },
+    {
+      "epoch": 0.45631554156233733,
+      "grad_norm": 0.5244348645210266,
+      "learning_rate": 6.613426368131608e-06,
+      "loss": 0.0209,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4566328680863723,
+      "grad_norm": 1.0642354488372803,
+      "learning_rate": 6.608181044705893e-06,
+      "loss": 0.0225,
+      "step": 1439
+    },
+    {
+      "epoch": 0.4569501946104073,
+      "grad_norm": 0.3568953573703766,
+      "learning_rate": 6.602933746455646e-06,
+      "loss": 0.023,
+      "step": 1440
+    },
+    {
+      "epoch": 0.4572675211344423,
+      "grad_norm": 0.411978155374527,
+      "learning_rate": 6.597684479824481e-06,
+      "loss": 0.0255,
+      "step": 1441
+    },
+    {
+      "epoch": 0.45758484765847734,
+      "grad_norm": 0.4599054753780365,
+      "learning_rate": 6.592433251258423e-06,
+      "loss": 0.0238,
+      "step": 1442
+    },
+    {
+      "epoch": 0.45790217418251233,
+      "grad_norm": 0.38168469071388245,
+      "learning_rate": 6.58718006720591e-06,
+      "loss": 0.0222,
+      "step": 1443
+    },
+    {
+      "epoch": 0.4582195007065473,
+      "grad_norm": 0.38115110993385315,
+      "learning_rate": 6.581924934117783e-06,
+      "loss": 0.0205,
+      "step": 1444
+    },
+    {
+      "epoch": 0.45853682723058237,
+      "grad_norm": 0.33050525188446045,
+      "learning_rate": 6.576667858447272e-06,
+      "loss": 0.0217,
+      "step": 1445
+    },
+    {
+      "epoch": 0.45885415375461736,
+      "grad_norm": 0.3156895339488983,
+      "learning_rate": 6.571408846649996e-06,
+      "loss": 0.0208,
+      "step": 1446
+    },
+    {
+      "epoch": 0.45917148027865234,
+      "grad_norm": 0.41157788038253784,
+      "learning_rate": 6.56614790518395e-06,
+      "loss": 0.0256,
+      "step": 1447
+    },
+    {
+      "epoch": 0.45948880680268733,
+      "grad_norm": 0.39286094903945923,
+      "learning_rate": 6.560885040509499e-06,
+      "loss": 0.0243,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4598061333267224,
+      "grad_norm": 0.4856831729412079,
+      "learning_rate": 6.555620259089369e-06,
+      "loss": 0.0235,
+      "step": 1449
+    },
+    {
+      "epoch": 0.46012345985075737,
+      "grad_norm": 0.4514527916908264,
+      "learning_rate": 6.55035356738864e-06,
+      "loss": 0.0235,
+      "step": 1450
+    },
+    {
+      "epoch": 0.46044078637479235,
+      "grad_norm": 0.4845404028892517,
+      "learning_rate": 6.545084971874738e-06,
+      "loss": 0.0195,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4607581128988274,
+      "grad_norm": 0.5388951897621155,
+      "learning_rate": 6.539814479017428e-06,
+      "loss": 0.0243,
+      "step": 1452
+    },
+    {
+      "epoch": 0.4610754394228624,
+      "grad_norm": 0.5087718367576599,
+      "learning_rate": 6.534542095288802e-06,
+      "loss": 0.0197,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4613927659468974,
+      "grad_norm": 0.8699803948402405,
+      "learning_rate": 6.529267827163277e-06,
+      "loss": 0.0226,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4617100924709324,
+      "grad_norm": 0.37085598707199097,
+      "learning_rate": 6.523991681117584e-06,
+      "loss": 0.0222,
+      "step": 1455
+    },
+    {
+      "epoch": 0.4620274189949674,
+      "grad_norm": 0.3756875693798065,
+      "learning_rate": 6.518713663630756e-06,
+      "loss": 0.0256,
+      "step": 1456
+    },
+    {
+      "epoch": 0.4623447455190024,
+      "grad_norm": 0.4006965160369873,
+      "learning_rate": 6.513433781184131e-06,
+      "loss": 0.0255,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4626620720430374,
+      "grad_norm": 0.38266339898109436,
+      "learning_rate": 6.508152040261329e-06,
+      "loss": 0.0233,
+      "step": 1458
+    },
+    {
+      "epoch": 0.46297939856707243,
+      "grad_norm": 0.3170243501663208,
+      "learning_rate": 6.502868447348257e-06,
+      "loss": 0.0169,
+      "step": 1459
+    },
+    {
+      "epoch": 0.4632967250911074,
+      "grad_norm": 0.3887498080730438,
+      "learning_rate": 6.497583008933097e-06,
+      "loss": 0.0224,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4636140516151424,
+      "grad_norm": 0.42865467071533203,
+      "learning_rate": 6.492295731506295e-06,
+      "loss": 0.0235,
+      "step": 1461
+    },
+    {
+      "epoch": 0.46393137813917745,
+      "grad_norm": 0.6379486322402954,
+      "learning_rate": 6.487006621560556e-06,
+      "loss": 0.0211,
+      "step": 1462
+    },
+    {
+      "epoch": 0.46424870466321244,
+      "grad_norm": 0.3886668086051941,
+      "learning_rate": 6.481715685590836e-06,
+      "loss": 0.0228,
+      "step": 1463
+    },
+    {
+      "epoch": 0.46456603118724743,
+      "grad_norm": 2.327456474304199,
+      "learning_rate": 6.476422930094333e-06,
+      "loss": 0.0186,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4648833577112824,
+      "grad_norm": 0.3477025330066681,
+      "learning_rate": 6.4711283615704755e-06,
+      "loss": 0.022,
+      "step": 1465
+    },
+    {
+      "epoch": 0.46520068423531746,
+      "grad_norm": 0.5507770776748657,
+      "learning_rate": 6.4658319865209276e-06,
+      "loss": 0.0208,
+      "step": 1466
+    },
+    {
+      "epoch": 0.46551801075935245,
+      "grad_norm": 0.439270555973053,
+      "learning_rate": 6.460533811449562e-06,
+      "loss": 0.0215,
+      "step": 1467
+    },
+    {
+      "epoch": 0.46583533728338744,
+      "grad_norm": 0.38275206089019775,
+      "learning_rate": 6.455233842862467e-06,
+      "loss": 0.0171,
+      "step": 1468
+    },
+    {
+      "epoch": 0.4661526638074225,
+      "grad_norm": 0.4057563543319702,
+      "learning_rate": 6.449932087267932e-06,
+      "loss": 0.0185,
+      "step": 1469
+    },
+    {
+      "epoch": 0.4664699903314575,
+      "grad_norm": 0.38364097476005554,
+      "learning_rate": 6.4446285511764395e-06,
+      "loss": 0.017,
+      "step": 1470
+    },
+    {
+      "epoch": 0.46678731685549246,
+      "grad_norm": 0.3925889730453491,
+      "learning_rate": 6.439323241100663e-06,
+      "loss": 0.0181,
+      "step": 1471
+    },
+    {
+      "epoch": 0.4671046433795275,
+      "grad_norm": 0.39403218030929565,
+      "learning_rate": 6.434016163555452e-06,
+      "loss": 0.0201,
+      "step": 1472
+    },
+    {
+      "epoch": 0.4674219699035625,
+      "grad_norm": 0.40882232785224915,
+      "learning_rate": 6.428707325057823e-06,
+      "loss": 0.0233,
+      "step": 1473
+    },
+    {
+      "epoch": 0.4677392964275975,
+      "grad_norm": 0.44441258907318115,
+      "learning_rate": 6.423396732126957e-06,
+      "loss": 0.0236,
+      "step": 1474
+    },
+    {
+      "epoch": 0.4680566229516325,
+      "grad_norm": 0.6555232405662537,
+      "learning_rate": 6.4180843912841925e-06,
+      "loss": 0.0205,
+      "step": 1475
+    },
+    {
+      "epoch": 0.4683739494756675,
+      "grad_norm": 0.3322071135044098,
+      "learning_rate": 6.412770309053011e-06,
+      "loss": 0.0191,
+      "step": 1476
+    },
+    {
+      "epoch": 0.4686912759997025,
+      "grad_norm": 0.46424970030784607,
+      "learning_rate": 6.407454491959036e-06,
+      "loss": 0.0219,
+      "step": 1477
+    },
+    {
+      "epoch": 0.4690086025237375,
+      "grad_norm": 0.5539832711219788,
+      "learning_rate": 6.402136946530014e-06,
+      "loss": 0.0225,
+      "step": 1478
+    },
+    {
+      "epoch": 0.46932592904777254,
+      "grad_norm": 1.0112429857254028,
+      "learning_rate": 6.396817679295823e-06,
+      "loss": 0.0217,
+      "step": 1479
+    },
+    {
+      "epoch": 0.4696432555718075,
+      "grad_norm": 0.5675466060638428,
+      "learning_rate": 6.391496696788448e-06,
+      "loss": 0.0245,
+      "step": 1480
+    },
+    {
+      "epoch": 0.4699605820958425,
+      "grad_norm": 0.4311521053314209,
+      "learning_rate": 6.386174005541986e-06,
+      "loss": 0.0235,
+      "step": 1481
+    },
+    {
+      "epoch": 0.4702779086198775,
+      "grad_norm": 0.3322450816631317,
+      "learning_rate": 6.38084961209263e-06,
+      "loss": 0.0212,
+      "step": 1482
+    },
+    {
+      "epoch": 0.47059523514391255,
+      "grad_norm": 0.3807448446750641,
+      "learning_rate": 6.375523522978659e-06,
+      "loss": 0.0197,
+      "step": 1483
+    },
+    {
+      "epoch": 0.47091256166794754,
+      "grad_norm": 0.32353633642196655,
+      "learning_rate": 6.3701957447404426e-06,
+      "loss": 0.0204,
+      "step": 1484
+    },
+    {
+      "epoch": 0.4712298881919825,
+      "grad_norm": 0.37132465839385986,
+      "learning_rate": 6.364866283920418e-06,
+      "loss": 0.0209,
+      "step": 1485
+    },
+    {
+      "epoch": 0.47154721471601757,
+      "grad_norm": 0.37269091606140137,
+      "learning_rate": 6.359535147063092e-06,
+      "loss": 0.0212,
+      "step": 1486
+    },
+    {
+      "epoch": 0.47186454124005256,
+      "grad_norm": 0.5509238839149475,
+      "learning_rate": 6.354202340715027e-06,
+      "loss": 0.025,
+      "step": 1487
+    },
+    {
+      "epoch": 0.47218186776408755,
+      "grad_norm": 0.3995044231414795,
+      "learning_rate": 6.348867871424837e-06,
+      "loss": 0.0191,
+      "step": 1488
+    },
+    {
+      "epoch": 0.4724991942881226,
+      "grad_norm": 0.6040096282958984,
+      "learning_rate": 6.34353174574318e-06,
+      "loss": 0.0213,
+      "step": 1489
+    },
+    {
+      "epoch": 0.4728165208121576,
+      "grad_norm": 0.44989195466041565,
+      "learning_rate": 6.338193970222745e-06,
+      "loss": 0.0184,
+      "step": 1490
+    },
+    {
+      "epoch": 0.47313384733619257,
+      "grad_norm": 0.9319937825202942,
+      "learning_rate": 6.332854551418247e-06,
+      "loss": 0.0241,
+      "step": 1491
+    },
+    {
+      "epoch": 0.47345117386022756,
+      "grad_norm": 0.4940868318080902,
+      "learning_rate": 6.327513495886423e-06,
+      "loss": 0.0184,
+      "step": 1492
+    },
+    {
+      "epoch": 0.4737685003842626,
+      "grad_norm": 0.42922475934028625,
+      "learning_rate": 6.322170810186013e-06,
+      "loss": 0.0186,
+      "step": 1493
+    },
+    {
+      "epoch": 0.4740858269082976,
+      "grad_norm": 0.39056891202926636,
+      "learning_rate": 6.316826500877763e-06,
+      "loss": 0.0206,
+      "step": 1494
+    },
+    {
+      "epoch": 0.4744031534323326,
+      "grad_norm": 0.44848722219467163,
+      "learning_rate": 6.3114805745244156e-06,
+      "loss": 0.022,
+      "step": 1495
+    },
+    {
+      "epoch": 0.4747204799563676,
+      "grad_norm": 0.5576976537704468,
+      "learning_rate": 6.306133037690693e-06,
+      "loss": 0.0209,
+      "step": 1496
+    },
+    {
+      "epoch": 0.4750378064804026,
+      "grad_norm": 0.4400976300239563,
+      "learning_rate": 6.300783896943299e-06,
+      "loss": 0.0204,
+      "step": 1497
+    },
+    {
+      "epoch": 0.4753551330044376,
+      "grad_norm": 0.43480384349823,
+      "learning_rate": 6.295433158850906e-06,
+      "loss": 0.0203,
+      "step": 1498
+    },
+    {
+      "epoch": 0.4756724595284726,
+      "grad_norm": 0.6192718744277954,
+      "learning_rate": 6.290080829984147e-06,
+      "loss": 0.0201,
+      "step": 1499
+    },
+    {
+      "epoch": 0.47598978605250764,
+      "grad_norm": 0.3313943147659302,
+      "learning_rate": 6.284726916915611e-06,
+      "loss": 0.0227,
+      "step": 1500
+    },
+    {
+      "epoch": 0.4763071125765426,
+      "grad_norm": 0.7278152108192444,
+      "learning_rate": 6.27937142621983e-06,
+      "loss": 0.0223,
+      "step": 1501
+    },
+    {
+      "epoch": 0.4766244391005776,
+      "grad_norm": 0.45221155881881714,
+      "learning_rate": 6.274014364473274e-06,
+      "loss": 0.0194,
+      "step": 1502
+    },
+    {
+      "epoch": 0.47694176562461266,
+      "grad_norm": 0.34876394271850586,
+      "learning_rate": 6.268655738254342e-06,
+      "loss": 0.0211,
+      "step": 1503
+    },
+    {
+      "epoch": 0.47725909214864765,
+      "grad_norm": 0.31355229020118713,
+      "learning_rate": 6.263295554143354e-06,
+      "loss": 0.0173,
+      "step": 1504
+    },
+    {
+      "epoch": 0.47757641867268263,
+      "grad_norm": 0.3405880630016327,
+      "learning_rate": 6.257933818722544e-06,
+      "loss": 0.0221,
+      "step": 1505
+    },
+    {
+      "epoch": 0.4778937451967177,
+      "grad_norm": 0.4402998685836792,
+      "learning_rate": 6.252570538576048e-06,
+      "loss": 0.0194,
+      "step": 1506
+    },
+    {
+      "epoch": 0.47821107172075267,
+      "grad_norm": 0.39403659105300903,
+      "learning_rate": 6.247205720289907e-06,
+      "loss": 0.025,
+      "step": 1507
+    },
+    {
+      "epoch": 0.47852839824478766,
+      "grad_norm": 0.6735592484474182,
+      "learning_rate": 6.241839370452041e-06,
+      "loss": 0.0269,
+      "step": 1508
+    },
+    {
+      "epoch": 0.47884572476882264,
+      "grad_norm": 0.45241019129753113,
+      "learning_rate": 6.236471495652255e-06,
+      "loss": 0.0181,
+      "step": 1509
+    },
+    {
+      "epoch": 0.4791630512928577,
+      "grad_norm": 0.38393908739089966,
+      "learning_rate": 6.2311021024822274e-06,
+      "loss": 0.0202,
+      "step": 1510
+    },
+    {
+      "epoch": 0.4794803778168927,
+      "grad_norm": 0.6353452205657959,
+      "learning_rate": 6.225731197535501e-06,
+      "loss": 0.0232,
+      "step": 1511
+    },
+    {
+      "epoch": 0.47979770434092767,
+      "grad_norm": 0.34179943799972534,
+      "learning_rate": 6.220358787407473e-06,
+      "loss": 0.0172,
+      "step": 1512
+    },
+    {
+      "epoch": 0.4801150308649627,
+      "grad_norm": 1.228218913078308,
+      "learning_rate": 6.2149848786953905e-06,
+      "loss": 0.0182,
+      "step": 1513
+    },
+    {
+      "epoch": 0.4804323573889977,
+      "grad_norm": 0.500300943851471,
+      "learning_rate": 6.209609477998339e-06,
+      "loss": 0.0238,
+      "step": 1514
+    },
+    {
+      "epoch": 0.4807496839130327,
+      "grad_norm": 0.6081173419952393,
+      "learning_rate": 6.204232591917241e-06,
+      "loss": 0.0201,
+      "step": 1515
+    },
+    {
+      "epoch": 0.4810670104370677,
+      "grad_norm": 0.46144214272499084,
+      "learning_rate": 6.198854227054838e-06,
+      "loss": 0.02,
+      "step": 1516
+    },
+    {
+      "epoch": 0.4813843369611027,
+      "grad_norm": 0.460873007774353,
+      "learning_rate": 6.19347439001569e-06,
+      "loss": 0.0254,
+      "step": 1517
+    },
+    {
+      "epoch": 0.4817016634851377,
+      "grad_norm": 0.4825160801410675,
+      "learning_rate": 6.188093087406162e-06,
+      "loss": 0.0212,
+      "step": 1518
+    },
+    {
+      "epoch": 0.4820189900091727,
+      "grad_norm": 0.43737396597862244,
+      "learning_rate": 6.1827103258344225e-06,
+      "loss": 0.0225,
+      "step": 1519
+    },
+    {
+      "epoch": 0.48233631653320774,
+      "grad_norm": 0.4665667414665222,
+      "learning_rate": 6.177326111910429e-06,
+      "loss": 0.0208,
+      "step": 1520
+    },
+    {
+      "epoch": 0.48265364305724273,
+      "grad_norm": 0.3708983063697815,
+      "learning_rate": 6.171940452245923e-06,
+      "loss": 0.0193,
+      "step": 1521
+    },
+    {
+      "epoch": 0.4829709695812777,
+      "grad_norm": 0.5465094447135925,
+      "learning_rate": 6.166553353454424e-06,
+      "loss": 0.022,
+      "step": 1522
+    },
+    {
+      "epoch": 0.48328829610531276,
+      "grad_norm": 0.3993150293827057,
+      "learning_rate": 6.161164822151213e-06,
+      "loss": 0.0223,
+      "step": 1523
+    },
+    {
+      "epoch": 0.48360562262934775,
+      "grad_norm": 0.9811030626296997,
+      "learning_rate": 6.1557748649533355e-06,
+      "loss": 0.0217,
+      "step": 1524
+    },
+    {
+      "epoch": 0.48392294915338274,
+      "grad_norm": 0.46173638105392456,
+      "learning_rate": 6.150383488479584e-06,
+      "loss": 0.0222,
+      "step": 1525
+    },
+    {
+      "epoch": 0.48424027567741773,
+      "grad_norm": 0.45639386773109436,
+      "learning_rate": 6.144990699350498e-06,
+      "loss": 0.021,
+      "step": 1526
+    },
+    {
+      "epoch": 0.4845576022014528,
+      "grad_norm": 0.5327286720275879,
+      "learning_rate": 6.13959650418835e-06,
+      "loss": 0.0194,
+      "step": 1527
+    },
+    {
+      "epoch": 0.48487492872548776,
+      "grad_norm": 1.6407263278961182,
+      "learning_rate": 6.134200909617135e-06,
+      "loss": 0.0216,
+      "step": 1528
+    },
+    {
+      "epoch": 0.48519225524952275,
+      "grad_norm": 0.3510459363460541,
+      "learning_rate": 6.128803922262573e-06,
+      "loss": 0.0193,
+      "step": 1529
+    },
+    {
+      "epoch": 0.4855095817735578,
+      "grad_norm": 0.41622984409332275,
+      "learning_rate": 6.12340554875209e-06,
+      "loss": 0.022,
+      "step": 1530
+    },
+    {
+      "epoch": 0.4858269082975928,
+      "grad_norm": 0.4226149916648865,
+      "learning_rate": 6.118005795714814e-06,
+      "loss": 0.0237,
+      "step": 1531
+    },
+    {
+      "epoch": 0.4861442348216278,
+      "grad_norm": 0.4828113317489624,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.0201,
+      "step": 1532
+    },
+    {
+      "epoch": 0.4864615613456628,
+      "grad_norm": 1.8855443000793457,
+      "learning_rate": 6.107202177584872e-06,
+      "loss": 0.0256,
+      "step": 1533
+    },
+    {
+      "epoch": 0.4867788878696978,
+      "grad_norm": 0.4086664617061615,
+      "learning_rate": 6.101798325758902e-06,
+      "loss": 0.0185,
+      "step": 1534
+    },
+    {
+      "epoch": 0.4870962143937328,
+      "grad_norm": 0.8287175297737122,
+      "learning_rate": 6.0963931209395165e-06,
+      "loss": 0.0266,
+      "step": 1535
+    },
+    {
+      "epoch": 0.4874135409177678,
+      "grad_norm": 0.44071587920188904,
+      "learning_rate": 6.090986569764237e-06,
+      "loss": 0.0247,
+      "step": 1536
+    },
+    {
+      "epoch": 0.48773086744180283,
+      "grad_norm": 0.5581042170524597,
+      "learning_rate": 6.085578678872233e-06,
+      "loss": 0.0252,
+      "step": 1537
+    },
+    {
+      "epoch": 0.4880481939658378,
+      "grad_norm": 0.47154611349105835,
+      "learning_rate": 6.08016945490432e-06,
+      "loss": 0.0217,
+      "step": 1538
+    },
+    {
+      "epoch": 0.4883655204898728,
+      "grad_norm": 0.5383191704750061,
+      "learning_rate": 6.0747589045029545e-06,
+      "loss": 0.0218,
+      "step": 1539
+    },
+    {
+      "epoch": 0.48868284701390785,
+      "grad_norm": 0.44215118885040283,
+      "learning_rate": 6.069347034312217e-06,
+      "loss": 0.0229,
+      "step": 1540
+    },
+    {
+      "epoch": 0.48900017353794284,
+      "grad_norm": 0.4712156057357788,
+      "learning_rate": 6.063933850977811e-06,
+      "loss": 0.0219,
+      "step": 1541
+    },
+    {
+      "epoch": 0.48931750006197783,
+      "grad_norm": 0.43483635783195496,
+      "learning_rate": 6.058519361147055e-06,
+      "loss": 0.0221,
+      "step": 1542
+    },
+    {
+      "epoch": 0.4896348265860128,
+      "grad_norm": 1.0566209554672241,
+      "learning_rate": 6.053103571468866e-06,
+      "loss": 0.0217,
+      "step": 1543
+    },
+    {
+      "epoch": 0.48995215311004786,
+      "grad_norm": 0.690752387046814,
+      "learning_rate": 6.047686488593761e-06,
+      "loss": 0.0239,
+      "step": 1544
+    },
+    {
+      "epoch": 0.49026947963408285,
+      "grad_norm": 0.4572145938873291,
+      "learning_rate": 6.042268119173844e-06,
+      "loss": 0.0207,
+      "step": 1545
+    },
+    {
+      "epoch": 0.49058680615811784,
+      "grad_norm": 0.32279157638549805,
+      "learning_rate": 6.036848469862801e-06,
+      "loss": 0.0186,
+      "step": 1546
+    },
+    {
+      "epoch": 0.4909041326821529,
+      "grad_norm": 0.7983816862106323,
+      "learning_rate": 6.031427547315889e-06,
+      "loss": 0.0233,
+      "step": 1547
+    },
+    {
+      "epoch": 0.49122145920618787,
+      "grad_norm": 0.3936874270439148,
+      "learning_rate": 6.026005358189922e-06,
+      "loss": 0.0206,
+      "step": 1548
+    },
+    {
+      "epoch": 0.49153878573022286,
+      "grad_norm": 4.518249988555908,
+      "learning_rate": 6.020581909143279e-06,
+      "loss": 0.0244,
+      "step": 1549
+    },
+    {
+      "epoch": 0.4918561122542579,
+      "grad_norm": 0.48843395709991455,
+      "learning_rate": 6.015157206835881e-06,
+      "loss": 0.0187,
+      "step": 1550
+    },
+    {
+      "epoch": 0.4921734387782929,
+      "grad_norm": 0.4149928092956543,
+      "learning_rate": 6.009731257929189e-06,
+      "loss": 0.0233,
+      "step": 1551
+    },
+    {
+      "epoch": 0.4924907653023279,
+      "grad_norm": 0.5580644011497498,
+      "learning_rate": 6.004304069086196e-06,
+      "loss": 0.0234,
+      "step": 1552
+    },
+    {
+      "epoch": 0.49280809182636287,
+      "grad_norm": 0.3997536301612854,
+      "learning_rate": 5.998875646971414e-06,
+      "loss": 0.0252,
+      "step": 1553
+    },
+    {
+      "epoch": 0.4931254183503979,
+      "grad_norm": 0.7834483981132507,
+      "learning_rate": 5.993445998250873e-06,
+      "loss": 0.0217,
+      "step": 1554
+    },
+    {
+      "epoch": 0.4934427448744329,
+      "grad_norm": 0.407829612493515,
+      "learning_rate": 5.988015129592108e-06,
+      "loss": 0.0223,
+      "step": 1555
+    },
+    {
+      "epoch": 0.4937600713984679,
+      "grad_norm": 0.44270750880241394,
+      "learning_rate": 5.982583047664151e-06,
+      "loss": 0.0213,
+      "step": 1556
+    },
+    {
+      "epoch": 0.49407739792250294,
+      "grad_norm": 0.4069477617740631,
+      "learning_rate": 5.977149759137528e-06,
+      "loss": 0.0241,
+      "step": 1557
+    },
+    {
+      "epoch": 0.4943947244465379,
+      "grad_norm": 0.4178355038166046,
+      "learning_rate": 5.97171527068424e-06,
+      "loss": 0.0254,
+      "step": 1558
+    },
+    {
+      "epoch": 0.4947120509705729,
+      "grad_norm": 0.35646361112594604,
+      "learning_rate": 5.9662795889777666e-06,
+      "loss": 0.0189,
+      "step": 1559
+    },
+    {
+      "epoch": 0.4950293774946079,
+      "grad_norm": 0.4340394139289856,
+      "learning_rate": 5.960842720693052e-06,
+      "loss": 0.0217,
+      "step": 1560
+    },
+    {
+      "epoch": 0.49534670401864295,
+      "grad_norm": 0.38348206877708435,
+      "learning_rate": 5.9554046725064975e-06,
+      "loss": 0.025,
+      "step": 1561
+    },
+    {
+      "epoch": 0.49566403054267794,
+      "grad_norm": 0.5364371538162231,
+      "learning_rate": 5.949965451095952e-06,
+      "loss": 0.0197,
+      "step": 1562
+    },
+    {
+      "epoch": 0.4959813570667129,
+      "grad_norm": 0.4698741137981415,
+      "learning_rate": 5.944525063140703e-06,
+      "loss": 0.0205,
+      "step": 1563
+    },
+    {
+      "epoch": 0.49629868359074797,
+      "grad_norm": 0.45352524518966675,
+      "learning_rate": 5.939083515321475e-06,
+      "loss": 0.021,
+      "step": 1564
+    },
+    {
+      "epoch": 0.49661601011478296,
+      "grad_norm": 0.4164828956127167,
+      "learning_rate": 5.933640814320417e-06,
+      "loss": 0.0216,
+      "step": 1565
+    },
+    {
+      "epoch": 0.49693333663881795,
+      "grad_norm": 0.4591805040836334,
+      "learning_rate": 5.92819696682109e-06,
+      "loss": 0.0208,
+      "step": 1566
+    },
+    {
+      "epoch": 0.497250663162853,
+      "grad_norm": 1.3027271032333374,
+      "learning_rate": 5.922751979508464e-06,
+      "loss": 0.0229,
+      "step": 1567
+    },
+    {
+      "epoch": 0.497567989686888,
+      "grad_norm": 0.4845893979072571,
+      "learning_rate": 5.917305859068912e-06,
+      "loss": 0.0215,
+      "step": 1568
+    },
+    {
+      "epoch": 0.49788531621092297,
+      "grad_norm": 0.4081409275531769,
+      "learning_rate": 5.911858612190196e-06,
+      "loss": 0.0238,
+      "step": 1569
+    },
+    {
+      "epoch": 0.49820264273495796,
+      "grad_norm": 0.8910696506500244,
+      "learning_rate": 5.906410245561459e-06,
+      "loss": 0.0244,
+      "step": 1570
+    },
+    {
+      "epoch": 0.498519969258993,
+      "grad_norm": 0.3903805613517761,
+      "learning_rate": 5.900960765873223e-06,
+      "loss": 0.022,
+      "step": 1571
+    },
+    {
+      "epoch": 0.498837295783028,
+      "grad_norm": 0.6407555341720581,
+      "learning_rate": 5.895510179817376e-06,
+      "loss": 0.021,
+      "step": 1572
+    },
+    {
+      "epoch": 0.499154622307063,
+      "grad_norm": 0.3809860646724701,
+      "learning_rate": 5.890058494087163e-06,
+      "loss": 0.022,
+      "step": 1573
+    },
+    {
+      "epoch": 0.499471948831098,
+      "grad_norm": 0.8184378743171692,
+      "learning_rate": 5.884605715377179e-06,
+      "loss": 0.0211,
+      "step": 1574
+    },
+    {
+      "epoch": 0.499789275355133,
+      "grad_norm": 0.3677801787853241,
+      "learning_rate": 5.879151850383364e-06,
+      "loss": 0.0203,
+      "step": 1575
+    },
+    {
+      "epoch": 0.500106601879168,
+      "grad_norm": 0.38399162888526917,
+      "learning_rate": 5.873696905802988e-06,
+      "loss": 0.022,
+      "step": 1576
+    },
+    {
+      "epoch": 0.500423928403203,
+      "grad_norm": 0.5217903852462769,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.0197,
+      "step": 1577
+    },
+    {
+      "epoch": 0.500741254927238,
+      "grad_norm": 0.4459115266799927,
+      "learning_rate": 5.862783804678271e-06,
+      "loss": 0.02,
+      "step": 1578
+    },
+    {
+      "epoch": 0.501058581451273,
+      "grad_norm": 0.37872937321662903,
+      "learning_rate": 5.857325661535069e-06,
+      "loss": 0.025,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5013759079753081,
+      "grad_norm": 0.4076676666736603,
+      "learning_rate": 5.8518664656075704e-06,
+      "loss": 0.0227,
+      "step": 1580
+    },
+    {
+      "epoch": 0.501693234499343,
+      "grad_norm": 1.12064790725708,
+      "learning_rate": 5.846406223599597e-06,
+      "loss": 0.0194,
+      "step": 1581
+    },
+    {
+      "epoch": 0.502010561023378,
+      "grad_norm": 0.4336223006248474,
+      "learning_rate": 5.84094494221625e-06,
+      "loss": 0.0214,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5023278875474131,
+      "grad_norm": 0.6931948065757751,
+      "learning_rate": 5.835482628163909e-06,
+      "loss": 0.0204,
+      "step": 1583
+    },
+    {
+      "epoch": 0.502645214071448,
+      "grad_norm": 0.6110202074050903,
+      "learning_rate": 5.830019288150222e-06,
+      "loss": 0.0217,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5029625405954831,
+      "grad_norm": 0.47392725944519043,
+      "learning_rate": 5.824554928884094e-06,
+      "loss": 0.0243,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5032798671195181,
+      "grad_norm": 0.5562558174133301,
+      "learning_rate": 5.819089557075689e-06,
+      "loss": 0.0212,
+      "step": 1586
+    },
+    {
+      "epoch": 0.503597193643553,
+      "grad_norm": 0.30724456906318665,
+      "learning_rate": 5.8136231794364075e-06,
+      "loss": 0.0178,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5039145201675881,
+      "grad_norm": 0.41055262088775635,
+      "learning_rate": 5.808155802678885e-06,
+      "loss": 0.0198,
+      "step": 1588
+    },
+    {
+      "epoch": 0.504231846691623,
+      "grad_norm": 0.49053382873535156,
+      "learning_rate": 5.802687433516989e-06,
+      "loss": 0.0239,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5045491732156581,
+      "grad_norm": 0.4997645914554596,
+      "learning_rate": 5.797218078665801e-06,
+      "loss": 0.0211,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5048664997396931,
+      "grad_norm": 0.4033114016056061,
+      "learning_rate": 5.791747744841615e-06,
+      "loss": 0.0198,
+      "step": 1591
+    },
+    {
+      "epoch": 0.505183826263728,
+      "grad_norm": 0.678817868232727,
+      "learning_rate": 5.786276438761928e-06,
+      "loss": 0.0201,
+      "step": 1592
+    },
+    {
+      "epoch": 0.5055011527877631,
+      "grad_norm": 0.36383405327796936,
+      "learning_rate": 5.780804167145426e-06,
+      "loss": 0.0174,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5058184793117981,
+      "grad_norm": 0.5318219661712646,
+      "learning_rate": 5.775330936711989e-06,
+      "loss": 0.0231,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5061358058358331,
+      "grad_norm": 0.5423073172569275,
+      "learning_rate": 5.769856754182668e-06,
+      "loss": 0.0217,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5064531323598681,
+      "grad_norm": 0.47389331459999084,
+      "learning_rate": 5.764381626279685e-06,
+      "loss": 0.0165,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5067704588839032,
+      "grad_norm": 0.4404006600379944,
+      "learning_rate": 5.7589055597264235e-06,
+      "loss": 0.0231,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5070877854079381,
+      "grad_norm": 0.35410892963409424,
+      "learning_rate": 5.753428561247416e-06,
+      "loss": 0.0224,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5074051119319731,
+      "grad_norm": 0.4725388288497925,
+      "learning_rate": 5.747950637568347e-06,
+      "loss": 0.0229,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5077224384560082,
+      "grad_norm": 0.3531433045864105,
+      "learning_rate": 5.742471795416028e-06,
+      "loss": 0.0191,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5080397649800431,
+      "grad_norm": 0.47159817814826965,
+      "learning_rate": 5.736992041518407e-06,
+      "loss": 0.0263,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5083570915040782,
+      "grad_norm": 0.3754887878894806,
+      "learning_rate": 5.7315113826045444e-06,
+      "loss": 0.024,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5086744180281132,
+      "grad_norm": 0.44312143325805664,
+      "learning_rate": 5.726029825404617e-06,
+      "loss": 0.0185,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5089917445521481,
+      "grad_norm": 0.3287600874900818,
+      "learning_rate": 5.720547376649901e-06,
+      "loss": 0.0202,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5093090710761832,
+      "grad_norm": 0.3902679979801178,
+      "learning_rate": 5.715064043072771e-06,
+      "loss": 0.0212,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5096263976002182,
+      "grad_norm": 0.4121951758861542,
+      "learning_rate": 5.7095798314066865e-06,
+      "loss": 0.0186,
+      "step": 1606
+    },
+    {
+      "epoch": 0.5099437241242532,
+      "grad_norm": 0.5619876384735107,
+      "learning_rate": 5.7040947483861845e-06,
+      "loss": 0.0227,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5102610506482882,
+      "grad_norm": 0.36465877294540405,
+      "learning_rate": 5.698608800746871e-06,
+      "loss": 0.0227,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5105783771723231,
+      "grad_norm": 0.4445246756076813,
+      "learning_rate": 5.693121995225417e-06,
+      "loss": 0.0205,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5108957036963582,
+      "grad_norm": 0.5697159171104431,
+      "learning_rate": 5.687634338559544e-06,
+      "loss": 0.0235,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5112130302203932,
+      "grad_norm": 0.4753643274307251,
+      "learning_rate": 5.6821458374880235e-06,
+      "loss": 0.0201,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5115303567444281,
+      "grad_norm": 0.8460371494293213,
+      "learning_rate": 5.6766564987506564e-06,
+      "loss": 0.0216,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5118476832684632,
+      "grad_norm": 0.4608052372932434,
+      "learning_rate": 5.671166329088278e-06,
+      "loss": 0.0245,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5121650097924982,
+      "grad_norm": 0.5140283107757568,
+      "learning_rate": 5.665675335242743e-06,
+      "loss": 0.0141,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5124823363165332,
+      "grad_norm": 0.545911431312561,
+      "learning_rate": 5.660183523956916e-06,
+      "loss": 0.0217,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5127996628405682,
+      "grad_norm": 0.3706926107406616,
+      "learning_rate": 5.654690901974667e-06,
+      "loss": 0.0228,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5131169893646033,
+      "grad_norm": 0.3601500689983368,
+      "learning_rate": 5.649197476040863e-06,
+      "loss": 0.0223,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5134343158886382,
+      "grad_norm": 1.260022759437561,
+      "learning_rate": 5.643703252901355e-06,
+      "loss": 0.0201,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5137516424126732,
+      "grad_norm": 0.35067862272262573,
+      "learning_rate": 5.638208239302975e-06,
+      "loss": 0.0205,
+      "step": 1619
+    },
+    {
+      "epoch": 0.5140689689367083,
+      "grad_norm": 0.3738689720630646,
+      "learning_rate": 5.632712441993523e-06,
+      "loss": 0.021,
+      "step": 1620
+    },
+    {
+      "epoch": 0.5143862954607432,
+      "grad_norm": 0.38874855637550354,
+      "learning_rate": 5.627215867721767e-06,
+      "loss": 0.0189,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5147036219847783,
+      "grad_norm": 0.4734913110733032,
+      "learning_rate": 5.621718523237427e-06,
+      "loss": 0.02,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5150209485088133,
+      "grad_norm": 0.3535957932472229,
+      "learning_rate": 5.616220415291162e-06,
+      "loss": 0.0164,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5153382750328482,
+      "grad_norm": 0.393298864364624,
+      "learning_rate": 5.610721550634579e-06,
+      "loss": 0.0194,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5156556015568833,
+      "grad_norm": 0.3934926688671112,
+      "learning_rate": 5.605221936020208e-06,
+      "loss": 0.0207,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5159729280809182,
+      "grad_norm": 0.6171572208404541,
+      "learning_rate": 5.599721578201499e-06,
+      "loss": 0.0233,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5162902546049533,
+      "grad_norm": 0.30965059995651245,
+      "learning_rate": 5.594220483932821e-06,
+      "loss": 0.0189,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5166075811289883,
+      "grad_norm": 0.5120173096656799,
+      "learning_rate": 5.588718659969437e-06,
+      "loss": 0.0183,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5169249076530232,
+      "grad_norm": 0.30773499608039856,
+      "learning_rate": 5.5832161130675175e-06,
+      "loss": 0.0157,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5172422341770583,
+      "grad_norm": 0.5780503749847412,
+      "learning_rate": 5.5777128499841145e-06,
+      "loss": 0.0191,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5175595607010933,
+      "grad_norm": 0.6167354583740234,
+      "learning_rate": 5.57220887747716e-06,
+      "loss": 0.0216,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5178768872251283,
+      "grad_norm": 0.39777395129203796,
+      "learning_rate": 5.566704202305458e-06,
+      "loss": 0.0239,
+      "step": 1632
+    },
+    {
+      "epoch": 0.5181942137491633,
+      "grad_norm": 0.525771975517273,
+      "learning_rate": 5.561198831228676e-06,
+      "loss": 0.0185,
+      "step": 1633
+    },
+    {
+      "epoch": 0.5185115402731983,
+      "grad_norm": 0.3647536635398865,
+      "learning_rate": 5.555692771007331e-06,
+      "loss": 0.0222,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5188288667972333,
+      "grad_norm": 0.397408664226532,
+      "learning_rate": 5.550186028402795e-06,
+      "loss": 0.0207,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5191461933212683,
+      "grad_norm": 0.3332580029964447,
+      "learning_rate": 5.544678610177271e-06,
+      "loss": 0.0194,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5194635198453034,
+      "grad_norm": 0.36164405941963196,
+      "learning_rate": 5.539170523093794e-06,
+      "loss": 0.0201,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5197808463693383,
+      "grad_norm": 0.503025472164154,
+      "learning_rate": 5.533661773916221e-06,
+      "loss": 0.0218,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5200981728933733,
+      "grad_norm": 0.6972259283065796,
+      "learning_rate": 5.528152369409221e-06,
+      "loss": 0.0239,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5204154994174084,
+      "grad_norm": 0.3416927754878998,
+      "learning_rate": 5.522642316338268e-06,
+      "loss": 0.0168,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5207328259414433,
+      "grad_norm": 0.47474542260169983,
+      "learning_rate": 5.517131621469633e-06,
+      "loss": 0.021,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5210501524654784,
+      "grad_norm": 0.5929709076881409,
+      "learning_rate": 5.511620291570374e-06,
+      "loss": 0.0226,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5213674789895133,
+      "grad_norm": 0.43984341621398926,
+      "learning_rate": 5.50610833340833e-06,
+      "loss": 0.0177,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5216848055135483,
+      "grad_norm": 0.48030033707618713,
+      "learning_rate": 5.500595753752111e-06,
+      "loss": 0.0238,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5220021320375834,
+      "grad_norm": 0.3018196225166321,
+      "learning_rate": 5.495082559371089e-06,
+      "loss": 0.0179,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5223194585616183,
+      "grad_norm": 0.3184017241001129,
+      "learning_rate": 5.4895687570353916e-06,
+      "loss": 0.0185,
+      "step": 1646
+    },
+    {
+      "epoch": 0.5226367850856534,
+      "grad_norm": 0.3547445237636566,
+      "learning_rate": 5.484054353515896e-06,
+      "loss": 0.018,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5229541116096884,
+      "grad_norm": 0.3798380196094513,
+      "learning_rate": 5.478539355584214e-06,
+      "loss": 0.0214,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5232714381337233,
+      "grad_norm": 0.5518311858177185,
+      "learning_rate": 5.473023770012686e-06,
+      "loss": 0.0193,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5235887646577584,
+      "grad_norm": 0.4111924171447754,
+      "learning_rate": 5.467507603574379e-06,
+      "loss": 0.0215,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5239060911817934,
+      "grad_norm": 0.2993650734424591,
+      "learning_rate": 5.461990863043071e-06,
+      "loss": 0.0167,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5242234177058284,
+      "grad_norm": 0.6353009343147278,
+      "learning_rate": 5.456473555193242e-06,
+      "loss": 0.0188,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5245407442298634,
+      "grad_norm": 1.3412119150161743,
+      "learning_rate": 5.450955686800074e-06,
+      "loss": 0.0222,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5248580707538985,
+      "grad_norm": 0.36179670691490173,
+      "learning_rate": 5.445437264639433e-06,
+      "loss": 0.0176,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5251753972779334,
+      "grad_norm": 0.34636926651000977,
+      "learning_rate": 5.439918295487866e-06,
+      "loss": 0.0192,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5254927238019684,
+      "grad_norm": 0.44843757152557373,
+      "learning_rate": 5.434398786122592e-06,
+      "loss": 0.0178,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5258100503260035,
+      "grad_norm": 0.6889106035232544,
+      "learning_rate": 5.428878743321498e-06,
+      "loss": 0.024,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5261273768500384,
+      "grad_norm": 0.5556723475456238,
+      "learning_rate": 5.423358173863117e-06,
+      "loss": 0.0198,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5264447033740735,
+      "grad_norm": 0.594514787197113,
+      "learning_rate": 5.417837084526636e-06,
+      "loss": 0.0202,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5267620298981084,
+      "grad_norm": 0.42548370361328125,
+      "learning_rate": 5.412315482091877e-06,
+      "loss": 0.0216,
+      "step": 1660
+    },
+    {
+      "epoch": 0.5270793564221434,
+      "grad_norm": 0.3927461504936218,
+      "learning_rate": 5.406793373339292e-06,
+      "loss": 0.0202,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5273966829461785,
+      "grad_norm": 0.48913854360580444,
+      "learning_rate": 5.401270765049959e-06,
+      "loss": 0.0194,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5277140094702134,
+      "grad_norm": 0.46120521426200867,
+      "learning_rate": 5.395747664005562e-06,
+      "loss": 0.0188,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5280313359942485,
+      "grad_norm": 0.47164568305015564,
+      "learning_rate": 5.390224076988396e-06,
+      "loss": 0.0232,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5283486625182835,
+      "grad_norm": 0.31116214394569397,
+      "learning_rate": 5.384700010781352e-06,
+      "loss": 0.0173,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5286659890423184,
+      "grad_norm": 0.4031047821044922,
+      "learning_rate": 5.379175472167908e-06,
+      "loss": 0.0203,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5289833155663535,
+      "grad_norm": 0.3319724202156067,
+      "learning_rate": 5.373650467932122e-06,
+      "loss": 0.0188,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5293006420903885,
+      "grad_norm": 0.32883304357528687,
+      "learning_rate": 5.3681250048586246e-06,
+      "loss": 0.023,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5296179686144235,
+      "grad_norm": 0.3976210355758667,
+      "learning_rate": 5.3625990897326095e-06,
+      "loss": 0.0206,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5299352951384585,
+      "grad_norm": 0.4127482771873474,
+      "learning_rate": 5.357072729339825e-06,
+      "loss": 0.0209,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5302526216624935,
+      "grad_norm": 0.43644988536834717,
+      "learning_rate": 5.3515459304665685e-06,
+      "loss": 0.0221,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5305699481865285,
+      "grad_norm": 0.42127880454063416,
+      "learning_rate": 5.346018699899672e-06,
+      "loss": 0.0179,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5308872747105635,
+      "grad_norm": 0.4117668867111206,
+      "learning_rate": 5.3404910444265015e-06,
+      "loss": 0.0182,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5312046012345986,
+      "grad_norm": 0.3330513536930084,
+      "learning_rate": 5.3349629708349414e-06,
+      "loss": 0.0187,
+      "step": 1674
+    },
+    {
+      "epoch": 0.5315219277586335,
+      "grad_norm": 0.578879714012146,
+      "learning_rate": 5.329434485913393e-06,
+      "loss": 0.0199,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5318392542826685,
+      "grad_norm": 0.6576412320137024,
+      "learning_rate": 5.323905596450759e-06,
+      "loss": 0.0247,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5321565808067035,
+      "grad_norm": 0.4574038088321686,
+      "learning_rate": 5.3183763092364434e-06,
+      "loss": 0.0194,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5324739073307385,
+      "grad_norm": 0.4376643896102905,
+      "learning_rate": 5.312846631060333e-06,
+      "loss": 0.0207,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5327912338547736,
+      "grad_norm": 0.4596840441226959,
+      "learning_rate": 5.307316568712799e-06,
+      "loss": 0.0222,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5331085603788085,
+      "grad_norm": 0.43733111023902893,
+      "learning_rate": 5.301786128984683e-06,
+      "loss": 0.0215,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5334258869028435,
+      "grad_norm": 0.3890819847583771,
+      "learning_rate": 5.29625531866729e-06,
+      "loss": 0.0185,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5337432134268786,
+      "grad_norm": 0.3348178267478943,
+      "learning_rate": 5.290724144552379e-06,
+      "loss": 0.0189,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5340605399509135,
+      "grad_norm": 0.3784427344799042,
+      "learning_rate": 5.28519261343216e-06,
+      "loss": 0.0183,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5343778664749486,
+      "grad_norm": 0.3042350709438324,
+      "learning_rate": 5.2796607320992755e-06,
+      "loss": 0.0149,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5346951929989836,
+      "grad_norm": 0.4684559106826782,
+      "learning_rate": 5.274128507346801e-06,
+      "loss": 0.021,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5350125195230185,
+      "grad_norm": 0.373166024684906,
+      "learning_rate": 5.268595945968233e-06,
+      "loss": 0.0213,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5353298460470536,
+      "grad_norm": 0.3420615792274475,
+      "learning_rate": 5.263063054757486e-06,
+      "loss": 0.0195,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5356471725710886,
+      "grad_norm": 0.42656421661376953,
+      "learning_rate": 5.25752984050887e-06,
+      "loss": 0.0192,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5359644990951236,
+      "grad_norm": 0.3336891829967499,
+      "learning_rate": 5.251996310017101e-06,
+      "loss": 0.0187,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5362818256191586,
+      "grad_norm": 0.31349629163742065,
+      "learning_rate": 5.2464624700772775e-06,
+      "loss": 0.0182,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5365991521431936,
+      "grad_norm": 0.9121575951576233,
+      "learning_rate": 5.240928327484879e-06,
+      "loss": 0.0168,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5369164786672286,
+      "grad_norm": 0.4607718288898468,
+      "learning_rate": 5.23539388903576e-06,
+      "loss": 0.02,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5372338051912636,
+      "grad_norm": 0.6139616966247559,
+      "learning_rate": 5.229859161526136e-06,
+      "loss": 0.0197,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5375511317152986,
+      "grad_norm": 0.3302242159843445,
+      "learning_rate": 5.224324151752575e-06,
+      "loss": 0.0202,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5378684582393336,
+      "grad_norm": 0.5411401987075806,
+      "learning_rate": 5.218788866511996e-06,
+      "loss": 0.0171,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5381857847633686,
+      "grad_norm": 0.555778443813324,
+      "learning_rate": 5.213253312601654e-06,
+      "loss": 0.0177,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5385031112874036,
+      "grad_norm": 0.9068291187286377,
+      "learning_rate": 5.207717496819134e-06,
+      "loss": 0.0186,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5388204378114386,
+      "grad_norm": 0.5624876618385315,
+      "learning_rate": 5.202181425962342e-06,
+      "loss": 0.0197,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5391377643354737,
+      "grad_norm": 0.41694313287734985,
+      "learning_rate": 5.196645106829499e-06,
+      "loss": 0.0197,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5394550908595086,
+      "grad_norm": 0.3325691521167755,
+      "learning_rate": 5.191108546219128e-06,
+      "loss": 0.0173,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5397724173835436,
+      "grad_norm": 0.3463689386844635,
+      "learning_rate": 5.185571750930052e-06,
+      "loss": 0.0189,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5400897439075787,
+      "grad_norm": 0.4142729341983795,
+      "learning_rate": 5.1800347277613825e-06,
+      "loss": 0.0154,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5404070704316136,
+      "grad_norm": 0.5036169290542603,
+      "learning_rate": 5.174497483512506e-06,
+      "loss": 0.018,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5407243969556487,
+      "grad_norm": 0.3513205051422119,
+      "learning_rate": 5.168960024983084e-06,
+      "loss": 0.0183,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5410417234796837,
+      "grad_norm": 0.3592582941055298,
+      "learning_rate": 5.1634223589730405e-06,
+      "loss": 0.0166,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5413590500037186,
+      "grad_norm": 0.31134000420570374,
+      "learning_rate": 5.157884492282555e-06,
+      "loss": 0.0174,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5416763765277537,
+      "grad_norm": 0.41898345947265625,
+      "learning_rate": 5.152346431712053e-06,
+      "loss": 0.0204,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5419937030517887,
+      "grad_norm": 0.5624595284461975,
+      "learning_rate": 5.146808184062196e-06,
+      "loss": 0.021,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5423110295758237,
+      "grad_norm": 0.3373849093914032,
+      "learning_rate": 5.141269756133879e-06,
+      "loss": 0.0205,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5426283560998587,
+      "grad_norm": 0.42047062516212463,
+      "learning_rate": 5.135731154728215e-06,
+      "loss": 0.0216,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5429456826238936,
+      "grad_norm": 0.4230741262435913,
+      "learning_rate": 5.130192386646533e-06,
+      "loss": 0.0195,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5432630091479287,
+      "grad_norm": 0.7019308805465698,
+      "learning_rate": 5.1246534586903655e-06,
+      "loss": 0.0167,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5435803356719637,
+      "grad_norm": 0.41375410556793213,
+      "learning_rate": 5.1191143776614385e-06,
+      "loss": 0.0147,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5438976621959987,
+      "grad_norm": 0.3792034983634949,
+      "learning_rate": 5.113575150361671e-06,
+      "loss": 0.0196,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5442149887200337,
+      "grad_norm": 0.4139844477176666,
+      "learning_rate": 5.108035783593156e-06,
+      "loss": 0.0181,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5445323152440688,
+      "grad_norm": 0.3922460079193115,
+      "learning_rate": 5.1024962841581626e-06,
+      "loss": 0.0196,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5448496417681037,
+      "grad_norm": 0.5936586260795593,
+      "learning_rate": 5.096956658859122e-06,
+      "loss": 0.0177,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5451669682921387,
+      "grad_norm": 0.32270196080207825,
+      "learning_rate": 5.091416914498619e-06,
+      "loss": 0.0197,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5454842948161738,
+      "grad_norm": 0.6860949993133545,
+      "learning_rate": 5.085877057879383e-06,
+      "loss": 0.0191,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5458016213402087,
+      "grad_norm": 0.45622923970222473,
+      "learning_rate": 5.080337095804283e-06,
+      "loss": 0.0223,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5461189478642438,
+      "grad_norm": 0.61833256483078,
+      "learning_rate": 5.074797035076319e-06,
+      "loss": 0.0218,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5464362743882788,
+      "grad_norm": 0.2893732786178589,
+      "learning_rate": 5.06925688249861e-06,
+      "loss": 0.016,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5467536009123137,
+      "grad_norm": 0.46447500586509705,
+      "learning_rate": 5.0637166448743855e-06,
+      "loss": 0.0208,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5470709274363488,
+      "grad_norm": 0.37167614698410034,
+      "learning_rate": 5.0581763290069865e-06,
+      "loss": 0.0194,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5473882539603838,
+      "grad_norm": 0.4014522433280945,
+      "learning_rate": 5.052635941699843e-06,
+      "loss": 0.0193,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5477055804844188,
+      "grad_norm": 0.337545245885849,
+      "learning_rate": 5.047095489756476e-06,
+      "loss": 0.0186,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5480229070084538,
+      "grad_norm": 0.3818226158618927,
+      "learning_rate": 5.041554979980487e-06,
+      "loss": 0.019,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5483402335324887,
+      "grad_norm": 0.2889942228794098,
+      "learning_rate": 5.036014419175543e-06,
+      "loss": 0.0174,
+      "step": 1728
+    },
+    {
+      "epoch": 0.5486575600565238,
+      "grad_norm": 0.3546144366264343,
+      "learning_rate": 5.030473814145383e-06,
+      "loss": 0.0197,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5489748865805588,
+      "grad_norm": 0.4713887572288513,
+      "learning_rate": 5.024933171693791e-06,
+      "loss": 0.0172,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5492922131045938,
+      "grad_norm": 0.3125307559967041,
+      "learning_rate": 5.019392498624602e-06,
+      "loss": 0.017,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5496095396286288,
+      "grad_norm": 0.3542378544807434,
+      "learning_rate": 5.0138518017416895e-06,
+      "loss": 0.0209,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5499268661526638,
+      "grad_norm": 0.7511146664619446,
+      "learning_rate": 5.008311087848949e-06,
+      "loss": 0.0224,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5502441926766988,
+      "grad_norm": 0.34068095684051514,
+      "learning_rate": 5.002770363750307e-06,
+      "loss": 0.0183,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5505615192007338,
+      "grad_norm": 0.5346512198448181,
+      "learning_rate": 4.997229636249695e-06,
+      "loss": 0.0201,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5508788457247689,
+      "grad_norm": 0.45094403624534607,
+      "learning_rate": 4.9916889121510535e-06,
+      "loss": 0.0202,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5511961722488038,
+      "grad_norm": 0.4324469566345215,
+      "learning_rate": 4.986148198258314e-06,
+      "loss": 0.0217,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5515134987728388,
+      "grad_norm": 0.36036667227745056,
+      "learning_rate": 4.980607501375399e-06,
+      "loss": 0.0159,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5518308252968739,
+      "grad_norm": 0.34378546476364136,
+      "learning_rate": 4.9750668283062104e-06,
+      "loss": 0.0214,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5521481518209088,
+      "grad_norm": 0.38431471586227417,
+      "learning_rate": 4.969526185854619e-06,
+      "loss": 0.0225,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5524654783449439,
+      "grad_norm": 0.3103681206703186,
+      "learning_rate": 4.963985580824458e-06,
+      "loss": 0.0189,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5527828048689789,
+      "grad_norm": 0.33627772331237793,
+      "learning_rate": 4.958445020019516e-06,
+      "loss": 0.0142,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5531001313930138,
+      "grad_norm": 0.363440603017807,
+      "learning_rate": 4.9529045102435255e-06,
+      "loss": 0.0164,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5534174579170489,
+      "grad_norm": 0.5307464599609375,
+      "learning_rate": 4.947364058300159e-06,
+      "loss": 0.0201,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5537347844410838,
+      "grad_norm": 0.4576621651649475,
+      "learning_rate": 4.941823670993016e-06,
+      "loss": 0.0199,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5540521109651189,
+      "grad_norm": 0.3848884403705597,
+      "learning_rate": 4.936283355125616e-06,
+      "loss": 0.0215,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5543694374891539,
+      "grad_norm": 0.4116488993167877,
+      "learning_rate": 4.930743117501393e-06,
+      "loss": 0.0218,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5546867640131888,
+      "grad_norm": 0.3505142033100128,
+      "learning_rate": 4.9252029649236835e-06,
+      "loss": 0.0194,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5550040905372239,
+      "grad_norm": 0.3550426661968231,
+      "learning_rate": 4.919662904195719e-06,
+      "loss": 0.0208,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5553214170612589,
+      "grad_norm": 0.38079461455345154,
+      "learning_rate": 4.914122942120619e-06,
+      "loss": 0.021,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5556387435852939,
+      "grad_norm": 0.4393172562122345,
+      "learning_rate": 4.908583085501383e-06,
+      "loss": 0.017,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5559560701093289,
+      "grad_norm": 0.27365630865097046,
+      "learning_rate": 4.903043341140879e-06,
+      "loss": 0.0174,
+      "step": 1752
+    },
+    {
+      "epoch": 0.556273396633364,
+      "grad_norm": 0.3354533910751343,
+      "learning_rate": 4.897503715841838e-06,
+      "loss": 0.0202,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5565907231573989,
+      "grad_norm": 0.4844367206096649,
+      "learning_rate": 4.891964216406844e-06,
+      "loss": 0.0188,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5569080496814339,
+      "grad_norm": 0.3147219121456146,
+      "learning_rate": 4.886424849638333e-06,
+      "loss": 0.0155,
+      "step": 1755
+    },
+    {
+      "epoch": 0.557225376205469,
+      "grad_norm": 0.4211744964122772,
+      "learning_rate": 4.880885622338564e-06,
+      "loss": 0.0204,
+      "step": 1756
+    },
+    {
+      "epoch": 0.5575427027295039,
+      "grad_norm": 0.33566656708717346,
+      "learning_rate": 4.875346541309637e-06,
+      "loss": 0.0172,
+      "step": 1757
+    },
+    {
+      "epoch": 0.557860029253539,
+      "grad_norm": 0.4216957986354828,
+      "learning_rate": 4.8698076133534685e-06,
+      "loss": 0.0191,
+      "step": 1758
+    },
+    {
+      "epoch": 0.558177355777574,
+      "grad_norm": 0.3412401080131531,
+      "learning_rate": 4.864268845271786e-06,
+      "loss": 0.0166,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5584946823016089,
+      "grad_norm": 0.47509831190109253,
+      "learning_rate": 4.858730243866123e-06,
+      "loss": 0.0181,
+      "step": 1760
+    },
+    {
+      "epoch": 0.558812008825644,
+      "grad_norm": 0.36668065190315247,
+      "learning_rate": 4.853191815937805e-06,
+      "loss": 0.0211,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5591293353496789,
+      "grad_norm": 0.540259599685669,
+      "learning_rate": 4.8476535682879475e-06,
+      "loss": 0.0169,
+      "step": 1762
+    },
+    {
+      "epoch": 0.559446661873714,
+      "grad_norm": 0.38041362166404724,
+      "learning_rate": 4.842115507717446e-06,
+      "loss": 0.0186,
+      "step": 1763
+    },
+    {
+      "epoch": 0.559763988397749,
+      "grad_norm": 0.4882596731185913,
+      "learning_rate": 4.83657764102696e-06,
+      "loss": 0.0221,
+      "step": 1764
+    },
+    {
+      "epoch": 0.5600813149217839,
+      "grad_norm": 0.308431476354599,
+      "learning_rate": 4.831039975016919e-06,
+      "loss": 0.0167,
+      "step": 1765
+    },
+    {
+      "epoch": 0.560398641445819,
+      "grad_norm": 0.2967756986618042,
+      "learning_rate": 4.825502516487497e-06,
+      "loss": 0.0177,
+      "step": 1766
+    },
+    {
+      "epoch": 0.560715967969854,
+      "grad_norm": 0.3402934670448303,
+      "learning_rate": 4.81996527223862e-06,
+      "loss": 0.0154,
+      "step": 1767
+    },
+    {
+      "epoch": 0.561033294493889,
+      "grad_norm": 0.3286725580692291,
+      "learning_rate": 4.814428249069949e-06,
+      "loss": 0.019,
+      "step": 1768
+    },
+    {
+      "epoch": 0.561350621017924,
+      "grad_norm": 0.5096375942230225,
+      "learning_rate": 4.808891453780874e-06,
+      "loss": 0.0243,
+      "step": 1769
+    },
+    {
+      "epoch": 0.561667947541959,
+      "grad_norm": 0.29528793692588806,
+      "learning_rate": 4.803354893170503e-06,
+      "loss": 0.0171,
+      "step": 1770
+    },
+    {
+      "epoch": 0.561985274065994,
+      "grad_norm": 0.3357822597026825,
+      "learning_rate": 4.797818574037659e-06,
+      "loss": 0.0196,
+      "step": 1771
+    },
+    {
+      "epoch": 0.562302600590029,
+      "grad_norm": 0.31477004289627075,
+      "learning_rate": 4.792282503180867e-06,
+      "loss": 0.0201,
+      "step": 1772
+    },
+    {
+      "epoch": 0.5626199271140641,
+      "grad_norm": 0.42078790068626404,
+      "learning_rate": 4.786746687398347e-06,
+      "loss": 0.0214,
+      "step": 1773
+    },
+    {
+      "epoch": 0.562937253638099,
+      "grad_norm": 0.3314855992794037,
+      "learning_rate": 4.781211133488004e-06,
+      "loss": 0.0167,
+      "step": 1774
+    },
+    {
+      "epoch": 0.563254580162134,
+      "grad_norm": 0.319039911031723,
+      "learning_rate": 4.775675848247427e-06,
+      "loss": 0.0175,
+      "step": 1775
+    },
+    {
+      "epoch": 0.5635719066861691,
+      "grad_norm": 0.4608457088470459,
+      "learning_rate": 4.770140838473867e-06,
+      "loss": 0.0155,
+      "step": 1776
+    },
+    {
+      "epoch": 0.563889233210204,
+      "grad_norm": 0.36440718173980713,
+      "learning_rate": 4.7646061109642416e-06,
+      "loss": 0.0216,
+      "step": 1777
+    },
+    {
+      "epoch": 0.564206559734239,
+      "grad_norm": 0.40081971883773804,
+      "learning_rate": 4.7590716725151236e-06,
+      "loss": 0.02,
+      "step": 1778
+    },
+    {
+      "epoch": 0.564523886258274,
+      "grad_norm": 0.28073081374168396,
+      "learning_rate": 4.753537529922724e-06,
+      "loss": 0.0171,
+      "step": 1779
+    },
+    {
+      "epoch": 0.564841212782309,
+      "grad_norm": 0.35247814655303955,
+      "learning_rate": 4.748003689982901e-06,
+      "loss": 0.0146,
+      "step": 1780
+    },
+    {
+      "epoch": 0.5651585393063441,
+      "grad_norm": 0.3507697880268097,
+      "learning_rate": 4.742470159491131e-06,
+      "loss": 0.0176,
+      "step": 1781
+    },
+    {
+      "epoch": 0.565475865830379,
+      "grad_norm": 0.43584078550338745,
+      "learning_rate": 4.736936945242515e-06,
+      "loss": 0.0181,
+      "step": 1782
+    },
+    {
+      "epoch": 0.565793192354414,
+      "grad_norm": 0.30565229058265686,
+      "learning_rate": 4.731404054031766e-06,
+      "loss": 0.0164,
+      "step": 1783
+    },
+    {
+      "epoch": 0.5661105188784491,
+      "grad_norm": 0.4089169204235077,
+      "learning_rate": 4.7258714926532e-06,
+      "loss": 0.0218,
+      "step": 1784
+    },
+    {
+      "epoch": 0.566427845402484,
+      "grad_norm": 0.30306902527809143,
+      "learning_rate": 4.720339267900728e-06,
+      "loss": 0.0184,
+      "step": 1785
+    },
+    {
+      "epoch": 0.5667451719265191,
+      "grad_norm": 0.38698190450668335,
+      "learning_rate": 4.714807386567843e-06,
+      "loss": 0.0192,
+      "step": 1786
+    },
+    {
+      "epoch": 0.5670624984505541,
+      "grad_norm": 0.37431445717811584,
+      "learning_rate": 4.7092758554476215e-06,
+      "loss": 0.0229,
+      "step": 1787
+    },
+    {
+      "epoch": 0.567379824974589,
+      "grad_norm": 0.3626854121685028,
+      "learning_rate": 4.703744681332711e-06,
+      "loss": 0.023,
+      "step": 1788
+    },
+    {
+      "epoch": 0.5676971514986241,
+      "grad_norm": 0.3253890573978424,
+      "learning_rate": 4.698213871015318e-06,
+      "loss": 0.0172,
+      "step": 1789
+    },
+    {
+      "epoch": 0.5680144780226591,
+      "grad_norm": 0.4929578900337219,
+      "learning_rate": 4.692683431287202e-06,
+      "loss": 0.0174,
+      "step": 1790
+    },
+    {
+      "epoch": 0.5683318045466941,
+      "grad_norm": 0.34769970178604126,
+      "learning_rate": 4.687153368939668e-06,
+      "loss": 0.0198,
+      "step": 1791
+    },
+    {
+      "epoch": 0.5686491310707291,
+      "grad_norm": 0.33412739634513855,
+      "learning_rate": 4.681623690763557e-06,
+      "loss": 0.0194,
+      "step": 1792
+    },
+    {
+      "epoch": 0.5689664575947642,
+      "grad_norm": 0.3509608209133148,
+      "learning_rate": 4.676094403549241e-06,
+      "loss": 0.017,
+      "step": 1793
+    },
+    {
+      "epoch": 0.5692837841187991,
+      "grad_norm": 0.3759349584579468,
+      "learning_rate": 4.670565514086607e-06,
+      "loss": 0.0207,
+      "step": 1794
+    },
+    {
+      "epoch": 0.5696011106428341,
+      "grad_norm": 0.3578685224056244,
+      "learning_rate": 4.66503702916506e-06,
+      "loss": 0.0219,
+      "step": 1795
+    },
+    {
+      "epoch": 0.5699184371668691,
+      "grad_norm": 0.34081071615219116,
+      "learning_rate": 4.6595089555735e-06,
+      "loss": 0.0166,
+      "step": 1796
+    },
+    {
+      "epoch": 0.5702357636909041,
+      "grad_norm": 0.38654112815856934,
+      "learning_rate": 4.6539813001003295e-06,
+      "loss": 0.0202,
+      "step": 1797
+    },
+    {
+      "epoch": 0.5705530902149392,
+      "grad_norm": 0.3973645567893982,
+      "learning_rate": 4.648454069533433e-06,
+      "loss": 0.0204,
+      "step": 1798
+    },
+    {
+      "epoch": 0.5708704167389741,
+      "grad_norm": 0.2973407804965973,
+      "learning_rate": 4.642927270660176e-06,
+      "loss": 0.0157,
+      "step": 1799
+    },
+    {
+      "epoch": 0.5711877432630091,
+      "grad_norm": 0.3247319757938385,
+      "learning_rate": 4.637400910267392e-06,
+      "loss": 0.0181,
+      "step": 1800
+    },
+    {
+      "epoch": 0.5715050697870442,
+      "grad_norm": 0.302016943693161,
+      "learning_rate": 4.631874995141376e-06,
+      "loss": 0.0153,
+      "step": 1801
+    },
+    {
+      "epoch": 0.5718223963110791,
+      "grad_norm": 0.28966957330703735,
+      "learning_rate": 4.626349532067879e-06,
+      "loss": 0.0156,
+      "step": 1802
+    },
+    {
+      "epoch": 0.5721397228351142,
+      "grad_norm": 0.422054260969162,
+      "learning_rate": 4.6208245278320926e-06,
+      "loss": 0.0236,
+      "step": 1803
+    },
+    {
+      "epoch": 0.5724570493591492,
+      "grad_norm": 0.5330262780189514,
+      "learning_rate": 4.615299989218648e-06,
+      "loss": 0.0174,
+      "step": 1804
+    },
+    {
+      "epoch": 0.5727743758831841,
+      "grad_norm": 0.5900294780731201,
+      "learning_rate": 4.6097759230116055e-06,
+      "loss": 0.0152,
+      "step": 1805
+    },
+    {
+      "epoch": 0.5730917024072192,
+      "grad_norm": 0.3148462474346161,
+      "learning_rate": 4.60425233599444e-06,
+      "loss": 0.0179,
+      "step": 1806
+    },
+    {
+      "epoch": 0.5734090289312542,
+      "grad_norm": 0.332144558429718,
+      "learning_rate": 4.598729234950044e-06,
+      "loss": 0.0186,
+      "step": 1807
+    },
+    {
+      "epoch": 0.5737263554552892,
+      "grad_norm": 0.4321858882904053,
+      "learning_rate": 4.59320662666071e-06,
+      "loss": 0.0194,
+      "step": 1808
+    },
+    {
+      "epoch": 0.5740436819793242,
+      "grad_norm": 0.7980892062187195,
+      "learning_rate": 4.587684517908125e-06,
+      "loss": 0.0156,
+      "step": 1809
+    },
+    {
+      "epoch": 0.5743610085033592,
+      "grad_norm": 0.2824123203754425,
+      "learning_rate": 4.582162915473366e-06,
+      "loss": 0.0174,
+      "step": 1810
+    },
+    {
+      "epoch": 0.5746783350273942,
+      "grad_norm": 0.3217780590057373,
+      "learning_rate": 4.576641826136884e-06,
+      "loss": 0.0181,
+      "step": 1811
+    },
+    {
+      "epoch": 0.5749956615514292,
+      "grad_norm": 0.49817126989364624,
+      "learning_rate": 4.571121256678503e-06,
+      "loss": 0.0192,
+      "step": 1812
+    },
+    {
+      "epoch": 0.5753129880754642,
+      "grad_norm": 0.404278039932251,
+      "learning_rate": 4.565601213877408e-06,
+      "loss": 0.0172,
+      "step": 1813
+    },
+    {
+      "epoch": 0.5756303145994992,
+      "grad_norm": 0.4896482825279236,
+      "learning_rate": 4.560081704512136e-06,
+      "loss": 0.0185,
+      "step": 1814
+    },
+    {
+      "epoch": 0.5759476411235342,
+      "grad_norm": 0.3928994834423065,
+      "learning_rate": 4.5545627353605705e-06,
+      "loss": 0.0162,
+      "step": 1815
+    },
+    {
+      "epoch": 0.5762649676475692,
+      "grad_norm": 0.377953439950943,
+      "learning_rate": 4.549044313199929e-06,
+      "loss": 0.0162,
+      "step": 1816
+    },
+    {
+      "epoch": 0.5765822941716042,
+      "grad_norm": 0.3721259832382202,
+      "learning_rate": 4.5435264448067595e-06,
+      "loss": 0.018,
+      "step": 1817
+    },
+    {
+      "epoch": 0.5768996206956393,
+      "grad_norm": 0.3186573088169098,
+      "learning_rate": 4.538009136956931e-06,
+      "loss": 0.0197,
+      "step": 1818
+    },
+    {
+      "epoch": 0.5772169472196742,
+      "grad_norm": 0.4137105941772461,
+      "learning_rate": 4.5324923964256215e-06,
+      "loss": 0.02,
+      "step": 1819
+    },
+    {
+      "epoch": 0.5775342737437092,
+      "grad_norm": 0.2993280589580536,
+      "learning_rate": 4.526976229987315e-06,
+      "loss": 0.0143,
+      "step": 1820
+    },
+    {
+      "epoch": 0.5778516002677443,
+      "grad_norm": 0.41219374537467957,
+      "learning_rate": 4.521460644415788e-06,
+      "loss": 0.0176,
+      "step": 1821
+    },
+    {
+      "epoch": 0.5781689267917792,
+      "grad_norm": 0.3846438527107239,
+      "learning_rate": 4.515945646484105e-06,
+      "loss": 0.0165,
+      "step": 1822
+    },
+    {
+      "epoch": 0.5784862533158143,
+      "grad_norm": 0.35758450627326965,
+      "learning_rate": 4.510431242964609e-06,
+      "loss": 0.0202,
+      "step": 1823
+    },
+    {
+      "epoch": 0.5788035798398493,
+      "grad_norm": 0.42949286103248596,
+      "learning_rate": 4.504917440628913e-06,
+      "loss": 0.021,
+      "step": 1824
+    },
+    {
+      "epoch": 0.5791209063638842,
+      "grad_norm": 0.5928534269332886,
+      "learning_rate": 4.499404246247892e-06,
+      "loss": 0.0194,
+      "step": 1825
+    },
+    {
+      "epoch": 0.5794382328879193,
+      "grad_norm": 0.351223349571228,
+      "learning_rate": 4.493891666591672e-06,
+      "loss": 0.0174,
+      "step": 1826
+    },
+    {
+      "epoch": 0.5797555594119543,
+      "grad_norm": 0.39287564158439636,
+      "learning_rate": 4.488379708429628e-06,
+      "loss": 0.0174,
+      "step": 1827
+    },
+    {
+      "epoch": 0.5800728859359893,
+      "grad_norm": 0.3472081422805786,
+      "learning_rate": 4.482868378530369e-06,
+      "loss": 0.018,
+      "step": 1828
+    },
+    {
+      "epoch": 0.5803902124600243,
+      "grad_norm": 0.40511420369148254,
+      "learning_rate": 4.477357683661734e-06,
+      "loss": 0.0185,
+      "step": 1829
+    },
+    {
+      "epoch": 0.5807075389840592,
+      "grad_norm": 0.27846759557724,
+      "learning_rate": 4.471847630590781e-06,
+      "loss": 0.0155,
+      "step": 1830
+    },
+    {
+      "epoch": 0.5810248655080943,
+      "grad_norm": 0.3140798509120941,
+      "learning_rate": 4.46633822608378e-06,
+      "loss": 0.0151,
+      "step": 1831
+    },
+    {
+      "epoch": 0.5813421920321293,
+      "grad_norm": 0.36867961287498474,
+      "learning_rate": 4.460829476906208e-06,
+      "loss": 0.0223,
+      "step": 1832
+    },
+    {
+      "epoch": 0.5816595185561643,
+      "grad_norm": 0.30136358737945557,
+      "learning_rate": 4.45532138982273e-06,
+      "loss": 0.0197,
+      "step": 1833
+    },
+    {
+      "epoch": 0.5819768450801993,
+      "grad_norm": 0.4835137724876404,
+      "learning_rate": 4.449813971597206e-06,
+      "loss": 0.019,
+      "step": 1834
+    },
+    {
+      "epoch": 0.5822941716042344,
+      "grad_norm": 0.33054250478744507,
+      "learning_rate": 4.444307228992671e-06,
+      "loss": 0.0149,
+      "step": 1835
+    },
+    {
+      "epoch": 0.5826114981282693,
+      "grad_norm": 1.3166013956069946,
+      "learning_rate": 4.4388011687713274e-06,
+      "loss": 0.0178,
+      "step": 1836
+    },
+    {
+      "epoch": 0.5829288246523043,
+      "grad_norm": 0.3779589533805847,
+      "learning_rate": 4.433295797694544e-06,
+      "loss": 0.0194,
+      "step": 1837
+    },
+    {
+      "epoch": 0.5832461511763394,
+      "grad_norm": 0.3328324556350708,
+      "learning_rate": 4.427791122522841e-06,
+      "loss": 0.0212,
+      "step": 1838
+    },
+    {
+      "epoch": 0.5835634777003743,
+      "grad_norm": 0.28207018971443176,
+      "learning_rate": 4.422287150015887e-06,
+      "loss": 0.0159,
+      "step": 1839
+    },
+    {
+      "epoch": 0.5838808042244094,
+      "grad_norm": 0.37970712780952454,
+      "learning_rate": 4.416783886932483e-06,
+      "loss": 0.0186,
+      "step": 1840
+    },
+    {
+      "epoch": 0.5841981307484444,
+      "grad_norm": 0.44644683599472046,
+      "learning_rate": 4.411281340030564e-06,
+      "loss": 0.0193,
+      "step": 1841
+    },
+    {
+      "epoch": 0.5845154572724793,
+      "grad_norm": 0.39640310406684875,
+      "learning_rate": 4.405779516067182e-06,
+      "loss": 0.0195,
+      "step": 1842
+    },
+    {
+      "epoch": 0.5848327837965144,
+      "grad_norm": 0.42071351408958435,
+      "learning_rate": 4.400278421798501e-06,
+      "loss": 0.0215,
+      "step": 1843
+    },
+    {
+      "epoch": 0.5851501103205494,
+      "grad_norm": 0.35830244421958923,
+      "learning_rate": 4.394778063979793e-06,
+      "loss": 0.0204,
+      "step": 1844
+    },
+    {
+      "epoch": 0.5854674368445844,
+      "grad_norm": 0.4621761739253998,
+      "learning_rate": 4.389278449365422e-06,
+      "loss": 0.0187,
+      "step": 1845
+    },
+    {
+      "epoch": 0.5857847633686194,
+      "grad_norm": 0.33336499333381653,
+      "learning_rate": 4.3837795847088394e-06,
+      "loss": 0.0139,
+      "step": 1846
+    },
+    {
+      "epoch": 0.5861020898926543,
+      "grad_norm": 0.32810717821121216,
+      "learning_rate": 4.3782814767625755e-06,
+      "loss": 0.0157,
+      "step": 1847
+    },
+    {
+      "epoch": 0.5864194164166894,
+      "grad_norm": 0.27847108244895935,
+      "learning_rate": 4.372784132278234e-06,
+      "loss": 0.0144,
+      "step": 1848
+    },
+    {
+      "epoch": 0.5867367429407244,
+      "grad_norm": 0.35818353295326233,
+      "learning_rate": 4.367287558006479e-06,
+      "loss": 0.0177,
+      "step": 1849
+    },
+    {
+      "epoch": 0.5870540694647594,
+      "grad_norm": 0.41418641805648804,
+      "learning_rate": 4.361791760697027e-06,
+      "loss": 0.0192,
+      "step": 1850
+    },
+    {
+      "epoch": 0.5873713959887944,
+      "grad_norm": 0.3078855276107788,
+      "learning_rate": 4.356296747098647e-06,
+      "loss": 0.0162,
+      "step": 1851
+    },
+    {
+      "epoch": 0.5876887225128294,
+      "grad_norm": 0.38694295287132263,
+      "learning_rate": 4.3508025239591385e-06,
+      "loss": 0.0164,
+      "step": 1852
+    },
+    {
+      "epoch": 0.5880060490368644,
+      "grad_norm": 0.3471567928791046,
+      "learning_rate": 4.345309098025333e-06,
+      "loss": 0.0197,
+      "step": 1853
+    },
+    {
+      "epoch": 0.5883233755608994,
+      "grad_norm": 0.3009127676486969,
+      "learning_rate": 4.339816476043085e-06,
+      "loss": 0.015,
+      "step": 1854
+    },
+    {
+      "epoch": 0.5886407020849345,
+      "grad_norm": 0.3233533799648285,
+      "learning_rate": 4.33432466475726e-06,
+      "loss": 0.0202,
+      "step": 1855
+    },
+    {
+      "epoch": 0.5889580286089694,
+      "grad_norm": 0.32282426953315735,
+      "learning_rate": 4.3288336709117246e-06,
+      "loss": 0.0188,
+      "step": 1856
+    },
+    {
+      "epoch": 0.5892753551330044,
+      "grad_norm": 0.4565737247467041,
+      "learning_rate": 4.323343501249346e-06,
+      "loss": 0.0197,
+      "step": 1857
+    },
+    {
+      "epoch": 0.5895926816570395,
+      "grad_norm": 0.3827863037586212,
+      "learning_rate": 4.317854162511978e-06,
+      "loss": 0.0196,
+      "step": 1858
+    },
+    {
+      "epoch": 0.5899100081810744,
+      "grad_norm": 0.3502645492553711,
+      "learning_rate": 4.3123656614404565e-06,
+      "loss": 0.0187,
+      "step": 1859
+    },
+    {
+      "epoch": 0.5902273347051095,
+      "grad_norm": 0.5492234826087952,
+      "learning_rate": 4.306878004774584e-06,
+      "loss": 0.0181,
+      "step": 1860
+    },
+    {
+      "epoch": 0.5905446612291445,
+      "grad_norm": 1.7816503047943115,
+      "learning_rate": 4.301391199253131e-06,
+      "loss": 0.0207,
+      "step": 1861
+    },
+    {
+      "epoch": 0.5908619877531794,
+      "grad_norm": 0.3262493312358856,
+      "learning_rate": 4.295905251613817e-06,
+      "loss": 0.0166,
+      "step": 1862
+    },
+    {
+      "epoch": 0.5911793142772145,
+      "grad_norm": 0.36608606576919556,
+      "learning_rate": 4.290420168593314e-06,
+      "loss": 0.0168,
+      "step": 1863
+    },
+    {
+      "epoch": 0.5914966408012494,
+      "grad_norm": 0.30079418420791626,
+      "learning_rate": 4.284935956927229e-06,
+      "loss": 0.0165,
+      "step": 1864
+    },
+    {
+      "epoch": 0.5918139673252845,
+      "grad_norm": 0.3564797043800354,
+      "learning_rate": 4.279452623350101e-06,
+      "loss": 0.0174,
+      "step": 1865
+    },
+    {
+      "epoch": 0.5921312938493195,
+      "grad_norm": 0.6890491843223572,
+      "learning_rate": 4.273970174595385e-06,
+      "loss": 0.0178,
+      "step": 1866
+    },
+    {
+      "epoch": 0.5924486203733544,
+      "grad_norm": 0.38451188802719116,
+      "learning_rate": 4.268488617395457e-06,
+      "loss": 0.0175,
+      "step": 1867
+    },
+    {
+      "epoch": 0.5927659468973895,
+      "grad_norm": 0.3831936717033386,
+      "learning_rate": 4.263007958481595e-06,
+      "loss": 0.0177,
+      "step": 1868
+    },
+    {
+      "epoch": 0.5930832734214245,
+      "grad_norm": 0.32724201679229736,
+      "learning_rate": 4.257528204583973e-06,
+      "loss": 0.0152,
+      "step": 1869
+    },
+    {
+      "epoch": 0.5934005999454595,
+      "grad_norm": 0.9824090600013733,
+      "learning_rate": 4.252049362431655e-06,
+      "loss": 0.0159,
+      "step": 1870
+    },
+    {
+      "epoch": 0.5937179264694945,
+      "grad_norm": 0.39863312244415283,
+      "learning_rate": 4.246571438752585e-06,
+      "loss": 0.0189,
+      "step": 1871
+    },
+    {
+      "epoch": 0.5940352529935296,
+      "grad_norm": 0.33342844247817993,
+      "learning_rate": 4.241094440273578e-06,
+      "loss": 0.0179,
+      "step": 1872
+    },
+    {
+      "epoch": 0.5943525795175645,
+      "grad_norm": 0.357719749212265,
+      "learning_rate": 4.235618373720315e-06,
+      "loss": 0.0178,
+      "step": 1873
+    },
+    {
+      "epoch": 0.5946699060415995,
+      "grad_norm": 0.5923636555671692,
+      "learning_rate": 4.230143245817332e-06,
+      "loss": 0.0193,
+      "step": 1874
+    },
+    {
+      "epoch": 0.5949872325656346,
+      "grad_norm": 0.3748299181461334,
+      "learning_rate": 4.224669063288012e-06,
+      "loss": 0.017,
+      "step": 1875
+    },
+    {
+      "epoch": 0.5953045590896695,
+      "grad_norm": 0.3360614776611328,
+      "learning_rate": 4.219195832854574e-06,
+      "loss": 0.0165,
+      "step": 1876
+    },
+    {
+      "epoch": 0.5956218856137045,
+      "grad_norm": 0.3752370774745941,
+      "learning_rate": 4.213723561238074e-06,
+      "loss": 0.0169,
+      "step": 1877
+    },
+    {
+      "epoch": 0.5959392121377396,
+      "grad_norm": 0.3340493142604828,
+      "learning_rate": 4.208252255158387e-06,
+      "loss": 0.0174,
+      "step": 1878
+    },
+    {
+      "epoch": 0.5962565386617745,
+      "grad_norm": 0.6768607497215271,
+      "learning_rate": 4.202781921334201e-06,
+      "loss": 0.0179,
+      "step": 1879
+    },
+    {
+      "epoch": 0.5965738651858096,
+      "grad_norm": 0.38332343101501465,
+      "learning_rate": 4.197312566483013e-06,
+      "loss": 0.0181,
+      "step": 1880
+    },
+    {
+      "epoch": 0.5968911917098445,
+      "grad_norm": 0.38498783111572266,
+      "learning_rate": 4.191844197321116e-06,
+      "loss": 0.0194,
+      "step": 1881
+    },
+    {
+      "epoch": 0.5972085182338795,
+      "grad_norm": 0.7729988098144531,
+      "learning_rate": 4.186376820563595e-06,
+      "loss": 0.0175,
+      "step": 1882
+    },
+    {
+      "epoch": 0.5975258447579146,
+      "grad_norm": 0.3526090383529663,
+      "learning_rate": 4.180910442924312e-06,
+      "loss": 0.0195,
+      "step": 1883
+    },
+    {
+      "epoch": 0.5978431712819495,
+      "grad_norm": 0.5022876858711243,
+      "learning_rate": 4.175445071115906e-06,
+      "loss": 0.0162,
+      "step": 1884
+    },
+    {
+      "epoch": 0.5981604978059846,
+      "grad_norm": 0.3365197479724884,
+      "learning_rate": 4.1699807118497815e-06,
+      "loss": 0.0196,
+      "step": 1885
+    },
+    {
+      "epoch": 0.5984778243300196,
+      "grad_norm": 0.2689717710018158,
+      "learning_rate": 4.164517371836094e-06,
+      "loss": 0.0173,
+      "step": 1886
+    },
+    {
+      "epoch": 0.5987951508540545,
+      "grad_norm": 0.31464099884033203,
+      "learning_rate": 4.1590550577837524e-06,
+      "loss": 0.015,
+      "step": 1887
+    },
+    {
+      "epoch": 0.5991124773780896,
+      "grad_norm": 0.3674316108226776,
+      "learning_rate": 4.153593776400405e-06,
+      "loss": 0.0173,
+      "step": 1888
+    },
+    {
+      "epoch": 0.5994298039021246,
+      "grad_norm": 0.3420359492301941,
+      "learning_rate": 4.148133534392431e-06,
+      "loss": 0.0157,
+      "step": 1889
+    },
+    {
+      "epoch": 0.5997471304261596,
+      "grad_norm": 0.32488444447517395,
+      "learning_rate": 4.142674338464933e-06,
+      "loss": 0.0151,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6000644569501946,
+      "grad_norm": 0.36253267526626587,
+      "learning_rate": 4.137216195321731e-06,
+      "loss": 0.023,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6003817834742297,
+      "grad_norm": 0.33637577295303345,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.019,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6006991099982646,
+      "grad_norm": 0.43741825222969055,
+      "learning_rate": 4.1263030941970124e-06,
+      "loss": 0.0169,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6010164365222996,
+      "grad_norm": 0.40051475167274475,
+      "learning_rate": 4.120848149616638e-06,
+      "loss": 0.0207,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6013337630463347,
+      "grad_norm": 0.7792407274246216,
+      "learning_rate": 4.115394284622824e-06,
+      "loss": 0.0156,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6016510895703696,
+      "grad_norm": 0.33454182744026184,
+      "learning_rate": 4.10994150591284e-06,
+      "loss": 0.0186,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6019684160944047,
+      "grad_norm": 0.44952741265296936,
+      "learning_rate": 4.104489820182626e-06,
+      "loss": 0.0164,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6022857426184397,
+      "grad_norm": 0.4245060384273529,
+      "learning_rate": 4.099039234126778e-06,
+      "loss": 0.0216,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6026030691424746,
+      "grad_norm": 0.43050241470336914,
+      "learning_rate": 4.093589754438543e-06,
+      "loss": 0.0203,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6029203956665097,
+      "grad_norm": 0.44346150755882263,
+      "learning_rate": 4.088141387809806e-06,
+      "loss": 0.0176,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6032377221905446,
+      "grad_norm": 0.39374828338623047,
+      "learning_rate": 4.0826941409310885e-06,
+      "loss": 0.0216,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6035550487145797,
+      "grad_norm": 0.3911575675010681,
+      "learning_rate": 4.0772480204915364e-06,
+      "loss": 0.018,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6038723752386147,
+      "grad_norm": 0.32082056999206543,
+      "learning_rate": 4.071803033178911e-06,
+      "loss": 0.018,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6041897017626496,
+      "grad_norm": 0.3726379871368408,
+      "learning_rate": 4.066359185679584e-06,
+      "loss": 0.0212,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6045070282866847,
+      "grad_norm": 0.5339725613594055,
+      "learning_rate": 4.060916484678526e-06,
+      "loss": 0.0195,
+      "step": 1905
+    },
+    {
+      "epoch": 0.6048243548107197,
+      "grad_norm": 0.31710970401763916,
+      "learning_rate": 4.0554749368593e-06,
+      "loss": 0.0159,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6051416813347547,
+      "grad_norm": 0.4115637242794037,
+      "learning_rate": 4.0500345489040515e-06,
+      "loss": 0.0192,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6054590078587897,
+      "grad_norm": 0.43987002968788147,
+      "learning_rate": 4.044595327493505e-06,
+      "loss": 0.0213,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6057763343828247,
+      "grad_norm": 0.3933052122592926,
+      "learning_rate": 4.039157279306949e-06,
+      "loss": 0.0153,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6060936609068597,
+      "grad_norm": 0.6967605948448181,
+      "learning_rate": 4.033720411022235e-06,
+      "loss": 0.0175,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6064109874308947,
+      "grad_norm": 0.3661051094532013,
+      "learning_rate": 4.028284729315762e-06,
+      "loss": 0.0224,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6067283139549298,
+      "grad_norm": 0.401399701833725,
+      "learning_rate": 4.0228502408624735e-06,
+      "loss": 0.0187,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6070456404789647,
+      "grad_norm": 0.2652651071548462,
+      "learning_rate": 4.017416952335849e-06,
+      "loss": 0.015,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6073629670029997,
+      "grad_norm": 0.7262048721313477,
+      "learning_rate": 4.011984870407893e-06,
+      "loss": 0.02,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6076802935270348,
+      "grad_norm": 0.34282732009887695,
+      "learning_rate": 4.00655400174913e-06,
+      "loss": 0.016,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6079976200510697,
+      "grad_norm": 0.49580395221710205,
+      "learning_rate": 4.0011243530285885e-06,
+      "loss": 0.0167,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6083149465751048,
+      "grad_norm": 0.5014708638191223,
+      "learning_rate": 3.995695930913806e-06,
+      "loss": 0.0171,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6086322730991397,
+      "grad_norm": 0.3929635286331177,
+      "learning_rate": 3.990268742070812e-06,
+      "loss": 0.0162,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6089495996231747,
+      "grad_norm": 0.5658152103424072,
+      "learning_rate": 3.98484279316412e-06,
+      "loss": 0.0169,
+      "step": 1919
+    },
+    {
+      "epoch": 0.6092669261472098,
+      "grad_norm": 0.8225061297416687,
+      "learning_rate": 3.979418090856723e-06,
+      "loss": 0.0189,
+      "step": 1920
+    },
+    {
+      "epoch": 0.6095842526712447,
+      "grad_norm": 0.3190343379974365,
+      "learning_rate": 3.9739946418100785e-06,
+      "loss": 0.0188,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6099015791952798,
+      "grad_norm": 0.3619401156902313,
+      "learning_rate": 3.968572452684113e-06,
+      "loss": 0.019,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6102189057193148,
+      "grad_norm": 0.41502177715301514,
+      "learning_rate": 3.963151530137199e-06,
+      "loss": 0.0187,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6105362322433497,
+      "grad_norm": 0.38101398944854736,
+      "learning_rate": 3.957731880826156e-06,
+      "loss": 0.0196,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6108535587673848,
+      "grad_norm": 0.29558947682380676,
+      "learning_rate": 3.952313511406242e-06,
+      "loss": 0.0141,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6111708852914198,
+      "grad_norm": 0.3555171489715576,
+      "learning_rate": 3.9468964285311375e-06,
+      "loss": 0.0185,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6114882118154548,
+      "grad_norm": 0.5933432579040527,
+      "learning_rate": 3.941480638852948e-06,
+      "loss": 0.0214,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6118055383394898,
+      "grad_norm": 0.5558792352676392,
+      "learning_rate": 3.936066149022191e-06,
+      "loss": 0.0215,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6121228648635249,
+      "grad_norm": 0.5972689986228943,
+      "learning_rate": 3.930652965687785e-06,
+      "loss": 0.0223,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6124401913875598,
+      "grad_norm": 0.4915259778499603,
+      "learning_rate": 3.925241095497047e-06,
+      "loss": 0.0207,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6127575179115948,
+      "grad_norm": 0.40416863560676575,
+      "learning_rate": 3.919830545095681e-06,
+      "loss": 0.0176,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6130748444356299,
+      "grad_norm": 0.3234327733516693,
+      "learning_rate": 3.914421321127769e-06,
+      "loss": 0.015,
+      "step": 1932
+    },
+    {
+      "epoch": 0.6133921709596648,
+      "grad_norm": 0.26694533228874207,
+      "learning_rate": 3.909013430235764e-06,
+      "loss": 0.0162,
+      "step": 1933
+    },
+    {
+      "epoch": 0.6137094974836999,
+      "grad_norm": 0.494841605424881,
+      "learning_rate": 3.903606879060483e-06,
+      "loss": 0.0185,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6140268240077348,
+      "grad_norm": 0.3341357707977295,
+      "learning_rate": 3.8982016742411005e-06,
+      "loss": 0.0174,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6143441505317698,
+      "grad_norm": 0.3706906735897064,
+      "learning_rate": 3.892797822415129e-06,
+      "loss": 0.0142,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6146614770558049,
+      "grad_norm": 0.5088722705841064,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.0168,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6149788035798398,
+      "grad_norm": 0.5534664392471313,
+      "learning_rate": 3.881994204285187e-06,
+      "loss": 0.0192,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6152961301038748,
+      "grad_norm": 0.4941696524620056,
+      "learning_rate": 3.876594451247913e-06,
+      "loss": 0.0216,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6156134566279099,
+      "grad_norm": 1.5360089540481567,
+      "learning_rate": 3.871196077737429e-06,
+      "loss": 0.0184,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6159307831519448,
+      "grad_norm": 0.36695775389671326,
+      "learning_rate": 3.865799090382866e-06,
+      "loss": 0.0154,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6162481096759799,
+      "grad_norm": 0.37227070331573486,
+      "learning_rate": 3.8604034958116515e-06,
+      "loss": 0.0165,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6165654362000149,
+      "grad_norm": 0.40412476658821106,
+      "learning_rate": 3.855009300649502e-06,
+      "loss": 0.0187,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6168827627240498,
+      "grad_norm": 0.34698861837387085,
+      "learning_rate": 3.849616511520415e-06,
+      "loss": 0.0163,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6172000892480849,
+      "grad_norm": 0.7037256360054016,
+      "learning_rate": 3.844225135046666e-06,
+      "loss": 0.0155,
+      "step": 1945
+    },
+    {
+      "epoch": 0.6175174157721199,
+      "grad_norm": 0.34823551774024963,
+      "learning_rate": 3.8388351778487884e-06,
+      "loss": 0.0165,
+      "step": 1946
+    },
+    {
+      "epoch": 0.6178347422961549,
+      "grad_norm": 0.4502529203891754,
+      "learning_rate": 3.833446646545577e-06,
+      "loss": 0.0173,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6181520688201899,
+      "grad_norm": 0.4536212384700775,
+      "learning_rate": 3.828059547754078e-06,
+      "loss": 0.0189,
+      "step": 1948
+    },
+    {
+      "epoch": 0.618469395344225,
+      "grad_norm": 0.3786444067955017,
+      "learning_rate": 3.822673888089572e-06,
+      "loss": 0.0196,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6187867218682599,
+      "grad_norm": 0.47480350732803345,
+      "learning_rate": 3.817289674165578e-06,
+      "loss": 0.0167,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6191040483922949,
+      "grad_norm": 0.323874294757843,
+      "learning_rate": 3.811906912593839e-06,
+      "loss": 0.0173,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6194213749163299,
+      "grad_norm": 0.5282031297683716,
+      "learning_rate": 3.806525609984312e-06,
+      "loss": 0.0184,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6197387014403649,
+      "grad_norm": 0.3333488404750824,
+      "learning_rate": 3.801145772945163e-06,
+      "loss": 0.0155,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6200560279644,
+      "grad_norm": 0.4559798538684845,
+      "learning_rate": 3.79576740808276e-06,
+      "loss": 0.0179,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6203733544884349,
+      "grad_norm": 0.34577229619026184,
+      "learning_rate": 3.790390522001662e-06,
+      "loss": 0.0168,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6206906810124699,
+      "grad_norm": 0.580481767654419,
+      "learning_rate": 3.785015121304612e-06,
+      "loss": 0.022,
+      "step": 1956
+    },
+    {
+      "epoch": 0.621008007536505,
+      "grad_norm": 0.31832155585289,
+      "learning_rate": 3.77964121259253e-06,
+      "loss": 0.0177,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6213253340605399,
+      "grad_norm": 1.0267103910446167,
+      "learning_rate": 3.7742688024645007e-06,
+      "loss": 0.0156,
+      "step": 1958
+    },
+    {
+      "epoch": 0.621642660584575,
+      "grad_norm": 0.34192973375320435,
+      "learning_rate": 3.768897897517774e-06,
+      "loss": 0.0193,
+      "step": 1959
+    },
+    {
+      "epoch": 0.62195998710861,
+      "grad_norm": 0.42662444710731506,
+      "learning_rate": 3.7635285043477465e-06,
+      "loss": 0.0189,
+      "step": 1960
+    },
+    {
+      "epoch": 0.6222773136326449,
+      "grad_norm": 0.3320058286190033,
+      "learning_rate": 3.7581606295479605e-06,
+      "loss": 0.0169,
+      "step": 1961
+    },
+    {
+      "epoch": 0.62259464015668,
+      "grad_norm": 0.35327625274658203,
+      "learning_rate": 3.752794279710094e-06,
+      "loss": 0.0158,
+      "step": 1962
+    },
+    {
+      "epoch": 0.622911966680715,
+      "grad_norm": 0.6692973375320435,
+      "learning_rate": 3.7474294614239514e-06,
+      "loss": 0.0175,
+      "step": 1963
+    },
+    {
+      "epoch": 0.62322929320475,
+      "grad_norm": 0.45618170499801636,
+      "learning_rate": 3.7420661812774577e-06,
+      "loss": 0.0156,
+      "step": 1964
+    },
+    {
+      "epoch": 0.623546619728785,
+      "grad_norm": 0.3191971480846405,
+      "learning_rate": 3.7367044458566485e-06,
+      "loss": 0.0188,
+      "step": 1965
+    },
+    {
+      "epoch": 0.62386394625282,
+      "grad_norm": 0.39543017745018005,
+      "learning_rate": 3.7313442617456606e-06,
+      "loss": 0.019,
+      "step": 1966
+    },
+    {
+      "epoch": 0.624181272776855,
+      "grad_norm": 0.32241928577423096,
+      "learning_rate": 3.7259856355267275e-06,
+      "loss": 0.0143,
+      "step": 1967
+    },
+    {
+      "epoch": 0.62449859930089,
+      "grad_norm": 0.339998722076416,
+      "learning_rate": 3.7206285737801718e-06,
+      "loss": 0.0169,
+      "step": 1968
+    },
+    {
+      "epoch": 0.624815925824925,
+      "grad_norm": 0.6769499182701111,
+      "learning_rate": 3.7152730830843904e-06,
+      "loss": 0.015,
+      "step": 1969
+    },
+    {
+      "epoch": 0.62513325234896,
+      "grad_norm": 0.41190305352211,
+      "learning_rate": 3.7099191700158545e-06,
+      "loss": 0.0182,
+      "step": 1970
+    },
+    {
+      "epoch": 0.625450578872995,
+      "grad_norm": 0.30510956048965454,
+      "learning_rate": 3.7045668411490955e-06,
+      "loss": 0.0194,
+      "step": 1971
+    },
+    {
+      "epoch": 0.62576790539703,
+      "grad_norm": 0.31207746267318726,
+      "learning_rate": 3.699216103056703e-06,
+      "loss": 0.0175,
+      "step": 1972
+    },
+    {
+      "epoch": 0.626085231921065,
+      "grad_norm": 0.37149518728256226,
+      "learning_rate": 3.6938669623093086e-06,
+      "loss": 0.0165,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6264025584451001,
+      "grad_norm": 0.3917272090911865,
+      "learning_rate": 3.688519425475586e-06,
+      "loss": 0.0224,
+      "step": 1974
+    },
+    {
+      "epoch": 0.626719884969135,
+      "grad_norm": 0.35287442803382874,
+      "learning_rate": 3.683173499122239e-06,
+      "loss": 0.0204,
+      "step": 1975
+    },
+    {
+      "epoch": 0.62703721149317,
+      "grad_norm": 0.34307757019996643,
+      "learning_rate": 3.6778291898139907e-06,
+      "loss": 0.0189,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6273545380172051,
+      "grad_norm": 0.34982767701148987,
+      "learning_rate": 3.6724865041135805e-06,
+      "loss": 0.0149,
+      "step": 1977
+    },
+    {
+      "epoch": 0.62767186454124,
+      "grad_norm": 0.3181287348270416,
+      "learning_rate": 3.667145448581754e-06,
+      "loss": 0.0142,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6279891910652751,
+      "grad_norm": 0.4341374933719635,
+      "learning_rate": 3.661806029777257e-06,
+      "loss": 0.0144,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6283065175893101,
+      "grad_norm": 0.32886481285095215,
+      "learning_rate": 3.6564682542568215e-06,
+      "loss": 0.017,
+      "step": 1980
+    },
+    {
+      "epoch": 0.628623844113345,
+      "grad_norm": 0.4191092252731323,
+      "learning_rate": 3.6511321285751643e-06,
+      "loss": 0.0184,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6289411706373801,
+      "grad_norm": 0.3749864399433136,
+      "learning_rate": 3.6457976592849753e-06,
+      "loss": 0.0187,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6292584971614151,
+      "grad_norm": 0.31717389822006226,
+      "learning_rate": 3.640464852936909e-06,
+      "loss": 0.0189,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6295758236854501,
+      "grad_norm": 0.29760774970054626,
+      "learning_rate": 3.635133716079583e-06,
+      "loss": 0.0191,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6298931502094851,
+      "grad_norm": 0.2856566309928894,
+      "learning_rate": 3.6298042552595604e-06,
+      "loss": 0.0171,
+      "step": 1985
+    },
+    {
+      "epoch": 0.63021047673352,
+      "grad_norm": 0.3433302342891693,
+      "learning_rate": 3.6244764770213426e-06,
+      "loss": 0.0165,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6305278032575551,
+      "grad_norm": 0.3605843782424927,
+      "learning_rate": 3.619150387907373e-06,
+      "loss": 0.0154,
+      "step": 1987
+    },
+    {
+      "epoch": 0.6308451297815901,
+      "grad_norm": 0.341530442237854,
+      "learning_rate": 3.613825994458016e-06,
+      "loss": 0.0154,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6311624563056251,
+      "grad_norm": 0.4772051274776459,
+      "learning_rate": 3.608503303211553e-06,
+      "loss": 0.0164,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6314797828296601,
+      "grad_norm": 0.41449597477912903,
+      "learning_rate": 3.603182320704179e-06,
+      "loss": 0.0186,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6317971093536952,
+      "grad_norm": 0.4609835147857666,
+      "learning_rate": 3.5978630534699873e-06,
+      "loss": 0.017,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6321144358777301,
+      "grad_norm": 0.29162850975990295,
+      "learning_rate": 3.5925455080409653e-06,
+      "loss": 0.0167,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6324317624017651,
+      "grad_norm": 0.3483189046382904,
+      "learning_rate": 3.5872296909469884e-06,
+      "loss": 0.0156,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6327490889258002,
+      "grad_norm": 0.4061286747455597,
+      "learning_rate": 3.5819156087158075e-06,
+      "loss": 0.0205,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6330664154498351,
+      "grad_norm": 0.3818877935409546,
+      "learning_rate": 3.576603267873045e-06,
+      "loss": 0.0145,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6333837419738702,
+      "grad_norm": 0.36665207147598267,
+      "learning_rate": 3.57129267494218e-06,
+      "loss": 0.018,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6337010684979052,
+      "grad_norm": 0.3804652690887451,
+      "learning_rate": 3.5659838364445505e-06,
+      "loss": 0.0189,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6340183950219401,
+      "grad_norm": 0.3495914936065674,
+      "learning_rate": 3.5606767588993375e-06,
+      "loss": 0.0183,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6343357215459752,
+      "grad_norm": 0.24691030383110046,
+      "learning_rate": 3.5553714488235613e-06,
+      "loss": 0.012,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6346530480700102,
+      "grad_norm": 0.31578126549720764,
+      "learning_rate": 3.550067912732069e-06,
+      "loss": 0.0197,
+      "step": 2000
+    },
+    {
+      "epoch": 0.6349703745940452,
+      "grad_norm": 0.4393080472946167,
+      "learning_rate": 3.5447661571375347e-06,
+      "loss": 0.0163,
+      "step": 2001
+    },
+    {
+      "epoch": 0.6352877011180802,
+      "grad_norm": 0.3581767976284027,
+      "learning_rate": 3.5394661885504388e-06,
+      "loss": 0.0156,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6356050276421151,
+      "grad_norm": 0.32103243470191956,
+      "learning_rate": 3.5341680134790732e-06,
+      "loss": 0.0149,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6359223541661502,
+      "grad_norm": 0.563320517539978,
+      "learning_rate": 3.528871638429524e-06,
+      "loss": 0.0184,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6362396806901852,
+      "grad_norm": 0.30426591634750366,
+      "learning_rate": 3.5235770699056702e-06,
+      "loss": 0.0177,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6365570072142201,
+      "grad_norm": 0.32721996307373047,
+      "learning_rate": 3.518284314409166e-06,
+      "loss": 0.0176,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6368743337382552,
+      "grad_norm": 0.3340579569339752,
+      "learning_rate": 3.512993378439446e-06,
+      "loss": 0.0185,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6371916602622902,
+      "grad_norm": 0.35821712017059326,
+      "learning_rate": 3.5077042684937064e-06,
+      "loss": 0.018,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6375089867863252,
+      "grad_norm": 0.35088595747947693,
+      "learning_rate": 3.502416991066904e-06,
+      "loss": 0.0168,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6378263133103602,
+      "grad_norm": 0.8139888048171997,
+      "learning_rate": 3.4971315526517445e-06,
+      "loss": 0.0191,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6381436398343953,
+      "grad_norm": 0.32912614941596985,
+      "learning_rate": 3.491847959738673e-06,
+      "loss": 0.0193,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6384609663584302,
+      "grad_norm": 0.541120171546936,
+      "learning_rate": 3.4865662188158713e-06,
+      "loss": 0.0191,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6387782928824652,
+      "grad_norm": 0.29765427112579346,
+      "learning_rate": 3.481286336369244e-06,
+      "loss": 0.0156,
+      "step": 2013
+    },
+    {
+      "epoch": 0.6390956194065003,
+      "grad_norm": 0.2911182940006256,
+      "learning_rate": 3.4760083188824166e-06,
+      "loss": 0.0133,
+      "step": 2014
+    },
+    {
+      "epoch": 0.6394129459305352,
+      "grad_norm": 0.2925530970096588,
+      "learning_rate": 3.470732172836725e-06,
+      "loss": 0.0159,
+      "step": 2015
+    },
+    {
+      "epoch": 0.6397302724545703,
+      "grad_norm": 0.3668912351131439,
+      "learning_rate": 3.4654579047111992e-06,
+      "loss": 0.0181,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6400475989786053,
+      "grad_norm": 0.36165982484817505,
+      "learning_rate": 3.4601855209825742e-06,
+      "loss": 0.0185,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6403649255026402,
+      "grad_norm": 0.37385743856430054,
+      "learning_rate": 3.4549150281252635e-06,
+      "loss": 0.0177,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6406822520266753,
+      "grad_norm": 0.3319050669670105,
+      "learning_rate": 3.449646432611362e-06,
+      "loss": 0.0157,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6409995785507102,
+      "grad_norm": 0.30417710542678833,
+      "learning_rate": 3.4443797409106326e-06,
+      "loss": 0.0152,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6413169050747453,
+      "grad_norm": 0.39012932777404785,
+      "learning_rate": 3.4391149594905015e-06,
+      "loss": 0.0152,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6416342315987803,
+      "grad_norm": 0.29158490896224976,
+      "learning_rate": 3.43385209481605e-06,
+      "loss": 0.0146,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6419515581228152,
+      "grad_norm": 0.30221083760261536,
+      "learning_rate": 3.428591153350004e-06,
+      "loss": 0.0142,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6422688846468503,
+      "grad_norm": 0.37840715050697327,
+      "learning_rate": 3.4233321415527275e-06,
+      "loss": 0.0187,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6425862111708853,
+      "grad_norm": 0.4418327212333679,
+      "learning_rate": 3.418075065882217e-06,
+      "loss": 0.0133,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6429035376949203,
+      "grad_norm": 0.3791719973087311,
+      "learning_rate": 3.412819932794091e-06,
+      "loss": 0.0134,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6432208642189553,
+      "grad_norm": 0.4080108404159546,
+      "learning_rate": 3.4075667487415785e-06,
+      "loss": 0.0171,
+      "step": 2027
+    },
+    {
+      "epoch": 0.6435381907429903,
+      "grad_norm": 0.3588544726371765,
+      "learning_rate": 3.402315520175521e-06,
+      "loss": 0.0176,
+      "step": 2028
+    },
+    {
+      "epoch": 0.6438555172670253,
+      "grad_norm": 0.31708693504333496,
+      "learning_rate": 3.3970662535443545e-06,
+      "loss": 0.0153,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6441728437910603,
+      "grad_norm": 0.32324641942977905,
+      "learning_rate": 3.391818955294108e-06,
+      "loss": 0.0155,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6444901703150954,
+      "grad_norm": 0.27354204654693604,
+      "learning_rate": 3.3865736318683926e-06,
+      "loss": 0.016,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6448074968391303,
+      "grad_norm": 0.400935560464859,
+      "learning_rate": 3.3813302897083955e-06,
+      "loss": 0.0162,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6451248233631653,
+      "grad_norm": 0.4504110813140869,
+      "learning_rate": 3.376088935252868e-06,
+      "loss": 0.0217,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6454421498872004,
+      "grad_norm": 0.30385300517082214,
+      "learning_rate": 3.3708495749381233e-06,
+      "loss": 0.0144,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6457594764112353,
+      "grad_norm": 0.36345189809799194,
+      "learning_rate": 3.365612215198024e-06,
+      "loss": 0.0191,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6460768029352704,
+      "grad_norm": 0.4550416171550751,
+      "learning_rate": 3.3603768624639786e-06,
+      "loss": 0.0181,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6463941294593053,
+      "grad_norm": 0.744445264339447,
+      "learning_rate": 3.355143523164928e-06,
+      "loss": 0.0176,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6467114559833403,
+      "grad_norm": 1.0131126642227173,
+      "learning_rate": 3.3499122037273423e-06,
+      "loss": 0.0236,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6470287825073754,
+      "grad_norm": 0.341100811958313,
+      "learning_rate": 3.3446829105752103e-06,
+      "loss": 0.0143,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6473461090314103,
+      "grad_norm": 1.3380036354064941,
+      "learning_rate": 3.339455650130034e-06,
+      "loss": 0.016,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6476634355554454,
+      "grad_norm": 0.34890320897102356,
+      "learning_rate": 3.3342304288108172e-06,
+      "loss": 0.0169,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6479807620794804,
+      "grad_norm": 0.42958852648735046,
+      "learning_rate": 3.3290072530340628e-06,
+      "loss": 0.0192,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6482980886035153,
+      "grad_norm": 0.7416742444038391,
+      "learning_rate": 3.3237861292137587e-06,
+      "loss": 0.0146,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6486154151275504,
+      "grad_norm": 0.28566300868988037,
+      "learning_rate": 3.3185670637613737e-06,
+      "loss": 0.0172,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6489327416515854,
+      "grad_norm": 0.3411003053188324,
+      "learning_rate": 3.3133500630858507e-06,
+      "loss": 0.0172,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6492500681756204,
+      "grad_norm": 0.34374678134918213,
+      "learning_rate": 3.308135133593595e-06,
+      "loss": 0.0168,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6495673946996554,
+      "grad_norm": 0.33419886231422424,
+      "learning_rate": 3.3029222816884697e-06,
+      "loss": 0.0137,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6498847212236905,
+      "grad_norm": 0.3334493339061737,
+      "learning_rate": 3.297711513771786e-06,
+      "loss": 0.0204,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6502020477477254,
+      "grad_norm": 0.38038066029548645,
+      "learning_rate": 3.2925028362422963e-06,
+      "loss": 0.02,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6505193742717604,
+      "grad_norm": 0.29700466990470886,
+      "learning_rate": 3.287296255496186e-06,
+      "loss": 0.0156,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6508367007957955,
+      "grad_norm": 0.4022567868232727,
+      "learning_rate": 3.2820917779270657e-06,
+      "loss": 0.0165,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6511540273198304,
+      "grad_norm": 0.4007360339164734,
+      "learning_rate": 3.2768894099259636e-06,
+      "loss": 0.0171,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6514713538438655,
+      "grad_norm": 0.3016543388366699,
+      "learning_rate": 3.271689157881317e-06,
+      "loss": 0.0141,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6517886803679004,
+      "grad_norm": 0.3930988311767578,
+      "learning_rate": 3.266491028178964e-06,
+      "loss": 0.0157,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6521060068919354,
+      "grad_norm": 0.3150501847267151,
+      "learning_rate": 3.261295027202139e-06,
+      "loss": 0.0147,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6524233334159705,
+      "grad_norm": 0.33003994822502136,
+      "learning_rate": 3.256101161331463e-06,
+      "loss": 0.0151,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6527406599400054,
+      "grad_norm": 0.3820667862892151,
+      "learning_rate": 3.250909436944928e-06,
+      "loss": 0.0178,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6530579864640405,
+      "grad_norm": 1.094849944114685,
+      "learning_rate": 3.2457198604179025e-06,
+      "loss": 0.0162,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6533753129880755,
+      "grad_norm": 0.6185505986213684,
+      "learning_rate": 3.2405324381231174e-06,
+      "loss": 0.0164,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6536926395121104,
+      "grad_norm": 0.36891767382621765,
+      "learning_rate": 3.2353471764306567e-06,
+      "loss": 0.0166,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6540099660361455,
+      "grad_norm": 0.5257048606872559,
+      "learning_rate": 3.2301640817079506e-06,
+      "loss": 0.0181,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6543272925601805,
+      "grad_norm": 0.38972312211990356,
+      "learning_rate": 3.2249831603197705e-06,
+      "loss": 0.0169,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6546446190842155,
+      "grad_norm": 0.32267650961875916,
+      "learning_rate": 3.219804418628216e-06,
+      "loss": 0.0185,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6549619456082505,
+      "grad_norm": 0.2571272552013397,
+      "learning_rate": 3.2146278629927143e-06,
+      "loss": 0.0125,
+      "step": 2064
+    },
+    {
+      "epoch": 0.6552792721322855,
+      "grad_norm": 0.3337092101573944,
+      "learning_rate": 3.2094534997700033e-06,
+      "loss": 0.0152,
+      "step": 2065
+    },
+    {
+      "epoch": 0.6555965986563205,
+      "grad_norm": 1.1272650957107544,
+      "learning_rate": 3.2042813353141333e-06,
+      "loss": 0.0174,
+      "step": 2066
+    },
+    {
+      "epoch": 0.6559139251803555,
+      "grad_norm": 0.325348824262619,
+      "learning_rate": 3.1991113759764493e-06,
+      "loss": 0.0183,
+      "step": 2067
+    },
+    {
+      "epoch": 0.6562312517043906,
+      "grad_norm": 0.3323421776294708,
+      "learning_rate": 3.1939436281055923e-06,
+      "loss": 0.0174,
+      "step": 2068
+    },
+    {
+      "epoch": 0.6565485782284255,
+      "grad_norm": 0.3264688551425934,
+      "learning_rate": 3.188778098047487e-06,
+      "loss": 0.0175,
+      "step": 2069
+    },
+    {
+      "epoch": 0.6568659047524605,
+      "grad_norm": 0.4318616986274719,
+      "learning_rate": 3.183614792145334e-06,
+      "loss": 0.0161,
+      "step": 2070
+    },
+    {
+      "epoch": 0.6571832312764955,
+      "grad_norm": 0.3870818614959717,
+      "learning_rate": 3.1784537167396025e-06,
+      "loss": 0.0165,
+      "step": 2071
+    },
+    {
+      "epoch": 0.6575005578005305,
+      "grad_norm": 0.47984957695007324,
+      "learning_rate": 3.173294878168025e-06,
+      "loss": 0.0167,
+      "step": 2072
+    },
+    {
+      "epoch": 0.6578178843245656,
+      "grad_norm": 0.3075462877750397,
+      "learning_rate": 3.168138282765584e-06,
+      "loss": 0.0165,
+      "step": 2073
+    },
+    {
+      "epoch": 0.6581352108486005,
+      "grad_norm": 0.2950163781642914,
+      "learning_rate": 3.1629839368645087e-06,
+      "loss": 0.0161,
+      "step": 2074
+    },
+    {
+      "epoch": 0.6584525373726355,
+      "grad_norm": 0.42081788182258606,
+      "learning_rate": 3.1578318467942672e-06,
+      "loss": 0.0143,
+      "step": 2075
+    },
+    {
+      "epoch": 0.6587698638966706,
+      "grad_norm": 1.5085655450820923,
+      "learning_rate": 3.1526820188815564e-06,
+      "loss": 0.0189,
+      "step": 2076
+    },
+    {
+      "epoch": 0.6590871904207055,
+      "grad_norm": 0.410441130399704,
+      "learning_rate": 3.1475344594502934e-06,
+      "loss": 0.0159,
+      "step": 2077
+    },
+    {
+      "epoch": 0.6594045169447406,
+      "grad_norm": 0.3890712261199951,
+      "learning_rate": 3.1423891748216106e-06,
+      "loss": 0.016,
+      "step": 2078
+    },
+    {
+      "epoch": 0.6597218434687756,
+      "grad_norm": 0.3868752419948578,
+      "learning_rate": 3.1372461713138504e-06,
+      "loss": 0.0189,
+      "step": 2079
+    },
+    {
+      "epoch": 0.6600391699928105,
+      "grad_norm": 0.30048370361328125,
+      "learning_rate": 3.13210545524255e-06,
+      "loss": 0.0177,
+      "step": 2080
+    },
+    {
+      "epoch": 0.6603564965168456,
+      "grad_norm": 0.37871989607810974,
+      "learning_rate": 3.12696703292044e-06,
+      "loss": 0.0165,
+      "step": 2081
+    },
+    {
+      "epoch": 0.6606738230408806,
+      "grad_norm": 0.2954692542552948,
+      "learning_rate": 3.121830910657433e-06,
+      "loss": 0.0141,
+      "step": 2082
+    },
+    {
+      "epoch": 0.6609911495649156,
+      "grad_norm": 1.1161175966262817,
+      "learning_rate": 3.1166970947606174e-06,
+      "loss": 0.0173,
+      "step": 2083
+    },
+    {
+      "epoch": 0.6613084760889506,
+      "grad_norm": 0.5158510804176331,
+      "learning_rate": 3.1115655915342497e-06,
+      "loss": 0.017,
+      "step": 2084
+    },
+    {
+      "epoch": 0.6616258026129856,
+      "grad_norm": 0.45906102657318115,
+      "learning_rate": 3.106436407279746e-06,
+      "loss": 0.018,
+      "step": 2085
+    },
+    {
+      "epoch": 0.6619431291370206,
+      "grad_norm": 0.30168408155441284,
+      "learning_rate": 3.1013095482956778e-06,
+      "loss": 0.0173,
+      "step": 2086
+    },
+    {
+      "epoch": 0.6622604556610556,
+      "grad_norm": 0.30551382899284363,
+      "learning_rate": 3.0961850208777527e-06,
+      "loss": 0.0157,
+      "step": 2087
+    },
+    {
+      "epoch": 0.6625777821850906,
+      "grad_norm": 0.3919103741645813,
+      "learning_rate": 3.091062831318825e-06,
+      "loss": 0.0148,
+      "step": 2088
+    },
+    {
+      "epoch": 0.6628951087091256,
+      "grad_norm": 0.3018123507499695,
+      "learning_rate": 3.0859429859088735e-06,
+      "loss": 0.0154,
+      "step": 2089
+    },
+    {
+      "epoch": 0.6632124352331606,
+      "grad_norm": 0.30155760049819946,
+      "learning_rate": 3.0808254909349987e-06,
+      "loss": 0.0172,
+      "step": 2090
+    },
+    {
+      "epoch": 0.6635297617571956,
+      "grad_norm": 0.3371751308441162,
+      "learning_rate": 3.075710352681416e-06,
+      "loss": 0.0166,
+      "step": 2091
+    },
+    {
+      "epoch": 0.6638470882812306,
+      "grad_norm": 0.3174290359020233,
+      "learning_rate": 3.070597577429444e-06,
+      "loss": 0.0169,
+      "step": 2092
+    },
+    {
+      "epoch": 0.6641644148052657,
+      "grad_norm": 0.7387028932571411,
+      "learning_rate": 3.0654871714575023e-06,
+      "loss": 0.0138,
+      "step": 2093
+    },
+    {
+      "epoch": 0.6644817413293006,
+      "grad_norm": 0.3602951765060425,
+      "learning_rate": 3.060379141041101e-06,
+      "loss": 0.0143,
+      "step": 2094
+    },
+    {
+      "epoch": 0.6647990678533356,
+      "grad_norm": 0.3577573001384735,
+      "learning_rate": 3.0552734924528304e-06,
+      "loss": 0.0173,
+      "step": 2095
+    },
+    {
+      "epoch": 0.6651163943773707,
+      "grad_norm": 0.3842978775501251,
+      "learning_rate": 3.0501702319623617e-06,
+      "loss": 0.0144,
+      "step": 2096
+    },
+    {
+      "epoch": 0.6654337209014056,
+      "grad_norm": 0.49353650212287903,
+      "learning_rate": 3.0450693658364243e-06,
+      "loss": 0.0171,
+      "step": 2097
+    },
+    {
+      "epoch": 0.6657510474254407,
+      "grad_norm": 0.2863524258136749,
+      "learning_rate": 3.039970900338816e-06,
+      "loss": 0.0172,
+      "step": 2098
+    },
+    {
+      "epoch": 0.6660683739494757,
+      "grad_norm": 0.5062177777290344,
+      "learning_rate": 3.0348748417303826e-06,
+      "loss": 0.0175,
+      "step": 2099
+    },
+    {
+      "epoch": 0.6663857004735106,
+      "grad_norm": 0.3373948633670807,
+      "learning_rate": 3.029781196269015e-06,
+      "loss": 0.019,
+      "step": 2100
+    },
+    {
+      "epoch": 0.6667030269975457,
+      "grad_norm": 0.3201659321784973,
+      "learning_rate": 3.0246899702096405e-06,
+      "loss": 0.0163,
+      "step": 2101
+    },
+    {
+      "epoch": 0.6670203535215807,
+      "grad_norm": 0.423923134803772,
+      "learning_rate": 3.019601169804216e-06,
+      "loss": 0.0143,
+      "step": 2102
+    },
+    {
+      "epoch": 0.6673376800456157,
+      "grad_norm": 0.6040564775466919,
+      "learning_rate": 3.0145148013017202e-06,
+      "loss": 0.0181,
+      "step": 2103
+    },
+    {
+      "epoch": 0.6676550065696507,
+      "grad_norm": 0.3996431827545166,
+      "learning_rate": 3.0094308709481436e-06,
+      "loss": 0.018,
+      "step": 2104
+    },
+    {
+      "epoch": 0.6679723330936856,
+      "grad_norm": 0.2921179533004761,
+      "learning_rate": 3.0043493849864835e-06,
+      "loss": 0.0136,
+      "step": 2105
+    },
+    {
+      "epoch": 0.6682896596177207,
+      "grad_norm": 0.31724169850349426,
+      "learning_rate": 2.999270349656738e-06,
+      "loss": 0.0149,
+      "step": 2106
+    },
+    {
+      "epoch": 0.6686069861417557,
+      "grad_norm": 0.3376412093639374,
+      "learning_rate": 2.9941937711958912e-06,
+      "loss": 0.0142,
+      "step": 2107
+    },
+    {
+      "epoch": 0.6689243126657907,
+      "grad_norm": 0.5097676515579224,
+      "learning_rate": 2.989119655837913e-06,
+      "loss": 0.0139,
+      "step": 2108
+    },
+    {
+      "epoch": 0.6692416391898257,
+      "grad_norm": 0.5805062651634216,
+      "learning_rate": 2.9840480098137498e-06,
+      "loss": 0.0174,
+      "step": 2109
+    },
+    {
+      "epoch": 0.6695589657138608,
+      "grad_norm": 0.4081941843032837,
+      "learning_rate": 2.978978839351313e-06,
+      "loss": 0.0197,
+      "step": 2110
+    },
+    {
+      "epoch": 0.6698762922378957,
+      "grad_norm": 1.9659669399261475,
+      "learning_rate": 2.973912150675475e-06,
+      "loss": 0.0161,
+      "step": 2111
+    },
+    {
+      "epoch": 0.6701936187619307,
+      "grad_norm": 0.41173362731933594,
+      "learning_rate": 2.968847950008062e-06,
+      "loss": 0.0192,
+      "step": 2112
+    },
+    {
+      "epoch": 0.6705109452859658,
+      "grad_norm": 0.33073171973228455,
+      "learning_rate": 2.963786243567843e-06,
+      "loss": 0.0166,
+      "step": 2113
+    },
+    {
+      "epoch": 0.6708282718100007,
+      "grad_norm": 0.3505552411079407,
+      "learning_rate": 2.9587270375705244e-06,
+      "loss": 0.0154,
+      "step": 2114
+    },
+    {
+      "epoch": 0.6711455983340358,
+      "grad_norm": 0.34168368577957153,
+      "learning_rate": 2.9536703382287433e-06,
+      "loss": 0.0141,
+      "step": 2115
+    },
+    {
+      "epoch": 0.6714629248580708,
+      "grad_norm": 0.3472922146320343,
+      "learning_rate": 2.94861615175206e-06,
+      "loss": 0.0155,
+      "step": 2116
+    },
+    {
+      "epoch": 0.6717802513821057,
+      "grad_norm": 0.3465234041213989,
+      "learning_rate": 2.9435644843469434e-06,
+      "loss": 0.0182,
+      "step": 2117
+    },
+    {
+      "epoch": 0.6720975779061408,
+      "grad_norm": 0.7214925289154053,
+      "learning_rate": 2.9385153422167747e-06,
+      "loss": 0.0184,
+      "step": 2118
+    },
+    {
+      "epoch": 0.6724149044301758,
+      "grad_norm": 0.34361156821250916,
+      "learning_rate": 2.9334687315618303e-06,
+      "loss": 0.0162,
+      "step": 2119
+    },
+    {
+      "epoch": 0.6727322309542108,
+      "grad_norm": 0.3094915747642517,
+      "learning_rate": 2.9284246585792812e-06,
+      "loss": 0.0143,
+      "step": 2120
+    },
+    {
+      "epoch": 0.6730495574782458,
+      "grad_norm": 0.6872573494911194,
+      "learning_rate": 2.923383129463181e-06,
+      "loss": 0.0132,
+      "step": 2121
+    },
+    {
+      "epoch": 0.6733668840022807,
+      "grad_norm": 0.29732489585876465,
+      "learning_rate": 2.9183441504044573e-06,
+      "loss": 0.0136,
+      "step": 2122
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 0.4968198835849762,
+      "learning_rate": 2.9133077275909112e-06,
+      "loss": 0.0187,
+      "step": 2123
+    },
+    {
+      "epoch": 0.6740015370503508,
+      "grad_norm": 0.4629189372062683,
+      "learning_rate": 2.9082738672071973e-06,
+      "loss": 0.0142,
+      "step": 2124
+    },
+    {
+      "epoch": 0.6743188635743858,
+      "grad_norm": 0.3642481863498688,
+      "learning_rate": 2.9032425754348332e-06,
+      "loss": 0.0157,
+      "step": 2125
+    },
+    {
+      "epoch": 0.6746361900984208,
+      "grad_norm": 0.5122303366661072,
+      "learning_rate": 2.8982138584521734e-06,
+      "loss": 0.0136,
+      "step": 2126
+    },
+    {
+      "epoch": 0.6749535166224558,
+      "grad_norm": 0.3100191354751587,
+      "learning_rate": 2.8931877224344142e-06,
+      "loss": 0.013,
+      "step": 2127
+    },
+    {
+      "epoch": 0.6752708431464908,
+      "grad_norm": 2.334325075149536,
+      "learning_rate": 2.888164173553585e-06,
+      "loss": 0.0154,
+      "step": 2128
+    },
+    {
+      "epoch": 0.6755881696705258,
+      "grad_norm": 0.8937236070632935,
+      "learning_rate": 2.8831432179785314e-06,
+      "loss": 0.0167,
+      "step": 2129
+    },
+    {
+      "epoch": 0.6759054961945609,
+      "grad_norm": 0.4089837968349457,
+      "learning_rate": 2.8781248618749235e-06,
+      "loss": 0.0175,
+      "step": 2130
+    },
+    {
+      "epoch": 0.6762228227185958,
+      "grad_norm": 0.3515603840351105,
+      "learning_rate": 2.8731091114052313e-06,
+      "loss": 0.0154,
+      "step": 2131
+    },
+    {
+      "epoch": 0.6765401492426308,
+      "grad_norm": 0.6663767695426941,
+      "learning_rate": 2.8680959727287316e-06,
+      "loss": 0.0187,
+      "step": 2132
+    },
+    {
+      "epoch": 0.6768574757666659,
+      "grad_norm": 0.3048708140850067,
+      "learning_rate": 2.863085452001487e-06,
+      "loss": 0.0143,
+      "step": 2133
+    },
+    {
+      "epoch": 0.6771748022907008,
+      "grad_norm": 0.3935924470424652,
+      "learning_rate": 2.858077555376353e-06,
+      "loss": 0.0183,
+      "step": 2134
+    },
+    {
+      "epoch": 0.6774921288147359,
+      "grad_norm": 0.3416304886341095,
+      "learning_rate": 2.853072289002954e-06,
+      "loss": 0.0166,
+      "step": 2135
+    },
+    {
+      "epoch": 0.6778094553387709,
+      "grad_norm": 0.364504337310791,
+      "learning_rate": 2.8480696590276946e-06,
+      "loss": 0.0187,
+      "step": 2136
+    },
+    {
+      "epoch": 0.6781267818628058,
+      "grad_norm": 0.33134713768959045,
+      "learning_rate": 2.843069671593734e-06,
+      "loss": 0.0154,
+      "step": 2137
+    },
+    {
+      "epoch": 0.6784441083868409,
+      "grad_norm": 0.3196084499359131,
+      "learning_rate": 2.8380723328409872e-06,
+      "loss": 0.0177,
+      "step": 2138
+    },
+    {
+      "epoch": 0.6787614349108758,
+      "grad_norm": 0.2857343554496765,
+      "learning_rate": 2.833077648906122e-06,
+      "loss": 0.0136,
+      "step": 2139
+    },
+    {
+      "epoch": 0.6790787614349109,
+      "grad_norm": 0.3807588517665863,
+      "learning_rate": 2.828085625922539e-06,
+      "loss": 0.0144,
+      "step": 2140
+    },
+    {
+      "epoch": 0.6793960879589459,
+      "grad_norm": 0.31665682792663574,
+      "learning_rate": 2.823096270020379e-06,
+      "loss": 0.0137,
+      "step": 2141
+    },
+    {
+      "epoch": 0.6797134144829808,
+      "grad_norm": 0.6582081913948059,
+      "learning_rate": 2.818109587326499e-06,
+      "loss": 0.0148,
+      "step": 2142
+    },
+    {
+      "epoch": 0.6800307410070159,
+      "grad_norm": 0.38207006454467773,
+      "learning_rate": 2.813125583964482e-06,
+      "loss": 0.0131,
+      "step": 2143
+    },
+    {
+      "epoch": 0.6803480675310509,
+      "grad_norm": 0.3686722218990326,
+      "learning_rate": 2.8081442660546126e-06,
+      "loss": 0.0169,
+      "step": 2144
+    },
+    {
+      "epoch": 0.6806653940550859,
+      "grad_norm": 0.4547528624534607,
+      "learning_rate": 2.803165639713885e-06,
+      "loss": 0.0217,
+      "step": 2145
+    },
+    {
+      "epoch": 0.6809827205791209,
+      "grad_norm": 0.45979824662208557,
+      "learning_rate": 2.798189711055984e-06,
+      "loss": 0.0196,
+      "step": 2146
+    },
+    {
+      "epoch": 0.681300047103156,
+      "grad_norm": 0.4516683518886566,
+      "learning_rate": 2.7932164861912805e-06,
+      "loss": 0.0177,
+      "step": 2147
+    },
+    {
+      "epoch": 0.6816173736271909,
+      "grad_norm": 0.35137662291526794,
+      "learning_rate": 2.7882459712268306e-06,
+      "loss": 0.016,
+      "step": 2148
+    },
+    {
+      "epoch": 0.6819347001512259,
+      "grad_norm": 0.36649513244628906,
+      "learning_rate": 2.7832781722663547e-06,
+      "loss": 0.0156,
+      "step": 2149
+    },
+    {
+      "epoch": 0.682252026675261,
+      "grad_norm": 0.485613077878952,
+      "learning_rate": 2.7783130954102484e-06,
+      "loss": 0.0138,
+      "step": 2150
+    },
+    {
+      "epoch": 0.6825693531992959,
+      "grad_norm": 0.5274218320846558,
+      "learning_rate": 2.7733507467555532e-06,
+      "loss": 0.0171,
+      "step": 2151
+    },
+    {
+      "epoch": 0.682886679723331,
+      "grad_norm": 0.3891105353832245,
+      "learning_rate": 2.768391132395971e-06,
+      "loss": 0.0165,
+      "step": 2152
+    },
+    {
+      "epoch": 0.683204006247366,
+      "grad_norm": 0.48889485001564026,
+      "learning_rate": 2.7634342584218364e-06,
+      "loss": 0.0137,
+      "step": 2153
+    },
+    {
+      "epoch": 0.6835213327714009,
+      "grad_norm": 0.313838392496109,
+      "learning_rate": 2.758480130920128e-06,
+      "loss": 0.0148,
+      "step": 2154
+    },
+    {
+      "epoch": 0.683838659295436,
+      "grad_norm": 0.3248315453529358,
+      "learning_rate": 2.753528755974443e-06,
+      "loss": 0.0162,
+      "step": 2155
+    },
+    {
+      "epoch": 0.6841559858194709,
+      "grad_norm": 0.3159582018852234,
+      "learning_rate": 2.7485801396650067e-06,
+      "loss": 0.0137,
+      "step": 2156
+    },
+    {
+      "epoch": 0.684473312343506,
+      "grad_norm": 0.3258107900619507,
+      "learning_rate": 2.7436342880686505e-06,
+      "loss": 0.0152,
+      "step": 2157
+    },
+    {
+      "epoch": 0.684790638867541,
+      "grad_norm": 0.2763664722442627,
+      "learning_rate": 2.7386912072588123e-06,
+      "loss": 0.0133,
+      "step": 2158
+    },
+    {
+      "epoch": 0.6851079653915759,
+      "grad_norm": 0.33939462900161743,
+      "learning_rate": 2.733750903305531e-06,
+      "loss": 0.0158,
+      "step": 2159
+    },
+    {
+      "epoch": 0.685425291915611,
+      "grad_norm": 0.36634159088134766,
+      "learning_rate": 2.7288133822754298e-06,
+      "loss": 0.0169,
+      "step": 2160
+    },
+    {
+      "epoch": 0.685742618439646,
+      "grad_norm": 0.39323440194129944,
+      "learning_rate": 2.7238786502317213e-06,
+      "loss": 0.0176,
+      "step": 2161
+    },
+    {
+      "epoch": 0.686059944963681,
+      "grad_norm": 0.30379176139831543,
+      "learning_rate": 2.718946713234185e-06,
+      "loss": 0.016,
+      "step": 2162
+    },
+    {
+      "epoch": 0.686377271487716,
+      "grad_norm": 0.3969588875770569,
+      "learning_rate": 2.714017577339178e-06,
+      "loss": 0.0164,
+      "step": 2163
+    },
+    {
+      "epoch": 0.686694598011751,
+      "grad_norm": 0.681617259979248,
+      "learning_rate": 2.709091248599608e-06,
+      "loss": 0.0143,
+      "step": 2164
+    },
+    {
+      "epoch": 0.687011924535786,
+      "grad_norm": 0.39065876603126526,
+      "learning_rate": 2.7041677330649408e-06,
+      "loss": 0.0137,
+      "step": 2165
+    },
+    {
+      "epoch": 0.687329251059821,
+      "grad_norm": 0.43625757098197937,
+      "learning_rate": 2.6992470367811917e-06,
+      "loss": 0.02,
+      "step": 2166
+    },
+    {
+      "epoch": 0.6876465775838561,
+      "grad_norm": 0.46815308928489685,
+      "learning_rate": 2.694329165790901e-06,
+      "loss": 0.0184,
+      "step": 2167
+    },
+    {
+      "epoch": 0.687963904107891,
+      "grad_norm": 0.3831689655780792,
+      "learning_rate": 2.6894141261331542e-06,
+      "loss": 0.0173,
+      "step": 2168
+    },
+    {
+      "epoch": 0.688281230631926,
+      "grad_norm": 1.4392800331115723,
+      "learning_rate": 2.6845019238435476e-06,
+      "loss": 0.016,
+      "step": 2169
+    },
+    {
+      "epoch": 0.6885985571559611,
+      "grad_norm": 0.4797901511192322,
+      "learning_rate": 2.6795925649542053e-06,
+      "loss": 0.0195,
+      "step": 2170
+    },
+    {
+      "epoch": 0.688915883679996,
+      "grad_norm": 0.8687662482261658,
+      "learning_rate": 2.674686055493748e-06,
+      "loss": 0.0153,
+      "step": 2171
+    },
+    {
+      "epoch": 0.689233210204031,
+      "grad_norm": 0.37890276312828064,
+      "learning_rate": 2.6697824014873076e-06,
+      "loss": 0.0164,
+      "step": 2172
+    },
+    {
+      "epoch": 0.689550536728066,
+      "grad_norm": 0.3307332992553711,
+      "learning_rate": 2.6648816089565e-06,
+      "loss": 0.0153,
+      "step": 2173
+    },
+    {
+      "epoch": 0.689867863252101,
+      "grad_norm": 0.4897457957267761,
+      "learning_rate": 2.659983683919434e-06,
+      "loss": 0.0151,
+      "step": 2174
+    },
+    {
+      "epoch": 0.6901851897761361,
+      "grad_norm": 0.353871613740921,
+      "learning_rate": 2.6550886323906982e-06,
+      "loss": 0.0172,
+      "step": 2175
+    },
+    {
+      "epoch": 0.690502516300171,
+      "grad_norm": 0.4975787103176117,
+      "learning_rate": 2.650196460381347e-06,
+      "loss": 0.0152,
+      "step": 2176
+    },
+    {
+      "epoch": 0.690819842824206,
+      "grad_norm": 0.35999345779418945,
+      "learning_rate": 2.645307173898901e-06,
+      "loss": 0.0161,
+      "step": 2177
+    },
+    {
+      "epoch": 0.6911371693482411,
+      "grad_norm": 0.33458954095840454,
+      "learning_rate": 2.640420778947337e-06,
+      "loss": 0.0167,
+      "step": 2178
+    },
+    {
+      "epoch": 0.691454495872276,
+      "grad_norm": 0.428806871175766,
+      "learning_rate": 2.6355372815270837e-06,
+      "loss": 0.0156,
+      "step": 2179
+    },
+    {
+      "epoch": 0.6917718223963111,
+      "grad_norm": 0.37776508927345276,
+      "learning_rate": 2.6306566876350072e-06,
+      "loss": 0.0129,
+      "step": 2180
+    },
+    {
+      "epoch": 0.6920891489203461,
+      "grad_norm": 0.34307435154914856,
+      "learning_rate": 2.6257790032644124e-06,
+      "loss": 0.0148,
+      "step": 2181
+    },
+    {
+      "epoch": 0.692406475444381,
+      "grad_norm": 0.4149021804332733,
+      "learning_rate": 2.6209042344050318e-06,
+      "loss": 0.0146,
+      "step": 2182
+    },
+    {
+      "epoch": 0.6927238019684161,
+      "grad_norm": 0.3160228431224823,
+      "learning_rate": 2.616032387043011e-06,
+      "loss": 0.0152,
+      "step": 2183
+    },
+    {
+      "epoch": 0.6930411284924511,
+      "grad_norm": 0.3062594532966614,
+      "learning_rate": 2.6111634671609166e-06,
+      "loss": 0.0134,
+      "step": 2184
+    },
+    {
+      "epoch": 0.6933584550164861,
+      "grad_norm": 0.3333306610584259,
+      "learning_rate": 2.606297480737713e-06,
+      "loss": 0.0163,
+      "step": 2185
+    },
+    {
+      "epoch": 0.6936757815405211,
+      "grad_norm": 0.32482069730758667,
+      "learning_rate": 2.601434433748771e-06,
+      "loss": 0.0141,
+      "step": 2186
+    },
+    {
+      "epoch": 0.6939931080645562,
+      "grad_norm": 0.39419203996658325,
+      "learning_rate": 2.5965743321658384e-06,
+      "loss": 0.0156,
+      "step": 2187
+    },
+    {
+      "epoch": 0.6943104345885911,
+      "grad_norm": 0.2942584455013275,
+      "learning_rate": 2.5917171819570596e-06,
+      "loss": 0.0149,
+      "step": 2188
+    },
+    {
+      "epoch": 0.6946277611126261,
+      "grad_norm": 0.2823149859905243,
+      "learning_rate": 2.5868629890869467e-06,
+      "loss": 0.0136,
+      "step": 2189
+    },
+    {
+      "epoch": 0.6949450876366612,
+      "grad_norm": 0.36929309368133545,
+      "learning_rate": 2.582011759516383e-06,
+      "loss": 0.0153,
+      "step": 2190
+    },
+    {
+      "epoch": 0.6952624141606961,
+      "grad_norm": 0.37845832109451294,
+      "learning_rate": 2.577163499202616e-06,
+      "loss": 0.0137,
+      "step": 2191
+    },
+    {
+      "epoch": 0.6955797406847312,
+      "grad_norm": 0.31392261385917664,
+      "learning_rate": 2.5723182140992385e-06,
+      "loss": 0.0145,
+      "step": 2192
+    },
+    {
+      "epoch": 0.6958970672087661,
+      "grad_norm": 0.3041812479496002,
+      "learning_rate": 2.567475910156201e-06,
+      "loss": 0.0139,
+      "step": 2193
+    },
+    {
+      "epoch": 0.6962143937328011,
+      "grad_norm": 0.4810658395290375,
+      "learning_rate": 2.5626365933197807e-06,
+      "loss": 0.0154,
+      "step": 2194
+    },
+    {
+      "epoch": 0.6965317202568362,
+      "grad_norm": 0.51142418384552,
+      "learning_rate": 2.5578002695325986e-06,
+      "loss": 0.0163,
+      "step": 2195
+    },
+    {
+      "epoch": 0.6968490467808711,
+      "grad_norm": 0.36141523718833923,
+      "learning_rate": 2.552966944733592e-06,
+      "loss": 0.0154,
+      "step": 2196
+    },
+    {
+      "epoch": 0.6971663733049062,
+      "grad_norm": 0.374225378036499,
+      "learning_rate": 2.5481366248580165e-06,
+      "loss": 0.0174,
+      "step": 2197
+    },
+    {
+      "epoch": 0.6974836998289412,
+      "grad_norm": 0.3976570665836334,
+      "learning_rate": 2.543309315837444e-06,
+      "loss": 0.0157,
+      "step": 2198
+    },
+    {
+      "epoch": 0.6978010263529761,
+      "grad_norm": 0.3860352039337158,
+      "learning_rate": 2.53848502359974e-06,
+      "loss": 0.0161,
+      "step": 2199
+    },
+    {
+      "epoch": 0.6981183528770112,
+      "grad_norm": 0.6633860468864441,
+      "learning_rate": 2.533663754069074e-06,
+      "loss": 0.0133,
+      "step": 2200
+    },
+    {
+      "epoch": 0.6984356794010462,
+      "grad_norm": 0.40059465169906616,
+      "learning_rate": 2.528845513165896e-06,
+      "loss": 0.015,
+      "step": 2201
+    },
+    {
+      "epoch": 0.6987530059250812,
+      "grad_norm": 0.5123053789138794,
+      "learning_rate": 2.5240303068069436e-06,
+      "loss": 0.0144,
+      "step": 2202
+    },
+    {
+      "epoch": 0.6990703324491162,
+      "grad_norm": 0.5301564931869507,
+      "learning_rate": 2.519218140905223e-06,
+      "loss": 0.0154,
+      "step": 2203
+    },
+    {
+      "epoch": 0.6993876589731512,
+      "grad_norm": 0.3864990770816803,
+      "learning_rate": 2.5144090213700103e-06,
+      "loss": 0.0155,
+      "step": 2204
+    },
+    {
+      "epoch": 0.6997049854971862,
+      "grad_norm": 0.46803587675094604,
+      "learning_rate": 2.5096029541068367e-06,
+      "loss": 0.0179,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7000223120212212,
+      "grad_norm": 0.7264665365219116,
+      "learning_rate": 2.504799945017491e-06,
+      "loss": 0.0162,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7003396385452563,
+      "grad_norm": 0.40292418003082275,
+      "learning_rate": 2.5000000000000015e-06,
+      "loss": 0.0171,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7006569650692912,
+      "grad_norm": 0.41257205605506897,
+      "learning_rate": 2.495203124948633e-06,
+      "loss": 0.0148,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7009742915933262,
+      "grad_norm": 0.8459236025810242,
+      "learning_rate": 2.4904093257538887e-06,
+      "loss": 0.0181,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7012916181173612,
+      "grad_norm": 0.4207494854927063,
+      "learning_rate": 2.4856186083024836e-06,
+      "loss": 0.0153,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7016089446413962,
+      "grad_norm": 0.4271310567855835,
+      "learning_rate": 2.480830978477358e-06,
+      "loss": 0.0153,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7019262711654313,
+      "grad_norm": 0.33121222257614136,
+      "learning_rate": 2.476046442157653e-06,
+      "loss": 0.0163,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7022435976894662,
+      "grad_norm": 0.7708227634429932,
+      "learning_rate": 2.4712650052187174e-06,
+      "loss": 0.0132,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7025609242135012,
+      "grad_norm": 0.38019660115242004,
+      "learning_rate": 2.4664866735320886e-06,
+      "loss": 0.0138,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7028782507375363,
+      "grad_norm": 1.0516713857650757,
+      "learning_rate": 2.4617114529654957e-06,
+      "loss": 0.0136,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7031955772615712,
+      "grad_norm": 0.3525136411190033,
+      "learning_rate": 2.4569393493828433e-06,
+      "loss": 0.0147,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7035129037856063,
+      "grad_norm": 0.3800276815891266,
+      "learning_rate": 2.4521703686442076e-06,
+      "loss": 0.0173,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7038302303096413,
+      "grad_norm": 0.3363269865512848,
+      "learning_rate": 2.4474045166058373e-06,
+      "loss": 0.0149,
+      "step": 2218
+    },
+    {
+      "epoch": 0.7041475568336762,
+      "grad_norm": 0.6546519994735718,
+      "learning_rate": 2.4426417991201294e-06,
+      "loss": 0.0156,
+      "step": 2219
+    },
+    {
+      "epoch": 0.7044648833577113,
+      "grad_norm": 0.38113677501678467,
+      "learning_rate": 2.4378822220356407e-06,
+      "loss": 0.0175,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7047822098817463,
+      "grad_norm": 1.1958566904067993,
+      "learning_rate": 2.4331257911970628e-06,
+      "loss": 0.0184,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7050995364057813,
+      "grad_norm": 0.7342499494552612,
+      "learning_rate": 2.428372512445233e-06,
+      "loss": 0.0124,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7054168629298163,
+      "grad_norm": 0.2954676151275635,
+      "learning_rate": 2.423622391617109e-06,
+      "loss": 0.0125,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7057341894538514,
+      "grad_norm": 0.37941333651542664,
+      "learning_rate": 2.4188754345457783e-06,
+      "loss": 0.0147,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7060515159778863,
+      "grad_norm": 0.42071616649627686,
+      "learning_rate": 2.4141316470604362e-06,
+      "loss": 0.0142,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7063688425019213,
+      "grad_norm": 0.3100742697715759,
+      "learning_rate": 2.409391034986394e-06,
+      "loss": 0.0138,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7066861690259563,
+      "grad_norm": 0.5162502527236938,
+      "learning_rate": 2.404653604145057e-06,
+      "loss": 0.0143,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7070034955499913,
+      "grad_norm": 0.3232317566871643,
+      "learning_rate": 2.3999193603539234e-06,
+      "loss": 0.0145,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7073208220740264,
+      "grad_norm": 0.6358088850975037,
+      "learning_rate": 2.3951883094265842e-06,
+      "loss": 0.0201,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7076381485980613,
+      "grad_norm": 0.3418027460575104,
+      "learning_rate": 2.390460457172703e-06,
+      "loss": 0.0182,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7079554751220963,
+      "grad_norm": 0.3283490240573883,
+      "learning_rate": 2.3857358093980217e-06,
+      "loss": 0.0149,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7082728016461314,
+      "grad_norm": 0.42807942628860474,
+      "learning_rate": 2.3810143719043393e-06,
+      "loss": 0.0207,
+      "step": 2232
+    },
+    {
+      "epoch": 0.7085901281701663,
+      "grad_norm": 0.5114642977714539,
+      "learning_rate": 2.3762961504895218e-06,
+      "loss": 0.0174,
+      "step": 2233
+    },
+    {
+      "epoch": 0.7089074546942014,
+      "grad_norm": 0.9146451950073242,
+      "learning_rate": 2.371581150947476e-06,
+      "loss": 0.0145,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7092247812182364,
+      "grad_norm": 0.3744998872280121,
+      "learning_rate": 2.3668693790681634e-06,
+      "loss": 0.015,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7095421077422713,
+      "grad_norm": 0.7724823355674744,
+      "learning_rate": 2.362160840637572e-06,
+      "loss": 0.0133,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7098594342663064,
+      "grad_norm": 0.3056487739086151,
+      "learning_rate": 2.357455541437723e-06,
+      "loss": 0.0138,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7101767607903414,
+      "grad_norm": 0.35063448548316956,
+      "learning_rate": 2.352753487246663e-06,
+      "loss": 0.0145,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7104940873143764,
+      "grad_norm": 0.409842848777771,
+      "learning_rate": 2.3480546838384466e-06,
+      "loss": 0.0168,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7108114138384114,
+      "grad_norm": 0.34883037209510803,
+      "learning_rate": 2.3433591369831455e-06,
+      "loss": 0.0181,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7111287403624464,
+      "grad_norm": 0.404796302318573,
+      "learning_rate": 2.338666852446823e-06,
+      "loss": 0.0155,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7114460668864814,
+      "grad_norm": 0.6545796990394592,
+      "learning_rate": 2.333977835991545e-06,
+      "loss": 0.0132,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7117633934105164,
+      "grad_norm": 0.4031830430030823,
+      "learning_rate": 2.3292920933753566e-06,
+      "loss": 0.0152,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7120807199345514,
+      "grad_norm": 0.9239604473114014,
+      "learning_rate": 2.3246096303522875e-06,
+      "loss": 0.0173,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7123980464585864,
+      "grad_norm": 0.30136194825172424,
+      "learning_rate": 2.3199304526723417e-06,
+      "loss": 0.0137,
+      "step": 2245
+    },
+    {
+      "epoch": 0.7127153729826214,
+      "grad_norm": 0.30847397446632385,
+      "learning_rate": 2.3152545660814835e-06,
+      "loss": 0.0165,
+      "step": 2246
+    },
+    {
+      "epoch": 0.7130326995066564,
+      "grad_norm": 0.4371183514595032,
+      "learning_rate": 2.310581976321638e-06,
+      "loss": 0.0156,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7133500260306914,
+      "grad_norm": 0.44709643721580505,
+      "learning_rate": 2.305912689130681e-06,
+      "loss": 0.0146,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7136673525547265,
+      "grad_norm": 0.34953901171684265,
+      "learning_rate": 2.3012467102424373e-06,
+      "loss": 0.0141,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7139846790787614,
+      "grad_norm": 0.2779126763343811,
+      "learning_rate": 2.296584045386662e-06,
+      "loss": 0.0135,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7143020056027964,
+      "grad_norm": 0.3999202251434326,
+      "learning_rate": 2.2919247002890473e-06,
+      "loss": 0.0186,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7146193321268315,
+      "grad_norm": 0.43698573112487793,
+      "learning_rate": 2.2872686806712037e-06,
+      "loss": 0.0154,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7149366586508664,
+      "grad_norm": 0.2771173417568207,
+      "learning_rate": 2.2826159922506603e-06,
+      "loss": 0.0129,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7152539851749015,
+      "grad_norm": 2.255387544631958,
+      "learning_rate": 2.2779666407408592e-06,
+      "loss": 0.0166,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7155713116989365,
+      "grad_norm": 0.3429323136806488,
+      "learning_rate": 2.2733206318511354e-06,
+      "loss": 0.0138,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7158886382229714,
+      "grad_norm": 0.3615747392177582,
+      "learning_rate": 2.268677971286732e-06,
+      "loss": 0.0176,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7162059647470065,
+      "grad_norm": 0.303710401058197,
+      "learning_rate": 2.264038664748765e-06,
+      "loss": 0.0141,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7165232912710415,
+      "grad_norm": 0.3559911549091339,
+      "learning_rate": 2.259402717934246e-06,
+      "loss": 0.0148,
+      "step": 2258
+    },
+    {
+      "epoch": 0.7168406177950765,
+      "grad_norm": 0.3069722056388855,
+      "learning_rate": 2.2547701365360504e-06,
+      "loss": 0.0147,
+      "step": 2259
+    },
+    {
+      "epoch": 0.7171579443191115,
+      "grad_norm": 0.2715959846973419,
+      "learning_rate": 2.250140926242928e-06,
+      "loss": 0.0114,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7174752708431464,
+      "grad_norm": 0.31905630230903625,
+      "learning_rate": 2.245515092739488e-06,
+      "loss": 0.0171,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7177925973671815,
+      "grad_norm": 0.34318119287490845,
+      "learning_rate": 2.2408926417061874e-06,
+      "loss": 0.0142,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7181099238912165,
+      "grad_norm": 0.31723517179489136,
+      "learning_rate": 2.236273578819337e-06,
+      "loss": 0.0169,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7184272504152515,
+      "grad_norm": 0.2600262463092804,
+      "learning_rate": 2.23165790975108e-06,
+      "loss": 0.0122,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7187445769392865,
+      "grad_norm": 0.46086031198501587,
+      "learning_rate": 2.227045640169399e-06,
+      "loss": 0.0183,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7190619034633216,
+      "grad_norm": 0.36903125047683716,
+      "learning_rate": 2.222436775738097e-06,
+      "loss": 0.0181,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7193792299873565,
+      "grad_norm": 0.40168410539627075,
+      "learning_rate": 2.217831322116797e-06,
+      "loss": 0.0146,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7196965565113915,
+      "grad_norm": 0.4546142816543579,
+      "learning_rate": 2.2132292849609315e-06,
+      "loss": 0.0155,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7200138830354266,
+      "grad_norm": 0.7269618511199951,
+      "learning_rate": 2.2086306699217418e-06,
+      "loss": 0.0178,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7203312095594615,
+      "grad_norm": 0.32575419545173645,
+      "learning_rate": 2.204035482646267e-06,
+      "loss": 0.0162,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7206485360834965,
+      "grad_norm": 0.4135962128639221,
+      "learning_rate": 2.19944372877733e-06,
+      "loss": 0.0153,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7209658626075316,
+      "grad_norm": 0.4032052457332611,
+      "learning_rate": 2.1948554139535472e-06,
+      "loss": 0.0162,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7212831891315665,
+      "grad_norm": 0.26178789138793945,
+      "learning_rate": 2.190270543809303e-06,
+      "loss": 0.0128,
+      "step": 2273
+    },
+    {
+      "epoch": 0.7216005156556016,
+      "grad_norm": 0.3340522050857544,
+      "learning_rate": 2.185689123974759e-06,
+      "loss": 0.0144,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7219178421796366,
+      "grad_norm": 0.3374699652194977,
+      "learning_rate": 2.181111160075834e-06,
+      "loss": 0.017,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7222351687036715,
+      "grad_norm": 0.4458087384700775,
+      "learning_rate": 2.1765366577342083e-06,
+      "loss": 0.0179,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7225524952277066,
+      "grad_norm": 0.4413011074066162,
+      "learning_rate": 2.171965622567308e-06,
+      "loss": 0.0169,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7228698217517415,
+      "grad_norm": 0.307363897562027,
+      "learning_rate": 2.167398060188299e-06,
+      "loss": 0.0137,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7231871482757766,
+      "grad_norm": 0.4098169207572937,
+      "learning_rate": 2.162833976206092e-06,
+      "loss": 0.0142,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7235044747998116,
+      "grad_norm": 2.996056079864502,
+      "learning_rate": 2.158273376225314e-06,
+      "loss": 0.0142,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7238218013238465,
+      "grad_norm": 0.43798860907554626,
+      "learning_rate": 2.153716265846325e-06,
+      "loss": 0.0161,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7241391278478816,
+      "grad_norm": 0.28707703948020935,
+      "learning_rate": 2.1491626506651914e-06,
+      "loss": 0.0111,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7244564543719166,
+      "grad_norm": 0.3009926676750183,
+      "learning_rate": 2.144612536273694e-06,
+      "loss": 0.0111,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7247737808959516,
+      "grad_norm": 0.3915623426437378,
+      "learning_rate": 2.1400659282593083e-06,
+      "loss": 0.0133,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7250911074199866,
+      "grad_norm": 0.32315778732299805,
+      "learning_rate": 2.13552283220521e-06,
+      "loss": 0.0146,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7254084339440217,
+      "grad_norm": 0.44342291355133057,
+      "learning_rate": 2.1309832536902596e-06,
+      "loss": 0.0137,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7257257604680566,
+      "grad_norm": 0.330182284116745,
+      "learning_rate": 2.1264471982889945e-06,
+      "loss": 0.0131,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7260430869920916,
+      "grad_norm": 0.3824295401573181,
+      "learning_rate": 2.1219146715716332e-06,
+      "loss": 0.0164,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7263604135161267,
+      "grad_norm": 0.4241666793823242,
+      "learning_rate": 2.1173856791040543e-06,
+      "loss": 0.0146,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7266777400401616,
+      "grad_norm": 0.4282054603099823,
+      "learning_rate": 2.1128602264478022e-06,
+      "loss": 0.0133,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7269950665641967,
+      "grad_norm": 0.4379190504550934,
+      "learning_rate": 2.1083383191600676e-06,
+      "loss": 0.0168,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7273123930882317,
+      "grad_norm": 0.3186231553554535,
+      "learning_rate": 2.103819962793695e-06,
+      "loss": 0.0139,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7276297196122666,
+      "grad_norm": 0.5427785515785217,
+      "learning_rate": 2.099305162897162e-06,
+      "loss": 0.0171,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7279470461363017,
+      "grad_norm": 0.3104928433895111,
+      "learning_rate": 2.0947939250145843e-06,
+      "loss": 0.0127,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7282643726603366,
+      "grad_norm": 0.6673256158828735,
+      "learning_rate": 2.0902862546856978e-06,
+      "loss": 0.0212,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7285816991843717,
+      "grad_norm": 0.41772598028182983,
+      "learning_rate": 2.085782157445864e-06,
+      "loss": 0.0142,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7288990257084067,
+      "grad_norm": 0.8223555684089661,
+      "learning_rate": 2.081281638826052e-06,
+      "loss": 0.014,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7292163522324416,
+      "grad_norm": 0.4050436317920685,
+      "learning_rate": 2.076784704352835e-06,
+      "loss": 0.0157,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7295336787564767,
+      "grad_norm": 0.47505250573158264,
+      "learning_rate": 2.0722913595483917e-06,
+      "loss": 0.0184,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7298510052805117,
+      "grad_norm": 0.3594221770763397,
+      "learning_rate": 2.0678016099304853e-06,
+      "loss": 0.0137,
+      "step": 2300
+    },
+    {
+      "epoch": 0.7301683318045467,
+      "grad_norm": 0.3560483753681183,
+      "learning_rate": 2.0633154610124707e-06,
+      "loss": 0.017,
+      "step": 2301
+    },
+    {
+      "epoch": 0.7304856583285817,
+      "grad_norm": 0.35778406262397766,
+      "learning_rate": 2.0588329183032745e-06,
+      "loss": 0.016,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7308029848526167,
+      "grad_norm": 0.35666322708129883,
+      "learning_rate": 2.054353987307402e-06,
+      "loss": 0.0169,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7311203113766517,
+      "grad_norm": 0.3991971015930176,
+      "learning_rate": 2.049878673524916e-06,
+      "loss": 0.0153,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7314376379006867,
+      "grad_norm": 0.3205359876155853,
+      "learning_rate": 2.0454069824514445e-06,
+      "loss": 0.0136,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7317549644247218,
+      "grad_norm": 0.3009489178657532,
+      "learning_rate": 2.0409389195781627e-06,
+      "loss": 0.0149,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7320722909487567,
+      "grad_norm": 0.41346052289009094,
+      "learning_rate": 2.0364744903917886e-06,
+      "loss": 0.0159,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7323896174727917,
+      "grad_norm": 0.5148558020591736,
+      "learning_rate": 2.032013700374585e-06,
+      "loss": 0.0131,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7327069439968268,
+      "grad_norm": 0.3071138560771942,
+      "learning_rate": 2.0275565550043376e-06,
+      "loss": 0.0145,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7330242705208617,
+      "grad_norm": 0.29141765832901,
+      "learning_rate": 2.0231030597543645e-06,
+      "loss": 0.012,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7333415970448968,
+      "grad_norm": 0.33958831429481506,
+      "learning_rate": 2.018653220093493e-06,
+      "loss": 0.0155,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7336589235689317,
+      "grad_norm": 0.26582571864128113,
+      "learning_rate": 2.0142070414860704e-06,
+      "loss": 0.0109,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7339762500929667,
+      "grad_norm": 0.5817447304725647,
+      "learning_rate": 2.0097645293919397e-06,
+      "loss": 0.016,
+      "step": 2313
+    },
+    {
+      "epoch": 0.7342935766170018,
+      "grad_norm": 0.44592636823654175,
+      "learning_rate": 2.0053256892664495e-06,
+      "loss": 0.0202,
+      "step": 2314
+    },
+    {
+      "epoch": 0.7346109031410367,
+      "grad_norm": 0.26412898302078247,
+      "learning_rate": 2.0008905265604316e-06,
+      "loss": 0.012,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7349282296650718,
+      "grad_norm": 0.41788673400878906,
+      "learning_rate": 1.996459046720209e-06,
+      "loss": 0.0118,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7352455561891068,
+      "grad_norm": 0.4503333866596222,
+      "learning_rate": 1.9920312551875777e-06,
+      "loss": 0.0148,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7355628827131417,
+      "grad_norm": 0.3732012212276459,
+      "learning_rate": 1.9876071573998036e-06,
+      "loss": 0.0172,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7358802092371768,
+      "grad_norm": 0.45048144459724426,
+      "learning_rate": 1.983186758789622e-06,
+      "loss": 0.0164,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7361975357612118,
+      "grad_norm": 0.5097677111625671,
+      "learning_rate": 1.9787700647852197e-06,
+      "loss": 0.0126,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7365148622852468,
+      "grad_norm": 0.3380756974220276,
+      "learning_rate": 1.9743570808102402e-06,
+      "loss": 0.0151,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7368321888092818,
+      "grad_norm": 0.36480513215065,
+      "learning_rate": 1.9699478122837647e-06,
+      "loss": 0.0153,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7371495153333169,
+      "grad_norm": 0.406808078289032,
+      "learning_rate": 1.9655422646203166e-06,
+      "loss": 0.0151,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7374668418573518,
+      "grad_norm": 0.3189714848995209,
+      "learning_rate": 1.9611404432298505e-06,
+      "loss": 0.0135,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7377841683813868,
+      "grad_norm": 0.31413108110427856,
+      "learning_rate": 1.9567423535177405e-06,
+      "loss": 0.0127,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7381014949054219,
+      "grad_norm": 0.32846975326538086,
+      "learning_rate": 1.9523480008847856e-06,
+      "loss": 0.0167,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7384188214294568,
+      "grad_norm": 0.3956926763057709,
+      "learning_rate": 1.947957390727185e-06,
+      "loss": 0.0174,
+      "step": 2327
+    },
+    {
+      "epoch": 0.7387361479534919,
+      "grad_norm": 0.37404265999794006,
+      "learning_rate": 1.9435705284365537e-06,
+      "loss": 0.0141,
+      "step": 2328
+    },
+    {
+      "epoch": 0.7390534744775268,
+      "grad_norm": 0.8427812457084656,
+      "learning_rate": 1.939187419399895e-06,
+      "loss": 0.0145,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7393708010015618,
+      "grad_norm": 0.40433818101882935,
+      "learning_rate": 1.9348080689996112e-06,
+      "loss": 0.0186,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7396881275255969,
+      "grad_norm": 0.30524200201034546,
+      "learning_rate": 1.9304324826134823e-06,
+      "loss": 0.0134,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7400054540496318,
+      "grad_norm": 0.46382248401641846,
+      "learning_rate": 1.9260606656146702e-06,
+      "loss": 0.0138,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7403227805736668,
+      "grad_norm": 0.27185261249542236,
+      "learning_rate": 1.9216926233717087e-06,
+      "loss": 0.0133,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7406401070977019,
+      "grad_norm": 0.3239748477935791,
+      "learning_rate": 1.9173283612484916e-06,
+      "loss": 0.0146,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7409574336217368,
+      "grad_norm": 0.6387577056884766,
+      "learning_rate": 1.9129678846042764e-06,
+      "loss": 0.014,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7412747601457719,
+      "grad_norm": 0.42953187227249146,
+      "learning_rate": 1.9086111987936677e-06,
+      "loss": 0.0184,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7415920866698069,
+      "grad_norm": 0.3009189963340759,
+      "learning_rate": 1.9042583091666173e-06,
+      "loss": 0.0138,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7419094131938418,
+      "grad_norm": 0.4307123124599457,
+      "learning_rate": 1.8999092210684122e-06,
+      "loss": 0.0161,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7422267397178769,
+      "grad_norm": 0.6371519565582275,
+      "learning_rate": 1.8955639398396759e-06,
+      "loss": 0.0161,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7425440662419119,
+      "grad_norm": 0.3582170307636261,
+      "learning_rate": 1.8912224708163561e-06,
+      "loss": 0.0174,
+      "step": 2340
+    },
+    {
+      "epoch": 0.7428613927659469,
+      "grad_norm": 0.5800724029541016,
+      "learning_rate": 1.8868848193297152e-06,
+      "loss": 0.0171,
+      "step": 2341
+    },
+    {
+      "epoch": 0.7431787192899819,
+      "grad_norm": 0.46966421604156494,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.0131,
+      "step": 2342
+    },
+    {
+      "epoch": 0.743496045814017,
+      "grad_norm": 0.3825952410697937,
+      "learning_rate": 1.8782209902680893e-06,
+      "loss": 0.0173,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7438133723380519,
+      "grad_norm": 0.2511214315891266,
+      "learning_rate": 1.8738948233321696e-06,
+      "loss": 0.0104,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7441306988620869,
+      "grad_norm": 0.49588170647621155,
+      "learning_rate": 1.8695724952110445e-06,
+      "loss": 0.0158,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7444480253861219,
+      "grad_norm": 0.29558849334716797,
+      "learning_rate": 1.8652540112124796e-06,
+      "loss": 0.0141,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7447653519101569,
+      "grad_norm": 0.33012712001800537,
+      "learning_rate": 1.8609393766395083e-06,
+      "loss": 0.0161,
+      "step": 2347
+    },
+    {
+      "epoch": 0.745082678434192,
+      "grad_norm": 0.302519828081131,
+      "learning_rate": 1.8566285967904462e-06,
+      "loss": 0.017,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7454000049582269,
+      "grad_norm": 0.4033617377281189,
+      "learning_rate": 1.8523216769588749e-06,
+      "loss": 0.0143,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7457173314822619,
+      "grad_norm": 0.3278765380382538,
+      "learning_rate": 1.848018622433629e-06,
+      "loss": 0.0137,
+      "step": 2350
+    },
+    {
+      "epoch": 0.746034658006297,
+      "grad_norm": 0.3197358548641205,
+      "learning_rate": 1.843719438498806e-06,
+      "loss": 0.0132,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7463519845303319,
+      "grad_norm": 0.3582751750946045,
+      "learning_rate": 1.8394241304337412e-06,
+      "loss": 0.0185,
+      "step": 2352
+    },
+    {
+      "epoch": 0.746669311054367,
+      "grad_norm": 0.3189921975135803,
+      "learning_rate": 1.8351327035130184e-06,
+      "loss": 0.0145,
+      "step": 2353
+    },
+    {
+      "epoch": 0.746986637578402,
+      "grad_norm": 0.29736804962158203,
+      "learning_rate": 1.8308451630064484e-06,
+      "loss": 0.0143,
+      "step": 2354
+    },
+    {
+      "epoch": 0.7473039641024369,
+      "grad_norm": 0.30976274609565735,
+      "learning_rate": 1.826561514179076e-06,
+      "loss": 0.0144,
+      "step": 2355
+    },
+    {
+      "epoch": 0.747621290626472,
+      "grad_norm": 0.40029004216194153,
+      "learning_rate": 1.822281762291162e-06,
+      "loss": 0.0113,
+      "step": 2356
+    },
+    {
+      "epoch": 0.747938617150507,
+      "grad_norm": 0.35772761702537537,
+      "learning_rate": 1.8180059125981826e-06,
+      "loss": 0.0135,
+      "step": 2357
+    },
+    {
+      "epoch": 0.748255943674542,
+      "grad_norm": 0.3627467155456543,
+      "learning_rate": 1.8137339703508267e-06,
+      "loss": 0.0167,
+      "step": 2358
+    },
+    {
+      "epoch": 0.748573270198577,
+      "grad_norm": 0.5716778039932251,
+      "learning_rate": 1.8094659407949778e-06,
+      "loss": 0.0137,
+      "step": 2359
+    },
+    {
+      "epoch": 0.748890596722612,
+      "grad_norm": 0.28971534967422485,
+      "learning_rate": 1.8052018291717216e-06,
+      "loss": 0.014,
+      "step": 2360
+    },
+    {
+      "epoch": 0.749207923246647,
+      "grad_norm": 0.36312246322631836,
+      "learning_rate": 1.8009416407173258e-06,
+      "loss": 0.0151,
+      "step": 2361
+    },
+    {
+      "epoch": 0.749525249770682,
+      "grad_norm": 0.3118540346622467,
+      "learning_rate": 1.7966853806632484e-06,
+      "loss": 0.0155,
+      "step": 2362
+    },
+    {
+      "epoch": 0.749842576294717,
+      "grad_norm": 0.3246113359928131,
+      "learning_rate": 1.7924330542361151e-06,
+      "loss": 0.0137,
+      "step": 2363
+    },
+    {
+      "epoch": 0.750159902818752,
+      "grad_norm": 0.3798889219760895,
+      "learning_rate": 1.7881846666577284e-06,
+      "loss": 0.0146,
+      "step": 2364
+    },
+    {
+      "epoch": 0.750477229342787,
+      "grad_norm": 0.3158401548862457,
+      "learning_rate": 1.7839402231450475e-06,
+      "loss": 0.015,
+      "step": 2365
+    },
+    {
+      "epoch": 0.750794555866822,
+      "grad_norm": 0.2810014486312866,
+      "learning_rate": 1.7796997289101947e-06,
+      "loss": 0.0127,
+      "step": 2366
+    },
+    {
+      "epoch": 0.751111882390857,
+      "grad_norm": 0.3579384982585907,
+      "learning_rate": 1.7754631891604384e-06,
+      "loss": 0.0138,
+      "step": 2367
+    },
+    {
+      "epoch": 0.7514292089148921,
+      "grad_norm": 0.5330855250358582,
+      "learning_rate": 1.7712306090981896e-06,
+      "loss": 0.0184,
+      "step": 2368
+    },
+    {
+      "epoch": 0.751746535438927,
+      "grad_norm": 0.300491064786911,
+      "learning_rate": 1.7670019939210025e-06,
+      "loss": 0.0128,
+      "step": 2369
+    },
+    {
+      "epoch": 0.752063861962962,
+      "grad_norm": 0.30303189158439636,
+      "learning_rate": 1.7627773488215566e-06,
+      "loss": 0.0142,
+      "step": 2370
+    },
+    {
+      "epoch": 0.7523811884869971,
+      "grad_norm": 0.5493279099464417,
+      "learning_rate": 1.7585566789876624e-06,
+      "loss": 0.0158,
+      "step": 2371
+    },
+    {
+      "epoch": 0.752698515011032,
+      "grad_norm": 0.4315148890018463,
+      "learning_rate": 1.7543399896022406e-06,
+      "loss": 0.0152,
+      "step": 2372
+    },
+    {
+      "epoch": 0.7530158415350671,
+      "grad_norm": 0.5824035406112671,
+      "learning_rate": 1.7501272858433326e-06,
+      "loss": 0.0178,
+      "step": 2373
+    },
+    {
+      "epoch": 0.7533331680591021,
+      "grad_norm": 0.34759777784347534,
+      "learning_rate": 1.745918572884079e-06,
+      "loss": 0.0139,
+      "step": 2374
+    },
+    {
+      "epoch": 0.753650494583137,
+      "grad_norm": 0.3633348047733307,
+      "learning_rate": 1.7417138558927244e-06,
+      "loss": 0.0164,
+      "step": 2375
+    },
+    {
+      "epoch": 0.7539678211071721,
+      "grad_norm": 0.37112629413604736,
+      "learning_rate": 1.7375131400326035e-06,
+      "loss": 0.0155,
+      "step": 2376
+    },
+    {
+      "epoch": 0.7542851476312071,
+      "grad_norm": 0.3750361204147339,
+      "learning_rate": 1.7333164304621364e-06,
+      "loss": 0.0143,
+      "step": 2377
+    },
+    {
+      "epoch": 0.7546024741552421,
+      "grad_norm": 0.29348060488700867,
+      "learning_rate": 1.7291237323348287e-06,
+      "loss": 0.0144,
+      "step": 2378
+    },
+    {
+      "epoch": 0.7549198006792771,
+      "grad_norm": 0.2986072897911072,
+      "learning_rate": 1.724935050799254e-06,
+      "loss": 0.0119,
+      "step": 2379
+    },
+    {
+      "epoch": 0.755237127203312,
+      "grad_norm": 0.45698851346969604,
+      "learning_rate": 1.7207503909990596e-06,
+      "loss": 0.0174,
+      "step": 2380
+    },
+    {
+      "epoch": 0.7555544537273471,
+      "grad_norm": 0.48649463057518005,
+      "learning_rate": 1.716569758072948e-06,
+      "loss": 0.0145,
+      "step": 2381
+    },
+    {
+      "epoch": 0.7558717802513821,
+      "grad_norm": 0.48085546493530273,
+      "learning_rate": 1.7123931571546826e-06,
+      "loss": 0.0136,
+      "step": 2382
+    },
+    {
+      "epoch": 0.7561891067754171,
+      "grad_norm": 0.34362903237342834,
+      "learning_rate": 1.7082205933730694e-06,
+      "loss": 0.0133,
+      "step": 2383
+    },
+    {
+      "epoch": 0.7565064332994521,
+      "grad_norm": 0.37269139289855957,
+      "learning_rate": 1.7040520718519644e-06,
+      "loss": 0.014,
+      "step": 2384
+    },
+    {
+      "epoch": 0.7568237598234872,
+      "grad_norm": 0.3034665882587433,
+      "learning_rate": 1.6998875977102513e-06,
+      "loss": 0.0169,
+      "step": 2385
+    },
+    {
+      "epoch": 0.7571410863475221,
+      "grad_norm": 0.426495760679245,
+      "learning_rate": 1.695727176061851e-06,
+      "loss": 0.0179,
+      "step": 2386
+    },
+    {
+      "epoch": 0.7574584128715571,
+      "grad_norm": 0.39625290036201477,
+      "learning_rate": 1.6915708120157042e-06,
+      "loss": 0.0151,
+      "step": 2387
+    },
+    {
+      "epoch": 0.7577757393955922,
+      "grad_norm": 0.31691229343414307,
+      "learning_rate": 1.6874185106757673e-06,
+      "loss": 0.0145,
+      "step": 2388
+    },
+    {
+      "epoch": 0.7580930659196271,
+      "grad_norm": 0.3359290659427643,
+      "learning_rate": 1.6832702771410142e-06,
+      "loss": 0.0125,
+      "step": 2389
+    },
+    {
+      "epoch": 0.7584103924436622,
+      "grad_norm": 0.32391563057899475,
+      "learning_rate": 1.679126116505415e-06,
+      "loss": 0.0177,
+      "step": 2390
+    },
+    {
+      "epoch": 0.7587277189676972,
+      "grad_norm": 0.28574100136756897,
+      "learning_rate": 1.674986033857947e-06,
+      "loss": 0.0123,
+      "step": 2391
+    },
+    {
+      "epoch": 0.7590450454917321,
+      "grad_norm": 0.3066690266132355,
+      "learning_rate": 1.6708500342825719e-06,
+      "loss": 0.0122,
+      "step": 2392
+    },
+    {
+      "epoch": 0.7593623720157672,
+      "grad_norm": 0.5026608109474182,
+      "learning_rate": 1.666718122858244e-06,
+      "loss": 0.0139,
+      "step": 2393
+    },
+    {
+      "epoch": 0.7596796985398022,
+      "grad_norm": 0.5132337808609009,
+      "learning_rate": 1.6625903046588927e-06,
+      "loss": 0.0143,
+      "step": 2394
+    },
+    {
+      "epoch": 0.7599970250638372,
+      "grad_norm": 0.6558234095573425,
+      "learning_rate": 1.6584665847534227e-06,
+      "loss": 0.0144,
+      "step": 2395
+    },
+    {
+      "epoch": 0.7603143515878722,
+      "grad_norm": 0.336299329996109,
+      "learning_rate": 1.6543469682057105e-06,
+      "loss": 0.0126,
+      "step": 2396
+    },
+    {
+      "epoch": 0.7606316781119071,
+      "grad_norm": 0.27623477578163147,
+      "learning_rate": 1.6502314600745828e-06,
+      "loss": 0.0139,
+      "step": 2397
+    },
+    {
+      "epoch": 0.7609490046359422,
+      "grad_norm": 0.2943379878997803,
+      "learning_rate": 1.646120065413832e-06,
+      "loss": 0.0141,
+      "step": 2398
+    },
+    {
+      "epoch": 0.7612663311599772,
+      "grad_norm": 0.34159597754478455,
+      "learning_rate": 1.6420127892721926e-06,
+      "loss": 0.0155,
+      "step": 2399
+    },
+    {
+      "epoch": 0.7615836576840121,
+      "grad_norm": 0.4143332839012146,
+      "learning_rate": 1.6379096366933473e-06,
+      "loss": 0.0142,
+      "step": 2400
+    },
+    {
+      "epoch": 0.7619009842080472,
+      "grad_norm": 0.2862209379673004,
+      "learning_rate": 1.633810612715908e-06,
+      "loss": 0.0127,
+      "step": 2401
+    },
+    {
+      "epoch": 0.7622183107320822,
+      "grad_norm": 0.4077364504337311,
+      "learning_rate": 1.6297157223734228e-06,
+      "loss": 0.0161,
+      "step": 2402
+    },
+    {
+      "epoch": 0.7625356372561172,
+      "grad_norm": 0.31948786973953247,
+      "learning_rate": 1.6256249706943628e-06,
+      "loss": 0.0155,
+      "step": 2403
+    },
+    {
+      "epoch": 0.7628529637801522,
+      "grad_norm": 0.30778801441192627,
+      "learning_rate": 1.6215383627021125e-06,
+      "loss": 0.0143,
+      "step": 2404
+    },
+    {
+      "epoch": 0.7631702903041873,
+      "grad_norm": 0.46970343589782715,
+      "learning_rate": 1.617455903414974e-06,
+      "loss": 0.0148,
+      "step": 2405
+    },
+    {
+      "epoch": 0.7634876168282222,
+      "grad_norm": 0.3304109275341034,
+      "learning_rate": 1.6133775978461514e-06,
+      "loss": 0.014,
+      "step": 2406
+    },
+    {
+      "epoch": 0.7638049433522572,
+      "grad_norm": 0.31018170714378357,
+      "learning_rate": 1.6093034510037476e-06,
+      "loss": 0.0122,
+      "step": 2407
+    },
+    {
+      "epoch": 0.7641222698762923,
+      "grad_norm": 0.27229562401771545,
+      "learning_rate": 1.6052334678907583e-06,
+      "loss": 0.015,
+      "step": 2408
+    },
+    {
+      "epoch": 0.7644395964003272,
+      "grad_norm": 0.34459733963012695,
+      "learning_rate": 1.6011676535050712e-06,
+      "loss": 0.0152,
+      "step": 2409
+    },
+    {
+      "epoch": 0.7647569229243623,
+      "grad_norm": 0.37441369891166687,
+      "learning_rate": 1.5971060128394483e-06,
+      "loss": 0.013,
+      "step": 2410
+    },
+    {
+      "epoch": 0.7650742494483973,
+      "grad_norm": 0.32356294989585876,
+      "learning_rate": 1.5930485508815302e-06,
+      "loss": 0.0132,
+      "step": 2411
+    },
+    {
+      "epoch": 0.7653915759724322,
+      "grad_norm": 0.406248539686203,
+      "learning_rate": 1.5889952726138274e-06,
+      "loss": 0.0165,
+      "step": 2412
+    },
+    {
+      "epoch": 0.7657089024964673,
+      "grad_norm": 0.35921716690063477,
+      "learning_rate": 1.5849461830137069e-06,
+      "loss": 0.0162,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7660262290205022,
+      "grad_norm": 0.3182511627674103,
+      "learning_rate": 1.5809012870533996e-06,
+      "loss": 0.0149,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7663435555445373,
+      "grad_norm": 0.33041903376579285,
+      "learning_rate": 1.5768605896999795e-06,
+      "loss": 0.0136,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7666608820685723,
+      "grad_norm": 0.3028944134712219,
+      "learning_rate": 1.5728240959153729e-06,
+      "loss": 0.0138,
+      "step": 2416
+    },
+    {
+      "epoch": 0.7669782085926072,
+      "grad_norm": 0.43638190627098083,
+      "learning_rate": 1.5687918106563326e-06,
+      "loss": 0.0162,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7672955351166423,
+      "grad_norm": 0.4108610451221466,
+      "learning_rate": 1.5647637388744546e-06,
+      "loss": 0.015,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7676128616406773,
+      "grad_norm": 0.37404143810272217,
+      "learning_rate": 1.5607398855161576e-06,
+      "loss": 0.0155,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7679301881647123,
+      "grad_norm": 0.4005305767059326,
+      "learning_rate": 1.5567202555226756e-06,
+      "loss": 0.0185,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7682475146887473,
+      "grad_norm": 0.30017367005348206,
+      "learning_rate": 1.5527048538300632e-06,
+      "loss": 0.0132,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7685648412127823,
+      "grad_norm": 0.3182089626789093,
+      "learning_rate": 1.5486936853691759e-06,
+      "loss": 0.0132,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7688821677368173,
+      "grad_norm": 0.295445054769516,
+      "learning_rate": 1.544686755065677e-06,
+      "loss": 0.0151,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7691994942608523,
+      "grad_norm": 0.32100117206573486,
+      "learning_rate": 1.5406840678400204e-06,
+      "loss": 0.0165,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7695168207848874,
+      "grad_norm": 0.30235016345977783,
+      "learning_rate": 1.536685628607454e-06,
+      "loss": 0.0141,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7698341473089223,
+      "grad_norm": 0.3539015054702759,
+      "learning_rate": 1.532691442278006e-06,
+      "loss": 0.0127,
+      "step": 2426
+    },
+    {
+      "epoch": 0.7701514738329573,
+      "grad_norm": 0.2801800072193146,
+      "learning_rate": 1.5287015137564815e-06,
+      "loss": 0.012,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7704688003569924,
+      "grad_norm": 0.2806015610694885,
+      "learning_rate": 1.5247158479424611e-06,
+      "loss": 0.0129,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7707861268810273,
+      "grad_norm": 0.330809623003006,
+      "learning_rate": 1.5207344497302862e-06,
+      "loss": 0.0143,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7711034534050624,
+      "grad_norm": 0.2515932619571686,
+      "learning_rate": 1.5167573240090626e-06,
+      "loss": 0.0122,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7714207799290973,
+      "grad_norm": 0.30109652876853943,
+      "learning_rate": 1.5127844756626437e-06,
+      "loss": 0.0141,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7717381064531323,
+      "grad_norm": 0.27825748920440674,
+      "learning_rate": 1.5088159095696365e-06,
+      "loss": 0.0144,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7720554329771674,
+      "grad_norm": 0.3394031822681427,
+      "learning_rate": 1.5048516306033834e-06,
+      "loss": 0.0182,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7723727595012023,
+      "grad_norm": 0.30876481533050537,
+      "learning_rate": 1.500891643631968e-06,
+      "loss": 0.0127,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7726900860252374,
+      "grad_norm": 0.3291650712490082,
+      "learning_rate": 1.4969359535181977e-06,
+      "loss": 0.0108,
+      "step": 2435
+    },
+    {
+      "epoch": 0.7730074125492724,
+      "grad_norm": 0.32963359355926514,
+      "learning_rate": 1.4929845651196101e-06,
+      "loss": 0.0132,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7733247390733073,
+      "grad_norm": 0.4894530177116394,
+      "learning_rate": 1.489037483288454e-06,
+      "loss": 0.015,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7736420655973424,
+      "grad_norm": 0.3310536742210388,
+      "learning_rate": 1.4850947128716914e-06,
+      "loss": 0.0158,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7739593921213774,
+      "grad_norm": 0.39202845096588135,
+      "learning_rate": 1.4811562587109946e-06,
+      "loss": 0.0153,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7742767186454124,
+      "grad_norm": 0.34008553624153137,
+      "learning_rate": 1.4772221256427277e-06,
+      "loss": 0.0114,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7745940451694474,
+      "grad_norm": 0.3594018220901489,
+      "learning_rate": 1.4732923184979563e-06,
+      "loss": 0.0109,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7749113716934825,
+      "grad_norm": 0.2841017544269562,
+      "learning_rate": 1.4693668421024277e-06,
+      "loss": 0.0118,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7752286982175174,
+      "grad_norm": 0.30123937129974365,
+      "learning_rate": 1.4654457012765766e-06,
+      "loss": 0.0121,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7755460247415524,
+      "grad_norm": 0.3106209635734558,
+      "learning_rate": 1.4615289008355077e-06,
+      "loss": 0.0134,
+      "step": 2444
+    },
+    {
+      "epoch": 0.7758633512655875,
+      "grad_norm": 0.34512394666671753,
+      "learning_rate": 1.4576164455890014e-06,
+      "loss": 0.0139,
+      "step": 2445
+    },
+    {
+      "epoch": 0.7761806777896224,
+      "grad_norm": 0.3337534964084625,
+      "learning_rate": 1.4537083403414986e-06,
+      "loss": 0.013,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7764980043136575,
+      "grad_norm": 0.7742713093757629,
+      "learning_rate": 1.4498045898920988e-06,
+      "loss": 0.0117,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7768153308376924,
+      "grad_norm": 0.3857761323451996,
+      "learning_rate": 1.4459051990345563e-06,
+      "loss": 0.0158,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7771326573617274,
+      "grad_norm": 0.3136574327945709,
+      "learning_rate": 1.4420101725572682e-06,
+      "loss": 0.0147,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7774499838857625,
+      "grad_norm": 0.24596168100833893,
+      "learning_rate": 1.438119515243277e-06,
+      "loss": 0.0116,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7777673104097974,
+      "grad_norm": 0.3011131286621094,
+      "learning_rate": 1.434233231870254e-06,
+      "loss": 0.0158,
+      "step": 2451
+    },
+    {
+      "epoch": 0.7780846369338325,
+      "grad_norm": 0.3561844229698181,
+      "learning_rate": 1.4303513272105057e-06,
+      "loss": 0.0144,
+      "step": 2452
+    },
+    {
+      "epoch": 0.7784019634578675,
+      "grad_norm": 0.3505239188671112,
+      "learning_rate": 1.4264738060309551e-06,
+      "loss": 0.0165,
+      "step": 2453
+    },
+    {
+      "epoch": 0.7787192899819024,
+      "grad_norm": 0.3318690061569214,
+      "learning_rate": 1.422600673093149e-06,
+      "loss": 0.015,
+      "step": 2454
+    },
+    {
+      "epoch": 0.7790366165059375,
+      "grad_norm": 0.32508471608161926,
+      "learning_rate": 1.4187319331532395e-06,
+      "loss": 0.0136,
+      "step": 2455
+    },
+    {
+      "epoch": 0.7793539430299725,
+      "grad_norm": 0.34000325202941895,
+      "learning_rate": 1.414867590961989e-06,
+      "loss": 0.013,
+      "step": 2456
+    },
+    {
+      "epoch": 0.7796712695540075,
+      "grad_norm": 0.27049747109413147,
+      "learning_rate": 1.4110076512647563e-06,
+      "loss": 0.0123,
+      "step": 2457
+    },
+    {
+      "epoch": 0.7799885960780425,
+      "grad_norm": 0.31445547938346863,
+      "learning_rate": 1.4071521188014937e-06,
+      "loss": 0.0109,
+      "step": 2458
+    },
+    {
+      "epoch": 0.7803059226020775,
+      "grad_norm": 0.3164269030094147,
+      "learning_rate": 1.4033009983067454e-06,
+      "loss": 0.0155,
+      "step": 2459
+    },
+    {
+      "epoch": 0.7806232491261125,
+      "grad_norm": 0.30799898505210876,
+      "learning_rate": 1.3994542945096323e-06,
+      "loss": 0.0116,
+      "step": 2460
+    },
+    {
+      "epoch": 0.7809405756501475,
+      "grad_norm": 0.30171382427215576,
+      "learning_rate": 1.3956120121338579e-06,
+      "loss": 0.0134,
+      "step": 2461
+    },
+    {
+      "epoch": 0.7812579021741826,
+      "grad_norm": 0.42864635586738586,
+      "learning_rate": 1.3917741558976894e-06,
+      "loss": 0.0193,
+      "step": 2462
+    },
+    {
+      "epoch": 0.7815752286982175,
+      "grad_norm": 0.31200334429740906,
+      "learning_rate": 1.3879407305139659e-06,
+      "loss": 0.0143,
+      "step": 2463
+    },
+    {
+      "epoch": 0.7818925552222525,
+      "grad_norm": 0.6565803289413452,
+      "learning_rate": 1.3841117406900778e-06,
+      "loss": 0.0164,
+      "step": 2464
+    },
+    {
+      "epoch": 0.7822098817462876,
+      "grad_norm": 0.4215078055858612,
+      "learning_rate": 1.3802871911279752e-06,
+      "loss": 0.0151,
+      "step": 2465
+    },
+    {
+      "epoch": 0.7825272082703225,
+      "grad_norm": 0.3392726182937622,
+      "learning_rate": 1.3764670865241557e-06,
+      "loss": 0.0146,
+      "step": 2466
+    },
+    {
+      "epoch": 0.7828445347943576,
+      "grad_norm": 0.28017330169677734,
+      "learning_rate": 1.3726514315696505e-06,
+      "loss": 0.0113,
+      "step": 2467
+    },
+    {
+      "epoch": 0.7831618613183925,
+      "grad_norm": 0.34483104944229126,
+      "learning_rate": 1.3688402309500353e-06,
+      "loss": 0.0132,
+      "step": 2468
+    },
+    {
+      "epoch": 0.7834791878424275,
+      "grad_norm": 0.2651517987251282,
+      "learning_rate": 1.3650334893454115e-06,
+      "loss": 0.0097,
+      "step": 2469
+    },
+    {
+      "epoch": 0.7837965143664626,
+      "grad_norm": 0.30205124616622925,
+      "learning_rate": 1.3612312114304082e-06,
+      "loss": 0.0121,
+      "step": 2470
+    },
+    {
+      "epoch": 0.7841138408904975,
+      "grad_norm": 0.6148396134376526,
+      "learning_rate": 1.3574334018741685e-06,
+      "loss": 0.012,
+      "step": 2471
+    },
+    {
+      "epoch": 0.7844311674145326,
+      "grad_norm": 0.7151223421096802,
+      "learning_rate": 1.3536400653403537e-06,
+      "loss": 0.0169,
+      "step": 2472
+    },
+    {
+      "epoch": 0.7847484939385676,
+      "grad_norm": 0.2600792646408081,
+      "learning_rate": 1.3498512064871272e-06,
+      "loss": 0.0124,
+      "step": 2473
+    },
+    {
+      "epoch": 0.7850658204626025,
+      "grad_norm": 0.3672007620334625,
+      "learning_rate": 1.3460668299671576e-06,
+      "loss": 0.0169,
+      "step": 2474
+    },
+    {
+      "epoch": 0.7853831469866376,
+      "grad_norm": 0.3608895540237427,
+      "learning_rate": 1.3422869404276094e-06,
+      "loss": 0.0138,
+      "step": 2475
+    },
+    {
+      "epoch": 0.7857004735106726,
+      "grad_norm": 0.3485403060913086,
+      "learning_rate": 1.3385115425101342e-06,
+      "loss": 0.0147,
+      "step": 2476
+    },
+    {
+      "epoch": 0.7860178000347076,
+      "grad_norm": 0.34955379366874695,
+      "learning_rate": 1.3347406408508695e-06,
+      "loss": 0.015,
+      "step": 2477
+    },
+    {
+      "epoch": 0.7863351265587426,
+      "grad_norm": 0.4677436053752899,
+      "learning_rate": 1.3309742400804304e-06,
+      "loss": 0.0161,
+      "step": 2478
+    },
+    {
+      "epoch": 0.7866524530827776,
+      "grad_norm": 0.2825945019721985,
+      "learning_rate": 1.3272123448239082e-06,
+      "loss": 0.0132,
+      "step": 2479
+    },
+    {
+      "epoch": 0.7869697796068126,
+      "grad_norm": 0.31685951352119446,
+      "learning_rate": 1.3234549597008572e-06,
+      "loss": 0.0151,
+      "step": 2480
+    },
+    {
+      "epoch": 0.7872871061308476,
+      "grad_norm": 0.6552386283874512,
+      "learning_rate": 1.3197020893252965e-06,
+      "loss": 0.0139,
+      "step": 2481
+    },
+    {
+      "epoch": 0.7876044326548827,
+      "grad_norm": 0.39144212007522583,
+      "learning_rate": 1.3159537383057008e-06,
+      "loss": 0.0146,
+      "step": 2482
+    },
+    {
+      "epoch": 0.7879217591789176,
+      "grad_norm": 0.32506516575813293,
+      "learning_rate": 1.3122099112449926e-06,
+      "loss": 0.0147,
+      "step": 2483
+    },
+    {
+      "epoch": 0.7882390857029526,
+      "grad_norm": 0.527711808681488,
+      "learning_rate": 1.3084706127405433e-06,
+      "loss": 0.0153,
+      "step": 2484
+    },
+    {
+      "epoch": 0.7885564122269876,
+      "grad_norm": 0.35382941365242004,
+      "learning_rate": 1.3047358473841575e-06,
+      "loss": 0.0137,
+      "step": 2485
+    },
+    {
+      "epoch": 0.7888737387510226,
+      "grad_norm": 0.6226952075958252,
+      "learning_rate": 1.3010056197620813e-06,
+      "loss": 0.0155,
+      "step": 2486
+    },
+    {
+      "epoch": 0.7891910652750577,
+      "grad_norm": 0.3379950225353241,
+      "learning_rate": 1.297279934454978e-06,
+      "loss": 0.0172,
+      "step": 2487
+    },
+    {
+      "epoch": 0.7895083917990926,
+      "grad_norm": 0.2975340485572815,
+      "learning_rate": 1.2935587960379424e-06,
+      "loss": 0.0145,
+      "step": 2488
+    },
+    {
+      "epoch": 0.7898257183231276,
+      "grad_norm": 0.3426641821861267,
+      "learning_rate": 1.2898422090804792e-06,
+      "loss": 0.0153,
+      "step": 2489
+    },
+    {
+      "epoch": 0.7901430448471627,
+      "grad_norm": 0.2762376666069031,
+      "learning_rate": 1.286130178146509e-06,
+      "loss": 0.0121,
+      "step": 2490
+    },
+    {
+      "epoch": 0.7904603713711976,
+      "grad_norm": 0.4973738193511963,
+      "learning_rate": 1.2824227077943563e-06,
+      "loss": 0.0135,
+      "step": 2491
+    },
+    {
+      "epoch": 0.7907776978952327,
+      "grad_norm": 0.37457481026649475,
+      "learning_rate": 1.2787198025767417e-06,
+      "loss": 0.0124,
+      "step": 2492
+    },
+    {
+      "epoch": 0.7910950244192677,
+      "grad_norm": 0.33032846450805664,
+      "learning_rate": 1.2750214670407856e-06,
+      "loss": 0.0146,
+      "step": 2493
+    },
+    {
+      "epoch": 0.7914123509433026,
+      "grad_norm": 0.4344386160373688,
+      "learning_rate": 1.271327705727991e-06,
+      "loss": 0.0127,
+      "step": 2494
+    },
+    {
+      "epoch": 0.7917296774673377,
+      "grad_norm": 0.33561450242996216,
+      "learning_rate": 1.2676385231742493e-06,
+      "loss": 0.0143,
+      "step": 2495
+    },
+    {
+      "epoch": 0.7920470039913727,
+      "grad_norm": 0.2640942335128784,
+      "learning_rate": 1.2639539239098258e-06,
+      "loss": 0.0136,
+      "step": 2496
+    },
+    {
+      "epoch": 0.7923643305154077,
+      "grad_norm": 0.29306045174598694,
+      "learning_rate": 1.2602739124593572e-06,
+      "loss": 0.0134,
+      "step": 2497
+    },
+    {
+      "epoch": 0.7926816570394427,
+      "grad_norm": 0.31503745913505554,
+      "learning_rate": 1.2565984933418495e-06,
+      "loss": 0.0125,
+      "step": 2498
+    },
+    {
+      "epoch": 0.7929989835634778,
+      "grad_norm": 0.3651042878627777,
+      "learning_rate": 1.2529276710706656e-06,
+      "loss": 0.0179,
+      "step": 2499
+    },
+    {
+      "epoch": 0.7933163100875127,
+      "grad_norm": 0.2740778625011444,
+      "learning_rate": 1.2492614501535283e-06,
+      "loss": 0.0131,
+      "step": 2500
+    },
+    {
+      "epoch": 0.7936336366115477,
+      "grad_norm": 0.33348000049591064,
+      "learning_rate": 1.2455998350925042e-06,
+      "loss": 0.0138,
+      "step": 2501
+    },
+    {
+      "epoch": 0.7939509631355827,
+      "grad_norm": 0.27049335837364197,
+      "learning_rate": 1.2419428303840092e-06,
+      "loss": 0.0149,
+      "step": 2502
+    },
+    {
+      "epoch": 0.7942682896596177,
+      "grad_norm": 0.34368011355400085,
+      "learning_rate": 1.238290440518794e-06,
+      "loss": 0.013,
+      "step": 2503
+    },
+    {
+      "epoch": 0.7945856161836528,
+      "grad_norm": 0.3631610572338104,
+      "learning_rate": 1.234642669981946e-06,
+      "loss": 0.0164,
+      "step": 2504
+    },
+    {
+      "epoch": 0.7949029427076877,
+      "grad_norm": 0.4950118660926819,
+      "learning_rate": 1.230999523252875e-06,
+      "loss": 0.0147,
+      "step": 2505
+    },
+    {
+      "epoch": 0.7952202692317227,
+      "grad_norm": 0.3908025026321411,
+      "learning_rate": 1.2273610048053186e-06,
+      "loss": 0.0164,
+      "step": 2506
+    },
+    {
+      "epoch": 0.7955375957557578,
+      "grad_norm": 0.6258262395858765,
+      "learning_rate": 1.223727119107327e-06,
+      "loss": 0.016,
+      "step": 2507
+    },
+    {
+      "epoch": 0.7958549222797927,
+      "grad_norm": 0.3070192337036133,
+      "learning_rate": 1.2200978706212606e-06,
+      "loss": 0.0107,
+      "step": 2508
+    },
+    {
+      "epoch": 0.7961722488038278,
+      "grad_norm": 0.4016229808330536,
+      "learning_rate": 1.2164732638037917e-06,
+      "loss": 0.0155,
+      "step": 2509
+    },
+    {
+      "epoch": 0.7964895753278628,
+      "grad_norm": 0.3548549711704254,
+      "learning_rate": 1.2128533031058838e-06,
+      "loss": 0.0152,
+      "step": 2510
+    },
+    {
+      "epoch": 0.7968069018518977,
+      "grad_norm": 0.2922239601612091,
+      "learning_rate": 1.2092379929728038e-06,
+      "loss": 0.0128,
+      "step": 2511
+    },
+    {
+      "epoch": 0.7971242283759328,
+      "grad_norm": 0.4365336000919342,
+      "learning_rate": 1.2056273378440998e-06,
+      "loss": 0.0176,
+      "step": 2512
+    },
+    {
+      "epoch": 0.7974415548999678,
+      "grad_norm": 0.33986228704452515,
+      "learning_rate": 1.2020213421536103e-06,
+      "loss": 0.013,
+      "step": 2513
+    },
+    {
+      "epoch": 0.7977588814240028,
+      "grad_norm": 0.39563462138175964,
+      "learning_rate": 1.1984200103294463e-06,
+      "loss": 0.0148,
+      "step": 2514
+    },
+    {
+      "epoch": 0.7980762079480378,
+      "grad_norm": 0.463929682970047,
+      "learning_rate": 1.1948233467939978e-06,
+      "loss": 0.0164,
+      "step": 2515
+    },
+    {
+      "epoch": 0.7983935344720728,
+      "grad_norm": 0.38284891843795776,
+      "learning_rate": 1.1912313559639172e-06,
+      "loss": 0.0146,
+      "step": 2516
+    },
+    {
+      "epoch": 0.7987108609961078,
+      "grad_norm": 0.4618520736694336,
+      "learning_rate": 1.187644042250119e-06,
+      "loss": 0.0143,
+      "step": 2517
+    },
+    {
+      "epoch": 0.7990281875201428,
+      "grad_norm": 2.4488625526428223,
+      "learning_rate": 1.1840614100577785e-06,
+      "loss": 0.0133,
+      "step": 2518
+    },
+    {
+      "epoch": 0.7993455140441778,
+      "grad_norm": 0.30640709400177,
+      "learning_rate": 1.1804834637863176e-06,
+      "loss": 0.0119,
+      "step": 2519
+    },
+    {
+      "epoch": 0.7996628405682128,
+      "grad_norm": 0.6243758797645569,
+      "learning_rate": 1.1769102078294077e-06,
+      "loss": 0.0123,
+      "step": 2520
+    },
+    {
+      "epoch": 0.7999801670922478,
+      "grad_norm": 0.3778286576271057,
+      "learning_rate": 1.173341646574956e-06,
+      "loss": 0.0134,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8002974936162828,
+      "grad_norm": 0.3506750166416168,
+      "learning_rate": 1.1697777844051105e-06,
+      "loss": 0.0142,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8006148201403178,
+      "grad_norm": 0.3806482255458832,
+      "learning_rate": 1.1662186256962427e-06,
+      "loss": 0.0134,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8009321466643529,
+      "grad_norm": 0.26420775055885315,
+      "learning_rate": 1.1626641748189544e-06,
+      "loss": 0.0101,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8012494731883878,
+      "grad_norm": 0.39315545558929443,
+      "learning_rate": 1.159114436138059e-06,
+      "loss": 0.0147,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8015667997124228,
+      "grad_norm": 0.32583558559417725,
+      "learning_rate": 1.1555694140125922e-06,
+      "loss": 0.0142,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8018841262364579,
+      "grad_norm": 0.3202246129512787,
+      "learning_rate": 1.1520291127957905e-06,
+      "loss": 0.0155,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8022014527604928,
+      "grad_norm": 0.29136598110198975,
+      "learning_rate": 1.1484935368350946e-06,
+      "loss": 0.013,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8025187792845279,
+      "grad_norm": 0.3037831485271454,
+      "learning_rate": 1.1449626904721472e-06,
+      "loss": 0.0116,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8028361058085629,
+      "grad_norm": 0.3002447783946991,
+      "learning_rate": 1.1414365780427761e-06,
+      "loss": 0.0123,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8031534323325978,
+      "grad_norm": 0.3085499107837677,
+      "learning_rate": 1.137915203877003e-06,
+      "loss": 0.0123,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8034707588566329,
+      "grad_norm": 0.44119712710380554,
+      "learning_rate": 1.1343985722990253e-06,
+      "loss": 0.0144,
+      "step": 2532
+    },
+    {
+      "epoch": 0.8037880853806679,
+      "grad_norm": 0.3067564368247986,
+      "learning_rate": 1.1308866876272206e-06,
+      "loss": 0.0136,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8041054119047029,
+      "grad_norm": 0.3368540406227112,
+      "learning_rate": 1.1273795541741334e-06,
+      "loss": 0.0149,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8044227384287379,
+      "grad_norm": 0.2975098490715027,
+      "learning_rate": 1.123877176246479e-06,
+      "loss": 0.0133,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8047400649527728,
+      "grad_norm": 0.37819764018058777,
+      "learning_rate": 1.1203795581451288e-06,
+      "loss": 0.0138,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8050573914768079,
+      "grad_norm": 0.39585766196250916,
+      "learning_rate": 1.1168867041651082e-06,
+      "loss": 0.0131,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8053747180008429,
+      "grad_norm": 0.33064329624176025,
+      "learning_rate": 1.1133986185955987e-06,
+      "loss": 0.0165,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8056920445248779,
+      "grad_norm": 0.30474236607551575,
+      "learning_rate": 1.109915305719918e-06,
+      "loss": 0.0126,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8060093710489129,
+      "grad_norm": 0.32977423071861267,
+      "learning_rate": 1.1064367698155303e-06,
+      "loss": 0.0124,
+      "step": 2540
+    },
+    {
+      "epoch": 0.806326697572948,
+      "grad_norm": 0.354495644569397,
+      "learning_rate": 1.1029630151540288e-06,
+      "loss": 0.012,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8066440240969829,
+      "grad_norm": 0.32604384422302246,
+      "learning_rate": 1.0994940460011382e-06,
+      "loss": 0.0126,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8069613506210179,
+      "grad_norm": 0.2772787809371948,
+      "learning_rate": 1.096029866616704e-06,
+      "loss": 0.0135,
+      "step": 2543
+    },
+    {
+      "epoch": 0.807278677145053,
+      "grad_norm": 0.48569127917289734,
+      "learning_rate": 1.0925704812546927e-06,
+      "loss": 0.0141,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8075960036690879,
+      "grad_norm": 0.2899952530860901,
+      "learning_rate": 1.0891158941631836e-06,
+      "loss": 0.0124,
+      "step": 2545
+    },
+    {
+      "epoch": 0.807913330193123,
+      "grad_norm": 0.5140385627746582,
+      "learning_rate": 1.0856661095843617e-06,
+      "loss": 0.015,
+      "step": 2546
+    },
+    {
+      "epoch": 0.808230656717158,
+      "grad_norm": 0.28233802318573,
+      "learning_rate": 1.0822211317545167e-06,
+      "loss": 0.01,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8085479832411929,
+      "grad_norm": 0.3965209126472473,
+      "learning_rate": 1.0787809649040314e-06,
+      "loss": 0.0146,
+      "step": 2548
+    },
+    {
+      "epoch": 0.808865309765228,
+      "grad_norm": 0.3478708267211914,
+      "learning_rate": 1.0753456132573886e-06,
+      "loss": 0.0134,
+      "step": 2549
+    },
+    {
+      "epoch": 0.809182636289263,
+      "grad_norm": 0.3578258156776428,
+      "learning_rate": 1.0719150810331497e-06,
+      "loss": 0.0138,
+      "step": 2550
+    },
+    {
+      "epoch": 0.809499962813298,
+      "grad_norm": 0.2746797800064087,
+      "learning_rate": 1.0684893724439655e-06,
+      "loss": 0.011,
+      "step": 2551
+    },
+    {
+      "epoch": 0.809817289337333,
+      "grad_norm": 0.37216416001319885,
+      "learning_rate": 1.065068491696556e-06,
+      "loss": 0.0151,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8101346158613679,
+      "grad_norm": 0.3191049098968506,
+      "learning_rate": 1.061652442991718e-06,
+      "loss": 0.0121,
+      "step": 2553
+    },
+    {
+      "epoch": 0.810451942385403,
+      "grad_norm": 0.3227132260799408,
+      "learning_rate": 1.0582412305243144e-06,
+      "loss": 0.012,
+      "step": 2554
+    },
+    {
+      "epoch": 0.810769268909438,
+      "grad_norm": 0.3264809250831604,
+      "learning_rate": 1.0548348584832656e-06,
+      "loss": 0.0179,
+      "step": 2555
+    },
+    {
+      "epoch": 0.811086595433473,
+      "grad_norm": 0.3146434426307678,
+      "learning_rate": 1.051433331051554e-06,
+      "loss": 0.0156,
+      "step": 2556
+    },
+    {
+      "epoch": 0.811403921957508,
+      "grad_norm": 0.3264912962913513,
+      "learning_rate": 1.0480366524062041e-06,
+      "loss": 0.0142,
+      "step": 2557
+    },
+    {
+      "epoch": 0.811721248481543,
+      "grad_norm": 0.278200626373291,
+      "learning_rate": 1.044644826718295e-06,
+      "loss": 0.0143,
+      "step": 2558
+    },
+    {
+      "epoch": 0.812038575005578,
+      "grad_norm": 0.3418426513671875,
+      "learning_rate": 1.0412578581529404e-06,
+      "loss": 0.0148,
+      "step": 2559
+    },
+    {
+      "epoch": 0.812355901529613,
+      "grad_norm": 0.25862735509872437,
+      "learning_rate": 1.037875750869292e-06,
+      "loss": 0.013,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8126732280536481,
+      "grad_norm": 0.30459871888160706,
+      "learning_rate": 1.0344985090205344e-06,
+      "loss": 0.0153,
+      "step": 2561
+    },
+    {
+      "epoch": 0.812990554577683,
+      "grad_norm": 0.32274380326271057,
+      "learning_rate": 1.0311261367538717e-06,
+      "loss": 0.0137,
+      "step": 2562
+    },
+    {
+      "epoch": 0.813307881101718,
+      "grad_norm": 0.36233755946159363,
+      "learning_rate": 1.0277586382105337e-06,
+      "loss": 0.0145,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8136252076257531,
+      "grad_norm": 0.3786611557006836,
+      "learning_rate": 1.0243960175257605e-06,
+      "loss": 0.0144,
+      "step": 2564
+    },
+    {
+      "epoch": 0.813942534149788,
+      "grad_norm": 0.3741892874240875,
+      "learning_rate": 1.0210382788288082e-06,
+      "loss": 0.0137,
+      "step": 2565
+    },
+    {
+      "epoch": 0.814259860673823,
+      "grad_norm": 0.25408482551574707,
+      "learning_rate": 1.0176854262429315e-06,
+      "loss": 0.0107,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8145771871978581,
+      "grad_norm": 0.3166499733924866,
+      "learning_rate": 1.0143374638853892e-06,
+      "loss": 0.014,
+      "step": 2567
+    },
+    {
+      "epoch": 0.814894513721893,
+      "grad_norm": 0.2933066189289093,
+      "learning_rate": 1.0109943958674317e-06,
+      "loss": 0.0131,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8152118402459281,
+      "grad_norm": 0.5521876811981201,
+      "learning_rate": 1.007656226294303e-06,
+      "loss": 0.0153,
+      "step": 2569
+    },
+    {
+      "epoch": 0.815529166769963,
+      "grad_norm": 0.31985822319984436,
+      "learning_rate": 1.004322959265231e-06,
+      "loss": 0.0129,
+      "step": 2570
+    },
+    {
+      "epoch": 0.815846493293998,
+      "grad_norm": 0.3906991183757782,
+      "learning_rate": 1.0009945988734205e-06,
+      "loss": 0.0139,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8161638198180331,
+      "grad_norm": 0.34185919165611267,
+      "learning_rate": 9.976711492060548e-07,
+      "loss": 0.0162,
+      "step": 2572
+    },
+    {
+      "epoch": 0.816481146342068,
+      "grad_norm": 0.3332889974117279,
+      "learning_rate": 9.943526143442827e-07,
+      "loss": 0.0176,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8167984728661031,
+      "grad_norm": 0.30006012320518494,
+      "learning_rate": 9.91038998363223e-07,
+      "loss": 0.0129,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8171157993901381,
+      "grad_norm": 0.2658500671386719,
+      "learning_rate": 9.877303053319482e-07,
+      "loss": 0.0148,
+      "step": 2575
+    },
+    {
+      "epoch": 0.817433125914173,
+      "grad_norm": 0.4013974368572235,
+      "learning_rate": 9.844265393134927e-07,
+      "loss": 0.0144,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8177504524382081,
+      "grad_norm": 0.31424617767333984,
+      "learning_rate": 9.81127704364831e-07,
+      "loss": 0.0141,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8180677789622431,
+      "grad_norm": 0.2701103091239929,
+      "learning_rate": 9.778338045368901e-07,
+      "loss": 0.0131,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8183851054862781,
+      "grad_norm": 0.289326548576355,
+      "learning_rate": 9.745448438745358e-07,
+      "loss": 0.0115,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8187024320103131,
+      "grad_norm": 0.351785272359848,
+      "learning_rate": 9.712608264165652e-07,
+      "loss": 0.0146,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8190197585343482,
+      "grad_norm": 0.2794457972049713,
+      "learning_rate": 9.679817561957095e-07,
+      "loss": 0.0146,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8193370850583831,
+      "grad_norm": 0.33724579215049744,
+      "learning_rate": 9.647076372386195e-07,
+      "loss": 0.0132,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8196544115824181,
+      "grad_norm": 0.356671541929245,
+      "learning_rate": 9.614384735658721e-07,
+      "loss": 0.0122,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8199717381064532,
+      "grad_norm": 0.47444474697113037,
+      "learning_rate": 9.581742691919526e-07,
+      "loss": 0.0135,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8202890646304881,
+      "grad_norm": 0.5099074840545654,
+      "learning_rate": 9.549150281252633e-07,
+      "loss": 0.0145,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8206063911545232,
+      "grad_norm": 0.3875661790370941,
+      "learning_rate": 9.51660754368105e-07,
+      "loss": 0.0145,
+      "step": 2586
+    },
+    {
+      "epoch": 0.8209237176785581,
+      "grad_norm": 0.27304697036743164,
+      "learning_rate": 9.484114519166838e-07,
+      "loss": 0.0145,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8212410442025931,
+      "grad_norm": 0.3219338655471802,
+      "learning_rate": 9.451671247610988e-07,
+      "loss": 0.0161,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8215583707266282,
+      "grad_norm": 0.39146003127098083,
+      "learning_rate": 9.41927776885338e-07,
+      "loss": 0.0142,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8218756972506631,
+      "grad_norm": 0.30209556221961975,
+      "learning_rate": 9.386934122672797e-07,
+      "loss": 0.0148,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8221930237746982,
+      "grad_norm": 0.4056265950202942,
+      "learning_rate": 9.354640348786765e-07,
+      "loss": 0.0152,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8225103502987332,
+      "grad_norm": 0.27461153268814087,
+      "learning_rate": 9.322396486851626e-07,
+      "loss": 0.0108,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8228276768227681,
+      "grad_norm": 0.28821390867233276,
+      "learning_rate": 9.290202576462393e-07,
+      "loss": 0.0112,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8231450033468032,
+      "grad_norm": 0.47933849692344666,
+      "learning_rate": 9.258058657152763e-07,
+      "loss": 0.0113,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8234623298708382,
+      "grad_norm": 0.48877254128456116,
+      "learning_rate": 9.225964768395012e-07,
+      "loss": 0.0122,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8237796563948732,
+      "grad_norm": 0.30717021226882935,
+      "learning_rate": 9.193920949600027e-07,
+      "loss": 0.0137,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8240969829189082,
+      "grad_norm": 0.3125111162662506,
+      "learning_rate": 9.161927240117174e-07,
+      "loss": 0.0152,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8244143094429432,
+      "grad_norm": 0.4730572998523712,
+      "learning_rate": 9.129983679234278e-07,
+      "loss": 0.0148,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8247316359669782,
+      "grad_norm": 0.304155170917511,
+      "learning_rate": 9.098090306177626e-07,
+      "loss": 0.013,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8250489624910132,
+      "grad_norm": 0.24590085446834564,
+      "learning_rate": 9.066247160111818e-07,
+      "loss": 0.0089,
+      "step": 2600
+    },
+    {
+      "epoch": 0.8253662890150483,
+      "grad_norm": 0.3371686041355133,
+      "learning_rate": 9.034454280139837e-07,
+      "loss": 0.0156,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8256836155390832,
+      "grad_norm": 0.4111330509185791,
+      "learning_rate": 9.002711705302886e-07,
+      "loss": 0.0152,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8260009420631182,
+      "grad_norm": 0.2706948518753052,
+      "learning_rate": 8.971019474580428e-07,
+      "loss": 0.0111,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8263182685871532,
+      "grad_norm": 0.30645784735679626,
+      "learning_rate": 8.939377626890084e-07,
+      "loss": 0.0112,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8266355951111882,
+      "grad_norm": 0.3036125600337982,
+      "learning_rate": 8.907786201087631e-07,
+      "loss": 0.0141,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8269529216352233,
+      "grad_norm": 0.38006728887557983,
+      "learning_rate": 8.876245235966884e-07,
+      "loss": 0.0124,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8272702481592582,
+      "grad_norm": 0.27994054555892944,
+      "learning_rate": 8.844754770259756e-07,
+      "loss": 0.0127,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8275875746832932,
+      "grad_norm": 0.3524767756462097,
+      "learning_rate": 8.813314842636089e-07,
+      "loss": 0.015,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8279049012073283,
+      "grad_norm": 0.332182377576828,
+      "learning_rate": 8.78192549170368e-07,
+      "loss": 0.0151,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8282222277313632,
+      "grad_norm": 0.28554603457450867,
+      "learning_rate": 8.750586756008256e-07,
+      "loss": 0.0157,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8285395542553983,
+      "grad_norm": 0.31627899408340454,
+      "learning_rate": 8.719298674033327e-07,
+      "loss": 0.0148,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8288568807794333,
+      "grad_norm": 0.2905786633491516,
+      "learning_rate": 8.688061284200266e-07,
+      "loss": 0.0113,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8291742073034682,
+      "grad_norm": 0.2594849765300751,
+      "learning_rate": 8.656874624868133e-07,
+      "loss": 0.011,
+      "step": 2613
+    },
+    {
+      "epoch": 0.8294915338275033,
+      "grad_norm": 0.29761698842048645,
+      "learning_rate": 8.62573873433375e-07,
+      "loss": 0.0124,
+      "step": 2614
+    },
+    {
+      "epoch": 0.8298088603515383,
+      "grad_norm": 0.482501357793808,
+      "learning_rate": 8.594653650831541e-07,
+      "loss": 0.0163,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8301261868755733,
+      "grad_norm": 0.43422311544418335,
+      "learning_rate": 8.563619412533592e-07,
+      "loss": 0.014,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8304435133996083,
+      "grad_norm": 0.28606298565864563,
+      "learning_rate": 8.532636057549537e-07,
+      "loss": 0.0124,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8307608399236434,
+      "grad_norm": 0.3103850483894348,
+      "learning_rate": 8.501703623926472e-07,
+      "loss": 0.0148,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8310781664476783,
+      "grad_norm": 0.37948864698410034,
+      "learning_rate": 8.47082214964905e-07,
+      "loss": 0.0166,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8313954929717133,
+      "grad_norm": 0.32329025864601135,
+      "learning_rate": 8.439991672639264e-07,
+      "loss": 0.0112,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8317128194957483,
+      "grad_norm": 0.3032236397266388,
+      "learning_rate": 8.409212230756564e-07,
+      "loss": 0.0151,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8320301460197833,
+      "grad_norm": 0.3033043146133423,
+      "learning_rate": 8.378483861797649e-07,
+      "loss": 0.0126,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8323474725438184,
+      "grad_norm": 0.3774949312210083,
+      "learning_rate": 8.347806603496556e-07,
+      "loss": 0.0146,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8326647990678533,
+      "grad_norm": 0.290289044380188,
+      "learning_rate": 8.317180493524557e-07,
+      "loss": 0.0141,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8329821255918883,
+      "grad_norm": 0.3350723683834076,
+      "learning_rate": 8.286605569490075e-07,
+      "loss": 0.014,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8332994521159234,
+      "grad_norm": 0.33123230934143066,
+      "learning_rate": 8.256081868938725e-07,
+      "loss": 0.0148,
+      "step": 2626
+    },
+    {
+      "epoch": 0.8336167786399583,
+      "grad_norm": 0.2772403657436371,
+      "learning_rate": 8.225609429353187e-07,
+      "loss": 0.0125,
+      "step": 2627
+    },
+    {
+      "epoch": 0.8339341051639934,
+      "grad_norm": 0.3809295892715454,
+      "learning_rate": 8.1951882881532e-07,
+      "loss": 0.013,
+      "step": 2628
+    },
+    {
+      "epoch": 0.8342514316880284,
+      "grad_norm": 0.3881082236766815,
+      "learning_rate": 8.164818482695508e-07,
+      "loss": 0.0152,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8345687582120633,
+      "grad_norm": 0.356620192527771,
+      "learning_rate": 8.134500050273841e-07,
+      "loss": 0.0131,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8348860847360984,
+      "grad_norm": 0.31949904561042786,
+      "learning_rate": 8.104233028118802e-07,
+      "loss": 0.013,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8352034112601334,
+      "grad_norm": 0.27666139602661133,
+      "learning_rate": 8.074017453397886e-07,
+      "loss": 0.0104,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8355207377841684,
+      "grad_norm": 0.4098544716835022,
+      "learning_rate": 8.043853363215437e-07,
+      "loss": 0.0158,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8358380643082034,
+      "grad_norm": 0.29557690024375916,
+      "learning_rate": 8.013740794612512e-07,
+      "loss": 0.0117,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8361553908322384,
+      "grad_norm": 0.3908742666244507,
+      "learning_rate": 7.983679784566967e-07,
+      "loss": 0.0151,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8364727173562734,
+      "grad_norm": 0.3445609509944916,
+      "learning_rate": 7.953670369993288e-07,
+      "loss": 0.0115,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8367900438803084,
+      "grad_norm": 0.3435474634170532,
+      "learning_rate": 7.923712587742672e-07,
+      "loss": 0.0143,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8371073704043434,
+      "grad_norm": 0.36400121450424194,
+      "learning_rate": 7.893806474602811e-07,
+      "loss": 0.0171,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8374246969283784,
+      "grad_norm": 0.33572742342948914,
+      "learning_rate": 7.863952067298042e-07,
+      "loss": 0.0134,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8377420234524134,
+      "grad_norm": 0.3048573136329651,
+      "learning_rate": 7.834149402489178e-07,
+      "loss": 0.01,
+      "step": 2640
+    },
+    {
+      "epoch": 0.8380593499764484,
+      "grad_norm": 0.5229014158248901,
+      "learning_rate": 7.804398516773465e-07,
+      "loss": 0.0116,
+      "step": 2641
+    },
+    {
+      "epoch": 0.8383766765004834,
+      "grad_norm": 0.3187839686870575,
+      "learning_rate": 7.774699446684608e-07,
+      "loss": 0.0144,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8386940030245185,
+      "grad_norm": 0.2594151794910431,
+      "learning_rate": 7.745052228692645e-07,
+      "loss": 0.0114,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8390113295485534,
+      "grad_norm": 0.32305243611335754,
+      "learning_rate": 7.715456899203983e-07,
+      "loss": 0.0144,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8393286560725884,
+      "grad_norm": 0.2983658015727997,
+      "learning_rate": 7.685913494561253e-07,
+      "loss": 0.0136,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8396459825966235,
+      "grad_norm": 0.31884413957595825,
+      "learning_rate": 7.656422051043383e-07,
+      "loss": 0.0147,
+      "step": 2646
+    },
+    {
+      "epoch": 0.8399633091206584,
+      "grad_norm": 0.24752821028232574,
+      "learning_rate": 7.626982604865457e-07,
+      "loss": 0.0112,
+      "step": 2647
+    },
+    {
+      "epoch": 0.8402806356446935,
+      "grad_norm": 0.3777827024459839,
+      "learning_rate": 7.597595192178702e-07,
+      "loss": 0.0116,
+      "step": 2648
+    },
+    {
+      "epoch": 0.8405979621687285,
+      "grad_norm": 0.3142775297164917,
+      "learning_rate": 7.568259849070486e-07,
+      "loss": 0.011,
+      "step": 2649
+    },
+    {
+      "epoch": 0.8409152886927634,
+      "grad_norm": 0.32579392194747925,
+      "learning_rate": 7.538976611564191e-07,
+      "loss": 0.0138,
+      "step": 2650
+    },
+    {
+      "epoch": 0.8412326152167985,
+      "grad_norm": 0.3628990054130554,
+      "learning_rate": 7.50974551561926e-07,
+      "loss": 0.0152,
+      "step": 2651
+    },
+    {
+      "epoch": 0.8415499417408335,
+      "grad_norm": 0.2748587429523468,
+      "learning_rate": 7.480566597131051e-07,
+      "loss": 0.012,
+      "step": 2652
+    },
+    {
+      "epoch": 0.8418672682648685,
+      "grad_norm": 0.9767450094223022,
+      "learning_rate": 7.451439891930917e-07,
+      "loss": 0.0147,
+      "step": 2653
+    },
+    {
+      "epoch": 0.8421845947889035,
+      "grad_norm": 0.3125697374343872,
+      "learning_rate": 7.422365435786022e-07,
+      "loss": 0.0135,
+      "step": 2654
+    },
+    {
+      "epoch": 0.8425019213129384,
+      "grad_norm": 0.3226590156555176,
+      "learning_rate": 7.393343264399439e-07,
+      "loss": 0.0104,
+      "step": 2655
+    },
+    {
+      "epoch": 0.8428192478369735,
+      "grad_norm": 0.3138394355773926,
+      "learning_rate": 7.364373413409964e-07,
+      "loss": 0.0148,
+      "step": 2656
+    },
+    {
+      "epoch": 0.8431365743610085,
+      "grad_norm": 0.3015248775482178,
+      "learning_rate": 7.33545591839222e-07,
+      "loss": 0.0116,
+      "step": 2657
+    },
+    {
+      "epoch": 0.8434539008850435,
+      "grad_norm": 0.3655318319797516,
+      "learning_rate": 7.306590814856474e-07,
+      "loss": 0.0174,
+      "step": 2658
+    },
+    {
+      "epoch": 0.8437712274090785,
+      "grad_norm": 0.34401243925094604,
+      "learning_rate": 7.277778138248681e-07,
+      "loss": 0.0151,
+      "step": 2659
+    },
+    {
+      "epoch": 0.8440885539331136,
+      "grad_norm": 0.34693798422813416,
+      "learning_rate": 7.24901792395043e-07,
+      "loss": 0.0117,
+      "step": 2660
+    },
+    {
+      "epoch": 0.8444058804571485,
+      "grad_norm": 0.34092068672180176,
+      "learning_rate": 7.220310207278863e-07,
+      "loss": 0.013,
+      "step": 2661
+    },
+    {
+      "epoch": 0.8447232069811835,
+      "grad_norm": 0.4595496654510498,
+      "learning_rate": 7.191655023486682e-07,
+      "loss": 0.0136,
+      "step": 2662
+    },
+    {
+      "epoch": 0.8450405335052186,
+      "grad_norm": 0.3220844566822052,
+      "learning_rate": 7.163052407762045e-07,
+      "loss": 0.0123,
+      "step": 2663
+    },
+    {
+      "epoch": 0.8453578600292535,
+      "grad_norm": 0.2672998011112213,
+      "learning_rate": 7.134502395228593e-07,
+      "loss": 0.0141,
+      "step": 2664
+    },
+    {
+      "epoch": 0.8456751865532885,
+      "grad_norm": 0.27619096636772156,
+      "learning_rate": 7.106005020945339e-07,
+      "loss": 0.0168,
+      "step": 2665
+    },
+    {
+      "epoch": 0.8459925130773236,
+      "grad_norm": 0.3456980288028717,
+      "learning_rate": 7.077560319906696e-07,
+      "loss": 0.0155,
+      "step": 2666
+    },
+    {
+      "epoch": 0.8463098396013585,
+      "grad_norm": 0.29126960039138794,
+      "learning_rate": 7.049168327042355e-07,
+      "loss": 0.0135,
+      "step": 2667
+    },
+    {
+      "epoch": 0.8466271661253936,
+      "grad_norm": 0.29286107420921326,
+      "learning_rate": 7.020829077217289e-07,
+      "loss": 0.0114,
+      "step": 2668
+    },
+    {
+      "epoch": 0.8469444926494286,
+      "grad_norm": 0.34655508399009705,
+      "learning_rate": 6.992542605231739e-07,
+      "loss": 0.0144,
+      "step": 2669
+    },
+    {
+      "epoch": 0.8472618191734635,
+      "grad_norm": 0.27362260222435,
+      "learning_rate": 6.96430894582108e-07,
+      "loss": 0.0101,
+      "step": 2670
+    },
+    {
+      "epoch": 0.8475791456974986,
+      "grad_norm": 0.31191667914390564,
+      "learning_rate": 6.93612813365589e-07,
+      "loss": 0.0135,
+      "step": 2671
+    },
+    {
+      "epoch": 0.8478964722215335,
+      "grad_norm": 0.42535537481307983,
+      "learning_rate": 6.908000203341802e-07,
+      "loss": 0.0177,
+      "step": 2672
+    },
+    {
+      "epoch": 0.8482137987455686,
+      "grad_norm": 0.33755335211753845,
+      "learning_rate": 6.879925189419556e-07,
+      "loss": 0.0133,
+      "step": 2673
+    },
+    {
+      "epoch": 0.8485311252696036,
+      "grad_norm": 0.341694176197052,
+      "learning_rate": 6.851903126364862e-07,
+      "loss": 0.0136,
+      "step": 2674
+    },
+    {
+      "epoch": 0.8488484517936385,
+      "grad_norm": 0.30101582407951355,
+      "learning_rate": 6.82393404858846e-07,
+      "loss": 0.0114,
+      "step": 2675
+    },
+    {
+      "epoch": 0.8491657783176736,
+      "grad_norm": 0.26999253034591675,
+      "learning_rate": 6.796017990435977e-07,
+      "loss": 0.0129,
+      "step": 2676
+    },
+    {
+      "epoch": 0.8494831048417086,
+      "grad_norm": 0.5647525787353516,
+      "learning_rate": 6.768154986187975e-07,
+      "loss": 0.0134,
+      "step": 2677
+    },
+    {
+      "epoch": 0.8498004313657436,
+      "grad_norm": 0.35429972410202026,
+      "learning_rate": 6.740345070059829e-07,
+      "loss": 0.0178,
+      "step": 2678
+    },
+    {
+      "epoch": 0.8501177578897786,
+      "grad_norm": 0.46233436465263367,
+      "learning_rate": 6.712588276201737e-07,
+      "loss": 0.0114,
+      "step": 2679
+    },
+    {
+      "epoch": 0.8504350844138137,
+      "grad_norm": 0.26998889446258545,
+      "learning_rate": 6.684884638698685e-07,
+      "loss": 0.0127,
+      "step": 2680
+    },
+    {
+      "epoch": 0.8507524109378486,
+      "grad_norm": 0.2778751254081726,
+      "learning_rate": 6.657234191570344e-07,
+      "loss": 0.0111,
+      "step": 2681
+    },
+    {
+      "epoch": 0.8510697374618836,
+      "grad_norm": 0.5455251932144165,
+      "learning_rate": 6.629636968771103e-07,
+      "loss": 0.0117,
+      "step": 2682
+    },
+    {
+      "epoch": 0.8513870639859187,
+      "grad_norm": 0.3082531988620758,
+      "learning_rate": 6.602093004189963e-07,
+      "loss": 0.0136,
+      "step": 2683
+    },
+    {
+      "epoch": 0.8517043905099536,
+      "grad_norm": 0.41919851303100586,
+      "learning_rate": 6.574602331650559e-07,
+      "loss": 0.0114,
+      "step": 2684
+    },
+    {
+      "epoch": 0.8520217170339887,
+      "grad_norm": 1.1842098236083984,
+      "learning_rate": 6.547164984911042e-07,
+      "loss": 0.0156,
+      "step": 2685
+    },
+    {
+      "epoch": 0.8523390435580237,
+      "grad_norm": 0.3882652819156647,
+      "learning_rate": 6.519780997664115e-07,
+      "loss": 0.012,
+      "step": 2686
+    },
+    {
+      "epoch": 0.8526563700820586,
+      "grad_norm": 0.3414646089076996,
+      "learning_rate": 6.492450403536959e-07,
+      "loss": 0.0138,
+      "step": 2687
+    },
+    {
+      "epoch": 0.8529736966060937,
+      "grad_norm": 0.34248214960098267,
+      "learning_rate": 6.465173236091138e-07,
+      "loss": 0.0144,
+      "step": 2688
+    },
+    {
+      "epoch": 0.8532910231301286,
+      "grad_norm": 0.42740803956985474,
+      "learning_rate": 6.437949528822674e-07,
+      "loss": 0.0204,
+      "step": 2689
+    },
+    {
+      "epoch": 0.8536083496541637,
+      "grad_norm": 0.522136926651001,
+      "learning_rate": 6.410779315161885e-07,
+      "loss": 0.0141,
+      "step": 2690
+    },
+    {
+      "epoch": 0.8539256761781987,
+      "grad_norm": 0.26047345995903015,
+      "learning_rate": 6.383662628473447e-07,
+      "loss": 0.0114,
+      "step": 2691
+    },
+    {
+      "epoch": 0.8542430027022336,
+      "grad_norm": 0.4682519733905792,
+      "learning_rate": 6.356599502056271e-07,
+      "loss": 0.0142,
+      "step": 2692
+    },
+    {
+      "epoch": 0.8545603292262687,
+      "grad_norm": 0.3418788015842438,
+      "learning_rate": 6.329589969143518e-07,
+      "loss": 0.0149,
+      "step": 2693
+    },
+    {
+      "epoch": 0.8548776557503037,
+      "grad_norm": 0.2848743796348572,
+      "learning_rate": 6.302634062902519e-07,
+      "loss": 0.012,
+      "step": 2694
+    },
+    {
+      "epoch": 0.8551949822743387,
+      "grad_norm": 0.4291332960128784,
+      "learning_rate": 6.275731816434766e-07,
+      "loss": 0.0103,
+      "step": 2695
+    },
+    {
+      "epoch": 0.8555123087983737,
+      "grad_norm": 0.3299597203731537,
+      "learning_rate": 6.248883262775862e-07,
+      "loss": 0.0131,
+      "step": 2696
+    },
+    {
+      "epoch": 0.8558296353224087,
+      "grad_norm": 0.5998820662498474,
+      "learning_rate": 6.222088434895462e-07,
+      "loss": 0.0143,
+      "step": 2697
+    },
+    {
+      "epoch": 0.8561469618464437,
+      "grad_norm": 0.30976974964141846,
+      "learning_rate": 6.195347365697246e-07,
+      "loss": 0.0122,
+      "step": 2698
+    },
+    {
+      "epoch": 0.8564642883704787,
+      "grad_norm": 0.2995264232158661,
+      "learning_rate": 6.16866008801888e-07,
+      "loss": 0.0106,
+      "step": 2699
+    },
+    {
+      "epoch": 0.8567816148945138,
+      "grad_norm": 0.30966717004776,
+      "learning_rate": 6.142026634631998e-07,
+      "loss": 0.0128,
+      "step": 2700
+    },
+    {
+      "epoch": 0.8570989414185487,
+      "grad_norm": 0.2559893727302551,
+      "learning_rate": 6.115447038242095e-07,
+      "loss": 0.0108,
+      "step": 2701
+    },
+    {
+      "epoch": 0.8574162679425837,
+      "grad_norm": 0.7940621376037598,
+      "learning_rate": 6.088921331488568e-07,
+      "loss": 0.0143,
+      "step": 2702
+    },
+    {
+      "epoch": 0.8577335944666188,
+      "grad_norm": 0.44165828824043274,
+      "learning_rate": 6.062449546944643e-07,
+      "loss": 0.0154,
+      "step": 2703
+    },
+    {
+      "epoch": 0.8580509209906537,
+      "grad_norm": 0.2752687633037567,
+      "learning_rate": 6.03603171711728e-07,
+      "loss": 0.012,
+      "step": 2704
+    },
+    {
+      "epoch": 0.8583682475146888,
+      "grad_norm": 0.2896328866481781,
+      "learning_rate": 6.009667874447244e-07,
+      "loss": 0.0127,
+      "step": 2705
+    },
+    {
+      "epoch": 0.8586855740387237,
+      "grad_norm": 0.3210100829601288,
+      "learning_rate": 5.983358051308952e-07,
+      "loss": 0.0157,
+      "step": 2706
+    },
+    {
+      "epoch": 0.8590029005627587,
+      "grad_norm": 0.28336453437805176,
+      "learning_rate": 5.957102280010546e-07,
+      "loss": 0.0118,
+      "step": 2707
+    },
+    {
+      "epoch": 0.8593202270867938,
+      "grad_norm": 0.2996644675731659,
+      "learning_rate": 5.930900592793715e-07,
+      "loss": 0.0137,
+      "step": 2708
+    },
+    {
+      "epoch": 0.8596375536108287,
+      "grad_norm": 0.6069369912147522,
+      "learning_rate": 5.904753021833804e-07,
+      "loss": 0.0137,
+      "step": 2709
+    },
+    {
+      "epoch": 0.8599548801348638,
+      "grad_norm": 0.30916112661361694,
+      "learning_rate": 5.878659599239655e-07,
+      "loss": 0.0121,
+      "step": 2710
+    },
+    {
+      "epoch": 0.8602722066588988,
+      "grad_norm": 0.37320324778556824,
+      "learning_rate": 5.852620357053651e-07,
+      "loss": 0.0132,
+      "step": 2711
+    },
+    {
+      "epoch": 0.8605895331829337,
+      "grad_norm": 0.3157472312450409,
+      "learning_rate": 5.82663532725164e-07,
+      "loss": 0.0174,
+      "step": 2712
+    },
+    {
+      "epoch": 0.8609068597069688,
+      "grad_norm": 0.33607903122901917,
+      "learning_rate": 5.800704541742857e-07,
+      "loss": 0.0127,
+      "step": 2713
+    },
+    {
+      "epoch": 0.8612241862310038,
+      "grad_norm": 1.0808379650115967,
+      "learning_rate": 5.774828032369983e-07,
+      "loss": 0.0147,
+      "step": 2714
+    },
+    {
+      "epoch": 0.8615415127550388,
+      "grad_norm": 0.3460705280303955,
+      "learning_rate": 5.749005830909005e-07,
+      "loss": 0.0141,
+      "step": 2715
+    },
+    {
+      "epoch": 0.8618588392790738,
+      "grad_norm": 0.2704014182090759,
+      "learning_rate": 5.723237969069246e-07,
+      "loss": 0.0134,
+      "step": 2716
+    },
+    {
+      "epoch": 0.8621761658031089,
+      "grad_norm": 0.2794088125228882,
+      "learning_rate": 5.697524478493288e-07,
+      "loss": 0.0144,
+      "step": 2717
+    },
+    {
+      "epoch": 0.8624934923271438,
+      "grad_norm": 0.3347330093383789,
+      "learning_rate": 5.671865390756948e-07,
+      "loss": 0.0131,
+      "step": 2718
+    },
+    {
+      "epoch": 0.8628108188511788,
+      "grad_norm": 0.2701948285102844,
+      "learning_rate": 5.646260737369225e-07,
+      "loss": 0.0111,
+      "step": 2719
+    },
+    {
+      "epoch": 0.8631281453752139,
+      "grad_norm": 0.37814491987228394,
+      "learning_rate": 5.620710549772295e-07,
+      "loss": 0.0143,
+      "step": 2720
+    },
+    {
+      "epoch": 0.8634454718992488,
+      "grad_norm": 0.28858959674835205,
+      "learning_rate": 5.595214859341464e-07,
+      "loss": 0.0112,
+      "step": 2721
+    },
+    {
+      "epoch": 0.8637627984232839,
+      "grad_norm": 0.3872847855091095,
+      "learning_rate": 5.569773697385061e-07,
+      "loss": 0.0156,
+      "step": 2722
+    },
+    {
+      "epoch": 0.8640801249473188,
+      "grad_norm": 0.3281511664390564,
+      "learning_rate": 5.54438709514451e-07,
+      "loss": 0.013,
+      "step": 2723
+    },
+    {
+      "epoch": 0.8643974514713538,
+      "grad_norm": 0.3460828959941864,
+      "learning_rate": 5.519055083794206e-07,
+      "loss": 0.013,
+      "step": 2724
+    },
+    {
+      "epoch": 0.8647147779953889,
+      "grad_norm": 0.28578415513038635,
+      "learning_rate": 5.493777694441521e-07,
+      "loss": 0.0105,
+      "step": 2725
+    },
+    {
+      "epoch": 0.8650321045194238,
+      "grad_norm": 0.2402612715959549,
+      "learning_rate": 5.468554958126737e-07,
+      "loss": 0.0123,
+      "step": 2726
+    },
+    {
+      "epoch": 0.8653494310434588,
+      "grad_norm": 0.2686838209629059,
+      "learning_rate": 5.443386905823045e-07,
+      "loss": 0.0131,
+      "step": 2727
+    },
+    {
+      "epoch": 0.8656667575674939,
+      "grad_norm": 0.2934703826904297,
+      "learning_rate": 5.418273568436466e-07,
+      "loss": 0.0134,
+      "step": 2728
+    },
+    {
+      "epoch": 0.8659840840915288,
+      "grad_norm": 0.26395538449287415,
+      "learning_rate": 5.393214976805833e-07,
+      "loss": 0.0125,
+      "step": 2729
+    },
+    {
+      "epoch": 0.8663014106155639,
+      "grad_norm": 0.2838364839553833,
+      "learning_rate": 5.368211161702763e-07,
+      "loss": 0.0102,
+      "step": 2730
+    },
+    {
+      "epoch": 0.8666187371395989,
+      "grad_norm": 0.2914049029350281,
+      "learning_rate": 5.343262153831596e-07,
+      "loss": 0.0119,
+      "step": 2731
+    },
+    {
+      "epoch": 0.8669360636636338,
+      "grad_norm": 0.34540829062461853,
+      "learning_rate": 5.318367983829393e-07,
+      "loss": 0.0132,
+      "step": 2732
+    },
+    {
+      "epoch": 0.8672533901876689,
+      "grad_norm": 0.34043240547180176,
+      "learning_rate": 5.293528682265831e-07,
+      "loss": 0.0148,
+      "step": 2733
+    },
+    {
+      "epoch": 0.8675707167117039,
+      "grad_norm": 0.3409547209739685,
+      "learning_rate": 5.268744279643262e-07,
+      "loss": 0.0087,
+      "step": 2734
+    },
+    {
+      "epoch": 0.8678880432357389,
+      "grad_norm": 0.3093658983707428,
+      "learning_rate": 5.244014806396569e-07,
+      "loss": 0.011,
+      "step": 2735
+    },
+    {
+      "epoch": 0.8682053697597739,
+      "grad_norm": 0.26403477787971497,
+      "learning_rate": 5.219340292893238e-07,
+      "loss": 0.0132,
+      "step": 2736
+    },
+    {
+      "epoch": 0.868522696283809,
+      "grad_norm": 0.5612727403640747,
+      "learning_rate": 5.194720769433226e-07,
+      "loss": 0.0144,
+      "step": 2737
+    },
+    {
+      "epoch": 0.8688400228078439,
+      "grad_norm": 0.3012463450431824,
+      "learning_rate": 5.17015626624896e-07,
+      "loss": 0.0139,
+      "step": 2738
+    },
+    {
+      "epoch": 0.8691573493318789,
+      "grad_norm": 0.3414081931114197,
+      "learning_rate": 5.145646813505339e-07,
+      "loss": 0.0137,
+      "step": 2739
+    },
+    {
+      "epoch": 0.8694746758559139,
+      "grad_norm": 0.3212358355522156,
+      "learning_rate": 5.121192441299628e-07,
+      "loss": 0.0138,
+      "step": 2740
+    },
+    {
+      "epoch": 0.8697920023799489,
+      "grad_norm": 0.3216117322444916,
+      "learning_rate": 5.096793179661463e-07,
+      "loss": 0.0116,
+      "step": 2741
+    },
+    {
+      "epoch": 0.870109328903984,
+      "grad_norm": 0.28926587104797363,
+      "learning_rate": 5.072449058552809e-07,
+      "loss": 0.0135,
+      "step": 2742
+    },
+    {
+      "epoch": 0.8704266554280189,
+      "grad_norm": 0.2835099697113037,
+      "learning_rate": 5.048160107867928e-07,
+      "loss": 0.0114,
+      "step": 2743
+    },
+    {
+      "epoch": 0.8707439819520539,
+      "grad_norm": 0.26874589920043945,
+      "learning_rate": 5.023926357433296e-07,
+      "loss": 0.0114,
+      "step": 2744
+    },
+    {
+      "epoch": 0.871061308476089,
+      "grad_norm": 0.2615184783935547,
+      "learning_rate": 4.999747837007662e-07,
+      "loss": 0.0115,
+      "step": 2745
+    },
+    {
+      "epoch": 0.8713786350001239,
+      "grad_norm": 0.2877953052520752,
+      "learning_rate": 4.97562457628189e-07,
+      "loss": 0.0125,
+      "step": 2746
+    },
+    {
+      "epoch": 0.871695961524159,
+      "grad_norm": 0.31606653332710266,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.014,
+      "step": 2747
+    },
+    {
+      "epoch": 0.872013288048194,
+      "grad_norm": 0.3179304897785187,
+      "learning_rate": 4.927543952354257e-07,
+      "loss": 0.0117,
+      "step": 2748
+    },
+    {
+      "epoch": 0.8723306145722289,
+      "grad_norm": 0.3574990928173065,
+      "learning_rate": 4.90358664819473e-07,
+      "loss": 0.0144,
+      "step": 2749
+    },
+    {
+      "epoch": 0.872647941096264,
+      "grad_norm": 0.3040415346622467,
+      "learning_rate": 4.879684721819728e-07,
+      "loss": 0.0136,
+      "step": 2750
+    },
+    {
+      "epoch": 0.872965267620299,
+      "grad_norm": 0.3670097887516022,
+      "learning_rate": 4.85583820258047e-07,
+      "loss": 0.0135,
+      "step": 2751
+    },
+    {
+      "epoch": 0.873282594144334,
+      "grad_norm": 0.3290829658508301,
+      "learning_rate": 4.832047119760186e-07,
+      "loss": 0.0148,
+      "step": 2752
+    },
+    {
+      "epoch": 0.873599920668369,
+      "grad_norm": 0.37180569767951965,
+      "learning_rate": 4.808311502573976e-07,
+      "loss": 0.0151,
+      "step": 2753
+    },
+    {
+      "epoch": 0.873917247192404,
+      "grad_norm": 0.3862403929233551,
+      "learning_rate": 4.784631380168875e-07,
+      "loss": 0.0135,
+      "step": 2754
+    },
+    {
+      "epoch": 0.874234573716439,
+      "grad_norm": 0.4051336944103241,
+      "learning_rate": 4.7610067816237336e-07,
+      "loss": 0.0126,
+      "step": 2755
+    },
+    {
+      "epoch": 0.874551900240474,
+      "grad_norm": 0.3356613516807556,
+      "learning_rate": 4.737437735949263e-07,
+      "loss": 0.0165,
+      "step": 2756
+    },
+    {
+      "epoch": 0.8748692267645091,
+      "grad_norm": 0.2633938193321228,
+      "learning_rate": 4.713924272087922e-07,
+      "loss": 0.0113,
+      "step": 2757
+    },
+    {
+      "epoch": 0.875186553288544,
+      "grad_norm": 0.3358660936355591,
+      "learning_rate": 4.690466418913914e-07,
+      "loss": 0.0114,
+      "step": 2758
+    },
+    {
+      "epoch": 0.875503879812579,
+      "grad_norm": 0.3369928002357483,
+      "learning_rate": 4.6670642052331793e-07,
+      "loss": 0.0138,
+      "step": 2759
+    },
+    {
+      "epoch": 0.875821206336614,
+      "grad_norm": 0.34340205788612366,
+      "learning_rate": 4.643717659783309e-07,
+      "loss": 0.012,
+      "step": 2760
+    },
+    {
+      "epoch": 0.876138532860649,
+      "grad_norm": 0.3562813401222229,
+      "learning_rate": 4.620426811233558e-07,
+      "loss": 0.0122,
+      "step": 2761
+    },
+    {
+      "epoch": 0.8764558593846841,
+      "grad_norm": 0.27377620339393616,
+      "learning_rate": 4.5971916881847543e-07,
+      "loss": 0.0146,
+      "step": 2762
+    },
+    {
+      "epoch": 0.876773185908719,
+      "grad_norm": 0.30182477831840515,
+      "learning_rate": 4.57401231916933e-07,
+      "loss": 0.0125,
+      "step": 2763
+    },
+    {
+      "epoch": 0.877090512432754,
+      "grad_norm": 0.35292133688926697,
+      "learning_rate": 4.5508887326512266e-07,
+      "loss": 0.0137,
+      "step": 2764
+    },
+    {
+      "epoch": 0.8774078389567891,
+      "grad_norm": 0.2793472707271576,
+      "learning_rate": 4.5278209570258914e-07,
+      "loss": 0.0114,
+      "step": 2765
+    },
+    {
+      "epoch": 0.877725165480824,
+      "grad_norm": 0.35573819279670715,
+      "learning_rate": 4.5048090206202657e-07,
+      "loss": 0.0129,
+      "step": 2766
+    },
+    {
+      "epoch": 0.8780424920048591,
+      "grad_norm": 0.29732802510261536,
+      "learning_rate": 4.481852951692672e-07,
+      "loss": 0.0168,
+      "step": 2767
+    },
+    {
+      "epoch": 0.8783598185288941,
+      "grad_norm": 0.3891223967075348,
+      "learning_rate": 4.458952778432857e-07,
+      "loss": 0.0115,
+      "step": 2768
+    },
+    {
+      "epoch": 0.878677145052929,
+      "grad_norm": 0.4266033470630646,
+      "learning_rate": 4.436108528961913e-07,
+      "loss": 0.0118,
+      "step": 2769
+    },
+    {
+      "epoch": 0.8789944715769641,
+      "grad_norm": 0.32447296380996704,
+      "learning_rate": 4.413320231332291e-07,
+      "loss": 0.0128,
+      "step": 2770
+    },
+    {
+      "epoch": 0.8793117981009991,
+      "grad_norm": 0.3583220839500427,
+      "learning_rate": 4.390587913527694e-07,
+      "loss": 0.0138,
+      "step": 2771
+    },
+    {
+      "epoch": 0.8796291246250341,
+      "grad_norm": 0.4123729467391968,
+      "learning_rate": 4.3679116034631096e-07,
+      "loss": 0.0106,
+      "step": 2772
+    },
+    {
+      "epoch": 0.8799464511490691,
+      "grad_norm": 0.3192569315433502,
+      "learning_rate": 4.34529132898473e-07,
+      "loss": 0.0133,
+      "step": 2773
+    },
+    {
+      "epoch": 0.8802637776731042,
+      "grad_norm": 0.33305734395980835,
+      "learning_rate": 4.322727117869951e-07,
+      "loss": 0.0122,
+      "step": 2774
+    },
+    {
+      "epoch": 0.8805811041971391,
+      "grad_norm": 0.4422624707221985,
+      "learning_rate": 4.3002189978273265e-07,
+      "loss": 0.0199,
+      "step": 2775
+    },
+    {
+      "epoch": 0.8808984307211741,
+      "grad_norm": 0.3541223704814911,
+      "learning_rate": 4.2777669964965074e-07,
+      "loss": 0.0137,
+      "step": 2776
+    },
+    {
+      "epoch": 0.8812157572452091,
+      "grad_norm": 0.34024664759635925,
+      "learning_rate": 4.255371141448272e-07,
+      "loss": 0.0109,
+      "step": 2777
+    },
+    {
+      "epoch": 0.8815330837692441,
+      "grad_norm": 0.37475040555000305,
+      "learning_rate": 4.2330314601843945e-07,
+      "loss": 0.0172,
+      "step": 2778
+    },
+    {
+      "epoch": 0.8818504102932792,
+      "grad_norm": 0.2543153762817383,
+      "learning_rate": 4.210747980137725e-07,
+      "loss": 0.0108,
+      "step": 2779
+    },
+    {
+      "epoch": 0.8821677368173141,
+      "grad_norm": 0.5705439448356628,
+      "learning_rate": 4.188520728672052e-07,
+      "loss": 0.0166,
+      "step": 2780
+    },
+    {
+      "epoch": 0.8824850633413491,
+      "grad_norm": 0.25810036063194275,
+      "learning_rate": 4.1663497330821536e-07,
+      "loss": 0.0119,
+      "step": 2781
+    },
+    {
+      "epoch": 0.8828023898653842,
+      "grad_norm": 0.3595520853996277,
+      "learning_rate": 4.144235020593723e-07,
+      "loss": 0.0147,
+      "step": 2782
+    },
+    {
+      "epoch": 0.8831197163894191,
+      "grad_norm": 0.31457337737083435,
+      "learning_rate": 4.122176618363305e-07,
+      "loss": 0.013,
+      "step": 2783
+    },
+    {
+      "epoch": 0.8834370429134542,
+      "grad_norm": 0.32256531715393066,
+      "learning_rate": 4.1001745534783423e-07,
+      "loss": 0.0167,
+      "step": 2784
+    },
+    {
+      "epoch": 0.8837543694374892,
+      "grad_norm": 0.3404534161090851,
+      "learning_rate": 4.07822885295705e-07,
+      "loss": 0.0163,
+      "step": 2785
+    },
+    {
+      "epoch": 0.8840716959615241,
+      "grad_norm": 0.34233778715133667,
+      "learning_rate": 4.0563395437484666e-07,
+      "loss": 0.0148,
+      "step": 2786
+    },
+    {
+      "epoch": 0.8843890224855592,
+      "grad_norm": 0.3514814078807831,
+      "learning_rate": 4.0345066527323616e-07,
+      "loss": 0.0133,
+      "step": 2787
+    },
+    {
+      "epoch": 0.8847063490095942,
+      "grad_norm": 0.3726923167705536,
+      "learning_rate": 4.0127302067192285e-07,
+      "loss": 0.0144,
+      "step": 2788
+    },
+    {
+      "epoch": 0.8850236755336292,
+      "grad_norm": 0.3278256058692932,
+      "learning_rate": 3.991010232450243e-07,
+      "loss": 0.0116,
+      "step": 2789
+    },
+    {
+      "epoch": 0.8853410020576642,
+      "grad_norm": 0.285565048456192,
+      "learning_rate": 3.969346756597242e-07,
+      "loss": 0.0105,
+      "step": 2790
+    },
+    {
+      "epoch": 0.8856583285816992,
+      "grad_norm": 0.29515665769577026,
+      "learning_rate": 3.9477398057626904e-07,
+      "loss": 0.0124,
+      "step": 2791
+    },
+    {
+      "epoch": 0.8859756551057342,
+      "grad_norm": 0.30805134773254395,
+      "learning_rate": 3.9261894064796136e-07,
+      "loss": 0.0105,
+      "step": 2792
+    },
+    {
+      "epoch": 0.8862929816297692,
+      "grad_norm": 0.35017481446266174,
+      "learning_rate": 3.904695585211632e-07,
+      "loss": 0.0135,
+      "step": 2793
+    },
+    {
+      "epoch": 0.8866103081538041,
+      "grad_norm": 0.2773647904396057,
+      "learning_rate": 3.8832583683528436e-07,
+      "loss": 0.0108,
+      "step": 2794
+    },
+    {
+      "epoch": 0.8869276346778392,
+      "grad_norm": 0.2870911955833435,
+      "learning_rate": 3.8618777822278854e-07,
+      "loss": 0.0121,
+      "step": 2795
+    },
+    {
+      "epoch": 0.8872449612018742,
+      "grad_norm": 0.2959360182285309,
+      "learning_rate": 3.8405538530917994e-07,
+      "loss": 0.0156,
+      "step": 2796
+    },
+    {
+      "epoch": 0.8875622877259092,
+      "grad_norm": 0.32125529646873474,
+      "learning_rate": 3.8192866071301125e-07,
+      "loss": 0.0123,
+      "step": 2797
+    },
+    {
+      "epoch": 0.8878796142499442,
+      "grad_norm": 0.31530460715293884,
+      "learning_rate": 3.798076070458684e-07,
+      "loss": 0.0135,
+      "step": 2798
+    },
+    {
+      "epoch": 0.8881969407739793,
+      "grad_norm": 0.5799832940101624,
+      "learning_rate": 3.7769222691237727e-07,
+      "loss": 0.0141,
+      "step": 2799
+    },
+    {
+      "epoch": 0.8885142672980142,
+      "grad_norm": 0.32777565717697144,
+      "learning_rate": 3.755825229101984e-07,
+      "loss": 0.0111,
+      "step": 2800
+    },
+    {
+      "epoch": 0.8888315938220492,
+      "grad_norm": 0.3359781503677368,
+      "learning_rate": 3.734784976300165e-07,
+      "loss": 0.016,
+      "step": 2801
+    },
+    {
+      "epoch": 0.8891489203460843,
+      "grad_norm": 0.31883785128593445,
+      "learning_rate": 3.7138015365554834e-07,
+      "loss": 0.0118,
+      "step": 2802
+    },
+    {
+      "epoch": 0.8894662468701192,
+      "grad_norm": 0.5425217151641846,
+      "learning_rate": 3.6928749356353044e-07,
+      "loss": 0.0116,
+      "step": 2803
+    },
+    {
+      "epoch": 0.8897835733941543,
+      "grad_norm": 0.40964391827583313,
+      "learning_rate": 3.6720051992372276e-07,
+      "loss": 0.0118,
+      "step": 2804
+    },
+    {
+      "epoch": 0.8901008999181893,
+      "grad_norm": 0.3499176502227783,
+      "learning_rate": 3.651192352988986e-07,
+      "loss": 0.0135,
+      "step": 2805
+    },
+    {
+      "epoch": 0.8904182264422242,
+      "grad_norm": 0.362533301115036,
+      "learning_rate": 3.6304364224484856e-07,
+      "loss": 0.0144,
+      "step": 2806
+    },
+    {
+      "epoch": 0.8907355529662593,
+      "grad_norm": 0.2788107395172119,
+      "learning_rate": 3.6097374331037326e-07,
+      "loss": 0.0113,
+      "step": 2807
+    },
+    {
+      "epoch": 0.8910528794902943,
+      "grad_norm": 0.3616269826889038,
+      "learning_rate": 3.589095410372784e-07,
+      "loss": 0.0159,
+      "step": 2808
+    },
+    {
+      "epoch": 0.8913702060143293,
+      "grad_norm": 0.3377329409122467,
+      "learning_rate": 3.56851037960379e-07,
+      "loss": 0.0134,
+      "step": 2809
+    },
+    {
+      "epoch": 0.8916875325383643,
+      "grad_norm": 0.4688379764556885,
+      "learning_rate": 3.5479823660748703e-07,
+      "loss": 0.0158,
+      "step": 2810
+    },
+    {
+      "epoch": 0.8920048590623992,
+      "grad_norm": 0.2854113280773163,
+      "learning_rate": 3.5275113949941595e-07,
+      "loss": 0.0127,
+      "step": 2811
+    },
+    {
+      "epoch": 0.8923221855864343,
+      "grad_norm": 0.3263043165206909,
+      "learning_rate": 3.5070974914997167e-07,
+      "loss": 0.011,
+      "step": 2812
+    },
+    {
+      "epoch": 0.8926395121104693,
+      "grad_norm": 0.46263107657432556,
+      "learning_rate": 3.486740680659562e-07,
+      "loss": 0.0144,
+      "step": 2813
+    },
+    {
+      "epoch": 0.8929568386345043,
+      "grad_norm": 0.2522924840450287,
+      "learning_rate": 3.466440987471559e-07,
+      "loss": 0.0116,
+      "step": 2814
+    },
+    {
+      "epoch": 0.8932741651585393,
+      "grad_norm": 0.26032546162605286,
+      "learning_rate": 3.446198436863479e-07,
+      "loss": 0.0108,
+      "step": 2815
+    },
+    {
+      "epoch": 0.8935914916825743,
+      "grad_norm": 0.2912227213382721,
+      "learning_rate": 3.426013053692878e-07,
+      "loss": 0.0115,
+      "step": 2816
+    },
+    {
+      "epoch": 0.8939088182066093,
+      "grad_norm": 0.315967857837677,
+      "learning_rate": 3.405884862747155e-07,
+      "loss": 0.0155,
+      "step": 2817
+    },
+    {
+      "epoch": 0.8942261447306443,
+      "grad_norm": 0.31835824251174927,
+      "learning_rate": 3.385813888743444e-07,
+      "loss": 0.0116,
+      "step": 2818
+    },
+    {
+      "epoch": 0.8945434712546794,
+      "grad_norm": 0.29475992918014526,
+      "learning_rate": 3.365800156328619e-07,
+      "loss": 0.0123,
+      "step": 2819
+    },
+    {
+      "epoch": 0.8948607977787143,
+      "grad_norm": 0.312912255525589,
+      "learning_rate": 3.345843690079292e-07,
+      "loss": 0.0094,
+      "step": 2820
+    },
+    {
+      "epoch": 0.8951781243027493,
+      "grad_norm": 0.3848724067211151,
+      "learning_rate": 3.325944514501711e-07,
+      "loss": 0.0142,
+      "step": 2821
+    },
+    {
+      "epoch": 0.8954954508267844,
+      "grad_norm": 0.41990676522254944,
+      "learning_rate": 3.306102654031823e-07,
+      "loss": 0.0142,
+      "step": 2822
+    },
+    {
+      "epoch": 0.8958127773508193,
+      "grad_norm": 0.32180270552635193,
+      "learning_rate": 3.2863181330351325e-07,
+      "loss": 0.0143,
+      "step": 2823
+    },
+    {
+      "epoch": 0.8961301038748544,
+      "grad_norm": 0.3766134977340698,
+      "learning_rate": 3.266590975806788e-07,
+      "loss": 0.0154,
+      "step": 2824
+    },
+    {
+      "epoch": 0.8964474303988894,
+      "grad_norm": 0.41397395730018616,
+      "learning_rate": 3.2469212065714573e-07,
+      "loss": 0.0137,
+      "step": 2825
+    },
+    {
+      "epoch": 0.8967647569229243,
+      "grad_norm": 0.26895374059677124,
+      "learning_rate": 3.227308849483357e-07,
+      "loss": 0.012,
+      "step": 2826
+    },
+    {
+      "epoch": 0.8970820834469594,
+      "grad_norm": 3.202096700668335,
+      "learning_rate": 3.2077539286261947e-07,
+      "loss": 0.012,
+      "step": 2827
+    },
+    {
+      "epoch": 0.8973994099709943,
+      "grad_norm": 0.37230560183525085,
+      "learning_rate": 3.18825646801314e-07,
+      "loss": 0.0166,
+      "step": 2828
+    },
+    {
+      "epoch": 0.8977167364950294,
+      "grad_norm": 0.3756243884563446,
+      "learning_rate": 3.168816491586829e-07,
+      "loss": 0.0147,
+      "step": 2829
+    },
+    {
+      "epoch": 0.8980340630190644,
+      "grad_norm": 0.37143099308013916,
+      "learning_rate": 3.1494340232192667e-07,
+      "loss": 0.0148,
+      "step": 2830
+    },
+    {
+      "epoch": 0.8983513895430993,
+      "grad_norm": 0.3310752809047699,
+      "learning_rate": 3.1301090867118746e-07,
+      "loss": 0.0162,
+      "step": 2831
+    },
+    {
+      "epoch": 0.8986687160671344,
+      "grad_norm": 0.6282012462615967,
+      "learning_rate": 3.110841705795409e-07,
+      "loss": 0.0111,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8989860425911694,
+      "grad_norm": 0.26996517181396484,
+      "learning_rate": 3.0916319041299603e-07,
+      "loss": 0.01,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8993033691152044,
+      "grad_norm": 0.5007233023643494,
+      "learning_rate": 3.0724797053048927e-07,
+      "loss": 0.0107,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8996206956392394,
+      "grad_norm": 0.27117136120796204,
+      "learning_rate": 3.0533851328388596e-07,
+      "loss": 0.0108,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8999380221632745,
+      "grad_norm": 0.35939428210258484,
+      "learning_rate": 3.0343482101797274e-07,
+      "loss": 0.0144,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9002553486873094,
+      "grad_norm": 0.45762449502944946,
+      "learning_rate": 3.015368960704584e-07,
+      "loss": 0.0159,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9005726752113444,
+      "grad_norm": 0.6579799056053162,
+      "learning_rate": 2.996447407719688e-07,
+      "loss": 0.0142,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9008900017353795,
+      "grad_norm": 0.434283047914505,
+      "learning_rate": 2.9775835744604465e-07,
+      "loss": 0.0146,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9012073282594144,
+      "grad_norm": 0.3620293140411377,
+      "learning_rate": 2.9587774840914016e-07,
+      "loss": 0.0125,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9015246547834495,
+      "grad_norm": 0.3555227220058441,
+      "learning_rate": 2.940029159706165e-07,
+      "loss": 0.0129,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9018419813074845,
+      "grad_norm": 0.2930890917778015,
+      "learning_rate": 2.921338624327441e-07,
+      "loss": 0.0108,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9021593078315194,
+      "grad_norm": 0.3032636344432831,
+      "learning_rate": 2.9027059009069323e-07,
+      "loss": 0.0144,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9024766343555545,
+      "grad_norm": 0.2578798532485962,
+      "learning_rate": 2.8841310123253865e-07,
+      "loss": 0.0117,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9027939608795894,
+      "grad_norm": 0.30277174711227417,
+      "learning_rate": 2.865613981392518e-07,
+      "loss": 0.0118,
+      "step": 2845
+    },
+    {
+      "epoch": 0.9031112874036245,
+      "grad_norm": 0.3689783811569214,
+      "learning_rate": 2.847154830846971e-07,
+      "loss": 0.0153,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9034286139276595,
+      "grad_norm": 0.2787918448448181,
+      "learning_rate": 2.828753583356364e-07,
+      "loss": 0.0138,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9037459404516944,
+      "grad_norm": 2.458923101425171,
+      "learning_rate": 2.8104102615171434e-07,
+      "loss": 0.0159,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9040632669757295,
+      "grad_norm": 0.23523010313510895,
+      "learning_rate": 2.792124887854686e-07,
+      "loss": 0.0113,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9043805934997645,
+      "grad_norm": 0.32133597135543823,
+      "learning_rate": 2.7738974848231624e-07,
+      "loss": 0.0121,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9046979200237995,
+      "grad_norm": 0.3230030834674835,
+      "learning_rate": 2.755728074805597e-07,
+      "loss": 0.0125,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9050152465478345,
+      "grad_norm": 0.3152256906032562,
+      "learning_rate": 2.737616680113758e-07,
+      "loss": 0.0153,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9053325730718695,
+      "grad_norm": 0.36492982506752014,
+      "learning_rate": 2.7195633229882114e-07,
+      "loss": 0.0114,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9056498995959045,
+      "grad_norm": 0.3265707194805145,
+      "learning_rate": 2.701568025598228e-07,
+      "loss": 0.0116,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9059672261199395,
+      "grad_norm": 0.3134469985961914,
+      "learning_rate": 2.6836308100417874e-07,
+      "loss": 0.0164,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9062845526439746,
+      "grad_norm": 0.38507169485092163,
+      "learning_rate": 2.6657516983455535e-07,
+      "loss": 0.0148,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9066018791680095,
+      "grad_norm": 0.3218468725681305,
+      "learning_rate": 2.64793071246483e-07,
+      "loss": 0.0135,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9069192056920445,
+      "grad_norm": 0.3644973635673523,
+      "learning_rate": 2.63016787428354e-07,
+      "loss": 0.012,
+      "step": 2858
+    },
+    {
+      "epoch": 0.9072365322160796,
+      "grad_norm": 0.35505250096321106,
+      "learning_rate": 2.612463205614213e-07,
+      "loss": 0.0136,
+      "step": 2859
+    },
+    {
+      "epoch": 0.9075538587401145,
+      "grad_norm": 0.2898251414299011,
+      "learning_rate": 2.5948167281979355e-07,
+      "loss": 0.0156,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9078711852641496,
+      "grad_norm": 0.43171972036361694,
+      "learning_rate": 2.5772284637043563e-07,
+      "loss": 0.0133,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9081885117881845,
+      "grad_norm": 0.3291448652744293,
+      "learning_rate": 2.559698433731617e-07,
+      "loss": 0.011,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9085058383122195,
+      "grad_norm": 0.5292889475822449,
+      "learning_rate": 2.542226659806363e-07,
+      "loss": 0.0138,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9088231648362546,
+      "grad_norm": 0.23574671149253845,
+      "learning_rate": 2.524813163383683e-07,
+      "loss": 0.0111,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9091404913602895,
+      "grad_norm": 0.3343794345855713,
+      "learning_rate": 2.5074579658471266e-07,
+      "loss": 0.0151,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9094578178843246,
+      "grad_norm": 0.537752091884613,
+      "learning_rate": 2.490161088508636e-07,
+      "loss": 0.0126,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9097751444083596,
+      "grad_norm": 0.2759327292442322,
+      "learning_rate": 2.472922552608559e-07,
+      "loss": 0.0094,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9100924709323945,
+      "grad_norm": 0.4368760883808136,
+      "learning_rate": 2.455742379315551e-07,
+      "loss": 0.0136,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9104097974564296,
+      "grad_norm": 0.4420575499534607,
+      "learning_rate": 2.438620589726648e-07,
+      "loss": 0.0108,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9107271239804646,
+      "grad_norm": 0.23793616890907288,
+      "learning_rate": 2.421557204867192e-07,
+      "loss": 0.0097,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9110444505044996,
+      "grad_norm": 0.293649286031723,
+      "learning_rate": 2.4045522456907613e-07,
+      "loss": 0.0134,
+      "step": 2871
+    },
+    {
+      "epoch": 0.9113617770285346,
+      "grad_norm": 0.4080897569656372,
+      "learning_rate": 2.3876057330792344e-07,
+      "loss": 0.012,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9116791035525696,
+      "grad_norm": 0.2864163815975189,
+      "learning_rate": 2.3707176878426886e-07,
+      "loss": 0.0114,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9119964300766046,
+      "grad_norm": 0.32141414284706116,
+      "learning_rate": 2.3538881307194282e-07,
+      "loss": 0.014,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9123137566006396,
+      "grad_norm": 0.31077641248703003,
+      "learning_rate": 2.3371170823759004e-07,
+      "loss": 0.0127,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9126310831246747,
+      "grad_norm": 0.40330958366394043,
+      "learning_rate": 2.3204045634067418e-07,
+      "loss": 0.0123,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9129484096487096,
+      "grad_norm": 0.3902370035648346,
+      "learning_rate": 2.3037505943346972e-07,
+      "loss": 0.0131,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9132657361727446,
+      "grad_norm": 0.38175931572914124,
+      "learning_rate": 2.2871551956106075e-07,
+      "loss": 0.0142,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9135830626967796,
+      "grad_norm": 0.2845118045806885,
+      "learning_rate": 2.2706183876134047e-07,
+      "loss": 0.0147,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9139003892208146,
+      "grad_norm": 0.4060806632041931,
+      "learning_rate": 2.2541401906500615e-07,
+      "loss": 0.0183,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9142177157448497,
+      "grad_norm": 0.3206123411655426,
+      "learning_rate": 2.2377206249555804e-07,
+      "loss": 0.0105,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9145350422688846,
+      "grad_norm": 0.33294519782066345,
+      "learning_rate": 2.2213597106929608e-07,
+      "loss": 0.0131,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9148523687929196,
+      "grad_norm": 0.3538065254688263,
+      "learning_rate": 2.2050574679531977e-07,
+      "loss": 0.0157,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9151696953169547,
+      "grad_norm": 0.2888796031475067,
+      "learning_rate": 2.1888139167552114e-07,
+      "loss": 0.0141,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9154870218409896,
+      "grad_norm": 0.31199514865875244,
+      "learning_rate": 2.1726290770458735e-07,
+      "loss": 0.0125,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9158043483650247,
+      "grad_norm": 0.2825830578804016,
+      "learning_rate": 2.1565029686999306e-07,
+      "loss": 0.0129,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9161216748890597,
+      "grad_norm": 0.3792151212692261,
+      "learning_rate": 2.140435611520042e-07,
+      "loss": 0.0163,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9164390014130946,
+      "grad_norm": 0.34082549810409546,
+      "learning_rate": 2.1244270252367028e-07,
+      "loss": 0.0133,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9167563279371297,
+      "grad_norm": 0.34899425506591797,
+      "learning_rate": 2.1084772295082268e-07,
+      "loss": 0.0173,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9170736544611647,
+      "grad_norm": 0.27188974618911743,
+      "learning_rate": 2.0925862439207633e-07,
+      "loss": 0.012,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9173909809851997,
+      "grad_norm": 0.3742457926273346,
+      "learning_rate": 2.0767540879882143e-07,
+      "loss": 0.0126,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9177083075092347,
+      "grad_norm": 0.287489116191864,
+      "learning_rate": 2.0609807811522665e-07,
+      "loss": 0.0138,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9180256340332698,
+      "grad_norm": 0.38513705134391785,
+      "learning_rate": 2.0452663427823093e-07,
+      "loss": 0.0156,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9183429605573047,
+      "grad_norm": 0.24860358238220215,
+      "learning_rate": 2.0296107921754793e-07,
+      "loss": 0.01,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9186602870813397,
+      "grad_norm": 0.5693565607070923,
+      "learning_rate": 2.0140141485565589e-07,
+      "loss": 0.0132,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9189776136053747,
+      "grad_norm": 0.41502290964126587,
+      "learning_rate": 1.998476431078039e-07,
+      "loss": 0.0156,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9192949401294097,
+      "grad_norm": 0.647026777267456,
+      "learning_rate": 1.982997658820013e-07,
+      "loss": 0.0118,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9196122666534448,
+      "grad_norm": 0.37479984760284424,
+      "learning_rate": 1.9675778507901976e-07,
+      "loss": 0.0123,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9199295931774797,
+      "grad_norm": 0.4911288619041443,
+      "learning_rate": 1.9522170259239192e-07,
+      "loss": 0.0169,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9202469197015147,
+      "grad_norm": 0.2920830547809601,
+      "learning_rate": 1.9369152030840553e-07,
+      "loss": 0.0141,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9205642462255498,
+      "grad_norm": 0.2544691860675812,
+      "learning_rate": 1.9216724010610477e-07,
+      "loss": 0.011,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9208815727495847,
+      "grad_norm": 0.3347272276878357,
+      "learning_rate": 1.9064886385728399e-07,
+      "loss": 0.0156,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9211988992736198,
+      "grad_norm": 0.33420926332473755,
+      "learning_rate": 1.8913639342648893e-07,
+      "loss": 0.0121,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9215162257976548,
+      "grad_norm": 0.31481629610061646,
+      "learning_rate": 1.8762983067101226e-07,
+      "loss": 0.012,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9218335523216897,
+      "grad_norm": 0.35609710216522217,
+      "learning_rate": 1.861291774408941e-07,
+      "loss": 0.016,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9221508788457248,
+      "grad_norm": 0.2689898610115051,
+      "learning_rate": 1.846344355789148e-07,
+      "loss": 0.0094,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9224682053697598,
+      "grad_norm": 0.3952021598815918,
+      "learning_rate": 1.8314560692059836e-07,
+      "loss": 0.0115,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9227855318937948,
+      "grad_norm": 0.30838143825531006,
+      "learning_rate": 1.816626932942056e-07,
+      "loss": 0.0126,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9231028584178298,
+      "grad_norm": 0.34041520953178406,
+      "learning_rate": 1.801856965207338e-07,
+      "loss": 0.0154,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9234201849418648,
+      "grad_norm": 0.28676292300224304,
+      "learning_rate": 1.7871461841391603e-07,
+      "loss": 0.0117,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9237375114658998,
+      "grad_norm": 0.28846433758735657,
+      "learning_rate": 1.77249460780215e-07,
+      "loss": 0.0121,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9240548379899348,
+      "grad_norm": 0.37211740016937256,
+      "learning_rate": 1.757902254188254e-07,
+      "loss": 0.0144,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9243721645139698,
+      "grad_norm": 0.37717708945274353,
+      "learning_rate": 1.743369141216683e-07,
+      "loss": 0.0158,
+      "step": 2913
+    },
+    {
+      "epoch": 0.9246894910380048,
+      "grad_norm": 0.2882973849773407,
+      "learning_rate": 1.728895286733906e-07,
+      "loss": 0.0145,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9250068175620398,
+      "grad_norm": 0.4582322835922241,
+      "learning_rate": 1.7144807085136105e-07,
+      "loss": 0.015,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9253241440860748,
+      "grad_norm": 0.29735979437828064,
+      "learning_rate": 1.7001254242567043e-07,
+      "loss": 0.012,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9256414706101098,
+      "grad_norm": 0.2790689468383789,
+      "learning_rate": 1.6858294515913031e-07,
+      "loss": 0.0117,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9259587971341449,
+      "grad_norm": 0.3359939455986023,
+      "learning_rate": 1.6715928080726417e-07,
+      "loss": 0.01,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9262761236581798,
+      "grad_norm": 0.26120322942733765,
+      "learning_rate": 1.6574155111831413e-07,
+      "loss": 0.0093,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9265934501822148,
+      "grad_norm": 0.49146977066993713,
+      "learning_rate": 1.6432975783323145e-07,
+      "loss": 0.0171,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9269107767062499,
+      "grad_norm": 0.2993941903114319,
+      "learning_rate": 1.6292390268568103e-07,
+      "loss": 0.0147,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9272281032302848,
+      "grad_norm": 0.3551235496997833,
+      "learning_rate": 1.615239874020319e-07,
+      "loss": 0.0143,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9275454297543199,
+      "grad_norm": 0.49446946382522583,
+      "learning_rate": 1.6013001370136283e-07,
+      "loss": 0.0171,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9278627562783549,
+      "grad_norm": 0.5602728724479675,
+      "learning_rate": 1.5874198329545398e-07,
+      "loss": 0.012,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9281800828023898,
+      "grad_norm": 0.45830243825912476,
+      "learning_rate": 1.5735989788878746e-07,
+      "loss": 0.0147,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9284974093264249,
+      "grad_norm": 0.3296537697315216,
+      "learning_rate": 1.559837591785468e-07,
+      "loss": 0.0134,
+      "step": 2926
+    },
+    {
+      "epoch": 0.9288147358504599,
+      "grad_norm": 0.33876484632492065,
+      "learning_rate": 1.5461356885461077e-07,
+      "loss": 0.0129,
+      "step": 2927
+    },
+    {
+      "epoch": 0.9291320623744949,
+      "grad_norm": 0.31973785161972046,
+      "learning_rate": 1.53249328599554e-07,
+      "loss": 0.0143,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9294493888985299,
+      "grad_norm": 0.33615636825561523,
+      "learning_rate": 1.518910400886453e-07,
+      "loss": 0.0122,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9297667154225648,
+      "grad_norm": 0.3848973214626312,
+      "learning_rate": 1.505387049898449e-07,
+      "loss": 0.0146,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9300840419465999,
+      "grad_norm": 0.27356937527656555,
+      "learning_rate": 1.491923249638022e-07,
+      "loss": 0.0109,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9304013684706349,
+      "grad_norm": 0.35712289810180664,
+      "learning_rate": 1.4785190166385244e-07,
+      "loss": 0.0149,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9307186949946699,
+      "grad_norm": 0.4543388783931732,
+      "learning_rate": 1.4651743673601894e-07,
+      "loss": 0.0101,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9310360215187049,
+      "grad_norm": 0.3583502471446991,
+      "learning_rate": 1.451889318190053e-07,
+      "loss": 0.014,
+      "step": 2934
+    },
+    {
+      "epoch": 0.93135334804274,
+      "grad_norm": 0.39497265219688416,
+      "learning_rate": 1.438663885441982e-07,
+      "loss": 0.0153,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9316706745667749,
+      "grad_norm": 0.32626479864120483,
+      "learning_rate": 1.4254980853566248e-07,
+      "loss": 0.0155,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9319880010908099,
+      "grad_norm": 0.36946702003479004,
+      "learning_rate": 1.4123919341014203e-07,
+      "loss": 0.0126,
+      "step": 2937
+    },
+    {
+      "epoch": 0.932305327614845,
+      "grad_norm": 1.3751246929168701,
+      "learning_rate": 1.399345447770517e-07,
+      "loss": 0.0125,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9326226541388799,
+      "grad_norm": 0.2399183064699173,
+      "learning_rate": 1.3863586423848385e-07,
+      "loss": 0.0092,
+      "step": 2939
+    },
+    {
+      "epoch": 0.932939980662915,
+      "grad_norm": 0.3453684449195862,
+      "learning_rate": 1.3734315338920056e-07,
+      "loss": 0.0158,
+      "step": 2940
+    },
+    {
+      "epoch": 0.93325730718695,
+      "grad_norm": 0.2616996467113495,
+      "learning_rate": 1.3605641381663315e-07,
+      "loss": 0.0112,
+      "step": 2941
+    },
+    {
+      "epoch": 0.9335746337109849,
+      "grad_norm": 0.5160927176475525,
+      "learning_rate": 1.3477564710088097e-07,
+      "loss": 0.0179,
+      "step": 2942
+    },
+    {
+      "epoch": 0.93389196023502,
+      "grad_norm": 0.2702154815196991,
+      "learning_rate": 1.335008548147071e-07,
+      "loss": 0.0105,
+      "step": 2943
+    },
+    {
+      "epoch": 0.934209286759055,
+      "grad_norm": 0.29220494627952576,
+      "learning_rate": 1.322320385235404e-07,
+      "loss": 0.0129,
+      "step": 2944
+    },
+    {
+      "epoch": 0.93452661328309,
+      "grad_norm": 0.25889289379119873,
+      "learning_rate": 1.3096919978546842e-07,
+      "loss": 0.0125,
+      "step": 2945
+    },
+    {
+      "epoch": 0.934843939807125,
+      "grad_norm": 0.6756712794303894,
+      "learning_rate": 1.297123401512418e-07,
+      "loss": 0.012,
+      "step": 2946
+    },
+    {
+      "epoch": 0.9351612663311599,
+      "grad_norm": 0.2942531704902649,
+      "learning_rate": 1.284614611642665e-07,
+      "loss": 0.0118,
+      "step": 2947
+    },
+    {
+      "epoch": 0.935478592855195,
+      "grad_norm": 0.30766573548316956,
+      "learning_rate": 1.2721656436060382e-07,
+      "loss": 0.012,
+      "step": 2948
+    },
+    {
+      "epoch": 0.93579591937923,
+      "grad_norm": 0.31924325227737427,
+      "learning_rate": 1.25977651268972e-07,
+      "loss": 0.0119,
+      "step": 2949
+    },
+    {
+      "epoch": 0.936113245903265,
+      "grad_norm": 0.3244701623916626,
+      "learning_rate": 1.247447234107385e-07,
+      "loss": 0.0128,
+      "step": 2950
+    },
+    {
+      "epoch": 0.9364305724273,
+      "grad_norm": 0.3868367075920105,
+      "learning_rate": 1.2351778229992228e-07,
+      "loss": 0.0134,
+      "step": 2951
+    },
+    {
+      "epoch": 0.936747898951335,
+      "grad_norm": 0.25746744871139526,
+      "learning_rate": 1.222968294431903e-07,
+      "loss": 0.0118,
+      "step": 2952
+    },
+    {
+      "epoch": 0.93706522547537,
+      "grad_norm": 0.36704742908477783,
+      "learning_rate": 1.2108186633985608e-07,
+      "loss": 0.0109,
+      "step": 2953
+    },
+    {
+      "epoch": 0.937382551999405,
+      "grad_norm": 0.34252506494522095,
+      "learning_rate": 1.1987289448187777e-07,
+      "loss": 0.0127,
+      "step": 2954
+    },
+    {
+      "epoch": 0.9376998785234401,
+      "grad_norm": 0.27779799699783325,
+      "learning_rate": 1.186699153538573e-07,
+      "loss": 0.011,
+      "step": 2955
+    },
+    {
+      "epoch": 0.938017205047475,
+      "grad_norm": 0.3672632575035095,
+      "learning_rate": 1.174729304330352e-07,
+      "loss": 0.0118,
+      "step": 2956
+    },
+    {
+      "epoch": 0.93833453157151,
+      "grad_norm": 0.42128559947013855,
+      "learning_rate": 1.1628194118929403e-07,
+      "loss": 0.0152,
+      "step": 2957
+    },
+    {
+      "epoch": 0.9386518580955451,
+      "grad_norm": 0.34167101979255676,
+      "learning_rate": 1.1509694908515168e-07,
+      "loss": 0.0141,
+      "step": 2958
+    },
+    {
+      "epoch": 0.93896918461958,
+      "grad_norm": 0.33780181407928467,
+      "learning_rate": 1.1391795557576302e-07,
+      "loss": 0.0135,
+      "step": 2959
+    },
+    {
+      "epoch": 0.939286511143615,
+      "grad_norm": 0.35637080669403076,
+      "learning_rate": 1.127449621089155e-07,
+      "loss": 0.0138,
+      "step": 2960
+    },
+    {
+      "epoch": 0.9396038376676501,
+      "grad_norm": 0.36049559712409973,
+      "learning_rate": 1.1157797012502914e-07,
+      "loss": 0.0126,
+      "step": 2961
+    },
+    {
+      "epoch": 0.939921164191685,
+      "grad_norm": 0.7412881851196289,
+      "learning_rate": 1.1041698105715537e-07,
+      "loss": 0.0145,
+      "step": 2962
+    },
+    {
+      "epoch": 0.9402384907157201,
+      "grad_norm": 0.3751037120819092,
+      "learning_rate": 1.0926199633097156e-07,
+      "loss": 0.0145,
+      "step": 2963
+    },
+    {
+      "epoch": 0.940555817239755,
+      "grad_norm": 0.454550176858902,
+      "learning_rate": 1.0811301736478486e-07,
+      "loss": 0.0144,
+      "step": 2964
+    },
+    {
+      "epoch": 0.94087314376379,
+      "grad_norm": 0.26977789402008057,
+      "learning_rate": 1.0697004556952439e-07,
+      "loss": 0.0095,
+      "step": 2965
+    },
+    {
+      "epoch": 0.9411904702878251,
+      "grad_norm": 0.33345726132392883,
+      "learning_rate": 1.0583308234874523e-07,
+      "loss": 0.0146,
+      "step": 2966
+    },
+    {
+      "epoch": 0.94150779681186,
+      "grad_norm": 0.2915416657924652,
+      "learning_rate": 1.0470212909862165e-07,
+      "loss": 0.01,
+      "step": 2967
+    },
+    {
+      "epoch": 0.9418251233358951,
+      "grad_norm": 0.8715503215789795,
+      "learning_rate": 1.035771872079494e-07,
+      "loss": 0.0137,
+      "step": 2968
+    },
+    {
+      "epoch": 0.9421424498599301,
+      "grad_norm": 0.315629780292511,
+      "learning_rate": 1.0245825805814291e-07,
+      "loss": 0.0156,
+      "step": 2969
+    },
+    {
+      "epoch": 0.942459776383965,
+      "grad_norm": 0.4920424222946167,
+      "learning_rate": 1.0134534302323029e-07,
+      "loss": 0.0138,
+      "step": 2970
+    },
+    {
+      "epoch": 0.9427771029080001,
+      "grad_norm": 0.2940930724143982,
+      "learning_rate": 1.0023844346985723e-07,
+      "loss": 0.012,
+      "step": 2971
+    },
+    {
+      "epoch": 0.9430944294320351,
+      "grad_norm": 0.3818264603614807,
+      "learning_rate": 9.913756075728088e-08,
+      "loss": 0.0124,
+      "step": 2972
+    },
+    {
+      "epoch": 0.9434117559560701,
+      "grad_norm": 0.41292238235473633,
+      "learning_rate": 9.804269623737095e-08,
+      "loss": 0.0125,
+      "step": 2973
+    },
+    {
+      "epoch": 0.9437290824801051,
+      "grad_norm": 0.35163575410842896,
+      "learning_rate": 9.695385125460477e-08,
+      "loss": 0.0119,
+      "step": 2974
+    },
+    {
+      "epoch": 0.9440464090041402,
+      "grad_norm": 0.3406285345554352,
+      "learning_rate": 9.587102714607166e-08,
+      "loss": 0.0123,
+      "step": 2975
+    },
+    {
+      "epoch": 0.9443637355281751,
+      "grad_norm": 0.29399093985557556,
+      "learning_rate": 9.479422524146243e-08,
+      "loss": 0.0113,
+      "step": 2976
+    },
+    {
+      "epoch": 0.9446810620522101,
+      "grad_norm": 0.24811449646949768,
+      "learning_rate": 9.372344686307655e-08,
+      "loss": 0.0118,
+      "step": 2977
+    },
+    {
+      "epoch": 0.9449983885762452,
+      "grad_norm": 0.3864191472530365,
+      "learning_rate": 9.265869332581556e-08,
+      "loss": 0.0126,
+      "step": 2978
+    },
+    {
+      "epoch": 0.9453157151002801,
+      "grad_norm": 0.32512059807777405,
+      "learning_rate": 9.159996593718135e-08,
+      "loss": 0.0147,
+      "step": 2979
+    },
+    {
+      "epoch": 0.9456330416243152,
+      "grad_norm": 0.29477831721305847,
+      "learning_rate": 9.054726599727837e-08,
+      "loss": 0.0137,
+      "step": 2980
+    },
+    {
+      "epoch": 0.9459503681483501,
+      "grad_norm": 0.36873120069503784,
+      "learning_rate": 8.950059479880591e-08,
+      "loss": 0.0131,
+      "step": 2981
+    },
+    {
+      "epoch": 0.9462676946723851,
+      "grad_norm": 0.363715797662735,
+      "learning_rate": 8.845995362706361e-08,
+      "loss": 0.0188,
+      "step": 2982
+    },
+    {
+      "epoch": 0.9465850211964202,
+      "grad_norm": 0.3998616933822632,
+      "learning_rate": 8.742534375994371e-08,
+      "loss": 0.0127,
+      "step": 2983
+    },
+    {
+      "epoch": 0.9469023477204551,
+      "grad_norm": 0.28993263840675354,
+      "learning_rate": 8.639676646793382e-08,
+      "loss": 0.0173,
+      "step": 2984
+    },
+    {
+      "epoch": 0.9472196742444902,
+      "grad_norm": 0.25663742423057556,
+      "learning_rate": 8.537422301411246e-08,
+      "loss": 0.0125,
+      "step": 2985
+    },
+    {
+      "epoch": 0.9475370007685252,
+      "grad_norm": 0.3331059515476227,
+      "learning_rate": 8.435771465414911e-08,
+      "loss": 0.0147,
+      "step": 2986
+    },
+    {
+      "epoch": 0.9478543272925601,
+      "grad_norm": 0.38453784584999084,
+      "learning_rate": 8.334724263630301e-08,
+      "loss": 0.0155,
+      "step": 2987
+    },
+    {
+      "epoch": 0.9481716538165952,
+      "grad_norm": 0.3487488329410553,
+      "learning_rate": 8.234280820141938e-08,
+      "loss": 0.0173,
+      "step": 2988
+    },
+    {
+      "epoch": 0.9484889803406302,
+      "grad_norm": 0.34723857045173645,
+      "learning_rate": 8.134441258293046e-08,
+      "loss": 0.0153,
+      "step": 2989
+    },
+    {
+      "epoch": 0.9488063068646652,
+      "grad_norm": 0.2766669690608978,
+      "learning_rate": 8.035205700685167e-08,
+      "loss": 0.0116,
+      "step": 2990
+    },
+    {
+      "epoch": 0.9491236333887002,
+      "grad_norm": 0.26739683747291565,
+      "learning_rate": 7.936574269178376e-08,
+      "loss": 0.011,
+      "step": 2991
+    },
+    {
+      "epoch": 0.9494409599127352,
+      "grad_norm": 0.3693123757839203,
+      "learning_rate": 7.838547084890569e-08,
+      "loss": 0.0124,
+      "step": 2992
+    },
+    {
+      "epoch": 0.9497582864367702,
+      "grad_norm": 0.3679165542125702,
+      "learning_rate": 7.741124268197952e-08,
+      "loss": 0.0146,
+      "step": 2993
+    },
+    {
+      "epoch": 0.9500756129608052,
+      "grad_norm": 1.6905666589736938,
+      "learning_rate": 7.644305938734276e-08,
+      "loss": 0.0139,
+      "step": 2994
+    },
+    {
+      "epoch": 0.9503929394848403,
+      "grad_norm": 0.39203622937202454,
+      "learning_rate": 7.548092215391267e-08,
+      "loss": 0.0135,
+      "step": 2995
+    },
+    {
+      "epoch": 0.9507102660088752,
+      "grad_norm": 0.33710578083992004,
+      "learning_rate": 7.452483216317973e-08,
+      "loss": 0.0115,
+      "step": 2996
+    },
+    {
+      "epoch": 0.9510275925329102,
+      "grad_norm": 0.3041110336780548,
+      "learning_rate": 7.35747905892109e-08,
+      "loss": 0.0116,
+      "step": 2997
+    },
+    {
+      "epoch": 0.9513449190569452,
+      "grad_norm": 0.33398330211639404,
+      "learning_rate": 7.263079859864298e-08,
+      "loss": 0.0158,
+      "step": 2998
+    },
+    {
+      "epoch": 0.9516622455809802,
+      "grad_norm": 0.2809068262577057,
+      "learning_rate": 7.169285735068531e-08,
+      "loss": 0.0131,
+      "step": 2999
+    },
+    {
+      "epoch": 0.9519795721050153,
+      "grad_norm": 0.34716999530792236,
+      "learning_rate": 7.076096799711773e-08,
+      "loss": 0.0154,
+      "step": 3000
+    },
+    {
+      "epoch": 0.9522968986290502,
+      "grad_norm": 0.3130287826061249,
+      "learning_rate": 6.983513168228762e-08,
+      "loss": 0.0121,
+      "step": 3001
+    },
+    {
+      "epoch": 0.9526142251530852,
+      "grad_norm": 0.2709348797798157,
+      "learning_rate": 6.891534954310886e-08,
+      "loss": 0.0119,
+      "step": 3002
+    },
+    {
+      "epoch": 0.9529315516771203,
+      "grad_norm": 0.30951032042503357,
+      "learning_rate": 6.80016227090624e-08,
+      "loss": 0.0123,
+      "step": 3003
+    },
+    {
+      "epoch": 0.9532488782011552,
+      "grad_norm": 0.3161711096763611,
+      "learning_rate": 6.709395230219184e-08,
+      "loss": 0.0143,
+      "step": 3004
+    },
+    {
+      "epoch": 0.9535662047251903,
+      "grad_norm": 0.2815410792827606,
+      "learning_rate": 6.61923394371039e-08,
+      "loss": 0.0101,
+      "step": 3005
+    },
+    {
+      "epoch": 0.9538835312492253,
+      "grad_norm": 0.31943976879119873,
+      "learning_rate": 6.529678522096738e-08,
+      "loss": 0.0134,
+      "step": 3006
+    },
+    {
+      "epoch": 0.9542008577732602,
+      "grad_norm": 0.36457282304763794,
+      "learning_rate": 6.440729075351149e-08,
+      "loss": 0.0133,
+      "step": 3007
+    },
+    {
+      "epoch": 0.9545181842972953,
+      "grad_norm": 0.3278045356273651,
+      "learning_rate": 6.352385712702191e-08,
+      "loss": 0.0123,
+      "step": 3008
+    },
+    {
+      "epoch": 0.9548355108213303,
+      "grad_norm": 0.3335006535053253,
+      "learning_rate": 6.264648542634366e-08,
+      "loss": 0.0136,
+      "step": 3009
+    },
+    {
+      "epoch": 0.9551528373453653,
+      "grad_norm": 0.32260629534721375,
+      "learning_rate": 6.177517672887712e-08,
+      "loss": 0.011,
+      "step": 3010
+    },
+    {
+      "epoch": 0.9554701638694003,
+      "grad_norm": 1.351538896560669,
+      "learning_rate": 6.090993210457807e-08,
+      "loss": 0.0127,
+      "step": 3011
+    },
+    {
+      "epoch": 0.9557874903934354,
+      "grad_norm": 0.2912500202655792,
+      "learning_rate": 6.005075261595495e-08,
+      "loss": 0.0119,
+      "step": 3012
+    },
+    {
+      "epoch": 0.9561048169174703,
+      "grad_norm": 0.36004918813705444,
+      "learning_rate": 5.9197639318068765e-08,
+      "loss": 0.0125,
+      "step": 3013
+    },
+    {
+      "epoch": 0.9564221434415053,
+      "grad_norm": 0.2622070014476776,
+      "learning_rate": 5.835059325853098e-08,
+      "loss": 0.0121,
+      "step": 3014
+    },
+    {
+      "epoch": 0.9567394699655403,
+      "grad_norm": 0.32132047414779663,
+      "learning_rate": 5.7509615477502333e-08,
+      "loss": 0.0134,
+      "step": 3015
+    },
+    {
+      "epoch": 0.9570567964895753,
+      "grad_norm": 0.2795184254646301,
+      "learning_rate": 5.667470700769229e-08,
+      "loss": 0.0115,
+      "step": 3016
+    },
+    {
+      "epoch": 0.9573741230136104,
+      "grad_norm": 0.3140132427215576,
+      "learning_rate": 5.584586887435739e-08,
+      "loss": 0.017,
+      "step": 3017
+    },
+    {
+      "epoch": 0.9576914495376453,
+      "grad_norm": 0.3219519853591919,
+      "learning_rate": 5.50231020952996e-08,
+      "loss": 0.0099,
+      "step": 3018
+    },
+    {
+      "epoch": 0.9580087760616803,
+      "grad_norm": 0.28494930267333984,
+      "learning_rate": 5.42064076808646e-08,
+      "loss": 0.0136,
+      "step": 3019
+    },
+    {
+      "epoch": 0.9583261025857154,
+      "grad_norm": 0.33482465147972107,
+      "learning_rate": 5.339578663394296e-08,
+      "loss": 0.0114,
+      "step": 3020
+    },
+    {
+      "epoch": 0.9586434291097503,
+      "grad_norm": 0.5866627097129822,
+      "learning_rate": 5.259123994996673e-08,
+      "loss": 0.0119,
+      "step": 3021
+    },
+    {
+      "epoch": 0.9589607556337854,
+      "grad_norm": 0.30981114506721497,
+      "learning_rate": 5.179276861690674e-08,
+      "loss": 0.0136,
+      "step": 3022
+    },
+    {
+      "epoch": 0.9592780821578204,
+      "grad_norm": 0.30690962076187134,
+      "learning_rate": 5.100037361527699e-08,
+      "loss": 0.0109,
+      "step": 3023
+    },
+    {
+      "epoch": 0.9595954086818553,
+      "grad_norm": 0.3643493056297302,
+      "learning_rate": 5.021405591812578e-08,
+      "loss": 0.0119,
+      "step": 3024
+    },
+    {
+      "epoch": 0.9599127352058904,
+      "grad_norm": 0.43249207735061646,
+      "learning_rate": 4.943381649104184e-08,
+      "loss": 0.0183,
+      "step": 3025
+    },
+    {
+      "epoch": 0.9602300617299254,
+      "grad_norm": 0.28991538286209106,
+      "learning_rate": 4.865965629214819e-08,
+      "loss": 0.0117,
+      "step": 3026
+    },
+    {
+      "epoch": 0.9605473882539604,
+      "grad_norm": 0.5113929510116577,
+      "learning_rate": 4.7891576272103835e-08,
+      "loss": 0.0146,
+      "step": 3027
+    },
+    {
+      "epoch": 0.9608647147779954,
+      "grad_norm": 0.28454387187957764,
+      "learning_rate": 4.712957737409929e-08,
+      "loss": 0.0105,
+      "step": 3028
+    },
+    {
+      "epoch": 0.9611820413020304,
+      "grad_norm": 0.31223204731941223,
+      "learning_rate": 4.6373660533859945e-08,
+      "loss": 0.0118,
+      "step": 3029
+    },
+    {
+      "epoch": 0.9614993678260654,
+      "grad_norm": 0.29386088252067566,
+      "learning_rate": 4.562382667964105e-08,
+      "loss": 0.0118,
+      "step": 3030
+    },
+    {
+      "epoch": 0.9618166943501004,
+      "grad_norm": 0.24705055356025696,
+      "learning_rate": 4.4880076732227716e-08,
+      "loss": 0.0123,
+      "step": 3031
+    },
+    {
+      "epoch": 0.9621340208741354,
+      "grad_norm": 0.3000153601169586,
+      "learning_rate": 4.41424116049366e-08,
+      "loss": 0.0117,
+      "step": 3032
+    },
+    {
+      "epoch": 0.9624513473981704,
+      "grad_norm": 0.27869173884391785,
+      "learning_rate": 4.3410832203608645e-08,
+      "loss": 0.0116,
+      "step": 3033
+    },
+    {
+      "epoch": 0.9627686739222054,
+      "grad_norm": 0.29608604311943054,
+      "learning_rate": 4.2685339426614126e-08,
+      "loss": 0.0111,
+      "step": 3034
+    },
+    {
+      "epoch": 0.9630860004462404,
+      "grad_norm": 0.27750444412231445,
+      "learning_rate": 4.196593416484873e-08,
+      "loss": 0.0133,
+      "step": 3035
+    },
+    {
+      "epoch": 0.9634033269702754,
+      "grad_norm": 0.5588961243629456,
+      "learning_rate": 4.1252617301732465e-08,
+      "loss": 0.0132,
+      "step": 3036
+    },
+    {
+      "epoch": 0.9637206534943105,
+      "grad_norm": 0.4256257712841034,
+      "learning_rate": 4.054538971320798e-08,
+      "loss": 0.0133,
+      "step": 3037
+    },
+    {
+      "epoch": 0.9640379800183454,
+      "grad_norm": 0.24566379189491272,
+      "learning_rate": 3.9844252267741136e-08,
+      "loss": 0.0117,
+      "step": 3038
+    },
+    {
+      "epoch": 0.9643553065423804,
+      "grad_norm": 0.32990530133247375,
+      "learning_rate": 3.9149205826319866e-08,
+      "loss": 0.0111,
+      "step": 3039
+    },
+    {
+      "epoch": 0.9646726330664155,
+      "grad_norm": 0.34242862462997437,
+      "learning_rate": 3.8460251242451454e-08,
+      "loss": 0.0137,
+      "step": 3040
+    },
+    {
+      "epoch": 0.9649899595904504,
+      "grad_norm": 0.3294157087802887,
+      "learning_rate": 3.777738936216358e-08,
+      "loss": 0.0118,
+      "step": 3041
+    },
+    {
+      "epoch": 0.9653072861144855,
+      "grad_norm": 0.4339422881603241,
+      "learning_rate": 3.7100621023999915e-08,
+      "loss": 0.0152,
+      "step": 3042
+    },
+    {
+      "epoch": 0.9656246126385205,
+      "grad_norm": 0.3988680839538574,
+      "learning_rate": 3.642994705902403e-08,
+      "loss": 0.0134,
+      "step": 3043
+    },
+    {
+      "epoch": 0.9659419391625554,
+      "grad_norm": 0.3493803143501282,
+      "learning_rate": 3.576536829081323e-08,
+      "loss": 0.0142,
+      "step": 3044
+    },
+    {
+      "epoch": 0.9662592656865905,
+      "grad_norm": 0.373218834400177,
+      "learning_rate": 3.510688553546304e-08,
+      "loss": 0.0135,
+      "step": 3045
+    },
+    {
+      "epoch": 0.9665765922106255,
+      "grad_norm": 0.36084824800491333,
+      "learning_rate": 3.4454499601579424e-08,
+      "loss": 0.0138,
+      "step": 3046
+    },
+    {
+      "epoch": 0.9668939187346605,
+      "grad_norm": 0.331142783164978,
+      "learning_rate": 3.3808211290284886e-08,
+      "loss": 0.0116,
+      "step": 3047
+    },
+    {
+      "epoch": 0.9672112452586955,
+      "grad_norm": 0.34245896339416504,
+      "learning_rate": 3.316802139521236e-08,
+      "loss": 0.0132,
+      "step": 3048
+    },
+    {
+      "epoch": 0.9675285717827306,
+      "grad_norm": 0.25367680191993713,
+      "learning_rate": 3.253393070250577e-08,
+      "loss": 0.0109,
+      "step": 3049
+    },
+    {
+      "epoch": 0.9678458983067655,
+      "grad_norm": 0.30187225341796875,
+      "learning_rate": 3.190593999082114e-08,
+      "loss": 0.0153,
+      "step": 3050
+    },
+    {
+      "epoch": 0.9681632248308005,
+      "grad_norm": 0.3006788194179535,
+      "learning_rate": 3.128405003132107e-08,
+      "loss": 0.0125,
+      "step": 3051
+    },
+    {
+      "epoch": 0.9684805513548355,
+      "grad_norm": 0.3008324205875397,
+      "learning_rate": 3.066826158767855e-08,
+      "loss": 0.0143,
+      "step": 3052
+    },
+    {
+      "epoch": 0.9687978778788705,
+      "grad_norm": 1.8125765323638916,
+      "learning_rate": 3.005857541607371e-08,
+      "loss": 0.0108,
+      "step": 3053
+    },
+    {
+      "epoch": 0.9691152044029056,
+      "grad_norm": 0.2715354263782501,
+      "learning_rate": 2.9454992265193216e-08,
+      "loss": 0.0108,
+      "step": 3054
+    },
+    {
+      "epoch": 0.9694325309269405,
+      "grad_norm": 0.32523658871650696,
+      "learning_rate": 2.8857512876228068e-08,
+      "loss": 0.0141,
+      "step": 3055
+    },
+    {
+      "epoch": 0.9697498574509755,
+      "grad_norm": 0.6788278818130493,
+      "learning_rate": 2.8266137982875807e-08,
+      "loss": 0.0119,
+      "step": 3056
+    },
+    {
+      "epoch": 0.9700671839750106,
+      "grad_norm": 0.3457123339176178,
+      "learning_rate": 2.7680868311336094e-08,
+      "loss": 0.0136,
+      "step": 3057
+    },
+    {
+      "epoch": 0.9703845104990455,
+      "grad_norm": 0.302805632352829,
+      "learning_rate": 2.710170458031236e-08,
+      "loss": 0.015,
+      "step": 3058
+    },
+    {
+      "epoch": 0.9707018370230805,
+      "grad_norm": 0.3718216121196747,
+      "learning_rate": 2.6528647501009585e-08,
+      "loss": 0.0145,
+      "step": 3059
+    },
+    {
+      "epoch": 0.9710191635471156,
+      "grad_norm": 0.3198786675930023,
+      "learning_rate": 2.596169777713431e-08,
+      "loss": 0.012,
+      "step": 3060
+    },
+    {
+      "epoch": 0.9713364900711505,
+      "grad_norm": 0.36759257316589355,
+      "learning_rate": 2.5400856104894066e-08,
+      "loss": 0.0141,
+      "step": 3061
+    },
+    {
+      "epoch": 0.9716538165951856,
+      "grad_norm": 0.6025020480155945,
+      "learning_rate": 2.4846123172992953e-08,
+      "loss": 0.0136,
+      "step": 3062
+    },
+    {
+      "epoch": 0.9719711431192206,
+      "grad_norm": 0.3011252284049988,
+      "learning_rate": 2.429749966263717e-08,
+      "loss": 0.0122,
+      "step": 3063
+    },
+    {
+      "epoch": 0.9722884696432555,
+      "grad_norm": 0.26475778222084045,
+      "learning_rate": 2.375498624752781e-08,
+      "loss": 0.0099,
+      "step": 3064
+    },
+    {
+      "epoch": 0.9726057961672906,
+      "grad_norm": 0.28063201904296875,
+      "learning_rate": 2.3218583593864196e-08,
+      "loss": 0.0132,
+      "step": 3065
+    },
+    {
+      "epoch": 0.9729231226913256,
+      "grad_norm": 0.44080743193626404,
+      "learning_rate": 2.26882923603422e-08,
+      "loss": 0.013,
+      "step": 3066
+    },
+    {
+      "epoch": 0.9732404492153606,
+      "grad_norm": 0.32764410972595215,
+      "learning_rate": 2.2164113198152038e-08,
+      "loss": 0.0129,
+      "step": 3067
+    },
+    {
+      "epoch": 0.9735577757393956,
+      "grad_norm": 0.45631566643714905,
+      "learning_rate": 2.1646046750978255e-08,
+      "loss": 0.0123,
+      "step": 3068
+    },
+    {
+      "epoch": 0.9738751022634305,
+      "grad_norm": 0.33998578786849976,
+      "learning_rate": 2.113409365499919e-08,
+      "loss": 0.0127,
+      "step": 3069
+    },
+    {
+      "epoch": 0.9741924287874656,
+      "grad_norm": 0.3630041778087616,
+      "learning_rate": 2.062825453888695e-08,
+      "loss": 0.0133,
+      "step": 3070
+    },
+    {
+      "epoch": 0.9745097553115006,
+      "grad_norm": 0.298919677734375,
+      "learning_rate": 2.012853002380466e-08,
+      "loss": 0.0119,
+      "step": 3071
+    },
+    {
+      "epoch": 0.9748270818355356,
+      "grad_norm": 0.3541850745677948,
+      "learning_rate": 1.9634920723408113e-08,
+      "loss": 0.0138,
+      "step": 3072
+    },
+    {
+      "epoch": 0.9751444083595706,
+      "grad_norm": 0.31958580017089844,
+      "learning_rate": 1.9147427243841333e-08,
+      "loss": 0.0141,
+      "step": 3073
+    },
+    {
+      "epoch": 0.9754617348836057,
+      "grad_norm": 0.33634552359580994,
+      "learning_rate": 1.8666050183741013e-08,
+      "loss": 0.0138,
+      "step": 3074
+    },
+    {
+      "epoch": 0.9757790614076406,
+      "grad_norm": 0.3300630450248718,
+      "learning_rate": 1.8190790134231528e-08,
+      "loss": 0.0109,
+      "step": 3075
+    },
+    {
+      "epoch": 0.9760963879316756,
+      "grad_norm": 0.31325897574424744,
+      "learning_rate": 1.7721647678924924e-08,
+      "loss": 0.0123,
+      "step": 3076
+    },
+    {
+      "epoch": 0.9764137144557107,
+      "grad_norm": 0.3071819245815277,
+      "learning_rate": 1.725862339392259e-08,
+      "loss": 0.0121,
+      "step": 3077
+    },
+    {
+      "epoch": 0.9767310409797456,
+      "grad_norm": 0.2996073067188263,
+      "learning_rate": 1.6801717847811926e-08,
+      "loss": 0.0109,
+      "step": 3078
+    },
+    {
+      "epoch": 0.9770483675037807,
+      "grad_norm": 0.3076895475387573,
+      "learning_rate": 1.6350931601666342e-08,
+      "loss": 0.0114,
+      "step": 3079
+    },
+    {
+      "epoch": 0.9773656940278157,
+      "grad_norm": 0.36274516582489014,
+      "learning_rate": 1.590626520904526e-08,
+      "loss": 0.0122,
+      "step": 3080
+    },
+    {
+      "epoch": 0.9776830205518506,
+      "grad_norm": 0.2968495786190033,
+      "learning_rate": 1.5467719215994103e-08,
+      "loss": 0.0132,
+      "step": 3081
+    },
+    {
+      "epoch": 0.9780003470758857,
+      "grad_norm": 0.24213305115699768,
+      "learning_rate": 1.5035294161039882e-08,
+      "loss": 0.009,
+      "step": 3082
+    },
+    {
+      "epoch": 0.9783176735999207,
+      "grad_norm": 0.24643196165561676,
+      "learning_rate": 1.4608990575195048e-08,
+      "loss": 0.0112,
+      "step": 3083
+    },
+    {
+      "epoch": 0.9786350001239557,
+      "grad_norm": 0.278582364320755,
+      "learning_rate": 1.4188808981955848e-08,
+      "loss": 0.0114,
+      "step": 3084
+    },
+    {
+      "epoch": 0.9789523266479907,
+      "grad_norm": 0.45594415068626404,
+      "learning_rate": 1.377474989729788e-08,
+      "loss": 0.011,
+      "step": 3085
+    },
+    {
+      "epoch": 0.9792696531720256,
+      "grad_norm": 0.34510594606399536,
+      "learning_rate": 1.3366813829681636e-08,
+      "loss": 0.0116,
+      "step": 3086
+    },
+    {
+      "epoch": 0.9795869796960607,
+      "grad_norm": 0.31516706943511963,
+      "learning_rate": 1.2965001280045852e-08,
+      "loss": 0.0159,
+      "step": 3087
+    },
+    {
+      "epoch": 0.9799043062200957,
+      "grad_norm": 0.3654651343822479,
+      "learning_rate": 1.2569312741811945e-08,
+      "loss": 0.011,
+      "step": 3088
+    },
+    {
+      "epoch": 0.9802216327441307,
+      "grad_norm": 0.29559722542762756,
+      "learning_rate": 1.2179748700879013e-08,
+      "loss": 0.0115,
+      "step": 3089
+    },
+    {
+      "epoch": 0.9805389592681657,
+      "grad_norm": 0.3548383116722107,
+      "learning_rate": 1.1796309635626612e-08,
+      "loss": 0.0127,
+      "step": 3090
+    },
+    {
+      "epoch": 0.9808562857922007,
+      "grad_norm": 0.26564937829971313,
+      "learning_rate": 1.1418996016913653e-08,
+      "loss": 0.0119,
+      "step": 3091
+    },
+    {
+      "epoch": 0.9811736123162357,
+      "grad_norm": 0.30672067403793335,
+      "learning_rate": 1.1047808308075059e-08,
+      "loss": 0.0117,
+      "step": 3092
+    },
+    {
+      "epoch": 0.9814909388402707,
+      "grad_norm": 0.37219634652137756,
+      "learning_rate": 1.0682746964925106e-08,
+      "loss": 0.0128,
+      "step": 3093
+    },
+    {
+      "epoch": 0.9818082653643058,
+      "grad_norm": 0.32418644428253174,
+      "learning_rate": 1.0323812435754088e-08,
+      "loss": 0.0122,
+      "step": 3094
+    },
+    {
+      "epoch": 0.9821255918883407,
+      "grad_norm": 0.31112027168273926,
+      "learning_rate": 9.971005161327763e-09,
+      "loss": 0.012,
+      "step": 3095
+    },
+    {
+      "epoch": 0.9824429184123757,
+      "grad_norm": 0.28247568011283875,
+      "learning_rate": 9.624325574890125e-09,
+      "loss": 0.0141,
+      "step": 3096
+    },
+    {
+      "epoch": 0.9827602449364108,
+      "grad_norm": 0.34231138229370117,
+      "learning_rate": 9.283774102157861e-09,
+      "loss": 0.0142,
+      "step": 3097
+    },
+    {
+      "epoch": 0.9830775714604457,
+      "grad_norm": 0.38761696219444275,
+      "learning_rate": 8.949351161324227e-09,
+      "loss": 0.0131,
+      "step": 3098
+    },
+    {
+      "epoch": 0.9833948979844808,
+      "grad_norm": 0.3858702480792999,
+      "learning_rate": 8.621057163055723e-09,
+      "loss": 0.0114,
+      "step": 3099
+    },
+    {
+      "epoch": 0.9837122245085158,
+      "grad_norm": 0.38423028588294983,
+      "learning_rate": 8.298892510493205e-09,
+      "loss": 0.0137,
+      "step": 3100
+    },
+    {
+      "epoch": 0.9840295510325507,
+      "grad_norm": 0.3103446662425995,
+      "learning_rate": 7.982857599250216e-09,
+      "loss": 0.0122,
+      "step": 3101
+    },
+    {
+      "epoch": 0.9843468775565858,
+      "grad_norm": 0.3012070059776306,
+      "learning_rate": 7.672952817413537e-09,
+      "loss": 0.0123,
+      "step": 3102
+    },
+    {
+      "epoch": 0.9846642040806207,
+      "grad_norm": 0.5990753173828125,
+      "learning_rate": 7.369178545542088e-09,
+      "loss": 0.0153,
+      "step": 3103
+    },
+    {
+      "epoch": 0.9849815306046558,
+      "grad_norm": 0.33136898279190063,
+      "learning_rate": 7.071535156666365e-09,
+      "loss": 0.0158,
+      "step": 3104
+    },
+    {
+      "epoch": 0.9852988571286908,
+      "grad_norm": 0.2837555408477783,
+      "learning_rate": 6.7800230162878845e-09,
+      "loss": 0.0122,
+      "step": 3105
+    },
+    {
+      "epoch": 0.9856161836527257,
+      "grad_norm": 0.33877289295196533,
+      "learning_rate": 6.4946424823808526e-09,
+      "loss": 0.0124,
+      "step": 3106
+    },
+    {
+      "epoch": 0.9859335101767608,
+      "grad_norm": 0.28253695368766785,
+      "learning_rate": 6.215393905388278e-09,
+      "loss": 0.0111,
+      "step": 3107
+    },
+    {
+      "epoch": 0.9862508367007958,
+      "grad_norm": 0.3461020588874817,
+      "learning_rate": 5.942277628223081e-09,
+      "loss": 0.0148,
+      "step": 3108
+    },
+    {
+      "epoch": 0.9865681632248308,
+      "grad_norm": 0.3422561585903168,
+      "learning_rate": 5.6752939862686486e-09,
+      "loss": 0.0128,
+      "step": 3109
+    },
+    {
+      "epoch": 0.9868854897488658,
+      "grad_norm": 0.4291917681694031,
+      "learning_rate": 5.414443307377171e-09,
+      "loss": 0.0136,
+      "step": 3110
+    },
+    {
+      "epoch": 0.9872028162729009,
+      "grad_norm": 0.2584308385848999,
+      "learning_rate": 5.159725911870195e-09,
+      "loss": 0.0112,
+      "step": 3111
+    },
+    {
+      "epoch": 0.9875201427969358,
+      "grad_norm": 0.3564316928386688,
+      "learning_rate": 4.911142112536959e-09,
+      "loss": 0.0115,
+      "step": 3112
+    },
+    {
+      "epoch": 0.9878374693209708,
+      "grad_norm": 0.35429832339286804,
+      "learning_rate": 4.6686922146349515e-09,
+      "loss": 0.0126,
+      "step": 3113
+    },
+    {
+      "epoch": 0.9881547958450059,
+      "grad_norm": 0.2901901602745056,
+      "learning_rate": 4.432376515889348e-09,
+      "loss": 0.0114,
+      "step": 3114
+    },
+    {
+      "epoch": 0.9884721223690408,
+      "grad_norm": 0.4784353971481323,
+      "learning_rate": 4.2021953064924624e-09,
+      "loss": 0.0128,
+      "step": 3115
+    },
+    {
+      "epoch": 0.9887894488930759,
+      "grad_norm": 0.35313042998313904,
+      "learning_rate": 3.978148869103748e-09,
+      "loss": 0.0124,
+      "step": 3116
+    },
+    {
+      "epoch": 0.9891067754171109,
+      "grad_norm": 0.29578253626823425,
+      "learning_rate": 3.760237478849793e-09,
+      "loss": 0.0137,
+      "step": 3117
+    },
+    {
+      "epoch": 0.9894241019411458,
+      "grad_norm": 0.33413925766944885,
+      "learning_rate": 3.548461403322101e-09,
+      "loss": 0.0144,
+      "step": 3118
+    },
+    {
+      "epoch": 0.9897414284651809,
+      "grad_norm": 0.3151579797267914,
+      "learning_rate": 3.3428209025793135e-09,
+      "loss": 0.016,
+      "step": 3119
+    },
+    {
+      "epoch": 0.9900587549892158,
+      "grad_norm": 0.34057173132896423,
+      "learning_rate": 3.143316229144433e-09,
+      "loss": 0.0137,
+      "step": 3120
+    },
+    {
+      "epoch": 0.9903760815132508,
+      "grad_norm": 0.256466269493103,
+      "learning_rate": 2.9499476280075987e-09,
+      "loss": 0.0102,
+      "step": 3121
+    },
+    {
+      "epoch": 0.9906934080372859,
+      "grad_norm": 0.9034223556518555,
+      "learning_rate": 2.7627153366222014e-09,
+      "loss": 0.0123,
+      "step": 3122
+    },
+    {
+      "epoch": 0.9910107345613208,
+      "grad_norm": 0.4869183599948883,
+      "learning_rate": 2.5816195849065473e-09,
+      "loss": 0.0121,
+      "step": 3123
+    },
+    {
+      "epoch": 0.9913280610853559,
+      "grad_norm": 0.3689258396625519,
+      "learning_rate": 2.4066605952444144e-09,
+      "loss": 0.0132,
+      "step": 3124
+    },
+    {
+      "epoch": 0.9916453876093909,
+      "grad_norm": 0.38226577639579773,
+      "learning_rate": 2.237838582483387e-09,
+      "loss": 0.015,
+      "step": 3125
+    },
+    {
+      "epoch": 0.9919627141334258,
+      "grad_norm": 0.28858059644699097,
+      "learning_rate": 2.0751537539337453e-09,
+      "loss": 0.0115,
+      "step": 3126
+    },
+    {
+      "epoch": 0.9922800406574609,
+      "grad_norm": 0.3843352794647217,
+      "learning_rate": 1.9186063093706853e-09,
+      "loss": 0.0118,
+      "step": 3127
+    },
+    {
+      "epoch": 0.9925973671814959,
+      "grad_norm": 0.3113584518432617,
+      "learning_rate": 1.7681964410320995e-09,
+      "loss": 0.0132,
+      "step": 3128
+    },
+    {
+      "epoch": 0.9929146937055309,
+      "grad_norm": 0.3980807960033417,
+      "learning_rate": 1.6239243336196863e-09,
+      "loss": 0.016,
+      "step": 3129
+    },
+    {
+      "epoch": 0.9932320202295659,
+      "grad_norm": 0.4269327223300934,
+      "learning_rate": 1.48579016429673e-09,
+      "loss": 0.0156,
+      "step": 3130
+    },
+    {
+      "epoch": 0.993549346753601,
+      "grad_norm": 0.44277694821357727,
+      "learning_rate": 1.3537941026914302e-09,
+      "loss": 0.0115,
+      "step": 3131
+    },
+    {
+      "epoch": 0.9938666732776359,
+      "grad_norm": 0.3191652297973633,
+      "learning_rate": 1.2279363108924635e-09,
+      "loss": 0.0124,
+      "step": 3132
+    },
+    {
+      "epoch": 0.9941839998016709,
+      "grad_norm": 0.3274737298488617,
+      "learning_rate": 1.1082169434517564e-09,
+      "loss": 0.0102,
+      "step": 3133
+    },
+    {
+      "epoch": 0.994501326325706,
+      "grad_norm": 0.42481911182403564,
+      "learning_rate": 9.946361473822664e-10,
+      "loss": 0.0159,
+      "step": 3134
+    },
+    {
+      "epoch": 0.9948186528497409,
+      "grad_norm": 0.2940199077129364,
+      "learning_rate": 8.871940621607567e-10,
+      "loss": 0.013,
+      "step": 3135
+    },
+    {
+      "epoch": 0.995135979373776,
+      "grad_norm": 0.2813743054866791,
+      "learning_rate": 7.858908197239112e-10,
+      "loss": 0.0113,
+      "step": 3136
+    },
+    {
+      "epoch": 0.9954533058978109,
+      "grad_norm": 0.43854883313179016,
+      "learning_rate": 6.907265444716649e-10,
+      "loss": 0.0124,
+      "step": 3137
+    },
+    {
+      "epoch": 0.9957706324218459,
+      "grad_norm": 0.9812107682228088,
+      "learning_rate": 6.017013532627625e-10,
+      "loss": 0.012,
+      "step": 3138
+    },
+    {
+      "epoch": 0.996087958945881,
+      "grad_norm": 0.462831974029541,
+      "learning_rate": 5.188153554208652e-10,
+      "loss": 0.0157,
+      "step": 3139
+    },
+    {
+      "epoch": 0.9964052854699159,
+      "grad_norm": 0.29490187764167786,
+      "learning_rate": 4.4206865272733436e-10,
+      "loss": 0.0137,
+      "step": 3140
+    },
+    {
+      "epoch": 0.996722611993951,
+      "grad_norm": 0.2594425678253174,
+      "learning_rate": 3.714613394267819e-10,
+      "loss": 0.0101,
+      "step": 3141
+    },
+    {
+      "epoch": 0.997039938517986,
+      "grad_norm": 0.3489019274711609,
+      "learning_rate": 3.0699350222429536e-10,
+      "loss": 0.0123,
+      "step": 3142
+    },
+    {
+      "epoch": 0.9973572650420209,
+      "grad_norm": 0.30260008573532104,
+      "learning_rate": 2.486652202848827e-10,
+      "loss": 0.0137,
+      "step": 3143
+    },
+    {
+      "epoch": 0.997674591566056,
+      "grad_norm": 0.3811684548854828,
+      "learning_rate": 1.9647656523513747e-10,
+      "loss": 0.0124,
+      "step": 3144
+    },
+    {
+      "epoch": 0.997991918090091,
+      "grad_norm": 0.364578515291214,
+      "learning_rate": 1.504276011621286e-10,
+      "loss": 0.0136,
+      "step": 3145
+    },
+    {
+      "epoch": 0.998309244614126,
+      "grad_norm": 0.3206532299518585,
+      "learning_rate": 1.1051838461340059e-10,
+      "loss": 0.0116,
+      "step": 3146
+    },
+    {
+      "epoch": 0.998626571138161,
+      "grad_norm": 0.3003668487071991,
+      "learning_rate": 7.674896459641812e-11,
+      "loss": 0.0132,
+      "step": 3147
+    },
+    {
+      "epoch": 0.998943897662196,
+      "grad_norm": 0.4509921669960022,
+      "learning_rate": 4.911938258023164e-11,
+      "loss": 0.0111,
+      "step": 3148
+    },
+    {
+      "epoch": 0.999261224186231,
+      "grad_norm": 0.3066161572933197,
+      "learning_rate": 2.7629672493256766e-11,
+      "loss": 0.0124,
+      "step": 3149
+    },
+    {
+      "epoch": 0.999578550710266,
+      "grad_norm": 0.2729567289352417,
+      "learning_rate": 1.2279860724384585e-11,
+      "loss": 0.0115,
+      "step": 3150
+    },
+    {
+      "epoch": 0.9998958772343011,
+      "grad_norm": 0.3247279226779938,
+      "learning_rate": 3.069966123536716e-12,
+      "loss": 0.013,
+      "step": 3151
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3151,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.465803263427425e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}