diff --git "a/checkpoint-1620/trainer_state.json" "b/checkpoint-1620/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1620/trainer_state.json"
@@ -0,0 +1,11401 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9996914532551682,
+  "eval_steps": 405,
+  "global_step": 1620,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.7734472751617432,
+      "learning_rate": 2e-05,
+      "loss": 1.2057,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "eval_loss": 2.256255865097046,
+      "eval_runtime": 187.1236,
+      "eval_samples_per_second": 146.839,
+      "eval_steps_per_second": 73.422,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.8534036874771118,
+      "learning_rate": 4e-05,
+      "loss": 1.1774,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.32288506627082825,
+      "learning_rate": 6e-05,
+      "loss": 1.2129,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.2195775806903839,
+      "learning_rate": 8e-05,
+      "loss": 1.1477,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18129795789718628,
+      "learning_rate": 0.0001,
+      "loss": 1.1727,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.14628687500953674,
+      "learning_rate": 0.00012,
+      "loss": 1.1005,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.18358688056468964,
+      "learning_rate": 0.00014,
+      "loss": 1.16,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 0.22286616265773773,
+      "learning_rate": 0.00016,
+      "loss": 1.113,
+      "step": 8
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.23289859294891357,
+      "learning_rate": 0.00018,
+      "loss": 1.1599,
+      "step": 9
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1709630936384201,
+      "learning_rate": 0.0002,
+      "loss": 1.1172,
+      "step": 10
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14665727317333221,
+      "learning_rate": 0.0001999998096215179,
+      "loss": 1.1128,
+      "step": 11
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12764032185077667,
+      "learning_rate": 0.00019999923848679644,
+      "loss": 1.1029,
+      "step": 12
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1657416969537735,
+      "learning_rate": 0.0001999982865980103,
+      "loss": 1.1231,
+      "step": 13
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14713911712169647,
+      "learning_rate": 0.00019999695395878382,
+      "loss": 1.071,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.15686815977096558,
+      "learning_rate": 0.00019999524057419117,
+      "loss": 1.0337,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12417972832918167,
+      "learning_rate": 0.0001999931464507561,
+      "loss": 1.0527,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12729552388191223,
+      "learning_rate": 0.0001999906715964522,
+      "loss": 1.0415,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13774332404136658,
+      "learning_rate": 0.00019998781602070264,
+      "loss": 1.1042,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14067314565181732,
+      "learning_rate": 0.0001999845797343802,
+      "loss": 1.1312,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.14144887030124664,
+      "learning_rate": 0.00019998096274980728,
+      "loss": 1.1298,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.1327802687883377,
+      "learning_rate": 0.00019997696508075577,
+      "loss": 1.0305,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.12503667175769806,
+      "learning_rate": 0.00019997258674244717,
+      "loss": 1.0402,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.136058509349823,
+      "learning_rate": 0.00019996782775155218,
+      "loss": 1.0485,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.13043010234832764,
+      "learning_rate": 0.00019996268812619107,
+      "loss": 1.0385,
+      "step": 24
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13245601952075958,
+      "learning_rate": 0.0001999571678859333,
+      "loss": 1.0182,
+      "step": 25
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.14026328921318054,
+      "learning_rate": 0.00019995126705179756,
+      "loss": 1.0793,
+      "step": 26
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12143687158823013,
+      "learning_rate": 0.00019994498564625173,
+      "loss": 1.0859,
+      "step": 27
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12939739227294922,
+      "learning_rate": 0.00019993832369321262,
+      "loss": 1.0857,
+      "step": 28
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13331376016139984,
+      "learning_rate": 0.00019993128121804614,
+      "loss": 1.0537,
+      "step": 29
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1249675303697586,
+      "learning_rate": 0.000199923858247567,
+      "loss": 1.0343,
+      "step": 30
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12867054343223572,
+      "learning_rate": 0.00019991605481003866,
+      "loss": 1.0118,
+      "step": 31
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13537080585956573,
+      "learning_rate": 0.00019990787093517325,
+      "loss": 1.0986,
+      "step": 32
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.13139110803604126,
+      "learning_rate": 0.00019989930665413147,
+      "loss": 1.1083,
+      "step": 33
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12539294362068176,
+      "learning_rate": 0.0001998903619995224,
+      "loss": 1.0161,
+      "step": 34
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1322510689496994,
+      "learning_rate": 0.00019988103700540344,
+      "loss": 1.0176,
+      "step": 35
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1400848776102066,
+      "learning_rate": 0.00019987133170728012,
+      "loss": 1.0538,
+      "step": 36
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1217857077717781,
+      "learning_rate": 0.0001998612461421061,
+      "loss": 1.0108,
+      "step": 37
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.1294267624616623,
+      "learning_rate": 0.0001998507803482828,
+      "loss": 1.0104,
+      "step": 38
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12792208790779114,
+      "learning_rate": 0.00019983993436565952,
+      "loss": 1.0303,
+      "step": 39
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.12307696044445038,
+      "learning_rate": 0.00019982870823553308,
+      "loss": 0.9574,
+      "step": 40
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13150550425052643,
+      "learning_rate": 0.00019981710200064773,
+      "loss": 1.0474,
+      "step": 41
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14002776145935059,
+      "learning_rate": 0.00019980511570519505,
+      "loss": 1.0499,
+      "step": 42
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.14099255204200745,
+      "learning_rate": 0.00019979274939481366,
+      "loss": 1.0715,
+      "step": 43
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1360781192779541,
+      "learning_rate": 0.00019978000311658917,
+      "loss": 1.0751,
+      "step": 44
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13030701875686646,
+      "learning_rate": 0.00019976687691905393,
+      "loss": 1.0228,
+      "step": 45
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1272677332162857,
+      "learning_rate": 0.00019975337085218682,
+      "loss": 1.0052,
+      "step": 46
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13296392560005188,
+      "learning_rate": 0.00019973948496741314,
+      "loss": 1.0381,
+      "step": 47
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1351320445537567,
+      "learning_rate": 0.0001997252193176044,
+      "loss": 1.0436,
+      "step": 48
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1394244134426117,
+      "learning_rate": 0.00019971057395707802,
+      "loss": 1.0773,
+      "step": 49
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.17834258079528809,
+      "learning_rate": 0.00019969554894159723,
+      "loss": 1.0194,
+      "step": 50
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12855498492717743,
+      "learning_rate": 0.00019968014432837083,
+      "loss": 1.0266,
+      "step": 51
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.13544076681137085,
+      "learning_rate": 0.00019966436017605297,
+      "loss": 1.0922,
+      "step": 52
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.1299847811460495,
+      "learning_rate": 0.0001996481965447429,
+      "loss": 1.0319,
+      "step": 53
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12845264375209808,
+      "learning_rate": 0.00019963165349598473,
+      "loss": 1.0398,
+      "step": 54
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12594841420650482,
+      "learning_rate": 0.00019961473109276733,
+      "loss": 1.0369,
+      "step": 55
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.12297705560922623,
+      "learning_rate": 0.00019959742939952392,
+      "loss": 1.0288,
+      "step": 56
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13264279067516327,
+      "learning_rate": 0.00019957974848213187,
+      "loss": 1.0861,
+      "step": 57
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1319752037525177,
+      "learning_rate": 0.00019956168840791255,
+      "loss": 1.0224,
+      "step": 58
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13023924827575684,
+      "learning_rate": 0.00019954324924563089,
+      "loss": 0.9957,
+      "step": 59
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12949976325035095,
+      "learning_rate": 0.00019952443106549533,
+      "loss": 1.0545,
+      "step": 60
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1323748528957367,
+      "learning_rate": 0.00019950523393915737,
+      "loss": 0.9907,
+      "step": 61
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13179315626621246,
+      "learning_rate": 0.00019948565793971148,
+      "loss": 1.0081,
+      "step": 62
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13959741592407227,
+      "learning_rate": 0.00019946570314169457,
+      "loss": 1.0009,
+      "step": 63
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13472723960876465,
+      "learning_rate": 0.00019944536962108593,
+      "loss": 1.0322,
+      "step": 64
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13210035860538483,
+      "learning_rate": 0.00019942465745530685,
+      "loss": 1.005,
+      "step": 65
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1369844377040863,
+      "learning_rate": 0.00019940356672322037,
+      "loss": 1.0521,
+      "step": 66
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14004011452198029,
+      "learning_rate": 0.0001993820975051309,
+      "loss": 0.9591,
+      "step": 67
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.12669473886489868,
+      "learning_rate": 0.00019936024988278402,
+      "loss": 0.9264,
+      "step": 68
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.1408959925174713,
+      "learning_rate": 0.00019933802393936603,
+      "loss": 1.0441,
+      "step": 69
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.14294607937335968,
+      "learning_rate": 0.00019931541975950378,
+      "loss": 0.9579,
+      "step": 70
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.13492931425571442,
+      "learning_rate": 0.00019929243742926425,
+      "loss": 0.9927,
+      "step": 71
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.128339946269989,
+      "learning_rate": 0.00019926907703615428,
+      "loss": 0.9802,
+      "step": 72
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13415645062923431,
+      "learning_rate": 0.00019924533866912017,
+      "loss": 1.0173,
+      "step": 73
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14114364981651306,
+      "learning_rate": 0.00019922122241854742,
+      "loss": 1.0295,
+      "step": 74
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1372201293706894,
+      "learning_rate": 0.0001991967283762603,
+      "loss": 1.0038,
+      "step": 75
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13632307946681976,
+      "learning_rate": 0.00019917185663552165,
+      "loss": 1.0085,
+      "step": 76
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13486582040786743,
+      "learning_rate": 0.0001991466072910323,
+      "loss": 1.0091,
+      "step": 77
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13797307014465332,
+      "learning_rate": 0.0001991209804389309,
+      "loss": 1.0357,
+      "step": 78
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1424788385629654,
+      "learning_rate": 0.00019909497617679348,
+      "loss": 1.0587,
+      "step": 79
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14115746319293976,
+      "learning_rate": 0.00019906859460363307,
+      "loss": 0.9996,
+      "step": 80
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13875117897987366,
+      "learning_rate": 0.00019904183581989936,
+      "loss": 1.0623,
+      "step": 81
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13902142643928528,
+      "learning_rate": 0.0001990146999274783,
+      "loss": 0.9763,
+      "step": 82
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.15071237087249756,
+      "learning_rate": 0.00019898718702969164,
+      "loss": 1.0864,
+      "step": 83
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13857468962669373,
+      "learning_rate": 0.0001989592972312967,
+      "loss": 1.0042,
+      "step": 84
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14365358650684357,
+      "learning_rate": 0.0001989310306384858,
+      "loss": 1.0037,
+      "step": 85
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.13876394927501678,
+      "learning_rate": 0.00019890238735888595,
+      "loss": 1.0278,
+      "step": 86
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14259740710258484,
+      "learning_rate": 0.0001988733675015585,
+      "loss": 1.0158,
+      "step": 87
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.14382067322731018,
+      "learning_rate": 0.0001988439711769985,
+      "loss": 1.068,
+      "step": 88
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.1407095491886139,
+      "learning_rate": 0.0001988141984971345,
+      "loss": 1.0579,
+      "step": 89
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14294175803661346,
+      "learning_rate": 0.00019878404957532814,
+      "loss": 0.936,
+      "step": 90
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.12836596369743347,
+      "learning_rate": 0.0001987535245263735,
+      "loss": 0.9527,
+      "step": 91
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14682206511497498,
+      "learning_rate": 0.00019872262346649677,
+      "loss": 1.0698,
+      "step": 92
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.142591193318367,
+      "learning_rate": 0.00019869134651335596,
+      "loss": 0.9883,
+      "step": 93
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1410866230726242,
+      "learning_rate": 0.0001986596937860402,
+      "loss": 0.9708,
+      "step": 94
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14686676859855652,
+      "learning_rate": 0.0001986276654050695,
+      "loss": 0.9976,
+      "step": 95
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14652645587921143,
+      "learning_rate": 0.0001985952614923941,
+      "loss": 0.9715,
+      "step": 96
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.1395326554775238,
+      "learning_rate": 0.0001985624821713942,
+      "loss": 1.0251,
+      "step": 97
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14578589797019958,
+      "learning_rate": 0.0001985293275668793,
+      "loss": 0.9243,
+      "step": 98
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13804090023040771,
+      "learning_rate": 0.00019849579780508792,
+      "loss": 1.0126,
+      "step": 99
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13700765371322632,
+      "learning_rate": 0.0001984618930136869,
+      "loss": 0.9827,
+      "step": 100
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14506202936172485,
+      "learning_rate": 0.00019842761332177115,
+      "loss": 1.055,
+      "step": 101
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.15084940195083618,
+      "learning_rate": 0.00019839295885986296,
+      "loss": 1.0056,
+      "step": 102
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.13431130349636078,
+      "learning_rate": 0.00019835792975991164,
+      "loss": 0.9394,
+      "step": 103
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14309975504875183,
+      "learning_rate": 0.00019832252615529288,
+      "loss": 1.0143,
+      "step": 104
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.14179019629955292,
+      "learning_rate": 0.00019828674818080837,
+      "loss": 0.9854,
+      "step": 105
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.13689136505126953,
+      "learning_rate": 0.00019825059597268526,
+      "loss": 0.9607,
+      "step": 106
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14776825904846191,
+      "learning_rate": 0.00019821406966857562,
+      "loss": 0.9413,
+      "step": 107
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14173467457294464,
+      "learning_rate": 0.00019817716940755586,
+      "loss": 0.9668,
+      "step": 108
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1513083279132843,
+      "learning_rate": 0.00019813989533012634,
+      "loss": 1.0297,
+      "step": 109
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1363351196050644,
+      "learning_rate": 0.00019810224757821064,
+      "loss": 0.979,
+      "step": 110
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15797564387321472,
+      "learning_rate": 0.00019806422629515524,
+      "loss": 1.0038,
+      "step": 111
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1440480500459671,
+      "learning_rate": 0.00019802583162572881,
+      "loss": 1.0032,
+      "step": 112
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1352243721485138,
+      "learning_rate": 0.00019798706371612173,
+      "loss": 1.008,
+      "step": 113
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.140641450881958,
+      "learning_rate": 0.00019794792271394553,
+      "loss": 0.9439,
+      "step": 114
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1501038521528244,
+      "learning_rate": 0.00019790840876823232,
+      "loss": 0.9764,
+      "step": 115
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1435052454471588,
+      "learning_rate": 0.0001978685220294341,
+      "loss": 1.0453,
+      "step": 116
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1418715864419937,
+      "learning_rate": 0.00019782826264942252,
+      "loss": 0.9875,
+      "step": 117
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15290676057338715,
+      "learning_rate": 0.0001977876307814879,
+      "loss": 0.98,
+      "step": 118
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.1427755206823349,
+      "learning_rate": 0.00019774662658033897,
+      "loss": 0.9779,
+      "step": 119
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.15220731496810913,
+      "learning_rate": 0.00019770525020210204,
+      "loss": 1.002,
+      "step": 120
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.14343520998954773,
+      "learning_rate": 0.00019766350180432056,
+      "loss": 0.9262,
+      "step": 121
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1511845737695694,
+      "learning_rate": 0.00019762138154595446,
+      "loss": 1.0116,
+      "step": 122
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15124031901359558,
+      "learning_rate": 0.00019757888958737957,
+      "loss": 1.0394,
+      "step": 123
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13628233969211578,
+      "learning_rate": 0.00019753602609038696,
+      "loss": 0.9627,
+      "step": 124
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14199230074882507,
+      "learning_rate": 0.00019749279121818235,
+      "loss": 0.9886,
+      "step": 125
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14725786447525024,
+      "learning_rate": 0.0001974491851353856,
+      "loss": 1.0272,
+      "step": 126
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14161822199821472,
+      "learning_rate": 0.00019740520800802988,
+      "loss": 0.9817,
+      "step": 127
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.13846537470817566,
+      "learning_rate": 0.00019736086000356114,
+      "loss": 0.9728,
+      "step": 128
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1412704735994339,
+      "learning_rate": 0.00019731614129083754,
+      "loss": 1.0123,
+      "step": 129
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.15130552649497986,
+      "learning_rate": 0.0001972710520401287,
+      "loss": 0.9652,
+      "step": 130
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.16430234909057617,
+      "learning_rate": 0.00019722559242311497,
+      "loss": 0.9892,
+      "step": 131
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1458190530538559,
+      "learning_rate": 0.0001971797626128871,
+      "loss": 1.0007,
+      "step": 132
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14452631771564484,
+      "learning_rate": 0.0001971335627839453,
+      "loss": 1.0166,
+      "step": 133
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14020834863185883,
+      "learning_rate": 0.00019708699311219861,
+      "loss": 1.0398,
+      "step": 134
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.1409185528755188,
+      "learning_rate": 0.0001970400537749643,
+      "loss": 0.9999,
+      "step": 135
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14768308401107788,
+      "learning_rate": 0.00019699274495096712,
+      "loss": 1.0088,
+      "step": 136
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.14993049204349518,
+      "learning_rate": 0.0001969450668203388,
+      "loss": 1.0369,
+      "step": 137
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1372499018907547,
+      "learning_rate": 0.00019689701956461707,
+      "loss": 0.9464,
+      "step": 138
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13801540434360504,
+      "learning_rate": 0.00019684860336674525,
+      "loss": 0.9098,
+      "step": 139
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14262749254703522,
+      "learning_rate": 0.0001967998184110713,
+      "loss": 0.9944,
+      "step": 140
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14317511022090912,
+      "learning_rate": 0.00019675066488334747,
+      "loss": 0.9878,
+      "step": 141
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15064817667007446,
+      "learning_rate": 0.00019670114297072914,
+      "loss": 0.9301,
+      "step": 142
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1389160007238388,
+      "learning_rate": 0.00019665125286177449,
+      "loss": 0.9237,
+      "step": 143
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.13915666937828064,
+      "learning_rate": 0.00019660099474644356,
+      "loss": 0.9225,
+      "step": 144
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14348295331001282,
+      "learning_rate": 0.0001965503688160976,
+      "loss": 0.9389,
+      "step": 145
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15439380705356598,
+      "learning_rate": 0.0001964993752634984,
+      "loss": 0.9823,
+      "step": 146
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1442396342754364,
+      "learning_rate": 0.0001964480142828075,
+      "loss": 1.0092,
+      "step": 147
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1483033150434494,
+      "learning_rate": 0.00019639628606958533,
+      "loss": 0.986,
+      "step": 148
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14599324762821198,
+      "learning_rate": 0.00019634419082079072,
+      "loss": 0.9266,
+      "step": 149
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15389233827590942,
+      "learning_rate": 0.00019629172873477995,
+      "loss": 0.9828,
+      "step": 150
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.14817312359809875,
+      "learning_rate": 0.00019623890001130602,
+      "loss": 1.042,
+      "step": 151
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.1483391374349594,
+      "learning_rate": 0.00019618570485151806,
+      "loss": 0.9176,
+      "step": 152
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.15675631165504456,
+      "learning_rate": 0.00019613214345796026,
+      "loss": 0.9743,
+      "step": 153
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15009979903697968,
+      "learning_rate": 0.0001960782160345714,
+      "loss": 1.007,
+      "step": 154
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1458912342786789,
+      "learning_rate": 0.0001960239227866839,
+      "loss": 0.9769,
+      "step": 155
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15029476583003998,
+      "learning_rate": 0.00019596926392102305,
+      "loss": 1.0033,
+      "step": 156
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14372846484184265,
+      "learning_rate": 0.00019591423964570632,
+      "loss": 0.978,
+      "step": 157
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1353691667318344,
+      "learning_rate": 0.00019585885017024248,
+      "loss": 0.8577,
+      "step": 158
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1537759006023407,
+      "learning_rate": 0.0001958030957055308,
+      "loss": 1.0079,
+      "step": 159
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14179471135139465,
+      "learning_rate": 0.00019574697646386027,
+      "loss": 0.9666,
+      "step": 160
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14804232120513916,
+      "learning_rate": 0.00019569049265890883,
+      "loss": 0.9901,
+      "step": 161
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.1432216763496399,
+      "learning_rate": 0.00019563364450574252,
+      "loss": 0.9493,
+      "step": 162
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14438655972480774,
+      "learning_rate": 0.0001955764322208146,
+      "loss": 1.0082,
+      "step": 163
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.14895915985107422,
+      "learning_rate": 0.0001955188560219648,
+      "loss": 0.9489,
+      "step": 164
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15027450025081635,
+      "learning_rate": 0.00019546091612841863,
+      "loss": 0.9324,
+      "step": 165
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15637291967868805,
+      "learning_rate": 0.00019540261276078616,
+      "loss": 0.9542,
+      "step": 166
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15349559485912323,
+      "learning_rate": 0.00019534394614106155,
+      "loss": 0.9551,
+      "step": 167
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15433311462402344,
+      "learning_rate": 0.00019528491649262208,
+      "loss": 0.9133,
+      "step": 168
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15646134316921234,
+      "learning_rate": 0.0001952255240402272,
+      "loss": 0.9334,
+      "step": 169
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.15300197899341583,
+      "learning_rate": 0.0001951657690100178,
+      "loss": 0.9762,
+      "step": 170
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15511955320835114,
+      "learning_rate": 0.00019510565162951537,
+      "loss": 0.9277,
+      "step": 171
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14775170385837555,
+      "learning_rate": 0.000195045172127621,
+      "loss": 0.8589,
+      "step": 172
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15104347467422485,
+      "learning_rate": 0.00019498433073461457,
+      "loss": 1.0288,
+      "step": 173
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15289169549942017,
+      "learning_rate": 0.00019492312768215397,
+      "loss": 0.9827,
+      "step": 174
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1526726484298706,
+      "learning_rate": 0.00019486156320327406,
+      "loss": 0.9362,
+      "step": 175
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1503329575061798,
+      "learning_rate": 0.0001947996375323859,
+      "loss": 0.986,
+      "step": 176
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14389175176620483,
+      "learning_rate": 0.0001947373509052758,
+      "loss": 0.9245,
+      "step": 177
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15165729820728302,
+      "learning_rate": 0.00019467470355910438,
+      "loss": 0.9635,
+      "step": 178
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15394453704357147,
+      "learning_rate": 0.00019461169573240582,
+      "loss": 0.9871,
+      "step": 179
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15355828404426575,
+      "learning_rate": 0.0001945483276650868,
+      "loss": 1.0069,
+      "step": 180
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.1516997516155243,
+      "learning_rate": 0.00019448459959842566,
+      "loss": 0.9297,
+      "step": 181
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15058712661266327,
+      "learning_rate": 0.0001944205117750714,
+      "loss": 1.0105,
+      "step": 182
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15325979888439178,
+      "learning_rate": 0.0001943560644390429,
+      "loss": 0.9716,
+      "step": 183
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.16299349069595337,
+      "learning_rate": 0.00019429125783572792,
+      "loss": 0.9757,
+      "step": 184
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.15402033925056458,
+      "learning_rate": 0.00019422609221188207,
+      "loss": 1.0291,
+      "step": 185
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.14746250212192535,
+      "learning_rate": 0.000194160567815628,
+      "loss": 0.9287,
+      "step": 186
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15267391502857208,
+      "learning_rate": 0.00019409468489645443,
+      "loss": 0.941,
+      "step": 187
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15115174651145935,
+      "learning_rate": 0.00019402844370521515,
+      "loss": 0.9743,
+      "step": 188
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1529087871313095,
+      "learning_rate": 0.00019396184449412815,
+      "loss": 0.9589,
+      "step": 189
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15190894901752472,
+      "learning_rate": 0.0001938948875167745,
+      "loss": 0.9811,
+      "step": 190
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.13938122987747192,
+      "learning_rate": 0.00019382757302809761,
+      "loss": 0.9324,
+      "step": 191
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16739986836910248,
+      "learning_rate": 0.00019375990128440204,
+      "loss": 0.9853,
+      "step": 192
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16093076765537262,
+      "learning_rate": 0.00019369187254335271,
+      "loss": 1.0001,
+      "step": 193
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14566665887832642,
+      "learning_rate": 0.00019362348706397373,
+      "loss": 0.9341,
+      "step": 194
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1497393101453781,
+      "learning_rate": 0.00019355474510664763,
+      "loss": 0.9934,
+      "step": 195
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1498127579689026,
+      "learning_rate": 0.00019348564693311417,
+      "loss": 0.9461,
+      "step": 196
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1455003172159195,
+      "learning_rate": 0.00019341619280646949,
+      "loss": 0.9603,
+      "step": 197
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.1546657532453537,
+      "learning_rate": 0.00019334638299116496,
+      "loss": 0.9494,
+      "step": 198
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.15734489262104034,
+      "learning_rate": 0.00019327621775300637,
+      "loss": 1.0197,
+      "step": 199
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14751465618610382,
+      "learning_rate": 0.00019320569735915271,
+      "loss": 0.9291,
+      "step": 200
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.14648941159248352,
+      "learning_rate": 0.00019313482207811532,
+      "loss": 0.9234,
+      "step": 201
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.16170018911361694,
+      "learning_rate": 0.00019306359217975677,
+      "loss": 1.0084,
+      "step": 202
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15328988432884216,
+      "learning_rate": 0.00019299200793528983,
+      "loss": 0.9446,
+      "step": 203
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15634268522262573,
+      "learning_rate": 0.0001929200696172765,
+      "loss": 0.9361,
+      "step": 204
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1598619818687439,
+      "learning_rate": 0.00019284777749962694,
+      "loss": 0.9569,
+      "step": 205
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15576934814453125,
+      "learning_rate": 0.00019277513185759844,
+      "loss": 0.9445,
+      "step": 206
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1431022435426712,
+      "learning_rate": 0.00019270213296779435,
+      "loss": 0.9186,
+      "step": 207
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.14549916982650757,
+      "learning_rate": 0.00019262878110816299,
+      "loss": 0.9247,
+      "step": 208
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1463356912136078,
+      "learning_rate": 0.0001925550765579967,
+      "loss": 0.9582,
+      "step": 209
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15328474342823029,
+      "learning_rate": 0.00019248101959793066,
+      "loss": 0.9533,
+      "step": 210
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15553633868694305,
+      "learning_rate": 0.00019240661050994194,
+      "loss": 1.0229,
+      "step": 211
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15532618761062622,
+      "learning_rate": 0.0001923318495773483,
+      "loss": 0.8567,
+      "step": 212
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15063565969467163,
+      "learning_rate": 0.00019225673708480717,
+      "loss": 0.9495,
+      "step": 213
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.16670912504196167,
+      "learning_rate": 0.00019218127331831469,
+      "loss": 0.9658,
+      "step": 214
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15556268393993378,
+      "learning_rate": 0.0001921054585652043,
+      "loss": 0.9402,
+      "step": 215
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.156576007604599,
+      "learning_rate": 0.00019202929311414602,
+      "loss": 0.982,
+      "step": 216
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.1444980800151825,
+      "learning_rate": 0.0001919527772551451,
+      "loss": 0.9305,
+      "step": 217
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.15206484496593475,
+      "learning_rate": 0.00019187591127954094,
+      "loss": 0.9156,
+      "step": 218
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15326672792434692,
+      "learning_rate": 0.00019179869548000618,
+      "loss": 0.9304,
+      "step": 219
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15003164112567902,
+      "learning_rate": 0.00019172113015054532,
+      "loss": 0.9729,
+      "step": 220
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15101422369480133,
+      "learning_rate": 0.00019164321558649375,
+      "loss": 0.8856,
+      "step": 221
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16672931611537933,
+      "learning_rate": 0.00019156495208451658,
+      "loss": 0.946,
+      "step": 222
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14862582087516785,
+      "learning_rate": 0.00019148633994260758,
+      "loss": 0.9108,
+      "step": 223
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1526026725769043,
+      "learning_rate": 0.00019140737946008796,
+      "loss": 0.9241,
+      "step": 224
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1648893803358078,
+      "learning_rate": 0.0001913280709376052,
+      "loss": 0.9751,
+      "step": 225
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15331527590751648,
+      "learning_rate": 0.0001912484146771321,
+      "loss": 0.9514,
+      "step": 226
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1558353304862976,
+      "learning_rate": 0.00019116841098196536,
+      "loss": 0.9946,
+      "step": 227
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14292731881141663,
+      "learning_rate": 0.00019108806015672464,
+      "loss": 0.9364,
+      "step": 228
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.14759871363639832,
+      "learning_rate": 0.00019100736250735127,
+      "loss": 0.9242,
+      "step": 229
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16730011999607086,
+      "learning_rate": 0.00019092631834110723,
+      "loss": 0.9632,
+      "step": 230
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.1640511155128479,
+      "learning_rate": 0.00019084492796657383,
+      "loss": 1.0195,
+      "step": 231
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15260474383831024,
+      "learning_rate": 0.00019076319169365053,
+      "loss": 0.9493,
+      "step": 232
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.15863056480884552,
+      "learning_rate": 0.00019068110983355393,
+      "loss": 0.935,
+      "step": 233
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.16079328954219818,
+      "learning_rate": 0.0001905986826988164,
+      "loss": 1.0088,
+      "step": 234
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14830029010772705,
+      "learning_rate": 0.00019051591060328496,
+      "loss": 0.9266,
+      "step": 235
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1541360318660736,
+      "learning_rate": 0.0001904327938621202,
+      "loss": 0.9457,
+      "step": 236
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1579032987356186,
+      "learning_rate": 0.0001903493327917949,
+      "loss": 1.0189,
+      "step": 237
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1537141352891922,
+      "learning_rate": 0.00019026552771009284,
+      "loss": 0.9036,
+      "step": 238
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15884777903556824,
+      "learning_rate": 0.00019018137893610774,
+      "loss": 0.9764,
+      "step": 239
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1545841544866562,
+      "learning_rate": 0.0001900968867902419,
+      "loss": 0.9426,
+      "step": 240
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15149977803230286,
+      "learning_rate": 0.00019001205159420513,
+      "loss": 0.9705,
+      "step": 241
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15349112451076508,
+      "learning_rate": 0.00018992687367101323,
+      "loss": 0.9414,
+      "step": 242
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15908096730709076,
+      "learning_rate": 0.0001898413533449871,
+      "loss": 0.9737,
+      "step": 243
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15833136439323425,
+      "learning_rate": 0.0001897554909417514,
+      "loss": 0.9684,
+      "step": 244
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14696678519248962,
+      "learning_rate": 0.00018966928678823317,
+      "loss": 0.872,
+      "step": 245
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14649872481822968,
+      "learning_rate": 0.0001895827412126607,
+      "loss": 0.8993,
+      "step": 246
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.16043460369110107,
+      "learning_rate": 0.00018949585454456234,
+      "loss": 0.9826,
+      "step": 247
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1589144617319107,
+      "learning_rate": 0.00018940862711476513,
+      "loss": 0.905,
+      "step": 248
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.15877105295658112,
+      "learning_rate": 0.00018932105925539356,
+      "loss": 0.9585,
+      "step": 249
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.14759941399097443,
+      "learning_rate": 0.00018923315129986835,
+      "loss": 0.9121,
+      "step": 250
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.1556202620267868,
+      "learning_rate": 0.0001891449035829052,
+      "loss": 0.9591,
+      "step": 251
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15696395933628082,
+      "learning_rate": 0.0001890563164405134,
+      "loss": 0.9747,
+      "step": 252
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16544775664806366,
+      "learning_rate": 0.00018896739020999467,
+      "loss": 0.9851,
+      "step": 253
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1663719266653061,
+      "learning_rate": 0.00018887812522994186,
+      "loss": 1.0153,
+      "step": 254
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16498468816280365,
+      "learning_rate": 0.0001887885218402375,
+      "loss": 0.9433,
+      "step": 255
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.16752475500106812,
+      "learning_rate": 0.0001886985803820529,
+      "loss": 0.9771,
+      "step": 256
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15939229726791382,
+      "learning_rate": 0.00018860830119784628,
+      "loss": 0.9311,
+      "step": 257
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15414445102214813,
+      "learning_rate": 0.000188517684631362,
+      "loss": 0.9275,
+      "step": 258
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.166487917304039,
+      "learning_rate": 0.0001884267310276289,
+      "loss": 0.9745,
+      "step": 259
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1624753773212433,
+      "learning_rate": 0.00018833544073295917,
+      "loss": 0.9021,
+      "step": 260
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1506534367799759,
+      "learning_rate": 0.000188243814094947,
+      "loss": 0.8681,
+      "step": 261
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.1566423624753952,
+      "learning_rate": 0.00018815185146246716,
+      "loss": 0.9354,
+      "step": 262
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15600961446762085,
+      "learning_rate": 0.0001880595531856738,
+      "loss": 0.8906,
+      "step": 263
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.14650306105613708,
+      "learning_rate": 0.00018796691961599904,
+      "loss": 0.9145,
+      "step": 264
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15429402887821198,
+      "learning_rate": 0.00018787395110615162,
+      "loss": 0.9536,
+      "step": 265
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15946565568447113,
+      "learning_rate": 0.00018778064801011564,
+      "loss": 0.9533,
+      "step": 266
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.15552061796188354,
+      "learning_rate": 0.0001876870106831491,
+      "loss": 0.942,
+      "step": 267
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15441003441810608,
+      "learning_rate": 0.00018759303948178268,
+      "loss": 0.882,
+      "step": 268
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1577693223953247,
+      "learning_rate": 0.00018749873476381828,
+      "loss": 0.9292,
+      "step": 269
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.17108196020126343,
+      "learning_rate": 0.00018740409688832764,
+      "loss": 1.0521,
+      "step": 270
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16209512948989868,
+      "learning_rate": 0.00018730912621565108,
+      "loss": 1.0012,
+      "step": 271
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.173130601644516,
+      "learning_rate": 0.00018721382310739608,
+      "loss": 1.0097,
+      "step": 272
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16810902953147888,
+      "learning_rate": 0.00018711818792643584,
+      "loss": 0.9621,
+      "step": 273
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.1646544188261032,
+      "learning_rate": 0.00018702222103690797,
+      "loss": 1.0051,
+      "step": 274
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14652374386787415,
+      "learning_rate": 0.00018692592280421307,
+      "loss": 0.8897,
+      "step": 275
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.159442737698555,
+      "learning_rate": 0.00018682929359501338,
+      "loss": 0.9605,
+      "step": 276
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15850922465324402,
+      "learning_rate": 0.00018673233377723136,
+      "loss": 0.9649,
+      "step": 277
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15739966928958893,
+      "learning_rate": 0.00018663504372004824,
+      "loss": 1.0098,
+      "step": 278
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15784309804439545,
+      "learning_rate": 0.00018653742379390274,
+      "loss": 0.8929,
+      "step": 279
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.14993084967136383,
+      "learning_rate": 0.00018643947437048944,
+      "loss": 0.9429,
+      "step": 280
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16533209383487701,
+      "learning_rate": 0.00018634119582275766,
+      "loss": 1.0104,
+      "step": 281
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.15955400466918945,
+      "learning_rate": 0.0001862425885249098,
+      "loss": 0.941,
+      "step": 282
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.16894498467445374,
+      "learning_rate": 0.0001861436528524,
+      "loss": 0.9568,
+      "step": 283
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15746143460273743,
+      "learning_rate": 0.00018604438918193273,
+      "loss": 0.8615,
+      "step": 284
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.157247856259346,
+      "learning_rate": 0.00018594479789146138,
+      "loss": 1.0079,
+      "step": 285
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15766699612140656,
+      "learning_rate": 0.00018584487936018661,
+      "loss": 0.9202,
+      "step": 286
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1572881042957306,
+      "learning_rate": 0.0001857446339685553,
+      "loss": 0.9844,
+      "step": 287
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1533818542957306,
+      "learning_rate": 0.0001856440620982587,
+      "loss": 0.9597,
+      "step": 288
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15828703343868256,
+      "learning_rate": 0.0001855431641322312,
+      "loss": 0.9511,
+      "step": 289
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15478287637233734,
+      "learning_rate": 0.00018544194045464886,
+      "loss": 0.9319,
+      "step": 290
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1608508974313736,
+      "learning_rate": 0.00018534039145092786,
+      "loss": 0.9778,
+      "step": 291
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15300685167312622,
+      "learning_rate": 0.00018523851750772318,
+      "loss": 0.9477,
+      "step": 292
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1596435308456421,
+      "learning_rate": 0.00018513631901292686,
+      "loss": 0.9663,
+      "step": 293
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16469767689704895,
+      "learning_rate": 0.0001850337963556668,
+      "loss": 0.92,
+      "step": 294
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.17223823070526123,
+      "learning_rate": 0.00018493094992630518,
+      "loss": 0.8333,
+      "step": 295
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.16195571422576904,
+      "learning_rate": 0.00018482778011643696,
+      "loss": 0.9373,
+      "step": 296
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.1574103832244873,
+      "learning_rate": 0.00018472428731888837,
+      "loss": 0.9453,
+      "step": 297
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15881569683551788,
+      "learning_rate": 0.00018462047192771542,
+      "loss": 0.9292,
+      "step": 298
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.15923351049423218,
+      "learning_rate": 0.0001845163343382024,
+      "loss": 0.9115,
+      "step": 299
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1724962294101715,
+      "learning_rate": 0.00018441187494686053,
+      "loss": 0.9755,
+      "step": 300
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15946060419082642,
+      "learning_rate": 0.00018430709415142617,
+      "loss": 0.9454,
+      "step": 301
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15819205343723297,
+      "learning_rate": 0.00018420199235085952,
+      "loss": 0.8806,
+      "step": 302
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16571106016635895,
+      "learning_rate": 0.00018409656994534298,
+      "loss": 0.9923,
+      "step": 303
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17277686297893524,
+      "learning_rate": 0.00018399082733627965,
+      "loss": 0.9279,
+      "step": 304
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16722162067890167,
+      "learning_rate": 0.00018388476492629197,
+      "loss": 0.9464,
+      "step": 305
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17181581258773804,
+      "learning_rate": 0.00018377838311921992,
+      "loss": 0.9954,
+      "step": 306
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15838095545768738,
+      "learning_rate": 0.00018367168232011962,
+      "loss": 0.9793,
+      "step": 307
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16448432207107544,
+      "learning_rate": 0.00018356466293526183,
+      "loss": 0.9781,
+      "step": 308
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15508323907852173,
+      "learning_rate": 0.00018345732537213027,
+      "loss": 0.9309,
+      "step": 309
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15612675249576569,
+      "learning_rate": 0.0001833496700394202,
+      "loss": 0.8913,
+      "step": 310
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1593405157327652,
+      "learning_rate": 0.00018324169734703683,
+      "loss": 0.933,
+      "step": 311
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.16736635565757751,
+      "learning_rate": 0.00018313340770609367,
+      "loss": 0.9598,
+      "step": 312
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.15413588285446167,
+      "learning_rate": 0.00018302480152891106,
+      "loss": 0.9331,
+      "step": 313
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.17913931608200073,
+      "learning_rate": 0.00018291587922901462,
+      "loss": 0.9263,
+      "step": 314
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.1654043346643448,
+      "learning_rate": 0.00018280664122113355,
+      "loss": 0.8511,
+      "step": 315
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18442225456237793,
+      "learning_rate": 0.00018269708792119921,
+      "loss": 0.9445,
+      "step": 316
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.18277233839035034,
+      "learning_rate": 0.00018258721974634341,
+      "loss": 0.9397,
+      "step": 317
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16975083947181702,
+      "learning_rate": 0.00018247703711489686,
+      "loss": 1.0132,
+      "step": 318
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16668036580085754,
+      "learning_rate": 0.00018236654044638764,
+      "loss": 0.9426,
+      "step": 319
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1658882200717926,
+      "learning_rate": 0.00018225573016153945,
+      "loss": 0.9698,
+      "step": 320
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.155363067984581,
+      "learning_rate": 0.00018214460668227023,
+      "loss": 0.9317,
+      "step": 321
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1638273149728775,
+      "learning_rate": 0.0001820331704316903,
+      "loss": 0.9486,
+      "step": 322
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15337038040161133,
+      "learning_rate": 0.00018192142183410102,
+      "loss": 1.004,
+      "step": 323
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15263631939888,
+      "learning_rate": 0.0001818093613149929,
+      "loss": 0.8769,
+      "step": 324
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1707790046930313,
+      "learning_rate": 0.0001816969893010442,
+      "loss": 0.9754,
+      "step": 325
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1663234382867813,
+      "learning_rate": 0.00018158430622011918,
+      "loss": 0.9241,
+      "step": 326
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15456202626228333,
+      "learning_rate": 0.00018147131250126653,
+      "loss": 0.9221,
+      "step": 327
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.1634504199028015,
+      "learning_rate": 0.00018135800857471768,
+      "loss": 0.975,
+      "step": 328
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16812089085578918,
+      "learning_rate": 0.00018124439487188525,
+      "loss": 1.0054,
+      "step": 329
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.15992523729801178,
+      "learning_rate": 0.00018113047182536127,
+      "loss": 0.9927,
+      "step": 330
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16373218595981598,
+      "learning_rate": 0.00018101623986891574,
+      "loss": 0.9717,
+      "step": 331
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.16420403122901917,
+      "learning_rate": 0.00018090169943749476,
+      "loss": 0.9147,
+      "step": 332
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1626630574464798,
+      "learning_rate": 0.000180786850967219,
+      "loss": 0.9763,
+      "step": 333
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1740325540304184,
+      "learning_rate": 0.000180671694895382,
+      "loss": 0.8994,
+      "step": 334
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16136254370212555,
+      "learning_rate": 0.00018055623166044854,
+      "loss": 0.9521,
+      "step": 335
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1706748902797699,
+      "learning_rate": 0.00018044046170205292,
+      "loss": 1.0206,
+      "step": 336
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16417153179645538,
+      "learning_rate": 0.0001803243854609973,
+      "loss": 0.8801,
+      "step": 337
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1656985580921173,
+      "learning_rate": 0.0001802080033792501,
+      "loss": 0.9826,
+      "step": 338
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16293616592884064,
+      "learning_rate": 0.00018009131589994418,
+      "loss": 0.91,
+      "step": 339
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15921129286289215,
+      "learning_rate": 0.00017997432346737524,
+      "loss": 0.8974,
+      "step": 340
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16984550654888153,
+      "learning_rate": 0.0001798570265270001,
+      "loss": 1.0132,
+      "step": 341
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16374579071998596,
+      "learning_rate": 0.00017973942552543503,
+      "loss": 0.917,
+      "step": 342
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15114158391952515,
+      "learning_rate": 0.00017962152091045405,
+      "loss": 0.827,
+      "step": 343
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16914744675159454,
+      "learning_rate": 0.0001795033131309872,
+      "loss": 0.9264,
+      "step": 344
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.1594720184803009,
+      "learning_rate": 0.0001793848026371188,
+      "loss": 0.8903,
+      "step": 345
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.17011883854866028,
+      "learning_rate": 0.00017926598988008582,
+      "loss": 0.8772,
+      "step": 346
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.15930189192295074,
+      "learning_rate": 0.00017914687531227615,
+      "loss": 0.9468,
+      "step": 347
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.16631504893302917,
+      "learning_rate": 0.00017902745938722675,
+      "loss": 0.9883,
+      "step": 348
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1664229780435562,
+      "learning_rate": 0.0001789077425596221,
+      "loss": 0.9036,
+      "step": 349
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15435925126075745,
+      "learning_rate": 0.00017878772528529232,
+      "loss": 0.8573,
+      "step": 350
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16924455761909485,
+      "learning_rate": 0.00017866740802121157,
+      "loss": 0.9597,
+      "step": 351
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16971321403980255,
+      "learning_rate": 0.00017854679122549625,
+      "loss": 1.0098,
+      "step": 352
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16085046529769897,
+      "learning_rate": 0.00017842587535740314,
+      "loss": 0.9609,
+      "step": 353
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18205095827579498,
+      "learning_rate": 0.00017830466087732786,
+      "loss": 0.965,
+      "step": 354
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16832834482192993,
+      "learning_rate": 0.000178183148246803,
+      "loss": 0.9174,
+      "step": 355
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.18558774888515472,
+      "learning_rate": 0.00017806133792849633,
+      "loss": 0.9444,
+      "step": 356
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17227703332901,
+      "learning_rate": 0.0001779392303862091,
+      "loss": 0.908,
+      "step": 357
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1650489717721939,
+      "learning_rate": 0.0001778168260848744,
+      "loss": 0.8979,
+      "step": 358
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.155917689204216,
+      "learning_rate": 0.000177694125490555,
+      "loss": 0.8964,
+      "step": 359
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16796348989009857,
+      "learning_rate": 0.000177571129070442,
+      "loss": 1.0064,
+      "step": 360
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.16753911972045898,
+      "learning_rate": 0.00017744783729285285,
+      "loss": 0.993,
+      "step": 361
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.17053554952144623,
+      "learning_rate": 0.00017732425062722958,
+      "loss": 0.9686,
+      "step": 362
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.15861544013023376,
+      "learning_rate": 0.000177200369544137,
+      "loss": 0.9141,
+      "step": 363
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.1590377241373062,
+      "learning_rate": 0.000177076194515261,
+      "loss": 0.9296,
+      "step": 364
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16656719148159027,
+      "learning_rate": 0.0001769517260134066,
+      "loss": 0.8941,
+      "step": 365
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16416172683238983,
+      "learning_rate": 0.00017682696451249634,
+      "loss": 0.8735,
+      "step": 366
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16383446753025055,
+      "learning_rate": 0.0001767019104875683,
+      "loss": 0.9162,
+      "step": 367
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16413764655590057,
+      "learning_rate": 0.00017657656441477437,
+      "loss": 0.9456,
+      "step": 368
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1698492467403412,
+      "learning_rate": 0.00017645092677137848,
+      "loss": 0.9625,
+      "step": 369
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16152215003967285,
+      "learning_rate": 0.00017632499803575474,
+      "loss": 0.9655,
+      "step": 370
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16848498582839966,
+      "learning_rate": 0.0001761987786873855,
+      "loss": 0.9607,
+      "step": 371
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1602499634027481,
+      "learning_rate": 0.00017607226920685976,
+      "loss": 0.9053,
+      "step": 372
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1640511155128479,
+      "learning_rate": 0.00017594547007587118,
+      "loss": 0.9321,
+      "step": 373
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16732460260391235,
+      "learning_rate": 0.0001758183817772163,
+      "loss": 0.9024,
+      "step": 374
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16569490730762482,
+      "learning_rate": 0.0001756910047947926,
+      "loss": 0.9004,
+      "step": 375
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.1634737104177475,
+      "learning_rate": 0.0001755633396135969,
+      "loss": 0.9061,
+      "step": 376
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17057223618030548,
+      "learning_rate": 0.00017543538671972318,
+      "loss": 0.9557,
+      "step": 377
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16236676275730133,
+      "learning_rate": 0.00017530714660036112,
+      "loss": 0.9317,
+      "step": 378
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.16234555840492249,
+      "learning_rate": 0.00017517861974379378,
+      "loss": 0.9667,
+      "step": 379
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.17065909504890442,
+      "learning_rate": 0.00017504980663939613,
+      "loss": 0.9639,
+      "step": 380
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1569371372461319,
+      "learning_rate": 0.0001749207077776331,
+      "loss": 0.9072,
+      "step": 381
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15749488770961761,
+      "learning_rate": 0.00017479132365005756,
+      "loss": 0.8986,
+      "step": 382
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1759043037891388,
+      "learning_rate": 0.0001746616547493086,
+      "loss": 0.9083,
+      "step": 383
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1809539794921875,
+      "learning_rate": 0.0001745317015691096,
+      "loss": 0.9743,
+      "step": 384
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17483428120613098,
+      "learning_rate": 0.0001744014646042663,
+      "loss": 0.9471,
+      "step": 385
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16812138259410858,
+      "learning_rate": 0.00017427094435066502,
+      "loss": 0.9628,
+      "step": 386
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15682841837406158,
+      "learning_rate": 0.00017414014130527077,
+      "loss": 0.9277,
+      "step": 387
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16021856665611267,
+      "learning_rate": 0.0001740090559661252,
+      "loss": 0.9214,
+      "step": 388
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17321635782718658,
+      "learning_rate": 0.00017387768883234489,
+      "loss": 0.9813,
+      "step": 389
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.1773107498884201,
+      "learning_rate": 0.00017374604040411935,
+      "loss": 0.9114,
+      "step": 390
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17130543291568756,
+      "learning_rate": 0.00017361411118270914,
+      "loss": 0.8627,
+      "step": 391
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.17349937558174133,
+      "learning_rate": 0.00017348190167044394,
+      "loss": 1.0018,
+      "step": 392
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16585971415042877,
+      "learning_rate": 0.0001733494123707207,
+      "loss": 0.9478,
+      "step": 393
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16625596582889557,
+      "learning_rate": 0.00017321664378800166,
+      "loss": 0.8661,
+      "step": 394
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.18391601741313934,
+      "learning_rate": 0.00017308359642781242,
+      "loss": 0.9185,
+      "step": 395
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.15480977296829224,
+      "learning_rate": 0.0001729502707967401,
+      "loss": 0.9044,
+      "step": 396
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.16025258600711823,
+      "learning_rate": 0.0001728166674024313,
+      "loss": 0.9077,
+      "step": 397
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17228160798549652,
+      "learning_rate": 0.00017268278675359018,
+      "loss": 0.9236,
+      "step": 398
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16289672255516052,
+      "learning_rate": 0.00017254862935997677,
+      "loss": 0.8719,
+      "step": 399
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16713880002498627,
+      "learning_rate": 0.00017241419573240462,
+      "loss": 0.9463,
+      "step": 400
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.169233039021492,
+      "learning_rate": 0.00017227948638273916,
+      "loss": 0.9359,
+      "step": 401
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16896389424800873,
+      "learning_rate": 0.00017214450182389559,
+      "loss": 0.9417,
+      "step": 402
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16022677719593048,
+      "learning_rate": 0.000172009242569837,
+      "loss": 0.9256,
+      "step": 403
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17156574130058289,
+      "learning_rate": 0.00017187370913557246,
+      "loss": 0.9312,
+      "step": 404
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1679723560810089,
+      "learning_rate": 0.00017173790203715494,
+      "loss": 0.892,
+      "step": 405
+    },
+    {
+      "epoch": 0.25,
+      "eval_loss": 1.6185349225997925,
+      "eval_runtime": 187.7472,
+      "eval_samples_per_second": 146.351,
+      "eval_steps_per_second": 73.178,
+      "step": 405
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17154468595981598,
+      "learning_rate": 0.00017160182179167942,
+      "loss": 0.9344,
+      "step": 406
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16832932829856873,
+      "learning_rate": 0.00017146546891728094,
+      "loss": 0.9465,
+      "step": 407
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16341401636600494,
+      "learning_rate": 0.00017132884393313255,
+      "loss": 0.9267,
+      "step": 408
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17056919634342194,
+      "learning_rate": 0.00017119194735944337,
+      "loss": 0.9032,
+      "step": 409
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.16983801126480103,
+      "learning_rate": 0.00017105477971745666,
+      "loss": 0.9181,
+      "step": 410
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17132684588432312,
+      "learning_rate": 0.0001709173415294478,
+      "loss": 0.9269,
+      "step": 411
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.17320628464221954,
+      "learning_rate": 0.0001707796333187222,
+      "loss": 0.9728,
+      "step": 412
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.1646837443113327,
+      "learning_rate": 0.0001706416556096135,
+      "loss": 0.9029,
+      "step": 413
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16305513679981232,
+      "learning_rate": 0.00017050340892748139,
+      "loss": 0.8826,
+      "step": 414
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17124176025390625,
+      "learning_rate": 0.00017036489379870982,
+      "loss": 0.9116,
+      "step": 415
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16273482143878937,
+      "learning_rate": 0.00017022611075070474,
+      "loss": 0.9454,
+      "step": 416
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16589228808879852,
+      "learning_rate": 0.00017008706031189225,
+      "loss": 0.8461,
+      "step": 417
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17099031805992126,
+      "learning_rate": 0.00016994774301171662,
+      "loss": 0.9651,
+      "step": 418
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16664941608905792,
+      "learning_rate": 0.00016980815938063816,
+      "loss": 0.9152,
+      "step": 419
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15957285463809967,
+      "learning_rate": 0.00016966830995013133,
+      "loss": 0.8779,
+      "step": 420
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17603972554206848,
+      "learning_rate": 0.0001695281952526824,
+      "loss": 0.9333,
+      "step": 421
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1547289788722992,
+      "learning_rate": 0.00016938781582178804,
+      "loss": 0.8524,
+      "step": 422
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16680575907230377,
+      "learning_rate": 0.0001692471721919526,
+      "loss": 0.9586,
+      "step": 423
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16918885707855225,
+      "learning_rate": 0.00016910626489868649,
+      "loss": 0.9521,
+      "step": 424
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.15893614292144775,
+      "learning_rate": 0.00016896509447850407,
+      "loss": 0.8761,
+      "step": 425
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16789183020591736,
+      "learning_rate": 0.00016882366146892154,
+      "loss": 0.9541,
+      "step": 426
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.16646571457386017,
+      "learning_rate": 0.00016868196640845495,
+      "loss": 0.9218,
+      "step": 427
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.1720345914363861,
+      "learning_rate": 0.0001685400098366181,
+      "loss": 0.9293,
+      "step": 428
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.17809993028640747,
+      "learning_rate": 0.00016839779229392053,
+      "loss": 0.8589,
+      "step": 429
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1753726303577423,
+      "learning_rate": 0.00016825531432186543,
+      "loss": 1.0154,
+      "step": 430
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.18114417791366577,
+      "learning_rate": 0.00016811257646294757,
+      "loss": 0.9269,
+      "step": 431
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1746157854795456,
+      "learning_rate": 0.00016796957926065134,
+      "loss": 0.9973,
+      "step": 432
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1642226129770279,
+      "learning_rate": 0.00016782632325944855,
+      "loss": 0.8985,
+      "step": 433
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17525264620780945,
+      "learning_rate": 0.00016768280900479634,
+      "loss": 0.9564,
+      "step": 434
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16471520066261292,
+      "learning_rate": 0.00016753903704313527,
+      "loss": 0.8895,
+      "step": 435
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17157848179340363,
+      "learning_rate": 0.0001673950079218871,
+      "loss": 0.8717,
+      "step": 436
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1647266447544098,
+      "learning_rate": 0.00016725072218945272,
+      "loss": 0.914,
+      "step": 437
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16844472289085388,
+      "learning_rate": 0.00016710618039521014,
+      "loss": 0.8691,
+      "step": 438
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17730486392974854,
+      "learning_rate": 0.0001669613830895123,
+      "loss": 0.9342,
+      "step": 439
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17128102481365204,
+      "learning_rate": 0.00016681633082368498,
+      "loss": 0.9787,
+      "step": 440
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.1719089299440384,
+      "learning_rate": 0.0001666710241500248,
+      "loss": 0.9025,
+      "step": 441
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17248529195785522,
+      "learning_rate": 0.0001665254636217971,
+      "loss": 0.9592,
+      "step": 442
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.162600576877594,
+      "learning_rate": 0.00016637964979323363,
+      "loss": 0.9149,
+      "step": 443
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.16336123645305634,
+      "learning_rate": 0.00016623358321953078,
+      "loss": 0.9096,
+      "step": 444
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.17300094664096832,
+      "learning_rate": 0.00016608726445684714,
+      "loss": 0.9412,
+      "step": 445
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17004786431789398,
+      "learning_rate": 0.00016594069406230166,
+      "loss": 0.9055,
+      "step": 446
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17904113233089447,
+      "learning_rate": 0.00016579387259397127,
+      "loss": 0.9453,
+      "step": 447
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15627439320087433,
+      "learning_rate": 0.00016564680061088895,
+      "loss": 0.8613,
+      "step": 448
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.18135209381580353,
+      "learning_rate": 0.00016549947867304152,
+      "loss": 0.893,
+      "step": 449
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1782214492559433,
+      "learning_rate": 0.0001653519073413675,
+      "loss": 0.926,
+      "step": 450
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.15894943475723267,
+      "learning_rate": 0.00016520408717775507,
+      "loss": 0.9143,
+      "step": 451
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16757187247276306,
+      "learning_rate": 0.00016505601874503971,
+      "loss": 0.8822,
+      "step": 452
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17122693359851837,
+      "learning_rate": 0.00016490770260700234,
+      "loss": 0.8792,
+      "step": 453
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1751956343650818,
+      "learning_rate": 0.00016475913932836699,
+      "loss": 0.9668,
+      "step": 454
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17547670006752014,
+      "learning_rate": 0.0001646103294747987,
+      "loss": 0.8525,
+      "step": 455
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17401553690433502,
+      "learning_rate": 0.0001644612736129013,
+      "loss": 0.9327,
+      "step": 456
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.16736406087875366,
+      "learning_rate": 0.00016431197231021543,
+      "loss": 0.8912,
+      "step": 457
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17626748979091644,
+      "learning_rate": 0.0001641624261352161,
+      "loss": 0.9135,
+      "step": 458
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1600714772939682,
+      "learning_rate": 0.0001640126356573109,
+      "loss": 0.8545,
+      "step": 459
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.1674841046333313,
+      "learning_rate": 0.00016386260144683745,
+      "loss": 0.9421,
+      "step": 460
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.17587649822235107,
+      "learning_rate": 0.00016371232407506146,
+      "loss": 0.9606,
+      "step": 461
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.18090274930000305,
+      "learning_rate": 0.00016356180411417447,
+      "loss": 0.9405,
+      "step": 462
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17311732470989227,
+      "learning_rate": 0.00016341104213729178,
+      "loss": 0.856,
+      "step": 463
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17478862404823303,
+      "learning_rate": 0.00016326003871845,
+      "loss": 0.9265,
+      "step": 464
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16457559168338776,
+      "learning_rate": 0.00016310879443260528,
+      "loss": 0.898,
+      "step": 465
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17601493000984192,
+      "learning_rate": 0.00016295730985563074,
+      "loss": 0.904,
+      "step": 466
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17504853010177612,
+      "learning_rate": 0.00016280558556431437,
+      "loss": 0.9869,
+      "step": 467
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1775156408548355,
+      "learning_rate": 0.00016265362213635712,
+      "loss": 0.8985,
+      "step": 468
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16909949481487274,
+      "learning_rate": 0.00016250142015037023,
+      "loss": 0.8772,
+      "step": 469
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1763755977153778,
+      "learning_rate": 0.00016234898018587337,
+      "loss": 0.8987,
+      "step": 470
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16016210615634918,
+      "learning_rate": 0.00016219630282329232,
+      "loss": 0.857,
+      "step": 471
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16906486451625824,
+      "learning_rate": 0.00016204338864395684,
+      "loss": 0.9844,
+      "step": 472
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17785392701625824,
+      "learning_rate": 0.0001618902382300982,
+      "loss": 0.9462,
+      "step": 473
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.16443286836147308,
+      "learning_rate": 0.00016173685216484737,
+      "loss": 0.9181,
+      "step": 474
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.17611455917358398,
+      "learning_rate": 0.0001615832310322324,
+      "loss": 0.8554,
+      "step": 475
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1714741587638855,
+      "learning_rate": 0.0001614293754171765,
+      "loss": 0.8871,
+      "step": 476
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1745806634426117,
+      "learning_rate": 0.0001612752859054956,
+      "loss": 0.9078,
+      "step": 477
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.1708574742078781,
+      "learning_rate": 0.0001611209630838963,
+      "loss": 0.9081,
+      "step": 478
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16905465722084045,
+      "learning_rate": 0.00016096640753997346,
+      "loss": 0.8768,
+      "step": 479
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1825428307056427,
+      "learning_rate": 0.00016081161986220807,
+      "loss": 0.8699,
+      "step": 480
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1652909368276596,
+      "learning_rate": 0.00016065660063996503,
+      "loss": 0.8567,
+      "step": 481
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1831086426973343,
+      "learning_rate": 0.00016050135046349073,
+      "loss": 0.9254,
+      "step": 482
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17882482707500458,
+      "learning_rate": 0.00016034586992391117,
+      "loss": 0.9289,
+      "step": 483
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16337734460830688,
+      "learning_rate": 0.0001601901596132292,
+      "loss": 0.8707,
+      "step": 484
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17880207300186157,
+      "learning_rate": 0.00016003422012432275,
+      "loss": 0.9375,
+      "step": 485
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17019447684288025,
+      "learning_rate": 0.00015987805205094227,
+      "loss": 0.9743,
+      "step": 486
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.16877618432044983,
+      "learning_rate": 0.00015972165598770853,
+      "loss": 0.8373,
+      "step": 487
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18040971457958221,
+      "learning_rate": 0.00015956503253011052,
+      "loss": 1.0085,
+      "step": 488
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18251942098140717,
+      "learning_rate": 0.0001594081822745029,
+      "loss": 0.9235,
+      "step": 489
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.17336182296276093,
+      "learning_rate": 0.00015925110581810394,
+      "loss": 0.8963,
+      "step": 490
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1648416668176651,
+      "learning_rate": 0.00015909380375899322,
+      "loss": 0.8421,
+      "step": 491
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.1580386757850647,
+      "learning_rate": 0.00015893627669610926,
+      "loss": 0.838,
+      "step": 492
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18445619940757751,
+      "learning_rate": 0.00015877852522924732,
+      "loss": 0.9191,
+      "step": 493
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.18491892516613007,
+      "learning_rate": 0.0001586205499590571,
+      "loss": 0.9483,
+      "step": 494
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1714458465576172,
+      "learning_rate": 0.00015846235148704048,
+      "loss": 0.8762,
+      "step": 495
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1694324016571045,
+      "learning_rate": 0.0001583039304155491,
+      "loss": 0.9081,
+      "step": 496
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17766743898391724,
+      "learning_rate": 0.00015814528734778226,
+      "loss": 0.8905,
+      "step": 497
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1762097030878067,
+      "learning_rate": 0.00015798642288778448,
+      "loss": 0.8871,
+      "step": 498
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.17503221333026886,
+      "learning_rate": 0.00015782733764044325,
+      "loss": 0.9171,
+      "step": 499
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16638228297233582,
+      "learning_rate": 0.00015766803221148673,
+      "loss": 0.8959,
+      "step": 500
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16727958619594574,
+      "learning_rate": 0.00015750850720748144,
+      "loss": 0.8988,
+      "step": 501
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1734267920255661,
+      "learning_rate": 0.00015734876323582995,
+      "loss": 0.9219,
+      "step": 502
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16582509875297546,
+      "learning_rate": 0.00015718880090476853,
+      "loss": 0.8467,
+      "step": 503
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18570883572101593,
+      "learning_rate": 0.00015702862082336489,
+      "loss": 0.9305,
+      "step": 504
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.18270942568778992,
+      "learning_rate": 0.0001568682236015159,
+      "loss": 0.9041,
+      "step": 505
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16850034892559052,
+      "learning_rate": 0.00015670760984994514,
+      "loss": 0.9188,
+      "step": 506
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16860127449035645,
+      "learning_rate": 0.0001565467801802006,
+      "loss": 0.9538,
+      "step": 507
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.16461698710918427,
+      "learning_rate": 0.0001563857352046525,
+      "loss": 0.8612,
+      "step": 508
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.1644446849822998,
+      "learning_rate": 0.00015622447553649077,
+      "loss": 0.859,
+      "step": 509
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.19424785673618317,
+      "learning_rate": 0.00015606300178972287,
+      "loss": 0.9066,
+      "step": 510
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1711163967847824,
+      "learning_rate": 0.0001559013145791713,
+      "loss": 0.9015,
+      "step": 511
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1647401750087738,
+      "learning_rate": 0.00015573941452047131,
+      "loss": 0.8472,
+      "step": 512
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1786355823278427,
+      "learning_rate": 0.00015557730223006878,
+      "loss": 0.986,
+      "step": 513
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17015327513217926,
+      "learning_rate": 0.0001554149783252175,
+      "loss": 0.8588,
+      "step": 514
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16389843821525574,
+      "learning_rate": 0.000155252443423977,
+      "loss": 0.9172,
+      "step": 515
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1643114537000656,
+      "learning_rate": 0.00015508969814521025,
+      "loss": 0.9479,
+      "step": 516
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1713220179080963,
+      "learning_rate": 0.0001549267431085813,
+      "loss": 0.8435,
+      "step": 517
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17502859234809875,
+      "learning_rate": 0.00015476357893455268,
+      "loss": 0.8979,
+      "step": 518
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1769992560148239,
+      "learning_rate": 0.00015460020624438346,
+      "loss": 0.994,
+      "step": 519
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.16803620755672455,
+      "learning_rate": 0.00015443662566012645,
+      "loss": 0.9697,
+      "step": 520
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1682910919189453,
+      "learning_rate": 0.0001542728378046262,
+      "loss": 0.9462,
+      "step": 521
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17835430800914764,
+      "learning_rate": 0.00015410884330151626,
+      "loss": 0.8888,
+      "step": 522
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1704440414905548,
+      "learning_rate": 0.00015394464277521726,
+      "loss": 0.9133,
+      "step": 523
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.1732661873102188,
+      "learning_rate": 0.00015378023685093406,
+      "loss": 0.8571,
+      "step": 524
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.17370902001857758,
+      "learning_rate": 0.00015361562615465367,
+      "loss": 0.9203,
+      "step": 525
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.18120653927326202,
+      "learning_rate": 0.00015345081131314274,
+      "loss": 0.8795,
+      "step": 526
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17480316758155823,
+      "learning_rate": 0.00015328579295394534,
+      "loss": 0.8935,
+      "step": 527
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17862842977046967,
+      "learning_rate": 0.00015312057170538035,
+      "loss": 0.9475,
+      "step": 528
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17572833597660065,
+      "learning_rate": 0.00015295514819653913,
+      "loss": 0.9468,
+      "step": 529
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17243683338165283,
+      "learning_rate": 0.00015278952305728324,
+      "loss": 0.8454,
+      "step": 530
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1824159026145935,
+      "learning_rate": 0.00015262369691824195,
+      "loss": 0.9583,
+      "step": 531
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1800861656665802,
+      "learning_rate": 0.00015245767041080981,
+      "loss": 0.9682,
+      "step": 532
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.16624125838279724,
+      "learning_rate": 0.00015229144416714434,
+      "loss": 0.8426,
+      "step": 533
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.167984277009964,
+      "learning_rate": 0.00015212501882016344,
+      "loss": 0.8874,
+      "step": 534
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17666511237621307,
+      "learning_rate": 0.00015195839500354335,
+      "loss": 0.9527,
+      "step": 535
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17988379299640656,
+      "learning_rate": 0.0001517915733517158,
+      "loss": 0.9543,
+      "step": 536
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17952507734298706,
+      "learning_rate": 0.00015162455449986578,
+      "loss": 0.9476,
+      "step": 537
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17034880816936493,
+      "learning_rate": 0.0001514573390839293,
+      "loss": 0.9112,
+      "step": 538
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17450974881649017,
+      "learning_rate": 0.00015128992774059063,
+      "loss": 0.9033,
+      "step": 539
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.19009935855865479,
+      "learning_rate": 0.00015112232110728015,
+      "loss": 0.9187,
+      "step": 540
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.17472144961357117,
+      "learning_rate": 0.00015095451982217176,
+      "loss": 0.872,
+      "step": 541
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.1820639818906784,
+      "learning_rate": 0.00015078652452418063,
+      "loss": 0.9888,
+      "step": 542
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16746963560581207,
+      "learning_rate": 0.00015061833585296045,
+      "loss": 0.8563,
+      "step": 543
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16884446144104004,
+      "learning_rate": 0.0001504499544489013,
+      "loss": 0.8622,
+      "step": 544
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1650804579257965,
+      "learning_rate": 0.0001502813809531272,
+      "loss": 0.8985,
+      "step": 545
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17889395356178284,
+      "learning_rate": 0.0001501126160074934,
+      "loss": 0.9372,
+      "step": 546
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.168239563703537,
+      "learning_rate": 0.0001499436602545842,
+      "loss": 0.9178,
+      "step": 547
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1722397655248642,
+      "learning_rate": 0.00014977451433771037,
+      "loss": 0.9347,
+      "step": 548
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17344911396503448,
+      "learning_rate": 0.0001496051789009068,
+      "loss": 0.954,
+      "step": 549
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.16188882291316986,
+      "learning_rate": 0.00014943565458893,
+      "loss": 0.8827,
+      "step": 550
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17988407611846924,
+      "learning_rate": 0.0001492659420472555,
+      "loss": 0.9189,
+      "step": 551
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18148158490657806,
+      "learning_rate": 0.00014909604192207568,
+      "loss": 0.9726,
+      "step": 552
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.1817133128643036,
+      "learning_rate": 0.0001489259548602971,
+      "loss": 0.9313,
+      "step": 553
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17899854481220245,
+      "learning_rate": 0.00014875568150953805,
+      "loss": 0.8836,
+      "step": 554
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17939497530460358,
+      "learning_rate": 0.0001485852225181262,
+      "loss": 0.8881,
+      "step": 555
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18853846192359924,
+      "learning_rate": 0.00014841457853509606,
+      "loss": 1.0299,
+      "step": 556
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17726485431194305,
+      "learning_rate": 0.00014824375021018645,
+      "loss": 0.9407,
+      "step": 557
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.18113796412944794,
+      "learning_rate": 0.0001480727381938381,
+      "loss": 0.8984,
+      "step": 558
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.17925652861595154,
+      "learning_rate": 0.00014790154313719115,
+      "loss": 0.9751,
+      "step": 559
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.18046168982982635,
+      "learning_rate": 0.00014773016569208283,
+      "loss": 0.9423,
+      "step": 560
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.166167750954628,
+      "learning_rate": 0.00014755860651104454,
+      "loss": 0.8804,
+      "step": 561
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17811396718025208,
+      "learning_rate": 0.00014738686624729986,
+      "loss": 0.9172,
+      "step": 562
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16356758773326874,
+      "learning_rate": 0.00014721494555476188,
+      "loss": 0.8995,
+      "step": 563
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17158697545528412,
+      "learning_rate": 0.00014704284508803048,
+      "loss": 0.9517,
+      "step": 564
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1741390973329544,
+      "learning_rate": 0.00014687056550239028,
+      "loss": 0.8454,
+      "step": 565
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17156049609184265,
+      "learning_rate": 0.00014669810745380777,
+      "loss": 0.8847,
+      "step": 566
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.16608455777168274,
+      "learning_rate": 0.00014652547159892896,
+      "loss": 0.9037,
+      "step": 567
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.15927013754844666,
+      "learning_rate": 0.00014635265859507687,
+      "loss": 0.8492,
+      "step": 568
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1725761592388153,
+      "learning_rate": 0.00014617966910024914,
+      "loss": 0.8995,
+      "step": 569
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17553302645683289,
+      "learning_rate": 0.00014600650377311522,
+      "loss": 0.8966,
+      "step": 570
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1696312129497528,
+      "learning_rate": 0.0001458331632730142,
+      "loss": 0.921,
+      "step": 571
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17569565773010254,
+      "learning_rate": 0.0001456596482599521,
+      "loss": 0.8826,
+      "step": 572
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17770029604434967,
+      "learning_rate": 0.0001454859593945994,
+      "loss": 0.9199,
+      "step": 573
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.1804373413324356,
+      "learning_rate": 0.00014531209733828856,
+      "loss": 0.8828,
+      "step": 574
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.17791211605072021,
+      "learning_rate": 0.0001451380627530115,
+      "loss": 0.8327,
+      "step": 575
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1708548218011856,
+      "learning_rate": 0.00014496385630141697,
+      "loss": 0.8759,
+      "step": 576
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18910300731658936,
+      "learning_rate": 0.0001447894786468082,
+      "loss": 0.9485,
+      "step": 577
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17465347051620483,
+      "learning_rate": 0.00014461493045314026,
+      "loss": 0.8879,
+      "step": 578
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17286266386508942,
+      "learning_rate": 0.00014444021238501752,
+      "loss": 0.8689,
+      "step": 579
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.18508301675319672,
+      "learning_rate": 0.0001442653251076912,
+      "loss": 0.9369,
+      "step": 580
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.19138920307159424,
+      "learning_rate": 0.00014409026928705683,
+      "loss": 0.9096,
+      "step": 581
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17769737541675568,
+      "learning_rate": 0.00014391504558965157,
+      "loss": 0.8675,
+      "step": 582
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16904981434345245,
+      "learning_rate": 0.00014373965468265194,
+      "loss": 0.8804,
+      "step": 583
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1705213487148285,
+      "learning_rate": 0.0001435640972338709,
+      "loss": 0.8282,
+      "step": 584
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.165312722325325,
+      "learning_rate": 0.00014338837391175582,
+      "loss": 0.9228,
+      "step": 585
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17032013833522797,
+      "learning_rate": 0.00014321248538538535,
+      "loss": 0.9109,
+      "step": 586
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17847879230976105,
+      "learning_rate": 0.00014303643232446743,
+      "loss": 0.9915,
+      "step": 587
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.1722143292427063,
+      "learning_rate": 0.00014286021539933625,
+      "loss": 0.8997,
+      "step": 588
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.16865919530391693,
+      "learning_rate": 0.00014268383528095007,
+      "loss": 0.8903,
+      "step": 589
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17730148136615753,
+      "learning_rate": 0.00014250729264088843,
+      "loss": 0.8211,
+      "step": 590
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.17798572778701782,
+      "learning_rate": 0.00014233058815134978,
+      "loss": 0.9271,
+      "step": 591
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17852333188056946,
+      "learning_rate": 0.00014215372248514875,
+      "loss": 0.8884,
+      "step": 592
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1794460415840149,
+      "learning_rate": 0.00014197669631571368,
+      "loss": 0.9006,
+      "step": 593
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.16837218403816223,
+      "learning_rate": 0.00014179951031708407,
+      "loss": 0.9137,
+      "step": 594
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18018725514411926,
+      "learning_rate": 0.00014162216516390788,
+      "loss": 0.9284,
+      "step": 595
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18733176589012146,
+      "learning_rate": 0.00014144466153143918,
+      "loss": 0.9632,
+      "step": 596
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18396121263504028,
+      "learning_rate": 0.00014126700009553538,
+      "loss": 0.8896,
+      "step": 597
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17854280769824982,
+      "learning_rate": 0.00014108918153265485,
+      "loss": 0.8739,
+      "step": 598
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.1756531000137329,
+      "learning_rate": 0.00014091120651985403,
+      "loss": 0.8871,
+      "step": 599
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18185968697071075,
+      "learning_rate": 0.00014073307573478526,
+      "loss": 0.9247,
+      "step": 600
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.2246873378753662,
+      "learning_rate": 0.00014055478985569394,
+      "loss": 0.909,
+      "step": 601
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17293086647987366,
+      "learning_rate": 0.00014037634956141586,
+      "loss": 0.889,
+      "step": 602
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.171371191740036,
+      "learning_rate": 0.00014019775553137493,
+      "loss": 0.8926,
+      "step": 603
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.18146918714046478,
+      "learning_rate": 0.00014001900844558038,
+      "loss": 0.9716,
+      "step": 604
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17303234338760376,
+      "learning_rate": 0.00013984010898462416,
+      "loss": 0.9067,
+      "step": 605
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17926719784736633,
+      "learning_rate": 0.00013966105782967843,
+      "loss": 0.9123,
+      "step": 606
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.17768074572086334,
+      "learning_rate": 0.00013948185566249293,
+      "loss": 0.8847,
+      "step": 607
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1646164506673813,
+      "learning_rate": 0.00013930250316539238,
+      "loss": 0.8169,
+      "step": 608
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19589723646640778,
+      "learning_rate": 0.0001391230010212739,
+      "loss": 0.9121,
+      "step": 609
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17763440310955048,
+      "learning_rate": 0.00013894334991360448,
+      "loss": 0.8878,
+      "step": 610
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.174228698015213,
+      "learning_rate": 0.0001387635505264181,
+      "loss": 0.9272,
+      "step": 611
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1815919280052185,
+      "learning_rate": 0.00013858360354431355,
+      "loss": 0.9209,
+      "step": 612
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17741326987743378,
+      "learning_rate": 0.00013840350965245142,
+      "loss": 0.9338,
+      "step": 613
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.171375572681427,
+      "learning_rate": 0.00013822326953655186,
+      "loss": 0.8965,
+      "step": 614
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18087345361709595,
+      "learning_rate": 0.00013804288388289152,
+      "loss": 0.9194,
+      "step": 615
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16721951961517334,
+      "learning_rate": 0.00013786235337830144,
+      "loss": 0.9325,
+      "step": 616
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.18986698985099792,
+      "learning_rate": 0.000137681678710164,
+      "loss": 0.9355,
+      "step": 617
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17893005907535553,
+      "learning_rate": 0.00013750086056641073,
+      "loss": 0.8736,
+      "step": 618
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16385990381240845,
+      "learning_rate": 0.00013731989963551913,
+      "loss": 0.9003,
+      "step": 619
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.16581682860851288,
+      "learning_rate": 0.00013713879660651068,
+      "loss": 0.8767,
+      "step": 620
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.17280465364456177,
+      "learning_rate": 0.00013695755216894771,
+      "loss": 0.9065,
+      "step": 621
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.1769782155752182,
+      "learning_rate": 0.00013677616701293104,
+      "loss": 0.8731,
+      "step": 622
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.19360144436359406,
+      "learning_rate": 0.0001365946418290973,
+      "loss": 0.9082,
+      "step": 623
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17435254156589508,
+      "learning_rate": 0.00013641297730861629,
+      "loss": 0.8821,
+      "step": 624
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17430394887924194,
+      "learning_rate": 0.00013623117414318826,
+      "loss": 0.8993,
+      "step": 625
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.18203990161418915,
+      "learning_rate": 0.00013604923302504147,
+      "loss": 0.9209,
+      "step": 626
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17175577580928802,
+      "learning_rate": 0.0001358671546469294,
+      "loss": 0.8536,
+      "step": 627
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19287161529064178,
+      "learning_rate": 0.00013568493970212816,
+      "loss": 0.9202,
+      "step": 628
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16792431473731995,
+      "learning_rate": 0.0001355025888844338,
+      "loss": 0.8965,
+      "step": 629
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17176759243011475,
+      "learning_rate": 0.0001353201028881598,
+      "loss": 0.9275,
+      "step": 630
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.16665878891944885,
+      "learning_rate": 0.0001351374824081343,
+      "loss": 0.8711,
+      "step": 631
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17521421611309052,
+      "learning_rate": 0.00013495472813969744,
+      "loss": 0.8694,
+      "step": 632
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.175704225897789,
+      "learning_rate": 0.00013477184077869892,
+      "loss": 0.9358,
+      "step": 633
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17732135951519012,
+      "learning_rate": 0.00013458882102149498,
+      "loss": 0.8635,
+      "step": 634
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17731735110282898,
+      "learning_rate": 0.0001344056695649462,
+      "loss": 0.9537,
+      "step": 635
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17888161540031433,
+      "learning_rate": 0.00013422238710641445,
+      "loss": 0.904,
+      "step": 636
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.17445015907287598,
+      "learning_rate": 0.0001340389743437605,
+      "loss": 0.8403,
+      "step": 637
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1798778623342514,
+      "learning_rate": 0.00013385543197534116,
+      "loss": 0.8858,
+      "step": 638
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.1699121743440628,
+      "learning_rate": 0.0001336717607000068,
+      "loss": 0.8649,
+      "step": 639
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.19028843939304352,
+      "learning_rate": 0.00013348796121709862,
+      "loss": 0.9483,
+      "step": 640
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17764434218406677,
+      "learning_rate": 0.00013330403422644595,
+      "loss": 0.9249,
+      "step": 641
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17139072716236115,
+      "learning_rate": 0.00013311998042836356,
+      "loss": 0.8698,
+      "step": 642
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18170565366744995,
+      "learning_rate": 0.00013293580052364915,
+      "loss": 0.9298,
+      "step": 643
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18292510509490967,
+      "learning_rate": 0.00013275149521358053,
+      "loss": 0.8919,
+      "step": 644
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1703965663909912,
+      "learning_rate": 0.000132567065199913,
+      "loss": 0.9093,
+      "step": 645
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.18531350791454315,
+      "learning_rate": 0.0001323825111848767,
+      "loss": 0.8954,
+      "step": 646
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17541749775409698,
+      "learning_rate": 0.00013219783387117385,
+      "loss": 0.8863,
+      "step": 647
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17715893685817719,
+      "learning_rate": 0.00013201303396197621,
+      "loss": 0.9174,
+      "step": 648
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16931463778018951,
+      "learning_rate": 0.0001318281121609223,
+      "loss": 0.8865,
+      "step": 649
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16968059539794922,
+      "learning_rate": 0.00013164306917211476,
+      "loss": 0.8782,
+      "step": 650
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17485713958740234,
+      "learning_rate": 0.00013145790570011766,
+      "loss": 0.8678,
+      "step": 651
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.179866224527359,
+      "learning_rate": 0.00013127262244995382,
+      "loss": 0.9431,
+      "step": 652
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17816640436649323,
+      "learning_rate": 0.0001310872201271021,
+      "loss": 0.8887,
+      "step": 653
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.17722396552562714,
+      "learning_rate": 0.00013090169943749476,
+      "loss": 0.9164,
+      "step": 654
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.16920991241931915,
+      "learning_rate": 0.00013071606108751475,
+      "loss": 0.8738,
+      "step": 655
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.1797451376914978,
+      "learning_rate": 0.000130530305783993,
+      "loss": 0.9565,
+      "step": 656
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1786402314901352,
+      "learning_rate": 0.00013034443423420581,
+      "loss": 0.9274,
+      "step": 657
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18438734114170074,
+      "learning_rate": 0.00013015844714587202,
+      "loss": 0.918,
+      "step": 658
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17035888135433197,
+      "learning_rate": 0.0001299723452271504,
+      "loss": 0.8672,
+      "step": 659
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18038679659366608,
+      "learning_rate": 0.000129786129186637,
+      "loss": 0.8836,
+      "step": 660
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19579140841960907,
+      "learning_rate": 0.00012959979973336237,
+      "loss": 0.9437,
+      "step": 661
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17777006328105927,
+      "learning_rate": 0.00012941335757678883,
+      "loss": 0.8357,
+      "step": 662
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17195580899715424,
+      "learning_rate": 0.0001292268034268079,
+      "loss": 0.9194,
+      "step": 663
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.19506369531154633,
+      "learning_rate": 0.00012904013799373756,
+      "loss": 0.9132,
+      "step": 664
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.177183598279953,
+      "learning_rate": 0.00012885336198831932,
+      "loss": 0.8635,
+      "step": 665
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1794150024652481,
+      "learning_rate": 0.00012866647612171594,
+      "loss": 0.8612,
+      "step": 666
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17331695556640625,
+      "learning_rate": 0.0001284794811055083,
+      "loss": 0.8683,
+      "step": 667
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17770697176456451,
+      "learning_rate": 0.000128292377651693,
+      "loss": 0.9118,
+      "step": 668
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.1756308227777481,
+      "learning_rate": 0.00012810516647267944,
+      "loss": 0.8872,
+      "step": 669
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.18574275076389313,
+      "learning_rate": 0.00012791784828128724,
+      "loss": 0.9066,
+      "step": 670
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17484493553638458,
+      "learning_rate": 0.00012773042379074344,
+      "loss": 0.8811,
+      "step": 671
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.17300565540790558,
+      "learning_rate": 0.00012754289371467986,
+      "loss": 0.8675,
+      "step": 672
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18177475035190582,
+      "learning_rate": 0.0001273552587671303,
+      "loss": 0.9322,
+      "step": 673
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.19294989109039307,
+      "learning_rate": 0.00012716751966252796,
+      "loss": 0.9435,
+      "step": 674
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18173274397850037,
+      "learning_rate": 0.00012697967711570242,
+      "loss": 0.8984,
+      "step": 675
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1820758581161499,
+      "learning_rate": 0.00012679173184187737,
+      "loss": 0.8679,
+      "step": 676
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18205545842647552,
+      "learning_rate": 0.00012660368455666752,
+      "loss": 0.8938,
+      "step": 677
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17626892030239105,
+      "learning_rate": 0.00012641553597607597,
+      "loss": 0.8673,
+      "step": 678
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17685864865779877,
+      "learning_rate": 0.00012622728681649153,
+      "loss": 0.8711,
+      "step": 679
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18767289817333221,
+      "learning_rate": 0.00012603893779468604,
+      "loss": 0.8907,
+      "step": 680
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17344599962234497,
+      "learning_rate": 0.00012585048962781146,
+      "loss": 0.8766,
+      "step": 681
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18164829909801483,
+      "learning_rate": 0.00012566194303339739,
+      "loss": 0.8901,
+      "step": 682
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17202521860599518,
+      "learning_rate": 0.00012547329872934803,
+      "loss": 0.8429,
+      "step": 683
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17770540714263916,
+      "learning_rate": 0.00012528455743393972,
+      "loss": 0.8939,
+      "step": 684
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17684073746204376,
+      "learning_rate": 0.00012509571986581814,
+      "loss": 0.8308,
+      "step": 685
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.18960794806480408,
+      "learning_rate": 0.00012490678674399543,
+      "loss": 0.9612,
+      "step": 686
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.17309339344501495,
+      "learning_rate": 0.00012471775878784768,
+      "loss": 0.8875,
+      "step": 687
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.1742386519908905,
+      "learning_rate": 0.00012452863671711187,
+      "loss": 0.9447,
+      "step": 688
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17292842268943787,
+      "learning_rate": 0.00012433942125188359,
+      "loss": 0.8738,
+      "step": 689
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17790433764457703,
+      "learning_rate": 0.0001241501131126138,
+      "loss": 0.9025,
+      "step": 690
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17391130328178406,
+      "learning_rate": 0.00012396071302010653,
+      "loss": 0.9116,
+      "step": 691
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18365596234798431,
+      "learning_rate": 0.00012377122169551568,
+      "loss": 0.883,
+      "step": 692
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.16463105380535126,
+      "learning_rate": 0.00012358163986034282,
+      "loss": 0.8347,
+      "step": 693
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18167784810066223,
+      "learning_rate": 0.00012339196823643386,
+      "loss": 0.8326,
+      "step": 694
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19045117497444153,
+      "learning_rate": 0.0001232022075459768,
+      "loss": 0.8822,
+      "step": 695
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18004459142684937,
+      "learning_rate": 0.00012301235851149865,
+      "loss": 0.855,
+      "step": 696
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18370255827903748,
+      "learning_rate": 0.00012282242185586285,
+      "loss": 0.8939,
+      "step": 697
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17824755609035492,
+      "learning_rate": 0.0001226323983022664,
+      "loss": 0.8844,
+      "step": 698
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18095892667770386,
+      "learning_rate": 0.00012244228857423726,
+      "loss": 0.8534,
+      "step": 699
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.1725408434867859,
+      "learning_rate": 0.00012225209339563145,
+      "loss": 0.9225,
+      "step": 700
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18814846873283386,
+      "learning_rate": 0.00012206181349063035,
+      "loss": 0.9068,
+      "step": 701
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.19980944693088531,
+      "learning_rate": 0.00012187144958373793,
+      "loss": 0.9415,
+      "step": 702
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.17679652571678162,
+      "learning_rate": 0.00012168100239977809,
+      "loss": 0.9042,
+      "step": 703
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.18936534225940704,
+      "learning_rate": 0.00012149047266389166,
+      "loss": 0.854,
+      "step": 704
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18569441139698029,
+      "learning_rate": 0.00012129986110153396,
+      "loss": 0.8997,
+      "step": 705
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17271539568901062,
+      "learning_rate": 0.00012110916843847172,
+      "loss": 0.9358,
+      "step": 706
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1842963993549347,
+      "learning_rate": 0.00012091839540078054,
+      "loss": 0.8747,
+      "step": 707
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1823119968175888,
+      "learning_rate": 0.00012072754271484207,
+      "loss": 0.8834,
+      "step": 708
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1758606880903244,
+      "learning_rate": 0.00012053661110734124,
+      "loss": 0.8985,
+      "step": 709
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.16988635063171387,
+      "learning_rate": 0.0001203456013052634,
+      "loss": 0.8587,
+      "step": 710
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17307110130786896,
+      "learning_rate": 0.00012015451403589164,
+      "loss": 0.8238,
+      "step": 711
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17483831942081451,
+      "learning_rate": 0.00011996335002680412,
+      "loss": 0.8625,
+      "step": 712
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17725145816802979,
+      "learning_rate": 0.00011977211000587108,
+      "loss": 0.9167,
+      "step": 713
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17913034558296204,
+      "learning_rate": 0.00011958079470125224,
+      "loss": 0.8905,
+      "step": 714
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18657708168029785,
+      "learning_rate": 0.00011938940484139387,
+      "loss": 0.8796,
+      "step": 715
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1729375571012497,
+      "learning_rate": 0.00011919794115502627,
+      "loss": 0.8638,
+      "step": 716
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18231768906116486,
+      "learning_rate": 0.00011900640437116073,
+      "loss": 0.8663,
+      "step": 717
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.18088145554065704,
+      "learning_rate": 0.00011881479521908694,
+      "loss": 0.8655,
+      "step": 718
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1904325634241104,
+      "learning_rate": 0.00011862311442837,
+      "loss": 0.8576,
+      "step": 719
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.17880167067050934,
+      "learning_rate": 0.00011843136272884794,
+      "loss": 0.8417,
+      "step": 720
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.1940956860780716,
+      "learning_rate": 0.00011823954085062868,
+      "loss": 0.9758,
+      "step": 721
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1720036119222641,
+      "learning_rate": 0.0001180476495240874,
+      "loss": 0.8878,
+      "step": 722
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17990410327911377,
+      "learning_rate": 0.00011785568947986367,
+      "loss": 0.8869,
+      "step": 723
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18490731716156006,
+      "learning_rate": 0.00011766366144885877,
+      "loss": 0.8983,
+      "step": 724
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.19024161994457245,
+      "learning_rate": 0.00011747156616223271,
+      "loss": 0.8851,
+      "step": 725
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1743820756673813,
+      "learning_rate": 0.00011727940435140177,
+      "loss": 0.8681,
+      "step": 726
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1824263483285904,
+      "learning_rate": 0.00011708717674803538,
+      "loss": 0.9058,
+      "step": 727
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1716662347316742,
+      "learning_rate": 0.00011689488408405353,
+      "loss": 0.8174,
+      "step": 728
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18255122005939484,
+      "learning_rate": 0.00011670252709162392,
+      "loss": 0.8675,
+      "step": 729
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18594492971897125,
+      "learning_rate": 0.00011651010650315923,
+      "loss": 0.8933,
+      "step": 730
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17157168686389923,
+      "learning_rate": 0.00011631762305131424,
+      "loss": 0.8353,
+      "step": 731
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.177988663315773,
+      "learning_rate": 0.00011612507746898307,
+      "loss": 0.9331,
+      "step": 732
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17265744507312775,
+      "learning_rate": 0.00011593247048929644,
+      "loss": 0.9162,
+      "step": 733
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.17633241415023804,
+      "learning_rate": 0.00011573980284561885,
+      "loss": 0.924,
+      "step": 734
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18592104315757751,
+      "learning_rate": 0.0001155470752715458,
+      "loss": 0.9535,
+      "step": 735
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.18250355124473572,
+      "learning_rate": 0.00011535428850090091,
+      "loss": 0.8504,
+      "step": 736
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.1778443604707718,
+      "learning_rate": 0.00011516144326773323,
+      "loss": 0.9034,
+      "step": 737
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16726194322109222,
+      "learning_rate": 0.00011496854030631443,
+      "loss": 0.8191,
+      "step": 738
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1768944412469864,
+      "learning_rate": 0.000114775580351136,
+      "loss": 0.8658,
+      "step": 739
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.19260743260383606,
+      "learning_rate": 0.00011458256413690633,
+      "loss": 0.9452,
+      "step": 740
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17385084927082062,
+      "learning_rate": 0.00011438949239854821,
+      "loss": 0.898,
+      "step": 741
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1651160567998886,
+      "learning_rate": 0.00011419636587119562,
+      "loss": 0.8276,
+      "step": 742
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.16847139596939087,
+      "learning_rate": 0.00011400318529019134,
+      "loss": 0.8164,
+      "step": 743
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.1743519902229309,
+      "learning_rate": 0.00011380995139108384,
+      "loss": 0.8796,
+      "step": 744
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17673030495643616,
+      "learning_rate": 0.00011361666490962468,
+      "loss": 0.8981,
+      "step": 745
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18919995427131653,
+      "learning_rate": 0.00011342332658176555,
+      "loss": 0.9286,
+      "step": 746
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18586495518684387,
+      "learning_rate": 0.00011322993714365566,
+      "loss": 0.8919,
+      "step": 747
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18239198625087738,
+      "learning_rate": 0.0001130364973316387,
+      "loss": 0.9094,
+      "step": 748
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17158791422843933,
+      "learning_rate": 0.0001128430078822503,
+      "loss": 0.8628,
+      "step": 749
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17712755501270294,
+      "learning_rate": 0.00011264946953221496,
+      "loss": 0.9332,
+      "step": 750
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18135789036750793,
+      "learning_rate": 0.00011245588301844343,
+      "loss": 0.8151,
+      "step": 751
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.18380603194236755,
+      "learning_rate": 0.00011226224907802985,
+      "loss": 0.8559,
+      "step": 752
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.17864134907722473,
+      "learning_rate": 0.00011206856844824896,
+      "loss": 0.8689,
+      "step": 753
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17704331874847412,
+      "learning_rate": 0.00011187484186655319,
+      "loss": 0.8361,
+      "step": 754
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17754654586315155,
+      "learning_rate": 0.00011168107007057005,
+      "loss": 0.9002,
+      "step": 755
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.19279205799102783,
+      "learning_rate": 0.0001114872537980991,
+      "loss": 0.9099,
+      "step": 756
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18683281540870667,
+      "learning_rate": 0.00011129339378710932,
+      "loss": 0.9149,
+      "step": 757
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1713230311870575,
+      "learning_rate": 0.00011109949077573623,
+      "loss": 0.9053,
+      "step": 758
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17414973676204681,
+      "learning_rate": 0.00011090554550227899,
+      "loss": 0.8581,
+      "step": 759
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18200905621051788,
+      "learning_rate": 0.00011071155870519777,
+      "loss": 0.8061,
+      "step": 760
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17745770514011383,
+      "learning_rate": 0.00011051753112311078,
+      "loss": 0.9049,
+      "step": 761
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1719997525215149,
+      "learning_rate": 0.00011032346349479161,
+      "loss": 0.8273,
+      "step": 762
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.16960744559764862,
+      "learning_rate": 0.00011012935655916623,
+      "loss": 0.8248,
+      "step": 763
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18298326432704926,
+      "learning_rate": 0.00010993521105531032,
+      "loss": 0.8446,
+      "step": 764
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.17146216332912445,
+      "learning_rate": 0.00010974102772244639,
+      "loss": 0.8294,
+      "step": 765
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18797147274017334,
+      "learning_rate": 0.00010954680729994102,
+      "loss": 0.951,
+      "step": 766
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.1836959719657898,
+      "learning_rate": 0.000109352550527302,
+      "loss": 0.9066,
+      "step": 767
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18029513955116272,
+      "learning_rate": 0.00010915825814417555,
+      "loss": 0.9148,
+      "step": 768
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.18929041922092438,
+      "learning_rate": 0.00010896393089034336,
+      "loss": 0.9077,
+      "step": 769
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18137846887111664,
+      "learning_rate": 0.00010876956950572006,
+      "loss": 0.8818,
+      "step": 770
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1807103157043457,
+      "learning_rate": 0.00010857517473035013,
+      "loss": 1.0707,
+      "step": 771
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17960844933986664,
+      "learning_rate": 0.00010838074730440524,
+      "loss": 0.8797,
+      "step": 772
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18237923085689545,
+      "learning_rate": 0.00010818628796818133,
+      "loss": 0.9048,
+      "step": 773
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19278337061405182,
+      "learning_rate": 0.00010799179746209587,
+      "loss": 0.884,
+      "step": 774
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.16928324103355408,
+      "learning_rate": 0.00010779727652668497,
+      "loss": 0.8389,
+      "step": 775
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.19637800753116608,
+      "learning_rate": 0.00010760272590260071,
+      "loss": 0.9733,
+      "step": 776
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17628687620162964,
+      "learning_rate": 0.00010740814633060811,
+      "loss": 0.8737,
+      "step": 777
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.17916975915431976,
+      "learning_rate": 0.00010721353855158243,
+      "loss": 0.8944,
+      "step": 778
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18450039625167847,
+      "learning_rate": 0.00010701890330650632,
+      "loss": 0.9252,
+      "step": 779
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18601484596729279,
+      "learning_rate": 0.0001068242413364671,
+      "loss": 0.9298,
+      "step": 780
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1758352816104889,
+      "learning_rate": 0.0001066295533826537,
+      "loss": 0.8417,
+      "step": 781
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18430499732494354,
+      "learning_rate": 0.00010643484018635417,
+      "loss": 0.8243,
+      "step": 782
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.1788029819726944,
+      "learning_rate": 0.00010624010248895242,
+      "loss": 0.8865,
+      "step": 783
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.18278548121452332,
+      "learning_rate": 0.00010604534103192589,
+      "loss": 0.9258,
+      "step": 784
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.175764799118042,
+      "learning_rate": 0.0001058505565568424,
+      "loss": 0.7853,
+      "step": 785
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17501765489578247,
+      "learning_rate": 0.00010565574980535733,
+      "loss": 0.8697,
+      "step": 786
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17194721102714539,
+      "learning_rate": 0.000105460921519211,
+      "loss": 0.8158,
+      "step": 787
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18962723016738892,
+      "learning_rate": 0.00010526607244022571,
+      "loss": 0.8724,
+      "step": 788
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18698546290397644,
+      "learning_rate": 0.00010507120331030285,
+      "loss": 0.9054,
+      "step": 789
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17789189517498016,
+      "learning_rate": 0.00010487631487142017,
+      "loss": 0.8194,
+      "step": 790
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18768589198589325,
+      "learning_rate": 0.0001046814078656291,
+      "loss": 0.8672,
+      "step": 791
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18617621064186096,
+      "learning_rate": 0.00010448648303505151,
+      "loss": 0.8952,
+      "step": 792
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1749228537082672,
+      "learning_rate": 0.00010429154112187731,
+      "loss": 0.8431,
+      "step": 793
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18151703476905823,
+      "learning_rate": 0.00010409658286836143,
+      "loss": 0.9007,
+      "step": 794
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.18821318447589874,
+      "learning_rate": 0.000103901609016821,
+      "loss": 0.8909,
+      "step": 795
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17506547272205353,
+      "learning_rate": 0.00010370662030963249,
+      "loss": 0.8545,
+      "step": 796
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17080792784690857,
+      "learning_rate": 0.00010351161748922904,
+      "loss": 0.9031,
+      "step": 797
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.16780482232570648,
+      "learning_rate": 0.00010331660129809741,
+      "loss": 0.7929,
+      "step": 798
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.17639389634132385,
+      "learning_rate": 0.00010312157247877538,
+      "loss": 0.8871,
+      "step": 799
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1892055720090866,
+      "learning_rate": 0.00010292653177384876,
+      "loss": 0.9069,
+      "step": 800
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.2034151256084442,
+      "learning_rate": 0.00010273147992594861,
+      "loss": 0.8588,
+      "step": 801
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.1780911684036255,
+      "learning_rate": 0.00010253641767774838,
+      "loss": 0.8366,
+      "step": 802
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18189097940921783,
+      "learning_rate": 0.00010234134577196122,
+      "loss": 0.8897,
+      "step": 803
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17693237960338593,
+      "learning_rate": 0.00010214626495133697,
+      "loss": 0.8325,
+      "step": 804
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16773997247219086,
+      "learning_rate": 0.00010195117595865947,
+      "loss": 0.793,
+      "step": 805
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.19267967343330383,
+      "learning_rate": 0.00010175607953674357,
+      "loss": 0.8647,
+      "step": 806
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18111740052700043,
+      "learning_rate": 0.00010156097642843258,
+      "loss": 0.904,
+      "step": 807
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17377494275569916,
+      "learning_rate": 0.0001013658673765951,
+      "loss": 0.8516,
+      "step": 808
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18672460317611694,
+      "learning_rate": 0.00010117075312412252,
+      "loss": 0.8618,
+      "step": 809
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.20235376060009003,
+      "learning_rate": 0.00010097563441392581,
+      "loss": 0.8585,
+      "step": 810
+    },
+    {
+      "epoch": 0.5,
+      "eval_loss": 1.5616835355758667,
+      "eval_runtime": 187.6128,
+      "eval_samples_per_second": 146.456,
+      "eval_steps_per_second": 73.231,
+      "step": 810
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1777012199163437,
+      "learning_rate": 0.00010078051198893316,
+      "loss": 0.8531,
+      "step": 811
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17983417212963104,
+      "learning_rate": 0.00010058538659208678,
+      "loss": 0.8601,
+      "step": 812
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18771332502365112,
+      "learning_rate": 0.00010039025896634016,
+      "loss": 0.864,
+      "step": 813
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.16632384061813354,
+      "learning_rate": 0.00010019512985465533,
+      "loss": 0.8191,
+      "step": 814
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18960148096084595,
+      "learning_rate": 0.0001,
+      "loss": 0.9061,
+      "step": 815
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.1767115592956543,
+      "learning_rate": 9.980487014534469e-05,
+      "loss": 0.8646,
+      "step": 816
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.18768620491027832,
+      "learning_rate": 9.960974103365986e-05,
+      "loss": 0.9369,
+      "step": 817
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.17243972420692444,
+      "learning_rate": 9.941461340791324e-05,
+      "loss": 0.9104,
+      "step": 818
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18164050579071045,
+      "learning_rate": 9.921948801106682e-05,
+      "loss": 0.8937,
+      "step": 819
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1948123723268509,
+      "learning_rate": 9.90243655860742e-05,
+      "loss": 0.8715,
+      "step": 820
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17371907830238342,
+      "learning_rate": 9.882924687587754e-05,
+      "loss": 0.8865,
+      "step": 821
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.19656354188919067,
+      "learning_rate": 9.863413262340491e-05,
+      "loss": 0.9355,
+      "step": 822
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18487712740898132,
+      "learning_rate": 9.843902357156745e-05,
+      "loss": 0.846,
+      "step": 823
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17178469896316528,
+      "learning_rate": 9.824392046325644e-05,
+      "loss": 0.8831,
+      "step": 824
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.20362038910388947,
+      "learning_rate": 9.804882404134056e-05,
+      "loss": 0.976,
+      "step": 825
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18884702026844025,
+      "learning_rate": 9.785373504866304e-05,
+      "loss": 0.9413,
+      "step": 826
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18527930974960327,
+      "learning_rate": 9.765865422803879e-05,
+      "loss": 0.8458,
+      "step": 827
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18456771969795227,
+      "learning_rate": 9.746358232225163e-05,
+      "loss": 0.8643,
+      "step": 828
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18191008269786835,
+      "learning_rate": 9.726852007405144e-05,
+      "loss": 0.9198,
+      "step": 829
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17499099671840668,
+      "learning_rate": 9.707346822615128e-05,
+      "loss": 0.8724,
+      "step": 830
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.18078701198101044,
+      "learning_rate": 9.687842752122463e-05,
+      "loss": 0.9122,
+      "step": 831
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.1872698813676834,
+      "learning_rate": 9.668339870190261e-05,
+      "loss": 0.8576,
+      "step": 832
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17838625609874725,
+      "learning_rate": 9.648838251077098e-05,
+      "loss": 0.8505,
+      "step": 833
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.17252275347709656,
+      "learning_rate": 9.629337969036752e-05,
+      "loss": 0.8265,
+      "step": 834
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1694963425397873,
+      "learning_rate": 9.6098390983179e-05,
+      "loss": 0.8019,
+      "step": 835
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18931111693382263,
+      "learning_rate": 9.590341713163858e-05,
+      "loss": 0.8416,
+      "step": 836
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18635179102420807,
+      "learning_rate": 9.57084588781227e-05,
+      "loss": 0.8878,
+      "step": 837
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18291349709033966,
+      "learning_rate": 9.551351696494854e-05,
+      "loss": 0.8077,
+      "step": 838
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18153534829616547,
+      "learning_rate": 9.531859213437094e-05,
+      "loss": 0.8647,
+      "step": 839
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.19093500077724457,
+      "learning_rate": 9.512368512857984e-05,
+      "loss": 0.9102,
+      "step": 840
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18069027364253998,
+      "learning_rate": 9.492879668969719e-05,
+      "loss": 0.8495,
+      "step": 841
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18871332705020905,
+      "learning_rate": 9.473392755977431e-05,
+      "loss": 0.8598,
+      "step": 842
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18910835683345795,
+      "learning_rate": 9.453907848078902e-05,
+      "loss": 0.8441,
+      "step": 843
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1749962866306305,
+      "learning_rate": 9.434425019464268e-05,
+      "loss": 0.842,
+      "step": 844
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1805962324142456,
+      "learning_rate": 9.414944344315764e-05,
+      "loss": 0.8393,
+      "step": 845
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.1837007999420166,
+      "learning_rate": 9.39546589680741e-05,
+      "loss": 0.8847,
+      "step": 846
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.18500761687755585,
+      "learning_rate": 9.37598975110476e-05,
+      "loss": 0.9155,
+      "step": 847
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.184367835521698,
+      "learning_rate": 9.356515981364588e-05,
+      "loss": 0.8879,
+      "step": 848
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.16617724299430847,
+      "learning_rate": 9.33704466173463e-05,
+      "loss": 0.7774,
+      "step": 849
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.17406000196933746,
+      "learning_rate": 9.317575866353292e-05,
+      "loss": 0.8896,
+      "step": 850
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.17451772093772888,
+      "learning_rate": 9.298109669349369e-05,
+      "loss": 0.8272,
+      "step": 851
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18715828657150269,
+      "learning_rate": 9.278646144841758e-05,
+      "loss": 0.8964,
+      "step": 852
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.181923970580101,
+      "learning_rate": 9.259185366939191e-05,
+      "loss": 0.8875,
+      "step": 853
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19103018939495087,
+      "learning_rate": 9.23972740973993e-05,
+      "loss": 0.8872,
+      "step": 854
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18528901040554047,
+      "learning_rate": 9.220272347331502e-05,
+      "loss": 0.8195,
+      "step": 855
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1798166185617447,
+      "learning_rate": 9.200820253790419e-05,
+      "loss": 0.8676,
+      "step": 856
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19125676155090332,
+      "learning_rate": 9.181371203181872e-05,
+      "loss": 0.899,
+      "step": 857
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18180611729621887,
+      "learning_rate": 9.161925269559479e-05,
+      "loss": 0.8844,
+      "step": 858
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18726137280464172,
+      "learning_rate": 9.14248252696499e-05,
+      "loss": 0.8079,
+      "step": 859
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18223150074481964,
+      "learning_rate": 9.123043049427995e-05,
+      "loss": 0.8821,
+      "step": 860
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.8702235817909241,
+      "learning_rate": 9.103606910965666e-05,
+      "loss": 0.9003,
+      "step": 861
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1906432956457138,
+      "learning_rate": 9.084174185582449e-05,
+      "loss": 0.9477,
+      "step": 862
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18080779910087585,
+      "learning_rate": 9.064744947269801e-05,
+      "loss": 0.8484,
+      "step": 863
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.1770569533109665,
+      "learning_rate": 9.0453192700059e-05,
+      "loss": 0.8382,
+      "step": 864
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.18079295754432678,
+      "learning_rate": 9.025897227755366e-05,
+      "loss": 0.8459,
+      "step": 865
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.19838999211788177,
+      "learning_rate": 9.006478894468973e-05,
+      "loss": 0.8576,
+      "step": 866
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19499678909778595,
+      "learning_rate": 8.987064344083381e-05,
+      "loss": 0.9357,
+      "step": 867
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1933078169822693,
+      "learning_rate": 8.967653650520841e-05,
+      "loss": 0.9509,
+      "step": 868
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.19150817394256592,
+      "learning_rate": 8.948246887688924e-05,
+      "loss": 0.8311,
+      "step": 869
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18107332289218903,
+      "learning_rate": 8.928844129480227e-05,
+      "loss": 0.805,
+      "step": 870
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17423181235790253,
+      "learning_rate": 8.909445449772102e-05,
+      "loss": 0.8464,
+      "step": 871
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1880725622177124,
+      "learning_rate": 8.890050922426381e-05,
+      "loss": 0.9009,
+      "step": 872
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17800618708133698,
+      "learning_rate": 8.870660621289066e-05,
+      "loss": 0.8457,
+      "step": 873
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.17739921808242798,
+      "learning_rate": 8.851274620190092e-05,
+      "loss": 0.8446,
+      "step": 874
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.21376144886016846,
+      "learning_rate": 8.831892992942999e-05,
+      "loss": 0.9154,
+      "step": 875
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.20101210474967957,
+      "learning_rate": 8.812515813344682e-05,
+      "loss": 0.8379,
+      "step": 876
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1853225827217102,
+      "learning_rate": 8.793143155175108e-05,
+      "loss": 0.7938,
+      "step": 877
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.1721082478761673,
+      "learning_rate": 8.773775092197017e-05,
+      "loss": 0.8137,
+      "step": 878
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.176341250538826,
+      "learning_rate": 8.754411698155658e-05,
+      "loss": 0.8465,
+      "step": 879
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18818019330501556,
+      "learning_rate": 8.735053046778506e-05,
+      "loss": 0.8539,
+      "step": 880
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18352611362934113,
+      "learning_rate": 8.71569921177497e-05,
+      "loss": 0.8382,
+      "step": 881
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18330875039100647,
+      "learning_rate": 8.696350266836128e-05,
+      "loss": 0.8411,
+      "step": 882
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.18866685032844543,
+      "learning_rate": 8.677006285634438e-05,
+      "loss": 0.8386,
+      "step": 883
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.200348362326622,
+      "learning_rate": 8.657667341823448e-05,
+      "loss": 0.892,
+      "step": 884
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17930357158184052,
+      "learning_rate": 8.638333509037536e-05,
+      "loss": 0.8692,
+      "step": 885
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18584062159061432,
+      "learning_rate": 8.619004860891618e-05,
+      "loss": 0.8348,
+      "step": 886
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17948465049266815,
+      "learning_rate": 8.599681470980867e-05,
+      "loss": 0.861,
+      "step": 887
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1800021380186081,
+      "learning_rate": 8.580363412880439e-05,
+      "loss": 0.9089,
+      "step": 888
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.17282156646251678,
+      "learning_rate": 8.56105076014518e-05,
+      "loss": 0.8524,
+      "step": 889
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1849554032087326,
+      "learning_rate": 8.541743586309365e-05,
+      "loss": 0.8621,
+      "step": 890
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18419396877288818,
+      "learning_rate": 8.522441964886403e-05,
+      "loss": 0.8367,
+      "step": 891
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1823270469903946,
+      "learning_rate": 8.503145969368562e-05,
+      "loss": 0.8002,
+      "step": 892
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18402813374996185,
+      "learning_rate": 8.48385567322668e-05,
+      "loss": 0.8401,
+      "step": 893
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1967848539352417,
+      "learning_rate": 8.464571149909914e-05,
+      "loss": 0.8983,
+      "step": 894
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.19050410389900208,
+      "learning_rate": 8.445292472845423e-05,
+      "loss": 0.8788,
+      "step": 895
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18483230471611023,
+      "learning_rate": 8.426019715438116e-05,
+      "loss": 0.7938,
+      "step": 896
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.1833154261112213,
+      "learning_rate": 8.406752951070359e-05,
+      "loss": 0.8859,
+      "step": 897
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18262380361557007,
+      "learning_rate": 8.387492253101695e-05,
+      "loss": 0.8383,
+      "step": 898
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.18968260288238525,
+      "learning_rate": 8.36823769486858e-05,
+      "loss": 0.9117,
+      "step": 899
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17808446288108826,
+      "learning_rate": 8.348989349684076e-05,
+      "loss": 0.8436,
+      "step": 900
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19946736097335815,
+      "learning_rate": 8.32974729083761e-05,
+      "loss": 0.8765,
+      "step": 901
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18614915013313293,
+      "learning_rate": 8.31051159159465e-05,
+      "loss": 0.9165,
+      "step": 902
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18090373277664185,
+      "learning_rate": 8.291282325196464e-05,
+      "loss": 0.9103,
+      "step": 903
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17819440364837646,
+      "learning_rate": 8.272059564859826e-05,
+      "loss": 0.9133,
+      "step": 904
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1872352659702301,
+      "learning_rate": 8.252843383776731e-05,
+      "loss": 0.8551,
+      "step": 905
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18670712411403656,
+      "learning_rate": 8.233633855114127e-05,
+      "loss": 0.9836,
+      "step": 906
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.2019394189119339,
+      "learning_rate": 8.214431052013634e-05,
+      "loss": 0.9269,
+      "step": 907
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.1772841513156891,
+      "learning_rate": 8.19523504759126e-05,
+      "loss": 0.8471,
+      "step": 908
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.17724165320396423,
+      "learning_rate": 8.176045914937133e-05,
+      "loss": 0.8407,
+      "step": 909
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18769262731075287,
+      "learning_rate": 8.156863727115211e-05,
+      "loss": 0.824,
+      "step": 910
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18256652355194092,
+      "learning_rate": 8.137688557163004e-05,
+      "loss": 0.8981,
+      "step": 911
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18411476910114288,
+      "learning_rate": 8.118520478091311e-05,
+      "loss": 0.8433,
+      "step": 912
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18739469349384308,
+      "learning_rate": 8.09935956288393e-05,
+      "loss": 0.9297,
+      "step": 913
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.19190461933612823,
+      "learning_rate": 8.080205884497375e-05,
+      "loss": 0.8467,
+      "step": 914
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.18624301254749298,
+      "learning_rate": 8.061059515860616e-05,
+      "loss": 0.9278,
+      "step": 915
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18778246641159058,
+      "learning_rate": 8.04192052987478e-05,
+      "loss": 0.8776,
+      "step": 916
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18942977488040924,
+      "learning_rate": 8.022788999412893e-05,
+      "loss": 0.8874,
+      "step": 917
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18389110267162323,
+      "learning_rate": 8.003664997319589e-05,
+      "loss": 0.839,
+      "step": 918
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18656383454799652,
+      "learning_rate": 7.984548596410834e-05,
+      "loss": 0.8718,
+      "step": 919
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17733711004257202,
+      "learning_rate": 7.965439869473664e-05,
+      "loss": 0.8573,
+      "step": 920
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18599346280097961,
+      "learning_rate": 7.946338889265879e-05,
+      "loss": 0.8077,
+      "step": 921
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1774177998304367,
+      "learning_rate": 7.927245728515794e-05,
+      "loss": 0.8324,
+      "step": 922
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1771276295185089,
+      "learning_rate": 7.908160459921948e-05,
+      "loss": 0.8895,
+      "step": 923
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17772090435028076,
+      "learning_rate": 7.889083156152831e-05,
+      "loss": 0.7986,
+      "step": 924
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17372700572013855,
+      "learning_rate": 7.870013889846608e-05,
+      "loss": 0.8645,
+      "step": 925
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.17801426351070404,
+      "learning_rate": 7.850952733610836e-05,
+      "loss": 0.8512,
+      "step": 926
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18646259605884552,
+      "learning_rate": 7.831899760022192e-05,
+      "loss": 0.8423,
+      "step": 927
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.2032555192708969,
+      "learning_rate": 7.812855041626206e-05,
+      "loss": 0.9032,
+      "step": 928
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.18744565546512604,
+      "learning_rate": 7.793818650936969e-05,
+      "loss": 0.8421,
+      "step": 929
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.19461603462696075,
+      "learning_rate": 7.774790660436858e-05,
+      "loss": 0.8681,
+      "step": 930
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.1867326945066452,
+      "learning_rate": 7.755771142576276e-05,
+      "loss": 0.8187,
+      "step": 931
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18859319388866425,
+      "learning_rate": 7.736760169773362e-05,
+      "loss": 0.8448,
+      "step": 932
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17433106899261475,
+      "learning_rate": 7.717757814413718e-05,
+      "loss": 0.8092,
+      "step": 933
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18000973761081696,
+      "learning_rate": 7.698764148850137e-05,
+      "loss": 0.8407,
+      "step": 934
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18358999490737915,
+      "learning_rate": 7.679779245402321e-05,
+      "loss": 0.8803,
+      "step": 935
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18168945610523224,
+      "learning_rate": 7.660803176356615e-05,
+      "loss": 0.8424,
+      "step": 936
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.1922016590833664,
+      "learning_rate": 7.641836013965719e-05,
+      "loss": 0.9138,
+      "step": 937
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.20511652529239655,
+      "learning_rate": 7.622877830448434e-05,
+      "loss": 0.906,
+      "step": 938
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18061776459217072,
+      "learning_rate": 7.603928697989352e-05,
+      "loss": 0.8846,
+      "step": 939
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.17372047901153564,
+      "learning_rate": 7.584988688738622e-05,
+      "loss": 0.896,
+      "step": 940
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18462759256362915,
+      "learning_rate": 7.566057874811642e-05,
+      "loss": 0.9125,
+      "step": 941
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18168705701828003,
+      "learning_rate": 7.547136328288814e-05,
+      "loss": 0.8952,
+      "step": 942
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18722060322761536,
+      "learning_rate": 7.528224121215235e-05,
+      "loss": 0.9264,
+      "step": 943
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18432766199111938,
+      "learning_rate": 7.509321325600456e-05,
+      "loss": 0.8485,
+      "step": 944
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18999868631362915,
+      "learning_rate": 7.490428013418187e-05,
+      "loss": 0.9213,
+      "step": 945
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.18130837380886078,
+      "learning_rate": 7.471544256606026e-05,
+      "loss": 0.9068,
+      "step": 946
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.19146692752838135,
+      "learning_rate": 7.452670127065202e-05,
+      "loss": 0.8527,
+      "step": 947
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17760176956653595,
+      "learning_rate": 7.433805696660266e-05,
+      "loss": 0.8655,
+      "step": 948
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1794421225786209,
+      "learning_rate": 7.414951037218853e-05,
+      "loss": 0.8332,
+      "step": 949
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18185631930828094,
+      "learning_rate": 7.396106220531398e-05,
+      "loss": 0.8691,
+      "step": 950
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18275043368339539,
+      "learning_rate": 7.377271318350846e-05,
+      "loss": 0.8473,
+      "step": 951
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1896333247423172,
+      "learning_rate": 7.358446402392405e-05,
+      "loss": 0.9124,
+      "step": 952
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17494642734527588,
+      "learning_rate": 7.339631544333249e-05,
+      "loss": 0.9173,
+      "step": 953
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18608129024505615,
+      "learning_rate": 7.320826815812261e-05,
+      "loss": 0.8215,
+      "step": 954
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18127049505710602,
+      "learning_rate": 7.302032288429756e-05,
+      "loss": 0.8354,
+      "step": 955
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1783689260482788,
+      "learning_rate": 7.283248033747209e-05,
+      "loss": 0.7718,
+      "step": 956
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.17510481178760529,
+      "learning_rate": 7.26447412328697e-05,
+      "loss": 0.8365,
+      "step": 957
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1876639574766159,
+      "learning_rate": 7.245710628532015e-05,
+      "loss": 0.9265,
+      "step": 958
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18069513142108917,
+      "learning_rate": 7.226957620925657e-05,
+      "loss": 0.8339,
+      "step": 959
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.18527735769748688,
+      "learning_rate": 7.208215171871277e-05,
+      "loss": 0.8808,
+      "step": 960
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1924070119857788,
+      "learning_rate": 7.189483352732056e-05,
+      "loss": 0.9016,
+      "step": 961
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19486652314662933,
+      "learning_rate": 7.170762234830699e-05,
+      "loss": 0.8891,
+      "step": 962
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.1852908879518509,
+      "learning_rate": 7.15205188944917e-05,
+      "loss": 0.8307,
+      "step": 963
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.19809791445732117,
+      "learning_rate": 7.133352387828407e-05,
+      "loss": 0.871,
+      "step": 964
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.193113312125206,
+      "learning_rate": 7.114663801168072e-05,
+      "loss": 0.8938,
+      "step": 965
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17784567177295685,
+      "learning_rate": 7.09598620062625e-05,
+      "loss": 0.8464,
+      "step": 966
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18583837151527405,
+      "learning_rate": 7.077319657319213e-05,
+      "loss": 0.9161,
+      "step": 967
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17241233587265015,
+      "learning_rate": 7.05866424232112e-05,
+      "loss": 0.8053,
+      "step": 968
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17903415858745575,
+      "learning_rate": 7.040020026663767e-05,
+      "loss": 0.8749,
+      "step": 969
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1805238574743271,
+      "learning_rate": 7.021387081336301e-05,
+      "loss": 0.8751,
+      "step": 970
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.1911240667104721,
+      "learning_rate": 7.002765477284961e-05,
+      "loss": 0.8382,
+      "step": 971
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19228386878967285,
+      "learning_rate": 6.9841552854128e-05,
+      "loss": 0.7998,
+      "step": 972
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19767771661281586,
+      "learning_rate": 6.96555657657942e-05,
+      "loss": 0.9781,
+      "step": 973
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19372697174549103,
+      "learning_rate": 6.946969421600702e-05,
+      "loss": 0.8438,
+      "step": 974
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19649174809455872,
+      "learning_rate": 6.92839389124853e-05,
+      "loss": 0.927,
+      "step": 975
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.20114201307296753,
+      "learning_rate": 6.909830056250527e-05,
+      "loss": 0.8821,
+      "step": 976
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18612854182720184,
+      "learning_rate": 6.891277987289794e-05,
+      "loss": 0.8875,
+      "step": 977
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.18017785251140594,
+      "learning_rate": 6.87273775500462e-05,
+      "loss": 0.8627,
+      "step": 978
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.19493426382541656,
+      "learning_rate": 6.854209429988235e-05,
+      "loss": 0.9042,
+      "step": 979
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.17783161997795105,
+      "learning_rate": 6.835693082788525e-05,
+      "loss": 0.8237,
+      "step": 980
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18601271510124207,
+      "learning_rate": 6.817188783907771e-05,
+      "loss": 0.8596,
+      "step": 981
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.19375163316726685,
+      "learning_rate": 6.798696603802381e-05,
+      "loss": 0.8528,
+      "step": 982
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18276815116405487,
+      "learning_rate": 6.780216612882619e-05,
+      "loss": 0.857,
+      "step": 983
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1879137009382248,
+      "learning_rate": 6.761748881512333e-05,
+      "loss": 0.8943,
+      "step": 984
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18253687024116516,
+      "learning_rate": 6.743293480008702e-05,
+      "loss": 0.8154,
+      "step": 985
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.17798271775245667,
+      "learning_rate": 6.72485047864195e-05,
+      "loss": 0.8721,
+      "step": 986
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18498069047927856,
+      "learning_rate": 6.706419947635086e-05,
+      "loss": 0.9134,
+      "step": 987
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18797878921031952,
+      "learning_rate": 6.688001957163646e-05,
+      "loss": 0.8378,
+      "step": 988
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1864275187253952,
+      "learning_rate": 6.669596577355408e-05,
+      "loss": 0.8796,
+      "step": 989
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1926390677690506,
+      "learning_rate": 6.651203878290139e-05,
+      "loss": 0.8927,
+      "step": 990
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1841459721326828,
+      "learning_rate": 6.632823929999318e-05,
+      "loss": 0.8988,
+      "step": 991
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18821528553962708,
+      "learning_rate": 6.61445680246589e-05,
+      "loss": 0.9059,
+      "step": 992
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1828063279390335,
+      "learning_rate": 6.596102565623954e-05,
+      "loss": 0.838,
+      "step": 993
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.18085777759552002,
+      "learning_rate": 6.577761289358558e-05,
+      "loss": 0.852,
+      "step": 994
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.1909693330526352,
+      "learning_rate": 6.559433043505383e-05,
+      "loss": 0.8746,
+      "step": 995
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.20024698972702026,
+      "learning_rate": 6.541117897850505e-05,
+      "loss": 0.8951,
+      "step": 996
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18915902078151703,
+      "learning_rate": 6.522815922130112e-05,
+      "loss": 0.9514,
+      "step": 997
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1937555968761444,
+      "learning_rate": 6.504527186030258e-05,
+      "loss": 0.8846,
+      "step": 998
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1910788118839264,
+      "learning_rate": 6.486251759186572e-05,
+      "loss": 0.8529,
+      "step": 999
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19415101408958435,
+      "learning_rate": 6.46798971118402e-05,
+      "loss": 0.8985,
+      "step": 1000
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18411551415920258,
+      "learning_rate": 6.449741111556622e-05,
+      "loss": 0.7691,
+      "step": 1001
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1851988434791565,
+      "learning_rate": 6.431506029787189e-05,
+      "loss": 0.8188,
+      "step": 1002
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18264465034008026,
+      "learning_rate": 6.413284535307061e-05,
+      "loss": 0.8604,
+      "step": 1003
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17648856341838837,
+      "learning_rate": 6.395076697495854e-05,
+      "loss": 0.8573,
+      "step": 1004
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18338896334171295,
+      "learning_rate": 6.376882585681175e-05,
+      "loss": 0.9133,
+      "step": 1005
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.17458754777908325,
+      "learning_rate": 6.358702269138375e-05,
+      "loss": 0.8653,
+      "step": 1006
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18256697058677673,
+      "learning_rate": 6.340535817090272e-05,
+      "loss": 0.8387,
+      "step": 1007
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19982388615608215,
+      "learning_rate": 6.322383298706897e-05,
+      "loss": 0.8984,
+      "step": 1008
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.19646942615509033,
+      "learning_rate": 6.304244783105231e-05,
+      "loss": 0.9245,
+      "step": 1009
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1775049865245819,
+      "learning_rate": 6.286120339348935e-05,
+      "loss": 0.8543,
+      "step": 1010
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.18325653672218323,
+      "learning_rate": 6.268010036448088e-05,
+      "loss": 0.8912,
+      "step": 1011
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.1884334534406662,
+      "learning_rate": 6.249913943358932e-05,
+      "loss": 0.8339,
+      "step": 1012
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1827017217874527,
+      "learning_rate": 6.231832128983599e-05,
+      "loss": 0.8002,
+      "step": 1013
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18097549676895142,
+      "learning_rate": 6.213764662169859e-05,
+      "loss": 0.8099,
+      "step": 1014
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17326505482196808,
+      "learning_rate": 6.19571161171085e-05,
+      "loss": 0.8476,
+      "step": 1015
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18046098947525024,
+      "learning_rate": 6.177673046344816e-05,
+      "loss": 0.8282,
+      "step": 1016
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1849922090768814,
+      "learning_rate": 6.159649034754857e-05,
+      "loss": 0.8582,
+      "step": 1017
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1950290948152542,
+      "learning_rate": 6.141639645568646e-05,
+      "loss": 0.8936,
+      "step": 1018
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1874220222234726,
+      "learning_rate": 6.123644947358191e-05,
+      "loss": 0.8371,
+      "step": 1019
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17714416980743408,
+      "learning_rate": 6.105665008639557e-05,
+      "loss": 0.847,
+      "step": 1020
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18766450881958008,
+      "learning_rate": 6.087699897872612e-05,
+      "loss": 0.8953,
+      "step": 1021
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1957690417766571,
+      "learning_rate": 6.069749683460765e-05,
+      "loss": 0.8899,
+      "step": 1022
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1880084127187729,
+      "learning_rate": 6.051814433750711e-05,
+      "loss": 0.8939,
+      "step": 1023
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.179568350315094,
+      "learning_rate": 6.033894217032159e-05,
+      "loss": 0.8272,
+      "step": 1024
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.17950114607810974,
+      "learning_rate": 6.015989101537586e-05,
+      "loss": 0.8771,
+      "step": 1025
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1966513693332672,
+      "learning_rate": 5.998099155441962e-05,
+      "loss": 0.8223,
+      "step": 1026
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1985529661178589,
+      "learning_rate": 5.9802244468625066e-05,
+      "loss": 0.8943,
+      "step": 1027
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.1800307184457779,
+      "learning_rate": 5.962365043858418e-05,
+      "loss": 0.8055,
+      "step": 1028
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.18859480321407318,
+      "learning_rate": 5.944521014430613e-05,
+      "loss": 0.8545,
+      "step": 1029
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18779581785202026,
+      "learning_rate": 5.926692426521474e-05,
+      "loss": 0.8365,
+      "step": 1030
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19234032928943634,
+      "learning_rate": 5.9088793480145996e-05,
+      "loss": 0.847,
+      "step": 1031
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1970566362142563,
+      "learning_rate": 5.8910818467345185e-05,
+      "loss": 0.8712,
+      "step": 1032
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1854643076658249,
+      "learning_rate": 5.8732999904464626e-05,
+      "loss": 0.845,
+      "step": 1033
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.1871088147163391,
+      "learning_rate": 5.855533846856085e-05,
+      "loss": 0.8454,
+      "step": 1034
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18273068964481354,
+      "learning_rate": 5.8377834836092136e-05,
+      "loss": 0.8508,
+      "step": 1035
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18777170777320862,
+      "learning_rate": 5.820048968291596e-05,
+      "loss": 0.8576,
+      "step": 1036
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18084120750427246,
+      "learning_rate": 5.8023303684286326e-05,
+      "loss": 0.7909,
+      "step": 1037
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19652147591114044,
+      "learning_rate": 5.7846277514851256e-05,
+      "loss": 0.8487,
+      "step": 1038
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17516165971755981,
+      "learning_rate": 5.7669411848650235e-05,
+      "loss": 0.8129,
+      "step": 1039
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18353667855262756,
+      "learning_rate": 5.749270735911158e-05,
+      "loss": 0.8634,
+      "step": 1040
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.19190074503421783,
+      "learning_rate": 5.731616471904998e-05,
+      "loss": 0.8668,
+      "step": 1041
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18586887419223785,
+      "learning_rate": 5.713978460066376e-05,
+      "loss": 0.8717,
+      "step": 1042
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17346972227096558,
+      "learning_rate": 5.696356767553257e-05,
+      "loss": 0.8177,
+      "step": 1043
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.17980901896953583,
+      "learning_rate": 5.6787514614614624e-05,
+      "loss": 0.8369,
+      "step": 1044
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.18174031376838684,
+      "learning_rate": 5.6611626088244194e-05,
+      "loss": 0.7948,
+      "step": 1045
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1818779855966568,
+      "learning_rate": 5.643590276612909e-05,
+      "loss": 0.8911,
+      "step": 1046
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1876106858253479,
+      "learning_rate": 5.6260345317348114e-05,
+      "loss": 0.9046,
+      "step": 1047
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18339186906814575,
+      "learning_rate": 5.6084954410348455e-05,
+      "loss": 0.7988,
+      "step": 1048
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1841948926448822,
+      "learning_rate": 5.590973071294322e-05,
+      "loss": 0.8826,
+      "step": 1049
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19951540231704712,
+      "learning_rate": 5.573467489230879e-05,
+      "loss": 0.9122,
+      "step": 1050
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18917638063430786,
+      "learning_rate": 5.5559787614982505e-05,
+      "loss": 0.8222,
+      "step": 1051
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.20554791390895844,
+      "learning_rate": 5.5385069546859766e-05,
+      "loss": 0.8833,
+      "step": 1052
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19960781931877136,
+      "learning_rate": 5.521052135319182e-05,
+      "loss": 0.9103,
+      "step": 1053
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.1844080239534378,
+      "learning_rate": 5.503614369858301e-05,
+      "loss": 0.8335,
+      "step": 1054
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18848901987075806,
+      "learning_rate": 5.486193724698854e-05,
+      "loss": 0.8903,
+      "step": 1055
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18261469900608063,
+      "learning_rate": 5.468790266171146e-05,
+      "loss": 0.8495,
+      "step": 1056
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19197224080562592,
+      "learning_rate": 5.451404060540064e-05,
+      "loss": 0.8688,
+      "step": 1057
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19706493616104126,
+      "learning_rate": 5.4340351740047914e-05,
+      "loss": 0.8399,
+      "step": 1058
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.18674053251743317,
+      "learning_rate": 5.416683672698582e-05,
+      "loss": 0.82,
+      "step": 1059
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19017022848129272,
+      "learning_rate": 5.399349622688479e-05,
+      "loss": 0.861,
+      "step": 1060
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.19184957444667816,
+      "learning_rate": 5.3820330899750895e-05,
+      "loss": 0.9052,
+      "step": 1061
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18365278840065002,
+      "learning_rate": 5.3647341404923134e-05,
+      "loss": 0.8791,
+      "step": 1062
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1919332593679428,
+      "learning_rate": 5.347452840107104e-05,
+      "loss": 0.8496,
+      "step": 1063
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18335145711898804,
+      "learning_rate": 5.3301892546192266e-05,
+      "loss": 0.8511,
+      "step": 1064
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17815229296684265,
+      "learning_rate": 5.312943449760975e-05,
+      "loss": 0.8611,
+      "step": 1065
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.1748436540365219,
+      "learning_rate": 5.295715491196951e-05,
+      "loss": 0.8342,
+      "step": 1066
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17804138362407684,
+      "learning_rate": 5.278505444523816e-05,
+      "loss": 0.8258,
+      "step": 1067
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18960395455360413,
+      "learning_rate": 5.261313375270014e-05,
+      "loss": 0.8422,
+      "step": 1068
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18084166944026947,
+      "learning_rate": 5.244139348895549e-05,
+      "loss": 0.8276,
+      "step": 1069
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17871583998203278,
+      "learning_rate": 5.226983430791722e-05,
+      "loss": 0.8411,
+      "step": 1070
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18520328402519226,
+      "learning_rate": 5.209845686280881e-05,
+      "loss": 0.8441,
+      "step": 1071
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18111847341060638,
+      "learning_rate": 5.192726180616191e-05,
+      "loss": 0.8352,
+      "step": 1072
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18117675185203552,
+      "learning_rate": 5.175624978981361e-05,
+      "loss": 0.837,
+      "step": 1073
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18395110964775085,
+      "learning_rate": 5.1585421464903994e-05,
+      "loss": 0.86,
+      "step": 1074
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.19681338965892792,
+      "learning_rate": 5.1414777481873796e-05,
+      "loss": 0.813,
+      "step": 1075
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.17966072261333466,
+      "learning_rate": 5.1244318490461976e-05,
+      "loss": 0.869,
+      "step": 1076
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.18585649132728577,
+      "learning_rate": 5.1074045139702956e-05,
+      "loss": 0.8669,
+      "step": 1077
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17946961522102356,
+      "learning_rate": 5.0903958077924366e-05,
+      "loss": 0.7999,
+      "step": 1078
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18911674618721008,
+      "learning_rate": 5.0734057952744504e-05,
+      "loss": 0.8712,
+      "step": 1079
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1970946043729782,
+      "learning_rate": 5.0564345411070025e-05,
+      "loss": 0.8436,
+      "step": 1080
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18350134789943695,
+      "learning_rate": 5.039482109909319e-05,
+      "loss": 0.8519,
+      "step": 1081
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.17878840863704681,
+      "learning_rate": 5.022548566228968e-05,
+      "loss": 0.8283,
+      "step": 1082
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18846851587295532,
+      "learning_rate": 5.005633974541584e-05,
+      "loss": 0.8887,
+      "step": 1083
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18276073038578033,
+      "learning_rate": 4.9887383992506634e-05,
+      "loss": 0.8586,
+      "step": 1084
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18765556812286377,
+      "learning_rate": 4.9718619046872825e-05,
+      "loss": 0.9009,
+      "step": 1085
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18436139822006226,
+      "learning_rate": 4.955004555109871e-05,
+      "loss": 0.7971,
+      "step": 1086
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1878422349691391,
+      "learning_rate": 4.938166414703958e-05,
+      "loss": 0.8467,
+      "step": 1087
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.1810942143201828,
+      "learning_rate": 4.921347547581939e-05,
+      "loss": 0.8353,
+      "step": 1088
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18749824166297913,
+      "learning_rate": 4.904548017782823e-05,
+      "loss": 0.8768,
+      "step": 1089
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19052742421627045,
+      "learning_rate": 4.8877678892719866e-05,
+      "loss": 0.8592,
+      "step": 1090
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19717518985271454,
+      "learning_rate": 4.87100722594094e-05,
+      "loss": 0.8945,
+      "step": 1091
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.19465872645378113,
+      "learning_rate": 4.8542660916070736e-05,
+      "loss": 0.8634,
+      "step": 1092
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.18439744412899017,
+      "learning_rate": 4.837544550013424e-05,
+      "loss": 0.8316,
+      "step": 1093
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1906382441520691,
+      "learning_rate": 4.8208426648284244e-05,
+      "loss": 0.8599,
+      "step": 1094
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18799233436584473,
+      "learning_rate": 4.804160499645667e-05,
+      "loss": 0.9088,
+      "step": 1095
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19213566184043884,
+      "learning_rate": 4.787498117983653e-05,
+      "loss": 0.8153,
+      "step": 1096
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18374043703079224,
+      "learning_rate": 4.7708555832855685e-05,
+      "loss": 0.8469,
+      "step": 1097
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1865880936384201,
+      "learning_rate": 4.7542329589190195e-05,
+      "loss": 0.8626,
+      "step": 1098
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18128207325935364,
+      "learning_rate": 4.737630308175806e-05,
+      "loss": 0.8925,
+      "step": 1099
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1784657984972,
+      "learning_rate": 4.721047694271676e-05,
+      "loss": 0.829,
+      "step": 1100
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18296921253204346,
+      "learning_rate": 4.704485180346089e-05,
+      "loss": 0.827,
+      "step": 1101
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18465541303157806,
+      "learning_rate": 4.687942829461969e-05,
+      "loss": 0.8943,
+      "step": 1102
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18593785166740417,
+      "learning_rate": 4.671420704605469e-05,
+      "loss": 0.8481,
+      "step": 1103
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18125657737255096,
+      "learning_rate": 4.654918868685726e-05,
+      "loss": 0.8117,
+      "step": 1104
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19158883392810822,
+      "learning_rate": 4.638437384534637e-05,
+      "loss": 0.8634,
+      "step": 1105
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.19072219729423523,
+      "learning_rate": 4.6219763149065974e-05,
+      "loss": 0.8288,
+      "step": 1106
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1918463259935379,
+      "learning_rate": 4.6055357224782766e-05,
+      "loss": 0.9037,
+      "step": 1107
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1800980418920517,
+      "learning_rate": 4.5891156698483716e-05,
+      "loss": 0.8725,
+      "step": 1108
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.1854526400566101,
+      "learning_rate": 4.572716219537385e-05,
+      "loss": 0.9072,
+      "step": 1109
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.18771405518054962,
+      "learning_rate": 4.556337433987359e-05,
+      "loss": 0.9138,
+      "step": 1110
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18725012242794037,
+      "learning_rate": 4.539979375561659e-05,
+      "loss": 0.9044,
+      "step": 1111
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1886708289384842,
+      "learning_rate": 4.523642106544732e-05,
+      "loss": 0.8482,
+      "step": 1112
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1814701110124588,
+      "learning_rate": 4.5073256891418735e-05,
+      "loss": 0.8768,
+      "step": 1113
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.2009040266275406,
+      "learning_rate": 4.491030185478976e-05,
+      "loss": 0.9422,
+      "step": 1114
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1815730780363083,
+      "learning_rate": 4.474755657602303e-05,
+      "loss": 0.8793,
+      "step": 1115
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1851624846458435,
+      "learning_rate": 4.4585021674782534e-05,
+      "loss": 0.8189,
+      "step": 1116
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1785864233970642,
+      "learning_rate": 4.4422697769931207e-05,
+      "loss": 0.8607,
+      "step": 1117
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1833125501871109,
+      "learning_rate": 4.426058547952867e-05,
+      "loss": 0.8437,
+      "step": 1118
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18532410264015198,
+      "learning_rate": 4.409868542082878e-05,
+      "loss": 0.8564,
+      "step": 1119
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1779632568359375,
+      "learning_rate": 4.393699821027716e-05,
+      "loss": 0.8329,
+      "step": 1120
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18608425557613373,
+      "learning_rate": 4.377552446350924e-05,
+      "loss": 0.7956,
+      "step": 1121
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18744583427906036,
+      "learning_rate": 4.361426479534753e-05,
+      "loss": 0.8915,
+      "step": 1122
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.18702028691768646,
+      "learning_rate": 4.345321981979942e-05,
+      "loss": 0.8243,
+      "step": 1123
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.1835690289735794,
+      "learning_rate": 4.3292390150054894e-05,
+      "loss": 0.8037,
+      "step": 1124
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.17932282388210297,
+      "learning_rate": 4.313177639848408e-05,
+      "loss": 0.8588,
+      "step": 1125
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.19243797659873962,
+      "learning_rate": 4.2971379176635097e-05,
+      "loss": 0.824,
+      "step": 1126
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.191883385181427,
+      "learning_rate": 4.281119909523149e-05,
+      "loss": 0.9015,
+      "step": 1127
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1800529509782791,
+      "learning_rate": 4.265123676417008e-05,
+      "loss": 0.8358,
+      "step": 1128
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19045579433441162,
+      "learning_rate": 4.249149279251858e-05,
+      "loss": 0.8996,
+      "step": 1129
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1921454817056656,
+      "learning_rate": 4.2331967788513295e-05,
+      "loss": 0.9056,
+      "step": 1130
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18252025544643402,
+      "learning_rate": 4.217266235955678e-05,
+      "loss": 0.8314,
+      "step": 1131
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.20346395671367645,
+      "learning_rate": 4.2013577112215566e-05,
+      "loss": 0.8989,
+      "step": 1132
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18643800914287567,
+      "learning_rate": 4.1854712652217745e-05,
+      "loss": 0.8771,
+      "step": 1133
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1962081640958786,
+      "learning_rate": 4.16960695844509e-05,
+      "loss": 0.8448,
+      "step": 1134
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18850386142730713,
+      "learning_rate": 4.153764851295954e-05,
+      "loss": 0.8535,
+      "step": 1135
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18714487552642822,
+      "learning_rate": 4.1379450040942905e-05,
+      "loss": 0.8879,
+      "step": 1136
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.17667607963085175,
+      "learning_rate": 4.12214747707527e-05,
+      "loss": 0.8011,
+      "step": 1137
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.19399698078632355,
+      "learning_rate": 4.1063723303890763e-05,
+      "loss": 0.9516,
+      "step": 1138
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.18556344509124756,
+      "learning_rate": 4.0906196241006814e-05,
+      "loss": 0.889,
+      "step": 1139
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1827767789363861,
+      "learning_rate": 4.074889418189608e-05,
+      "loss": 0.8489,
+      "step": 1140
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1933080106973648,
+      "learning_rate": 4.05918177254971e-05,
+      "loss": 0.848,
+      "step": 1141
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.1840435117483139,
+      "learning_rate": 4.0434967469889486e-05,
+      "loss": 0.8944,
+      "step": 1142
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.20891857147216797,
+      "learning_rate": 4.0278344012291455e-05,
+      "loss": 0.9077,
+      "step": 1143
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19896891713142395,
+      "learning_rate": 4.012194794905775e-05,
+      "loss": 0.8783,
+      "step": 1144
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17555470764636993,
+      "learning_rate": 3.996577987567727e-05,
+      "loss": 0.7849,
+      "step": 1145
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19006510078907013,
+      "learning_rate": 3.9809840386770825e-05,
+      "loss": 0.8864,
+      "step": 1146
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19303017854690552,
+      "learning_rate": 3.965413007608888e-05,
+      "loss": 0.8912,
+      "step": 1147
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1810441017150879,
+      "learning_rate": 3.949864953650929e-05,
+      "loss": 0.8273,
+      "step": 1148
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18066197633743286,
+      "learning_rate": 3.934339936003499e-05,
+      "loss": 0.8516,
+      "step": 1149
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1802552491426468,
+      "learning_rate": 3.9188380137791936e-05,
+      "loss": 0.8239,
+      "step": 1150
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18922318518161774,
+      "learning_rate": 3.9033592460026545e-05,
+      "loss": 0.8771,
+      "step": 1151
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18907152116298676,
+      "learning_rate": 3.8879036916103704e-05,
+      "loss": 0.8606,
+      "step": 1152
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.1793530136346817,
+      "learning_rate": 3.872471409450441e-05,
+      "loss": 0.859,
+      "step": 1153
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18526653945446014,
+      "learning_rate": 3.85706245828235e-05,
+      "loss": 0.8493,
+      "step": 1154
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.17269764840602875,
+      "learning_rate": 3.8416768967767635e-05,
+      "loss": 0.8206,
+      "step": 1155
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18234558403491974,
+      "learning_rate": 3.826314783515269e-05,
+      "loss": 0.8899,
+      "step": 1156
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.19679400324821472,
+      "learning_rate": 3.810976176990184e-05,
+      "loss": 0.9014,
+      "step": 1157
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.18051548302173615,
+      "learning_rate": 3.795661135604319e-05,
+      "loss": 0.8271,
+      "step": 1158
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18018056452274323,
+      "learning_rate": 3.7803697176707685e-05,
+      "loss": 0.8214,
+      "step": 1159
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19069695472717285,
+      "learning_rate": 3.7651019814126654e-05,
+      "loss": 0.857,
+      "step": 1160
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1932668685913086,
+      "learning_rate": 3.74985798496298e-05,
+      "loss": 0.8357,
+      "step": 1161
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18192018568515778,
+      "learning_rate": 3.734637786364288e-05,
+      "loss": 0.8888,
+      "step": 1162
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18109020590782166,
+      "learning_rate": 3.7194414435685595e-05,
+      "loss": 0.7768,
+      "step": 1163
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18755914270877838,
+      "learning_rate": 3.704269014436932e-05,
+      "loss": 0.8527,
+      "step": 1164
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.19775067269802094,
+      "learning_rate": 3.689120556739475e-05,
+      "loss": 0.8128,
+      "step": 1165
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18509656190872192,
+      "learning_rate": 3.6739961281549995e-05,
+      "loss": 0.8055,
+      "step": 1166
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18040892481803894,
+      "learning_rate": 3.658895786270825e-05,
+      "loss": 0.8695,
+      "step": 1167
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18310925364494324,
+      "learning_rate": 3.643819588582553e-05,
+      "loss": 0.8854,
+      "step": 1168
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18773789703845978,
+      "learning_rate": 3.628767592493857e-05,
+      "loss": 0.8215,
+      "step": 1169
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.18685688078403473,
+      "learning_rate": 3.613739855316257e-05,
+      "loss": 0.805,
+      "step": 1170
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1834569126367569,
+      "learning_rate": 3.598736434268909e-05,
+      "loss": 0.8338,
+      "step": 1171
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1872432827949524,
+      "learning_rate": 3.583757386478389e-05,
+      "loss": 0.8782,
+      "step": 1172
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.1946275234222412,
+      "learning_rate": 3.568802768978463e-05,
+      "loss": 0.8515,
+      "step": 1173
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.20066288113594055,
+      "learning_rate": 3.553872638709871e-05,
+      "loss": 0.8582,
+      "step": 1174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1834530234336853,
+      "learning_rate": 3.538967052520134e-05,
+      "loss": 0.8431,
+      "step": 1175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1824205368757248,
+      "learning_rate": 3.5240860671633024e-05,
+      "loss": 0.8884,
+      "step": 1176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18376876413822174,
+      "learning_rate": 3.509229739299769e-05,
+      "loss": 0.8416,
+      "step": 1177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19210116565227509,
+      "learning_rate": 3.494398125496032e-05,
+      "loss": 0.7968,
+      "step": 1178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18106195330619812,
+      "learning_rate": 3.479591282224496e-05,
+      "loss": 0.8549,
+      "step": 1179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.179016575217247,
+      "learning_rate": 3.46480926586325e-05,
+      "loss": 0.857,
+      "step": 1180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.18134085834026337,
+      "learning_rate": 3.450052132695849e-05,
+      "loss": 0.7819,
+      "step": 1181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17881114780902863,
+      "learning_rate": 3.4353199389111065e-05,
+      "loss": 0.7633,
+      "step": 1182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1760081946849823,
+      "learning_rate": 3.4206127406028745e-05,
+      "loss": 0.8288,
+      "step": 1183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1765538603067398,
+      "learning_rate": 3.405930593769836e-05,
+      "loss": 0.8281,
+      "step": 1184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.1904032826423645,
+      "learning_rate": 3.3912735543152864e-05,
+      "loss": 0.904,
+      "step": 1185
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.21091416478157043,
+      "learning_rate": 3.3766416780469256e-05,
+      "loss": 0.8742,
+      "step": 1186
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19084270298480988,
+      "learning_rate": 3.362035020676637e-05,
+      "loss": 0.909,
+      "step": 1187
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19029565155506134,
+      "learning_rate": 3.3474536378202926e-05,
+      "loss": 0.8604,
+      "step": 1188
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.17569898068904877,
+      "learning_rate": 3.3328975849975205e-05,
+      "loss": 0.8402,
+      "step": 1189
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.20008669793605804,
+      "learning_rate": 3.3183669176315045e-05,
+      "loss": 0.9094,
+      "step": 1190
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.19999565184116364,
+      "learning_rate": 3.303861691048774e-05,
+      "loss": 0.8914,
+      "step": 1191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1818060427904129,
+      "learning_rate": 3.289381960478988e-05,
+      "loss": 0.8628,
+      "step": 1192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18711596727371216,
+      "learning_rate": 3.2749277810547286e-05,
+      "loss": 0.8208,
+      "step": 1193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1935541033744812,
+      "learning_rate": 3.2604992078112926e-05,
+      "loss": 0.8843,
+      "step": 1194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20281094312667847,
+      "learning_rate": 3.2460962956864724e-05,
+      "loss": 0.7895,
+      "step": 1195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18525005877017975,
+      "learning_rate": 3.231719099520367e-05,
+      "loss": 0.8824,
+      "step": 1196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1925046443939209,
+      "learning_rate": 3.217367674055148e-05,
+      "loss": 0.8434,
+      "step": 1197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18043692409992218,
+      "learning_rate": 3.2030420739348665e-05,
+      "loss": 0.8322,
+      "step": 1198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18874524533748627,
+      "learning_rate": 3.188742353705244e-05,
+      "loss": 0.8367,
+      "step": 1199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18166741728782654,
+      "learning_rate": 3.174468567813461e-05,
+      "loss": 0.8356,
+      "step": 1200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18626669049263,
+      "learning_rate": 3.16022077060795e-05,
+      "loss": 0.8492,
+      "step": 1201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18880659341812134,
+      "learning_rate": 3.145999016338193e-05,
+      "loss": 0.8337,
+      "step": 1202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20019130408763885,
+      "learning_rate": 3.1318033591545046e-05,
+      "loss": 0.8581,
+      "step": 1203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20769058167934418,
+      "learning_rate": 3.117633853107846e-05,
+      "loss": 0.8972,
+      "step": 1204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.1961383819580078,
+      "learning_rate": 3.103490552149595e-05,
+      "loss": 0.8952,
+      "step": 1205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.20016257464885712,
+      "learning_rate": 3.089373510131354e-05,
+      "loss": 0.9237,
+      "step": 1206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.18094752728939056,
+      "learning_rate": 3.0752827808047445e-05,
+      "loss": 0.8017,
+      "step": 1207
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1957404911518097,
+      "learning_rate": 3.0612184178211965e-05,
+      "loss": 0.8767,
+      "step": 1208
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1770680546760559,
+      "learning_rate": 3.0471804747317613e-05,
+      "loss": 0.8393,
+      "step": 1209
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1798279881477356,
+      "learning_rate": 3.033169004986873e-05,
+      "loss": 0.8396,
+      "step": 1210
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1768520027399063,
+      "learning_rate": 3.0191840619361855e-05,
+      "loss": 0.8378,
+      "step": 1211
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17671707272529602,
+      "learning_rate": 3.005225698828338e-05,
+      "loss": 0.7792,
+      "step": 1212
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19124135375022888,
+      "learning_rate": 2.9912939688107756e-05,
+      "loss": 0.8695,
+      "step": 1213
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1776977926492691,
+      "learning_rate": 2.9773889249295294e-05,
+      "loss": 0.8072,
+      "step": 1214
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19133800268173218,
+      "learning_rate": 2.9635106201290207e-05,
+      "loss": 0.8593,
+      "step": 1215
+    },
+    {
+      "epoch": 0.75,
+      "eval_loss": 1.5269761085510254,
+      "eval_runtime": 187.4098,
+      "eval_samples_per_second": 146.615,
+      "eval_steps_per_second": 73.31,
+      "step": 1215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18236102163791656,
+      "learning_rate": 2.9496591072518597e-05,
+      "loss": 0.8014,
+      "step": 1216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19258786737918854,
+      "learning_rate": 2.935834439038653e-05,
+      "loss": 0.8448,
+      "step": 1217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.17276649177074432,
+      "learning_rate": 2.9220366681277855e-05,
+      "loss": 0.8128,
+      "step": 1218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.18369680643081665,
+      "learning_rate": 2.908265847055226e-05,
+      "loss": 0.829,
+      "step": 1219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1840008944272995,
+      "learning_rate": 2.894522028254334e-05,
+      "loss": 0.8379,
+      "step": 1220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.19129714369773865,
+      "learning_rate": 2.8808052640556637e-05,
+      "loss": 0.9178,
+      "step": 1221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1785333752632141,
+      "learning_rate": 2.8671156066867465e-05,
+      "loss": 0.7429,
+      "step": 1222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.1841101348400116,
+      "learning_rate": 2.8534531082719073e-05,
+      "loss": 0.8585,
+      "step": 1223
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19334495067596436,
+      "learning_rate": 2.839817820832056e-05,
+      "loss": 0.8254,
+      "step": 1224
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.19443805515766144,
+      "learning_rate": 2.826209796284506e-05,
+      "loss": 0.8647,
+      "step": 1225
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20505933463573456,
+      "learning_rate": 2.812629086442754e-05,
+      "loss": 0.8017,
+      "step": 1226
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18697763979434967,
+      "learning_rate": 2.7990757430163038e-05,
+      "loss": 0.8361,
+      "step": 1227
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20007814466953278,
+      "learning_rate": 2.7855498176104434e-05,
+      "loss": 0.849,
+      "step": 1228
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.187923863530159,
+      "learning_rate": 2.7720513617260856e-05,
+      "loss": 0.8348,
+      "step": 1229
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1997993290424347,
+      "learning_rate": 2.7585804267595384e-05,
+      "loss": 0.9098,
+      "step": 1230
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18425752222537994,
+      "learning_rate": 2.7451370640023232e-05,
+      "loss": 0.8415,
+      "step": 1231
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.17538057267665863,
+      "learning_rate": 2.731721324640982e-05,
+      "loss": 0.8384,
+      "step": 1232
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.21126823127269745,
+      "learning_rate": 2.7183332597568746e-05,
+      "loss": 0.8909,
+      "step": 1233
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.1960807889699936,
+      "learning_rate": 2.7049729203259934e-05,
+      "loss": 0.8657,
+      "step": 1234
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.2083304077386856,
+      "learning_rate": 2.691640357218759e-05,
+      "loss": 0.8443,
+      "step": 1235
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18388773500919342,
+      "learning_rate": 2.6783356211998355e-05,
+      "loss": 0.8561,
+      "step": 1236
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.20308782160282135,
+      "learning_rate": 2.66505876292793e-05,
+      "loss": 0.8579,
+      "step": 1237
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.174764946103096,
+      "learning_rate": 2.6518098329556075e-05,
+      "loss": 0.8344,
+      "step": 1238
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.18486882746219635,
+      "learning_rate": 2.638588881729088e-05,
+      "loss": 0.9006,
+      "step": 1239
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19115668535232544,
+      "learning_rate": 2.6253959595880673e-05,
+      "loss": 0.8442,
+      "step": 1240
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1864568144083023,
+      "learning_rate": 2.612231116765511e-05,
+      "loss": 0.8532,
+      "step": 1241
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19725145399570465,
+      "learning_rate": 2.599094403387481e-05,
+      "loss": 0.9175,
+      "step": 1242
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18937453627586365,
+      "learning_rate": 2.5859858694729245e-05,
+      "loss": 0.871,
+      "step": 1243
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19133366644382477,
+      "learning_rate": 2.5729055649334987e-05,
+      "loss": 0.8699,
+      "step": 1244
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1894661784172058,
+      "learning_rate": 2.5598535395733737e-05,
+      "loss": 0.838,
+      "step": 1245
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1935572475194931,
+      "learning_rate": 2.5468298430890437e-05,
+      "loss": 0.9157,
+      "step": 1246
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.20360638201236725,
+      "learning_rate": 2.533834525069142e-05,
+      "loss": 0.9336,
+      "step": 1247
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19841620326042175,
+      "learning_rate": 2.5208676349942462e-05,
+      "loss": 0.8803,
+      "step": 1248
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1718897819519043,
+      "learning_rate": 2.50792922223669e-05,
+      "loss": 0.8037,
+      "step": 1249
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18919512629508972,
+      "learning_rate": 2.495019336060387e-05,
+      "loss": 0.8316,
+      "step": 1250
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1818518340587616,
+      "learning_rate": 2.482138025620625e-05,
+      "loss": 0.807,
+      "step": 1251
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.19060440361499786,
+      "learning_rate": 2.4692853399638917e-05,
+      "loss": 0.8692,
+      "step": 1252
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1854245513677597,
+      "learning_rate": 2.4564613280276806e-05,
+      "loss": 0.8366,
+      "step": 1253
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.1819692701101303,
+      "learning_rate": 2.4436660386403122e-05,
+      "loss": 0.7966,
+      "step": 1254
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.18258213996887207,
+      "learning_rate": 2.4308995205207406e-05,
+      "loss": 0.8346,
+      "step": 1255
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19831517338752747,
+      "learning_rate": 2.418161822278374e-05,
+      "loss": 0.8836,
+      "step": 1256
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1940324753522873,
+      "learning_rate": 2.405452992412882e-05,
+      "loss": 0.8881,
+      "step": 1257
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.17954248189926147,
+      "learning_rate": 2.3927730793140246e-05,
+      "loss": 0.8408,
+      "step": 1258
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1928507685661316,
+      "learning_rate": 2.380122131261452e-05,
+      "loss": 0.812,
+      "step": 1259
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18852391839027405,
+      "learning_rate": 2.367500196424529e-05,
+      "loss": 0.8399,
+      "step": 1260
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1832701712846756,
+      "learning_rate": 2.3549073228621522e-05,
+      "loss": 0.8638,
+      "step": 1261
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1859746277332306,
+      "learning_rate": 2.3423435585225616e-05,
+      "loss": 0.8244,
+      "step": 1262
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18503403663635254,
+      "learning_rate": 2.329808951243174e-05,
+      "loss": 0.8776,
+      "step": 1263
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18097741901874542,
+      "learning_rate": 2.3173035487503703e-05,
+      "loss": 0.8592,
+      "step": 1264
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.19478042423725128,
+      "learning_rate": 2.3048273986593416e-05,
+      "loss": 0.8522,
+      "step": 1265
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18363747000694275,
+      "learning_rate": 2.2923805484739035e-05,
+      "loss": 0.7866,
+      "step": 1266
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18185609579086304,
+      "learning_rate": 2.279963045586302e-05,
+      "loss": 0.8548,
+      "step": 1267
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18947353959083557,
+      "learning_rate": 2.267574937277046e-05,
+      "loss": 0.8559,
+      "step": 1268
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18562479317188263,
+      "learning_rate": 2.2552162707147184e-05,
+      "loss": 0.8392,
+      "step": 1269
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.1905825287103653,
+      "learning_rate": 2.242887092955801e-05,
+      "loss": 0.8785,
+      "step": 1270
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18808040022850037,
+      "learning_rate": 2.2305874509445023e-05,
+      "loss": 0.8473,
+      "step": 1271
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.18851177394390106,
+      "learning_rate": 2.2183173915125656e-05,
+      "loss": 0.8049,
+      "step": 1272
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19175463914871216,
+      "learning_rate": 2.206076961379091e-05,
+      "loss": 0.8558,
+      "step": 1273
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18208831548690796,
+      "learning_rate": 2.1938662071503692e-05,
+      "loss": 0.8067,
+      "step": 1274
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19766081869602203,
+      "learning_rate": 2.181685175319702e-05,
+      "loss": 0.8561,
+      "step": 1275
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18860626220703125,
+      "learning_rate": 2.1695339122672155e-05,
+      "loss": 0.8342,
+      "step": 1276
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18599668145179749,
+      "learning_rate": 2.1574124642596883e-05,
+      "loss": 0.8667,
+      "step": 1277
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18972142040729523,
+      "learning_rate": 2.1453208774503754e-05,
+      "loss": 0.8764,
+      "step": 1278
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.21071122586727142,
+      "learning_rate": 2.133259197878841e-05,
+      "loss": 0.8684,
+      "step": 1279
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1776166409254074,
+      "learning_rate": 2.121227471470768e-05,
+      "loss": 0.7855,
+      "step": 1280
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1972324252128601,
+      "learning_rate": 2.109225744037795e-05,
+      "loss": 0.9253,
+      "step": 1281
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17811790108680725,
+      "learning_rate": 2.0972540612773262e-05,
+      "loss": 0.8294,
+      "step": 1282
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1759762316942215,
+      "learning_rate": 2.085312468772387e-05,
+      "loss": 0.8266,
+      "step": 1283
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.17470064759254456,
+      "learning_rate": 2.0734010119914192e-05,
+      "loss": 0.842,
+      "step": 1284
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.1845446527004242,
+      "learning_rate": 2.0615197362881234e-05,
+      "loss": 0.8645,
+      "step": 1285
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18126532435417175,
+      "learning_rate": 2.0496686869012847e-05,
+      "loss": 0.7934,
+      "step": 1286
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.18096975982189178,
+      "learning_rate": 2.0378479089545966e-05,
+      "loss": 0.8106,
+      "step": 1287
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.19099704921245575,
+      "learning_rate": 2.0260574474564987e-05,
+      "loss": 0.8638,
+      "step": 1288
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.199772447347641,
+      "learning_rate": 2.0142973472999928e-05,
+      "loss": 0.8491,
+      "step": 1289
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1808193475008011,
+      "learning_rate": 2.002567653262479e-05,
+      "loss": 0.8198,
+      "step": 1290
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1927025020122528,
+      "learning_rate": 1.9908684100055842e-05,
+      "loss": 0.8192,
+      "step": 1291
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18885312974452972,
+      "learning_rate": 1.979199662074991e-05,
+      "loss": 0.8431,
+      "step": 1292
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19639408588409424,
+      "learning_rate": 1.9675614539002708e-05,
+      "loss": 0.8376,
+      "step": 1293
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.21020396053791046,
+      "learning_rate": 1.955953829794711e-05,
+      "loss": 0.8289,
+      "step": 1294
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18382123112678528,
+      "learning_rate": 1.944376833955147e-05,
+      "loss": 0.8547,
+      "step": 1295
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1936212182044983,
+      "learning_rate": 1.9328305104618015e-05,
+      "loss": 0.8806,
+      "step": 1296
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.1913760006427765,
+      "learning_rate": 1.921314903278102e-05,
+      "loss": 0.8077,
+      "step": 1297
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18246234953403473,
+      "learning_rate": 1.9098300562505266e-05,
+      "loss": 0.8262,
+      "step": 1298
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.210890531539917,
+      "learning_rate": 1.898376013108428e-05,
+      "loss": 0.8834,
+      "step": 1299
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.19394394755363464,
+      "learning_rate": 1.8869528174638752e-05,
+      "loss": 0.8665,
+      "step": 1300
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20866113901138306,
+      "learning_rate": 1.8755605128114795e-05,
+      "loss": 0.9086,
+      "step": 1301
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.20150724053382874,
+      "learning_rate": 1.8641991425282345e-05,
+      "loss": 0.8301,
+      "step": 1302
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.18174578249454498,
+      "learning_rate": 1.8528687498733476e-05,
+      "loss": 0.8279,
+      "step": 1303
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.177090585231781,
+      "learning_rate": 1.8415693779880817e-05,
+      "loss": 0.8303,
+      "step": 1304
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.20232273638248444,
+      "learning_rate": 1.8303010698955804e-05,
+      "loss": 0.8258,
+      "step": 1305
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19295179843902588,
+      "learning_rate": 1.819063868500711e-05,
+      "loss": 0.9087,
+      "step": 1306
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18751981854438782,
+      "learning_rate": 1.8078578165898995e-05,
+      "loss": 0.8224,
+      "step": 1307
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18430137634277344,
+      "learning_rate": 1.7966829568309707e-05,
+      "loss": 0.8395,
+      "step": 1308
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1893179565668106,
+      "learning_rate": 1.7855393317729807e-05,
+      "loss": 0.8991,
+      "step": 1309
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19309356808662415,
+      "learning_rate": 1.774426983846058e-05,
+      "loss": 0.8092,
+      "step": 1310
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1843278855085373,
+      "learning_rate": 1.7633459553612387e-05,
+      "loss": 0.8033,
+      "step": 1311
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.17787937819957733,
+      "learning_rate": 1.7522962885103145e-05,
+      "loss": 0.8297,
+      "step": 1312
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19011004269123077,
+      "learning_rate": 1.7412780253656602e-05,
+      "loss": 0.8605,
+      "step": 1313
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18457487225532532,
+      "learning_rate": 1.7302912078800803e-05,
+      "loss": 0.835,
+      "step": 1314
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18622943758964539,
+      "learning_rate": 1.7193358778866465e-05,
+      "loss": 0.8646,
+      "step": 1315
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.19360771775245667,
+      "learning_rate": 1.708412077098539e-05,
+      "loss": 0.8534,
+      "step": 1316
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1863589584827423,
+      "learning_rate": 1.6975198471088972e-05,
+      "loss": 0.8234,
+      "step": 1317
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1854516565799713,
+      "learning_rate": 1.686659229390637e-05,
+      "loss": 0.8707,
+      "step": 1318
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.18933233618736267,
+      "learning_rate": 1.6758302652963175e-05,
+      "loss": 0.8188,
+      "step": 1319
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.1956777721643448,
+      "learning_rate": 1.6650329960579792e-05,
+      "loss": 0.8848,
+      "step": 1320
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1868615746498108,
+      "learning_rate": 1.6542674627869737e-05,
+      "loss": 0.8736,
+      "step": 1321
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20109330117702484,
+      "learning_rate": 1.643533706473819e-05,
+      "loss": 0.9563,
+      "step": 1322
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21118614077568054,
+      "learning_rate": 1.632831767988039e-05,
+      "loss": 0.8755,
+      "step": 1323
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1894291341304779,
+      "learning_rate": 1.6221616880780076e-05,
+      "loss": 0.8507,
+      "step": 1324
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18251925706863403,
+      "learning_rate": 1.6115235073708025e-05,
+      "loss": 0.8328,
+      "step": 1325
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18801167607307434,
+      "learning_rate": 1.600917266372035e-05,
+      "loss": 0.7795,
+      "step": 1326
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17382827401161194,
+      "learning_rate": 1.5903430054657077e-05,
+      "loss": 0.7695,
+      "step": 1327
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18468500673770905,
+      "learning_rate": 1.57980076491405e-05,
+      "loss": 0.8004,
+      "step": 1328
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19160085916519165,
+      "learning_rate": 1.5692905848573836e-05,
+      "loss": 0.8333,
+      "step": 1329
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.18871837854385376,
+      "learning_rate": 1.5588125053139468e-05,
+      "loss": 0.8482,
+      "step": 1330
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17362573742866516,
+      "learning_rate": 1.54836656617976e-05,
+      "loss": 0.792,
+      "step": 1331
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.17980434000492096,
+      "learning_rate": 1.53795280722846e-05,
+      "loss": 0.8159,
+      "step": 1332
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.21310344338417053,
+      "learning_rate": 1.5275712681111644e-05,
+      "loss": 0.8179,
+      "step": 1333
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.20284922420978546,
+      "learning_rate": 1.5172219883563033e-05,
+      "loss": 0.8624,
+      "step": 1334
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.19041256606578827,
+      "learning_rate": 1.5069050073694813e-05,
+      "loss": 0.8605,
+      "step": 1335
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.1864355057477951,
+      "learning_rate": 1.496620364433321e-05,
+      "loss": 0.8239,
+      "step": 1336
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18724216520786285,
+      "learning_rate": 1.486368098707317e-05,
+      "loss": 0.897,
+      "step": 1337
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.17345544695854187,
+      "learning_rate": 1.4761482492276846e-05,
+      "loss": 0.7494,
+      "step": 1338
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19252581894397736,
+      "learning_rate": 1.4659608549072135e-05,
+      "loss": 0.8897,
+      "step": 1339
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18758520483970642,
+      "learning_rate": 1.4558059545351143e-05,
+      "loss": 0.808,
+      "step": 1340
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1917487531900406,
+      "learning_rate": 1.4456835867768815e-05,
+      "loss": 0.8111,
+      "step": 1341
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19719301164150238,
+      "learning_rate": 1.4355937901741323e-05,
+      "loss": 0.8474,
+      "step": 1342
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19279302656650543,
+      "learning_rate": 1.4255366031444717e-05,
+      "loss": 0.8453,
+      "step": 1343
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18393969535827637,
+      "learning_rate": 1.415512063981339e-05,
+      "loss": 0.8567,
+      "step": 1344
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1856771856546402,
+      "learning_rate": 1.4055202108538657e-05,
+      "loss": 0.7951,
+      "step": 1345
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1898553967475891,
+      "learning_rate": 1.3955610818067266e-05,
+      "loss": 0.8593,
+      "step": 1346
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18659019470214844,
+      "learning_rate": 1.3856347147600013e-05,
+      "loss": 0.8263,
+      "step": 1347
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18961209058761597,
+      "learning_rate": 1.3757411475090231e-05,
+      "loss": 0.8072,
+      "step": 1348
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.1855127364397049,
+      "learning_rate": 1.3658804177242345e-05,
+      "loss": 0.8452,
+      "step": 1349
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19525891542434692,
+      "learning_rate": 1.3560525629510568e-05,
+      "loss": 0.8411,
+      "step": 1350
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18369780480861664,
+      "learning_rate": 1.3462576206097299e-05,
+      "loss": 0.8546,
+      "step": 1351
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.18788158893585205,
+      "learning_rate": 1.3364956279951768e-05,
+      "loss": 0.8737,
+      "step": 1352
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.19446510076522827,
+      "learning_rate": 1.3267666222768637e-05,
+      "loss": 0.8721,
+      "step": 1353
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18963198363780975,
+      "learning_rate": 1.3170706404986644e-05,
+      "loss": 0.8291,
+      "step": 1354
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18187525868415833,
+      "learning_rate": 1.307407719578696e-05,
+      "loss": 0.8138,
+      "step": 1355
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20309653878211975,
+      "learning_rate": 1.2977778963092068e-05,
+      "loss": 0.849,
+      "step": 1356
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1924443244934082,
+      "learning_rate": 1.288181207356417e-05,
+      "loss": 0.8645,
+      "step": 1357
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18161898851394653,
+      "learning_rate": 1.2786176892603929e-05,
+      "loss": 0.785,
+      "step": 1358
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20091351866722107,
+      "learning_rate": 1.2690873784348923e-05,
+      "loss": 0.9138,
+      "step": 1359
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18364280462265015,
+      "learning_rate": 1.259590311167238e-05,
+      "loss": 0.8678,
+      "step": 1360
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19146834313869476,
+      "learning_rate": 1.2501265236181737e-05,
+      "loss": 0.7759,
+      "step": 1361
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.19955946505069733,
+      "learning_rate": 1.2406960518217326e-05,
+      "loss": 0.8324,
+      "step": 1362
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18070782721042633,
+      "learning_rate": 1.2312989316850931e-05,
+      "loss": 0.7756,
+      "step": 1363
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1785333752632141,
+      "learning_rate": 1.221935198988441e-05,
+      "loss": 0.8768,
+      "step": 1364
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.20576317608356476,
+      "learning_rate": 1.2126048893848396e-05,
+      "loss": 0.8736,
+      "step": 1365
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.17525583505630493,
+      "learning_rate": 1.2033080384000983e-05,
+      "loss": 0.7515,
+      "step": 1366
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18548227846622467,
+      "learning_rate": 1.19404468143262e-05,
+      "loss": 0.8047,
+      "step": 1367
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.18451763689517975,
+      "learning_rate": 1.1848148537532843e-05,
+      "loss": 0.7808,
+      "step": 1368
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.1921512931585312,
+      "learning_rate": 1.1756185905053019e-05,
+      "loss": 0.9321,
+      "step": 1369
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18506480753421783,
+      "learning_rate": 1.166455926704082e-05,
+      "loss": 0.8376,
+      "step": 1370
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18874765932559967,
+      "learning_rate": 1.157326897237111e-05,
+      "loss": 0.8793,
+      "step": 1371
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18793274462223053,
+      "learning_rate": 1.1482315368638042e-05,
+      "loss": 0.7701,
+      "step": 1372
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1925497204065323,
+      "learning_rate": 1.1391698802153727e-05,
+      "loss": 0.778,
+      "step": 1373
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20436991751194,
+      "learning_rate": 1.1301419617947117e-05,
+      "loss": 0.8635,
+      "step": 1374
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1809050291776657,
+      "learning_rate": 1.1211478159762478e-05,
+      "loss": 0.7658,
+      "step": 1375
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18188363313674927,
+      "learning_rate": 1.112187477005817e-05,
+      "loss": 0.8481,
+      "step": 1376
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19081351161003113,
+      "learning_rate": 1.1032609790005344e-05,
+      "loss": 0.8718,
+      "step": 1377
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1823447048664093,
+      "learning_rate": 1.0943683559486605e-05,
+      "loss": 0.8108,
+      "step": 1378
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.19918982684612274,
+      "learning_rate": 1.0855096417094811e-05,
+      "loss": 0.8482,
+      "step": 1379
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.1928330510854721,
+      "learning_rate": 1.0766848700131648e-05,
+      "loss": 0.8152,
+      "step": 1380
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20285867154598236,
+      "learning_rate": 1.0678940744606469e-05,
+      "loss": 0.8722,
+      "step": 1381
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18436557054519653,
+      "learning_rate": 1.0591372885234885e-05,
+      "loss": 0.8071,
+      "step": 1382
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.18045209348201752,
+      "learning_rate": 1.0504145455437664e-05,
+      "loss": 0.8039,
+      "step": 1383
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.199941948056221,
+      "learning_rate": 1.0417258787339313e-05,
+      "loss": 0.86,
+      "step": 1384
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.20726729929447174,
+      "learning_rate": 1.0330713211766863e-05,
+      "loss": 0.8948,
+      "step": 1385
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1846933662891388,
+      "learning_rate": 1.0244509058248609e-05,
+      "loss": 0.8423,
+      "step": 1386
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19070808589458466,
+      "learning_rate": 1.0158646655012905e-05,
+      "loss": 0.8105,
+      "step": 1387
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18341971933841705,
+      "learning_rate": 1.0073126328986804e-05,
+      "loss": 0.8185,
+      "step": 1388
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19135870039463043,
+      "learning_rate": 9.98794840579491e-06,
+      "loss": 0.8552,
+      "step": 1389
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18689709901809692,
+      "learning_rate": 9.903113209758096e-06,
+      "loss": 0.8027,
+      "step": 1390
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1949356198310852,
+      "learning_rate": 9.818621063892275e-06,
+      "loss": 0.9225,
+      "step": 1391
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19429528713226318,
+      "learning_rate": 9.734472289907182e-06,
+      "loss": 0.8527,
+      "step": 1392
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18282286822795868,
+      "learning_rate": 9.650667208205144e-06,
+      "loss": 0.8571,
+      "step": 1393
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19118870794773102,
+      "learning_rate": 9.567206137879803e-06,
+      "loss": 0.8868,
+      "step": 1394
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18210899829864502,
+      "learning_rate": 9.484089396715056e-06,
+      "loss": 0.8973,
+      "step": 1395
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1832556277513504,
+      "learning_rate": 9.401317301183655e-06,
+      "loss": 0.8641,
+      "step": 1396
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1935901641845703,
+      "learning_rate": 9.318890166446115e-06,
+      "loss": 0.9098,
+      "step": 1397
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.19078364968299866,
+      "learning_rate": 9.236808306349497e-06,
+      "loss": 0.8388,
+      "step": 1398
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18594953417778015,
+      "learning_rate": 9.155072033426193e-06,
+      "loss": 0.8472,
+      "step": 1399
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.1824297159910202,
+      "learning_rate": 9.073681658892775e-06,
+      "loss": 0.8226,
+      "step": 1400
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.18082523345947266,
+      "learning_rate": 8.992637492648737e-06,
+      "loss": 0.795,
+      "step": 1401
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19871748983860016,
+      "learning_rate": 8.911939843275396e-06,
+      "loss": 0.7771,
+      "step": 1402
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.20180968940258026,
+      "learning_rate": 8.831589018034658e-06,
+      "loss": 0.8898,
+      "step": 1403
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18637341260910034,
+      "learning_rate": 8.751585322867917e-06,
+      "loss": 0.8235,
+      "step": 1404
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18972526490688324,
+      "learning_rate": 8.671929062394802e-06,
+      "loss": 0.8531,
+      "step": 1405
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1775706261396408,
+      "learning_rate": 8.592620539912066e-06,
+      "loss": 0.8275,
+      "step": 1406
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19284261763095856,
+      "learning_rate": 8.513660057392415e-06,
+      "loss": 0.8487,
+      "step": 1407
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.1854104846715927,
+      "learning_rate": 8.43504791548344e-06,
+      "loss": 0.8691,
+      "step": 1408
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19354654848575592,
+      "learning_rate": 8.35678441350628e-06,
+      "loss": 0.8538,
+      "step": 1409
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19303570687770844,
+      "learning_rate": 8.278869849454718e-06,
+      "loss": 0.822,
+      "step": 1410
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18061821162700653,
+      "learning_rate": 8.201304519993836e-06,
+      "loss": 0.8165,
+      "step": 1411
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.18833564221858978,
+      "learning_rate": 8.124088720459067e-06,
+      "loss": 0.8278,
+      "step": 1412
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19145245850086212,
+      "learning_rate": 8.047222744854943e-06,
+      "loss": 0.7919,
+      "step": 1413
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19037508964538574,
+      "learning_rate": 7.970706885854007e-06,
+      "loss": 0.8485,
+      "step": 1414
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19154959917068481,
+      "learning_rate": 7.894541434795699e-06,
+      "loss": 0.8404,
+      "step": 1415
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.194344624876976,
+      "learning_rate": 7.81872668168533e-06,
+      "loss": 0.8709,
+      "step": 1416
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.19423475861549377,
+      "learning_rate": 7.74326291519284e-06,
+      "loss": 0.7956,
+      "step": 1417
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18557237088680267,
+      "learning_rate": 7.668150422651743e-06,
+      "loss": 0.845,
+      "step": 1418
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1937812715768814,
+      "learning_rate": 7.593389490058078e-06,
+      "loss": 0.8643,
+      "step": 1419
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18862222135066986,
+      "learning_rate": 7.5189804020693536e-06,
+      "loss": 0.8278,
+      "step": 1420
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19189707934856415,
+      "learning_rate": 7.444923442003326e-06,
+      "loss": 0.8598,
+      "step": 1421
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1830250471830368,
+      "learning_rate": 7.3712188918370285e-06,
+      "loss": 0.8113,
+      "step": 1422
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18655569851398468,
+      "learning_rate": 7.297867032205663e-06,
+      "loss": 0.8051,
+      "step": 1423
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1855732798576355,
+      "learning_rate": 7.224868142401542e-06,
+      "loss": 0.8982,
+      "step": 1424
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18631410598754883,
+      "learning_rate": 7.152222500373052e-06,
+      "loss": 0.784,
+      "step": 1425
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19200319051742554,
+      "learning_rate": 7.079930382723521e-06,
+      "loss": 0.8495,
+      "step": 1426
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18701863288879395,
+      "learning_rate": 7.007992064710189e-06,
+      "loss": 0.8589,
+      "step": 1427
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19145749509334564,
+      "learning_rate": 6.936407820243251e-06,
+      "loss": 0.8652,
+      "step": 1428
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1971212774515152,
+      "learning_rate": 6.86517792188468e-06,
+      "loss": 0.8825,
+      "step": 1429
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19131560623645782,
+      "learning_rate": 6.794302640847294e-06,
+      "loss": 0.8699,
+      "step": 1430
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.19057752192020416,
+      "learning_rate": 6.7237822469936485e-06,
+      "loss": 0.8739,
+      "step": 1431
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.18546807765960693,
+      "learning_rate": 6.653617008835034e-06,
+      "loss": 0.8343,
+      "step": 1432
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1828879415988922,
+      "learning_rate": 6.583807193530523e-06,
+      "loss": 0.8477,
+      "step": 1433
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.1874518245458603,
+      "learning_rate": 6.514353066885837e-06,
+      "loss": 0.7684,
+      "step": 1434
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.20050178468227386,
+      "learning_rate": 6.4452548933523815e-06,
+      "loss": 0.9102,
+      "step": 1435
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17574195563793182,
+      "learning_rate": 6.37651293602628e-06,
+      "loss": 0.8018,
+      "step": 1436
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18918807804584503,
+      "learning_rate": 6.308127456647317e-06,
+      "loss": 0.8279,
+      "step": 1437
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.17920467257499695,
+      "learning_rate": 6.240098715597975e-06,
+      "loss": 0.798,
+      "step": 1438
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18810401856899261,
+      "learning_rate": 6.172426971902412e-06,
+      "loss": 0.8363,
+      "step": 1439
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18779538571834564,
+      "learning_rate": 6.1051124832254944e-06,
+      "loss": 0.824,
+      "step": 1440
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19366438686847687,
+      "learning_rate": 6.038155505871856e-06,
+      "loss": 0.8525,
+      "step": 1441
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19134458899497986,
+      "learning_rate": 5.971556294784841e-06,
+      "loss": 0.8893,
+      "step": 1442
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.21351316571235657,
+      "learning_rate": 5.9053151035455725e-06,
+      "loss": 0.8489,
+      "step": 1443
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19345346093177795,
+      "learning_rate": 5.839432184372018e-06,
+      "loss": 0.854,
+      "step": 1444
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18848451972007751,
+      "learning_rate": 5.77390778811796e-06,
+      "loss": 0.8094,
+      "step": 1445
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18633106350898743,
+      "learning_rate": 5.708742164272107e-06,
+      "loss": 0.8254,
+      "step": 1446
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18038615584373474,
+      "learning_rate": 5.643935560957114e-06,
+      "loss": 0.8254,
+      "step": 1447
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19941388070583344,
+      "learning_rate": 5.579488224928619e-06,
+      "loss": 0.9333,
+      "step": 1448
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.18492132425308228,
+      "learning_rate": 5.515400401574378e-06,
+      "loss": 0.8724,
+      "step": 1449
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.19234874844551086,
+      "learning_rate": 5.451672334913216e-06,
+      "loss": 0.8398,
+      "step": 1450
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19831697642803192,
+      "learning_rate": 5.388304267594191e-06,
+      "loss": 0.8411,
+      "step": 1451
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19975081086158752,
+      "learning_rate": 5.325296440895622e-06,
+      "loss": 0.8858,
+      "step": 1452
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1901513934135437,
+      "learning_rate": 5.262649094724226e-06,
+      "loss": 0.8308,
+      "step": 1453
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19020718336105347,
+      "learning_rate": 5.200362467614106e-06,
+      "loss": 0.7991,
+      "step": 1454
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18462246656417847,
+      "learning_rate": 5.138436796725943e-06,
+      "loss": 0.8164,
+      "step": 1455
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.17772462964057922,
+      "learning_rate": 5.0768723178460266e-06,
+      "loss": 0.8149,
+      "step": 1456
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18094763159751892,
+      "learning_rate": 5.015669265385436e-06,
+      "loss": 0.8447,
+      "step": 1457
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.19172245264053345,
+      "learning_rate": 4.954827872379031e-06,
+      "loss": 0.8531,
+      "step": 1458
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.22510623931884766,
+      "learning_rate": 4.8943483704846475e-06,
+      "loss": 0.8627,
+      "step": 1459
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18390975892543793,
+      "learning_rate": 4.834230989982213e-06,
+      "loss": 0.7941,
+      "step": 1460
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18206605315208435,
+      "learning_rate": 4.774475959772817e-06,
+      "loss": 0.7954,
+      "step": 1461
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20802518725395203,
+      "learning_rate": 4.715083507377949e-06,
+      "loss": 0.8824,
+      "step": 1462
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1775992214679718,
+      "learning_rate": 4.656053858938459e-06,
+      "loss": 0.8206,
+      "step": 1463
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.18683761358261108,
+      "learning_rate": 4.597387239213868e-06,
+      "loss": 0.8158,
+      "step": 1464
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.20285680890083313,
+      "learning_rate": 4.539083871581396e-06,
+      "loss": 0.8568,
+      "step": 1465
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.1975213885307312,
+      "learning_rate": 4.481143978035196e-06,
+      "loss": 0.8916,
+      "step": 1466
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19542734324932098,
+      "learning_rate": 4.423567779185445e-06,
+      "loss": 0.8731,
+      "step": 1467
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18391336500644684,
+      "learning_rate": 4.366355494257523e-06,
+      "loss": 0.8002,
+      "step": 1468
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1848813146352768,
+      "learning_rate": 4.309507341091179e-06,
+      "loss": 0.8138,
+      "step": 1469
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18202924728393555,
+      "learning_rate": 4.253023536139733e-06,
+      "loss": 0.8054,
+      "step": 1470
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19840873777866364,
+      "learning_rate": 4.196904294469228e-06,
+      "loss": 0.8679,
+      "step": 1471
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.17790162563323975,
+      "learning_rate": 4.14114982975754e-06,
+      "loss": 0.8152,
+      "step": 1472
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.194023996591568,
+      "learning_rate": 4.085760354293677e-06,
+      "loss": 0.8589,
+      "step": 1473
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18887946009635925,
+      "learning_rate": 4.03073607897696e-06,
+      "loss": 0.8565,
+      "step": 1474
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18685729801654816,
+      "learning_rate": 3.9760772133161315e-06,
+      "loss": 0.8343,
+      "step": 1475
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19194410741329193,
+      "learning_rate": 3.921783965428627e-06,
+      "loss": 0.8912,
+      "step": 1476
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.20712026953697205,
+      "learning_rate": 3.8678565420397605e-06,
+      "loss": 0.8922,
+      "step": 1477
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.18340593576431274,
+      "learning_rate": 3.814295148481961e-06,
+      "loss": 0.8391,
+      "step": 1478
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19487400352954865,
+      "learning_rate": 3.7610999886939722e-06,
+      "loss": 0.8496,
+      "step": 1479
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.2025231122970581,
+      "learning_rate": 3.7082712652200867e-06,
+      "loss": 0.8798,
+      "step": 1480
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.19572308659553528,
+      "learning_rate": 3.6558091792092906e-06,
+      "loss": 0.8337,
+      "step": 1481
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.1867847442626953,
+      "learning_rate": 3.6037139304146762e-06,
+      "loss": 0.8279,
+      "step": 1482
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19239585101604462,
+      "learning_rate": 3.5519857171925185e-06,
+      "loss": 0.8617,
+      "step": 1483
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18097510933876038,
+      "learning_rate": 3.50062473650159e-06,
+      "loss": 0.8477,
+      "step": 1484
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18740305304527283,
+      "learning_rate": 3.449631183902413e-06,
+      "loss": 0.8153,
+      "step": 1485
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19302509725093842,
+      "learning_rate": 3.399005253556464e-06,
+      "loss": 0.8648,
+      "step": 1486
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19576258957386017,
+      "learning_rate": 3.3487471382255275e-06,
+      "loss": 0.9037,
+      "step": 1487
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18345601856708527,
+      "learning_rate": 3.298857029270863e-06,
+      "loss": 0.8159,
+      "step": 1488
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1863374263048172,
+      "learning_rate": 3.249335116652541e-06,
+      "loss": 0.8135,
+      "step": 1489
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1797042042016983,
+      "learning_rate": 3.2001815889286856e-06,
+      "loss": 0.8274,
+      "step": 1490
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1892089694738388,
+      "learning_rate": 3.1513966332547883e-06,
+      "loss": 0.7845,
+      "step": 1491
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18437382578849792,
+      "learning_rate": 3.1029804353829517e-06,
+      "loss": 0.8437,
+      "step": 1492
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.19400900602340698,
+      "learning_rate": 3.0549331796612235e-06,
+      "loss": 0.8488,
+      "step": 1493
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1854696273803711,
+      "learning_rate": 3.0072550490328753e-06,
+      "loss": 0.7935,
+      "step": 1494
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18639519810676575,
+      "learning_rate": 2.959946225035726e-06,
+      "loss": 0.8802,
+      "step": 1495
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.18679916858673096,
+      "learning_rate": 2.913006887801395e-06,
+      "loss": 0.8483,
+      "step": 1496
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1891224980354309,
+      "learning_rate": 2.866437216054696e-06,
+      "loss": 0.8171,
+      "step": 1497
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.1865210384130478,
+      "learning_rate": 2.8202373871128895e-06,
+      "loss": 0.8446,
+      "step": 1498
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18717628717422485,
+      "learning_rate": 2.774407576885041e-06,
+      "loss": 0.8113,
+      "step": 1499
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18212980031967163,
+      "learning_rate": 2.728947959871353e-06,
+      "loss": 0.8735,
+      "step": 1500
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18781137466430664,
+      "learning_rate": 2.683858709162468e-06,
+      "loss": 0.8812,
+      "step": 1501
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.20542103052139282,
+      "learning_rate": 2.639139996438844e-06,
+      "loss": 0.8563,
+      "step": 1502
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.2054072469472885,
+      "learning_rate": 2.5947919919701136e-06,
+      "loss": 0.9012,
+      "step": 1503
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.1905122548341751,
+      "learning_rate": 2.5508148646143927e-06,
+      "loss": 0.8266,
+      "step": 1504
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18972791731357574,
+      "learning_rate": 2.5072087818176382e-06,
+      "loss": 0.8243,
+      "step": 1505
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19506774842739105,
+      "learning_rate": 2.4639739096130707e-06,
+      "loss": 0.9081,
+      "step": 1506
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19311931729316711,
+      "learning_rate": 2.421110412620453e-06,
+      "loss": 0.8152,
+      "step": 1507
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19823303818702698,
+      "learning_rate": 2.3786184540455448e-06,
+      "loss": 0.8052,
+      "step": 1508
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18085291981697083,
+      "learning_rate": 2.3364981956794487e-06,
+      "loss": 0.8011,
+      "step": 1509
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.17964868247509003,
+      "learning_rate": 2.294749797897955e-06,
+      "loss": 0.8142,
+      "step": 1510
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19236071407794952,
+      "learning_rate": 2.25337341966102e-06,
+      "loss": 0.847,
+      "step": 1511
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.18684546649456024,
+      "learning_rate": 2.212369218512078e-06,
+      "loss": 0.8767,
+      "step": 1512
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19908437132835388,
+      "learning_rate": 2.171737350577485e-06,
+      "loss": 0.8869,
+      "step": 1513
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.190731480717659,
+      "learning_rate": 2.1314779705658982e-06,
+      "loss": 0.8711,
+      "step": 1514
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.19613629579544067,
+      "learning_rate": 2.091591231767709e-06,
+      "loss": 0.8517,
+      "step": 1515
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18415379524230957,
+      "learning_rate": 2.0520772860544767e-06,
+      "loss": 0.8196,
+      "step": 1516
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.20092222094535828,
+      "learning_rate": 2.012936283878275e-06,
+      "loss": 0.8588,
+      "step": 1517
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19058625400066376,
+      "learning_rate": 1.9741683742712012e-06,
+      "loss": 0.8535,
+      "step": 1518
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19487521052360535,
+      "learning_rate": 1.9357737048447786e-06,
+      "loss": 0.8191,
+      "step": 1519
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18148715794086456,
+      "learning_rate": 1.8977524217893783e-06,
+      "loss": 0.8515,
+      "step": 1520
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19199548661708832,
+      "learning_rate": 1.8601046698736858e-06,
+      "loss": 0.8423,
+      "step": 1521
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.17849117517471313,
+      "learning_rate": 1.822830592444147e-06,
+      "loss": 0.8801,
+      "step": 1522
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19173994660377502,
+      "learning_rate": 1.7859303314243791e-06,
+      "loss": 0.9195,
+      "step": 1523
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.196760356426239,
+      "learning_rate": 1.7494040273147382e-06,
+      "loss": 0.8901,
+      "step": 1524
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18716773390769958,
+      "learning_rate": 1.7132518191916414e-06,
+      "loss": 0.8743,
+      "step": 1525
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18632376194000244,
+      "learning_rate": 1.6774738447071559e-06,
+      "loss": 0.8172,
+      "step": 1526
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19075815379619598,
+      "learning_rate": 1.642070240088378e-06,
+      "loss": 0.8216,
+      "step": 1527
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1906934678554535,
+      "learning_rate": 1.6070411401370334e-06,
+      "loss": 0.8492,
+      "step": 1528
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.18529535830020905,
+      "learning_rate": 1.5723866782288543e-06,
+      "loss": 0.855,
+      "step": 1529
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.19201524555683136,
+      "learning_rate": 1.5381069863131037e-06,
+      "loss": 0.887,
+      "step": 1530
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.1988871693611145,
+      "learning_rate": 1.5042021949120966e-06,
+      "loss": 0.8154,
+      "step": 1531
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19770891964435577,
+      "learning_rate": 1.4706724331207122e-06,
+      "loss": 0.7946,
+      "step": 1532
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19996926188468933,
+      "learning_rate": 1.4375178286058167e-06,
+      "loss": 0.8212,
+      "step": 1533
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.17670512199401855,
+      "learning_rate": 1.4047385076059071e-06,
+      "loss": 0.8156,
+      "step": 1534
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1882072389125824,
+      "learning_rate": 1.3723345949305245e-06,
+      "loss": 0.8264,
+      "step": 1535
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18426646292209625,
+      "learning_rate": 1.3403062139598076e-06,
+      "loss": 0.8729,
+      "step": 1536
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1895371824502945,
+      "learning_rate": 1.3086534866440513e-06,
+      "loss": 0.8512,
+      "step": 1537
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1917715072631836,
+      "learning_rate": 1.2773765335032383e-06,
+      "loss": 0.7989,
+      "step": 1538
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19698894023895264,
+      "learning_rate": 1.2464754736265182e-06,
+      "loss": 0.8456,
+      "step": 1539
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19695410132408142,
+      "learning_rate": 1.2159504246718522e-06,
+      "loss": 0.8665,
+      "step": 1540
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.19373415410518646,
+      "learning_rate": 1.1858015028654801e-06,
+      "loss": 0.8755,
+      "step": 1541
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.2013823539018631,
+      "learning_rate": 1.1560288230015203e-06,
+      "loss": 0.8756,
+      "step": 1542
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18109995126724243,
+      "learning_rate": 1.1266324984415266e-06,
+      "loss": 0.8512,
+      "step": 1543
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18117167055606842,
+      "learning_rate": 1.0976126411140431e-06,
+      "loss": 0.7767,
+      "step": 1544
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18761976063251495,
+      "learning_rate": 1.0689693615142161e-06,
+      "loss": 0.8142,
+      "step": 1545
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.1847393959760666,
+      "learning_rate": 1.0407027687033166e-06,
+      "loss": 0.8127,
+      "step": 1546
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.18555408716201782,
+      "learning_rate": 1.0128129703083633e-06,
+      "loss": 0.8698,
+      "step": 1547
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19472791254520416,
+      "learning_rate": 9.85300072521711e-07,
+      "loss": 0.8666,
+      "step": 1548
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1941266655921936,
+      "learning_rate": 9.58164180100629e-07,
+      "loss": 0.8508,
+      "step": 1549
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19015352427959442,
+      "learning_rate": 9.314053963669245e-07,
+      "loss": 0.8784,
+      "step": 1550
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18552695214748383,
+      "learning_rate": 9.0502382320653e-07,
+      "loss": 0.8351,
+      "step": 1551
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20146411657333374,
+      "learning_rate": 8.790195610691054e-07,
+      "loss": 0.877,
+      "step": 1552
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17787180840969086,
+      "learning_rate": 8.533927089677152e-07,
+      "loss": 0.791,
+      "step": 1553
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.20449447631835938,
+      "learning_rate": 8.281433644783621e-07,
+      "loss": 0.8711,
+      "step": 1554
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.1916695535182953,
+      "learning_rate": 8.032716237396987e-07,
+      "loss": 0.8282,
+      "step": 1555
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18444563448429108,
+      "learning_rate": 7.787775814526054e-07,
+      "loss": 0.8393,
+      "step": 1556
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18714573979377747,
+      "learning_rate": 7.546613308798466e-07,
+      "loss": 0.8594,
+      "step": 1557
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19066014885902405,
+      "learning_rate": 7.309229638457371e-07,
+      "loss": 0.9008,
+      "step": 1558
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18513765931129456,
+      "learning_rate": 7.075625707357536e-07,
+      "loss": 0.8332,
+      "step": 1559
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.17529350519180298,
+      "learning_rate": 6.845802404962243e-07,
+      "loss": 0.8613,
+      "step": 1560
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.19312287867069244,
+      "learning_rate": 6.619760606339731e-07,
+      "loss": 0.8898,
+      "step": 1561
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18521131575107574,
+      "learning_rate": 6.397501172159869e-07,
+      "loss": 0.8167,
+      "step": 1562
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.18218614161014557,
+      "learning_rate": 6.179024948690937e-07,
+      "loss": 0.8304,
+      "step": 1563
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1887504905462265,
+      "learning_rate": 5.964332767796399e-07,
+      "loss": 0.8706,
+      "step": 1564
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18959863483905792,
+      "learning_rate": 5.753425446931582e-07,
+      "loss": 0.8757,
+      "step": 1565
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18973901867866516,
+      "learning_rate": 5.546303789140894e-07,
+      "loss": 0.862,
+      "step": 1566
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18475745618343353,
+      "learning_rate": 5.342968583054498e-07,
+      "loss": 0.8447,
+      "step": 1567
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.18344885110855103,
+      "learning_rate": 5.143420602885308e-07,
+      "loss": 0.8047,
+      "step": 1568
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1916637271642685,
+      "learning_rate": 4.947660608426108e-07,
+      "loss": 0.8343,
+      "step": 1569
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.17382416129112244,
+      "learning_rate": 4.7556893450466653e-07,
+      "loss": 0.7685,
+      "step": 1570
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19896021485328674,
+      "learning_rate": 4.567507543691174e-07,
+      "loss": 0.8999,
+      "step": 1571
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1786774843931198,
+      "learning_rate": 4.383115920874814e-07,
+      "loss": 0.7936,
+      "step": 1572
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19705265760421753,
+      "learning_rate": 4.202515178681421e-07,
+      "loss": 0.9283,
+      "step": 1573
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19207850098609924,
+      "learning_rate": 4.025706004760932e-07,
+      "loss": 0.8473,
+      "step": 1574
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.20526379346847534,
+      "learning_rate": 3.8526890723268314e-07,
+      "loss": 0.8519,
+      "step": 1575
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19999198615550995,
+      "learning_rate": 3.6834650401528223e-07,
+      "loss": 0.8611,
+      "step": 1576
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.196251779794693,
+      "learning_rate": 3.51803455257127e-07,
+      "loss": 0.8502,
+      "step": 1577
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.19108015298843384,
+      "learning_rate": 3.3563982394704266e-07,
+      "loss": 0.7887,
+      "step": 1578
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.1788042187690735,
+      "learning_rate": 3.1985567162917673e-07,
+      "loss": 0.8769,
+      "step": 1579
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17838014662265778,
+      "learning_rate": 3.044510584027771e-07,
+      "loss": 0.8598,
+      "step": 1580
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18201227486133575,
+      "learning_rate": 2.8942604292199193e-07,
+      "loss": 0.8505,
+      "step": 1581
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1932731419801712,
+      "learning_rate": 2.747806823956034e-07,
+      "loss": 0.842,
+      "step": 1582
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18500591814517975,
+      "learning_rate": 2.6051503258684993e-07,
+      "loss": 0.8331,
+      "step": 1583
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18833038210868835,
+      "learning_rate": 2.4662914781318213e-07,
+      "loss": 0.8105,
+      "step": 1584
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19542557001113892,
+      "learning_rate": 2.3312308094607382e-07,
+      "loss": 0.8778,
+      "step": 1585
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18996158242225647,
+      "learning_rate": 2.1999688341083346e-07,
+      "loss": 0.8909,
+      "step": 1586
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19269198179244995,
+      "learning_rate": 2.0725060518633764e-07,
+      "loss": 0.8453,
+      "step": 1587
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1828884780406952,
+      "learning_rate": 1.948842948049534e-07,
+      "loss": 0.8752,
+      "step": 1588
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.19072295725345612,
+      "learning_rate": 1.828979993522717e-07,
+      "loss": 0.8841,
+      "step": 1589
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1883150041103363,
+      "learning_rate": 1.7129176446692984e-07,
+      "loss": 0.8682,
+      "step": 1590
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17624877393245697,
+      "learning_rate": 1.6006563434047828e-07,
+      "loss": 0.7998,
+      "step": 1591
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.1911545991897583,
+      "learning_rate": 1.4921965171720287e-07,
+      "loss": 0.8787,
+      "step": 1592
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.17575067281723022,
+      "learning_rate": 1.3875385789392515e-07,
+      "loss": 0.8219,
+      "step": 1593
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18420179188251495,
+      "learning_rate": 1.2866829271989124e-07,
+      "loss": 0.833,
+      "step": 1594
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18418926000595093,
+      "learning_rate": 1.1896299459658311e-07,
+      "loss": 0.8024,
+      "step": 1595
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.18934619426727295,
+      "learning_rate": 1.0963800047760763e-07,
+      "loss": 0.8634,
+      "step": 1596
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19058296084403992,
+      "learning_rate": 1.0069334586854107e-07,
+      "loss": 0.8622,
+      "step": 1597
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18763324618339539,
+      "learning_rate": 9.212906482675143e-08,
+      "loss": 0.8161,
+      "step": 1598
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18716083467006683,
+      "learning_rate": 8.394518996135414e-08,
+      "loss": 0.8357,
+      "step": 1599
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19783692061901093,
+      "learning_rate": 7.614175243301213e-08,
+      "loss": 0.863,
+      "step": 1600
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17088575661182404,
+      "learning_rate": 6.871878195386927e-08,
+      "loss": 0.8017,
+      "step": 1601
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18808765709400177,
+      "learning_rate": 6.16763067873949e-08,
+      "loss": 0.7873,
+      "step": 1602
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18691466748714447,
+      "learning_rate": 5.5014353748295e-08,
+      "loss": 0.8197,
+      "step": 1603
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1908489465713501,
+      "learning_rate": 4.873294820244567e-08,
+      "loss": 0.8337,
+      "step": 1604
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.1986658275127411,
+      "learning_rate": 4.283211406670429e-08,
+      "loss": 0.866,
+      "step": 1605
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.20240667462348938,
+      "learning_rate": 3.731187380893175e-08,
+      "loss": 0.8634,
+      "step": 1606
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18944308161735535,
+      "learning_rate": 3.217224844782596e-08,
+      "loss": 0.8072,
+      "step": 1607
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.19379983842372894,
+      "learning_rate": 2.7413257552855177e-08,
+      "loss": 0.8459,
+      "step": 1608
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.17960891127586365,
+      "learning_rate": 2.303491924421364e-08,
+      "loss": 0.812,
+      "step": 1609
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18622823059558868,
+      "learning_rate": 1.9037250192732726e-08,
+      "loss": 0.8537,
+      "step": 1610
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18758083879947662,
+      "learning_rate": 1.5420265619803254e-08,
+      "loss": 0.8032,
+      "step": 1611
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.18631847202777863,
+      "learning_rate": 1.2183979297364368e-08,
+      "loss": 0.8346,
+      "step": 1612
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1840689778327942,
+      "learning_rate": 9.32840354779252e-09,
+      "loss": 0.7965,
+      "step": 1613
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.20787417888641357,
+      "learning_rate": 6.8535492439014694e-09,
+      "loss": 0.8893,
+      "step": 1614
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1837613731622696,
+      "learning_rate": 4.759425808853468e-09,
+      "loss": 0.8367,
+      "step": 1615
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19482795894145966,
+      "learning_rate": 3.046041216181461e-09,
+      "loss": 0.8424,
+      "step": 1616
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18247024714946747,
+      "learning_rate": 1.7134019897113717e-09,
+      "loss": 0.7786,
+      "step": 1617
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.18248644471168518,
+      "learning_rate": 7.615132035510008e-10,
+      "loss": 0.8388,
+      "step": 1618
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1934431791305542,
+      "learning_rate": 1.903784821122301e-10,
+      "loss": 0.8494,
+      "step": 1619
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.1892014741897583,
+      "learning_rate": 0.0,
+      "loss": 0.8367,
+      "step": 1620
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 1.5195412635803223,
+      "eval_runtime": 187.7834,
+      "eval_samples_per_second": 146.323,
+      "eval_steps_per_second": 73.164,
+      "step": 1620
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1620,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "total_flos": 4.5950204436597965e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}