diff --git "a/checkpoint-2372/trainer_state.json" "b/checkpoint-2372/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2372/trainer_state.json"
@@ -0,0 +1,16637 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998946148171567,
+  "eval_steps": 500,
+  "global_step": 2372,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00042154073137316893,
+      "grad_norm": 1.554450511932373,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.9124,
+      "step": 1
+    },
+    {
+      "epoch": 0.0008430814627463379,
+      "grad_norm": 1.5286387205123901,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.8797,
+      "step": 2
+    },
+    {
+      "epoch": 0.001264622194119507,
+      "grad_norm": 1.5313383340835571,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.9107,
+      "step": 3
+    },
+    {
+      "epoch": 0.0016861629254926757,
+      "grad_norm": 1.668462872505188,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.8902,
+      "step": 4
+    },
+    {
+      "epoch": 0.0021077036568658445,
+      "grad_norm": 1.5774749517440796,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.9171,
+      "step": 5
+    },
+    {
+      "epoch": 0.002529244388239014,
+      "grad_norm": 1.6463191509246826,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.8578,
+      "step": 6
+    },
+    {
+      "epoch": 0.0029507851196121826,
+      "grad_norm": 1.5910251140594482,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 0.8858,
+      "step": 7
+    },
+    {
+      "epoch": 0.0033723258509853514,
+      "grad_norm": 1.573474645614624,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.8712,
+      "step": 8
+    },
+    {
+      "epoch": 0.0037938665823585203,
+      "grad_norm": 1.6730811595916748,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 0.8757,
+      "step": 9
+    },
+    {
+      "epoch": 0.004215407313731689,
+      "grad_norm": 1.6088658571243286,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.9272,
+      "step": 10
+    },
+    {
+      "epoch": 0.004636948045104858,
+      "grad_norm": 1.4241929054260254,
+      "learning_rate": 5.5e-07,
+      "loss": 0.8652,
+      "step": 11
+    },
+    {
+      "epoch": 0.005058488776478028,
+      "grad_norm": 1.4424278736114502,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8822,
+      "step": 12
+    },
+    {
+      "epoch": 0.005480029507851196,
+      "grad_norm": 1.4737945795059204,
+      "learning_rate": 6.5e-07,
+      "loss": 0.8871,
+      "step": 13
+    },
+    {
+      "epoch": 0.005901570239224365,
+      "grad_norm": 1.3290244340896606,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.8616,
+      "step": 14
+    },
+    {
+      "epoch": 0.006323110970597534,
+      "grad_norm": 1.3889096975326538,
+      "learning_rate": 7.5e-07,
+      "loss": 0.8775,
+      "step": 15
+    },
+    {
+      "epoch": 0.006744651701970703,
+      "grad_norm": 1.319577693939209,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.8907,
+      "step": 16
+    },
+    {
+      "epoch": 0.007166192433343872,
+      "grad_norm": 1.1662847995758057,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.8981,
+      "step": 17
+    },
+    {
+      "epoch": 0.0075877331647170405,
+      "grad_norm": 1.197760820388794,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.8417,
+      "step": 18
+    },
+    {
+      "epoch": 0.00800927389609021,
+      "grad_norm": 1.0497264862060547,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.8447,
+      "step": 19
+    },
+    {
+      "epoch": 0.008430814627463378,
+      "grad_norm": 1.058746337890625,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.8676,
+      "step": 20
+    },
+    {
+      "epoch": 0.008852355358836548,
+      "grad_norm": 0.942598819732666,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.8565,
+      "step": 21
+    },
+    {
+      "epoch": 0.009273896090209717,
+      "grad_norm": 0.9345218539237976,
+      "learning_rate": 1.1e-06,
+      "loss": 0.859,
+      "step": 22
+    },
+    {
+      "epoch": 0.009695436821582885,
+      "grad_norm": 0.8632031679153442,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.8711,
+      "step": 23
+    },
+    {
+      "epoch": 0.010116977552956055,
+      "grad_norm": 0.794277012348175,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.8619,
+      "step": 24
+    },
+    {
+      "epoch": 0.010538518284329224,
+      "grad_norm": 0.7953000664710999,
+      "learning_rate": 1.25e-06,
+      "loss": 0.8261,
+      "step": 25
+    },
+    {
+      "epoch": 0.010960059015702392,
+      "grad_norm": 0.7987999320030212,
+      "learning_rate": 1.3e-06,
+      "loss": 0.843,
+      "step": 26
+    },
+    {
+      "epoch": 0.01138159974707556,
+      "grad_norm": 0.7570469975471497,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.8293,
+      "step": 27
+    },
+    {
+      "epoch": 0.01180314047844873,
+      "grad_norm": 0.7138553857803345,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.8295,
+      "step": 28
+    },
+    {
+      "epoch": 0.012224681209821899,
+      "grad_norm": 0.6718113422393799,
+      "learning_rate": 1.45e-06,
+      "loss": 0.8251,
+      "step": 29
+    },
+    {
+      "epoch": 0.012646221941195067,
+      "grad_norm": 0.6300117373466492,
+      "learning_rate": 1.5e-06,
+      "loss": 0.7955,
+      "step": 30
+    },
+    {
+      "epoch": 0.013067762672568237,
+      "grad_norm": 0.6070649027824402,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.7859,
+      "step": 31
+    },
+    {
+      "epoch": 0.013489303403941406,
+      "grad_norm": 0.5400272011756897,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.7935,
+      "step": 32
+    },
+    {
+      "epoch": 0.013910844135314574,
+      "grad_norm": 0.5371619462966919,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.8271,
+      "step": 33
+    },
+    {
+      "epoch": 0.014332384866687744,
+      "grad_norm": 0.48049041628837585,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.7997,
+      "step": 34
+    },
+    {
+      "epoch": 0.014753925598060913,
+      "grad_norm": 0.47074759006500244,
+      "learning_rate": 1.75e-06,
+      "loss": 0.789,
+      "step": 35
+    },
+    {
+      "epoch": 0.015175466329434081,
+      "grad_norm": 0.4344562292098999,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.7876,
+      "step": 36
+    },
+    {
+      "epoch": 0.015597007060807251,
+      "grad_norm": 0.4061878025531769,
+      "learning_rate": 1.85e-06,
+      "loss": 0.777,
+      "step": 37
+    },
+    {
+      "epoch": 0.01601854779218042,
+      "grad_norm": 0.39230242371559143,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.8033,
+      "step": 38
+    },
+    {
+      "epoch": 0.016440088523553588,
+      "grad_norm": 0.3969016373157501,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.7837,
+      "step": 39
+    },
+    {
+      "epoch": 0.016861629254926756,
+      "grad_norm": 0.3729524612426758,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7795,
+      "step": 40
+    },
+    {
+      "epoch": 0.017283169986299925,
+      "grad_norm": 0.3697626292705536,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7919,
+      "step": 41
+    },
+    {
+      "epoch": 0.017704710717673097,
+      "grad_norm": 0.34239786863327026,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.7544,
+      "step": 42
+    },
+    {
+      "epoch": 0.018126251449046265,
+      "grad_norm": 0.36224567890167236,
+      "learning_rate": 2.15e-06,
+      "loss": 0.7769,
+      "step": 43
+    },
+    {
+      "epoch": 0.018547792180419433,
+      "grad_norm": 0.3495021462440491,
+      "learning_rate": 2.2e-06,
+      "loss": 0.746,
+      "step": 44
+    },
+    {
+      "epoch": 0.0189693329117926,
+      "grad_norm": 0.33883053064346313,
+      "learning_rate": 2.25e-06,
+      "loss": 0.77,
+      "step": 45
+    },
+    {
+      "epoch": 0.01939087364316577,
+      "grad_norm": 0.30793020129203796,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.7325,
+      "step": 46
+    },
+    {
+      "epoch": 0.01981241437453894,
+      "grad_norm": 0.31899771094322205,
+      "learning_rate": 2.35e-06,
+      "loss": 0.7739,
+      "step": 47
+    },
+    {
+      "epoch": 0.02023395510591211,
+      "grad_norm": 0.31002360582351685,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.7711,
+      "step": 48
+    },
+    {
+      "epoch": 0.02065549583728528,
+      "grad_norm": 0.29887646436691284,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.7484,
+      "step": 49
+    },
+    {
+      "epoch": 0.021077036568658447,
+      "grad_norm": 0.2904997766017914,
+      "learning_rate": 2.5e-06,
+      "loss": 0.755,
+      "step": 50
+    },
+    {
+      "epoch": 0.021498577300031615,
+      "grad_norm": 0.2875346541404724,
+      "learning_rate": 2.55e-06,
+      "loss": 0.7764,
+      "step": 51
+    },
+    {
+      "epoch": 0.021920118031404784,
+      "grad_norm": 0.2731809914112091,
+      "learning_rate": 2.6e-06,
+      "loss": 0.7518,
+      "step": 52
+    },
+    {
+      "epoch": 0.022341658762777952,
+      "grad_norm": 0.25800859928131104,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.7376,
+      "step": 53
+    },
+    {
+      "epoch": 0.02276319949415112,
+      "grad_norm": 0.2637141942977905,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.7376,
+      "step": 54
+    },
+    {
+      "epoch": 0.023184740225524293,
+      "grad_norm": 0.2621852457523346,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.7443,
+      "step": 55
+    },
+    {
+      "epoch": 0.02360628095689746,
+      "grad_norm": 0.24826474487781525,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7312,
+      "step": 56
+    },
+    {
+      "epoch": 0.02402782168827063,
+      "grad_norm": 0.2539840340614319,
+      "learning_rate": 2.85e-06,
+      "loss": 0.7306,
+      "step": 57
+    },
+    {
+      "epoch": 0.024449362419643798,
+      "grad_norm": 0.25040122866630554,
+      "learning_rate": 2.9e-06,
+      "loss": 0.7192,
+      "step": 58
+    },
+    {
+      "epoch": 0.024870903151016966,
+      "grad_norm": 0.23641498386859894,
+      "learning_rate": 2.95e-06,
+      "loss": 0.7518,
+      "step": 59
+    },
+    {
+      "epoch": 0.025292443882390134,
+      "grad_norm": 0.2318529635667801,
+      "learning_rate": 3e-06,
+      "loss": 0.7407,
+      "step": 60
+    },
+    {
+      "epoch": 0.025713984613763306,
+      "grad_norm": 0.22770705819129944,
+      "learning_rate": 3.05e-06,
+      "loss": 0.7297,
+      "step": 61
+    },
+    {
+      "epoch": 0.026135525345136475,
+      "grad_norm": 0.22948750853538513,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.7299,
+      "step": 62
+    },
+    {
+      "epoch": 0.026557066076509643,
+      "grad_norm": 0.2390872687101364,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.7325,
+      "step": 63
+    },
+    {
+      "epoch": 0.02697860680788281,
+      "grad_norm": 0.2283066362142563,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.7297,
+      "step": 64
+    },
+    {
+      "epoch": 0.02740014753925598,
+      "grad_norm": 0.2239808738231659,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.7686,
+      "step": 65
+    },
+    {
+      "epoch": 0.027821688270629148,
+      "grad_norm": 0.21871443092823029,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.7136,
+      "step": 66
+    },
+    {
+      "epoch": 0.02824322900200232,
+      "grad_norm": 0.2268369197845459,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.7636,
+      "step": 67
+    },
+    {
+      "epoch": 0.02866476973337549,
+      "grad_norm": 0.25275668501853943,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.777,
+      "step": 68
+    },
+    {
+      "epoch": 0.029086310464748657,
+      "grad_norm": 0.22813616693019867,
+      "learning_rate": 3.45e-06,
+      "loss": 0.7453,
+      "step": 69
+    },
+    {
+      "epoch": 0.029507851196121825,
+      "grad_norm": 0.22262635827064514,
+      "learning_rate": 3.5e-06,
+      "loss": 0.76,
+      "step": 70
+    },
+    {
+      "epoch": 0.029929391927494994,
+      "grad_norm": 0.20812402665615082,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.7437,
+      "step": 71
+    },
+    {
+      "epoch": 0.030350932658868162,
+      "grad_norm": 0.20619122684001923,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7258,
+      "step": 72
+    },
+    {
+      "epoch": 0.03077247339024133,
+      "grad_norm": 0.20562823116779327,
+      "learning_rate": 3.65e-06,
+      "loss": 0.7273,
+      "step": 73
+    },
+    {
+      "epoch": 0.031194014121614502,
+      "grad_norm": 0.1989251971244812,
+      "learning_rate": 3.7e-06,
+      "loss": 0.7462,
+      "step": 74
+    },
+    {
+      "epoch": 0.03161555485298767,
+      "grad_norm": 0.20220854878425598,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.7316,
+      "step": 75
+    },
+    {
+      "epoch": 0.03203709558436084,
+      "grad_norm": 0.19772911071777344,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.7414,
+      "step": 76
+    },
+    {
+      "epoch": 0.03245863631573401,
+      "grad_norm": 0.19900061190128326,
+      "learning_rate": 3.85e-06,
+      "loss": 0.6924,
+      "step": 77
+    },
+    {
+      "epoch": 0.032880177047107176,
+      "grad_norm": 0.18977713584899902,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.7117,
+      "step": 78
+    },
+    {
+      "epoch": 0.033301717778480344,
+      "grad_norm": 0.19983789324760437,
+      "learning_rate": 3.95e-06,
+      "loss": 0.7174,
+      "step": 79
+    },
+    {
+      "epoch": 0.03372325850985351,
+      "grad_norm": 0.1869533509016037,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.75,
+      "step": 80
+    },
+    {
+      "epoch": 0.03414479924122668,
+      "grad_norm": 0.1975327730178833,
+      "learning_rate": 4.05e-06,
+      "loss": 0.7586,
+      "step": 81
+    },
+    {
+      "epoch": 0.03456633997259985,
+      "grad_norm": 0.18499982357025146,
+      "learning_rate": 4.1e-06,
+      "loss": 0.7194,
+      "step": 82
+    },
+    {
+      "epoch": 0.034987880703973025,
+      "grad_norm": 0.1828499436378479,
+      "learning_rate": 4.15e-06,
+      "loss": 0.7232,
+      "step": 83
+    },
+    {
+      "epoch": 0.03540942143534619,
+      "grad_norm": 0.18891121447086334,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7064,
+      "step": 84
+    },
+    {
+      "epoch": 0.03583096216671936,
+      "grad_norm": 0.1964845210313797,
+      "learning_rate": 4.25e-06,
+      "loss": 0.72,
+      "step": 85
+    },
+    {
+      "epoch": 0.03625250289809253,
+      "grad_norm": 0.17995160818099976,
+      "learning_rate": 4.3e-06,
+      "loss": 0.7146,
+      "step": 86
+    },
+    {
+      "epoch": 0.0366740436294657,
+      "grad_norm": 0.18630728125572205,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.7168,
+      "step": 87
+    },
+    {
+      "epoch": 0.03709558436083887,
+      "grad_norm": 0.18258942663669586,
+      "learning_rate": 4.4e-06,
+      "loss": 0.714,
+      "step": 88
+    },
+    {
+      "epoch": 0.037517125092212035,
+      "grad_norm": 0.17748843133449554,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.7154,
+      "step": 89
+    },
+    {
+      "epoch": 0.0379386658235852,
+      "grad_norm": 0.1819748431444168,
+      "learning_rate": 4.5e-06,
+      "loss": 0.7111,
+      "step": 90
+    },
+    {
+      "epoch": 0.03836020655495837,
+      "grad_norm": 0.18879565596580505,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.6919,
+      "step": 91
+    },
+    {
+      "epoch": 0.03878174728633154,
+      "grad_norm": 0.18279008567333221,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.7635,
+      "step": 92
+    },
+    {
+      "epoch": 0.03920328801770471,
+      "grad_norm": 0.1807212084531784,
+      "learning_rate": 4.65e-06,
+      "loss": 0.714,
+      "step": 93
+    },
+    {
+      "epoch": 0.03962482874907788,
+      "grad_norm": 0.18220968544483185,
+      "learning_rate": 4.7e-06,
+      "loss": 0.7166,
+      "step": 94
+    },
+    {
+      "epoch": 0.040046369480451045,
+      "grad_norm": 0.18105052411556244,
+      "learning_rate": 4.75e-06,
+      "loss": 0.716,
+      "step": 95
+    },
+    {
+      "epoch": 0.04046791021182422,
+      "grad_norm": 0.18787191808223724,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7285,
+      "step": 96
+    },
+    {
+      "epoch": 0.04088945094319739,
+      "grad_norm": 0.176447331905365,
+      "learning_rate": 4.85e-06,
+      "loss": 0.6979,
+      "step": 97
+    },
+    {
+      "epoch": 0.04131099167457056,
+      "grad_norm": 0.17359976470470428,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.7048,
+      "step": 98
+    },
+    {
+      "epoch": 0.041732532405943726,
+      "grad_norm": 0.1798487901687622,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7714,
+      "step": 99
+    },
+    {
+      "epoch": 0.042154073137316894,
+      "grad_norm": 0.18481405079364777,
+      "learning_rate": 5e-06,
+      "loss": 0.7103,
+      "step": 100
+    },
+    {
+      "epoch": 0.04257561386869006,
+      "grad_norm": 0.1738777607679367,
+      "learning_rate": 4.9999997493714225e-06,
+      "loss": 0.709,
+      "step": 101
+    },
+    {
+      "epoch": 0.04299715460006323,
+      "grad_norm": 0.1754748374223709,
+      "learning_rate": 4.999998997485738e-06,
+      "loss": 0.7277,
+      "step": 102
+    },
+    {
+      "epoch": 0.0434186953314364,
+      "grad_norm": 0.17723973095417023,
+      "learning_rate": 4.999997744343098e-06,
+      "loss": 0.6855,
+      "step": 103
+    },
+    {
+      "epoch": 0.04384023606280957,
+      "grad_norm": 0.17454974353313446,
+      "learning_rate": 4.999995989943754e-06,
+      "loss": 0.7123,
+      "step": 104
+    },
+    {
+      "epoch": 0.044261776794182736,
+      "grad_norm": 0.1745677888393402,
+      "learning_rate": 4.999993734288058e-06,
+      "loss": 0.7089,
+      "step": 105
+    },
+    {
+      "epoch": 0.044683317525555905,
+      "grad_norm": 0.16998346149921417,
+      "learning_rate": 4.999990977376461e-06,
+      "loss": 0.7123,
+      "step": 106
+    },
+    {
+      "epoch": 0.04510485825692907,
+      "grad_norm": 0.17023731768131256,
+      "learning_rate": 4.9999877192095166e-06,
+      "loss": 0.6998,
+      "step": 107
+    },
+    {
+      "epoch": 0.04552639898830224,
+      "grad_norm": 0.1744530200958252,
+      "learning_rate": 4.999983959787878e-06,
+      "loss": 0.7051,
+      "step": 108
+    },
+    {
+      "epoch": 0.04594793971967542,
+      "grad_norm": 0.17939868569374084,
+      "learning_rate": 4.999979699112301e-06,
+      "loss": 0.7117,
+      "step": 109
+    },
+    {
+      "epoch": 0.046369480451048585,
+      "grad_norm": 0.18414615094661713,
+      "learning_rate": 4.9999749371836355e-06,
+      "loss": 0.7399,
+      "step": 110
+    },
+    {
+      "epoch": 0.04679102118242175,
+      "grad_norm": 0.17194895446300507,
+      "learning_rate": 4.99996967400284e-06,
+      "loss": 0.7011,
+      "step": 111
+    },
+    {
+      "epoch": 0.04721256191379492,
+      "grad_norm": 0.17224766314029694,
+      "learning_rate": 4.999963909570969e-06,
+      "loss": 0.706,
+      "step": 112
+    },
+    {
+      "epoch": 0.04763410264516809,
+      "grad_norm": 0.18243743479251862,
+      "learning_rate": 4.999957643889176e-06,
+      "loss": 0.6706,
+      "step": 113
+    },
+    {
+      "epoch": 0.04805564337654126,
+      "grad_norm": 0.1826334446668625,
+      "learning_rate": 4.999950876958721e-06,
+      "loss": 0.6943,
+      "step": 114
+    },
+    {
+      "epoch": 0.04847718410791443,
+      "grad_norm": 0.17811767756938934,
+      "learning_rate": 4.9999436087809584e-06,
+      "loss": 0.7036,
+      "step": 115
+    },
+    {
+      "epoch": 0.048898724839287595,
+      "grad_norm": 0.1884814202785492,
+      "learning_rate": 4.999935839357345e-06,
+      "loss": 0.71,
+      "step": 116
+    },
+    {
+      "epoch": 0.049320265570660764,
+      "grad_norm": 0.17991511523723602,
+      "learning_rate": 4.99992756868944e-06,
+      "loss": 0.7033,
+      "step": 117
+    },
+    {
+      "epoch": 0.04974180630203393,
+      "grad_norm": 0.1852027177810669,
+      "learning_rate": 4.999918796778902e-06,
+      "loss": 0.6701,
+      "step": 118
+    },
+    {
+      "epoch": 0.0501633470334071,
+      "grad_norm": 0.17148995399475098,
+      "learning_rate": 4.999909523627488e-06,
+      "loss": 0.7017,
+      "step": 119
+    },
+    {
+      "epoch": 0.05058488776478027,
+      "grad_norm": 0.18856461346149445,
+      "learning_rate": 4.999899749237059e-06,
+      "loss": 0.7128,
+      "step": 120
+    },
+    {
+      "epoch": 0.051006428496153444,
+      "grad_norm": 0.18930798768997192,
+      "learning_rate": 4.999889473609574e-06,
+      "loss": 0.7093,
+      "step": 121
+    },
+    {
+      "epoch": 0.05142796922752661,
+      "grad_norm": 0.17779304087162018,
+      "learning_rate": 4.999878696747093e-06,
+      "loss": 0.6867,
+      "step": 122
+    },
+    {
+      "epoch": 0.05184950995889978,
+      "grad_norm": 0.1712871789932251,
+      "learning_rate": 4.999867418651778e-06,
+      "loss": 0.6628,
+      "step": 123
+    },
+    {
+      "epoch": 0.05227105069027295,
+      "grad_norm": 0.1712663471698761,
+      "learning_rate": 4.999855639325888e-06,
+      "loss": 0.6866,
+      "step": 124
+    },
+    {
+      "epoch": 0.05269259142164612,
+      "grad_norm": 0.1759692132472992,
+      "learning_rate": 4.9998433587717875e-06,
+      "loss": 0.6659,
+      "step": 125
+    },
+    {
+      "epoch": 0.053114132153019286,
+      "grad_norm": 0.17685872316360474,
+      "learning_rate": 4.999830576991937e-06,
+      "loss": 0.7081,
+      "step": 126
+    },
+    {
+      "epoch": 0.053535672884392455,
+      "grad_norm": 0.17527182400226593,
+      "learning_rate": 4.9998172939889016e-06,
+      "loss": 0.7222,
+      "step": 127
+    },
+    {
+      "epoch": 0.05395721361576562,
+      "grad_norm": 0.17963087558746338,
+      "learning_rate": 4.999803509765341e-06,
+      "loss": 0.6822,
+      "step": 128
+    },
+    {
+      "epoch": 0.05437875434713879,
+      "grad_norm": 0.179128959774971,
+      "learning_rate": 4.999789224324021e-06,
+      "loss": 0.6903,
+      "step": 129
+    },
+    {
+      "epoch": 0.05480029507851196,
+      "grad_norm": 0.1729232668876648,
+      "learning_rate": 4.999774437667806e-06,
+      "loss": 0.6871,
+      "step": 130
+    },
+    {
+      "epoch": 0.05522183580988513,
+      "grad_norm": 0.18572039902210236,
+      "learning_rate": 4.99975914979966e-06,
+      "loss": 0.7034,
+      "step": 131
+    },
+    {
+      "epoch": 0.055643376541258296,
+      "grad_norm": 0.1708957850933075,
+      "learning_rate": 4.99974336072265e-06,
+      "loss": 0.7316,
+      "step": 132
+    },
+    {
+      "epoch": 0.056064917272631465,
+      "grad_norm": 0.17814520001411438,
+      "learning_rate": 4.9997270704399395e-06,
+      "loss": 0.7147,
+      "step": 133
+    },
+    {
+      "epoch": 0.05648645800400464,
+      "grad_norm": 0.18104323744773865,
+      "learning_rate": 4.999710278954796e-06,
+      "loss": 0.6796,
+      "step": 134
+    },
+    {
+      "epoch": 0.05690799873537781,
+      "grad_norm": 0.17645549774169922,
+      "learning_rate": 4.999692986270586e-06,
+      "loss": 0.7012,
+      "step": 135
+    },
+    {
+      "epoch": 0.05732953946675098,
+      "grad_norm": 0.18215152621269226,
+      "learning_rate": 4.999675192390776e-06,
+      "loss": 0.6851,
+      "step": 136
+    },
+    {
+      "epoch": 0.057751080198124145,
+      "grad_norm": 0.16767218708992004,
+      "learning_rate": 4.999656897318935e-06,
+      "loss": 0.7002,
+      "step": 137
+    },
+    {
+      "epoch": 0.058172620929497314,
+      "grad_norm": 0.18999888002872467,
+      "learning_rate": 4.99963810105873e-06,
+      "loss": 0.6831,
+      "step": 138
+    },
+    {
+      "epoch": 0.05859416166087048,
+      "grad_norm": 0.17262063920497894,
+      "learning_rate": 4.999618803613931e-06,
+      "loss": 0.6632,
+      "step": 139
+    },
+    {
+      "epoch": 0.05901570239224365,
+      "grad_norm": 0.18284167349338531,
+      "learning_rate": 4.999599004988407e-06,
+      "loss": 0.6976,
+      "step": 140
+    },
+    {
+      "epoch": 0.05943724312361682,
+      "grad_norm": 0.19742456078529358,
+      "learning_rate": 4.999578705186126e-06,
+      "loss": 0.7024,
+      "step": 141
+    },
+    {
+      "epoch": 0.05985878385498999,
+      "grad_norm": 0.17782630026340485,
+      "learning_rate": 4.9995579042111595e-06,
+      "loss": 0.6669,
+      "step": 142
+    },
+    {
+      "epoch": 0.060280324586363156,
+      "grad_norm": 0.1842881292104721,
+      "learning_rate": 4.9995366020676785e-06,
+      "loss": 0.7113,
+      "step": 143
+    },
+    {
+      "epoch": 0.060701865317736324,
+      "grad_norm": 0.20911230146884918,
+      "learning_rate": 4.999514798759954e-06,
+      "loss": 0.6731,
+      "step": 144
+    },
+    {
+      "epoch": 0.06112340604910949,
+      "grad_norm": 0.17967282235622406,
+      "learning_rate": 4.999492494292357e-06,
+      "loss": 0.6911,
+      "step": 145
+    },
+    {
+      "epoch": 0.06154494678048266,
+      "grad_norm": 0.18562951683998108,
+      "learning_rate": 4.99946968866936e-06,
+      "loss": 0.6793,
+      "step": 146
+    },
+    {
+      "epoch": 0.061966487511855836,
+      "grad_norm": 0.18126961588859558,
+      "learning_rate": 4.999446381895535e-06,
+      "loss": 0.6966,
+      "step": 147
+    },
+    {
+      "epoch": 0.062388028243229005,
+      "grad_norm": 0.18489444255828857,
+      "learning_rate": 4.999422573975556e-06,
+      "loss": 0.6882,
+      "step": 148
+    },
+    {
+      "epoch": 0.06280956897460217,
+      "grad_norm": 0.19031500816345215,
+      "learning_rate": 4.999398264914196e-06,
+      "loss": 0.6814,
+      "step": 149
+    },
+    {
+      "epoch": 0.06323110970597534,
+      "grad_norm": 0.17611351609230042,
+      "learning_rate": 4.999373454716329e-06,
+      "loss": 0.6931,
+      "step": 150
+    },
+    {
+      "epoch": 0.06365265043734851,
+      "grad_norm": 0.17956188321113586,
+      "learning_rate": 4.999348143386931e-06,
+      "loss": 0.678,
+      "step": 151
+    },
+    {
+      "epoch": 0.06407419116872168,
+      "grad_norm": 0.17854593694210052,
+      "learning_rate": 4.999322330931074e-06,
+      "loss": 0.6815,
+      "step": 152
+    },
+    {
+      "epoch": 0.06449573190009485,
+      "grad_norm": 0.1742367446422577,
+      "learning_rate": 4.999296017353935e-06,
+      "loss": 0.6993,
+      "step": 153
+    },
+    {
+      "epoch": 0.06491727263146801,
+      "grad_norm": 0.1765994429588318,
+      "learning_rate": 4.999269202660791e-06,
+      "loss": 0.6779,
+      "step": 154
+    },
+    {
+      "epoch": 0.06533881336284118,
+      "grad_norm": 0.1769678294658661,
+      "learning_rate": 4.999241886857018e-06,
+      "loss": 0.6833,
+      "step": 155
+    },
+    {
+      "epoch": 0.06576035409421435,
+      "grad_norm": 0.17637397348880768,
+      "learning_rate": 4.999214069948092e-06,
+      "loss": 0.6837,
+      "step": 156
+    },
+    {
+      "epoch": 0.06618189482558752,
+      "grad_norm": 0.1915796846151352,
+      "learning_rate": 4.99918575193959e-06,
+      "loss": 0.68,
+      "step": 157
+    },
+    {
+      "epoch": 0.06660343555696069,
+      "grad_norm": 0.17869912087917328,
+      "learning_rate": 4.999156932837192e-06,
+      "loss": 0.6947,
+      "step": 158
+    },
+    {
+      "epoch": 0.06702497628833386,
+      "grad_norm": 0.17902164161205292,
+      "learning_rate": 4.999127612646674e-06,
+      "loss": 0.6911,
+      "step": 159
+    },
+    {
+      "epoch": 0.06744651701970703,
+      "grad_norm": 0.18028411269187927,
+      "learning_rate": 4.999097791373915e-06,
+      "loss": 0.6795,
+      "step": 160
+    },
+    {
+      "epoch": 0.0678680577510802,
+      "grad_norm": 0.1825106143951416,
+      "learning_rate": 4.999067469024897e-06,
+      "loss": 0.6786,
+      "step": 161
+    },
+    {
+      "epoch": 0.06828959848245336,
+      "grad_norm": 0.1832597702741623,
+      "learning_rate": 4.999036645605696e-06,
+      "loss": 0.679,
+      "step": 162
+    },
+    {
+      "epoch": 0.06871113921382653,
+      "grad_norm": 0.18045204877853394,
+      "learning_rate": 4.9990053211224946e-06,
+      "loss": 0.6817,
+      "step": 163
+    },
+    {
+      "epoch": 0.0691326799451997,
+      "grad_norm": 0.1871611773967743,
+      "learning_rate": 4.998973495581572e-06,
+      "loss": 0.7141,
+      "step": 164
+    },
+    {
+      "epoch": 0.06955422067657287,
+      "grad_norm": 0.17999723553657532,
+      "learning_rate": 4.998941168989311e-06,
+      "loss": 0.6645,
+      "step": 165
+    },
+    {
+      "epoch": 0.06997576140794605,
+      "grad_norm": 0.17337147891521454,
+      "learning_rate": 4.998908341352191e-06,
+      "loss": 0.669,
+      "step": 166
+    },
+    {
+      "epoch": 0.07039730213931922,
+      "grad_norm": 0.17942361533641815,
+      "learning_rate": 4.998875012676796e-06,
+      "loss": 0.6935,
+      "step": 167
+    },
+    {
+      "epoch": 0.07081884287069239,
+      "grad_norm": 0.17995475232601166,
+      "learning_rate": 4.998841182969808e-06,
+      "loss": 0.7071,
+      "step": 168
+    },
+    {
+      "epoch": 0.07124038360206555,
+      "grad_norm": 0.1790342926979065,
+      "learning_rate": 4.99880685223801e-06,
+      "loss": 0.6829,
+      "step": 169
+    },
+    {
+      "epoch": 0.07166192433343872,
+      "grad_norm": 0.18393510580062866,
+      "learning_rate": 4.998772020488284e-06,
+      "loss": 0.6833,
+      "step": 170
+    },
+    {
+      "epoch": 0.07208346506481189,
+      "grad_norm": 0.17788998782634735,
+      "learning_rate": 4.998736687727616e-06,
+      "loss": 0.6839,
+      "step": 171
+    },
+    {
+      "epoch": 0.07250500579618506,
+      "grad_norm": 0.1787301003932953,
+      "learning_rate": 4.998700853963088e-06,
+      "loss": 0.6961,
+      "step": 172
+    },
+    {
+      "epoch": 0.07292654652755823,
+      "grad_norm": 0.1817287653684616,
+      "learning_rate": 4.998664519201887e-06,
+      "loss": 0.6671,
+      "step": 173
+    },
+    {
+      "epoch": 0.0733480872589314,
+      "grad_norm": 0.18745344877243042,
+      "learning_rate": 4.998627683451297e-06,
+      "loss": 0.7037,
+      "step": 174
+    },
+    {
+      "epoch": 0.07376962799030456,
+      "grad_norm": 0.18669237196445465,
+      "learning_rate": 4.998590346718704e-06,
+      "loss": 0.6899,
+      "step": 175
+    },
+    {
+      "epoch": 0.07419116872167773,
+      "grad_norm": 0.1815669983625412,
+      "learning_rate": 4.998552509011594e-06,
+      "loss": 0.6693,
+      "step": 176
+    },
+    {
+      "epoch": 0.0746127094530509,
+      "grad_norm": 0.18319211900234222,
+      "learning_rate": 4.998514170337554e-06,
+      "loss": 0.6811,
+      "step": 177
+    },
+    {
+      "epoch": 0.07503425018442407,
+      "grad_norm": 0.1861065775156021,
+      "learning_rate": 4.99847533070427e-06,
+      "loss": 0.6855,
+      "step": 178
+    },
+    {
+      "epoch": 0.07545579091579724,
+      "grad_norm": 0.17398186028003693,
+      "learning_rate": 4.998435990119531e-06,
+      "loss": 0.6787,
+      "step": 179
+    },
+    {
+      "epoch": 0.0758773316471704,
+      "grad_norm": 0.17608197033405304,
+      "learning_rate": 4.998396148591223e-06,
+      "loss": 0.686,
+      "step": 180
+    },
+    {
+      "epoch": 0.07629887237854358,
+      "grad_norm": 0.17787930369377136,
+      "learning_rate": 4.9983558061273364e-06,
+      "loss": 0.6869,
+      "step": 181
+    },
+    {
+      "epoch": 0.07672041310991674,
+      "grad_norm": 0.18847839534282684,
+      "learning_rate": 4.998314962735958e-06,
+      "loss": 0.6865,
+      "step": 182
+    },
+    {
+      "epoch": 0.07714195384128991,
+      "grad_norm": 0.1885470151901245,
+      "learning_rate": 4.998273618425279e-06,
+      "loss": 0.6971,
+      "step": 183
+    },
+    {
+      "epoch": 0.07756349457266308,
+      "grad_norm": 0.1825360357761383,
+      "learning_rate": 4.9982317732035876e-06,
+      "loss": 0.6923,
+      "step": 184
+    },
+    {
+      "epoch": 0.07798503530403625,
+      "grad_norm": 0.18757036328315735,
+      "learning_rate": 4.998189427079274e-06,
+      "loss": 0.6774,
+      "step": 185
+    },
+    {
+      "epoch": 0.07840657603540942,
+      "grad_norm": 0.1757952719926834,
+      "learning_rate": 4.998146580060829e-06,
+      "loss": 0.6945,
+      "step": 186
+    },
+    {
+      "epoch": 0.07882811676678259,
+      "grad_norm": 0.1819627583026886,
+      "learning_rate": 4.998103232156844e-06,
+      "loss": 0.6961,
+      "step": 187
+    },
+    {
+      "epoch": 0.07924965749815575,
+      "grad_norm": 0.17991092801094055,
+      "learning_rate": 4.998059383376009e-06,
+      "loss": 0.6896,
+      "step": 188
+    },
+    {
+      "epoch": 0.07967119822952892,
+      "grad_norm": 0.18240217864513397,
+      "learning_rate": 4.9980150337271185e-06,
+      "loss": 0.6678,
+      "step": 189
+    },
+    {
+      "epoch": 0.08009273896090209,
+      "grad_norm": 0.17979981005191803,
+      "learning_rate": 4.9979701832190616e-06,
+      "loss": 0.655,
+      "step": 190
+    },
+    {
+      "epoch": 0.08051427969227527,
+      "grad_norm": 0.18346832692623138,
+      "learning_rate": 4.997924831860832e-06,
+      "loss": 0.6631,
+      "step": 191
+    },
+    {
+      "epoch": 0.08093582042364844,
+      "grad_norm": 0.18479032814502716,
+      "learning_rate": 4.997878979661524e-06,
+      "loss": 0.6576,
+      "step": 192
+    },
+    {
+      "epoch": 0.08135736115502161,
+      "grad_norm": 0.19124893844127655,
+      "learning_rate": 4.99783262663033e-06,
+      "loss": 0.6619,
+      "step": 193
+    },
+    {
+      "epoch": 0.08177890188639478,
+      "grad_norm": 0.18158017098903656,
+      "learning_rate": 4.997785772776544e-06,
+      "loss": 0.6817,
+      "step": 194
+    },
+    {
+      "epoch": 0.08220044261776795,
+      "grad_norm": 0.19892625510692596,
+      "learning_rate": 4.997738418109561e-06,
+      "loss": 0.687,
+      "step": 195
+    },
+    {
+      "epoch": 0.08262198334914111,
+      "grad_norm": 0.1827138215303421,
+      "learning_rate": 4.997690562638874e-06,
+      "loss": 0.6801,
+      "step": 196
+    },
+    {
+      "epoch": 0.08304352408051428,
+      "grad_norm": 0.18990923464298248,
+      "learning_rate": 4.9976422063740805e-06,
+      "loss": 0.6696,
+      "step": 197
+    },
+    {
+      "epoch": 0.08346506481188745,
+      "grad_norm": 0.18232716619968414,
+      "learning_rate": 4.997593349324874e-06,
+      "loss": 0.6382,
+      "step": 198
+    },
+    {
+      "epoch": 0.08388660554326062,
+      "grad_norm": 0.18175926804542542,
+      "learning_rate": 4.997543991501052e-06,
+      "loss": 0.6997,
+      "step": 199
+    },
+    {
+      "epoch": 0.08430814627463379,
+      "grad_norm": 0.1816951483488083,
+      "learning_rate": 4.997494132912511e-06,
+      "loss": 0.6835,
+      "step": 200
+    },
+    {
+      "epoch": 0.08472968700600696,
+      "grad_norm": 0.18778060376644135,
+      "learning_rate": 4.997443773569246e-06,
+      "loss": 0.6473,
+      "step": 201
+    },
+    {
+      "epoch": 0.08515122773738013,
+      "grad_norm": 0.17751739919185638,
+      "learning_rate": 4.997392913481355e-06,
+      "loss": 0.6771,
+      "step": 202
+    },
+    {
+      "epoch": 0.0855727684687533,
+      "grad_norm": 0.18700537085533142,
+      "learning_rate": 4.997341552659037e-06,
+      "loss": 0.6668,
+      "step": 203
+    },
+    {
+      "epoch": 0.08599430920012646,
+      "grad_norm": 0.18602705001831055,
+      "learning_rate": 4.997289691112588e-06,
+      "loss": 0.6575,
+      "step": 204
+    },
+    {
+      "epoch": 0.08641584993149963,
+      "grad_norm": 0.18992920219898224,
+      "learning_rate": 4.9972373288524075e-06,
+      "loss": 0.6763,
+      "step": 205
+    },
+    {
+      "epoch": 0.0868373906628728,
+      "grad_norm": 0.189517080783844,
+      "learning_rate": 4.997184465888994e-06,
+      "loss": 0.7008,
+      "step": 206
+    },
+    {
+      "epoch": 0.08725893139424597,
+      "grad_norm": 0.18478715419769287,
+      "learning_rate": 4.997131102232947e-06,
+      "loss": 0.679,
+      "step": 207
+    },
+    {
+      "epoch": 0.08768047212561914,
+      "grad_norm": 0.18573221564292908,
+      "learning_rate": 4.997077237894965e-06,
+      "loss": 0.6756,
+      "step": 208
+    },
+    {
+      "epoch": 0.0881020128569923,
+      "grad_norm": 0.18967847526073456,
+      "learning_rate": 4.99702287288585e-06,
+      "loss": 0.6552,
+      "step": 209
+    },
+    {
+      "epoch": 0.08852355358836547,
+      "grad_norm": 0.18442094326019287,
+      "learning_rate": 4.9969680072165e-06,
+      "loss": 0.6823,
+      "step": 210
+    },
+    {
+      "epoch": 0.08894509431973864,
+      "grad_norm": 0.1894274652004242,
+      "learning_rate": 4.996912640897917e-06,
+      "loss": 0.6745,
+      "step": 211
+    },
+    {
+      "epoch": 0.08936663505111181,
+      "grad_norm": 0.18941868841648102,
+      "learning_rate": 4.996856773941202e-06,
+      "loss": 0.6971,
+      "step": 212
+    },
+    {
+      "epoch": 0.08978817578248498,
+      "grad_norm": 0.18658168613910675,
+      "learning_rate": 4.996800406357557e-06,
+      "loss": 0.6622,
+      "step": 213
+    },
+    {
+      "epoch": 0.09020971651385815,
+      "grad_norm": 0.18819528818130493,
+      "learning_rate": 4.996743538158283e-06,
+      "loss": 0.6848,
+      "step": 214
+    },
+    {
+      "epoch": 0.09063125724523131,
+      "grad_norm": 0.1927175223827362,
+      "learning_rate": 4.9966861693547815e-06,
+      "loss": 0.6805,
+      "step": 215
+    },
+    {
+      "epoch": 0.09105279797660448,
+      "grad_norm": 0.191157728433609,
+      "learning_rate": 4.996628299958557e-06,
+      "loss": 0.6763,
+      "step": 216
+    },
+    {
+      "epoch": 0.09147433870797766,
+      "grad_norm": 0.18605469167232513,
+      "learning_rate": 4.996569929981211e-06,
+      "loss": 0.6719,
+      "step": 217
+    },
+    {
+      "epoch": 0.09189587943935083,
+      "grad_norm": 0.18317627906799316,
+      "learning_rate": 4.996511059434448e-06,
+      "loss": 0.6728,
+      "step": 218
+    },
+    {
+      "epoch": 0.092317420170724,
+      "grad_norm": 0.18342594802379608,
+      "learning_rate": 4.9964516883300705e-06,
+      "loss": 0.6835,
+      "step": 219
+    },
+    {
+      "epoch": 0.09273896090209717,
+      "grad_norm": 0.19015051424503326,
+      "learning_rate": 4.996391816679984e-06,
+      "loss": 0.6815,
+      "step": 220
+    },
+    {
+      "epoch": 0.09316050163347034,
+      "grad_norm": 0.19501470029354095,
+      "learning_rate": 4.996331444496192e-06,
+      "loss": 0.685,
+      "step": 221
+    },
+    {
+      "epoch": 0.0935820423648435,
+      "grad_norm": 0.19016507267951965,
+      "learning_rate": 4.996270571790799e-06,
+      "loss": 0.6654,
+      "step": 222
+    },
+    {
+      "epoch": 0.09400358309621668,
+      "grad_norm": 0.1969563215970993,
+      "learning_rate": 4.996209198576011e-06,
+      "loss": 0.6625,
+      "step": 223
+    },
+    {
+      "epoch": 0.09442512382758984,
+      "grad_norm": 0.19229756295681,
+      "learning_rate": 4.9961473248641325e-06,
+      "loss": 0.6867,
+      "step": 224
+    },
+    {
+      "epoch": 0.09484666455896301,
+      "grad_norm": 0.18767569959163666,
+      "learning_rate": 4.9960849506675705e-06,
+      "loss": 0.6595,
+      "step": 225
+    },
+    {
+      "epoch": 0.09526820529033618,
+      "grad_norm": 0.19153764843940735,
+      "learning_rate": 4.9960220759988295e-06,
+      "loss": 0.6709,
+      "step": 226
+    },
+    {
+      "epoch": 0.09568974602170935,
+      "grad_norm": 0.20141515135765076,
+      "learning_rate": 4.995958700870519e-06,
+      "loss": 0.6791,
+      "step": 227
+    },
+    {
+      "epoch": 0.09611128675308252,
+      "grad_norm": 0.1969662606716156,
+      "learning_rate": 4.9958948252953435e-06,
+      "loss": 0.6792,
+      "step": 228
+    },
+    {
+      "epoch": 0.09653282748445569,
+      "grad_norm": 0.19556061923503876,
+      "learning_rate": 4.99583044928611e-06,
+      "loss": 0.6436,
+      "step": 229
+    },
+    {
+      "epoch": 0.09695436821582885,
+      "grad_norm": 0.20046646893024445,
+      "learning_rate": 4.9957655728557275e-06,
+      "loss": 0.6762,
+      "step": 230
+    },
+    {
+      "epoch": 0.09737590894720202,
+      "grad_norm": 0.18485157191753387,
+      "learning_rate": 4.995700196017204e-06,
+      "loss": 0.6658,
+      "step": 231
+    },
+    {
+      "epoch": 0.09779744967857519,
+      "grad_norm": 0.1980086714029312,
+      "learning_rate": 4.9956343187836455e-06,
+      "loss": 0.6516,
+      "step": 232
+    },
+    {
+      "epoch": 0.09821899040994836,
+      "grad_norm": 0.18426518142223358,
+      "learning_rate": 4.995567941168264e-06,
+      "loss": 0.6844,
+      "step": 233
+    },
+    {
+      "epoch": 0.09864053114132153,
+      "grad_norm": 0.1942421793937683,
+      "learning_rate": 4.995501063184365e-06,
+      "loss": 0.6946,
+      "step": 234
+    },
+    {
+      "epoch": 0.0990620718726947,
+      "grad_norm": 0.1975025087594986,
+      "learning_rate": 4.99543368484536e-06,
+      "loss": 0.6498,
+      "step": 235
+    },
+    {
+      "epoch": 0.09948361260406786,
+      "grad_norm": 0.18677708506584167,
+      "learning_rate": 4.995365806164758e-06,
+      "loss": 0.662,
+      "step": 236
+    },
+    {
+      "epoch": 0.09990515333544103,
+      "grad_norm": 0.1851079761981964,
+      "learning_rate": 4.995297427156169e-06,
+      "loss": 0.6773,
+      "step": 237
+    },
+    {
+      "epoch": 0.1003266940668142,
+      "grad_norm": 0.18994666635990143,
+      "learning_rate": 4.995228547833301e-06,
+      "loss": 0.6852,
+      "step": 238
+    },
+    {
+      "epoch": 0.10074823479818737,
+      "grad_norm": 0.19767063856124878,
+      "learning_rate": 4.995159168209968e-06,
+      "loss": 0.6863,
+      "step": 239
+    },
+    {
+      "epoch": 0.10116977552956054,
+      "grad_norm": 0.2009822577238083,
+      "learning_rate": 4.995089288300079e-06,
+      "loss": 0.695,
+      "step": 240
+    },
+    {
+      "epoch": 0.1015913162609337,
+      "grad_norm": 0.19976560771465302,
+      "learning_rate": 4.995018908117644e-06,
+      "loss": 0.6773,
+      "step": 241
+    },
+    {
+      "epoch": 0.10201285699230689,
+      "grad_norm": 0.19675956666469574,
+      "learning_rate": 4.994948027676778e-06,
+      "loss": 0.6595,
+      "step": 242
+    },
+    {
+      "epoch": 0.10243439772368006,
+      "grad_norm": 0.19577495753765106,
+      "learning_rate": 4.99487664699169e-06,
+      "loss": 0.6716,
+      "step": 243
+    },
+    {
+      "epoch": 0.10285593845505323,
+      "grad_norm": 0.1974613070487976,
+      "learning_rate": 4.994804766076691e-06,
+      "loss": 0.6548,
+      "step": 244
+    },
+    {
+      "epoch": 0.1032774791864264,
+      "grad_norm": 0.19054162502288818,
+      "learning_rate": 4.994732384946196e-06,
+      "loss": 0.6846,
+      "step": 245
+    },
+    {
+      "epoch": 0.10369901991779956,
+      "grad_norm": 0.1930086314678192,
+      "learning_rate": 4.994659503614716e-06,
+      "loss": 0.6636,
+      "step": 246
+    },
+    {
+      "epoch": 0.10412056064917273,
+      "grad_norm": 0.19504211843013763,
+      "learning_rate": 4.994586122096865e-06,
+      "loss": 0.6964,
+      "step": 247
+    },
+    {
+      "epoch": 0.1045421013805459,
+      "grad_norm": 0.19730274379253387,
+      "learning_rate": 4.994512240407354e-06,
+      "loss": 0.6822,
+      "step": 248
+    },
+    {
+      "epoch": 0.10496364211191907,
+      "grad_norm": 0.1991763710975647,
+      "learning_rate": 4.994437858561e-06,
+      "loss": 0.6736,
+      "step": 249
+    },
+    {
+      "epoch": 0.10538518284329224,
+      "grad_norm": 0.20081613957881927,
+      "learning_rate": 4.994362976572713e-06,
+      "loss": 0.6878,
+      "step": 250
+    },
+    {
+      "epoch": 0.1058067235746654,
+      "grad_norm": 0.21170829236507416,
+      "learning_rate": 4.994287594457511e-06,
+      "loss": 0.6787,
+      "step": 251
+    },
+    {
+      "epoch": 0.10622826430603857,
+      "grad_norm": 0.19960327446460724,
+      "learning_rate": 4.9942117122305045e-06,
+      "loss": 0.6702,
+      "step": 252
+    },
+    {
+      "epoch": 0.10664980503741174,
+      "grad_norm": 0.19778785109519958,
+      "learning_rate": 4.9941353299069105e-06,
+      "loss": 0.6963,
+      "step": 253
+    },
+    {
+      "epoch": 0.10707134576878491,
+      "grad_norm": 0.1955125480890274,
+      "learning_rate": 4.994058447502043e-06,
+      "loss": 0.6776,
+      "step": 254
+    },
+    {
+      "epoch": 0.10749288650015808,
+      "grad_norm": 0.19576016068458557,
+      "learning_rate": 4.9939810650313175e-06,
+      "loss": 0.6761,
+      "step": 255
+    },
+    {
+      "epoch": 0.10791442723153125,
+      "grad_norm": 0.19699929654598236,
+      "learning_rate": 4.993903182510249e-06,
+      "loss": 0.6787,
+      "step": 256
+    },
+    {
+      "epoch": 0.10833596796290441,
+      "grad_norm": 0.18686431646347046,
+      "learning_rate": 4.993824799954454e-06,
+      "loss": 0.6692,
+      "step": 257
+    },
+    {
+      "epoch": 0.10875750869427758,
+      "grad_norm": 0.19229714572429657,
+      "learning_rate": 4.9937459173796475e-06,
+      "loss": 0.664,
+      "step": 258
+    },
+    {
+      "epoch": 0.10917904942565075,
+      "grad_norm": 0.19343610107898712,
+      "learning_rate": 4.9936665348016465e-06,
+      "loss": 0.6621,
+      "step": 259
+    },
+    {
+      "epoch": 0.10960059015702392,
+      "grad_norm": 0.19266656041145325,
+      "learning_rate": 4.9935866522363665e-06,
+      "loss": 0.7024,
+      "step": 260
+    },
+    {
+      "epoch": 0.11002213088839709,
+      "grad_norm": 0.19614985585212708,
+      "learning_rate": 4.993506269699824e-06,
+      "loss": 0.6448,
+      "step": 261
+    },
+    {
+      "epoch": 0.11044367161977026,
+      "grad_norm": 0.19521160423755646,
+      "learning_rate": 4.993425387208137e-06,
+      "loss": 0.698,
+      "step": 262
+    },
+    {
+      "epoch": 0.11086521235114342,
+      "grad_norm": 0.21780386567115784,
+      "learning_rate": 4.993344004777523e-06,
+      "loss": 0.6868,
+      "step": 263
+    },
+    {
+      "epoch": 0.11128675308251659,
+      "grad_norm": 0.18932631611824036,
+      "learning_rate": 4.993262122424298e-06,
+      "loss": 0.6513,
+      "step": 264
+    },
+    {
+      "epoch": 0.11170829381388976,
+      "grad_norm": 0.20421689748764038,
+      "learning_rate": 4.9931797401648806e-06,
+      "loss": 0.6753,
+      "step": 265
+    },
+    {
+      "epoch": 0.11212983454526293,
+      "grad_norm": 0.20207132399082184,
+      "learning_rate": 4.993096858015788e-06,
+      "loss": 0.6821,
+      "step": 266
+    },
+    {
+      "epoch": 0.1125513752766361,
+      "grad_norm": 0.20157171785831451,
+      "learning_rate": 4.993013475993638e-06,
+      "loss": 0.6889,
+      "step": 267
+    },
+    {
+      "epoch": 0.11297291600800928,
+      "grad_norm": 0.20024839043617249,
+      "learning_rate": 4.992929594115151e-06,
+      "loss": 0.6243,
+      "step": 268
+    },
+    {
+      "epoch": 0.11339445673938245,
+      "grad_norm": 0.21691858768463135,
+      "learning_rate": 4.992845212397144e-06,
+      "loss": 0.6851,
+      "step": 269
+    },
+    {
+      "epoch": 0.11381599747075562,
+      "grad_norm": 0.19579987227916718,
+      "learning_rate": 4.9927603308565356e-06,
+      "loss": 0.6868,
+      "step": 270
+    },
+    {
+      "epoch": 0.11423753820212879,
+      "grad_norm": 0.19934384524822235,
+      "learning_rate": 4.992674949510345e-06,
+      "loss": 0.6815,
+      "step": 271
+    },
+    {
+      "epoch": 0.11465907893350195,
+      "grad_norm": 0.21055586636066437,
+      "learning_rate": 4.992589068375691e-06,
+      "loss": 0.6504,
+      "step": 272
+    },
+    {
+      "epoch": 0.11508061966487512,
+      "grad_norm": 0.20408973097801208,
+      "learning_rate": 4.992502687469795e-06,
+      "loss": 0.6768,
+      "step": 273
+    },
+    {
+      "epoch": 0.11550216039624829,
+      "grad_norm": 0.19852286577224731,
+      "learning_rate": 4.992415806809975e-06,
+      "loss": 0.6867,
+      "step": 274
+    },
+    {
+      "epoch": 0.11592370112762146,
+      "grad_norm": 0.19237147271633148,
+      "learning_rate": 4.99232842641365e-06,
+      "loss": 0.6666,
+      "step": 275
+    },
+    {
+      "epoch": 0.11634524185899463,
+      "grad_norm": 0.2004595398902893,
+      "learning_rate": 4.992240546298342e-06,
+      "loss": 0.6778,
+      "step": 276
+    },
+    {
+      "epoch": 0.1167667825903678,
+      "grad_norm": 0.20304861664772034,
+      "learning_rate": 4.992152166481669e-06,
+      "loss": 0.674,
+      "step": 277
+    },
+    {
+      "epoch": 0.11718832332174096,
+      "grad_norm": 0.19285213947296143,
+      "learning_rate": 4.992063286981354e-06,
+      "loss": 0.65,
+      "step": 278
+    },
+    {
+      "epoch": 0.11760986405311413,
+      "grad_norm": 0.20184293389320374,
+      "learning_rate": 4.991973907815215e-06,
+      "loss": 0.685,
+      "step": 279
+    },
+    {
+      "epoch": 0.1180314047844873,
+      "grad_norm": 0.19268806278705597,
+      "learning_rate": 4.991884029001175e-06,
+      "loss": 0.6626,
+      "step": 280
+    },
+    {
+      "epoch": 0.11845294551586047,
+      "grad_norm": 0.20452861487865448,
+      "learning_rate": 4.991793650557254e-06,
+      "loss": 0.6763,
+      "step": 281
+    },
+    {
+      "epoch": 0.11887448624723364,
+      "grad_norm": 0.20973370969295502,
+      "learning_rate": 4.991702772501572e-06,
+      "loss": 0.6571,
+      "step": 282
+    },
+    {
+      "epoch": 0.1192960269786068,
+      "grad_norm": 0.20668712258338928,
+      "learning_rate": 4.9916113948523526e-06,
+      "loss": 0.6783,
+      "step": 283
+    },
+    {
+      "epoch": 0.11971756770997997,
+      "grad_norm": 0.1931544989347458,
+      "learning_rate": 4.991519517627916e-06,
+      "loss": 0.6747,
+      "step": 284
+    },
+    {
+      "epoch": 0.12013910844135314,
+      "grad_norm": 0.2019011378288269,
+      "learning_rate": 4.991427140846683e-06,
+      "loss": 0.6406,
+      "step": 285
+    },
+    {
+      "epoch": 0.12056064917272631,
+      "grad_norm": 0.20072971284389496,
+      "learning_rate": 4.991334264527177e-06,
+      "loss": 0.6724,
+      "step": 286
+    },
+    {
+      "epoch": 0.12098218990409948,
+      "grad_norm": 0.2066023200750351,
+      "learning_rate": 4.991240888688019e-06,
+      "loss": 0.6412,
+      "step": 287
+    },
+    {
+      "epoch": 0.12140373063547265,
+      "grad_norm": 0.20346936583518982,
+      "learning_rate": 4.991147013347932e-06,
+      "loss": 0.6452,
+      "step": 288
+    },
+    {
+      "epoch": 0.12182527136684582,
+      "grad_norm": 0.19868072867393494,
+      "learning_rate": 4.991052638525737e-06,
+      "loss": 0.6579,
+      "step": 289
+    },
+    {
+      "epoch": 0.12224681209821898,
+      "grad_norm": 0.202948659658432,
+      "learning_rate": 4.990957764240358e-06,
+      "loss": 0.6521,
+      "step": 290
+    },
+    {
+      "epoch": 0.12266835282959215,
+      "grad_norm": 0.1964743435382843,
+      "learning_rate": 4.990862390510817e-06,
+      "loss": 0.6219,
+      "step": 291
+    },
+    {
+      "epoch": 0.12308989356096532,
+      "grad_norm": 0.20336145162582397,
+      "learning_rate": 4.990766517356236e-06,
+      "loss": 0.6606,
+      "step": 292
+    },
+    {
+      "epoch": 0.1235114342923385,
+      "grad_norm": 0.20082779228687286,
+      "learning_rate": 4.990670144795837e-06,
+      "loss": 0.6823,
+      "step": 293
+    },
+    {
+      "epoch": 0.12393297502371167,
+      "grad_norm": 0.20099136233329773,
+      "learning_rate": 4.990573272848946e-06,
+      "loss": 0.6808,
+      "step": 294
+    },
+    {
+      "epoch": 0.12435451575508484,
+      "grad_norm": 0.19388745725154877,
+      "learning_rate": 4.990475901534983e-06,
+      "loss": 0.675,
+      "step": 295
+    },
+    {
+      "epoch": 0.12477605648645801,
+      "grad_norm": 0.2019960582256317,
+      "learning_rate": 4.990378030873474e-06,
+      "loss": 0.6561,
+      "step": 296
+    },
+    {
+      "epoch": 0.12519759721783116,
+      "grad_norm": 0.21345113217830658,
+      "learning_rate": 4.99027966088404e-06,
+      "loss": 0.6692,
+      "step": 297
+    },
+    {
+      "epoch": 0.12561913794920435,
+      "grad_norm": 0.2115856260061264,
+      "learning_rate": 4.9901807915864055e-06,
+      "loss": 0.7,
+      "step": 298
+    },
+    {
+      "epoch": 0.1260406786805775,
+      "grad_norm": 0.20527495443820953,
+      "learning_rate": 4.990081423000394e-06,
+      "loss": 0.6672,
+      "step": 299
+    },
+    {
+      "epoch": 0.12646221941195068,
+      "grad_norm": 0.20394611358642578,
+      "learning_rate": 4.989981555145928e-06,
+      "loss": 0.6641,
+      "step": 300
+    },
+    {
+      "epoch": 0.12688376014332384,
+      "grad_norm": 0.20698480308055878,
+      "learning_rate": 4.989881188043035e-06,
+      "loss": 0.6649,
+      "step": 301
+    },
+    {
+      "epoch": 0.12730530087469702,
+      "grad_norm": 0.20958511531352997,
+      "learning_rate": 4.989780321711834e-06,
+      "loss": 0.6548,
+      "step": 302
+    },
+    {
+      "epoch": 0.12772684160607017,
+      "grad_norm": 0.2013298124074936,
+      "learning_rate": 4.989678956172553e-06,
+      "loss": 0.6671,
+      "step": 303
+    },
+    {
+      "epoch": 0.12814838233744336,
+      "grad_norm": 0.20705978572368622,
+      "learning_rate": 4.989577091445512e-06,
+      "loss": 0.6418,
+      "step": 304
+    },
+    {
+      "epoch": 0.1285699230688165,
+      "grad_norm": 0.2003588080406189,
+      "learning_rate": 4.9894747275511395e-06,
+      "loss": 0.6668,
+      "step": 305
+    },
+    {
+      "epoch": 0.1289914638001897,
+      "grad_norm": 0.20356950163841248,
+      "learning_rate": 4.9893718645099575e-06,
+      "loss": 0.6639,
+      "step": 306
+    },
+    {
+      "epoch": 0.12941300453156288,
+      "grad_norm": 0.20462188124656677,
+      "learning_rate": 4.98926850234259e-06,
+      "loss": 0.6681,
+      "step": 307
+    },
+    {
+      "epoch": 0.12983454526293603,
+      "grad_norm": 0.2045106738805771,
+      "learning_rate": 4.989164641069763e-06,
+      "loss": 0.6948,
+      "step": 308
+    },
+    {
+      "epoch": 0.1302560859943092,
+      "grad_norm": 0.2094605416059494,
+      "learning_rate": 4.9890602807123e-06,
+      "loss": 0.6622,
+      "step": 309
+    },
+    {
+      "epoch": 0.13067762672568237,
+      "grad_norm": 0.2059636414051056,
+      "learning_rate": 4.988955421291125e-06,
+      "loss": 0.6773,
+      "step": 310
+    },
+    {
+      "epoch": 0.13109916745705555,
+      "grad_norm": 0.1962340921163559,
+      "learning_rate": 4.988850062827263e-06,
+      "loss": 0.6415,
+      "step": 311
+    },
+    {
+      "epoch": 0.1315207081884287,
+      "grad_norm": 0.20182925462722778,
+      "learning_rate": 4.98874420534184e-06,
+      "loss": 0.659,
+      "step": 312
+    },
+    {
+      "epoch": 0.13194224891980189,
+      "grad_norm": 0.20462483167648315,
+      "learning_rate": 4.9886378488560785e-06,
+      "loss": 0.6682,
+      "step": 313
+    },
+    {
+      "epoch": 0.13236378965117504,
+      "grad_norm": 0.20841017365455627,
+      "learning_rate": 4.988530993391305e-06,
+      "loss": 0.6634,
+      "step": 314
+    },
+    {
+      "epoch": 0.13278533038254822,
+      "grad_norm": 0.20585286617279053,
+      "learning_rate": 4.988423638968943e-06,
+      "loss": 0.6696,
+      "step": 315
+    },
+    {
+      "epoch": 0.13320687111392138,
+      "grad_norm": 0.20328669250011444,
+      "learning_rate": 4.9883157856105195e-06,
+      "loss": 0.6781,
+      "step": 316
+    },
+    {
+      "epoch": 0.13362841184529456,
+      "grad_norm": 0.20803019404411316,
+      "learning_rate": 4.988207433337657e-06,
+      "loss": 0.656,
+      "step": 317
+    },
+    {
+      "epoch": 0.1340499525766677,
+      "grad_norm": 0.201886847615242,
+      "learning_rate": 4.988098582172081e-06,
+      "loss": 0.6313,
+      "step": 318
+    },
+    {
+      "epoch": 0.1344714933080409,
+      "grad_norm": 0.20804308354854584,
+      "learning_rate": 4.987989232135618e-06,
+      "loss": 0.6797,
+      "step": 319
+    },
+    {
+      "epoch": 0.13489303403941405,
+      "grad_norm": 0.20710523426532745,
+      "learning_rate": 4.987879383250191e-06,
+      "loss": 0.6789,
+      "step": 320
+    },
+    {
+      "epoch": 0.13531457477078723,
+      "grad_norm": 0.20678341388702393,
+      "learning_rate": 4.9877690355378265e-06,
+      "loss": 0.6597,
+      "step": 321
+    },
+    {
+      "epoch": 0.1357361155021604,
+      "grad_norm": 0.2014772742986679,
+      "learning_rate": 4.9876581890206485e-06,
+      "loss": 0.6496,
+      "step": 322
+    },
+    {
+      "epoch": 0.13615765623353357,
+      "grad_norm": 0.20414437353610992,
+      "learning_rate": 4.9875468437208825e-06,
+      "loss": 0.6291,
+      "step": 323
+    },
+    {
+      "epoch": 0.13657919696490672,
+      "grad_norm": 0.20725557208061218,
+      "learning_rate": 4.987434999660854e-06,
+      "loss": 0.6868,
+      "step": 324
+    },
+    {
+      "epoch": 0.1370007376962799,
+      "grad_norm": 0.20924833416938782,
+      "learning_rate": 4.987322656862987e-06,
+      "loss": 0.6747,
+      "step": 325
+    },
+    {
+      "epoch": 0.13742227842765306,
+      "grad_norm": 0.20984621345996857,
+      "learning_rate": 4.987209815349808e-06,
+      "loss": 0.6834,
+      "step": 326
+    },
+    {
+      "epoch": 0.13784381915902624,
+      "grad_norm": 0.21052715182304382,
+      "learning_rate": 4.98709647514394e-06,
+      "loss": 0.6422,
+      "step": 327
+    },
+    {
+      "epoch": 0.1382653598903994,
+      "grad_norm": 0.21226079761981964,
+      "learning_rate": 4.9869826362681105e-06,
+      "loss": 0.6533,
+      "step": 328
+    },
+    {
+      "epoch": 0.13868690062177258,
+      "grad_norm": 0.2016143947839737,
+      "learning_rate": 4.986868298745142e-06,
+      "loss": 0.6442,
+      "step": 329
+    },
+    {
+      "epoch": 0.13910844135314573,
+      "grad_norm": 0.19947828352451324,
+      "learning_rate": 4.986753462597962e-06,
+      "loss": 0.6563,
+      "step": 330
+    },
+    {
+      "epoch": 0.13952998208451892,
+      "grad_norm": 0.21439361572265625,
+      "learning_rate": 4.986638127849593e-06,
+      "loss": 0.6433,
+      "step": 331
+    },
+    {
+      "epoch": 0.1399515228158921,
+      "grad_norm": 0.20104160904884338,
+      "learning_rate": 4.9865222945231625e-06,
+      "loss": 0.637,
+      "step": 332
+    },
+    {
+      "epoch": 0.14037306354726525,
+      "grad_norm": 0.20007939636707306,
+      "learning_rate": 4.986405962641893e-06,
+      "loss": 0.6474,
+      "step": 333
+    },
+    {
+      "epoch": 0.14079460427863844,
+      "grad_norm": 0.2133202701807022,
+      "learning_rate": 4.9862891322291105e-06,
+      "loss": 0.6733,
+      "step": 334
+    },
+    {
+      "epoch": 0.1412161450100116,
+      "grad_norm": 0.2044401466846466,
+      "learning_rate": 4.9861718033082405e-06,
+      "loss": 0.6726,
+      "step": 335
+    },
+    {
+      "epoch": 0.14163768574138477,
+      "grad_norm": 0.20695999264717102,
+      "learning_rate": 4.986053975902807e-06,
+      "loss": 0.6794,
+      "step": 336
+    },
+    {
+      "epoch": 0.14205922647275793,
+      "grad_norm": 0.2152855545282364,
+      "learning_rate": 4.985935650036435e-06,
+      "loss": 0.6639,
+      "step": 337
+    },
+    {
+      "epoch": 0.1424807672041311,
+      "grad_norm": 0.21518713235855103,
+      "learning_rate": 4.985816825732848e-06,
+      "loss": 0.6671,
+      "step": 338
+    },
+    {
+      "epoch": 0.14290230793550426,
+      "grad_norm": 0.21271373331546783,
+      "learning_rate": 4.985697503015872e-06,
+      "loss": 0.6698,
+      "step": 339
+    },
+    {
+      "epoch": 0.14332384866687745,
+      "grad_norm": 0.2174816131591797,
+      "learning_rate": 4.985577681909431e-06,
+      "loss": 0.6376,
+      "step": 340
+    },
+    {
+      "epoch": 0.1437453893982506,
+      "grad_norm": 0.20762839913368225,
+      "learning_rate": 4.98545736243755e-06,
+      "loss": 0.6492,
+      "step": 341
+    },
+    {
+      "epoch": 0.14416693012962378,
+      "grad_norm": 0.21048258244991302,
+      "learning_rate": 4.985336544624353e-06,
+      "loss": 0.6494,
+      "step": 342
+    },
+    {
+      "epoch": 0.14458847086099694,
+      "grad_norm": 0.208738774061203,
+      "learning_rate": 4.985215228494065e-06,
+      "loss": 0.6738,
+      "step": 343
+    },
+    {
+      "epoch": 0.14501001159237012,
+      "grad_norm": 0.21118545532226562,
+      "learning_rate": 4.9850934140710084e-06,
+      "loss": 0.6526,
+      "step": 344
+    },
+    {
+      "epoch": 0.14543155232374327,
+      "grad_norm": 0.2104908674955368,
+      "learning_rate": 4.984971101379609e-06,
+      "loss": 0.6596,
+      "step": 345
+    },
+    {
+      "epoch": 0.14585309305511646,
+      "grad_norm": 0.21006746590137482,
+      "learning_rate": 4.9848482904443905e-06,
+      "loss": 0.6528,
+      "step": 346
+    },
+    {
+      "epoch": 0.1462746337864896,
+      "grad_norm": 0.21185743808746338,
+      "learning_rate": 4.984724981289976e-06,
+      "loss": 0.6642,
+      "step": 347
+    },
+    {
+      "epoch": 0.1466961745178628,
+      "grad_norm": 0.21709086000919342,
+      "learning_rate": 4.9846011739410905e-06,
+      "loss": 0.6833,
+      "step": 348
+    },
+    {
+      "epoch": 0.14711771524923595,
+      "grad_norm": 0.20782820880413055,
+      "learning_rate": 4.9844768684225575e-06,
+      "loss": 0.6474,
+      "step": 349
+    },
+    {
+      "epoch": 0.14753925598060913,
+      "grad_norm": 0.20482183992862701,
+      "learning_rate": 4.9843520647593005e-06,
+      "loss": 0.6421,
+      "step": 350
+    },
+    {
+      "epoch": 0.14796079671198228,
+      "grad_norm": 0.2143651843070984,
+      "learning_rate": 4.984226762976342e-06,
+      "loss": 0.6596,
+      "step": 351
+    },
+    {
+      "epoch": 0.14838233744335547,
+      "grad_norm": 0.22368572652339935,
+      "learning_rate": 4.984100963098807e-06,
+      "loss": 0.651,
+      "step": 352
+    },
+    {
+      "epoch": 0.14880387817472862,
+      "grad_norm": 0.20261022448539734,
+      "learning_rate": 4.983974665151918e-06,
+      "loss": 0.6363,
+      "step": 353
+    },
+    {
+      "epoch": 0.1492254189061018,
+      "grad_norm": 0.21051320433616638,
+      "learning_rate": 4.983847869160997e-06,
+      "loss": 0.6535,
+      "step": 354
+    },
+    {
+      "epoch": 0.14964695963747496,
+      "grad_norm": 0.20940569043159485,
+      "learning_rate": 4.98372057515147e-06,
+      "loss": 0.6523,
+      "step": 355
+    },
+    {
+      "epoch": 0.15006850036884814,
+      "grad_norm": 0.2194211781024933,
+      "learning_rate": 4.9835927831488565e-06,
+      "loss": 0.6533,
+      "step": 356
+    },
+    {
+      "epoch": 0.15049004110022132,
+      "grad_norm": 0.22318407893180847,
+      "learning_rate": 4.98346449317878e-06,
+      "loss": 0.6957,
+      "step": 357
+    },
+    {
+      "epoch": 0.15091158183159448,
+      "grad_norm": 0.2162528932094574,
+      "learning_rate": 4.983335705266964e-06,
+      "loss": 0.6691,
+      "step": 358
+    },
+    {
+      "epoch": 0.15133312256296766,
+      "grad_norm": 0.20046716928482056,
+      "learning_rate": 4.983206419439231e-06,
+      "loss": 0.6504,
+      "step": 359
+    },
+    {
+      "epoch": 0.1517546632943408,
+      "grad_norm": 0.23220473527908325,
+      "learning_rate": 4.9830766357215025e-06,
+      "loss": 0.6377,
+      "step": 360
+    },
+    {
+      "epoch": 0.152176204025714,
+      "grad_norm": 0.21880929172039032,
+      "learning_rate": 4.982946354139799e-06,
+      "loss": 0.6386,
+      "step": 361
+    },
+    {
+      "epoch": 0.15259774475708715,
+      "grad_norm": 0.21767985820770264,
+      "learning_rate": 4.982815574720246e-06,
+      "loss": 0.6569,
+      "step": 362
+    },
+    {
+      "epoch": 0.15301928548846033,
+      "grad_norm": 0.22018642723560333,
+      "learning_rate": 4.9826842974890614e-06,
+      "loss": 0.6602,
+      "step": 363
+    },
+    {
+      "epoch": 0.1534408262198335,
+      "grad_norm": 0.22948899865150452,
+      "learning_rate": 4.982552522472569e-06,
+      "loss": 0.6506,
+      "step": 364
+    },
+    {
+      "epoch": 0.15386236695120667,
+      "grad_norm": 0.20289656519889832,
+      "learning_rate": 4.9824202496971894e-06,
+      "loss": 0.6558,
+      "step": 365
+    },
+    {
+      "epoch": 0.15428390768257982,
+      "grad_norm": 0.22870121896266937,
+      "learning_rate": 4.982287479189444e-06,
+      "loss": 0.6854,
+      "step": 366
+    },
+    {
+      "epoch": 0.154705448413953,
+      "grad_norm": 0.23659977316856384,
+      "learning_rate": 4.982154210975952e-06,
+      "loss": 0.6698,
+      "step": 367
+    },
+    {
+      "epoch": 0.15512698914532616,
+      "grad_norm": 0.22445474565029144,
+      "learning_rate": 4.982020445083436e-06,
+      "loss": 0.6731,
+      "step": 368
+    },
+    {
+      "epoch": 0.15554852987669934,
+      "grad_norm": 0.21690113842487335,
+      "learning_rate": 4.981886181538717e-06,
+      "loss": 0.6479,
+      "step": 369
+    },
+    {
+      "epoch": 0.1559700706080725,
+      "grad_norm": 0.21330410242080688,
+      "learning_rate": 4.9817514203687124e-06,
+      "loss": 0.6647,
+      "step": 370
+    },
+    {
+      "epoch": 0.15639161133944568,
+      "grad_norm": 0.21377027034759521,
+      "learning_rate": 4.981616161600444e-06,
+      "loss": 0.657,
+      "step": 371
+    },
+    {
+      "epoch": 0.15681315207081883,
+      "grad_norm": 0.21915587782859802,
+      "learning_rate": 4.981480405261032e-06,
+      "loss": 0.665,
+      "step": 372
+    },
+    {
+      "epoch": 0.15723469280219202,
+      "grad_norm": 0.22189389169216156,
+      "learning_rate": 4.981344151377695e-06,
+      "loss": 0.6587,
+      "step": 373
+    },
+    {
+      "epoch": 0.15765623353356517,
+      "grad_norm": 0.21328085660934448,
+      "learning_rate": 4.981207399977752e-06,
+      "loss": 0.6425,
+      "step": 374
+    },
+    {
+      "epoch": 0.15807777426493835,
+      "grad_norm": 0.22045306861400604,
+      "learning_rate": 4.981070151088624e-06,
+      "loss": 0.6506,
+      "step": 375
+    },
+    {
+      "epoch": 0.1584993149963115,
+      "grad_norm": 0.2125290185213089,
+      "learning_rate": 4.980932404737827e-06,
+      "loss": 0.6815,
+      "step": 376
+    },
+    {
+      "epoch": 0.1589208557276847,
+      "grad_norm": 0.21428261697292328,
+      "learning_rate": 4.980794160952982e-06,
+      "loss": 0.65,
+      "step": 377
+    },
+    {
+      "epoch": 0.15934239645905784,
+      "grad_norm": 0.22129037976264954,
+      "learning_rate": 4.980655419761806e-06,
+      "loss": 0.6659,
+      "step": 378
+    },
+    {
+      "epoch": 0.15976393719043103,
+      "grad_norm": 0.22464123368263245,
+      "learning_rate": 4.980516181192117e-06,
+      "loss": 0.6667,
+      "step": 379
+    },
+    {
+      "epoch": 0.16018547792180418,
+      "grad_norm": 0.21550893783569336,
+      "learning_rate": 4.980376445271834e-06,
+      "loss": 0.6552,
+      "step": 380
+    },
+    {
+      "epoch": 0.16060701865317736,
+      "grad_norm": 0.21697524189949036,
+      "learning_rate": 4.980236212028973e-06,
+      "loss": 0.6615,
+      "step": 381
+    },
+    {
+      "epoch": 0.16102855938455055,
+      "grad_norm": 0.21036921441555023,
+      "learning_rate": 4.980095481491651e-06,
+      "loss": 0.6402,
+      "step": 382
+    },
+    {
+      "epoch": 0.1614501001159237,
+      "grad_norm": 0.21921782195568085,
+      "learning_rate": 4.979954253688086e-06,
+      "loss": 0.6404,
+      "step": 383
+    },
+    {
+      "epoch": 0.16187164084729688,
+      "grad_norm": 0.2197704017162323,
+      "learning_rate": 4.9798125286465935e-06,
+      "loss": 0.66,
+      "step": 384
+    },
+    {
+      "epoch": 0.16229318157867004,
+      "grad_norm": 0.21849073469638824,
+      "learning_rate": 4.979670306395591e-06,
+      "loss": 0.6416,
+      "step": 385
+    },
+    {
+      "epoch": 0.16271472231004322,
+      "grad_norm": 0.21315787732601166,
+      "learning_rate": 4.979527586963594e-06,
+      "loss": 0.653,
+      "step": 386
+    },
+    {
+      "epoch": 0.16313626304141637,
+      "grad_norm": 0.2077677696943283,
+      "learning_rate": 4.9793843703792175e-06,
+      "loss": 0.6138,
+      "step": 387
+    },
+    {
+      "epoch": 0.16355780377278956,
+      "grad_norm": 0.21379877626895905,
+      "learning_rate": 4.979240656671177e-06,
+      "loss": 0.6698,
+      "step": 388
+    },
+    {
+      "epoch": 0.1639793445041627,
+      "grad_norm": 0.214260071516037,
+      "learning_rate": 4.979096445868288e-06,
+      "loss": 0.6734,
+      "step": 389
+    },
+    {
+      "epoch": 0.1644008852355359,
+      "grad_norm": 0.219817116856575,
+      "learning_rate": 4.978951737999466e-06,
+      "loss": 0.6582,
+      "step": 390
+    },
+    {
+      "epoch": 0.16482242596690905,
+      "grad_norm": 0.2153392732143402,
+      "learning_rate": 4.978806533093723e-06,
+      "loss": 0.6577,
+      "step": 391
+    },
+    {
+      "epoch": 0.16524396669828223,
+      "grad_norm": 0.2169211059808731,
+      "learning_rate": 4.978660831180175e-06,
+      "loss": 0.6246,
+      "step": 392
+    },
+    {
+      "epoch": 0.16566550742965538,
+      "grad_norm": 0.22149506211280823,
+      "learning_rate": 4.978514632288035e-06,
+      "loss": 0.6434,
+      "step": 393
+    },
+    {
+      "epoch": 0.16608704816102857,
+      "grad_norm": 0.21007981896400452,
+      "learning_rate": 4.978367936446616e-06,
+      "loss": 0.6483,
+      "step": 394
+    },
+    {
+      "epoch": 0.16650858889240172,
+      "grad_norm": 0.22675804793834686,
+      "learning_rate": 4.978220743685331e-06,
+      "loss": 0.6221,
+      "step": 395
+    },
+    {
+      "epoch": 0.1669301296237749,
+      "grad_norm": 0.2154684215784073,
+      "learning_rate": 4.978073054033694e-06,
+      "loss": 0.6469,
+      "step": 396
+    },
+    {
+      "epoch": 0.16735167035514806,
+      "grad_norm": 0.21647655963897705,
+      "learning_rate": 4.977924867521316e-06,
+      "loss": 0.6452,
+      "step": 397
+    },
+    {
+      "epoch": 0.16777321108652124,
+      "grad_norm": 0.2090582400560379,
+      "learning_rate": 4.9777761841779074e-06,
+      "loss": 0.6469,
+      "step": 398
+    },
+    {
+      "epoch": 0.1681947518178944,
+      "grad_norm": 0.22375378012657166,
+      "learning_rate": 4.977627004033281e-06,
+      "loss": 0.6522,
+      "step": 399
+    },
+    {
+      "epoch": 0.16861629254926758,
+      "grad_norm": 0.2127104252576828,
+      "learning_rate": 4.9774773271173494e-06,
+      "loss": 0.6689,
+      "step": 400
+    },
+    {
+      "epoch": 0.16903783328064073,
+      "grad_norm": 0.2096283733844757,
+      "learning_rate": 4.9773271534601205e-06,
+      "loss": 0.649,
+      "step": 401
+    },
+    {
+      "epoch": 0.1694593740120139,
+      "grad_norm": 0.20685532689094543,
+      "learning_rate": 4.977176483091706e-06,
+      "loss": 0.6583,
+      "step": 402
+    },
+    {
+      "epoch": 0.16988091474338707,
+      "grad_norm": 0.22643961012363434,
+      "learning_rate": 4.977025316042316e-06,
+      "loss": 0.6447,
+      "step": 403
+    },
+    {
+      "epoch": 0.17030245547476025,
+      "grad_norm": 0.21273617446422577,
+      "learning_rate": 4.976873652342259e-06,
+      "loss": 0.6276,
+      "step": 404
+    },
+    {
+      "epoch": 0.1707239962061334,
+      "grad_norm": 0.215843066573143,
+      "learning_rate": 4.976721492021945e-06,
+      "loss": 0.6395,
+      "step": 405
+    },
+    {
+      "epoch": 0.1711455369375066,
+      "grad_norm": 0.2145632803440094,
+      "learning_rate": 4.976568835111882e-06,
+      "loss": 0.6568,
+      "step": 406
+    },
+    {
+      "epoch": 0.17156707766887974,
+      "grad_norm": 0.22427786886692047,
+      "learning_rate": 4.976415681642678e-06,
+      "loss": 0.6351,
+      "step": 407
+    },
+    {
+      "epoch": 0.17198861840025292,
+      "grad_norm": 0.21854615211486816,
+      "learning_rate": 4.97626203164504e-06,
+      "loss": 0.6957,
+      "step": 408
+    },
+    {
+      "epoch": 0.1724101591316261,
+      "grad_norm": 0.21164420247077942,
+      "learning_rate": 4.976107885149777e-06,
+      "loss": 0.6592,
+      "step": 409
+    },
+    {
+      "epoch": 0.17283169986299926,
+      "grad_norm": 0.22965668141841888,
+      "learning_rate": 4.9759532421877956e-06,
+      "loss": 0.6589,
+      "step": 410
+    },
+    {
+      "epoch": 0.17325324059437244,
+      "grad_norm": 0.2169698178768158,
+      "learning_rate": 4.9757981027901004e-06,
+      "loss": 0.6339,
+      "step": 411
+    },
+    {
+      "epoch": 0.1736747813257456,
+      "grad_norm": 0.2144223302602768,
+      "learning_rate": 4.975642466987799e-06,
+      "loss": 0.6302,
+      "step": 412
+    },
+    {
+      "epoch": 0.17409632205711878,
+      "grad_norm": 0.2207185924053192,
+      "learning_rate": 4.975486334812096e-06,
+      "loss": 0.6293,
+      "step": 413
+    },
+    {
+      "epoch": 0.17451786278849193,
+      "grad_norm": 0.22312521934509277,
+      "learning_rate": 4.975329706294297e-06,
+      "loss": 0.6574,
+      "step": 414
+    },
+    {
+      "epoch": 0.17493940351986512,
+      "grad_norm": 0.21296381950378418,
+      "learning_rate": 4.975172581465806e-06,
+      "loss": 0.6432,
+      "step": 415
+    },
+    {
+      "epoch": 0.17536094425123827,
+      "grad_norm": 0.21974453330039978,
+      "learning_rate": 4.975014960358126e-06,
+      "loss": 0.6406,
+      "step": 416
+    },
+    {
+      "epoch": 0.17578248498261145,
+      "grad_norm": 0.22318316996097565,
+      "learning_rate": 4.974856843002863e-06,
+      "loss": 0.6444,
+      "step": 417
+    },
+    {
+      "epoch": 0.1762040257139846,
+      "grad_norm": 0.23577533662319183,
+      "learning_rate": 4.974698229431718e-06,
+      "loss": 0.657,
+      "step": 418
+    },
+    {
+      "epoch": 0.1766255664453578,
+      "grad_norm": 0.23191891610622406,
+      "learning_rate": 4.9745391196764936e-06,
+      "loss": 0.6507,
+      "step": 419
+    },
+    {
+      "epoch": 0.17704710717673094,
+      "grad_norm": 0.22186735272407532,
+      "learning_rate": 4.974379513769093e-06,
+      "loss": 0.6517,
+      "step": 420
+    },
+    {
+      "epoch": 0.17746864790810413,
+      "grad_norm": 0.23127466440200806,
+      "learning_rate": 4.974219411741516e-06,
+      "loss": 0.6619,
+      "step": 421
+    },
+    {
+      "epoch": 0.17789018863947728,
+      "grad_norm": 0.22906289994716644,
+      "learning_rate": 4.974058813625865e-06,
+      "loss": 0.6456,
+      "step": 422
+    },
+    {
+      "epoch": 0.17831172937085046,
+      "grad_norm": 0.22450274229049683,
+      "learning_rate": 4.973897719454339e-06,
+      "loss": 0.6628,
+      "step": 423
+    },
+    {
+      "epoch": 0.17873327010222362,
+      "grad_norm": 0.2400558739900589,
+      "learning_rate": 4.973736129259239e-06,
+      "loss": 0.6606,
+      "step": 424
+    },
+    {
+      "epoch": 0.1791548108335968,
+      "grad_norm": 0.23657605051994324,
+      "learning_rate": 4.973574043072963e-06,
+      "loss": 0.6195,
+      "step": 425
+    },
+    {
+      "epoch": 0.17957635156496995,
+      "grad_norm": 0.21887663006782532,
+      "learning_rate": 4.973411460928011e-06,
+      "loss": 0.6379,
+      "step": 426
+    },
+    {
+      "epoch": 0.17999789229634314,
+      "grad_norm": 0.22087790071964264,
+      "learning_rate": 4.973248382856981e-06,
+      "loss": 0.6206,
+      "step": 427
+    },
+    {
+      "epoch": 0.1804194330277163,
+      "grad_norm": 0.2348846197128296,
+      "learning_rate": 4.973084808892571e-06,
+      "loss": 0.6543,
+      "step": 428
+    },
+    {
+      "epoch": 0.18084097375908947,
+      "grad_norm": 0.23617230355739594,
+      "learning_rate": 4.972920739067576e-06,
+      "loss": 0.6679,
+      "step": 429
+    },
+    {
+      "epoch": 0.18126251449046263,
+      "grad_norm": 0.22668734192848206,
+      "learning_rate": 4.972756173414895e-06,
+      "loss": 0.6269,
+      "step": 430
+    },
+    {
+      "epoch": 0.1816840552218358,
+      "grad_norm": 0.2268986999988556,
+      "learning_rate": 4.972591111967523e-06,
+      "loss": 0.6393,
+      "step": 431
+    },
+    {
+      "epoch": 0.18210559595320897,
+      "grad_norm": 0.22111493349075317,
+      "learning_rate": 4.972425554758554e-06,
+      "loss": 0.6618,
+      "step": 432
+    },
+    {
+      "epoch": 0.18252713668458215,
+      "grad_norm": 0.22980272769927979,
+      "learning_rate": 4.972259501821184e-06,
+      "loss": 0.6473,
+      "step": 433
+    },
+    {
+      "epoch": 0.18294867741595533,
+      "grad_norm": 0.2237127721309662,
+      "learning_rate": 4.9720929531887075e-06,
+      "loss": 0.6571,
+      "step": 434
+    },
+    {
+      "epoch": 0.18337021814732848,
+      "grad_norm": 0.22420798242092133,
+      "learning_rate": 4.971925908894517e-06,
+      "loss": 0.6362,
+      "step": 435
+    },
+    {
+      "epoch": 0.18379175887870167,
+      "grad_norm": 0.21993198990821838,
+      "learning_rate": 4.971758368972105e-06,
+      "loss": 0.6634,
+      "step": 436
+    },
+    {
+      "epoch": 0.18421329961007482,
+      "grad_norm": 0.22855877876281738,
+      "learning_rate": 4.9715903334550654e-06,
+      "loss": 0.6481,
+      "step": 437
+    },
+    {
+      "epoch": 0.184634840341448,
+      "grad_norm": 0.22450317442417145,
+      "learning_rate": 4.9714218023770885e-06,
+      "loss": 0.6404,
+      "step": 438
+    },
+    {
+      "epoch": 0.18505638107282116,
+      "grad_norm": 0.22329369187355042,
+      "learning_rate": 4.971252775771965e-06,
+      "loss": 0.659,
+      "step": 439
+    },
+    {
+      "epoch": 0.18547792180419434,
+      "grad_norm": 0.22804588079452515,
+      "learning_rate": 4.971083253673587e-06,
+      "loss": 0.6469,
+      "step": 440
+    },
+    {
+      "epoch": 0.1858994625355675,
+      "grad_norm": 0.22212441265583038,
+      "learning_rate": 4.970913236115942e-06,
+      "loss": 0.6594,
+      "step": 441
+    },
+    {
+      "epoch": 0.18632100326694068,
+      "grad_norm": 0.2262367606163025,
+      "learning_rate": 4.97074272313312e-06,
+      "loss": 0.6328,
+      "step": 442
+    },
+    {
+      "epoch": 0.18674254399831383,
+      "grad_norm": 0.22934353351593018,
+      "learning_rate": 4.97057171475931e-06,
+      "loss": 0.6617,
+      "step": 443
+    },
+    {
+      "epoch": 0.187164084729687,
+      "grad_norm": 0.2225915640592575,
+      "learning_rate": 4.970400211028798e-06,
+      "loss": 0.6272,
+      "step": 444
+    },
+    {
+      "epoch": 0.18758562546106017,
+      "grad_norm": 0.2209312468767166,
+      "learning_rate": 4.970228211975973e-06,
+      "loss": 0.6557,
+      "step": 445
+    },
+    {
+      "epoch": 0.18800716619243335,
+      "grad_norm": 0.21495817601680756,
+      "learning_rate": 4.970055717635321e-06,
+      "loss": 0.631,
+      "step": 446
+    },
+    {
+      "epoch": 0.1884287069238065,
+      "grad_norm": 0.23232516646385193,
+      "learning_rate": 4.969882728041425e-06,
+      "loss": 0.67,
+      "step": 447
+    },
+    {
+      "epoch": 0.1888502476551797,
+      "grad_norm": 0.22030207514762878,
+      "learning_rate": 4.969709243228972e-06,
+      "loss": 0.6181,
+      "step": 448
+    },
+    {
+      "epoch": 0.18927178838655284,
+      "grad_norm": 0.2210187464952469,
+      "learning_rate": 4.969535263232747e-06,
+      "loss": 0.6505,
+      "step": 449
+    },
+    {
+      "epoch": 0.18969332911792602,
+      "grad_norm": 0.21762901544570923,
+      "learning_rate": 4.969360788087632e-06,
+      "loss": 0.6316,
+      "step": 450
+    },
+    {
+      "epoch": 0.19011486984929918,
+      "grad_norm": 0.21815869212150574,
+      "learning_rate": 4.96918581782861e-06,
+      "loss": 0.6379,
+      "step": 451
+    },
+    {
+      "epoch": 0.19053641058067236,
+      "grad_norm": 0.22199398279190063,
+      "learning_rate": 4.969010352490764e-06,
+      "loss": 0.6366,
+      "step": 452
+    },
+    {
+      "epoch": 0.19095795131204552,
+      "grad_norm": 0.22442398965358734,
+      "learning_rate": 4.968834392109274e-06,
+      "loss": 0.6407,
+      "step": 453
+    },
+    {
+      "epoch": 0.1913794920434187,
+      "grad_norm": 0.21754616498947144,
+      "learning_rate": 4.968657936719422e-06,
+      "loss": 0.6285,
+      "step": 454
+    },
+    {
+      "epoch": 0.19180103277479185,
+      "grad_norm": 0.20777983963489532,
+      "learning_rate": 4.968480986356586e-06,
+      "loss": 0.6651,
+      "step": 455
+    },
+    {
+      "epoch": 0.19222257350616503,
+      "grad_norm": 0.22279435396194458,
+      "learning_rate": 4.968303541056247e-06,
+      "loss": 0.6319,
+      "step": 456
+    },
+    {
+      "epoch": 0.1926441142375382,
+      "grad_norm": 0.2189827561378479,
+      "learning_rate": 4.968125600853982e-06,
+      "loss": 0.6609,
+      "step": 457
+    },
+    {
+      "epoch": 0.19306565496891137,
+      "grad_norm": 0.22233273088932037,
+      "learning_rate": 4.967947165785467e-06,
+      "loss": 0.6571,
+      "step": 458
+    },
+    {
+      "epoch": 0.19348719570028455,
+      "grad_norm": 0.21394787728786469,
+      "learning_rate": 4.967768235886483e-06,
+      "loss": 0.6478,
+      "step": 459
+    },
+    {
+      "epoch": 0.1939087364316577,
+      "grad_norm": 0.23329754173755646,
+      "learning_rate": 4.967588811192902e-06,
+      "loss": 0.662,
+      "step": 460
+    },
+    {
+      "epoch": 0.1943302771630309,
+      "grad_norm": 0.2127532660961151,
+      "learning_rate": 4.967408891740702e-06,
+      "loss": 0.6594,
+      "step": 461
+    },
+    {
+      "epoch": 0.19475181789440404,
+      "grad_norm": 0.22482381761074066,
+      "learning_rate": 4.967228477565955e-06,
+      "loss": 0.6587,
+      "step": 462
+    },
+    {
+      "epoch": 0.19517335862577723,
+      "grad_norm": 0.23118707537651062,
+      "learning_rate": 4.967047568704836e-06,
+      "loss": 0.6379,
+      "step": 463
+    },
+    {
+      "epoch": 0.19559489935715038,
+      "grad_norm": 0.21856175363063812,
+      "learning_rate": 4.966866165193617e-06,
+      "loss": 0.6586,
+      "step": 464
+    },
+    {
+      "epoch": 0.19601644008852356,
+      "grad_norm": 0.21749050915241241,
+      "learning_rate": 4.966684267068671e-06,
+      "loss": 0.6654,
+      "step": 465
+    },
+    {
+      "epoch": 0.19643798081989672,
+      "grad_norm": 0.22064447402954102,
+      "learning_rate": 4.9665018743664676e-06,
+      "loss": 0.6404,
+      "step": 466
+    },
+    {
+      "epoch": 0.1968595215512699,
+      "grad_norm": 0.2192775160074234,
+      "learning_rate": 4.9663189871235784e-06,
+      "loss": 0.6662,
+      "step": 467
+    },
+    {
+      "epoch": 0.19728106228264305,
+      "grad_norm": 0.22765064239501953,
+      "learning_rate": 4.966135605376672e-06,
+      "loss": 0.6206,
+      "step": 468
+    },
+    {
+      "epoch": 0.19770260301401624,
+      "grad_norm": 0.21945759654045105,
+      "learning_rate": 4.965951729162517e-06,
+      "loss": 0.6589,
+      "step": 469
+    },
+    {
+      "epoch": 0.1981241437453894,
+      "grad_norm": 0.22714141011238098,
+      "learning_rate": 4.965767358517982e-06,
+      "loss": 0.6597,
+      "step": 470
+    },
+    {
+      "epoch": 0.19854568447676257,
+      "grad_norm": 0.21871019899845123,
+      "learning_rate": 4.9655824934800325e-06,
+      "loss": 0.6449,
+      "step": 471
+    },
+    {
+      "epoch": 0.19896722520813573,
+      "grad_norm": 0.23086382448673248,
+      "learning_rate": 4.965397134085736e-06,
+      "loss": 0.6607,
+      "step": 472
+    },
+    {
+      "epoch": 0.1993887659395089,
+      "grad_norm": 0.22550630569458008,
+      "learning_rate": 4.965211280372256e-06,
+      "loss": 0.6494,
+      "step": 473
+    },
+    {
+      "epoch": 0.19981030667088207,
+      "grad_norm": 0.2192370444536209,
+      "learning_rate": 4.965024932376857e-06,
+      "loss": 0.6421,
+      "step": 474
+    },
+    {
+      "epoch": 0.20023184740225525,
+      "grad_norm": 0.23266327381134033,
+      "learning_rate": 4.9648380901369044e-06,
+      "loss": 0.6672,
+      "step": 475
+    },
+    {
+      "epoch": 0.2006533881336284,
+      "grad_norm": 0.21692770719528198,
+      "learning_rate": 4.964650753689857e-06,
+      "loss": 0.6454,
+      "step": 476
+    },
+    {
+      "epoch": 0.20107492886500158,
+      "grad_norm": 0.22346735000610352,
+      "learning_rate": 4.96446292307328e-06,
+      "loss": 0.6438,
+      "step": 477
+    },
+    {
+      "epoch": 0.20149646959637474,
+      "grad_norm": 0.22075144946575165,
+      "learning_rate": 4.964274598324831e-06,
+      "loss": 0.6704,
+      "step": 478
+    },
+    {
+      "epoch": 0.20191801032774792,
+      "grad_norm": 0.22169655561447144,
+      "learning_rate": 4.964085779482271e-06,
+      "loss": 0.6341,
+      "step": 479
+    },
+    {
+      "epoch": 0.20233955105912108,
+      "grad_norm": 0.2210836112499237,
+      "learning_rate": 4.963896466583459e-06,
+      "loss": 0.6558,
+      "step": 480
+    },
+    {
+      "epoch": 0.20276109179049426,
+      "grad_norm": 0.22162769734859467,
+      "learning_rate": 4.963706659666352e-06,
+      "loss": 0.6541,
+      "step": 481
+    },
+    {
+      "epoch": 0.2031826325218674,
+      "grad_norm": 0.2314087152481079,
+      "learning_rate": 4.963516358769007e-06,
+      "loss": 0.6314,
+      "step": 482
+    },
+    {
+      "epoch": 0.2036041732532406,
+      "grad_norm": 0.22203563153743744,
+      "learning_rate": 4.963325563929581e-06,
+      "loss": 0.6438,
+      "step": 483
+    },
+    {
+      "epoch": 0.20402571398461378,
+      "grad_norm": 0.22781838476657867,
+      "learning_rate": 4.963134275186327e-06,
+      "loss": 0.6444,
+      "step": 484
+    },
+    {
+      "epoch": 0.20444725471598693,
+      "grad_norm": 0.2267368882894516,
+      "learning_rate": 4.9629424925776e-06,
+      "loss": 0.6477,
+      "step": 485
+    },
+    {
+      "epoch": 0.2048687954473601,
+      "grad_norm": 0.2233951836824417,
+      "learning_rate": 4.962750216141854e-06,
+      "loss": 0.6492,
+      "step": 486
+    },
+    {
+      "epoch": 0.20529033617873327,
+      "grad_norm": 0.23200178146362305,
+      "learning_rate": 4.9625574459176385e-06,
+      "loss": 0.6959,
+      "step": 487
+    },
+    {
+      "epoch": 0.20571187691010645,
+      "grad_norm": 0.23076531291007996,
+      "learning_rate": 4.962364181943606e-06,
+      "loss": 0.6435,
+      "step": 488
+    },
+    {
+      "epoch": 0.2061334176414796,
+      "grad_norm": 0.23725484311580658,
+      "learning_rate": 4.962170424258508e-06,
+      "loss": 0.6642,
+      "step": 489
+    },
+    {
+      "epoch": 0.2065549583728528,
+      "grad_norm": 0.2263781726360321,
+      "learning_rate": 4.961976172901189e-06,
+      "loss": 0.6212,
+      "step": 490
+    },
+    {
+      "epoch": 0.20697649910422594,
+      "grad_norm": 0.22753490507602692,
+      "learning_rate": 4.961781427910602e-06,
+      "loss": 0.6547,
+      "step": 491
+    },
+    {
+      "epoch": 0.20739803983559912,
+      "grad_norm": 0.22493496537208557,
+      "learning_rate": 4.961586189325791e-06,
+      "loss": 0.6449,
+      "step": 492
+    },
+    {
+      "epoch": 0.20781958056697228,
+      "grad_norm": 0.23242434859275818,
+      "learning_rate": 4.961390457185902e-06,
+      "loss": 0.6636,
+      "step": 493
+    },
+    {
+      "epoch": 0.20824112129834546,
+      "grad_norm": 0.2310808002948761,
+      "learning_rate": 4.961194231530181e-06,
+      "loss": 0.649,
+      "step": 494
+    },
+    {
+      "epoch": 0.20866266202971862,
+      "grad_norm": 0.22858233749866486,
+      "learning_rate": 4.960997512397971e-06,
+      "loss": 0.6558,
+      "step": 495
+    },
+    {
+      "epoch": 0.2090842027610918,
+      "grad_norm": 0.22522948682308197,
+      "learning_rate": 4.960800299828715e-06,
+      "loss": 0.6699,
+      "step": 496
+    },
+    {
+      "epoch": 0.20950574349246495,
+      "grad_norm": 0.2328474521636963,
+      "learning_rate": 4.960602593861955e-06,
+      "loss": 0.6489,
+      "step": 497
+    },
+    {
+      "epoch": 0.20992728422383813,
+      "grad_norm": 0.2228248119354248,
+      "learning_rate": 4.96040439453733e-06,
+      "loss": 0.6617,
+      "step": 498
+    },
+    {
+      "epoch": 0.2103488249552113,
+      "grad_norm": 0.23502282798290253,
+      "learning_rate": 4.960205701894583e-06,
+      "loss": 0.6553,
+      "step": 499
+    },
+    {
+      "epoch": 0.21077036568658447,
+      "grad_norm": 0.22323139011859894,
+      "learning_rate": 4.960006515973548e-06,
+      "loss": 0.6329,
+      "step": 500
+    },
+    {
+      "epoch": 0.21119190641795763,
+      "grad_norm": 0.2224983423948288,
+      "learning_rate": 4.959806836814166e-06,
+      "loss": 0.6451,
+      "step": 501
+    },
+    {
+      "epoch": 0.2116134471493308,
+      "grad_norm": 0.22500398755073547,
+      "learning_rate": 4.9596066644564715e-06,
+      "loss": 0.6219,
+      "step": 502
+    },
+    {
+      "epoch": 0.21203498788070396,
+      "grad_norm": 0.22492237389087677,
+      "learning_rate": 4.9594059989406e-06,
+      "loss": 0.6564,
+      "step": 503
+    },
+    {
+      "epoch": 0.21245652861207714,
+      "grad_norm": 0.23181870579719543,
+      "learning_rate": 4.959204840306785e-06,
+      "loss": 0.6504,
+      "step": 504
+    },
+    {
+      "epoch": 0.2128780693434503,
+      "grad_norm": 0.22693991661071777,
+      "learning_rate": 4.95900318859536e-06,
+      "loss": 0.647,
+      "step": 505
+    },
+    {
+      "epoch": 0.21329961007482348,
+      "grad_norm": 0.23504593968391418,
+      "learning_rate": 4.958801043846757e-06,
+      "loss": 0.6242,
+      "step": 506
+    },
+    {
+      "epoch": 0.21372115080619664,
+      "grad_norm": 0.2442081719636917,
+      "learning_rate": 4.9585984061015065e-06,
+      "loss": 0.6764,
+      "step": 507
+    },
+    {
+      "epoch": 0.21414269153756982,
+      "grad_norm": 0.23746952414512634,
+      "learning_rate": 4.958395275400237e-06,
+      "loss": 0.6612,
+      "step": 508
+    },
+    {
+      "epoch": 0.214564232268943,
+      "grad_norm": 0.23237504065036774,
+      "learning_rate": 4.958191651783678e-06,
+      "loss": 0.6793,
+      "step": 509
+    },
+    {
+      "epoch": 0.21498577300031615,
+      "grad_norm": 0.23623941838741302,
+      "learning_rate": 4.957987535292656e-06,
+      "loss": 0.6412,
+      "step": 510
+    },
+    {
+      "epoch": 0.21540731373168934,
+      "grad_norm": 0.23477403819561005,
+      "learning_rate": 4.957782925968096e-06,
+      "loss": 0.6673,
+      "step": 511
+    },
+    {
+      "epoch": 0.2158288544630625,
+      "grad_norm": 0.21841122210025787,
+      "learning_rate": 4.9575778238510245e-06,
+      "loss": 0.64,
+      "step": 512
+    },
+    {
+      "epoch": 0.21625039519443567,
+      "grad_norm": 0.22514000535011292,
+      "learning_rate": 4.957372228982564e-06,
+      "loss": 0.6522,
+      "step": 513
+    },
+    {
+      "epoch": 0.21667193592580883,
+      "grad_norm": 0.2274819165468216,
+      "learning_rate": 4.957166141403938e-06,
+      "loss": 0.6697,
+      "step": 514
+    },
+    {
+      "epoch": 0.217093476657182,
+      "grad_norm": 0.24151133000850677,
+      "learning_rate": 4.956959561156466e-06,
+      "loss": 0.6647,
+      "step": 515
+    },
+    {
+      "epoch": 0.21751501738855517,
+      "grad_norm": 0.2281079739332199,
+      "learning_rate": 4.956752488281569e-06,
+      "loss": 0.6029,
+      "step": 516
+    },
+    {
+      "epoch": 0.21793655811992835,
+      "grad_norm": 0.24034132063388824,
+      "learning_rate": 4.956544922820765e-06,
+      "loss": 0.6634,
+      "step": 517
+    },
+    {
+      "epoch": 0.2183580988513015,
+      "grad_norm": 0.2356782704591751,
+      "learning_rate": 4.956336864815673e-06,
+      "loss": 0.6233,
+      "step": 518
+    },
+    {
+      "epoch": 0.21877963958267468,
+      "grad_norm": 0.23048047721385956,
+      "learning_rate": 4.956128314308007e-06,
+      "loss": 0.6509,
+      "step": 519
+    },
+    {
+      "epoch": 0.21920118031404784,
+      "grad_norm": 0.23716945946216583,
+      "learning_rate": 4.9559192713395835e-06,
+      "loss": 0.6414,
+      "step": 520
+    },
+    {
+      "epoch": 0.21962272104542102,
+      "grad_norm": 0.22305086255073547,
+      "learning_rate": 4.955709735952317e-06,
+      "loss": 0.6242,
+      "step": 521
+    },
+    {
+      "epoch": 0.22004426177679418,
+      "grad_norm": 0.25302743911743164,
+      "learning_rate": 4.955499708188217e-06,
+      "loss": 0.6416,
+      "step": 522
+    },
+    {
+      "epoch": 0.22046580250816736,
+      "grad_norm": 0.2288902997970581,
+      "learning_rate": 4.9552891880893975e-06,
+      "loss": 0.6564,
+      "step": 523
+    },
+    {
+      "epoch": 0.2208873432395405,
+      "grad_norm": 0.22555950284004211,
+      "learning_rate": 4.955078175698067e-06,
+      "loss": 0.6402,
+      "step": 524
+    },
+    {
+      "epoch": 0.2213088839709137,
+      "grad_norm": 0.22635763883590698,
+      "learning_rate": 4.954866671056534e-06,
+      "loss": 0.6776,
+      "step": 525
+    },
+    {
+      "epoch": 0.22173042470228685,
+      "grad_norm": 0.2513132393360138,
+      "learning_rate": 4.954654674207207e-06,
+      "loss": 0.6291,
+      "step": 526
+    },
+    {
+      "epoch": 0.22215196543366003,
+      "grad_norm": 0.24248895049095154,
+      "learning_rate": 4.95444218519259e-06,
+      "loss": 0.6473,
+      "step": 527
+    },
+    {
+      "epoch": 0.22257350616503319,
+      "grad_norm": 0.23155072331428528,
+      "learning_rate": 4.9542292040552905e-06,
+      "loss": 0.6617,
+      "step": 528
+    },
+    {
+      "epoch": 0.22299504689640637,
+      "grad_norm": 0.22736796736717224,
+      "learning_rate": 4.95401573083801e-06,
+      "loss": 0.6436,
+      "step": 529
+    },
+    {
+      "epoch": 0.22341658762777952,
+      "grad_norm": 0.23843534290790558,
+      "learning_rate": 4.953801765583549e-06,
+      "loss": 0.6523,
+      "step": 530
+    },
+    {
+      "epoch": 0.2238381283591527,
+      "grad_norm": 0.2373620867729187,
+      "learning_rate": 4.953587308334812e-06,
+      "loss": 0.6434,
+      "step": 531
+    },
+    {
+      "epoch": 0.22425966909052586,
+      "grad_norm": 0.22916729748249054,
+      "learning_rate": 4.953372359134795e-06,
+      "loss": 0.6298,
+      "step": 532
+    },
+    {
+      "epoch": 0.22468120982189904,
+      "grad_norm": 0.23192116618156433,
+      "learning_rate": 4.953156918026597e-06,
+      "loss": 0.6307,
+      "step": 533
+    },
+    {
+      "epoch": 0.2251027505532722,
+      "grad_norm": 0.23496942222118378,
+      "learning_rate": 4.9529409850534134e-06,
+      "loss": 0.6604,
+      "step": 534
+    },
+    {
+      "epoch": 0.22552429128464538,
+      "grad_norm": 0.2322310507297516,
+      "learning_rate": 4.952724560258543e-06,
+      "loss": 0.6559,
+      "step": 535
+    },
+    {
+      "epoch": 0.22594583201601856,
+      "grad_norm": 0.23275437951087952,
+      "learning_rate": 4.952507643685375e-06,
+      "loss": 0.6479,
+      "step": 536
+    },
+    {
+      "epoch": 0.22636737274739172,
+      "grad_norm": 0.23254382610321045,
+      "learning_rate": 4.952290235377405e-06,
+      "loss": 0.6294,
+      "step": 537
+    },
+    {
+      "epoch": 0.2267889134787649,
+      "grad_norm": 0.23082607984542847,
+      "learning_rate": 4.952072335378223e-06,
+      "loss": 0.607,
+      "step": 538
+    },
+    {
+      "epoch": 0.22721045421013805,
+      "grad_norm": 0.23705212771892548,
+      "learning_rate": 4.951853943731518e-06,
+      "loss": 0.6502,
+      "step": 539
+    },
+    {
+      "epoch": 0.22763199494151123,
+      "grad_norm": 0.23072996735572815,
+      "learning_rate": 4.951635060481079e-06,
+      "loss": 0.6508,
+      "step": 540
+    },
+    {
+      "epoch": 0.2280535356728844,
+      "grad_norm": 0.21911437809467316,
+      "learning_rate": 4.951415685670793e-06,
+      "loss": 0.6112,
+      "step": 541
+    },
+    {
+      "epoch": 0.22847507640425757,
+      "grad_norm": 0.22660993039608002,
+      "learning_rate": 4.951195819344644e-06,
+      "loss": 0.6417,
+      "step": 542
+    },
+    {
+      "epoch": 0.22889661713563073,
+      "grad_norm": 0.23092220723628998,
+      "learning_rate": 4.950975461546716e-06,
+      "loss": 0.6524,
+      "step": 543
+    },
+    {
+      "epoch": 0.2293181578670039,
+      "grad_norm": 0.23236800730228424,
+      "learning_rate": 4.9507546123211926e-06,
+      "loss": 0.6595,
+      "step": 544
+    },
+    {
+      "epoch": 0.22973969859837706,
+      "grad_norm": 0.2335730642080307,
+      "learning_rate": 4.950533271712353e-06,
+      "loss": 0.6396,
+      "step": 545
+    },
+    {
+      "epoch": 0.23016123932975024,
+      "grad_norm": 0.23026856780052185,
+      "learning_rate": 4.950311439764579e-06,
+      "loss": 0.6514,
+      "step": 546
+    },
+    {
+      "epoch": 0.2305827800611234,
+      "grad_norm": 0.22758536040782928,
+      "learning_rate": 4.950089116522347e-06,
+      "loss": 0.6575,
+      "step": 547
+    },
+    {
+      "epoch": 0.23100432079249658,
+      "grad_norm": 0.24229104816913605,
+      "learning_rate": 4.949866302030232e-06,
+      "loss": 0.6465,
+      "step": 548
+    },
+    {
+      "epoch": 0.23142586152386974,
+      "grad_norm": 0.24203070998191833,
+      "learning_rate": 4.949642996332912e-06,
+      "loss": 0.6248,
+      "step": 549
+    },
+    {
+      "epoch": 0.23184740225524292,
+      "grad_norm": 0.23183302581310272,
+      "learning_rate": 4.949419199475159e-06,
+      "loss": 0.6478,
+      "step": 550
+    },
+    {
+      "epoch": 0.23226894298661607,
+      "grad_norm": 0.2225986272096634,
+      "learning_rate": 4.949194911501843e-06,
+      "loss": 0.6524,
+      "step": 551
+    },
+    {
+      "epoch": 0.23269048371798925,
+      "grad_norm": 0.22970671951770782,
+      "learning_rate": 4.948970132457938e-06,
+      "loss": 0.6479,
+      "step": 552
+    },
+    {
+      "epoch": 0.2331120244493624,
+      "grad_norm": 0.22436730563640594,
+      "learning_rate": 4.94874486238851e-06,
+      "loss": 0.6435,
+      "step": 553
+    },
+    {
+      "epoch": 0.2335335651807356,
+      "grad_norm": 0.2407492846250534,
+      "learning_rate": 4.948519101338729e-06,
+      "loss": 0.6237,
+      "step": 554
+    },
+    {
+      "epoch": 0.23395510591210875,
+      "grad_norm": 0.23236878216266632,
+      "learning_rate": 4.948292849353858e-06,
+      "loss": 0.6445,
+      "step": 555
+    },
+    {
+      "epoch": 0.23437664664348193,
+      "grad_norm": 0.22568278014659882,
+      "learning_rate": 4.948066106479262e-06,
+      "loss": 0.6358,
+      "step": 556
+    },
+    {
+      "epoch": 0.23479818737485508,
+      "grad_norm": 0.2250591665506363,
+      "learning_rate": 4.947838872760404e-06,
+      "loss": 0.6168,
+      "step": 557
+    },
+    {
+      "epoch": 0.23521972810622827,
+      "grad_norm": 0.23011131584644318,
+      "learning_rate": 4.947611148242846e-06,
+      "loss": 0.6548,
+      "step": 558
+    },
+    {
+      "epoch": 0.23564126883760142,
+      "grad_norm": 0.22665046155452728,
+      "learning_rate": 4.947382932972245e-06,
+      "loss": 0.6691,
+      "step": 559
+    },
+    {
+      "epoch": 0.2360628095689746,
+      "grad_norm": 0.2275232970714569,
+      "learning_rate": 4.94715422699436e-06,
+      "loss": 0.6531,
+      "step": 560
+    },
+    {
+      "epoch": 0.23648435030034778,
+      "grad_norm": 0.23100855946540833,
+      "learning_rate": 4.946925030355049e-06,
+      "loss": 0.6181,
+      "step": 561
+    },
+    {
+      "epoch": 0.23690589103172094,
+      "grad_norm": 0.2354455590248108,
+      "learning_rate": 4.9466953431002634e-06,
+      "loss": 0.6543,
+      "step": 562
+    },
+    {
+      "epoch": 0.23732743176309412,
+      "grad_norm": 0.23215655982494354,
+      "learning_rate": 4.946465165276058e-06,
+      "loss": 0.6615,
+      "step": 563
+    },
+    {
+      "epoch": 0.23774897249446728,
+      "grad_norm": 0.2307896912097931,
+      "learning_rate": 4.9462344969285835e-06,
+      "loss": 0.6296,
+      "step": 564
+    },
+    {
+      "epoch": 0.23817051322584046,
+      "grad_norm": 0.2466116100549698,
+      "learning_rate": 4.946003338104091e-06,
+      "loss": 0.6794,
+      "step": 565
+    },
+    {
+      "epoch": 0.2385920539572136,
+      "grad_norm": 0.22987677156925201,
+      "learning_rate": 4.945771688848926e-06,
+      "loss": 0.6555,
+      "step": 566
+    },
+    {
+      "epoch": 0.2390135946885868,
+      "grad_norm": 0.22888435423374176,
+      "learning_rate": 4.945539549209537e-06,
+      "loss": 0.6651,
+      "step": 567
+    },
+    {
+      "epoch": 0.23943513541995995,
+      "grad_norm": 0.22943688929080963,
+      "learning_rate": 4.9453069192324675e-06,
+      "loss": 0.6517,
+      "step": 568
+    },
+    {
+      "epoch": 0.23985667615133313,
+      "grad_norm": 0.2373216152191162,
+      "learning_rate": 4.94507379896436e-06,
+      "loss": 0.6348,
+      "step": 569
+    },
+    {
+      "epoch": 0.24027821688270629,
+      "grad_norm": 0.23128359019756317,
+      "learning_rate": 4.944840188451958e-06,
+      "loss": 0.6548,
+      "step": 570
+    },
+    {
+      "epoch": 0.24069975761407947,
+      "grad_norm": 0.23508068919181824,
+      "learning_rate": 4.944606087742099e-06,
+      "loss": 0.63,
+      "step": 571
+    },
+    {
+      "epoch": 0.24112129834545262,
+      "grad_norm": 0.24441313743591309,
+      "learning_rate": 4.944371496881721e-06,
+      "loss": 0.6485,
+      "step": 572
+    },
+    {
+      "epoch": 0.2415428390768258,
+      "grad_norm": 0.23405586183071136,
+      "learning_rate": 4.944136415917862e-06,
+      "loss": 0.6395,
+      "step": 573
+    },
+    {
+      "epoch": 0.24196437980819896,
+      "grad_norm": 0.22768528759479523,
+      "learning_rate": 4.9439008448976545e-06,
+      "loss": 0.6287,
+      "step": 574
+    },
+    {
+      "epoch": 0.24238592053957214,
+      "grad_norm": 0.23460225760936737,
+      "learning_rate": 4.943664783868331e-06,
+      "loss": 0.6458,
+      "step": 575
+    },
+    {
+      "epoch": 0.2428074612709453,
+      "grad_norm": 0.23477911949157715,
+      "learning_rate": 4.9434282328772245e-06,
+      "loss": 0.6618,
+      "step": 576
+    },
+    {
+      "epoch": 0.24322900200231848,
+      "grad_norm": 0.2367464154958725,
+      "learning_rate": 4.943191191971762e-06,
+      "loss": 0.6503,
+      "step": 577
+    },
+    {
+      "epoch": 0.24365054273369163,
+      "grad_norm": 0.23600925505161285,
+      "learning_rate": 4.9429536611994715e-06,
+      "loss": 0.6617,
+      "step": 578
+    },
+    {
+      "epoch": 0.24407208346506482,
+      "grad_norm": 0.2309107929468155,
+      "learning_rate": 4.942715640607979e-06,
+      "loss": 0.6212,
+      "step": 579
+    },
+    {
+      "epoch": 0.24449362419643797,
+      "grad_norm": 0.24014215171337128,
+      "learning_rate": 4.942477130245009e-06,
+      "loss": 0.6437,
+      "step": 580
+    },
+    {
+      "epoch": 0.24491516492781115,
+      "grad_norm": 0.23044876754283905,
+      "learning_rate": 4.942238130158381e-06,
+      "loss": 0.6494,
+      "step": 581
+    },
+    {
+      "epoch": 0.2453367056591843,
+      "grad_norm": 0.22551876306533813,
+      "learning_rate": 4.941998640396018e-06,
+      "loss": 0.6247,
+      "step": 582
+    },
+    {
+      "epoch": 0.2457582463905575,
+      "grad_norm": 0.24673812091350555,
+      "learning_rate": 4.9417586610059366e-06,
+      "loss": 0.6417,
+      "step": 583
+    },
+    {
+      "epoch": 0.24617978712193064,
+      "grad_norm": 0.23272879421710968,
+      "learning_rate": 4.941518192036255e-06,
+      "loss": 0.6613,
+      "step": 584
+    },
+    {
+      "epoch": 0.24660132785330383,
+      "grad_norm": 0.22514313459396362,
+      "learning_rate": 4.941277233535185e-06,
+      "loss": 0.639,
+      "step": 585
+    },
+    {
+      "epoch": 0.247022868584677,
+      "grad_norm": 0.24539217352867126,
+      "learning_rate": 4.941035785551043e-06,
+      "loss": 0.6453,
+      "step": 586
+    },
+    {
+      "epoch": 0.24744440931605016,
+      "grad_norm": 0.24861027300357819,
+      "learning_rate": 4.940793848132237e-06,
+      "loss": 0.6413,
+      "step": 587
+    },
+    {
+      "epoch": 0.24786595004742334,
+      "grad_norm": 0.23630225658416748,
+      "learning_rate": 4.940551421327279e-06,
+      "loss": 0.638,
+      "step": 588
+    },
+    {
+      "epoch": 0.2482874907787965,
+      "grad_norm": 0.24242550134658813,
+      "learning_rate": 4.940308505184773e-06,
+      "loss": 0.6313,
+      "step": 589
+    },
+    {
+      "epoch": 0.24870903151016968,
+      "grad_norm": 0.23441772162914276,
+      "learning_rate": 4.940065099753427e-06,
+      "loss": 0.6603,
+      "step": 590
+    },
+    {
+      "epoch": 0.24913057224154284,
+      "grad_norm": 0.230148583650589,
+      "learning_rate": 4.939821205082045e-06,
+      "loss": 0.6576,
+      "step": 591
+    },
+    {
+      "epoch": 0.24955211297291602,
+      "grad_norm": 0.24094484746456146,
+      "learning_rate": 4.939576821219525e-06,
+      "loss": 0.6595,
+      "step": 592
+    },
+    {
+      "epoch": 0.24997365370428917,
+      "grad_norm": 0.24190938472747803,
+      "learning_rate": 4.93933194821487e-06,
+      "loss": 0.6572,
+      "step": 593
+    },
+    {
+      "epoch": 0.2503951944356623,
+      "grad_norm": 0.23510822653770447,
+      "learning_rate": 4.939086586117177e-06,
+      "loss": 0.6401,
+      "step": 594
+    },
+    {
+      "epoch": 0.2508167351670355,
+      "grad_norm": 0.24858558177947998,
+      "learning_rate": 4.93884073497564e-06,
+      "loss": 0.646,
+      "step": 595
+    },
+    {
+      "epoch": 0.2512382758984087,
+      "grad_norm": 0.25159332156181335,
+      "learning_rate": 4.9385943948395556e-06,
+      "loss": 0.6773,
+      "step": 596
+    },
+    {
+      "epoch": 0.2516598166297819,
+      "grad_norm": 0.23862655460834503,
+      "learning_rate": 4.938347565758313e-06,
+      "loss": 0.6484,
+      "step": 597
+    },
+    {
+      "epoch": 0.252081357361155,
+      "grad_norm": 0.24019469320774078,
+      "learning_rate": 4.938100247781405e-06,
+      "loss": 0.6456,
+      "step": 598
+    },
+    {
+      "epoch": 0.2525028980925282,
+      "grad_norm": 0.2486957311630249,
+      "learning_rate": 4.937852440958417e-06,
+      "loss": 0.6572,
+      "step": 599
+    },
+    {
+      "epoch": 0.25292443882390137,
+      "grad_norm": 0.2364828735589981,
+      "learning_rate": 4.937604145339037e-06,
+      "loss": 0.6173,
+      "step": 600
+    },
+    {
+      "epoch": 0.25334597955527455,
+      "grad_norm": 0.24158920347690582,
+      "learning_rate": 4.937355360973047e-06,
+      "loss": 0.6515,
+      "step": 601
+    },
+    {
+      "epoch": 0.2537675202866477,
+      "grad_norm": 0.23431281745433807,
+      "learning_rate": 4.93710608791033e-06,
+      "loss": 0.6425,
+      "step": 602
+    },
+    {
+      "epoch": 0.25418906101802086,
+      "grad_norm": 0.23290280997753143,
+      "learning_rate": 4.936856326200867e-06,
+      "loss": 0.6521,
+      "step": 603
+    },
+    {
+      "epoch": 0.25461060174939404,
+      "grad_norm": 0.23566517233848572,
+      "learning_rate": 4.936606075894734e-06,
+      "loss": 0.6633,
+      "step": 604
+    },
+    {
+      "epoch": 0.2550321424807672,
+      "grad_norm": 0.2343212217092514,
+      "learning_rate": 4.936355337042109e-06,
+      "loss": 0.6336,
+      "step": 605
+    },
+    {
+      "epoch": 0.25545368321214035,
+      "grad_norm": 0.2344362884759903,
+      "learning_rate": 4.936104109693264e-06,
+      "loss": 0.6393,
+      "step": 606
+    },
+    {
+      "epoch": 0.25587522394351353,
+      "grad_norm": 0.2336214780807495,
+      "learning_rate": 4.9358523938985715e-06,
+      "loss": 0.6687,
+      "step": 607
+    },
+    {
+      "epoch": 0.2562967646748867,
+      "grad_norm": 0.24387873709201813,
+      "learning_rate": 4.9356001897085e-06,
+      "loss": 0.6457,
+      "step": 608
+    },
+    {
+      "epoch": 0.2567183054062599,
+      "grad_norm": 0.22620625793933868,
+      "learning_rate": 4.93534749717362e-06,
+      "loss": 0.6354,
+      "step": 609
+    },
+    {
+      "epoch": 0.257139846137633,
+      "grad_norm": 0.2580990791320801,
+      "learning_rate": 4.9350943163445956e-06,
+      "loss": 0.6321,
+      "step": 610
+    },
+    {
+      "epoch": 0.2575613868690062,
+      "grad_norm": 0.2368079125881195,
+      "learning_rate": 4.9348406472721884e-06,
+      "loss": 0.6405,
+      "step": 611
+    },
+    {
+      "epoch": 0.2579829276003794,
+      "grad_norm": 0.24162867665290833,
+      "learning_rate": 4.9345864900072625e-06,
+      "loss": 0.6569,
+      "step": 612
+    },
+    {
+      "epoch": 0.25840446833175257,
+      "grad_norm": 0.2327560931444168,
+      "learning_rate": 4.934331844600777e-06,
+      "loss": 0.6281,
+      "step": 613
+    },
+    {
+      "epoch": 0.25882600906312575,
+      "grad_norm": 0.2410680055618286,
+      "learning_rate": 4.934076711103787e-06,
+      "loss": 0.6386,
+      "step": 614
+    },
+    {
+      "epoch": 0.2592475497944989,
+      "grad_norm": 0.236967995762825,
+      "learning_rate": 4.933821089567448e-06,
+      "loss": 0.6716,
+      "step": 615
+    },
+    {
+      "epoch": 0.25966909052587206,
+      "grad_norm": 0.23831325769424438,
+      "learning_rate": 4.933564980043016e-06,
+      "loss": 0.6403,
+      "step": 616
+    },
+    {
+      "epoch": 0.26009063125724524,
+      "grad_norm": 0.23473335802555084,
+      "learning_rate": 4.933308382581837e-06,
+      "loss": 0.6506,
+      "step": 617
+    },
+    {
+      "epoch": 0.2605121719886184,
+      "grad_norm": 0.22875827550888062,
+      "learning_rate": 4.9330512972353634e-06,
+      "loss": 0.6468,
+      "step": 618
+    },
+    {
+      "epoch": 0.26093371271999155,
+      "grad_norm": 0.23231644928455353,
+      "learning_rate": 4.9327937240551394e-06,
+      "loss": 0.6653,
+      "step": 619
+    },
+    {
+      "epoch": 0.26135525345136473,
+      "grad_norm": 0.23044337332248688,
+      "learning_rate": 4.932535663092809e-06,
+      "loss": 0.6392,
+      "step": 620
+    },
+    {
+      "epoch": 0.2617767941827379,
+      "grad_norm": 0.24206477403640747,
+      "learning_rate": 4.9322771144001155e-06,
+      "loss": 0.6447,
+      "step": 621
+    },
+    {
+      "epoch": 0.2621983349141111,
+      "grad_norm": 0.22915469110012054,
+      "learning_rate": 4.9320180780288975e-06,
+      "loss": 0.639,
+      "step": 622
+    },
+    {
+      "epoch": 0.2626198756454842,
+      "grad_norm": 0.2405308037996292,
+      "learning_rate": 4.931758554031093e-06,
+      "loss": 0.64,
+      "step": 623
+    },
+    {
+      "epoch": 0.2630414163768574,
+      "grad_norm": 0.24254557490348816,
+      "learning_rate": 4.931498542458738e-06,
+      "loss": 0.6644,
+      "step": 624
+    },
+    {
+      "epoch": 0.2634629571082306,
+      "grad_norm": 0.24402983486652374,
+      "learning_rate": 4.931238043363964e-06,
+      "loss": 0.6504,
+      "step": 625
+    },
+    {
+      "epoch": 0.26388449783960377,
+      "grad_norm": 0.23167647421360016,
+      "learning_rate": 4.930977056799003e-06,
+      "loss": 0.6241,
+      "step": 626
+    },
+    {
+      "epoch": 0.2643060385709769,
+      "grad_norm": 0.23658999800682068,
+      "learning_rate": 4.930715582816183e-06,
+      "loss": 0.6177,
+      "step": 627
+    },
+    {
+      "epoch": 0.2647275793023501,
+      "grad_norm": 0.23379571735858917,
+      "learning_rate": 4.930453621467931e-06,
+      "loss": 0.6213,
+      "step": 628
+    },
+    {
+      "epoch": 0.26514912003372326,
+      "grad_norm": 0.25034937262535095,
+      "learning_rate": 4.9301911728067705e-06,
+      "loss": 0.6495,
+      "step": 629
+    },
+    {
+      "epoch": 0.26557066076509644,
+      "grad_norm": 0.23181910812854767,
+      "learning_rate": 4.929928236885324e-06,
+      "loss": 0.6373,
+      "step": 630
+    },
+    {
+      "epoch": 0.26599220149646957,
+      "grad_norm": 0.2382071316242218,
+      "learning_rate": 4.9296648137563095e-06,
+      "loss": 0.6619,
+      "step": 631
+    },
+    {
+      "epoch": 0.26641374222784275,
+      "grad_norm": 0.2502119541168213,
+      "learning_rate": 4.929400903472545e-06,
+      "loss": 0.6367,
+      "step": 632
+    },
+    {
+      "epoch": 0.26683528295921594,
+      "grad_norm": 0.2474728375673294,
+      "learning_rate": 4.929136506086944e-06,
+      "loss": 0.6489,
+      "step": 633
+    },
+    {
+      "epoch": 0.2672568236905891,
+      "grad_norm": 0.2416396290063858,
+      "learning_rate": 4.928871621652521e-06,
+      "loss": 0.6524,
+      "step": 634
+    },
+    {
+      "epoch": 0.26767836442196224,
+      "grad_norm": 0.2540903389453888,
+      "learning_rate": 4.928606250222384e-06,
+      "loss": 0.6196,
+      "step": 635
+    },
+    {
+      "epoch": 0.2680999051533354,
+      "grad_norm": 0.24426867067813873,
+      "learning_rate": 4.928340391849742e-06,
+      "loss": 0.6482,
+      "step": 636
+    },
+    {
+      "epoch": 0.2685214458847086,
+      "grad_norm": 0.2426331788301468,
+      "learning_rate": 4.9280740465879005e-06,
+      "loss": 0.6472,
+      "step": 637
+    },
+    {
+      "epoch": 0.2689429866160818,
+      "grad_norm": 0.23949024081230164,
+      "learning_rate": 4.9278072144902624e-06,
+      "loss": 0.6422,
+      "step": 638
+    },
+    {
+      "epoch": 0.269364527347455,
+      "grad_norm": 0.24121366441249847,
+      "learning_rate": 4.927539895610328e-06,
+      "loss": 0.6446,
+      "step": 639
+    },
+    {
+      "epoch": 0.2697860680788281,
+      "grad_norm": 0.23991113901138306,
+      "learning_rate": 4.927272090001695e-06,
+      "loss": 0.6287,
+      "step": 640
+    },
+    {
+      "epoch": 0.2702076088102013,
+      "grad_norm": 0.23528017103672028,
+      "learning_rate": 4.92700379771806e-06,
+      "loss": 0.6643,
+      "step": 641
+    },
+    {
+      "epoch": 0.27062914954157447,
+      "grad_norm": 0.24310113489627838,
+      "learning_rate": 4.9267350188132155e-06,
+      "loss": 0.6189,
+      "step": 642
+    },
+    {
+      "epoch": 0.27105069027294765,
+      "grad_norm": 0.24128517508506775,
+      "learning_rate": 4.926465753341054e-06,
+      "loss": 0.6429,
+      "step": 643
+    },
+    {
+      "epoch": 0.2714722310043208,
+      "grad_norm": 0.2314596027135849,
+      "learning_rate": 4.926196001355563e-06,
+      "loss": 0.642,
+      "step": 644
+    },
+    {
+      "epoch": 0.27189377173569396,
+      "grad_norm": 0.23932240903377533,
+      "learning_rate": 4.925925762910828e-06,
+      "loss": 0.6416,
+      "step": 645
+    },
+    {
+      "epoch": 0.27231531246706714,
+      "grad_norm": 0.2377786934375763,
+      "learning_rate": 4.925655038061033e-06,
+      "loss": 0.6403,
+      "step": 646
+    },
+    {
+      "epoch": 0.2727368531984403,
+      "grad_norm": 0.2351432591676712,
+      "learning_rate": 4.925383826860459e-06,
+      "loss": 0.6564,
+      "step": 647
+    },
+    {
+      "epoch": 0.27315839392981345,
+      "grad_norm": 0.2569757103919983,
+      "learning_rate": 4.9251121293634855e-06,
+      "loss": 0.6586,
+      "step": 648
+    },
+    {
+      "epoch": 0.27357993466118663,
+      "grad_norm": 0.24319124221801758,
+      "learning_rate": 4.924839945624588e-06,
+      "loss": 0.6443,
+      "step": 649
+    },
+    {
+      "epoch": 0.2740014753925598,
+      "grad_norm": 0.23898649215698242,
+      "learning_rate": 4.92456727569834e-06,
+      "loss": 0.6427,
+      "step": 650
+    },
+    {
+      "epoch": 0.274423016123933,
+      "grad_norm": 0.23469041287899017,
+      "learning_rate": 4.9242941196394124e-06,
+      "loss": 0.6249,
+      "step": 651
+    },
+    {
+      "epoch": 0.2748445568553061,
+      "grad_norm": 0.23133525252342224,
+      "learning_rate": 4.9240204775025746e-06,
+      "loss": 0.6377,
+      "step": 652
+    },
+    {
+      "epoch": 0.2752660975866793,
+      "grad_norm": 0.23608164489269257,
+      "learning_rate": 4.923746349342692e-06,
+      "loss": 0.6566,
+      "step": 653
+    },
+    {
+      "epoch": 0.2756876383180525,
+      "grad_norm": 0.24548587203025818,
+      "learning_rate": 4.923471735214728e-06,
+      "loss": 0.639,
+      "step": 654
+    },
+    {
+      "epoch": 0.27610917904942567,
+      "grad_norm": 0.2368946373462677,
+      "learning_rate": 4.923196635173744e-06,
+      "loss": 0.626,
+      "step": 655
+    },
+    {
+      "epoch": 0.2765307197807988,
+      "grad_norm": 0.23965714871883392,
+      "learning_rate": 4.922921049274897e-06,
+      "loss": 0.6172,
+      "step": 656
+    },
+    {
+      "epoch": 0.276952260512172,
+      "grad_norm": 0.24123302102088928,
+      "learning_rate": 4.922644977573446e-06,
+      "loss": 0.6523,
+      "step": 657
+    },
+    {
+      "epoch": 0.27737380124354516,
+      "grad_norm": 0.2342904508113861,
+      "learning_rate": 4.92236842012474e-06,
+      "loss": 0.6317,
+      "step": 658
+    },
+    {
+      "epoch": 0.27779534197491834,
+      "grad_norm": 0.24967117607593536,
+      "learning_rate": 4.922091376984232e-06,
+      "loss": 0.6577,
+      "step": 659
+    },
+    {
+      "epoch": 0.27821688270629147,
+      "grad_norm": 0.2520374655723572,
+      "learning_rate": 4.921813848207469e-06,
+      "loss": 0.6535,
+      "step": 660
+    },
+    {
+      "epoch": 0.27863842343766465,
+      "grad_norm": 0.251691609621048,
+      "learning_rate": 4.921535833850098e-06,
+      "loss": 0.6594,
+      "step": 661
+    },
+    {
+      "epoch": 0.27905996416903783,
+      "grad_norm": 0.2477567046880722,
+      "learning_rate": 4.92125733396786e-06,
+      "loss": 0.6614,
+      "step": 662
+    },
+    {
+      "epoch": 0.279481504900411,
+      "grad_norm": 0.24444128572940826,
+      "learning_rate": 4.920978348616595e-06,
+      "loss": 0.627,
+      "step": 663
+    },
+    {
+      "epoch": 0.2799030456317842,
+      "grad_norm": 0.24467122554779053,
+      "learning_rate": 4.920698877852242e-06,
+      "loss": 0.6509,
+      "step": 664
+    },
+    {
+      "epoch": 0.2803245863631573,
+      "grad_norm": 0.24350443482398987,
+      "learning_rate": 4.920418921730834e-06,
+      "loss": 0.6169,
+      "step": 665
+    },
+    {
+      "epoch": 0.2807461270945305,
+      "grad_norm": 0.2476222813129425,
+      "learning_rate": 4.920138480308503e-06,
+      "loss": 0.6631,
+      "step": 666
+    },
+    {
+      "epoch": 0.2811676678259037,
+      "grad_norm": 0.2540128827095032,
+      "learning_rate": 4.91985755364148e-06,
+      "loss": 0.6398,
+      "step": 667
+    },
+    {
+      "epoch": 0.28158920855727687,
+      "grad_norm": 0.25022536516189575,
+      "learning_rate": 4.919576141786089e-06,
+      "loss": 0.6465,
+      "step": 668
+    },
+    {
+      "epoch": 0.28201074928865,
+      "grad_norm": 0.2424604594707489,
+      "learning_rate": 4.919294244798757e-06,
+      "loss": 0.6331,
+      "step": 669
+    },
+    {
+      "epoch": 0.2824322900200232,
+      "grad_norm": 0.24882696568965912,
+      "learning_rate": 4.919011862736002e-06,
+      "loss": 0.6221,
+      "step": 670
+    },
+    {
+      "epoch": 0.28285383075139636,
+      "grad_norm": 0.2524561583995819,
+      "learning_rate": 4.918728995654445e-06,
+      "loss": 0.649,
+      "step": 671
+    },
+    {
+      "epoch": 0.28327537148276954,
+      "grad_norm": 0.24799224734306335,
+      "learning_rate": 4.9184456436107996e-06,
+      "loss": 0.669,
+      "step": 672
+    },
+    {
+      "epoch": 0.28369691221414267,
+      "grad_norm": 0.2447243630886078,
+      "learning_rate": 4.918161806661881e-06,
+      "loss": 0.6357,
+      "step": 673
+    },
+    {
+      "epoch": 0.28411845294551585,
+      "grad_norm": 0.2402827888727188,
+      "learning_rate": 4.917877484864597e-06,
+      "loss": 0.6458,
+      "step": 674
+    },
+    {
+      "epoch": 0.28453999367688904,
+      "grad_norm": 0.2410714477300644,
+      "learning_rate": 4.917592678275956e-06,
+      "loss": 0.6231,
+      "step": 675
+    },
+    {
+      "epoch": 0.2849615344082622,
+      "grad_norm": 0.2530798614025116,
+      "learning_rate": 4.917307386953063e-06,
+      "loss": 0.6479,
+      "step": 676
+    },
+    {
+      "epoch": 0.28538307513963534,
+      "grad_norm": 0.24118180572986603,
+      "learning_rate": 4.917021610953119e-06,
+      "loss": 0.6201,
+      "step": 677
+    },
+    {
+      "epoch": 0.2858046158710085,
+      "grad_norm": 0.2527044415473938,
+      "learning_rate": 4.916735350333422e-06,
+      "loss": 0.6538,
+      "step": 678
+    },
+    {
+      "epoch": 0.2862261566023817,
+      "grad_norm": 0.2409856915473938,
+      "learning_rate": 4.916448605151371e-06,
+      "loss": 0.6569,
+      "step": 679
+    },
+    {
+      "epoch": 0.2866476973337549,
+      "grad_norm": 0.2408577799797058,
+      "learning_rate": 4.916161375464455e-06,
+      "loss": 0.6129,
+      "step": 680
+    },
+    {
+      "epoch": 0.287069238065128,
+      "grad_norm": 0.2356063276529312,
+      "learning_rate": 4.915873661330269e-06,
+      "loss": 0.6367,
+      "step": 681
+    },
+    {
+      "epoch": 0.2874907787965012,
+      "grad_norm": 0.23885254561901093,
+      "learning_rate": 4.915585462806497e-06,
+      "loss": 0.6056,
+      "step": 682
+    },
+    {
+      "epoch": 0.2879123195278744,
+      "grad_norm": 0.24226675927639008,
+      "learning_rate": 4.915296779950926e-06,
+      "loss": 0.6319,
+      "step": 683
+    },
+    {
+      "epoch": 0.28833386025924757,
+      "grad_norm": 0.22968819737434387,
+      "learning_rate": 4.915007612821436e-06,
+      "loss": 0.656,
+      "step": 684
+    },
+    {
+      "epoch": 0.2887554009906207,
+      "grad_norm": 0.23501384258270264,
+      "learning_rate": 4.914717961476006e-06,
+      "loss": 0.6303,
+      "step": 685
+    },
+    {
+      "epoch": 0.2891769417219939,
+      "grad_norm": 0.24456937611103058,
+      "learning_rate": 4.9144278259727135e-06,
+      "loss": 0.6654,
+      "step": 686
+    },
+    {
+      "epoch": 0.28959848245336706,
+      "grad_norm": 0.25023457407951355,
+      "learning_rate": 4.9141372063697305e-06,
+      "loss": 0.6341,
+      "step": 687
+    },
+    {
+      "epoch": 0.29002002318474024,
+      "grad_norm": 0.23992198705673218,
+      "learning_rate": 4.913846102725326e-06,
+      "loss": 0.6535,
+      "step": 688
+    },
+    {
+      "epoch": 0.2904415639161134,
+      "grad_norm": 0.2526721656322479,
+      "learning_rate": 4.913554515097868e-06,
+      "loss": 0.6284,
+      "step": 689
+    },
+    {
+      "epoch": 0.29086310464748655,
+      "grad_norm": 0.24390150606632233,
+      "learning_rate": 4.913262443545821e-06,
+      "loss": 0.6087,
+      "step": 690
+    },
+    {
+      "epoch": 0.29128464537885973,
+      "grad_norm": 0.23285752534866333,
+      "learning_rate": 4.912969888127746e-06,
+      "loss": 0.6377,
+      "step": 691
+    },
+    {
+      "epoch": 0.2917061861102329,
+      "grad_norm": 0.23910963535308838,
+      "learning_rate": 4.912676848902301e-06,
+      "loss": 0.639,
+      "step": 692
+    },
+    {
+      "epoch": 0.2921277268416061,
+      "grad_norm": 0.24280737340450287,
+      "learning_rate": 4.912383325928242e-06,
+      "loss": 0.6493,
+      "step": 693
+    },
+    {
+      "epoch": 0.2925492675729792,
+      "grad_norm": 0.25216546654701233,
+      "learning_rate": 4.912089319264421e-06,
+      "loss": 0.626,
+      "step": 694
+    },
+    {
+      "epoch": 0.2929708083043524,
+      "grad_norm": 0.2513390779495239,
+      "learning_rate": 4.9117948289697845e-06,
+      "loss": 0.658,
+      "step": 695
+    },
+    {
+      "epoch": 0.2933923490357256,
+      "grad_norm": 0.25428324937820435,
+      "learning_rate": 4.911499855103382e-06,
+      "loss": 0.6141,
+      "step": 696
+    },
+    {
+      "epoch": 0.29381388976709877,
+      "grad_norm": 0.24803313612937927,
+      "learning_rate": 4.911204397724357e-06,
+      "loss": 0.6153,
+      "step": 697
+    },
+    {
+      "epoch": 0.2942354304984719,
+      "grad_norm": 0.25573164224624634,
+      "learning_rate": 4.910908456891946e-06,
+      "loss": 0.6282,
+      "step": 698
+    },
+    {
+      "epoch": 0.2946569712298451,
+      "grad_norm": 0.24843843281269073,
+      "learning_rate": 4.910612032665488e-06,
+      "loss": 0.6163,
+      "step": 699
+    },
+    {
+      "epoch": 0.29507851196121826,
+      "grad_norm": 0.2477579414844513,
+      "learning_rate": 4.910315125104418e-06,
+      "loss": 0.6404,
+      "step": 700
+    },
+    {
+      "epoch": 0.29550005269259144,
+      "grad_norm": 0.24367770552635193,
+      "learning_rate": 4.910017734268264e-06,
+      "loss": 0.6545,
+      "step": 701
+    },
+    {
+      "epoch": 0.29592159342396457,
+      "grad_norm": 0.2565951347351074,
+      "learning_rate": 4.909719860216658e-06,
+      "loss": 0.6398,
+      "step": 702
+    },
+    {
+      "epoch": 0.29634313415533775,
+      "grad_norm": 0.2409091293811798,
+      "learning_rate": 4.90942150300932e-06,
+      "loss": 0.6171,
+      "step": 703
+    },
+    {
+      "epoch": 0.29676467488671093,
+      "grad_norm": 0.26979008316993713,
+      "learning_rate": 4.909122662706074e-06,
+      "loss": 0.6314,
+      "step": 704
+    },
+    {
+      "epoch": 0.2971862156180841,
+      "grad_norm": 0.24505124986171722,
+      "learning_rate": 4.908823339366838e-06,
+      "loss": 0.6547,
+      "step": 705
+    },
+    {
+      "epoch": 0.29760775634945724,
+      "grad_norm": 0.24928990006446838,
+      "learning_rate": 4.908523533051627e-06,
+      "loss": 0.6214,
+      "step": 706
+    },
+    {
+      "epoch": 0.2980292970808304,
+      "grad_norm": 0.2416117638349533,
+      "learning_rate": 4.908223243820553e-06,
+      "loss": 0.6086,
+      "step": 707
+    },
+    {
+      "epoch": 0.2984508378122036,
+      "grad_norm": 0.259198933839798,
+      "learning_rate": 4.907922471733824e-06,
+      "loss": 0.6152,
+      "step": 708
+    },
+    {
+      "epoch": 0.2988723785435768,
+      "grad_norm": 0.24309048056602478,
+      "learning_rate": 4.907621216851748e-06,
+      "loss": 0.6251,
+      "step": 709
+    },
+    {
+      "epoch": 0.2992939192749499,
+      "grad_norm": 0.24457012116909027,
+      "learning_rate": 4.907319479234726e-06,
+      "loss": 0.6515,
+      "step": 710
+    },
+    {
+      "epoch": 0.2997154600063231,
+      "grad_norm": 0.241489976644516,
+      "learning_rate": 4.907017258943256e-06,
+      "loss": 0.6434,
+      "step": 711
+    },
+    {
+      "epoch": 0.3001370007376963,
+      "grad_norm": 0.2522222399711609,
+      "learning_rate": 4.906714556037936e-06,
+      "loss": 0.6491,
+      "step": 712
+    },
+    {
+      "epoch": 0.30055854146906946,
+      "grad_norm": 0.25545766949653625,
+      "learning_rate": 4.906411370579458e-06,
+      "loss": 0.6328,
+      "step": 713
+    },
+    {
+      "epoch": 0.30098008220044264,
+      "grad_norm": 0.25071826577186584,
+      "learning_rate": 4.906107702628612e-06,
+      "loss": 0.6062,
+      "step": 714
+    },
+    {
+      "epoch": 0.30140162293181577,
+      "grad_norm": 0.23720961809158325,
+      "learning_rate": 4.905803552246284e-06,
+      "loss": 0.6551,
+      "step": 715
+    },
+    {
+      "epoch": 0.30182316366318895,
+      "grad_norm": 0.2490943819284439,
+      "learning_rate": 4.905498919493457e-06,
+      "loss": 0.6455,
+      "step": 716
+    },
+    {
+      "epoch": 0.30224470439456214,
+      "grad_norm": 0.24237141013145447,
+      "learning_rate": 4.90519380443121e-06,
+      "loss": 0.6218,
+      "step": 717
+    },
+    {
+      "epoch": 0.3026662451259353,
+      "grad_norm": 0.22884462773799896,
+      "learning_rate": 4.90488820712072e-06,
+      "loss": 0.6464,
+      "step": 718
+    },
+    {
+      "epoch": 0.30308778585730844,
+      "grad_norm": 0.23946937918663025,
+      "learning_rate": 4.904582127623262e-06,
+      "loss": 0.6151,
+      "step": 719
+    },
+    {
+      "epoch": 0.3035093265886816,
+      "grad_norm": 0.2361740618944168,
+      "learning_rate": 4.904275566000202e-06,
+      "loss": 0.6212,
+      "step": 720
+    },
+    {
+      "epoch": 0.3039308673200548,
+      "grad_norm": 0.24641628563404083,
+      "learning_rate": 4.90396852231301e-06,
+      "loss": 0.6388,
+      "step": 721
+    },
+    {
+      "epoch": 0.304352408051428,
+      "grad_norm": 0.23151759803295135,
+      "learning_rate": 4.903660996623247e-06,
+      "loss": 0.6162,
+      "step": 722
+    },
+    {
+      "epoch": 0.3047739487828011,
+      "grad_norm": 0.23785744607448578,
+      "learning_rate": 4.903352988992574e-06,
+      "loss": 0.6513,
+      "step": 723
+    },
+    {
+      "epoch": 0.3051954895141743,
+      "grad_norm": 0.24509042501449585,
+      "learning_rate": 4.903044499482747e-06,
+      "loss": 0.6387,
+      "step": 724
+    },
+    {
+      "epoch": 0.3056170302455475,
+      "grad_norm": 0.2480020523071289,
+      "learning_rate": 4.902735528155619e-06,
+      "loss": 0.6377,
+      "step": 725
+    },
+    {
+      "epoch": 0.30603857097692067,
+      "grad_norm": 0.2569882273674011,
+      "learning_rate": 4.902426075073139e-06,
+      "loss": 0.6641,
+      "step": 726
+    },
+    {
+      "epoch": 0.3064601117082938,
+      "grad_norm": 0.24334286153316498,
+      "learning_rate": 4.902116140297354e-06,
+      "loss": 0.639,
+      "step": 727
+    },
+    {
+      "epoch": 0.306881652439667,
+      "grad_norm": 0.241581529378891,
+      "learning_rate": 4.901805723890407e-06,
+      "loss": 0.6092,
+      "step": 728
+    },
+    {
+      "epoch": 0.30730319317104016,
+      "grad_norm": 0.23744048178195953,
+      "learning_rate": 4.901494825914537e-06,
+      "loss": 0.6537,
+      "step": 729
+    },
+    {
+      "epoch": 0.30772473390241334,
+      "grad_norm": 0.24378854036331177,
+      "learning_rate": 4.901183446432079e-06,
+      "loss": 0.6003,
+      "step": 730
+    },
+    {
+      "epoch": 0.30814627463378647,
+      "grad_norm": 0.24870693683624268,
+      "learning_rate": 4.9008715855054674e-06,
+      "loss": 0.6252,
+      "step": 731
+    },
+    {
+      "epoch": 0.30856781536515965,
+      "grad_norm": 0.24152925610542297,
+      "learning_rate": 4.9005592431972306e-06,
+      "loss": 0.6372,
+      "step": 732
+    },
+    {
+      "epoch": 0.30898935609653283,
+      "grad_norm": 0.25459250807762146,
+      "learning_rate": 4.900246419569993e-06,
+      "loss": 0.6274,
+      "step": 733
+    },
+    {
+      "epoch": 0.309410896827906,
+      "grad_norm": 0.24936500191688538,
+      "learning_rate": 4.899933114686477e-06,
+      "loss": 0.6015,
+      "step": 734
+    },
+    {
+      "epoch": 0.30983243755927914,
+      "grad_norm": 0.2448284775018692,
+      "learning_rate": 4.8996193286095015e-06,
+      "loss": 0.611,
+      "step": 735
+    },
+    {
+      "epoch": 0.3102539782906523,
+      "grad_norm": 0.2678229808807373,
+      "learning_rate": 4.899305061401983e-06,
+      "loss": 0.6311,
+      "step": 736
+    },
+    {
+      "epoch": 0.3106755190220255,
+      "grad_norm": 0.26180657744407654,
+      "learning_rate": 4.89899031312693e-06,
+      "loss": 0.6033,
+      "step": 737
+    },
+    {
+      "epoch": 0.3110970597533987,
+      "grad_norm": 0.2456626296043396,
+      "learning_rate": 4.898675083847453e-06,
+      "loss": 0.6353,
+      "step": 738
+    },
+    {
+      "epoch": 0.31151860048477187,
+      "grad_norm": 0.25847429037094116,
+      "learning_rate": 4.898359373626755e-06,
+      "loss": 0.6465,
+      "step": 739
+    },
+    {
+      "epoch": 0.311940141216145,
+      "grad_norm": 0.24212178587913513,
+      "learning_rate": 4.898043182528136e-06,
+      "loss": 0.6338,
+      "step": 740
+    },
+    {
+      "epoch": 0.3123616819475182,
+      "grad_norm": 0.25338494777679443,
+      "learning_rate": 4.897726510614995e-06,
+      "loss": 0.6451,
+      "step": 741
+    },
+    {
+      "epoch": 0.31278322267889136,
+      "grad_norm": 0.25079071521759033,
+      "learning_rate": 4.897409357950826e-06,
+      "loss": 0.6608,
+      "step": 742
+    },
+    {
+      "epoch": 0.31320476341026454,
+      "grad_norm": 0.26444000005722046,
+      "learning_rate": 4.897091724599217e-06,
+      "loss": 0.6153,
+      "step": 743
+    },
+    {
+      "epoch": 0.31362630414163767,
+      "grad_norm": 0.24816632270812988,
+      "learning_rate": 4.896773610623855e-06,
+      "loss": 0.6135,
+      "step": 744
+    },
+    {
+      "epoch": 0.31404784487301085,
+      "grad_norm": 0.24375613033771515,
+      "learning_rate": 4.896455016088524e-06,
+      "loss": 0.6487,
+      "step": 745
+    },
+    {
+      "epoch": 0.31446938560438403,
+      "grad_norm": 0.24404917657375336,
+      "learning_rate": 4.896135941057102e-06,
+      "loss": 0.6519,
+      "step": 746
+    },
+    {
+      "epoch": 0.3148909263357572,
+      "grad_norm": 0.24915026128292084,
+      "learning_rate": 4.8958163855935645e-06,
+      "loss": 0.623,
+      "step": 747
+    },
+    {
+      "epoch": 0.31531246706713034,
+      "grad_norm": 0.25012120604515076,
+      "learning_rate": 4.895496349761983e-06,
+      "loss": 0.6493,
+      "step": 748
+    },
+    {
+      "epoch": 0.3157340077985035,
+      "grad_norm": 0.250488817691803,
+      "learning_rate": 4.895175833626527e-06,
+      "loss": 0.6508,
+      "step": 749
+    },
+    {
+      "epoch": 0.3161555485298767,
+      "grad_norm": 0.2555750012397766,
+      "learning_rate": 4.894854837251459e-06,
+      "loss": 0.6038,
+      "step": 750
+    },
+    {
+      "epoch": 0.3165770892612499,
+      "grad_norm": 0.24569761753082275,
+      "learning_rate": 4.894533360701141e-06,
+      "loss": 0.625,
+      "step": 751
+    },
+    {
+      "epoch": 0.316998629992623,
+      "grad_norm": 0.25181856751441956,
+      "learning_rate": 4.89421140404003e-06,
+      "loss": 0.6228,
+      "step": 752
+    },
+    {
+      "epoch": 0.3174201707239962,
+      "grad_norm": 0.2686469256877899,
+      "learning_rate": 4.893888967332679e-06,
+      "loss": 0.6522,
+      "step": 753
+    },
+    {
+      "epoch": 0.3178417114553694,
+      "grad_norm": 0.24826131761074066,
+      "learning_rate": 4.893566050643738e-06,
+      "loss": 0.6144,
+      "step": 754
+    },
+    {
+      "epoch": 0.31826325218674256,
+      "grad_norm": 0.24306431412696838,
+      "learning_rate": 4.89324265403795e-06,
+      "loss": 0.6756,
+      "step": 755
+    },
+    {
+      "epoch": 0.3186847929181157,
+      "grad_norm": 0.24478738009929657,
+      "learning_rate": 4.892918777580161e-06,
+      "loss": 0.654,
+      "step": 756
+    },
+    {
+      "epoch": 0.31910633364948887,
+      "grad_norm": 0.2635796070098877,
+      "learning_rate": 4.892594421335306e-06,
+      "loss": 0.6443,
+      "step": 757
+    },
+    {
+      "epoch": 0.31952787438086205,
+      "grad_norm": 0.2445783019065857,
+      "learning_rate": 4.892269585368421e-06,
+      "loss": 0.6463,
+      "step": 758
+    },
+    {
+      "epoch": 0.31994941511223524,
+      "grad_norm": 0.2572665512561798,
+      "learning_rate": 4.891944269744636e-06,
+      "loss": 0.6554,
+      "step": 759
+    },
+    {
+      "epoch": 0.32037095584360836,
+      "grad_norm": 0.2534368634223938,
+      "learning_rate": 4.8916184745291785e-06,
+      "loss": 0.6257,
+      "step": 760
+    },
+    {
+      "epoch": 0.32079249657498154,
+      "grad_norm": 0.2623680531978607,
+      "learning_rate": 4.89129219978737e-06,
+      "loss": 0.6369,
+      "step": 761
+    },
+    {
+      "epoch": 0.3212140373063547,
+      "grad_norm": 0.28523921966552734,
+      "learning_rate": 4.89096544558463e-06,
+      "loss": 0.6242,
+      "step": 762
+    },
+    {
+      "epoch": 0.3216355780377279,
+      "grad_norm": 0.26124414801597595,
+      "learning_rate": 4.890638211986474e-06,
+      "loss": 0.6406,
+      "step": 763
+    },
+    {
+      "epoch": 0.3220571187691011,
+      "grad_norm": 0.26042789220809937,
+      "learning_rate": 4.890310499058513e-06,
+      "loss": 0.629,
+      "step": 764
+    },
+    {
+      "epoch": 0.3224786595004742,
+      "grad_norm": 0.27142202854156494,
+      "learning_rate": 4.889982306866454e-06,
+      "loss": 0.6162,
+      "step": 765
+    },
+    {
+      "epoch": 0.3229002002318474,
+      "grad_norm": 0.25213897228240967,
+      "learning_rate": 4.889653635476102e-06,
+      "loss": 0.6265,
+      "step": 766
+    },
+    {
+      "epoch": 0.3233217409632206,
+      "grad_norm": 0.2568320333957672,
+      "learning_rate": 4.889324484953355e-06,
+      "loss": 0.6362,
+      "step": 767
+    },
+    {
+      "epoch": 0.32374328169459377,
+      "grad_norm": 0.2552599012851715,
+      "learning_rate": 4.8889948553642095e-06,
+      "loss": 0.6526,
+      "step": 768
+    },
+    {
+      "epoch": 0.3241648224259669,
+      "grad_norm": 0.25189757347106934,
+      "learning_rate": 4.888664746774756e-06,
+      "loss": 0.6356,
+      "step": 769
+    },
+    {
+      "epoch": 0.3245863631573401,
+      "grad_norm": 0.24278956651687622,
+      "learning_rate": 4.8883341592511844e-06,
+      "loss": 0.6367,
+      "step": 770
+    },
+    {
+      "epoch": 0.32500790388871326,
+      "grad_norm": 0.2523997724056244,
+      "learning_rate": 4.888003092859776e-06,
+      "loss": 0.6454,
+      "step": 771
+    },
+    {
+      "epoch": 0.32542944462008644,
+      "grad_norm": 0.2542168200016022,
+      "learning_rate": 4.887671547666912e-06,
+      "loss": 0.6168,
+      "step": 772
+    },
+    {
+      "epoch": 0.32585098535145957,
+      "grad_norm": 0.2451404333114624,
+      "learning_rate": 4.887339523739068e-06,
+      "loss": 0.6295,
+      "step": 773
+    },
+    {
+      "epoch": 0.32627252608283275,
+      "grad_norm": 0.2537959814071655,
+      "learning_rate": 4.887007021142815e-06,
+      "loss": 0.6039,
+      "step": 774
+    },
+    {
+      "epoch": 0.32669406681420593,
+      "grad_norm": 0.2613690197467804,
+      "learning_rate": 4.886674039944822e-06,
+      "loss": 0.6246,
+      "step": 775
+    },
+    {
+      "epoch": 0.3271156075455791,
+      "grad_norm": 0.24343545734882355,
+      "learning_rate": 4.886340580211851e-06,
+      "loss": 0.6165,
+      "step": 776
+    },
+    {
+      "epoch": 0.32753714827695224,
+      "grad_norm": 0.2493949979543686,
+      "learning_rate": 4.886006642010763e-06,
+      "loss": 0.6433,
+      "step": 777
+    },
+    {
+      "epoch": 0.3279586890083254,
+      "grad_norm": 0.24980653822422028,
+      "learning_rate": 4.885672225408513e-06,
+      "loss": 0.6134,
+      "step": 778
+    },
+    {
+      "epoch": 0.3283802297396986,
+      "grad_norm": 0.24947059154510498,
+      "learning_rate": 4.885337330472153e-06,
+      "loss": 0.6178,
+      "step": 779
+    },
+    {
+      "epoch": 0.3288017704710718,
+      "grad_norm": 0.2452121376991272,
+      "learning_rate": 4.8850019572688305e-06,
+      "loss": 0.626,
+      "step": 780
+    },
+    {
+      "epoch": 0.3292233112024449,
+      "grad_norm": 0.2447056919336319,
+      "learning_rate": 4.884666105865787e-06,
+      "loss": 0.6363,
+      "step": 781
+    },
+    {
+      "epoch": 0.3296448519338181,
+      "grad_norm": 0.2492496818304062,
+      "learning_rate": 4.8843297763303644e-06,
+      "loss": 0.6582,
+      "step": 782
+    },
+    {
+      "epoch": 0.3300663926651913,
+      "grad_norm": 0.2591716945171356,
+      "learning_rate": 4.883992968729996e-06,
+      "loss": 0.5915,
+      "step": 783
+    },
+    {
+      "epoch": 0.33048793339656446,
+      "grad_norm": 0.24453039467334747,
+      "learning_rate": 4.883655683132212e-06,
+      "loss": 0.6341,
+      "step": 784
+    },
+    {
+      "epoch": 0.3309094741279376,
+      "grad_norm": 0.2636326849460602,
+      "learning_rate": 4.883317919604641e-06,
+      "loss": 0.6385,
+      "step": 785
+    },
+    {
+      "epoch": 0.33133101485931077,
+      "grad_norm": 0.24589572846889496,
+      "learning_rate": 4.882979678215004e-06,
+      "loss": 0.6167,
+      "step": 786
+    },
+    {
+      "epoch": 0.33175255559068395,
+      "grad_norm": 0.26419368386268616,
+      "learning_rate": 4.8826409590311205e-06,
+      "loss": 0.6182,
+      "step": 787
+    },
+    {
+      "epoch": 0.33217409632205713,
+      "grad_norm": 0.2559862434864044,
+      "learning_rate": 4.882301762120905e-06,
+      "loss": 0.6363,
+      "step": 788
+    },
+    {
+      "epoch": 0.3325956370534303,
+      "grad_norm": 0.25816717743873596,
+      "learning_rate": 4.881962087552366e-06,
+      "loss": 0.6167,
+      "step": 789
+    },
+    {
+      "epoch": 0.33301717778480344,
+      "grad_norm": 0.2528778314590454,
+      "learning_rate": 4.88162193539361e-06,
+      "loss": 0.6355,
+      "step": 790
+    },
+    {
+      "epoch": 0.3334387185161766,
+      "grad_norm": 0.2598842680454254,
+      "learning_rate": 4.881281305712838e-06,
+      "loss": 0.6468,
+      "step": 791
+    },
+    {
+      "epoch": 0.3338602592475498,
+      "grad_norm": 0.2565869092941284,
+      "learning_rate": 4.880940198578348e-06,
+      "loss": 0.6284,
+      "step": 792
+    },
+    {
+      "epoch": 0.334281799978923,
+      "grad_norm": 0.2511397898197174,
+      "learning_rate": 4.880598614058531e-06,
+      "loss": 0.6279,
+      "step": 793
+    },
+    {
+      "epoch": 0.3347033407102961,
+      "grad_norm": 0.2486419975757599,
+      "learning_rate": 4.880256552221879e-06,
+      "loss": 0.6546,
+      "step": 794
+    },
+    {
+      "epoch": 0.3351248814416693,
+      "grad_norm": 0.25715431571006775,
+      "learning_rate": 4.879914013136975e-06,
+      "loss": 0.6235,
+      "step": 795
+    },
+    {
+      "epoch": 0.3355464221730425,
+      "grad_norm": 0.2668071985244751,
+      "learning_rate": 4.879570996872497e-06,
+      "loss": 0.6425,
+      "step": 796
+    },
+    {
+      "epoch": 0.33596796290441566,
+      "grad_norm": 0.25566986203193665,
+      "learning_rate": 4.879227503497224e-06,
+      "loss": 0.6031,
+      "step": 797
+    },
+    {
+      "epoch": 0.3363895036357888,
+      "grad_norm": 0.2451721429824829,
+      "learning_rate": 4.878883533080025e-06,
+      "loss": 0.6505,
+      "step": 798
+    },
+    {
+      "epoch": 0.33681104436716197,
+      "grad_norm": 0.2558004856109619,
+      "learning_rate": 4.878539085689869e-06,
+      "loss": 0.6288,
+      "step": 799
+    },
+    {
+      "epoch": 0.33723258509853515,
+      "grad_norm": 0.2570480406284332,
+      "learning_rate": 4.878194161395816e-06,
+      "loss": 0.6425,
+      "step": 800
+    },
+    {
+      "epoch": 0.33765412582990834,
+      "grad_norm": 0.25120779871940613,
+      "learning_rate": 4.877848760267026e-06,
+      "loss": 0.6341,
+      "step": 801
+    },
+    {
+      "epoch": 0.33807566656128146,
+      "grad_norm": 0.25851771235466003,
+      "learning_rate": 4.877502882372753e-06,
+      "loss": 0.6311,
+      "step": 802
+    },
+    {
+      "epoch": 0.33849720729265464,
+      "grad_norm": 0.2400793433189392,
+      "learning_rate": 4.877156527782348e-06,
+      "loss": 0.6388,
+      "step": 803
+    },
+    {
+      "epoch": 0.3389187480240278,
+      "grad_norm": 0.24545061588287354,
+      "learning_rate": 4.876809696565252e-06,
+      "loss": 0.646,
+      "step": 804
+    },
+    {
+      "epoch": 0.339340288755401,
+      "grad_norm": 0.25027358531951904,
+      "learning_rate": 4.876462388791009e-06,
+      "loss": 0.6561,
+      "step": 805
+    },
+    {
+      "epoch": 0.33976182948677414,
+      "grad_norm": 0.255466103553772,
+      "learning_rate": 4.876114604529254e-06,
+      "loss": 0.6484,
+      "step": 806
+    },
+    {
+      "epoch": 0.3401833702181473,
+      "grad_norm": 0.25834470987319946,
+      "learning_rate": 4.875766343849719e-06,
+      "loss": 0.6465,
+      "step": 807
+    },
+    {
+      "epoch": 0.3406049109495205,
+      "grad_norm": 0.27207326889038086,
+      "learning_rate": 4.875417606822232e-06,
+      "loss": 0.628,
+      "step": 808
+    },
+    {
+      "epoch": 0.3410264516808937,
+      "grad_norm": 0.2536725401878357,
+      "learning_rate": 4.875068393516714e-06,
+      "loss": 0.6439,
+      "step": 809
+    },
+    {
+      "epoch": 0.3414479924122668,
+      "grad_norm": 0.2758451998233795,
+      "learning_rate": 4.874718704003185e-06,
+      "loss": 0.6498,
+      "step": 810
+    },
+    {
+      "epoch": 0.34186953314364,
+      "grad_norm": 0.26022541522979736,
+      "learning_rate": 4.874368538351758e-06,
+      "loss": 0.6394,
+      "step": 811
+    },
+    {
+      "epoch": 0.3422910738750132,
+      "grad_norm": 0.26493576169013977,
+      "learning_rate": 4.874017896632642e-06,
+      "loss": 0.6342,
+      "step": 812
+    },
+    {
+      "epoch": 0.34271261460638636,
+      "grad_norm": 0.2755403518676758,
+      "learning_rate": 4.873666778916142e-06,
+      "loss": 0.6193,
+      "step": 813
+    },
+    {
+      "epoch": 0.3431341553377595,
+      "grad_norm": 0.2545062303543091,
+      "learning_rate": 4.873315185272658e-06,
+      "loss": 0.6154,
+      "step": 814
+    },
+    {
+      "epoch": 0.34355569606913267,
+      "grad_norm": 0.26998916268348694,
+      "learning_rate": 4.872963115772686e-06,
+      "loss": 0.6476,
+      "step": 815
+    },
+    {
+      "epoch": 0.34397723680050585,
+      "grad_norm": 0.25837239623069763,
+      "learning_rate": 4.872610570486816e-06,
+      "loss": 0.6146,
+      "step": 816
+    },
+    {
+      "epoch": 0.34439877753187903,
+      "grad_norm": 0.2706039845943451,
+      "learning_rate": 4.872257549485735e-06,
+      "loss": 0.6183,
+      "step": 817
+    },
+    {
+      "epoch": 0.3448203182632522,
+      "grad_norm": 0.25320756435394287,
+      "learning_rate": 4.871904052840224e-06,
+      "loss": 0.6373,
+      "step": 818
+    },
+    {
+      "epoch": 0.34524185899462534,
+      "grad_norm": 0.2520752549171448,
+      "learning_rate": 4.871550080621162e-06,
+      "loss": 0.6214,
+      "step": 819
+    },
+    {
+      "epoch": 0.3456633997259985,
+      "grad_norm": 0.2538292407989502,
+      "learning_rate": 4.871195632899519e-06,
+      "loss": 0.6273,
+      "step": 820
+    },
+    {
+      "epoch": 0.3460849404573717,
+      "grad_norm": 0.2511141300201416,
+      "learning_rate": 4.870840709746364e-06,
+      "loss": 0.6395,
+      "step": 821
+    },
+    {
+      "epoch": 0.3465064811887449,
+      "grad_norm": 0.2537423372268677,
+      "learning_rate": 4.8704853112328595e-06,
+      "loss": 0.6459,
+      "step": 822
+    },
+    {
+      "epoch": 0.346928021920118,
+      "grad_norm": 0.24753199517726898,
+      "learning_rate": 4.870129437430266e-06,
+      "loss": 0.6243,
+      "step": 823
+    },
+    {
+      "epoch": 0.3473495626514912,
+      "grad_norm": 0.25171908736228943,
+      "learning_rate": 4.869773088409934e-06,
+      "loss": 0.6251,
+      "step": 824
+    },
+    {
+      "epoch": 0.3477711033828644,
+      "grad_norm": 0.2594963014125824,
+      "learning_rate": 4.8694162642433145e-06,
+      "loss": 0.6305,
+      "step": 825
+    },
+    {
+      "epoch": 0.34819264411423756,
+      "grad_norm": 0.24246972799301147,
+      "learning_rate": 4.869058965001951e-06,
+      "loss": 0.6108,
+      "step": 826
+    },
+    {
+      "epoch": 0.3486141848456107,
+      "grad_norm": 0.2587549090385437,
+      "learning_rate": 4.8687011907574846e-06,
+      "loss": 0.6291,
+      "step": 827
+    },
+    {
+      "epoch": 0.34903572557698387,
+      "grad_norm": 0.24805046617984772,
+      "learning_rate": 4.8683429415816485e-06,
+      "loss": 0.6091,
+      "step": 828
+    },
+    {
+      "epoch": 0.34945726630835705,
+      "grad_norm": 0.25509005784988403,
+      "learning_rate": 4.867984217546272e-06,
+      "loss": 0.6467,
+      "step": 829
+    },
+    {
+      "epoch": 0.34987880703973023,
+      "grad_norm": 0.2585395574569702,
+      "learning_rate": 4.867625018723283e-06,
+      "loss": 0.6102,
+      "step": 830
+    },
+    {
+      "epoch": 0.35030034777110336,
+      "grad_norm": 0.24894072115421295,
+      "learning_rate": 4.8672653451847e-06,
+      "loss": 0.6445,
+      "step": 831
+    },
+    {
+      "epoch": 0.35072188850247654,
+      "grad_norm": 0.2756640613079071,
+      "learning_rate": 4.866905197002637e-06,
+      "loss": 0.6197,
+      "step": 832
+    },
+    {
+      "epoch": 0.3511434292338497,
+      "grad_norm": 0.26490357518196106,
+      "learning_rate": 4.866544574249308e-06,
+      "loss": 0.6191,
+      "step": 833
+    },
+    {
+      "epoch": 0.3515649699652229,
+      "grad_norm": 0.24230548739433289,
+      "learning_rate": 4.866183476997017e-06,
+      "loss": 0.6319,
+      "step": 834
+    },
+    {
+      "epoch": 0.35198651069659603,
+      "grad_norm": 0.2637353241443634,
+      "learning_rate": 4.865821905318167e-06,
+      "loss": 0.6571,
+      "step": 835
+    },
+    {
+      "epoch": 0.3524080514279692,
+      "grad_norm": 0.2556266486644745,
+      "learning_rate": 4.865459859285251e-06,
+      "loss": 0.6432,
+      "step": 836
+    },
+    {
+      "epoch": 0.3528295921593424,
+      "grad_norm": 0.25549042224884033,
+      "learning_rate": 4.865097338970862e-06,
+      "loss": 0.6322,
+      "step": 837
+    },
+    {
+      "epoch": 0.3532511328907156,
+      "grad_norm": 0.2747564911842346,
+      "learning_rate": 4.864734344447686e-06,
+      "loss": 0.6418,
+      "step": 838
+    },
+    {
+      "epoch": 0.3536726736220887,
+      "grad_norm": 0.2574198544025421,
+      "learning_rate": 4.864370875788505e-06,
+      "loss": 0.6198,
+      "step": 839
+    },
+    {
+      "epoch": 0.3540942143534619,
+      "grad_norm": 0.25986191630363464,
+      "learning_rate": 4.864006933066196e-06,
+      "loss": 0.6157,
+      "step": 840
+    },
+    {
+      "epoch": 0.35451575508483507,
+      "grad_norm": 0.27981826663017273,
+      "learning_rate": 4.863642516353729e-06,
+      "loss": 0.6505,
+      "step": 841
+    },
+    {
+      "epoch": 0.35493729581620825,
+      "grad_norm": 0.2704963982105255,
+      "learning_rate": 4.863277625724172e-06,
+      "loss": 0.6485,
+      "step": 842
+    },
+    {
+      "epoch": 0.35535883654758144,
+      "grad_norm": 0.27085819840431213,
+      "learning_rate": 4.862912261250685e-06,
+      "loss": 0.6521,
+      "step": 843
+    },
+    {
+      "epoch": 0.35578037727895456,
+      "grad_norm": 0.2682935297489166,
+      "learning_rate": 4.862546423006526e-06,
+      "loss": 0.6403,
+      "step": 844
+    },
+    {
+      "epoch": 0.35620191801032774,
+      "grad_norm": 0.2632185220718384,
+      "learning_rate": 4.862180111065047e-06,
+      "loss": 0.6316,
+      "step": 845
+    },
+    {
+      "epoch": 0.3566234587417009,
+      "grad_norm": 0.25609323382377625,
+      "learning_rate": 4.861813325499693e-06,
+      "loss": 0.6518,
+      "step": 846
+    },
+    {
+      "epoch": 0.3570449994730741,
+      "grad_norm": 0.25910910964012146,
+      "learning_rate": 4.861446066384007e-06,
+      "loss": 0.6625,
+      "step": 847
+    },
+    {
+      "epoch": 0.35746654020444724,
+      "grad_norm": 0.2763497233390808,
+      "learning_rate": 4.861078333791624e-06,
+      "loss": 0.6157,
+      "step": 848
+    },
+    {
+      "epoch": 0.3578880809358204,
+      "grad_norm": 0.2565842866897583,
+      "learning_rate": 4.860710127796276e-06,
+      "loss": 0.6446,
+      "step": 849
+    },
+    {
+      "epoch": 0.3583096216671936,
+      "grad_norm": 0.26684844493865967,
+      "learning_rate": 4.860341448471791e-06,
+      "loss": 0.6393,
+      "step": 850
+    },
+    {
+      "epoch": 0.3587311623985668,
+      "grad_norm": 0.25313448905944824,
+      "learning_rate": 4.859972295892088e-06,
+      "loss": 0.6541,
+      "step": 851
+    },
+    {
+      "epoch": 0.3591527031299399,
+      "grad_norm": 0.2515822649002075,
+      "learning_rate": 4.859602670131185e-06,
+      "loss": 0.653,
+      "step": 852
+    },
+    {
+      "epoch": 0.3595742438613131,
+      "grad_norm": 0.2599954605102539,
+      "learning_rate": 4.8592325712631914e-06,
+      "loss": 0.5938,
+      "step": 853
+    },
+    {
+      "epoch": 0.3599957845926863,
+      "grad_norm": 0.27787014842033386,
+      "learning_rate": 4.858861999362313e-06,
+      "loss": 0.6227,
+      "step": 854
+    },
+    {
+      "epoch": 0.36041732532405946,
+      "grad_norm": 0.2670431137084961,
+      "learning_rate": 4.858490954502852e-06,
+      "loss": 0.6296,
+      "step": 855
+    },
+    {
+      "epoch": 0.3608388660554326,
+      "grad_norm": 0.25765085220336914,
+      "learning_rate": 4.858119436759204e-06,
+      "loss": 0.6446,
+      "step": 856
+    },
+    {
+      "epoch": 0.36126040678680577,
+      "grad_norm": 0.2589804530143738,
+      "learning_rate": 4.857747446205858e-06,
+      "loss": 0.6274,
+      "step": 857
+    },
+    {
+      "epoch": 0.36168194751817895,
+      "grad_norm": 0.2487720400094986,
+      "learning_rate": 4.8573749829174e-06,
+      "loss": 0.6248,
+      "step": 858
+    },
+    {
+      "epoch": 0.36210348824955213,
+      "grad_norm": 0.2504652142524719,
+      "learning_rate": 4.85700204696851e-06,
+      "loss": 0.6076,
+      "step": 859
+    },
+    {
+      "epoch": 0.36252502898092526,
+      "grad_norm": 0.24419501423835754,
+      "learning_rate": 4.856628638433962e-06,
+      "loss": 0.6378,
+      "step": 860
+    },
+    {
+      "epoch": 0.36294656971229844,
+      "grad_norm": 0.259529173374176,
+      "learning_rate": 4.856254757388626e-06,
+      "loss": 0.6306,
+      "step": 861
+    },
+    {
+      "epoch": 0.3633681104436716,
+      "grad_norm": 0.26039767265319824,
+      "learning_rate": 4.855880403907466e-06,
+      "loss": 0.6507,
+      "step": 862
+    },
+    {
+      "epoch": 0.3637896511750448,
+      "grad_norm": 0.23905687034130096,
+      "learning_rate": 4.855505578065542e-06,
+      "loss": 0.6185,
+      "step": 863
+    },
+    {
+      "epoch": 0.36421119190641793,
+      "grad_norm": 0.25929901003837585,
+      "learning_rate": 4.855130279938006e-06,
+      "loss": 0.6156,
+      "step": 864
+    },
+    {
+      "epoch": 0.3646327326377911,
+      "grad_norm": 0.24727587401866913,
+      "learning_rate": 4.854754509600107e-06,
+      "loss": 0.6191,
+      "step": 865
+    },
+    {
+      "epoch": 0.3650542733691643,
+      "grad_norm": 0.2667529582977295,
+      "learning_rate": 4.8543782671271885e-06,
+      "loss": 0.6253,
+      "step": 866
+    },
+    {
+      "epoch": 0.3654758141005375,
+      "grad_norm": 0.25281596183776855,
+      "learning_rate": 4.8540015525946884e-06,
+      "loss": 0.6449,
+      "step": 867
+    },
+    {
+      "epoch": 0.36589735483191066,
+      "grad_norm": 0.25489482283592224,
+      "learning_rate": 4.853624366078137e-06,
+      "loss": 0.6571,
+      "step": 868
+    },
+    {
+      "epoch": 0.3663188955632838,
+      "grad_norm": 0.2643046975135803,
+      "learning_rate": 4.8532467076531646e-06,
+      "loss": 0.645,
+      "step": 869
+    },
+    {
+      "epoch": 0.36674043629465697,
+      "grad_norm": 0.24946551024913788,
+      "learning_rate": 4.852868577395489e-06,
+      "loss": 0.6307,
+      "step": 870
+    },
+    {
+      "epoch": 0.36716197702603015,
+      "grad_norm": 0.2534451484680176,
+      "learning_rate": 4.85248997538093e-06,
+      "loss": 0.6357,
+      "step": 871
+    },
+    {
+      "epoch": 0.36758351775740333,
+      "grad_norm": 0.25324976444244385,
+      "learning_rate": 4.852110901685396e-06,
+      "loss": 0.6223,
+      "step": 872
+    },
+    {
+      "epoch": 0.36800505848877646,
+      "grad_norm": 0.2721559405326843,
+      "learning_rate": 4.851731356384893e-06,
+      "loss": 0.6169,
+      "step": 873
+    },
+    {
+      "epoch": 0.36842659922014964,
+      "grad_norm": 0.253261536359787,
+      "learning_rate": 4.851351339555522e-06,
+      "loss": 0.6312,
+      "step": 874
+    },
+    {
+      "epoch": 0.3688481399515228,
+      "grad_norm": 0.2624780833721161,
+      "learning_rate": 4.850970851273476e-06,
+      "loss": 0.6403,
+      "step": 875
+    },
+    {
+      "epoch": 0.369269680682896,
+      "grad_norm": 0.25246620178222656,
+      "learning_rate": 4.850589891615044e-06,
+      "loss": 0.6373,
+      "step": 876
+    },
+    {
+      "epoch": 0.36969122141426913,
+      "grad_norm": 0.24897249042987823,
+      "learning_rate": 4.85020846065661e-06,
+      "loss": 0.6202,
+      "step": 877
+    },
+    {
+      "epoch": 0.3701127621456423,
+      "grad_norm": 0.2582487165927887,
+      "learning_rate": 4.849826558474653e-06,
+      "loss": 0.6474,
+      "step": 878
+    },
+    {
+      "epoch": 0.3705343028770155,
+      "grad_norm": 0.2585439085960388,
+      "learning_rate": 4.849444185145743e-06,
+      "loss": 0.6001,
+      "step": 879
+    },
+    {
+      "epoch": 0.3709558436083887,
+      "grad_norm": 0.2579975724220276,
+      "learning_rate": 4.849061340746549e-06,
+      "loss": 0.6488,
+      "step": 880
+    },
+    {
+      "epoch": 0.3713773843397618,
+      "grad_norm": 0.24333162605762482,
+      "learning_rate": 4.848678025353832e-06,
+      "loss": 0.618,
+      "step": 881
+    },
+    {
+      "epoch": 0.371798925071135,
+      "grad_norm": 0.297209769487381,
+      "learning_rate": 4.848294239044448e-06,
+      "loss": 0.6101,
+      "step": 882
+    },
+    {
+      "epoch": 0.37222046580250817,
+      "grad_norm": 0.2524620294570923,
+      "learning_rate": 4.847909981895347e-06,
+      "loss": 0.6352,
+      "step": 883
+    },
+    {
+      "epoch": 0.37264200653388135,
+      "grad_norm": 0.26648691296577454,
+      "learning_rate": 4.847525253983573e-06,
+      "loss": 0.6176,
+      "step": 884
+    },
+    {
+      "epoch": 0.3730635472652545,
+      "grad_norm": 0.2886981964111328,
+      "learning_rate": 4.8471400553862655e-06,
+      "loss": 0.6219,
+      "step": 885
+    },
+    {
+      "epoch": 0.37348508799662766,
+      "grad_norm": 0.25700879096984863,
+      "learning_rate": 4.846754386180658e-06,
+      "loss": 0.6268,
+      "step": 886
+    },
+    {
+      "epoch": 0.37390662872800084,
+      "grad_norm": 0.26508307456970215,
+      "learning_rate": 4.846368246444078e-06,
+      "loss": 0.6552,
+      "step": 887
+    },
+    {
+      "epoch": 0.374328169459374,
+      "grad_norm": 0.2605459988117218,
+      "learning_rate": 4.8459816362539495e-06,
+      "loss": 0.6327,
+      "step": 888
+    },
+    {
+      "epoch": 0.37474971019074715,
+      "grad_norm": 0.2567014992237091,
+      "learning_rate": 4.845594555687786e-06,
+      "loss": 0.6208,
+      "step": 889
+    },
+    {
+      "epoch": 0.37517125092212034,
+      "grad_norm": 0.28781476616859436,
+      "learning_rate": 4.845207004823199e-06,
+      "loss": 0.6409,
+      "step": 890
+    },
+    {
+      "epoch": 0.3755927916534935,
+      "grad_norm": 0.24735136330127716,
+      "learning_rate": 4.844818983737896e-06,
+      "loss": 0.6269,
+      "step": 891
+    },
+    {
+      "epoch": 0.3760143323848667,
+      "grad_norm": 0.25130677223205566,
+      "learning_rate": 4.844430492509674e-06,
+      "loss": 0.6312,
+      "step": 892
+    },
+    {
+      "epoch": 0.3764358731162399,
+      "grad_norm": 0.2650367021560669,
+      "learning_rate": 4.844041531216427e-06,
+      "loss": 0.6274,
+      "step": 893
+    },
+    {
+      "epoch": 0.376857413847613,
+      "grad_norm": 0.2622985541820526,
+      "learning_rate": 4.843652099936142e-06,
+      "loss": 0.6257,
+      "step": 894
+    },
+    {
+      "epoch": 0.3772789545789862,
+      "grad_norm": 0.24184496700763702,
+      "learning_rate": 4.843262198746903e-06,
+      "loss": 0.6426,
+      "step": 895
+    },
+    {
+      "epoch": 0.3777004953103594,
+      "grad_norm": 0.2562757134437561,
+      "learning_rate": 4.842871827726886e-06,
+      "loss": 0.6095,
+      "step": 896
+    },
+    {
+      "epoch": 0.37812203604173256,
+      "grad_norm": 0.25449368357658386,
+      "learning_rate": 4.84248098695436e-06,
+      "loss": 0.6024,
+      "step": 897
+    },
+    {
+      "epoch": 0.3785435767731057,
+      "grad_norm": 0.2706413269042969,
+      "learning_rate": 4.84208967650769e-06,
+      "loss": 0.6158,
+      "step": 898
+    },
+    {
+      "epoch": 0.37896511750447887,
+      "grad_norm": 0.27312538027763367,
+      "learning_rate": 4.841697896465337e-06,
+      "loss": 0.618,
+      "step": 899
+    },
+    {
+      "epoch": 0.37938665823585205,
+      "grad_norm": 0.26674240827560425,
+      "learning_rate": 4.841305646905852e-06,
+      "loss": 0.6546,
+      "step": 900
+    },
+    {
+      "epoch": 0.37980819896722523,
+      "grad_norm": 0.2656947076320648,
+      "learning_rate": 4.840912927907882e-06,
+      "loss": 0.6346,
+      "step": 901
+    },
+    {
+      "epoch": 0.38022973969859836,
+      "grad_norm": 0.24790920317173004,
+      "learning_rate": 4.840519739550169e-06,
+      "loss": 0.6328,
+      "step": 902
+    },
+    {
+      "epoch": 0.38065128042997154,
+      "grad_norm": 0.2540440559387207,
+      "learning_rate": 4.8401260819115494e-06,
+      "loss": 0.6491,
+      "step": 903
+    },
+    {
+      "epoch": 0.3810728211613447,
+      "grad_norm": 0.2641192674636841,
+      "learning_rate": 4.8397319550709504e-06,
+      "loss": 0.6378,
+      "step": 904
+    },
+    {
+      "epoch": 0.3814943618927179,
+      "grad_norm": 0.25249913334846497,
+      "learning_rate": 4.839337359107396e-06,
+      "loss": 0.6412,
+      "step": 905
+    },
+    {
+      "epoch": 0.38191590262409103,
+      "grad_norm": 0.25797387957572937,
+      "learning_rate": 4.8389422941000065e-06,
+      "loss": 0.6211,
+      "step": 906
+    },
+    {
+      "epoch": 0.3823374433554642,
+      "grad_norm": 0.24320074915885925,
+      "learning_rate": 4.8385467601279915e-06,
+      "loss": 0.6305,
+      "step": 907
+    },
+    {
+      "epoch": 0.3827589840868374,
+      "grad_norm": 0.24869081377983093,
+      "learning_rate": 4.8381507572706555e-06,
+      "loss": 0.6262,
+      "step": 908
+    },
+    {
+      "epoch": 0.3831805248182106,
+      "grad_norm": 0.2462390661239624,
+      "learning_rate": 4.8377542856074e-06,
+      "loss": 0.6381,
+      "step": 909
+    },
+    {
+      "epoch": 0.3836020655495837,
+      "grad_norm": 0.26460525393486023,
+      "learning_rate": 4.837357345217718e-06,
+      "loss": 0.6338,
+      "step": 910
+    },
+    {
+      "epoch": 0.3840236062809569,
+      "grad_norm": 0.2555061876773834,
+      "learning_rate": 4.836959936181198e-06,
+      "loss": 0.6261,
+      "step": 911
+    },
+    {
+      "epoch": 0.38444514701233007,
+      "grad_norm": 0.26664793491363525,
+      "learning_rate": 4.836562058577521e-06,
+      "loss": 0.6248,
+      "step": 912
+    },
+    {
+      "epoch": 0.38486668774370325,
+      "grad_norm": 0.25871822237968445,
+      "learning_rate": 4.836163712486462e-06,
+      "loss": 0.6578,
+      "step": 913
+    },
+    {
+      "epoch": 0.3852882284750764,
+      "grad_norm": 0.25666993856430054,
+      "learning_rate": 4.835764897987892e-06,
+      "loss": 0.6487,
+      "step": 914
+    },
+    {
+      "epoch": 0.38570976920644956,
+      "grad_norm": 0.26849183440208435,
+      "learning_rate": 4.835365615161774e-06,
+      "loss": 0.6591,
+      "step": 915
+    },
+    {
+      "epoch": 0.38613130993782274,
+      "grad_norm": 0.28308624029159546,
+      "learning_rate": 4.834965864088164e-06,
+      "loss": 0.6279,
+      "step": 916
+    },
+    {
+      "epoch": 0.3865528506691959,
+      "grad_norm": 0.244406059384346,
+      "learning_rate": 4.834565644847215e-06,
+      "loss": 0.5834,
+      "step": 917
+    },
+    {
+      "epoch": 0.3869743914005691,
+      "grad_norm": 0.2602737545967102,
+      "learning_rate": 4.83416495751917e-06,
+      "loss": 0.6424,
+      "step": 918
+    },
+    {
+      "epoch": 0.38739593213194223,
+      "grad_norm": 0.25824710726737976,
+      "learning_rate": 4.833763802184371e-06,
+      "loss": 0.6539,
+      "step": 919
+    },
+    {
+      "epoch": 0.3878174728633154,
+      "grad_norm": 0.28233054280281067,
+      "learning_rate": 4.833362178923249e-06,
+      "loss": 0.6425,
+      "step": 920
+    },
+    {
+      "epoch": 0.3882390135946886,
+      "grad_norm": 0.25876104831695557,
+      "learning_rate": 4.83296008781633e-06,
+      "loss": 0.6141,
+      "step": 921
+    },
+    {
+      "epoch": 0.3886605543260618,
+      "grad_norm": 0.2694402039051056,
+      "learning_rate": 4.832557528944235e-06,
+      "loss": 0.6208,
+      "step": 922
+    },
+    {
+      "epoch": 0.3890820950574349,
+      "grad_norm": 0.2811453938484192,
+      "learning_rate": 4.832154502387678e-06,
+      "loss": 0.6169,
+      "step": 923
+    },
+    {
+      "epoch": 0.3895036357888081,
+      "grad_norm": 0.27431371808052063,
+      "learning_rate": 4.831751008227468e-06,
+      "loss": 0.6315,
+      "step": 924
+    },
+    {
+      "epoch": 0.38992517652018127,
+      "grad_norm": 0.2558780610561371,
+      "learning_rate": 4.831347046544505e-06,
+      "loss": 0.6419,
+      "step": 925
+    },
+    {
+      "epoch": 0.39034671725155445,
+      "grad_norm": 0.2636081874370575,
+      "learning_rate": 4.8309426174197875e-06,
+      "loss": 0.6255,
+      "step": 926
+    },
+    {
+      "epoch": 0.3907682579829276,
+      "grad_norm": 0.2561451196670532,
+      "learning_rate": 4.830537720934401e-06,
+      "loss": 0.6306,
+      "step": 927
+    },
+    {
+      "epoch": 0.39118979871430076,
+      "grad_norm": 0.26738831400871277,
+      "learning_rate": 4.830132357169531e-06,
+      "loss": 0.6358,
+      "step": 928
+    },
+    {
+      "epoch": 0.39161133944567394,
+      "grad_norm": 0.26140525937080383,
+      "learning_rate": 4.829726526206453e-06,
+      "loss": 0.6347,
+      "step": 929
+    },
+    {
+      "epoch": 0.3920328801770471,
+      "grad_norm": 0.2580152750015259,
+      "learning_rate": 4.829320228126537e-06,
+      "loss": 0.6141,
+      "step": 930
+    },
+    {
+      "epoch": 0.39245442090842025,
+      "grad_norm": 0.2830435633659363,
+      "learning_rate": 4.828913463011248e-06,
+      "loss": 0.634,
+      "step": 931
+    },
+    {
+      "epoch": 0.39287596163979344,
+      "grad_norm": 0.2561737895011902,
+      "learning_rate": 4.8285062309421424e-06,
+      "loss": 0.6337,
+      "step": 932
+    },
+    {
+      "epoch": 0.3932975023711666,
+      "grad_norm": 0.26395025849342346,
+      "learning_rate": 4.828098532000872e-06,
+      "loss": 0.6218,
+      "step": 933
+    },
+    {
+      "epoch": 0.3937190431025398,
+      "grad_norm": 0.25036484003067017,
+      "learning_rate": 4.827690366269182e-06,
+      "loss": 0.6037,
+      "step": 934
+    },
+    {
+      "epoch": 0.3941405838339129,
+      "grad_norm": 0.2546205520629883,
+      "learning_rate": 4.82728173382891e-06,
+      "loss": 0.6038,
+      "step": 935
+    },
+    {
+      "epoch": 0.3945621245652861,
+      "grad_norm": 0.26110392808914185,
+      "learning_rate": 4.826872634761989e-06,
+      "loss": 0.6198,
+      "step": 936
+    },
+    {
+      "epoch": 0.3949836652966593,
+      "grad_norm": 0.2652420401573181,
+      "learning_rate": 4.826463069150443e-06,
+      "loss": 0.6357,
+      "step": 937
+    },
+    {
+      "epoch": 0.3954052060280325,
+      "grad_norm": 0.275619238615036,
+      "learning_rate": 4.826053037076393e-06,
+      "loss": 0.5979,
+      "step": 938
+    },
+    {
+      "epoch": 0.3958267467594056,
+      "grad_norm": 0.2655697166919708,
+      "learning_rate": 4.82564253862205e-06,
+      "loss": 0.6333,
+      "step": 939
+    },
+    {
+      "epoch": 0.3962482874907788,
+      "grad_norm": 0.25900575518608093,
+      "learning_rate": 4.825231573869721e-06,
+      "loss": 0.6359,
+      "step": 940
+    },
+    {
+      "epoch": 0.39666982822215197,
+      "grad_norm": 0.25702881813049316,
+      "learning_rate": 4.824820142901805e-06,
+      "loss": 0.667,
+      "step": 941
+    },
+    {
+      "epoch": 0.39709136895352515,
+      "grad_norm": 0.2654663622379303,
+      "learning_rate": 4.824408245800794e-06,
+      "loss": 0.6149,
+      "step": 942
+    },
+    {
+      "epoch": 0.39751290968489833,
+      "grad_norm": 0.25416263937950134,
+      "learning_rate": 4.823995882649279e-06,
+      "loss": 0.6481,
+      "step": 943
+    },
+    {
+      "epoch": 0.39793445041627146,
+      "grad_norm": 0.2957819104194641,
+      "learning_rate": 4.823583053529934e-06,
+      "loss": 0.6442,
+      "step": 944
+    },
+    {
+      "epoch": 0.39835599114764464,
+      "grad_norm": 0.2590720057487488,
+      "learning_rate": 4.823169758525537e-06,
+      "loss": 0.646,
+      "step": 945
+    },
+    {
+      "epoch": 0.3987775318790178,
+      "grad_norm": 0.2745110094547272,
+      "learning_rate": 4.8227559977189516e-06,
+      "loss": 0.6227,
+      "step": 946
+    },
+    {
+      "epoch": 0.399199072610391,
+      "grad_norm": 0.28078117966651917,
+      "learning_rate": 4.822341771193142e-06,
+      "loss": 0.6345,
+      "step": 947
+    },
+    {
+      "epoch": 0.39962061334176413,
+      "grad_norm": 0.26438355445861816,
+      "learning_rate": 4.821927079031157e-06,
+      "loss": 0.6151,
+      "step": 948
+    },
+    {
+      "epoch": 0.4000421540731373,
+      "grad_norm": 0.26167407631874084,
+      "learning_rate": 4.821511921316147e-06,
+      "loss": 0.617,
+      "step": 949
+    },
+    {
+      "epoch": 0.4004636948045105,
+      "grad_norm": 0.2908693552017212,
+      "learning_rate": 4.821096298131351e-06,
+      "loss": 0.6236,
+      "step": 950
+    },
+    {
+      "epoch": 0.4008852355358837,
+      "grad_norm": 0.2645222544670105,
+      "learning_rate": 4.820680209560104e-06,
+      "loss": 0.6482,
+      "step": 951
+    },
+    {
+      "epoch": 0.4013067762672568,
+      "grad_norm": 0.2635226547718048,
+      "learning_rate": 4.820263655685831e-06,
+      "loss": 0.6192,
+      "step": 952
+    },
+    {
+      "epoch": 0.40172831699863,
+      "grad_norm": 0.3159831166267395,
+      "learning_rate": 4.8198466365920525e-06,
+      "loss": 0.6171,
+      "step": 953
+    },
+    {
+      "epoch": 0.40214985773000317,
+      "grad_norm": 0.2780272960662842,
+      "learning_rate": 4.819429152362383e-06,
+      "loss": 0.6223,
+      "step": 954
+    },
+    {
+      "epoch": 0.40257139846137635,
+      "grad_norm": 0.2659270465373993,
+      "learning_rate": 4.81901120308053e-06,
+      "loss": 0.635,
+      "step": 955
+    },
+    {
+      "epoch": 0.4029929391927495,
+      "grad_norm": 0.32857468724250793,
+      "learning_rate": 4.818592788830292e-06,
+      "loss": 0.6236,
+      "step": 956
+    },
+    {
+      "epoch": 0.40341447992412266,
+      "grad_norm": 0.36082008481025696,
+      "learning_rate": 4.8181739096955614e-06,
+      "loss": 0.5739,
+      "step": 957
+    },
+    {
+      "epoch": 0.40383602065549584,
+      "grad_norm": 0.2774188220500946,
+      "learning_rate": 4.817754565760327e-06,
+      "loss": 0.6393,
+      "step": 958
+    },
+    {
+      "epoch": 0.404257561386869,
+      "grad_norm": 0.38916826248168945,
+      "learning_rate": 4.817334757108667e-06,
+      "loss": 0.6197,
+      "step": 959
+    },
+    {
+      "epoch": 0.40467910211824215,
+      "grad_norm": 0.33336907625198364,
+      "learning_rate": 4.816914483824755e-06,
+      "loss": 0.6377,
+      "step": 960
+    },
+    {
+      "epoch": 0.40510064284961533,
+      "grad_norm": 0.2591753304004669,
+      "learning_rate": 4.8164937459928565e-06,
+      "loss": 0.6189,
+      "step": 961
+    },
+    {
+      "epoch": 0.4055221835809885,
+      "grad_norm": 0.3399209976196289,
+      "learning_rate": 4.816072543697331e-06,
+      "loss": 0.6243,
+      "step": 962
+    },
+    {
+      "epoch": 0.4059437243123617,
+      "grad_norm": 0.3264703154563904,
+      "learning_rate": 4.815650877022631e-06,
+      "loss": 0.6107,
+      "step": 963
+    },
+    {
+      "epoch": 0.4063652650437348,
+      "grad_norm": 0.26772332191467285,
+      "learning_rate": 4.815228746053301e-06,
+      "loss": 0.6011,
+      "step": 964
+    },
+    {
+      "epoch": 0.406786805775108,
+      "grad_norm": 0.34515243768692017,
+      "learning_rate": 4.814806150873979e-06,
+      "loss": 0.6466,
+      "step": 965
+    },
+    {
+      "epoch": 0.4072083465064812,
+      "grad_norm": 0.30386337637901306,
+      "learning_rate": 4.814383091569398e-06,
+      "loss": 0.6219,
+      "step": 966
+    },
+    {
+      "epoch": 0.40762988723785437,
+      "grad_norm": 0.252386212348938,
+      "learning_rate": 4.813959568224383e-06,
+      "loss": 0.6343,
+      "step": 967
+    },
+    {
+      "epoch": 0.40805142796922755,
+      "grad_norm": 0.2739870548248291,
+      "learning_rate": 4.813535580923849e-06,
+      "loss": 0.6363,
+      "step": 968
+    },
+    {
+      "epoch": 0.4084729687006007,
+      "grad_norm": 0.27867093682289124,
+      "learning_rate": 4.81311112975281e-06,
+      "loss": 0.6255,
+      "step": 969
+    },
+    {
+      "epoch": 0.40889450943197386,
+      "grad_norm": 0.26044467091560364,
+      "learning_rate": 4.8126862147963675e-06,
+      "loss": 0.6192,
+      "step": 970
+    },
+    {
+      "epoch": 0.40931605016334704,
+      "grad_norm": 0.24662238359451294,
+      "learning_rate": 4.812260836139718e-06,
+      "loss": 0.6205,
+      "step": 971
+    },
+    {
+      "epoch": 0.4097375908947202,
+      "grad_norm": 0.26094821095466614,
+      "learning_rate": 4.8118349938681525e-06,
+      "loss": 0.6368,
+      "step": 972
+    },
+    {
+      "epoch": 0.41015913162609335,
+      "grad_norm": 0.2669769823551178,
+      "learning_rate": 4.811408688067053e-06,
+      "loss": 0.6391,
+      "step": 973
+    },
+    {
+      "epoch": 0.41058067235746654,
+      "grad_norm": 0.26848191022872925,
+      "learning_rate": 4.810981918821894e-06,
+      "loss": 0.6263,
+      "step": 974
+    },
+    {
+      "epoch": 0.4110022130888397,
+      "grad_norm": 0.2604810893535614,
+      "learning_rate": 4.810554686218246e-06,
+      "loss": 0.6212,
+      "step": 975
+    },
+    {
+      "epoch": 0.4114237538202129,
+      "grad_norm": 0.26288801431655884,
+      "learning_rate": 4.810126990341769e-06,
+      "loss": 0.6204,
+      "step": 976
+    },
+    {
+      "epoch": 0.411845294551586,
+      "grad_norm": 0.3077448010444641,
+      "learning_rate": 4.809698831278217e-06,
+      "loss": 0.629,
+      "step": 977
+    },
+    {
+      "epoch": 0.4122668352829592,
+      "grad_norm": 0.2948167622089386,
+      "learning_rate": 4.809270209113438e-06,
+      "loss": 0.6324,
+      "step": 978
+    },
+    {
+      "epoch": 0.4126883760143324,
+      "grad_norm": 0.2664027810096741,
+      "learning_rate": 4.8088411239333725e-06,
+      "loss": 0.6253,
+      "step": 979
+    },
+    {
+      "epoch": 0.4131099167457056,
+      "grad_norm": 0.27179285883903503,
+      "learning_rate": 4.808411575824051e-06,
+      "loss": 0.6277,
+      "step": 980
+    },
+    {
+      "epoch": 0.4135314574770787,
+      "grad_norm": 0.2673485279083252,
+      "learning_rate": 4.807981564871601e-06,
+      "loss": 0.6243,
+      "step": 981
+    },
+    {
+      "epoch": 0.4139529982084519,
+      "grad_norm": 0.2943292260169983,
+      "learning_rate": 4.807551091162241e-06,
+      "loss": 0.6337,
+      "step": 982
+    },
+    {
+      "epoch": 0.41437453893982507,
+      "grad_norm": 0.24857105314731598,
+      "learning_rate": 4.807120154782281e-06,
+      "loss": 0.6193,
+      "step": 983
+    },
+    {
+      "epoch": 0.41479607967119825,
+      "grad_norm": 0.28324785828590393,
+      "learning_rate": 4.806688755818126e-06,
+      "loss": 0.6116,
+      "step": 984
+    },
+    {
+      "epoch": 0.4152176204025714,
+      "grad_norm": 0.2738904058933258,
+      "learning_rate": 4.806256894356273e-06,
+      "loss": 0.6279,
+      "step": 985
+    },
+    {
+      "epoch": 0.41563916113394456,
+      "grad_norm": 0.2727424204349518,
+      "learning_rate": 4.805824570483309e-06,
+      "loss": 0.6232,
+      "step": 986
+    },
+    {
+      "epoch": 0.41606070186531774,
+      "grad_norm": 0.2537044286727905,
+      "learning_rate": 4.80539178428592e-06,
+      "loss": 0.6316,
+      "step": 987
+    },
+    {
+      "epoch": 0.4164822425966909,
+      "grad_norm": 0.25446152687072754,
+      "learning_rate": 4.804958535850877e-06,
+      "loss": 0.6313,
+      "step": 988
+    },
+    {
+      "epoch": 0.41690378332806405,
+      "grad_norm": 0.30367958545684814,
+      "learning_rate": 4.804524825265051e-06,
+      "loss": 0.6262,
+      "step": 989
+    },
+    {
+      "epoch": 0.41732532405943723,
+      "grad_norm": 0.2823420464992523,
+      "learning_rate": 4.804090652615401e-06,
+      "loss": 0.6256,
+      "step": 990
+    },
+    {
+      "epoch": 0.4177468647908104,
+      "grad_norm": 0.27978891134262085,
+      "learning_rate": 4.803656017988978e-06,
+      "loss": 0.6501,
+      "step": 991
+    },
+    {
+      "epoch": 0.4181684055221836,
+      "grad_norm": 0.28915122151374817,
+      "learning_rate": 4.80322092147293e-06,
+      "loss": 0.6079,
+      "step": 992
+    },
+    {
+      "epoch": 0.4185899462535568,
+      "grad_norm": 0.2608501613140106,
+      "learning_rate": 4.802785363154493e-06,
+      "loss": 0.6165,
+      "step": 993
+    },
+    {
+      "epoch": 0.4190114869849299,
+      "grad_norm": 0.2665726840496063,
+      "learning_rate": 4.802349343121e-06,
+      "loss": 0.6212,
+      "step": 994
+    },
+    {
+      "epoch": 0.4194330277163031,
+      "grad_norm": 0.26818427443504333,
+      "learning_rate": 4.801912861459873e-06,
+      "loss": 0.6494,
+      "step": 995
+    },
+    {
+      "epoch": 0.41985456844767627,
+      "grad_norm": 0.26791125535964966,
+      "learning_rate": 4.801475918258628e-06,
+      "loss": 0.6191,
+      "step": 996
+    },
+    {
+      "epoch": 0.42027610917904945,
+      "grad_norm": 0.26152119040489197,
+      "learning_rate": 4.801038513604872e-06,
+      "loss": 0.6274,
+      "step": 997
+    },
+    {
+      "epoch": 0.4206976499104226,
+      "grad_norm": 0.26130810379981995,
+      "learning_rate": 4.8006006475863075e-06,
+      "loss": 0.6156,
+      "step": 998
+    },
+    {
+      "epoch": 0.42111919064179576,
+      "grad_norm": 0.2579680383205414,
+      "learning_rate": 4.800162320290728e-06,
+      "loss": 0.6235,
+      "step": 999
+    },
+    {
+      "epoch": 0.42154073137316894,
+      "grad_norm": 0.27043697237968445,
+      "learning_rate": 4.799723531806019e-06,
+      "loss": 0.5876,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4219622721045421,
+      "grad_norm": 0.27036502957344055,
+      "learning_rate": 4.7992842822201586e-06,
+      "loss": 0.6473,
+      "step": 1001
+    },
+    {
+      "epoch": 0.42238381283591525,
+      "grad_norm": 0.260535329580307,
+      "learning_rate": 4.798844571621217e-06,
+      "loss": 0.6272,
+      "step": 1002
+    },
+    {
+      "epoch": 0.42280535356728843,
+      "grad_norm": 0.2565985321998596,
+      "learning_rate": 4.798404400097359e-06,
+      "loss": 0.6298,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4232268942986616,
+      "grad_norm": 0.2548902928829193,
+      "learning_rate": 4.79796376773684e-06,
+      "loss": 0.6208,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4236484350300348,
+      "grad_norm": 0.2636111378669739,
+      "learning_rate": 4.7975226746280064e-06,
+      "loss": 0.6088,
+      "step": 1005
+    },
+    {
+      "epoch": 0.4240699757614079,
+      "grad_norm": 0.25958186388015747,
+      "learning_rate": 4.797081120859301e-06,
+      "loss": 0.6421,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4244915164927811,
+      "grad_norm": 0.26245561242103577,
+      "learning_rate": 4.796639106519253e-06,
+      "loss": 0.612,
+      "step": 1007
+    },
+    {
+      "epoch": 0.4249130572241543,
+      "grad_norm": 0.25324252247810364,
+      "learning_rate": 4.7961966316964916e-06,
+      "loss": 0.6278,
+      "step": 1008
+    },
+    {
+      "epoch": 0.42533459795552747,
+      "grad_norm": 0.26179778575897217,
+      "learning_rate": 4.795753696479733e-06,
+      "loss": 0.6126,
+      "step": 1009
+    },
+    {
+      "epoch": 0.4257561386869006,
+      "grad_norm": 0.25410422682762146,
+      "learning_rate": 4.795310300957784e-06,
+      "loss": 0.6127,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4261776794182738,
+      "grad_norm": 0.2644502520561218,
+      "learning_rate": 4.7948664452195515e-06,
+      "loss": 0.6303,
+      "step": 1011
+    },
+    {
+      "epoch": 0.42659922014964696,
+      "grad_norm": 0.2584587633609772,
+      "learning_rate": 4.794422129354026e-06,
+      "loss": 0.6257,
+      "step": 1012
+    },
+    {
+      "epoch": 0.42702076088102014,
+      "grad_norm": 0.2599596083164215,
+      "learning_rate": 4.793977353450297e-06,
+      "loss": 0.5903,
+      "step": 1013
+    },
+    {
+      "epoch": 0.42744230161239327,
+      "grad_norm": 0.25887182354927063,
+      "learning_rate": 4.793532117597541e-06,
+      "loss": 0.6244,
+      "step": 1014
+    },
+    {
+      "epoch": 0.42786384234376645,
+      "grad_norm": 0.2563137710094452,
+      "learning_rate": 4.7930864218850305e-06,
+      "loss": 0.6033,
+      "step": 1015
+    },
+    {
+      "epoch": 0.42828538307513964,
+      "grad_norm": 0.2676146626472473,
+      "learning_rate": 4.792640266402128e-06,
+      "loss": 0.6297,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4287069238065128,
+      "grad_norm": 0.2720690369606018,
+      "learning_rate": 4.792193651238289e-06,
+      "loss": 0.6335,
+      "step": 1017
+    },
+    {
+      "epoch": 0.429128464537886,
+      "grad_norm": 0.25424066185951233,
+      "learning_rate": 4.791746576483062e-06,
+      "loss": 0.6278,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4295500052692591,
+      "grad_norm": 0.2661586105823517,
+      "learning_rate": 4.791299042226085e-06,
+      "loss": 0.6293,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4299715460006323,
+      "grad_norm": 0.25833815336227417,
+      "learning_rate": 4.790851048557092e-06,
+      "loss": 0.6113,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4303930867320055,
+      "grad_norm": 0.2604774534702301,
+      "learning_rate": 4.790402595565906e-06,
+      "loss": 0.6309,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4308146274633787,
+      "grad_norm": 0.2568994462490082,
+      "learning_rate": 4.789953683342443e-06,
+      "loss": 0.6344,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4312361681947518,
+      "grad_norm": 0.2731021046638489,
+      "learning_rate": 4.789504311976712e-06,
+      "loss": 0.6195,
+      "step": 1023
+    },
+    {
+      "epoch": 0.431657708926125,
+      "grad_norm": 0.2635532021522522,
+      "learning_rate": 4.789054481558811e-06,
+      "loss": 0.6193,
+      "step": 1024
+    },
+    {
+      "epoch": 0.43207924965749817,
+      "grad_norm": 0.2546440362930298,
+      "learning_rate": 4.788604192178935e-06,
+      "loss": 0.6274,
+      "step": 1025
+    },
+    {
+      "epoch": 0.43250079038887135,
+      "grad_norm": 0.268265962600708,
+      "learning_rate": 4.788153443927367e-06,
+      "loss": 0.6401,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4329223311202445,
+      "grad_norm": 0.2618175745010376,
+      "learning_rate": 4.787702236894483e-06,
+      "loss": 0.65,
+      "step": 1027
+    },
+    {
+      "epoch": 0.43334387185161766,
+      "grad_norm": 0.25307247042655945,
+      "learning_rate": 4.787250571170752e-06,
+      "loss": 0.624,
+      "step": 1028
+    },
+    {
+      "epoch": 0.43376541258299084,
+      "grad_norm": 0.24586071074008942,
+      "learning_rate": 4.786798446846734e-06,
+      "loss": 0.6287,
+      "step": 1029
+    },
+    {
+      "epoch": 0.434186953314364,
+      "grad_norm": 0.2568615972995758,
+      "learning_rate": 4.786345864013081e-06,
+      "loss": 0.6345,
+      "step": 1030
+    },
+    {
+      "epoch": 0.43460849404573715,
+      "grad_norm": 0.2736034691333771,
+      "learning_rate": 4.785892822760538e-06,
+      "loss": 0.6079,
+      "step": 1031
+    },
+    {
+      "epoch": 0.43503003477711033,
+      "grad_norm": 0.26009607315063477,
+      "learning_rate": 4.7854393231799415e-06,
+      "loss": 0.6154,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4354515755084835,
+      "grad_norm": 0.26082906126976013,
+      "learning_rate": 4.784985365362217e-06,
+      "loss": 0.6118,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4358731162398567,
+      "grad_norm": 0.26791390776634216,
+      "learning_rate": 4.784530949398386e-06,
+      "loss": 0.6348,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4362946569712298,
+      "grad_norm": 0.24652516841888428,
+      "learning_rate": 4.78407607537956e-06,
+      "loss": 0.661,
+      "step": 1035
+    },
+    {
+      "epoch": 0.436716197702603,
+      "grad_norm": 0.2581693232059479,
+      "learning_rate": 4.783620743396943e-06,
+      "loss": 0.6159,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4371377384339762,
+      "grad_norm": 0.2631494998931885,
+      "learning_rate": 4.783164953541829e-06,
+      "loss": 0.6167,
+      "step": 1037
+    },
+    {
+      "epoch": 0.43755927916534937,
+      "grad_norm": 0.2587696611881256,
+      "learning_rate": 4.782708705905607e-06,
+      "loss": 0.619,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4379808198967225,
+      "grad_norm": 0.255253404378891,
+      "learning_rate": 4.7822520005797545e-06,
+      "loss": 0.6425,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4384023606280957,
+      "grad_norm": 0.26016899943351746,
+      "learning_rate": 4.7817948376558435e-06,
+      "loss": 0.6172,
+      "step": 1040
+    },
+    {
+      "epoch": 0.43882390135946886,
+      "grad_norm": 0.2678396701812744,
+      "learning_rate": 4.7813372172255345e-06,
+      "loss": 0.6272,
+      "step": 1041
+    },
+    {
+      "epoch": 0.43924544209084204,
+      "grad_norm": 0.2512117922306061,
+      "learning_rate": 4.780879139380584e-06,
+      "loss": 0.6242,
+      "step": 1042
+    },
+    {
+      "epoch": 0.43966698282221517,
+      "grad_norm": 0.25928163528442383,
+      "learning_rate": 4.780420604212836e-06,
+      "loss": 0.6097,
+      "step": 1043
+    },
+    {
+      "epoch": 0.44008852355358835,
+      "grad_norm": 0.2669723629951477,
+      "learning_rate": 4.77996161181423e-06,
+      "loss": 0.6439,
+      "step": 1044
+    },
+    {
+      "epoch": 0.44051006428496153,
+      "grad_norm": 0.2671312093734741,
+      "learning_rate": 4.779502162276793e-06,
+      "loss": 0.6315,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4409316050163347,
+      "grad_norm": 0.25989487767219543,
+      "learning_rate": 4.7790422556926485e-06,
+      "loss": 0.6374,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4413531457477079,
+      "grad_norm": 0.2662140429019928,
+      "learning_rate": 4.778581892154007e-06,
+      "loss": 0.6199,
+      "step": 1047
+    },
+    {
+      "epoch": 0.441774686479081,
+      "grad_norm": 0.2540116608142853,
+      "learning_rate": 4.778121071753174e-06,
+      "loss": 0.5967,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4421962272104542,
+      "grad_norm": 0.2513158917427063,
+      "learning_rate": 4.777659794582544e-06,
+      "loss": 0.6175,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4426177679418274,
+      "grad_norm": 0.2592232823371887,
+      "learning_rate": 4.777198060734606e-06,
+      "loss": 0.5877,
+      "step": 1050
+    },
+    {
+      "epoch": 0.44303930867320057,
+      "grad_norm": 0.25538673996925354,
+      "learning_rate": 4.776735870301939e-06,
+      "loss": 0.635,
+      "step": 1051
+    },
+    {
+      "epoch": 0.4434608494045737,
+      "grad_norm": 0.2596547603607178,
+      "learning_rate": 4.776273223377211e-06,
+      "loss": 0.64,
+      "step": 1052
+    },
+    {
+      "epoch": 0.4438823901359469,
+      "grad_norm": 0.2586513161659241,
+      "learning_rate": 4.775810120053186e-06,
+      "loss": 0.6121,
+      "step": 1053
+    },
+    {
+      "epoch": 0.44430393086732006,
+      "grad_norm": 0.2675786316394806,
+      "learning_rate": 4.775346560422718e-06,
+      "loss": 0.6537,
+      "step": 1054
+    },
+    {
+      "epoch": 0.44472547159869324,
+      "grad_norm": 0.2632947266101837,
+      "learning_rate": 4.7748825445787505e-06,
+      "loss": 0.6269,
+      "step": 1055
+    },
+    {
+      "epoch": 0.44514701233006637,
+      "grad_norm": 0.25514236092567444,
+      "learning_rate": 4.7744180726143215e-06,
+      "loss": 0.6271,
+      "step": 1056
+    },
+    {
+      "epoch": 0.44556855306143955,
+      "grad_norm": 0.26467758417129517,
+      "learning_rate": 4.773953144622558e-06,
+      "loss": 0.6083,
+      "step": 1057
+    },
+    {
+      "epoch": 0.44599009379281274,
+      "grad_norm": 0.25567513704299927,
+      "learning_rate": 4.773487760696679e-06,
+      "loss": 0.632,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4464116345241859,
+      "grad_norm": 0.26852452754974365,
+      "learning_rate": 4.773021920929996e-06,
+      "loss": 0.6156,
+      "step": 1059
+    },
+    {
+      "epoch": 0.44683317525555905,
+      "grad_norm": 0.26093602180480957,
+      "learning_rate": 4.772555625415912e-06,
+      "loss": 0.6063,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4472547159869322,
+      "grad_norm": 0.2546708285808563,
+      "learning_rate": 4.77208887424792e-06,
+      "loss": 0.6205,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4476762567183054,
+      "grad_norm": 0.2539382576942444,
+      "learning_rate": 4.771621667519604e-06,
+      "loss": 0.6236,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4480977974496786,
+      "grad_norm": 0.26151296496391296,
+      "learning_rate": 4.7711540053246415e-06,
+      "loss": 0.5942,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4485193381810517,
+      "grad_norm": 0.2815280258655548,
+      "learning_rate": 4.770685887756799e-06,
+      "loss": 0.6457,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4489408789124249,
+      "grad_norm": 0.269229918718338,
+      "learning_rate": 4.770217314909937e-06,
+      "loss": 0.6418,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4493624196437981,
+      "grad_norm": 0.26152241230010986,
+      "learning_rate": 4.7697482868780025e-06,
+      "loss": 0.6066,
+      "step": 1066
+    },
+    {
+      "epoch": 0.44978396037517127,
+      "grad_norm": 0.2684823274612427,
+      "learning_rate": 4.769278803755041e-06,
+      "loss": 0.6238,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4502055011065444,
+      "grad_norm": 0.25364938378334045,
+      "learning_rate": 4.768808865635183e-06,
+      "loss": 0.6417,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4506270418379176,
+      "grad_norm": 0.2659594714641571,
+      "learning_rate": 4.7683384726126525e-06,
+      "loss": 0.615,
+      "step": 1069
+    },
+    {
+      "epoch": 0.45104858256929076,
+      "grad_norm": 0.2671301066875458,
+      "learning_rate": 4.767867624781765e-06,
+      "loss": 0.6399,
+      "step": 1070
+    },
+    {
+      "epoch": 0.45147012330066394,
+      "grad_norm": 0.26167336106300354,
+      "learning_rate": 4.7673963222369264e-06,
+      "loss": 0.6278,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4518916640320371,
+      "grad_norm": 0.2554144561290741,
+      "learning_rate": 4.766924565072636e-06,
+      "loss": 0.644,
+      "step": 1072
+    },
+    {
+      "epoch": 0.45231320476341025,
+      "grad_norm": 0.2789387106895447,
+      "learning_rate": 4.766452353383479e-06,
+      "loss": 0.6336,
+      "step": 1073
+    },
+    {
+      "epoch": 0.45273474549478343,
+      "grad_norm": 0.2620246708393097,
+      "learning_rate": 4.765979687264138e-06,
+      "loss": 0.6323,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4531562862261566,
+      "grad_norm": 0.25983020663261414,
+      "learning_rate": 4.765506566809385e-06,
+      "loss": 0.6488,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4535778269575298,
+      "grad_norm": 0.26858508586883545,
+      "learning_rate": 4.765032992114078e-06,
+      "loss": 0.6129,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4539993676889029,
+      "grad_norm": 0.2651866376399994,
+      "learning_rate": 4.764558963273173e-06,
+      "loss": 0.6083,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4544209084202761,
+      "grad_norm": 0.2506422996520996,
+      "learning_rate": 4.764084480381713e-06,
+      "loss": 0.6206,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4548424491516493,
+      "grad_norm": 0.27116668224334717,
+      "learning_rate": 4.7636095435348336e-06,
+      "loss": 0.6445,
+      "step": 1079
+    },
+    {
+      "epoch": 0.45526398988302247,
+      "grad_norm": 0.26328805088996887,
+      "learning_rate": 4.763134152827762e-06,
+      "loss": 0.6463,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4556855306143956,
+      "grad_norm": 0.26561427116394043,
+      "learning_rate": 4.762658308355813e-06,
+      "loss": 0.5944,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4561070713457688,
+      "grad_norm": 0.2735762894153595,
+      "learning_rate": 4.762182010214398e-06,
+      "loss": 0.6099,
+      "step": 1082
+    },
+    {
+      "epoch": 0.45652861207714196,
+      "grad_norm": 0.27155566215515137,
+      "learning_rate": 4.761705258499012e-06,
+      "loss": 0.6211,
+      "step": 1083
+    },
+    {
+      "epoch": 0.45695015280851514,
+      "grad_norm": 0.2594447135925293,
+      "learning_rate": 4.761228053305249e-06,
+      "loss": 0.633,
+      "step": 1084
+    },
+    {
+      "epoch": 0.45737169353988827,
+      "grad_norm": 0.25998398661613464,
+      "learning_rate": 4.7607503947287885e-06,
+      "loss": 0.6284,
+      "step": 1085
+    },
+    {
+      "epoch": 0.45779323427126145,
+      "grad_norm": 0.26641547679901123,
+      "learning_rate": 4.760272282865402e-06,
+      "loss": 0.5986,
+      "step": 1086
+    },
+    {
+      "epoch": 0.45821477500263463,
+      "grad_norm": 0.26664137840270996,
+      "learning_rate": 4.759793717810951e-06,
+      "loss": 0.614,
+      "step": 1087
+    },
+    {
+      "epoch": 0.4586363157340078,
+      "grad_norm": 0.2598848342895508,
+      "learning_rate": 4.759314699661392e-06,
+      "loss": 0.617,
+      "step": 1088
+    },
+    {
+      "epoch": 0.45905785646538094,
+      "grad_norm": 0.28089427947998047,
+      "learning_rate": 4.758835228512768e-06,
+      "loss": 0.6155,
+      "step": 1089
+    },
+    {
+      "epoch": 0.4594793971967541,
+      "grad_norm": 0.28749001026153564,
+      "learning_rate": 4.758355304461215e-06,
+      "loss": 0.6243,
+      "step": 1090
+    },
+    {
+      "epoch": 0.4599009379281273,
+      "grad_norm": 0.27035224437713623,
+      "learning_rate": 4.757874927602959e-06,
+      "loss": 0.6113,
+      "step": 1091
+    },
+    {
+      "epoch": 0.4603224786595005,
+      "grad_norm": 0.2721879482269287,
+      "learning_rate": 4.757394098034316e-06,
+      "loss": 0.6232,
+      "step": 1092
+    },
+    {
+      "epoch": 0.4607440193908736,
+      "grad_norm": 0.2591506838798523,
+      "learning_rate": 4.756912815851694e-06,
+      "loss": 0.6237,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4611655601222468,
+      "grad_norm": 0.2687780261039734,
+      "learning_rate": 4.756431081151593e-06,
+      "loss": 0.6327,
+      "step": 1094
+    },
+    {
+      "epoch": 0.46158710085362,
+      "grad_norm": 0.25412189960479736,
+      "learning_rate": 4.7559488940306e-06,
+      "loss": 0.5993,
+      "step": 1095
+    },
+    {
+      "epoch": 0.46200864158499316,
+      "grad_norm": 0.2526768147945404,
+      "learning_rate": 4.7554662545853966e-06,
+      "loss": 0.6207,
+      "step": 1096
+    },
+    {
+      "epoch": 0.46243018231636634,
+      "grad_norm": 0.26229342818260193,
+      "learning_rate": 4.754983162912753e-06,
+      "loss": 0.6353,
+      "step": 1097
+    },
+    {
+      "epoch": 0.46285172304773947,
+      "grad_norm": 0.25999030470848083,
+      "learning_rate": 4.754499619109529e-06,
+      "loss": 0.6407,
+      "step": 1098
+    },
+    {
+      "epoch": 0.46327326377911265,
+      "grad_norm": 0.26542598009109497,
+      "learning_rate": 4.754015623272679e-06,
+      "loss": 0.6319,
+      "step": 1099
+    },
+    {
+      "epoch": 0.46369480451048584,
+      "grad_norm": 0.25090083479881287,
+      "learning_rate": 4.753531175499243e-06,
+      "loss": 0.5904,
+      "step": 1100
+    },
+    {
+      "epoch": 0.464116345241859,
+      "grad_norm": 0.2685970962047577,
+      "learning_rate": 4.753046275886356e-06,
+      "loss": 0.638,
+      "step": 1101
+    },
+    {
+      "epoch": 0.46453788597323215,
+      "grad_norm": 0.2622133493423462,
+      "learning_rate": 4.7525609245312414e-06,
+      "loss": 0.6111,
+      "step": 1102
+    },
+    {
+      "epoch": 0.4649594267046053,
+      "grad_norm": 0.26265624165534973,
+      "learning_rate": 4.752075121531213e-06,
+      "loss": 0.6259,
+      "step": 1103
+    },
+    {
+      "epoch": 0.4653809674359785,
+      "grad_norm": 0.2703480124473572,
+      "learning_rate": 4.751588866983676e-06,
+      "loss": 0.615,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4658025081673517,
+      "grad_norm": 0.2632228136062622,
+      "learning_rate": 4.751102160986126e-06,
+      "loss": 0.6091,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4662240488987248,
+      "grad_norm": 0.2622014582157135,
+      "learning_rate": 4.750615003636149e-06,
+      "loss": 0.6336,
+      "step": 1106
+    },
+    {
+      "epoch": 0.466645589630098,
+      "grad_norm": 0.2672812342643738,
+      "learning_rate": 4.750127395031421e-06,
+      "loss": 0.6219,
+      "step": 1107
+    },
+    {
+      "epoch": 0.4670671303614712,
+      "grad_norm": 0.26772648096084595,
+      "learning_rate": 4.749639335269709e-06,
+      "loss": 0.6113,
+      "step": 1108
+    },
+    {
+      "epoch": 0.46748867109284437,
+      "grad_norm": 0.26772385835647583,
+      "learning_rate": 4.74915082444887e-06,
+      "loss": 0.6246,
+      "step": 1109
+    },
+    {
+      "epoch": 0.4679102118242175,
+      "grad_norm": 0.2761198580265045,
+      "learning_rate": 4.748661862666854e-06,
+      "loss": 0.6341,
+      "step": 1110
+    },
+    {
+      "epoch": 0.4683317525555907,
+      "grad_norm": 0.2883282005786896,
+      "learning_rate": 4.748172450021695e-06,
+      "loss": 0.6078,
+      "step": 1111
+    },
+    {
+      "epoch": 0.46875329328696386,
+      "grad_norm": 0.26267868280410767,
+      "learning_rate": 4.747682586611526e-06,
+      "loss": 0.6168,
+      "step": 1112
+    },
+    {
+      "epoch": 0.46917483401833704,
+      "grad_norm": 0.2641350030899048,
+      "learning_rate": 4.747192272534564e-06,
+      "loss": 0.6288,
+      "step": 1113
+    },
+    {
+      "epoch": 0.46959637474971017,
+      "grad_norm": 0.2652049660682678,
+      "learning_rate": 4.746701507889119e-06,
+      "loss": 0.6109,
+      "step": 1114
+    },
+    {
+      "epoch": 0.47001791548108335,
+      "grad_norm": 0.2701943516731262,
+      "learning_rate": 4.746210292773589e-06,
+      "loss": 0.631,
+      "step": 1115
+    },
+    {
+      "epoch": 0.47043945621245653,
+      "grad_norm": 0.26050370931625366,
+      "learning_rate": 4.745718627286467e-06,
+      "loss": 0.6046,
+      "step": 1116
+    },
+    {
+      "epoch": 0.4708609969438297,
+      "grad_norm": 0.2500555217266083,
+      "learning_rate": 4.745226511526331e-06,
+      "loss": 0.621,
+      "step": 1117
+    },
+    {
+      "epoch": 0.47128253767520284,
+      "grad_norm": 0.27831533551216125,
+      "learning_rate": 4.744733945591852e-06,
+      "loss": 0.6063,
+      "step": 1118
+    },
+    {
+      "epoch": 0.471704078406576,
+      "grad_norm": 0.27488645911216736,
+      "learning_rate": 4.7442409295817905e-06,
+      "loss": 0.6426,
+      "step": 1119
+    },
+    {
+      "epoch": 0.4721256191379492,
+      "grad_norm": 0.26801246404647827,
+      "learning_rate": 4.743747463594999e-06,
+      "loss": 0.6219,
+      "step": 1120
+    },
+    {
+      "epoch": 0.4725471598693224,
+      "grad_norm": 0.2606232464313507,
+      "learning_rate": 4.743253547730419e-06,
+      "loss": 0.6343,
+      "step": 1121
+    },
+    {
+      "epoch": 0.47296870060069557,
+      "grad_norm": 0.26671120524406433,
+      "learning_rate": 4.74275918208708e-06,
+      "loss": 0.6246,
+      "step": 1122
+    },
+    {
+      "epoch": 0.4733902413320687,
+      "grad_norm": 0.2592431306838989,
+      "learning_rate": 4.742264366764106e-06,
+      "loss": 0.6178,
+      "step": 1123
+    },
+    {
+      "epoch": 0.4738117820634419,
+      "grad_norm": 0.2565414309501648,
+      "learning_rate": 4.741769101860707e-06,
+      "loss": 0.6211,
+      "step": 1124
+    },
+    {
+      "epoch": 0.47423332279481506,
+      "grad_norm": 0.2641512453556061,
+      "learning_rate": 4.741273387476186e-06,
+      "loss": 0.6228,
+      "step": 1125
+    },
+    {
+      "epoch": 0.47465486352618824,
+      "grad_norm": 0.2560514211654663,
+      "learning_rate": 4.740777223709934e-06,
+      "loss": 0.6173,
+      "step": 1126
+    },
+    {
+      "epoch": 0.47507640425756137,
+      "grad_norm": 0.2703777253627777,
+      "learning_rate": 4.740280610661436e-06,
+      "loss": 0.6566,
+      "step": 1127
+    },
+    {
+      "epoch": 0.47549794498893455,
+      "grad_norm": 0.2565274238586426,
+      "learning_rate": 4.739783548430262e-06,
+      "loss": 0.6207,
+      "step": 1128
+    },
+    {
+      "epoch": 0.47591948572030773,
+      "grad_norm": 0.2626293897628784,
+      "learning_rate": 4.739286037116075e-06,
+      "loss": 0.6205,
+      "step": 1129
+    },
+    {
+      "epoch": 0.4763410264516809,
+      "grad_norm": 0.2706518769264221,
+      "learning_rate": 4.738788076818627e-06,
+      "loss": 0.6454,
+      "step": 1130
+    },
+    {
+      "epoch": 0.47676256718305404,
+      "grad_norm": 0.2505090534687042,
+      "learning_rate": 4.738289667637761e-06,
+      "loss": 0.6198,
+      "step": 1131
+    },
+    {
+      "epoch": 0.4771841079144272,
+      "grad_norm": 0.26570942997932434,
+      "learning_rate": 4.73779080967341e-06,
+      "loss": 0.6382,
+      "step": 1132
+    },
+    {
+      "epoch": 0.4776056486458004,
+      "grad_norm": 0.26299405097961426,
+      "learning_rate": 4.737291503025595e-06,
+      "loss": 0.6518,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4780271893771736,
+      "grad_norm": 0.2567993998527527,
+      "learning_rate": 4.73679174779443e-06,
+      "loss": 0.621,
+      "step": 1134
+    },
+    {
+      "epoch": 0.4784487301085467,
+      "grad_norm": 0.27800050377845764,
+      "learning_rate": 4.736291544080116e-06,
+      "loss": 0.5977,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4788702708399199,
+      "grad_norm": 0.26815560460090637,
+      "learning_rate": 4.735790891982946e-06,
+      "loss": 0.6233,
+      "step": 1136
+    },
+    {
+      "epoch": 0.4792918115712931,
+      "grad_norm": 0.26686331629753113,
+      "learning_rate": 4.735289791603303e-06,
+      "loss": 0.6315,
+      "step": 1137
+    },
+    {
+      "epoch": 0.47971335230266626,
+      "grad_norm": 0.2701219618320465,
+      "learning_rate": 4.734788243041656e-06,
+      "loss": 0.6328,
+      "step": 1138
+    },
+    {
+      "epoch": 0.4801348930340394,
+      "grad_norm": 0.26330623030662537,
+      "learning_rate": 4.7342862463985705e-06,
+      "loss": 0.6183,
+      "step": 1139
+    },
+    {
+      "epoch": 0.48055643376541257,
+      "grad_norm": 0.25904008746147156,
+      "learning_rate": 4.733783801774696e-06,
+      "loss": 0.6172,
+      "step": 1140
+    },
+    {
+      "epoch": 0.48097797449678575,
+      "grad_norm": 0.26612216234207153,
+      "learning_rate": 4.733280909270775e-06,
+      "loss": 0.6304,
+      "step": 1141
+    },
+    {
+      "epoch": 0.48139951522815894,
+      "grad_norm": 0.258950799703598,
+      "learning_rate": 4.73277756898764e-06,
+      "loss": 0.6151,
+      "step": 1142
+    },
+    {
+      "epoch": 0.48182105595953206,
+      "grad_norm": 0.2604837417602539,
+      "learning_rate": 4.73227378102621e-06,
+      "loss": 0.6077,
+      "step": 1143
+    },
+    {
+      "epoch": 0.48224259669090525,
+      "grad_norm": 0.27270665764808655,
+      "learning_rate": 4.731769545487496e-06,
+      "loss": 0.6295,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4826641374222784,
+      "grad_norm": 0.27092665433883667,
+      "learning_rate": 4.731264862472601e-06,
+      "loss": 0.6293,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4830856781536516,
+      "grad_norm": 0.2547491192817688,
+      "learning_rate": 4.730759732082713e-06,
+      "loss": 0.6165,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4835072188850248,
+      "grad_norm": 0.2536171078681946,
+      "learning_rate": 4.730254154419114e-06,
+      "loss": 0.6339,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4839287596163979,
+      "grad_norm": 0.3112848401069641,
+      "learning_rate": 4.729748129583171e-06,
+      "loss": 0.5989,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4843503003477711,
+      "grad_norm": 0.2582842707633972,
+      "learning_rate": 4.729241657676347e-06,
+      "loss": 0.63,
+      "step": 1149
+    },
+    {
+      "epoch": 0.4847718410791443,
+      "grad_norm": 0.28158506751060486,
+      "learning_rate": 4.728734738800188e-06,
+      "loss": 0.6203,
+      "step": 1150
+    },
+    {
+      "epoch": 0.48519338181051747,
+      "grad_norm": 0.27585726976394653,
+      "learning_rate": 4.728227373056335e-06,
+      "loss": 0.5935,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4856149225418906,
+      "grad_norm": 0.255629301071167,
+      "learning_rate": 4.727719560546514e-06,
+      "loss": 0.6066,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4860364632732638,
+      "grad_norm": 0.2700357139110565,
+      "learning_rate": 4.727211301372545e-06,
+      "loss": 0.6083,
+      "step": 1153
+    },
+    {
+      "epoch": 0.48645800400463696,
+      "grad_norm": 0.25964388251304626,
+      "learning_rate": 4.7267025956363345e-06,
+      "loss": 0.6329,
+      "step": 1154
+    },
+    {
+      "epoch": 0.48687954473601014,
+      "grad_norm": 0.2649286985397339,
+      "learning_rate": 4.726193443439879e-06,
+      "loss": 0.6189,
+      "step": 1155
+    },
+    {
+      "epoch": 0.48730108546738327,
+      "grad_norm": 0.27843374013900757,
+      "learning_rate": 4.725683844885266e-06,
+      "loss": 0.6194,
+      "step": 1156
+    },
+    {
+      "epoch": 0.48772262619875645,
+      "grad_norm": 0.2973468005657196,
+      "learning_rate": 4.72517380007467e-06,
+      "loss": 0.624,
+      "step": 1157
+    },
+    {
+      "epoch": 0.48814416693012963,
+      "grad_norm": 0.2636312246322632,
+      "learning_rate": 4.724663309110359e-06,
+      "loss": 0.6249,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4885657076615028,
+      "grad_norm": 0.2726534903049469,
+      "learning_rate": 4.724152372094685e-06,
+      "loss": 0.6207,
+      "step": 1159
+    },
+    {
+      "epoch": 0.48898724839287594,
+      "grad_norm": 0.28269731998443604,
+      "learning_rate": 4.7236409891300936e-06,
+      "loss": 0.6207,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4894087891242491,
+      "grad_norm": 0.2639431357383728,
+      "learning_rate": 4.7231291603191184e-06,
+      "loss": 0.583,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4898303298556223,
+      "grad_norm": 0.2738626301288605,
+      "learning_rate": 4.722616885764384e-06,
+      "loss": 0.6006,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4902518705869955,
+      "grad_norm": 0.2973650395870209,
+      "learning_rate": 4.7221041655686e-06,
+      "loss": 0.6225,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4906734113183686,
+      "grad_norm": 0.2661890387535095,
+      "learning_rate": 4.721590999834571e-06,
+      "loss": 0.6161,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4910949520497418,
+      "grad_norm": 0.2920544743537903,
+      "learning_rate": 4.721077388665186e-06,
+      "loss": 0.6088,
+      "step": 1165
+    },
+    {
+      "epoch": 0.491516492781115,
+      "grad_norm": 0.2753474712371826,
+      "learning_rate": 4.720563332163428e-06,
+      "loss": 0.6053,
+      "step": 1166
+    },
+    {
+      "epoch": 0.49193803351248816,
+      "grad_norm": 0.2735542058944702,
+      "learning_rate": 4.720048830432364e-06,
+      "loss": 0.597,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4923595742438613,
+      "grad_norm": 0.3518235683441162,
+      "learning_rate": 4.7195338835751545e-06,
+      "loss": 0.623,
+      "step": 1168
+    },
+    {
+      "epoch": 0.49278111497523447,
+      "grad_norm": 0.26030486822128296,
+      "learning_rate": 4.7190184916950484e-06,
+      "loss": 0.6369,
+      "step": 1169
+    },
+    {
+      "epoch": 0.49320265570660765,
+      "grad_norm": 0.26839718222618103,
+      "learning_rate": 4.718502654895383e-06,
+      "loss": 0.6006,
+      "step": 1170
+    },
+    {
+      "epoch": 0.49362419643798083,
+      "grad_norm": 0.2796967327594757,
+      "learning_rate": 4.7179863732795835e-06,
+      "loss": 0.6441,
+      "step": 1171
+    },
+    {
+      "epoch": 0.494045737169354,
+      "grad_norm": 0.2542371153831482,
+      "learning_rate": 4.7174696469511675e-06,
+      "loss": 0.6417,
+      "step": 1172
+    },
+    {
+      "epoch": 0.49446727790072714,
+      "grad_norm": 0.2567559778690338,
+      "learning_rate": 4.71695247601374e-06,
+      "loss": 0.6154,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4948888186321003,
+      "grad_norm": 0.2611733675003052,
+      "learning_rate": 4.716434860570994e-06,
+      "loss": 0.6396,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4953103593634735,
+      "grad_norm": 0.2692919373512268,
+      "learning_rate": 4.7159168007267144e-06,
+      "loss": 0.6252,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4957319000948467,
+      "grad_norm": 0.2635321021080017,
+      "learning_rate": 4.715398296584773e-06,
+      "loss": 0.6166,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4961534408262198,
+      "grad_norm": 0.2747281491756439,
+      "learning_rate": 4.7148793482491314e-06,
+      "loss": 0.6074,
+      "step": 1177
+    },
+    {
+      "epoch": 0.496574981557593,
+      "grad_norm": 0.27167364954948425,
+      "learning_rate": 4.71435995582384e-06,
+      "loss": 0.6175,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4969965222889662,
+      "grad_norm": 0.27348098158836365,
+      "learning_rate": 4.713840119413039e-06,
+      "loss": 0.6353,
+      "step": 1179
+    },
+    {
+      "epoch": 0.49741806302033936,
+      "grad_norm": 0.2701179087162018,
+      "learning_rate": 4.713319839120957e-06,
+      "loss": 0.6189,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4978396037517125,
+      "grad_norm": 0.28248587250709534,
+      "learning_rate": 4.712799115051911e-06,
+      "loss": 0.6056,
+      "step": 1181
+    },
+    {
+      "epoch": 0.49826114448308567,
+      "grad_norm": 0.2572071850299835,
+      "learning_rate": 4.7122779473103075e-06,
+      "loss": 0.6542,
+      "step": 1182
+    },
+    {
+      "epoch": 0.49868268521445885,
+      "grad_norm": 0.2606280446052551,
+      "learning_rate": 4.711756336000643e-06,
+      "loss": 0.6199,
+      "step": 1183
+    },
+    {
+      "epoch": 0.49910422594583204,
+      "grad_norm": 0.2715243399143219,
+      "learning_rate": 4.711234281227502e-06,
+      "loss": 0.6235,
+      "step": 1184
+    },
+    {
+      "epoch": 0.49952576667720516,
+      "grad_norm": 0.2728928327560425,
+      "learning_rate": 4.710711783095559e-06,
+      "loss": 0.6058,
+      "step": 1185
+    },
+    {
+      "epoch": 0.49994730740857835,
+      "grad_norm": 0.2675876021385193,
+      "learning_rate": 4.710188841709575e-06,
+      "loss": 0.629,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5003688481399515,
+      "grad_norm": 0.2776484489440918,
+      "learning_rate": 4.709665457174399e-06,
+      "loss": 0.6052,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5007903888713247,
+      "grad_norm": 0.26823145151138306,
+      "learning_rate": 4.709141629594975e-06,
+      "loss": 0.5789,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5012119296026979,
+      "grad_norm": 0.29227155447006226,
+      "learning_rate": 4.7086173590763305e-06,
+      "loss": 0.6003,
+      "step": 1189
+    },
+    {
+      "epoch": 0.501633470334071,
+      "grad_norm": 0.2729746997356415,
+      "learning_rate": 4.708092645723584e-06,
+      "loss": 0.6348,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5020550110654441,
+      "grad_norm": 0.2588585615158081,
+      "learning_rate": 4.70756748964194e-06,
+      "loss": 0.6124,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5024765517968174,
+      "grad_norm": 0.2697394788265228,
+      "learning_rate": 4.707041890936696e-06,
+      "loss": 0.6089,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5028980925281905,
+      "grad_norm": 0.259132444858551,
+      "learning_rate": 4.706515849713234e-06,
+      "loss": 0.5953,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5033196332595637,
+      "grad_norm": 0.2552300691604614,
+      "learning_rate": 4.705989366077028e-06,
+      "loss": 0.6016,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5037411739909369,
+      "grad_norm": 0.2820501923561096,
+      "learning_rate": 4.70546244013364e-06,
+      "loss": 0.6113,
+      "step": 1195
+    },
+    {
+      "epoch": 0.50416271472231,
+      "grad_norm": 0.25056329369544983,
+      "learning_rate": 4.704935071988719e-06,
+      "loss": 0.625,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5045842554536832,
+      "grad_norm": 0.2605217695236206,
+      "learning_rate": 4.704407261748003e-06,
+      "loss": 0.6212,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5050057961850564,
+      "grad_norm": 0.27459844946861267,
+      "learning_rate": 4.703879009517322e-06,
+      "loss": 0.6344,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5054273369164296,
+      "grad_norm": 0.2750336527824402,
+      "learning_rate": 4.703350315402591e-06,
+      "loss": 0.6229,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5058488776478027,
+      "grad_norm": 0.2644559442996979,
+      "learning_rate": 4.702821179509815e-06,
+      "loss": 0.6228,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5062704183791759,
+      "grad_norm": 0.3024531900882721,
+      "learning_rate": 4.702291601945086e-06,
+      "loss": 0.5882,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5066919591105491,
+      "grad_norm": 0.27915382385253906,
+      "learning_rate": 4.701761582814586e-06,
+      "loss": 0.6138,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5071134998419222,
+      "grad_norm": 0.27671054005622864,
+      "learning_rate": 4.7012311222245866e-06,
+      "loss": 0.5977,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5075350405732953,
+      "grad_norm": 0.29683148860931396,
+      "learning_rate": 4.700700220281446e-06,
+      "loss": 0.6166,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5079565813046686,
+      "grad_norm": 0.26660504937171936,
+      "learning_rate": 4.700168877091612e-06,
+      "loss": 0.6251,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5083781220360417,
+      "grad_norm": 0.2591957151889801,
+      "learning_rate": 4.699637092761619e-06,
+      "loss": 0.6149,
+      "step": 1206
+    },
+    {
+      "epoch": 0.508799662767415,
+      "grad_norm": 0.2905402183532715,
+      "learning_rate": 4.699104867398093e-06,
+      "loss": 0.6,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5092212034987881,
+      "grad_norm": 0.28757134079933167,
+      "learning_rate": 4.6985722011077465e-06,
+      "loss": 0.6183,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5096427442301612,
+      "grad_norm": 0.26568982005119324,
+      "learning_rate": 4.698039093997379e-06,
+      "loss": 0.6165,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5100642849615344,
+      "grad_norm": 0.27066823840141296,
+      "learning_rate": 4.697505546173881e-06,
+      "loss": 0.5983,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5104858256929076,
+      "grad_norm": 0.2730790674686432,
+      "learning_rate": 4.696971557744231e-06,
+      "loss": 0.6478,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5109073664242807,
+      "grad_norm": 0.26784396171569824,
+      "learning_rate": 4.696437128815494e-06,
+      "loss": 0.625,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5113289071556539,
+      "grad_norm": 0.27559807896614075,
+      "learning_rate": 4.6959022594948265e-06,
+      "loss": 0.6228,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5117504478870271,
+      "grad_norm": 0.2691324055194855,
+      "learning_rate": 4.6953669498894695e-06,
+      "loss": 0.6239,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5121719886184003,
+      "grad_norm": 0.2766820192337036,
+      "learning_rate": 4.694831200106754e-06,
+      "loss": 0.5974,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5125935293497734,
+      "grad_norm": 0.2721300721168518,
+      "learning_rate": 4.694295010254101e-06,
+      "loss": 0.6012,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5130150700811466,
+      "grad_norm": 0.27212944626808167,
+      "learning_rate": 4.693758380439016e-06,
+      "loss": 0.6383,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5134366108125198,
+      "grad_norm": 0.25905007123947144,
+      "learning_rate": 4.693221310769097e-06,
+      "loss": 0.6465,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5138581515438929,
+      "grad_norm": 0.26136910915374756,
+      "learning_rate": 4.692683801352026e-06,
+      "loss": 0.613,
+      "step": 1219
+    },
+    {
+      "epoch": 0.514279692275266,
+      "grad_norm": 0.2776919901371002,
+      "learning_rate": 4.692145852295577e-06,
+      "loss": 0.6029,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5147012330066393,
+      "grad_norm": 0.26061469316482544,
+      "learning_rate": 4.6916074637076085e-06,
+      "loss": 0.6294,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5151227737380124,
+      "grad_norm": 0.2639305293560028,
+      "learning_rate": 4.69106863569607e-06,
+      "loss": 0.607,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5155443144693856,
+      "grad_norm": 0.2754727900028229,
+      "learning_rate": 4.690529368368999e-06,
+      "loss": 0.5958,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5159658552007588,
+      "grad_norm": 0.26845815777778625,
+      "learning_rate": 4.6899896618345185e-06,
+      "loss": 0.6294,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5163873959321319,
+      "grad_norm": 0.28922829031944275,
+      "learning_rate": 4.689449516200842e-06,
+      "loss": 0.6337,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5168089366635051,
+      "grad_norm": 0.27070170640945435,
+      "learning_rate": 4.688908931576269e-06,
+      "loss": 0.636,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5172304773948783,
+      "grad_norm": 0.26557695865631104,
+      "learning_rate": 4.68836790806919e-06,
+      "loss": 0.6313,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5176520181262515,
+      "grad_norm": 0.2571782171726227,
+      "learning_rate": 4.6878264457880815e-06,
+      "loss": 0.6316,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5180735588576246,
+      "grad_norm": 0.2846078872680664,
+      "learning_rate": 4.687284544841508e-06,
+      "loss": 0.6299,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5184950995889978,
+      "grad_norm": 0.28038179874420166,
+      "learning_rate": 4.68674220533812e-06,
+      "loss": 0.6239,
+      "step": 1230
+    },
+    {
+      "epoch": 0.518916640320371,
+      "grad_norm": 0.26650097966194153,
+      "learning_rate": 4.686199427386662e-06,
+      "loss": 0.628,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5193381810517441,
+      "grad_norm": 0.26156511902809143,
+      "learning_rate": 4.68565621109596e-06,
+      "loss": 0.6342,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5197597217831172,
+      "grad_norm": 0.26265615224838257,
+      "learning_rate": 4.685112556574931e-06,
+      "loss": 0.6094,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5201812625144905,
+      "grad_norm": 0.2677500247955322,
+      "learning_rate": 4.68456846393258e-06,
+      "loss": 0.6113,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5206028032458636,
+      "grad_norm": 0.26116621494293213,
+      "learning_rate": 4.684023933277998e-06,
+      "loss": 0.6254,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5210243439772368,
+      "grad_norm": 0.26652956008911133,
+      "learning_rate": 4.683478964720366e-06,
+      "loss": 0.6179,
+      "step": 1236
+    },
+    {
+      "epoch": 0.52144588470861,
+      "grad_norm": 0.2626531422138214,
+      "learning_rate": 4.68293355836895e-06,
+      "loss": 0.6542,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5218674254399831,
+      "grad_norm": 0.2718144953250885,
+      "learning_rate": 4.682387714333108e-06,
+      "loss": 0.5996,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5222889661713563,
+      "grad_norm": 0.27493616938591003,
+      "learning_rate": 4.681841432722281e-06,
+      "loss": 0.6105,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5227105069027295,
+      "grad_norm": 0.2722738981246948,
+      "learning_rate": 4.681294713646002e-06,
+      "loss": 0.6091,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5231320476341026,
+      "grad_norm": 0.2790028750896454,
+      "learning_rate": 4.680747557213888e-06,
+      "loss": 0.6217,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5235535883654758,
+      "grad_norm": 0.26496145129203796,
+      "learning_rate": 4.680199963535646e-06,
+      "loss": 0.6097,
+      "step": 1242
+    },
+    {
+      "epoch": 0.523975129096849,
+      "grad_norm": 0.2740166485309601,
+      "learning_rate": 4.679651932721071e-06,
+      "loss": 0.6401,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5243966698282222,
+      "grad_norm": 0.2687187194824219,
+      "learning_rate": 4.679103464880044e-06,
+      "loss": 0.6188,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5248182105595953,
+      "grad_norm": 0.2666671872138977,
+      "learning_rate": 4.678554560122534e-06,
+      "loss": 0.6176,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5252397512909684,
+      "grad_norm": 0.26051363348960876,
+      "learning_rate": 4.678005218558599e-06,
+      "loss": 0.6409,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5256612920223417,
+      "grad_norm": 0.27537602186203003,
+      "learning_rate": 4.677455440298383e-06,
+      "loss": 0.625,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5260828327537148,
+      "grad_norm": 0.2916540801525116,
+      "learning_rate": 4.676905225452117e-06,
+      "loss": 0.6291,
+      "step": 1248
+    },
+    {
+      "epoch": 0.526504373485088,
+      "grad_norm": 0.2708275318145752,
+      "learning_rate": 4.676354574130123e-06,
+      "loss": 0.6136,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5269259142164612,
+      "grad_norm": 0.27206695079803467,
+      "learning_rate": 4.675803486442807e-06,
+      "loss": 0.607,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5273474549478343,
+      "grad_norm": 0.2734690308570862,
+      "learning_rate": 4.6752519625006625e-06,
+      "loss": 0.6065,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5277689956792075,
+      "grad_norm": 0.2904685139656067,
+      "learning_rate": 4.674700002414274e-06,
+      "loss": 0.5856,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5281905364105807,
+      "grad_norm": 0.2815206050872803,
+      "learning_rate": 4.674147606294308e-06,
+      "loss": 0.6145,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5286120771419538,
+      "grad_norm": 0.26160579919815063,
+      "learning_rate": 4.673594774251525e-06,
+      "loss": 0.6331,
+      "step": 1254
+    },
+    {
+      "epoch": 0.529033617873327,
+      "grad_norm": 0.27661266922950745,
+      "learning_rate": 4.673041506396767e-06,
+      "loss": 0.6137,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5294551586047002,
+      "grad_norm": 0.2790932059288025,
+      "learning_rate": 4.672487802840966e-06,
+      "loss": 0.6175,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5298766993360734,
+      "grad_norm": 0.2758411765098572,
+      "learning_rate": 4.671933663695142e-06,
+      "loss": 0.6006,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5302982400674465,
+      "grad_norm": 0.26209884881973267,
+      "learning_rate": 4.671379089070402e-06,
+      "loss": 0.632,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5307197807988197,
+      "grad_norm": 0.2769412696361542,
+      "learning_rate": 4.670824079077937e-06,
+      "loss": 0.5768,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5311413215301929,
+      "grad_norm": 0.2690974473953247,
+      "learning_rate": 4.670268633829032e-06,
+      "loss": 0.6218,
+      "step": 1260
+    },
+    {
+      "epoch": 0.531562862261566,
+      "grad_norm": 0.26128584146499634,
+      "learning_rate": 4.669712753435052e-06,
+      "loss": 0.6099,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5319844029929391,
+      "grad_norm": 0.26777008175849915,
+      "learning_rate": 4.669156438007454e-06,
+      "loss": 0.6289,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5324059437243124,
+      "grad_norm": 0.2801518440246582,
+      "learning_rate": 4.668599687657781e-06,
+      "loss": 0.6217,
+      "step": 1263
+    },
+    {
+      "epoch": 0.5328274844556855,
+      "grad_norm": 0.28167295455932617,
+      "learning_rate": 4.6680425024976635e-06,
+      "loss": 0.6024,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5332490251870587,
+      "grad_norm": 0.2719642221927643,
+      "learning_rate": 4.667484882638818e-06,
+      "loss": 0.6301,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5336705659184319,
+      "grad_norm": 0.25974738597869873,
+      "learning_rate": 4.666926828193048e-06,
+      "loss": 0.6071,
+      "step": 1266
+    },
+    {
+      "epoch": 0.534092106649805,
+      "grad_norm": 0.271354079246521,
+      "learning_rate": 4.666368339272246e-06,
+      "loss": 0.625,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5345136473811782,
+      "grad_norm": 0.2817900478839874,
+      "learning_rate": 4.665809415988391e-06,
+      "loss": 0.6105,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5349351881125514,
+      "grad_norm": 0.28142639994621277,
+      "learning_rate": 4.665250058453549e-06,
+      "loss": 0.6025,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5353567288439245,
+      "grad_norm": 0.2750336527824402,
+      "learning_rate": 4.664690266779871e-06,
+      "loss": 0.5902,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5357782695752977,
+      "grad_norm": 0.28132110834121704,
+      "learning_rate": 4.664130041079599e-06,
+      "loss": 0.654,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5361998103066709,
+      "grad_norm": 0.2764010727405548,
+      "learning_rate": 4.663569381465057e-06,
+      "loss": 0.6378,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5366213510380441,
+      "grad_norm": 0.2571348249912262,
+      "learning_rate": 4.6630082880486625e-06,
+      "loss": 0.6298,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5370428917694172,
+      "grad_norm": 0.2853141129016876,
+      "learning_rate": 4.6624467609429135e-06,
+      "loss": 0.592,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5374644325007903,
+      "grad_norm": 0.2846934497356415,
+      "learning_rate": 4.661884800260399e-06,
+      "loss": 0.6011,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5378859732321636,
+      "grad_norm": 0.2740291655063629,
+      "learning_rate": 4.661322406113794e-06,
+      "loss": 0.6189,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5383075139635367,
+      "grad_norm": 0.2646265923976898,
+      "learning_rate": 4.660759578615859e-06,
+      "loss": 0.6108,
+      "step": 1277
+    },
+    {
+      "epoch": 0.53872905469491,
+      "grad_norm": 0.2702798843383789,
+      "learning_rate": 4.660196317879443e-06,
+      "loss": 0.6119,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5391505954262831,
+      "grad_norm": 0.2724854052066803,
+      "learning_rate": 4.659632624017482e-06,
+      "loss": 0.6253,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5395721361576562,
+      "grad_norm": 0.2663208842277527,
+      "learning_rate": 4.659068497142998e-06,
+      "loss": 0.627,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5399936768890294,
+      "grad_norm": 0.26867082715034485,
+      "learning_rate": 4.6585039373690986e-06,
+      "loss": 0.6279,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5404152176204026,
+      "grad_norm": 0.27773624658584595,
+      "learning_rate": 4.657938944808981e-06,
+      "loss": 0.5959,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5408367583517757,
+      "grad_norm": 0.2634912431240082,
+      "learning_rate": 4.65737351957593e-06,
+      "loss": 0.6357,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5412582990831489,
+      "grad_norm": 0.2709040939807892,
+      "learning_rate": 4.6568076617833104e-06,
+      "loss": 0.6215,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5416798398145221,
+      "grad_norm": 0.2677391469478607,
+      "learning_rate": 4.656241371544581e-06,
+      "loss": 0.617,
+      "step": 1285
+    },
+    {
+      "epoch": 0.5421013805458953,
+      "grad_norm": 0.2628321349620819,
+      "learning_rate": 4.655674648973285e-06,
+      "loss": 0.6216,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5425229212772684,
+      "grad_norm": 0.26534026861190796,
+      "learning_rate": 4.65510749418305e-06,
+      "loss": 0.5938,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5429444620086415,
+      "grad_norm": 0.2653622627258301,
+      "learning_rate": 4.654539907287594e-06,
+      "loss": 0.6245,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5433660027400148,
+      "grad_norm": 0.263202965259552,
+      "learning_rate": 4.653971888400719e-06,
+      "loss": 0.644,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5437875434713879,
+      "grad_norm": 0.283340185880661,
+      "learning_rate": 4.653403437636314e-06,
+      "loss": 0.6118,
+      "step": 1290
+    },
+    {
+      "epoch": 0.544209084202761,
+      "grad_norm": 0.26572883129119873,
+      "learning_rate": 4.6528345551083565e-06,
+      "loss": 0.6024,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5446306249341343,
+      "grad_norm": 0.26857027411460876,
+      "learning_rate": 4.652265240930907e-06,
+      "loss": 0.6376,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5450521656655074,
+      "grad_norm": 0.278123140335083,
+      "learning_rate": 4.651695495218116e-06,
+      "loss": 0.6101,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5454737063968806,
+      "grad_norm": 0.27677711844444275,
+      "learning_rate": 4.651125318084219e-06,
+      "loss": 0.63,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5458952471282538,
+      "grad_norm": 0.3076222240924835,
+      "learning_rate": 4.6505547096435374e-06,
+      "loss": 0.6067,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5463167878596269,
+      "grad_norm": 0.2680820822715759,
+      "learning_rate": 4.649983670010481e-06,
+      "loss": 0.6128,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5467383285910001,
+      "grad_norm": 0.2639501392841339,
+      "learning_rate": 4.649412199299544e-06,
+      "loss": 0.6173,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5471598693223733,
+      "grad_norm": 0.2687246799468994,
+      "learning_rate": 4.648840297625308e-06,
+      "loss": 0.6373,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5475814100537465,
+      "grad_norm": 0.2728808522224426,
+      "learning_rate": 4.648267965102442e-06,
+      "loss": 0.6089,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5480029507851196,
+      "grad_norm": 0.27639055252075195,
+      "learning_rate": 4.6476952018456974e-06,
+      "loss": 0.6113,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5484244915164928,
+      "grad_norm": 0.2534872591495514,
+      "learning_rate": 4.647122007969919e-06,
+      "loss": 0.621,
+      "step": 1301
+    },
+    {
+      "epoch": 0.548846032247866,
+      "grad_norm": 0.254555881023407,
+      "learning_rate": 4.646548383590031e-06,
+      "loss": 0.636,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5492675729792391,
+      "grad_norm": 0.27459999918937683,
+      "learning_rate": 4.645974328821048e-06,
+      "loss": 0.606,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5496891137106122,
+      "grad_norm": 0.2772482931613922,
+      "learning_rate": 4.6453998437780675e-06,
+      "loss": 0.6275,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5501106544419855,
+      "grad_norm": 0.27909979224205017,
+      "learning_rate": 4.644824928576278e-06,
+      "loss": 0.6096,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5505321951733586,
+      "grad_norm": 0.29173335433006287,
+      "learning_rate": 4.6442495833309505e-06,
+      "loss": 0.6369,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5509537359047318,
+      "grad_norm": 0.27623939514160156,
+      "learning_rate": 4.643673808157444e-06,
+      "loss": 0.6031,
+      "step": 1307
+    },
+    {
+      "epoch": 0.551375276636105,
+      "grad_norm": 0.2823559045791626,
+      "learning_rate": 4.643097603171202e-06,
+      "loss": 0.6291,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5517968173674781,
+      "grad_norm": 0.27116885781288147,
+      "learning_rate": 4.642520968487755e-06,
+      "loss": 0.62,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5522183580988513,
+      "grad_norm": 0.2853292226791382,
+      "learning_rate": 4.641943904222722e-06,
+      "loss": 0.6337,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5526398988302245,
+      "grad_norm": 0.27805304527282715,
+      "learning_rate": 4.641366410491804e-06,
+      "loss": 0.6042,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5530614395615976,
+      "grad_norm": 0.2828497290611267,
+      "learning_rate": 4.6407884874107914e-06,
+      "loss": 0.5965,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5534829802929708,
+      "grad_norm": 0.33296999335289,
+      "learning_rate": 4.640210135095559e-06,
+      "loss": 0.6094,
+      "step": 1313
+    },
+    {
+      "epoch": 0.553904521024344,
+      "grad_norm": 0.28261128067970276,
+      "learning_rate": 4.6396313536620665e-06,
+      "loss": 0.6339,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5543260617557172,
+      "grad_norm": 0.3172251284122467,
+      "learning_rate": 4.639052143226364e-06,
+      "loss": 0.611,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5547476024870903,
+      "grad_norm": 0.2962484359741211,
+      "learning_rate": 4.638472503904583e-06,
+      "loss": 0.64,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5551691432184634,
+      "grad_norm": 0.27523574233055115,
+      "learning_rate": 4.637892435812944e-06,
+      "loss": 0.6277,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5555906839498367,
+      "grad_norm": 0.28877633810043335,
+      "learning_rate": 4.6373119390677515e-06,
+      "loss": 0.5888,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5560122246812098,
+      "grad_norm": 0.2957388460636139,
+      "learning_rate": 4.6367310137853975e-06,
+      "loss": 0.6116,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5564337654125829,
+      "grad_norm": 0.2656603455543518,
+      "learning_rate": 4.636149660082358e-06,
+      "loss": 0.6043,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5568553061439562,
+      "grad_norm": 0.27609819173812866,
+      "learning_rate": 4.635567878075197e-06,
+      "loss": 0.6034,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5572768468753293,
+      "grad_norm": 0.2761106491088867,
+      "learning_rate": 4.634985667880563e-06,
+      "loss": 0.6135,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5576983876067025,
+      "grad_norm": 0.2718981206417084,
+      "learning_rate": 4.6344030296151915e-06,
+      "loss": 0.6282,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5581199283380757,
+      "grad_norm": 0.264858603477478,
+      "learning_rate": 4.6338199633959025e-06,
+      "loss": 0.6012,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5585414690694488,
+      "grad_norm": 0.281182199716568,
+      "learning_rate": 4.633236469339603e-06,
+      "loss": 0.6261,
+      "step": 1325
+    },
+    {
+      "epoch": 0.558963009800822,
+      "grad_norm": 0.28161755204200745,
+      "learning_rate": 4.632652547563284e-06,
+      "loss": 0.6213,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5593845505321952,
+      "grad_norm": 0.28579044342041016,
+      "learning_rate": 4.632068198184024e-06,
+      "loss": 0.5848,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5598060912635684,
+      "grad_norm": 0.28025031089782715,
+      "learning_rate": 4.631483421318989e-06,
+      "loss": 0.6444,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5602276319949415,
+      "grad_norm": 0.26460105180740356,
+      "learning_rate": 4.630898217085425e-06,
+      "loss": 0.6134,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5606491727263146,
+      "grad_norm": 0.2625086307525635,
+      "learning_rate": 4.630312585600668e-06,
+      "loss": 0.6358,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5610707134576879,
+      "grad_norm": 0.27133291959762573,
+      "learning_rate": 4.62972652698214e-06,
+      "loss": 0.6357,
+      "step": 1331
+    },
+    {
+      "epoch": 0.561492254189061,
+      "grad_norm": 0.267043799161911,
+      "learning_rate": 4.629140041347347e-06,
+      "loss": 0.623,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5619137949204341,
+      "grad_norm": 0.27161791920661926,
+      "learning_rate": 4.628553128813882e-06,
+      "loss": 0.6226,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5623353356518074,
+      "grad_norm": 0.27926748991012573,
+      "learning_rate": 4.62796578949942e-06,
+      "loss": 0.5837,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5627568763831805,
+      "grad_norm": 0.29721203446388245,
+      "learning_rate": 4.627378023521726e-06,
+      "loss": 0.6374,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5631784171145537,
+      "grad_norm": 0.2746407091617584,
+      "learning_rate": 4.626789830998649e-06,
+      "loss": 0.6241,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5635999578459269,
+      "grad_norm": 0.2687798738479614,
+      "learning_rate": 4.626201212048123e-06,
+      "loss": 0.6414,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5640214985773,
+      "grad_norm": 0.27230778336524963,
+      "learning_rate": 4.625612166788167e-06,
+      "loss": 0.5958,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5644430393086732,
+      "grad_norm": 0.274969220161438,
+      "learning_rate": 4.6250226953368875e-06,
+      "loss": 0.6172,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5648645800400464,
+      "grad_norm": 0.283153235912323,
+      "learning_rate": 4.624432797812474e-06,
+      "loss": 0.6026,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5652861207714195,
+      "grad_norm": 0.27577412128448486,
+      "learning_rate": 4.623842474333202e-06,
+      "loss": 0.6189,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5657076615027927,
+      "grad_norm": 0.27123355865478516,
+      "learning_rate": 4.623251725017436e-06,
+      "loss": 0.6146,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5661292022341659,
+      "grad_norm": 0.2641772925853729,
+      "learning_rate": 4.622660549983621e-06,
+      "loss": 0.6335,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5665507429655391,
+      "grad_norm": 0.2624492943286896,
+      "learning_rate": 4.622068949350289e-06,
+      "loss": 0.6314,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5669722836969122,
+      "grad_norm": 0.2726256847381592,
+      "learning_rate": 4.621476923236058e-06,
+      "loss": 0.6422,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5673938244282853,
+      "grad_norm": 0.26924243569374084,
+      "learning_rate": 4.62088447175963e-06,
+      "loss": 0.6421,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5678153651596586,
+      "grad_norm": 0.25986015796661377,
+      "learning_rate": 4.620291595039797e-06,
+      "loss": 0.5958,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5682369058910317,
+      "grad_norm": 0.2508310377597809,
+      "learning_rate": 4.619698293195427e-06,
+      "loss": 0.6323,
+      "step": 1348
+    },
+    {
+      "epoch": 0.568658446622405,
+      "grad_norm": 0.28102511167526245,
+      "learning_rate": 4.619104566345483e-06,
+      "loss": 0.5965,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5690799873537781,
+      "grad_norm": 0.2653999924659729,
+      "learning_rate": 4.618510414609006e-06,
+      "loss": 0.6317,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5695015280851512,
+      "grad_norm": 0.2717050313949585,
+      "learning_rate": 4.617915838105128e-06,
+      "loss": 0.6244,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5699230688165244,
+      "grad_norm": 0.27203893661499023,
+      "learning_rate": 4.617320836953061e-06,
+      "loss": 0.6342,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5703446095478976,
+      "grad_norm": 0.28115665912628174,
+      "learning_rate": 4.616725411272105e-06,
+      "loss": 0.5911,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5707661502792707,
+      "grad_norm": 0.2593139708042145,
+      "learning_rate": 4.616129561181645e-06,
+      "loss": 0.6203,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5711876910106439,
+      "grad_norm": 0.2681427597999573,
+      "learning_rate": 4.615533286801151e-06,
+      "loss": 0.6269,
+      "step": 1355
+    },
+    {
+      "epoch": 0.571609231742017,
+      "grad_norm": 0.2708660662174225,
+      "learning_rate": 4.614936588250176e-06,
+      "loss": 0.6163,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5720307724733903,
+      "grad_norm": 0.2663513422012329,
+      "learning_rate": 4.614339465648361e-06,
+      "loss": 0.6329,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5724523132047634,
+      "grad_norm": 0.2724343538284302,
+      "learning_rate": 4.613741919115432e-06,
+      "loss": 0.5891,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5728738539361365,
+      "grad_norm": 0.2695017457008362,
+      "learning_rate": 4.613143948771196e-06,
+      "loss": 0.6104,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5732953946675098,
+      "grad_norm": 0.27143579721450806,
+      "learning_rate": 4.61254555473555e-06,
+      "loss": 0.6004,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5737169353988829,
+      "grad_norm": 0.2669190764427185,
+      "learning_rate": 4.611946737128472e-06,
+      "loss": 0.6345,
+      "step": 1361
+    },
+    {
+      "epoch": 0.574138476130256,
+      "grad_norm": 0.27886879444122314,
+      "learning_rate": 4.611347496070029e-06,
+      "loss": 0.6092,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5745600168616293,
+      "grad_norm": 0.26956266164779663,
+      "learning_rate": 4.6107478316803675e-06,
+      "loss": 0.6073,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5749815575930024,
+      "grad_norm": 0.272056519985199,
+      "learning_rate": 4.610147744079725e-06,
+      "loss": 0.6121,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5754030983243756,
+      "grad_norm": 0.27157264947891235,
+      "learning_rate": 4.609547233388419e-06,
+      "loss": 0.6248,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5758246390557488,
+      "grad_norm": 0.2761085331439972,
+      "learning_rate": 4.608946299726854e-06,
+      "loss": 0.6282,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5762461797871219,
+      "grad_norm": 0.27115628123283386,
+      "learning_rate": 4.608344943215519e-06,
+      "loss": 0.5872,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5766677205184951,
+      "grad_norm": 0.28046920895576477,
+      "learning_rate": 4.607743163974987e-06,
+      "loss": 0.6274,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5770892612498683,
+      "grad_norm": 0.2628866732120514,
+      "learning_rate": 4.607140962125918e-06,
+      "loss": 0.6123,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5775108019812414,
+      "grad_norm": 0.2672618329524994,
+      "learning_rate": 4.606538337789053e-06,
+      "loss": 0.6141,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5779323427126146,
+      "grad_norm": 0.29874375462532043,
+      "learning_rate": 4.605935291085222e-06,
+      "loss": 0.6231,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5783538834439877,
+      "grad_norm": 0.2824569046497345,
+      "learning_rate": 4.605331822135336e-06,
+      "loss": 0.6285,
+      "step": 1372
+    },
+    {
+      "epoch": 0.578775424175361,
+      "grad_norm": 0.2896886467933655,
+      "learning_rate": 4.604727931060393e-06,
+      "loss": 0.6215,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5791969649067341,
+      "grad_norm": 0.2804042100906372,
+      "learning_rate": 4.604123617981477e-06,
+      "loss": 0.6106,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5796185056381072,
+      "grad_norm": 0.26808956265449524,
+      "learning_rate": 4.6035188830197495e-06,
+      "loss": 0.6223,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5800400463694805,
+      "grad_norm": 0.27078598737716675,
+      "learning_rate": 4.602913726296466e-06,
+      "loss": 0.6171,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5804615871008536,
+      "grad_norm": 0.2836013436317444,
+      "learning_rate": 4.60230814793296e-06,
+      "loss": 0.6154,
+      "step": 1377
+    },
+    {
+      "epoch": 0.5808831278322268,
+      "grad_norm": 0.27606114745140076,
+      "learning_rate": 4.601702148050653e-06,
+      "loss": 0.637,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5813046685636,
+      "grad_norm": 0.25860852003097534,
+      "learning_rate": 4.601095726771049e-06,
+      "loss": 0.6156,
+      "step": 1379
+    },
+    {
+      "epoch": 0.5817262092949731,
+      "grad_norm": 0.2803390622138977,
+      "learning_rate": 4.600488884215737e-06,
+      "loss": 0.6103,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5821477500263463,
+      "grad_norm": 0.27952250838279724,
+      "learning_rate": 4.599881620506391e-06,
+      "loss": 0.6215,
+      "step": 1381
+    },
+    {
+      "epoch": 0.5825692907577195,
+      "grad_norm": 0.2727219760417938,
+      "learning_rate": 4.599273935764768e-06,
+      "loss": 0.6532,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5829908314890926,
+      "grad_norm": 0.2749989628791809,
+      "learning_rate": 4.598665830112713e-06,
+      "loss": 0.6342,
+      "step": 1383
+    },
+    {
+      "epoch": 0.5834123722204658,
+      "grad_norm": 0.27826014161109924,
+      "learning_rate": 4.598057303672151e-06,
+      "loss": 0.6182,
+      "step": 1384
+    },
+    {
+      "epoch": 0.583833912951839,
+      "grad_norm": 0.26101773977279663,
+      "learning_rate": 4.597448356565094e-06,
+      "loss": 0.6149,
+      "step": 1385
+    },
+    {
+      "epoch": 0.5842554536832122,
+      "grad_norm": 0.27540817856788635,
+      "learning_rate": 4.596838988913637e-06,
+      "loss": 0.6178,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5846769944145853,
+      "grad_norm": 0.34848058223724365,
+      "learning_rate": 4.59622920083996e-06,
+      "loss": 0.5944,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5850985351459584,
+      "grad_norm": 0.26726233959198,
+      "learning_rate": 4.595618992466328e-06,
+      "loss": 0.5768,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5855200758773317,
+      "grad_norm": 0.30718880891799927,
+      "learning_rate": 4.595008363915089e-06,
+      "loss": 0.6099,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5859416166087048,
+      "grad_norm": 0.29525643587112427,
+      "learning_rate": 4.594397315308676e-06,
+      "loss": 0.6279,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5863631573400779,
+      "grad_norm": 0.27441391348838806,
+      "learning_rate": 4.593785846769606e-06,
+      "loss": 0.6305,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5867846980714512,
+      "grad_norm": 0.30598530173301697,
+      "learning_rate": 4.593173958420481e-06,
+      "loss": 0.6165,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5872062388028243,
+      "grad_norm": 0.28334683179855347,
+      "learning_rate": 4.592561650383984e-06,
+      "loss": 0.6385,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5876277795341975,
+      "grad_norm": 0.26329177618026733,
+      "learning_rate": 4.5919489227828865e-06,
+      "loss": 0.6177,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5880493202655707,
+      "grad_norm": 0.2867093086242676,
+      "learning_rate": 4.591335775740042e-06,
+      "loss": 0.6081,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5884708609969438,
+      "grad_norm": 0.28705722093582153,
+      "learning_rate": 4.590722209378387e-06,
+      "loss": 0.6131,
+      "step": 1396
+    },
+    {
+      "epoch": 0.588892401728317,
+      "grad_norm": 0.2669881582260132,
+      "learning_rate": 4.5901082238209445e-06,
+      "loss": 0.6209,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5893139424596902,
+      "grad_norm": 0.26534005999565125,
+      "learning_rate": 4.589493819190821e-06,
+      "loss": 0.6578,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5897354831910633,
+      "grad_norm": 0.2904300391674042,
+      "learning_rate": 4.588878995611205e-06,
+      "loss": 0.6183,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5901570239224365,
+      "grad_norm": 0.26092708110809326,
+      "learning_rate": 4.58826375320537e-06,
+      "loss": 0.6138,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5905785646538096,
+      "grad_norm": 0.2678774297237396,
+      "learning_rate": 4.587648092096675e-06,
+      "loss": 0.6136,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5910001053851829,
+      "grad_norm": 0.29587626457214355,
+      "learning_rate": 4.587032012408561e-06,
+      "loss": 0.6095,
+      "step": 1402
+    },
+    {
+      "epoch": 0.591421646116556,
+      "grad_norm": 0.2828860282897949,
+      "learning_rate": 4.586415514264555e-06,
+      "loss": 0.595,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5918431868479291,
+      "grad_norm": 0.28194284439086914,
+      "learning_rate": 4.585798597788266e-06,
+      "loss": 0.6387,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5922647275793024,
+      "grad_norm": 0.32987740635871887,
+      "learning_rate": 4.585181263103387e-06,
+      "loss": 0.6452,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5926862683106755,
+      "grad_norm": 0.2724097669124603,
+      "learning_rate": 4.584563510333696e-06,
+      "loss": 0.6049,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5931078090420487,
+      "grad_norm": 0.26263338327407837,
+      "learning_rate": 4.583945339603053e-06,
+      "loss": 0.6346,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5935293497734219,
+      "grad_norm": 0.29090073704719543,
+      "learning_rate": 4.583326751035405e-06,
+      "loss": 0.6365,
+      "step": 1408
+    },
+    {
+      "epoch": 0.593950890504795,
+      "grad_norm": 0.32534724473953247,
+      "learning_rate": 4.5827077447547795e-06,
+      "loss": 0.614,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5943724312361682,
+      "grad_norm": 0.2802751958370209,
+      "learning_rate": 4.58208832088529e-06,
+      "loss": 0.6399,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5947939719675414,
+      "grad_norm": 0.3451883792877197,
+      "learning_rate": 4.581468479551131e-06,
+      "loss": 0.6064,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5952155126989145,
+      "grad_norm": 0.26382121443748474,
+      "learning_rate": 4.580848220876584e-06,
+      "loss": 0.6242,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5956370534302877,
+      "grad_norm": 0.2767588198184967,
+      "learning_rate": 4.580227544986011e-06,
+      "loss": 0.6323,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5960585941616608,
+      "grad_norm": 0.2695329785346985,
+      "learning_rate": 4.579606452003862e-06,
+      "loss": 0.6181,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5964801348930341,
+      "grad_norm": 0.2818808853626251,
+      "learning_rate": 4.578984942054665e-06,
+      "loss": 0.6354,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5969016756244072,
+      "grad_norm": 0.2704218029975891,
+      "learning_rate": 4.578363015263036e-06,
+      "loss": 0.6338,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5973232163557803,
+      "grad_norm": 0.27999523282051086,
+      "learning_rate": 4.577740671753675e-06,
+      "loss": 0.6065,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5977447570871536,
+      "grad_norm": 0.277739942073822,
+      "learning_rate": 4.57711791165136e-06,
+      "loss": 0.6323,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5981662978185267,
+      "grad_norm": 0.2799640893936157,
+      "learning_rate": 4.576494735080957e-06,
+      "loss": 0.5866,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5985878385498998,
+      "grad_norm": 0.271397203207016,
+      "learning_rate": 4.575871142167417e-06,
+      "loss": 0.6213,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5990093792812731,
+      "grad_norm": 0.27047935128211975,
+      "learning_rate": 4.57524713303577e-06,
+      "loss": 0.6331,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5994309200126462,
+      "grad_norm": 0.2855691909790039,
+      "learning_rate": 4.574622707811132e-06,
+      "loss": 0.6189,
+      "step": 1422
+    },
+    {
+      "epoch": 0.5998524607440194,
+      "grad_norm": 0.26830536127090454,
+      "learning_rate": 4.573997866618703e-06,
+      "loss": 0.63,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6002740014753926,
+      "grad_norm": 0.27421873807907104,
+      "learning_rate": 4.5733726095837634e-06,
+      "loss": 0.649,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6006955422067657,
+      "grad_norm": 0.32462403178215027,
+      "learning_rate": 4.5727469368316825e-06,
+      "loss": 0.6303,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6011170829381389,
+      "grad_norm": 0.2736404240131378,
+      "learning_rate": 4.572120848487906e-06,
+      "loss": 0.6275,
+      "step": 1426
+    },
+    {
+      "epoch": 0.601538623669512,
+      "grad_norm": 0.2630680203437805,
+      "learning_rate": 4.571494344677968e-06,
+      "loss": 0.6186,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6019601644008853,
+      "grad_norm": 0.35493215918540955,
+      "learning_rate": 4.570867425527484e-06,
+      "loss": 0.6075,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6023817051322584,
+      "grad_norm": 0.27686911821365356,
+      "learning_rate": 4.570240091162154e-06,
+      "loss": 0.6367,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6028032458636315,
+      "grad_norm": 0.2893412113189697,
+      "learning_rate": 4.569612341707759e-06,
+      "loss": 0.6339,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6032247865950048,
+      "grad_norm": 0.3259349763393402,
+      "learning_rate": 4.568984177290164e-06,
+      "loss": 0.6071,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6036463273263779,
+      "grad_norm": 0.28279799222946167,
+      "learning_rate": 4.56835559803532e-06,
+      "loss": 0.6377,
+      "step": 1432
+    },
+    {
+      "epoch": 0.604067868057751,
+      "grad_norm": 0.28336918354034424,
+      "learning_rate": 4.567726604069257e-06,
+      "loss": 0.6047,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6044894087891243,
+      "grad_norm": 0.28855544328689575,
+      "learning_rate": 4.567097195518092e-06,
+      "loss": 0.6104,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6049109495204974,
+      "grad_norm": 0.2689586877822876,
+      "learning_rate": 4.566467372508021e-06,
+      "loss": 0.6279,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6053324902518706,
+      "grad_norm": 0.2705938518047333,
+      "learning_rate": 4.5658371351653265e-06,
+      "loss": 0.5944,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6057540309832438,
+      "grad_norm": 0.2617529332637787,
+      "learning_rate": 4.565206483616373e-06,
+      "loss": 0.6149,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6061755717146169,
+      "grad_norm": 0.2758113145828247,
+      "learning_rate": 4.564575417987608e-06,
+      "loss": 0.6355,
+      "step": 1438
+    },
+    {
+      "epoch": 0.6065971124459901,
+      "grad_norm": 0.2844107151031494,
+      "learning_rate": 4.5639439384055615e-06,
+      "loss": 0.6108,
+      "step": 1439
+    },
+    {
+      "epoch": 0.6070186531773633,
+      "grad_norm": 0.2876218855381012,
+      "learning_rate": 4.563312044996847e-06,
+      "loss": 0.6357,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6074401939087364,
+      "grad_norm": 0.26658907532691956,
+      "learning_rate": 4.5626797378881615e-06,
+      "loss": 0.634,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6078617346401096,
+      "grad_norm": 0.26826658844947815,
+      "learning_rate": 4.562047017206283e-06,
+      "loss": 0.5936,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6082832753714827,
+      "grad_norm": 0.28318578004837036,
+      "learning_rate": 4.561413883078077e-06,
+      "loss": 0.6081,
+      "step": 1443
+    },
+    {
+      "epoch": 0.608704816102856,
+      "grad_norm": 0.2692621052265167,
+      "learning_rate": 4.560780335630484e-06,
+      "loss": 0.5849,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6091263568342291,
+      "grad_norm": 0.2663171887397766,
+      "learning_rate": 4.560146374990536e-06,
+      "loss": 0.6154,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6095478975656022,
+      "grad_norm": 0.26806384325027466,
+      "learning_rate": 4.5595120012853425e-06,
+      "loss": 0.6142,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6099694382969755,
+      "grad_norm": 0.26338180899620056,
+      "learning_rate": 4.558877214642095e-06,
+      "loss": 0.6183,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6103909790283486,
+      "grad_norm": 0.2887309789657593,
+      "learning_rate": 4.558242015188075e-06,
+      "loss": 0.6238,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6108125197597217,
+      "grad_norm": 0.28384050726890564,
+      "learning_rate": 4.557606403050638e-06,
+      "loss": 0.5964,
+      "step": 1449
+    },
+    {
+      "epoch": 0.611234060491095,
+      "grad_norm": 0.2647939920425415,
+      "learning_rate": 4.556970378357227e-06,
+      "loss": 0.5983,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6116556012224681,
+      "grad_norm": 0.2683292627334595,
+      "learning_rate": 4.556333941235368e-06,
+      "loss": 0.6309,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6120771419538413,
+      "grad_norm": 0.29334157705307007,
+      "learning_rate": 4.555697091812666e-06,
+      "loss": 0.6013,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6124986826852145,
+      "grad_norm": 0.2800478935241699,
+      "learning_rate": 4.555059830216813e-06,
+      "loss": 0.6178,
+      "step": 1453
+    },
+    {
+      "epoch": 0.6129202234165876,
+      "grad_norm": 0.2783694863319397,
+      "learning_rate": 4.554422156575581e-06,
+      "loss": 0.6373,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6133417641479608,
+      "grad_norm": 0.2783585786819458,
+      "learning_rate": 4.553784071016826e-06,
+      "loss": 0.6337,
+      "step": 1455
+    },
+    {
+      "epoch": 0.613763304879334,
+      "grad_norm": 0.27096617221832275,
+      "learning_rate": 4.553145573668486e-06,
+      "loss": 0.6279,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6141848456107072,
+      "grad_norm": 0.267915815114975,
+      "learning_rate": 4.5525066646585805e-06,
+      "loss": 0.6164,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6146063863420803,
+      "grad_norm": 0.2673802971839905,
+      "learning_rate": 4.551867344115214e-06,
+      "loss": 0.6355,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6150279270734534,
+      "grad_norm": 0.2651333212852478,
+      "learning_rate": 4.551227612166571e-06,
+      "loss": 0.6326,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6154494678048267,
+      "grad_norm": 0.3178790211677551,
+      "learning_rate": 4.550587468940921e-06,
+      "loss": 0.6102,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6158710085361998,
+      "grad_norm": 0.2939439117908478,
+      "learning_rate": 4.549946914566612e-06,
+      "loss": 0.6204,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6162925492675729,
+      "grad_norm": 0.2602866590023041,
+      "learning_rate": 4.549305949172078e-06,
+      "loss": 0.6299,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6167140899989462,
+      "grad_norm": 0.26795539259910583,
+      "learning_rate": 4.5486645728858355e-06,
+      "loss": 0.6131,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6171356307303193,
+      "grad_norm": 0.28208860754966736,
+      "learning_rate": 4.548022785836483e-06,
+      "loss": 0.59,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6175571714616925,
+      "grad_norm": 0.2706642150878906,
+      "learning_rate": 4.547380588152698e-06,
+      "loss": 0.6256,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6179787121930657,
+      "grad_norm": 0.28007298707962036,
+      "learning_rate": 4.546737979963244e-06,
+      "loss": 0.6228,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6184002529244388,
+      "grad_norm": 0.27777689695358276,
+      "learning_rate": 4.546094961396967e-06,
+      "loss": 0.5908,
+      "step": 1467
+    },
+    {
+      "epoch": 0.618821793655812,
+      "grad_norm": 0.2847370207309723,
+      "learning_rate": 4.545451532582793e-06,
+      "loss": 0.6218,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6192433343871852,
+      "grad_norm": 0.28691092133522034,
+      "learning_rate": 4.544807693649731e-06,
+      "loss": 0.636,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6196648751185583,
+      "grad_norm": 0.2734300494194031,
+      "learning_rate": 4.544163444726873e-06,
+      "loss": 0.6121,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6200864158499315,
+      "grad_norm": 0.27876222133636475,
+      "learning_rate": 4.543518785943394e-06,
+      "loss": 0.6277,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6205079565813046,
+      "grad_norm": 0.28242984414100647,
+      "learning_rate": 4.5428737174285474e-06,
+      "loss": 0.5956,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6209294973126779,
+      "grad_norm": 0.273242324590683,
+      "learning_rate": 4.5422282393116725e-06,
+      "loss": 0.6177,
+      "step": 1473
+    },
+    {
+      "epoch": 0.621351038044051,
+      "grad_norm": 0.27772805094718933,
+      "learning_rate": 4.54158235172219e-06,
+      "loss": 0.6294,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6217725787754241,
+      "grad_norm": 0.2782262861728668,
+      "learning_rate": 4.540936054789602e-06,
+      "loss": 0.6131,
+      "step": 1475
+    },
+    {
+      "epoch": 0.6221941195067974,
+      "grad_norm": 0.2707035541534424,
+      "learning_rate": 4.540289348643493e-06,
+      "loss": 0.6204,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6226156602381705,
+      "grad_norm": 0.2733548581600189,
+      "learning_rate": 4.539642233413528e-06,
+      "loss": 0.6325,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6230372009695437,
+      "grad_norm": 0.2936704456806183,
+      "learning_rate": 4.538994709229457e-06,
+      "loss": 0.5916,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6234587417009169,
+      "grad_norm": 0.26808828115463257,
+      "learning_rate": 4.538346776221111e-06,
+      "loss": 0.6068,
+      "step": 1479
+    },
+    {
+      "epoch": 0.62388028243229,
+      "grad_norm": 0.2771219313144684,
+      "learning_rate": 4.537698434518401e-06,
+      "loss": 0.6029,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6243018231636632,
+      "grad_norm": 0.2813969552516937,
+      "learning_rate": 4.537049684251322e-06,
+      "loss": 0.6189,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6247233638950364,
+      "grad_norm": 0.29234641790390015,
+      "learning_rate": 4.536400525549948e-06,
+      "loss": 0.6017,
+      "step": 1482
+    },
+    {
+      "epoch": 0.6251449046264095,
+      "grad_norm": 0.27003294229507446,
+      "learning_rate": 4.5357509585444415e-06,
+      "loss": 0.6131,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6255664453577827,
+      "grad_norm": 0.28071534633636475,
+      "learning_rate": 4.53510098336504e-06,
+      "loss": 0.627,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6259879860891558,
+      "grad_norm": 0.27603864669799805,
+      "learning_rate": 4.534450600142064e-06,
+      "loss": 0.6086,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6264095268205291,
+      "grad_norm": 0.27458807826042175,
+      "learning_rate": 4.533799809005921e-06,
+      "loss": 0.5987,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6268310675519022,
+      "grad_norm": 0.28000253438949585,
+      "learning_rate": 4.533148610087093e-06,
+      "loss": 0.5872,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6272526082832753,
+      "grad_norm": 0.27066949009895325,
+      "learning_rate": 4.53249700351615e-06,
+      "loss": 0.5847,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6276741490146486,
+      "grad_norm": 0.27049657702445984,
+      "learning_rate": 4.531844989423738e-06,
+      "loss": 0.6182,
+      "step": 1489
+    },
+    {
+      "epoch": 0.6280956897460217,
+      "grad_norm": 0.2775096893310547,
+      "learning_rate": 4.531192567940591e-06,
+      "loss": 0.6314,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6285172304773948,
+      "grad_norm": 0.274361789226532,
+      "learning_rate": 4.530539739197519e-06,
+      "loss": 0.5897,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6289387712087681,
+      "grad_norm": 0.27788275480270386,
+      "learning_rate": 4.529886503325416e-06,
+      "loss": 0.627,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6293603119401412,
+      "grad_norm": 0.27773517370224,
+      "learning_rate": 4.529232860455259e-06,
+      "loss": 0.5992,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6297818526715144,
+      "grad_norm": 0.2709919214248657,
+      "learning_rate": 4.528578810718105e-06,
+      "loss": 0.6278,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6302033934028876,
+      "grad_norm": 0.26856496930122375,
+      "learning_rate": 4.527924354245093e-06,
+      "loss": 0.6096,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6306249341342607,
+      "grad_norm": 0.273642361164093,
+      "learning_rate": 4.527269491167441e-06,
+      "loss": 0.601,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6310464748656339,
+      "grad_norm": 0.2719555199146271,
+      "learning_rate": 4.5266142216164534e-06,
+      "loss": 0.5962,
+      "step": 1497
+    },
+    {
+      "epoch": 0.631468015597007,
+      "grad_norm": 0.2666900157928467,
+      "learning_rate": 4.525958545723514e-06,
+      "loss": 0.6321,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6318895563283802,
+      "grad_norm": 0.27161481976509094,
+      "learning_rate": 4.5253024636200855e-06,
+      "loss": 0.6045,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6323110970597534,
+      "grad_norm": 0.269593745470047,
+      "learning_rate": 4.524645975437717e-06,
+      "loss": 0.6193,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6327326377911265,
+      "grad_norm": 0.26981121301651,
+      "learning_rate": 4.523989081308034e-06,
+      "loss": 0.6238,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6331541785224998,
+      "grad_norm": 0.2852029502391815,
+      "learning_rate": 4.523331781362746e-06,
+      "loss": 0.5843,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6335757192538729,
+      "grad_norm": 0.26800987124443054,
+      "learning_rate": 4.522674075733644e-06,
+      "loss": 0.6485,
+      "step": 1503
+    },
+    {
+      "epoch": 0.633997259985246,
+      "grad_norm": 0.26402559876441956,
+      "learning_rate": 4.5220159645526e-06,
+      "loss": 0.637,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6344188007166193,
+      "grad_norm": 0.28142592310905457,
+      "learning_rate": 4.521357447951567e-06,
+      "loss": 0.638,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6348403414479924,
+      "grad_norm": 0.2762514054775238,
+      "learning_rate": 4.520698526062579e-06,
+      "loss": 0.6063,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6352618821793656,
+      "grad_norm": 0.27199673652648926,
+      "learning_rate": 4.520039199017752e-06,
+      "loss": 0.6026,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6356834229107388,
+      "grad_norm": 0.2824305593967438,
+      "learning_rate": 4.519379466949283e-06,
+      "loss": 0.6126,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6361049636421119,
+      "grad_norm": 0.27216824889183044,
+      "learning_rate": 4.51871932998945e-06,
+      "loss": 0.6105,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6365265043734851,
+      "grad_norm": 0.2837934195995331,
+      "learning_rate": 4.518058788270614e-06,
+      "loss": 0.5923,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6369480451048583,
+      "grad_norm": 0.27647504210472107,
+      "learning_rate": 4.5173978419252135e-06,
+      "loss": 0.6132,
+      "step": 1511
+    },
+    {
+      "epoch": 0.6373695858362314,
+      "grad_norm": 0.28019583225250244,
+      "learning_rate": 4.51673649108577e-06,
+      "loss": 0.6115,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6377911265676046,
+      "grad_norm": 0.28185561299324036,
+      "learning_rate": 4.516074735884888e-06,
+      "loss": 0.6146,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6382126672989777,
+      "grad_norm": 0.29335102438926697,
+      "learning_rate": 4.5154125764552504e-06,
+      "loss": 0.5891,
+      "step": 1514
+    },
+    {
+      "epoch": 0.638634208030351,
+      "grad_norm": 0.28951752185821533,
+      "learning_rate": 4.514750012929621e-06,
+      "loss": 0.6138,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6390557487617241,
+      "grad_norm": 0.2925238013267517,
+      "learning_rate": 4.514087045440847e-06,
+      "loss": 0.6185,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6394772894930972,
+      "grad_norm": 0.29035359621047974,
+      "learning_rate": 4.513423674121855e-06,
+      "loss": 0.6293,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6398988302244705,
+      "grad_norm": 0.29539164900779724,
+      "learning_rate": 4.512759899105653e-06,
+      "loss": 0.6239,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6403203709558436,
+      "grad_norm": 0.2987770438194275,
+      "learning_rate": 4.512095720525329e-06,
+      "loss": 0.6182,
+      "step": 1519
+    },
+    {
+      "epoch": 0.6407419116872167,
+      "grad_norm": 0.2673036456108093,
+      "learning_rate": 4.511431138514054e-06,
+      "loss": 0.6192,
+      "step": 1520
+    },
+    {
+      "epoch": 0.64116345241859,
+      "grad_norm": 0.30469295382499695,
+      "learning_rate": 4.510766153205077e-06,
+      "loss": 0.6167,
+      "step": 1521
+    },
+    {
+      "epoch": 0.6415849931499631,
+      "grad_norm": 0.2974946200847626,
+      "learning_rate": 4.510100764731731e-06,
+      "loss": 0.6141,
+      "step": 1522
+    },
+    {
+      "epoch": 0.6420065338813363,
+      "grad_norm": 0.2762567400932312,
+      "learning_rate": 4.509434973227427e-06,
+      "loss": 0.6023,
+      "step": 1523
+    },
+    {
+      "epoch": 0.6424280746127095,
+      "grad_norm": 0.30400657653808594,
+      "learning_rate": 4.508768778825658e-06,
+      "loss": 0.6145,
+      "step": 1524
+    },
+    {
+      "epoch": 0.6428496153440826,
+      "grad_norm": 0.3083377480506897,
+      "learning_rate": 4.50810218166e-06,
+      "loss": 0.5834,
+      "step": 1525
+    },
+    {
+      "epoch": 0.6432711560754558,
+      "grad_norm": 0.2626950144767761,
+      "learning_rate": 4.507435181864106e-06,
+      "loss": 0.6237,
+      "step": 1526
+    },
+    {
+      "epoch": 0.643692696806829,
+      "grad_norm": 0.27182537317276,
+      "learning_rate": 4.5067677795717116e-06,
+      "loss": 0.5978,
+      "step": 1527
+    },
+    {
+      "epoch": 0.6441142375382022,
+      "grad_norm": 0.2731630504131317,
+      "learning_rate": 4.506099974916632e-06,
+      "loss": 0.6393,
+      "step": 1528
+    },
+    {
+      "epoch": 0.6445357782695753,
+      "grad_norm": 0.28954973816871643,
+      "learning_rate": 4.505431768032767e-06,
+      "loss": 0.6282,
+      "step": 1529
+    },
+    {
+      "epoch": 0.6449573190009484,
+      "grad_norm": 0.2655770480632782,
+      "learning_rate": 4.504763159054089e-06,
+      "loss": 0.599,
+      "step": 1530
+    },
+    {
+      "epoch": 0.6453788597323217,
+      "grad_norm": 0.2699226140975952,
+      "learning_rate": 4.5040941481146605e-06,
+      "loss": 0.6117,
+      "step": 1531
+    },
+    {
+      "epoch": 0.6458004004636948,
+      "grad_norm": 0.2721829116344452,
+      "learning_rate": 4.503424735348618e-06,
+      "loss": 0.6175,
+      "step": 1532
+    },
+    {
+      "epoch": 0.6462219411950679,
+      "grad_norm": 0.284728467464447,
+      "learning_rate": 4.502754920890181e-06,
+      "loss": 0.6212,
+      "step": 1533
+    },
+    {
+      "epoch": 0.6466434819264412,
+      "grad_norm": 0.26562049984931946,
+      "learning_rate": 4.502084704873649e-06,
+      "loss": 0.6224,
+      "step": 1534
+    },
+    {
+      "epoch": 0.6470650226578143,
+      "grad_norm": 0.2700594961643219,
+      "learning_rate": 4.501414087433404e-06,
+      "loss": 0.6279,
+      "step": 1535
+    },
+    {
+      "epoch": 0.6474865633891875,
+      "grad_norm": 0.31072473526000977,
+      "learning_rate": 4.500743068703905e-06,
+      "loss": 0.6061,
+      "step": 1536
+    },
+    {
+      "epoch": 0.6479081041205607,
+      "grad_norm": 0.29663199186325073,
+      "learning_rate": 4.5000716488196915e-06,
+      "loss": 0.5961,
+      "step": 1537
+    },
+    {
+      "epoch": 0.6483296448519338,
+      "grad_norm": 0.3037363886833191,
+      "learning_rate": 4.4993998279153885e-06,
+      "loss": 0.6213,
+      "step": 1538
+    },
+    {
+      "epoch": 0.648751185583307,
+      "grad_norm": 0.30323055386543274,
+      "learning_rate": 4.4987276061256965e-06,
+      "loss": 0.6202,
+      "step": 1539
+    },
+    {
+      "epoch": 0.6491727263146801,
+      "grad_norm": 0.2785603404045105,
+      "learning_rate": 4.498054983585398e-06,
+      "loss": 0.5867,
+      "step": 1540
+    },
+    {
+      "epoch": 0.6495942670460533,
+      "grad_norm": 0.2782061696052551,
+      "learning_rate": 4.497381960429355e-06,
+      "loss": 0.6235,
+      "step": 1541
+    },
+    {
+      "epoch": 0.6500158077774265,
+      "grad_norm": 0.29752299189567566,
+      "learning_rate": 4.496708536792512e-06,
+      "loss": 0.6339,
+      "step": 1542
+    },
+    {
+      "epoch": 0.6504373485087996,
+      "grad_norm": 0.2694087326526642,
+      "learning_rate": 4.496034712809891e-06,
+      "loss": 0.6221,
+      "step": 1543
+    },
+    {
+      "epoch": 0.6508588892401729,
+      "grad_norm": 0.2693852484226227,
+      "learning_rate": 4.495360488616598e-06,
+      "loss": 0.6295,
+      "step": 1544
+    },
+    {
+      "epoch": 0.651280429971546,
+      "grad_norm": 0.290547639131546,
+      "learning_rate": 4.494685864347813e-06,
+      "loss": 0.6006,
+      "step": 1545
+    },
+    {
+      "epoch": 0.6517019707029191,
+      "grad_norm": 0.27716904878616333,
+      "learning_rate": 4.494010840138804e-06,
+      "loss": 0.6015,
+      "step": 1546
+    },
+    {
+      "epoch": 0.6521235114342924,
+      "grad_norm": 0.26897433400154114,
+      "learning_rate": 4.493335416124912e-06,
+      "loss": 0.6091,
+      "step": 1547
+    },
+    {
+      "epoch": 0.6525450521656655,
+      "grad_norm": 0.2837153375148773,
+      "learning_rate": 4.4926595924415635e-06,
+      "loss": 0.6033,
+      "step": 1548
+    },
+    {
+      "epoch": 0.6529665928970386,
+      "grad_norm": 0.2710731625556946,
+      "learning_rate": 4.4919833692242635e-06,
+      "loss": 0.6008,
+      "step": 1549
+    },
+    {
+      "epoch": 0.6533881336284119,
+      "grad_norm": 0.28248926997184753,
+      "learning_rate": 4.491306746608595e-06,
+      "loss": 0.6209,
+      "step": 1550
+    },
+    {
+      "epoch": 0.653809674359785,
+      "grad_norm": 0.2770392894744873,
+      "learning_rate": 4.490629724730223e-06,
+      "loss": 0.5995,
+      "step": 1551
+    },
+    {
+      "epoch": 0.6542312150911582,
+      "grad_norm": 0.27291423082351685,
+      "learning_rate": 4.489952303724894e-06,
+      "loss": 0.6107,
+      "step": 1552
+    },
+    {
+      "epoch": 0.6546527558225314,
+      "grad_norm": 0.27929964661598206,
+      "learning_rate": 4.489274483728431e-06,
+      "loss": 0.601,
+      "step": 1553
+    },
+    {
+      "epoch": 0.6550742965539045,
+      "grad_norm": 0.2734281122684479,
+      "learning_rate": 4.48859626487674e-06,
+      "loss": 0.5962,
+      "step": 1554
+    },
+    {
+      "epoch": 0.6554958372852777,
+      "grad_norm": 0.28047844767570496,
+      "learning_rate": 4.487917647305804e-06,
+      "loss": 0.6208,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6559173780166508,
+      "grad_norm": 0.275879830121994,
+      "learning_rate": 4.487238631151691e-06,
+      "loss": 0.6191,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6563389187480241,
+      "grad_norm": 0.26818183064460754,
+      "learning_rate": 4.4865592165505424e-06,
+      "loss": 0.6054,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6567604594793972,
+      "grad_norm": 0.26944100856781006,
+      "learning_rate": 4.485879403638585e-06,
+      "loss": 0.6024,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6571820002107703,
+      "grad_norm": 0.2800596058368683,
+      "learning_rate": 4.485199192552123e-06,
+      "loss": 0.6088,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6576035409421436,
+      "grad_norm": 0.26889878511428833,
+      "learning_rate": 4.484518583427539e-06,
+      "loss": 0.6045,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6580250816735167,
+      "grad_norm": 0.265600323677063,
+      "learning_rate": 4.483837576401299e-06,
+      "loss": 0.6105,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6584466224048898,
+      "grad_norm": 0.28204530477523804,
+      "learning_rate": 4.483156171609946e-06,
+      "loss": 0.6207,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6588681631362631,
+      "grad_norm": 0.2679276764392853,
+      "learning_rate": 4.4824743691901025e-06,
+      "loss": 0.6232,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6592897038676362,
+      "grad_norm": 0.27385732531547546,
+      "learning_rate": 4.481792169278474e-06,
+      "loss": 0.633,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6597112445990094,
+      "grad_norm": 0.2697983682155609,
+      "learning_rate": 4.481109572011842e-06,
+      "loss": 0.6154,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6601327853303826,
+      "grad_norm": 0.27938970923423767,
+      "learning_rate": 4.48042657752707e-06,
+      "loss": 0.6089,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6605543260617557,
+      "grad_norm": 0.26816418766975403,
+      "learning_rate": 4.479743185961101e-06,
+      "loss": 0.6198,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6609758667931289,
+      "grad_norm": 0.269928514957428,
+      "learning_rate": 4.479059397450955e-06,
+      "loss": 0.6093,
+      "step": 1568
+    },
+    {
+      "epoch": 0.661397407524502,
+      "grad_norm": 0.2774163484573364,
+      "learning_rate": 4.478375212133735e-06,
+      "loss": 0.6196,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6618189482558752,
+      "grad_norm": 0.2701868414878845,
+      "learning_rate": 4.477690630146622e-06,
+      "loss": 0.5941,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6622404889872484,
+      "grad_norm": 0.2725176215171814,
+      "learning_rate": 4.4770056516268755e-06,
+      "loss": 0.6138,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6626620297186215,
+      "grad_norm": 0.27007344365119934,
+      "learning_rate": 4.476320276711836e-06,
+      "loss": 0.6425,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6630835704499948,
+      "grad_norm": 0.2673550546169281,
+      "learning_rate": 4.475634505538925e-06,
+      "loss": 0.6205,
+      "step": 1573
+    },
+    {
+      "epoch": 0.6635051111813679,
+      "grad_norm": 0.2659700810909271,
+      "learning_rate": 4.4749483382456395e-06,
+      "loss": 0.5921,
+      "step": 1574
+    },
+    {
+      "epoch": 0.663926651912741,
+      "grad_norm": 0.2815890610218048,
+      "learning_rate": 4.47426177496956e-06,
+      "loss": 0.6092,
+      "step": 1575
+    },
+    {
+      "epoch": 0.6643481926441143,
+      "grad_norm": 0.2716395854949951,
+      "learning_rate": 4.473574815848341e-06,
+      "loss": 0.6044,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6647697333754874,
+      "grad_norm": 0.2714964747428894,
+      "learning_rate": 4.472887461019723e-06,
+      "loss": 0.6357,
+      "step": 1577
+    },
+    {
+      "epoch": 0.6651912741068606,
+      "grad_norm": 0.26573747396469116,
+      "learning_rate": 4.4721997106215224e-06,
+      "loss": 0.5995,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6656128148382338,
+      "grad_norm": 0.28311651945114136,
+      "learning_rate": 4.471511564791634e-06,
+      "loss": 0.616,
+      "step": 1579
+    },
+    {
+      "epoch": 0.6660343555696069,
+      "grad_norm": 0.281512051820755,
+      "learning_rate": 4.470823023668032e-06,
+      "loss": 0.6069,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6664558963009801,
+      "grad_norm": 0.2670106589794159,
+      "learning_rate": 4.470134087388773e-06,
+      "loss": 0.6189,
+      "step": 1581
+    },
+    {
+      "epoch": 0.6668774370323532,
+      "grad_norm": 0.26908373832702637,
+      "learning_rate": 4.4694447560919896e-06,
+      "loss": 0.599,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6672989777637264,
+      "grad_norm": 0.28021809458732605,
+      "learning_rate": 4.468755029915896e-06,
+      "loss": 0.6116,
+      "step": 1583
+    },
+    {
+      "epoch": 0.6677205184950996,
+      "grad_norm": 0.2837699353694916,
+      "learning_rate": 4.468064908998781e-06,
+      "loss": 0.605,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6681420592264727,
+      "grad_norm": 0.2869711220264435,
+      "learning_rate": 4.4673743934790195e-06,
+      "loss": 0.6222,
+      "step": 1585
+    },
+    {
+      "epoch": 0.668563599957846,
+      "grad_norm": 0.27720531821250916,
+      "learning_rate": 4.46668348349506e-06,
+      "loss": 0.6344,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6689851406892191,
+      "grad_norm": 0.2767402231693268,
+      "learning_rate": 4.465992179185432e-06,
+      "loss": 0.6305,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6694066814205922,
+      "grad_norm": 0.2671508193016052,
+      "learning_rate": 4.465300480688745e-06,
+      "loss": 0.6026,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6698282221519655,
+      "grad_norm": 0.28183871507644653,
+      "learning_rate": 4.4646083881436855e-06,
+      "loss": 0.5923,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6702497628833386,
+      "grad_norm": 0.27600982785224915,
+      "learning_rate": 4.4639159016890195e-06,
+      "loss": 0.6335,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6706713036147117,
+      "grad_norm": 0.2945455312728882,
+      "learning_rate": 4.463223021463595e-06,
+      "loss": 0.6129,
+      "step": 1591
+    },
+    {
+      "epoch": 0.671092844346085,
+      "grad_norm": 0.2954849302768707,
+      "learning_rate": 4.462529747606334e-06,
+      "loss": 0.6308,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6715143850774581,
+      "grad_norm": 0.272697776556015,
+      "learning_rate": 4.46183608025624e-06,
+      "loss": 0.6243,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6719359258088313,
+      "grad_norm": 0.28029197454452515,
+      "learning_rate": 4.461142019552397e-06,
+      "loss": 0.624,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6723574665402045,
+      "grad_norm": 0.27908846735954285,
+      "learning_rate": 4.4604475656339645e-06,
+      "loss": 0.6188,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6727790072715776,
+      "grad_norm": 0.2787590026855469,
+      "learning_rate": 4.459752718640183e-06,
+      "loss": 0.6203,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6732005480029508,
+      "grad_norm": 0.3122871220111847,
+      "learning_rate": 4.459057478710372e-06,
+      "loss": 0.6221,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6736220887343239,
+      "grad_norm": 0.3150908350944519,
+      "learning_rate": 4.458361845983929e-06,
+      "loss": 0.6024,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6740436294656971,
+      "grad_norm": 0.28091809153556824,
+      "learning_rate": 4.45766582060033e-06,
+      "loss": 0.6219,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6744651701970703,
+      "grad_norm": 0.3276417553424835,
+      "learning_rate": 4.456969402699129e-06,
+      "loss": 0.5989,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6748867109284434,
+      "grad_norm": 0.28086593747138977,
+      "learning_rate": 4.456272592419961e-06,
+      "loss": 0.6066,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6753082516598167,
+      "grad_norm": 0.29043135046958923,
+      "learning_rate": 4.455575389902539e-06,
+      "loss": 0.6236,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6757297923911898,
+      "grad_norm": 0.2842647135257721,
+      "learning_rate": 4.4548777952866515e-06,
+      "loss": 0.6256,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6761513331225629,
+      "grad_norm": 0.3000682592391968,
+      "learning_rate": 4.454179808712173e-06,
+      "loss": 0.603,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6765728738539362,
+      "grad_norm": 0.26292684674263,
+      "learning_rate": 4.453481430319046e-06,
+      "loss": 0.6529,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6769944145853093,
+      "grad_norm": 0.28136998414993286,
+      "learning_rate": 4.452782660247302e-06,
+      "loss": 0.5994,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6774159553166825,
+      "grad_norm": 0.2886597812175751,
+      "learning_rate": 4.4520834986370435e-06,
+      "loss": 0.6141,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6778374960480557,
+      "grad_norm": 0.2778439521789551,
+      "learning_rate": 4.451383945628456e-06,
+      "loss": 0.6098,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6782590367794288,
+      "grad_norm": 0.2698838412761688,
+      "learning_rate": 4.450684001361802e-06,
+      "loss": 0.6022,
+      "step": 1609
+    },
+    {
+      "epoch": 0.678680577510802,
+      "grad_norm": 0.3036693334579468,
+      "learning_rate": 4.449983665977421e-06,
+      "loss": 0.5953,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6791021182421751,
+      "grad_norm": 0.280348002910614,
+      "learning_rate": 4.449282939615734e-06,
+      "loss": 0.606,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6795236589735483,
+      "grad_norm": 0.2767488658428192,
+      "learning_rate": 4.448581822417236e-06,
+      "loss": 0.6062,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6799451997049215,
+      "grad_norm": 0.29645538330078125,
+      "learning_rate": 4.447880314522506e-06,
+      "loss": 0.5987,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6803667404362946,
+      "grad_norm": 0.2836908996105194,
+      "learning_rate": 4.447178416072197e-06,
+      "loss": 0.6341,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6807882811676679,
+      "grad_norm": 0.26265034079551697,
+      "learning_rate": 4.44647612720704e-06,
+      "loss": 0.5932,
+      "step": 1615
+    },
+    {
+      "epoch": 0.681209821899041,
+      "grad_norm": 0.28913354873657227,
+      "learning_rate": 4.445773448067849e-06,
+      "loss": 0.6349,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6816313626304141,
+      "grad_norm": 0.2771572470664978,
+      "learning_rate": 4.445070378795512e-06,
+      "loss": 0.6177,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6820529033617874,
+      "grad_norm": 0.26578274369239807,
+      "learning_rate": 4.444366919530995e-06,
+      "loss": 0.6289,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6824744440931605,
+      "grad_norm": 0.2910264730453491,
+      "learning_rate": 4.443663070415345e-06,
+      "loss": 0.6338,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6828959848245336,
+      "grad_norm": 0.27525678277015686,
+      "learning_rate": 4.442958831589686e-06,
+      "loss": 0.6194,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6833175255559069,
+      "grad_norm": 0.2677125930786133,
+      "learning_rate": 4.442254203195219e-06,
+      "loss": 0.6259,
+      "step": 1621
+    },
+    {
+      "epoch": 0.68373906628728,
+      "grad_norm": 0.2736016809940338,
+      "learning_rate": 4.441549185373225e-06,
+      "loss": 0.6112,
+      "step": 1622
+    },
+    {
+      "epoch": 0.6841606070186532,
+      "grad_norm": 0.2876531481742859,
+      "learning_rate": 4.440843778265061e-06,
+      "loss": 0.607,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6845821477500263,
+      "grad_norm": 0.28119125962257385,
+      "learning_rate": 4.440137982012163e-06,
+      "loss": 0.6166,
+      "step": 1624
+    },
+    {
+      "epoch": 0.6850036884813995,
+      "grad_norm": 0.27747398614883423,
+      "learning_rate": 4.439431796756047e-06,
+      "loss": 0.6046,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6854252292127727,
+      "grad_norm": 0.26655852794647217,
+      "learning_rate": 4.4387252226383035e-06,
+      "loss": 0.6229,
+      "step": 1626
+    },
+    {
+      "epoch": 0.6858467699441458,
+      "grad_norm": 0.3029691278934479,
+      "learning_rate": 4.438018259800604e-06,
+      "loss": 0.5992,
+      "step": 1627
+    },
+    {
+      "epoch": 0.686268310675519,
+      "grad_norm": 0.26988986134529114,
+      "learning_rate": 4.4373109083846944e-06,
+      "loss": 0.6118,
+      "step": 1628
+    },
+    {
+      "epoch": 0.6866898514068922,
+      "grad_norm": 0.27183255553245544,
+      "learning_rate": 4.436603168532403e-06,
+      "loss": 0.6044,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6871113921382653,
+      "grad_norm": 0.2642101049423218,
+      "learning_rate": 4.435895040385631e-06,
+      "loss": 0.6069,
+      "step": 1630
+    },
+    {
+      "epoch": 0.6875329328696386,
+      "grad_norm": 0.30380725860595703,
+      "learning_rate": 4.435186524086364e-06,
+      "loss": 0.5908,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6879544736010117,
+      "grad_norm": 0.27173638343811035,
+      "learning_rate": 4.434477619776658e-06,
+      "loss": 0.6317,
+      "step": 1632
+    },
+    {
+      "epoch": 0.6883760143323848,
+      "grad_norm": 0.286288321018219,
+      "learning_rate": 4.433768327598651e-06,
+      "loss": 0.5705,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6887975550637581,
+      "grad_norm": 0.2951652407646179,
+      "learning_rate": 4.43305864769456e-06,
+      "loss": 0.6033,
+      "step": 1634
+    },
+    {
+      "epoch": 0.6892190957951312,
+      "grad_norm": 0.2801284193992615,
+      "learning_rate": 4.4323485802066756e-06,
+      "loss": 0.616,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6896406365265044,
+      "grad_norm": 0.28426387906074524,
+      "learning_rate": 4.43163812527737e-06,
+      "loss": 0.6123,
+      "step": 1636
+    },
+    {
+      "epoch": 0.6900621772578776,
+      "grad_norm": 0.2763877511024475,
+      "learning_rate": 4.430927283049092e-06,
+      "loss": 0.625,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6904837179892507,
+      "grad_norm": 0.2745909094810486,
+      "learning_rate": 4.430216053664364e-06,
+      "loss": 0.6147,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6909052587206239,
+      "grad_norm": 0.2839808166027069,
+      "learning_rate": 4.4295044372657934e-06,
+      "loss": 0.6203,
+      "step": 1639
+    },
+    {
+      "epoch": 0.691326799451997,
+      "grad_norm": 0.28000640869140625,
+      "learning_rate": 4.428792433996059e-06,
+      "loss": 0.5878,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6917483401833702,
+      "grad_norm": 0.27341917157173157,
+      "learning_rate": 4.42808004399792e-06,
+      "loss": 0.5943,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6921698809147434,
+      "grad_norm": 0.27175790071487427,
+      "learning_rate": 4.427367267414213e-06,
+      "loss": 0.6246,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6925914216461165,
+      "grad_norm": 0.27792319655418396,
+      "learning_rate": 4.426654104387853e-06,
+      "loss": 0.6092,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6930129623774898,
+      "grad_norm": 0.2790071964263916,
+      "learning_rate": 4.425940555061828e-06,
+      "loss": 0.5965,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6934345031088629,
+      "grad_norm": 0.270605206489563,
+      "learning_rate": 4.42522661957921e-06,
+      "loss": 0.6445,
+      "step": 1645
+    },
+    {
+      "epoch": 0.693856043840236,
+      "grad_norm": 0.2771587669849396,
+      "learning_rate": 4.424512298083143e-06,
+      "loss": 0.6208,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6942775845716093,
+      "grad_norm": 0.2780877649784088,
+      "learning_rate": 4.4237975907168514e-06,
+      "loss": 0.6218,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6946991253029824,
+      "grad_norm": 0.2858000099658966,
+      "learning_rate": 4.423082497623636e-06,
+      "loss": 0.6081,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6951206660343555,
+      "grad_norm": 0.27489861845970154,
+      "learning_rate": 4.422367018946875e-06,
+      "loss": 0.5982,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6955422067657288,
+      "grad_norm": 0.27834510803222656,
+      "learning_rate": 4.421651154830024e-06,
+      "loss": 0.5961,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6959637474971019,
+      "grad_norm": 0.2735705077648163,
+      "learning_rate": 4.420934905416614e-06,
+      "loss": 0.5981,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6963852882284751,
+      "grad_norm": 0.28923502564430237,
+      "learning_rate": 4.420218270850258e-06,
+      "loss": 0.6034,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6968068289598482,
+      "grad_norm": 0.2764311134815216,
+      "learning_rate": 4.419501251274641e-06,
+      "loss": 0.6532,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6972283696912214,
+      "grad_norm": 0.29379788041114807,
+      "learning_rate": 4.4187838468335295e-06,
+      "loss": 0.5971,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6976499104225946,
+      "grad_norm": 0.28338396549224854,
+      "learning_rate": 4.418066057670764e-06,
+      "loss": 0.5937,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6980714511539677,
+      "grad_norm": 0.2692616879940033,
+      "learning_rate": 4.417347883930263e-06,
+      "loss": 0.615,
+      "step": 1656
+    },
+    {
+      "epoch": 0.698492991885341,
+      "grad_norm": 0.28758561611175537,
+      "learning_rate": 4.416629325756022e-06,
+      "loss": 0.6122,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6989145326167141,
+      "grad_norm": 0.2703517973423004,
+      "learning_rate": 4.415910383292115e-06,
+      "loss": 0.6437,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6993360733480872,
+      "grad_norm": 0.28340327739715576,
+      "learning_rate": 4.415191056682692e-06,
+      "loss": 0.6016,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6997576140794605,
+      "grad_norm": 0.2763698697090149,
+      "learning_rate": 4.41447134607198e-06,
+      "loss": 0.5967,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7001791548108336,
+      "grad_norm": 0.278862327337265,
+      "learning_rate": 4.413751251604282e-06,
+      "loss": 0.6285,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7006006955422067,
+      "grad_norm": 0.27666908502578735,
+      "learning_rate": 4.413030773423981e-06,
+      "loss": 0.626,
+      "step": 1662
+    },
+    {
+      "epoch": 0.70102223627358,
+      "grad_norm": 0.2785831391811371,
+      "learning_rate": 4.412309911675533e-06,
+      "loss": 0.5918,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7014437770049531,
+      "grad_norm": 0.2737468481063843,
+      "learning_rate": 4.411588666503473e-06,
+      "loss": 0.6294,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7018653177363263,
+      "grad_norm": 0.2960760295391083,
+      "learning_rate": 4.410867038052414e-06,
+      "loss": 0.6145,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7022868584676994,
+      "grad_norm": 0.273303359746933,
+      "learning_rate": 4.410145026467045e-06,
+      "loss": 0.5866,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7027083991990726,
+      "grad_norm": 0.2917192876338959,
+      "learning_rate": 4.409422631892128e-06,
+      "loss": 0.6049,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7031299399304458,
+      "grad_norm": 0.27532312273979187,
+      "learning_rate": 4.408699854472509e-06,
+      "loss": 0.6077,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7035514806618189,
+      "grad_norm": 0.27879154682159424,
+      "learning_rate": 4.407976694353105e-06,
+      "loss": 0.5919,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7039730213931921,
+      "grad_norm": 0.2655871510505676,
+      "learning_rate": 4.407253151678912e-06,
+      "loss": 0.5945,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7043945621245653,
+      "grad_norm": 0.2663457989692688,
+      "learning_rate": 4.4065292265950025e-06,
+      "loss": 0.6147,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7048161028559384,
+      "grad_norm": 0.2696913778781891,
+      "learning_rate": 4.405804919246525e-06,
+      "loss": 0.6132,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7052376435873117,
+      "grad_norm": 0.2775292992591858,
+      "learning_rate": 4.405080229778705e-06,
+      "loss": 0.6162,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7056591843186848,
+      "grad_norm": 0.2779128849506378,
+      "learning_rate": 4.404355158336847e-06,
+      "loss": 0.6332,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7060807250500579,
+      "grad_norm": 0.28609174489974976,
+      "learning_rate": 4.403629705066328e-06,
+      "loss": 0.5831,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7065022657814312,
+      "grad_norm": 0.28518691658973694,
+      "learning_rate": 4.402903870112603e-06,
+      "loss": 0.5973,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7069238065128043,
+      "grad_norm": 0.2682948112487793,
+      "learning_rate": 4.4021776536212065e-06,
+      "loss": 0.5832,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7073453472441774,
+      "grad_norm": 0.2743569016456604,
+      "learning_rate": 4.401451055737745e-06,
+      "loss": 0.6216,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7077668879755507,
+      "grad_norm": 0.2641680836677551,
+      "learning_rate": 4.4007240766079026e-06,
+      "loss": 0.6088,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7081884287069238,
+      "grad_norm": 0.2898157835006714,
+      "learning_rate": 4.399996716377442e-06,
+      "loss": 0.5973,
+      "step": 1680
+    },
+    {
+      "epoch": 0.708609969438297,
+      "grad_norm": 0.27880534529685974,
+      "learning_rate": 4.399268975192201e-06,
+      "loss": 0.624,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7090315101696701,
+      "grad_norm": 0.2757487893104553,
+      "learning_rate": 4.398540853198094e-06,
+      "loss": 0.6265,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7094530509010433,
+      "grad_norm": 0.26808667182922363,
+      "learning_rate": 4.3978123505411115e-06,
+      "loss": 0.5906,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7098745916324165,
+      "grad_norm": 0.2868767976760864,
+      "learning_rate": 4.39708346736732e-06,
+      "loss": 0.6192,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7102961323637896,
+      "grad_norm": 0.2838127017021179,
+      "learning_rate": 4.396354203822862e-06,
+      "loss": 0.5847,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7107176730951629,
+      "grad_norm": 0.2981458902359009,
+      "learning_rate": 4.3956245600539594e-06,
+      "loss": 0.6143,
+      "step": 1686
+    },
+    {
+      "epoch": 0.711139213826536,
+      "grad_norm": 0.27322861552238464,
+      "learning_rate": 4.3948945362069045e-06,
+      "loss": 0.5792,
+      "step": 1687
+    },
+    {
+      "epoch": 0.7115607545579091,
+      "grad_norm": 0.2746187746524811,
+      "learning_rate": 4.394164132428071e-06,
+      "loss": 0.5881,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7119822952892824,
+      "grad_norm": 0.279295414686203,
+      "learning_rate": 4.393433348863907e-06,
+      "loss": 0.6198,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7124038360206555,
+      "grad_norm": 0.29463455080986023,
+      "learning_rate": 4.3927021856609365e-06,
+      "loss": 0.6194,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7128253767520286,
+      "grad_norm": 0.28050824999809265,
+      "learning_rate": 4.3919706429657595e-06,
+      "loss": 0.6023,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7132469174834019,
+      "grad_norm": 0.29703789949417114,
+      "learning_rate": 4.391238720925053e-06,
+      "loss": 0.6013,
+      "step": 1692
+    },
+    {
+      "epoch": 0.713668458214775,
+      "grad_norm": 0.2685180604457855,
+      "learning_rate": 4.390506419685569e-06,
+      "loss": 0.6258,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7140899989461482,
+      "grad_norm": 0.27768176794052124,
+      "learning_rate": 4.389773739394137e-06,
+      "loss": 0.6099,
+      "step": 1694
+    },
+    {
+      "epoch": 0.7145115396775213,
+      "grad_norm": 0.28206533193588257,
+      "learning_rate": 4.389040680197659e-06,
+      "loss": 0.6266,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7149330804088945,
+      "grad_norm": 0.28915563225746155,
+      "learning_rate": 4.3883072422431186e-06,
+      "loss": 0.6149,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7153546211402677,
+      "grad_norm": 0.2777349650859833,
+      "learning_rate": 4.38757342567757e-06,
+      "loss": 0.6244,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7157761618716408,
+      "grad_norm": 0.27584996819496155,
+      "learning_rate": 4.3868392306481465e-06,
+      "loss": 0.6029,
+      "step": 1698
+    },
+    {
+      "epoch": 0.716197702603014,
+      "grad_norm": 0.2841974198818207,
+      "learning_rate": 4.386104657302056e-06,
+      "loss": 0.6189,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7166192433343872,
+      "grad_norm": 0.27435341477394104,
+      "learning_rate": 4.385369705786582e-06,
+      "loss": 0.6106,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7170407840657603,
+      "grad_norm": 0.28609856963157654,
+      "learning_rate": 4.384634376249085e-06,
+      "loss": 0.598,
+      "step": 1701
+    },
+    {
+      "epoch": 0.7174623247971336,
+      "grad_norm": 0.2744964361190796,
+      "learning_rate": 4.383898668837001e-06,
+      "loss": 0.6086,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7178838655285067,
+      "grad_norm": 0.28845101594924927,
+      "learning_rate": 4.383162583697842e-06,
+      "loss": 0.6018,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7183054062598798,
+      "grad_norm": 0.27615106105804443,
+      "learning_rate": 4.382426120979193e-06,
+      "loss": 0.6421,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7187269469912531,
+      "grad_norm": 0.28876346349716187,
+      "learning_rate": 4.381689280828718e-06,
+      "loss": 0.588,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7191484877226262,
+      "grad_norm": 0.2773711383342743,
+      "learning_rate": 4.380952063394157e-06,
+      "loss": 0.5934,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7195700284539994,
+      "grad_norm": 0.29457443952560425,
+      "learning_rate": 4.3802144688233215e-06,
+      "loss": 0.5994,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7199915691853725,
+      "grad_norm": 0.27908065915107727,
+      "learning_rate": 4.379476497264104e-06,
+      "loss": 0.6314,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7204131099167457,
+      "grad_norm": 0.2871860861778259,
+      "learning_rate": 4.3787381488644674e-06,
+      "loss": 0.6184,
+      "step": 1709
+    },
+    {
+      "epoch": 0.7208346506481189,
+      "grad_norm": 0.26911434531211853,
+      "learning_rate": 4.3779994237724555e-06,
+      "loss": 0.6107,
+      "step": 1710
+    },
+    {
+      "epoch": 0.721256191379492,
+      "grad_norm": 0.2794326841831207,
+      "learning_rate": 4.377260322136182e-06,
+      "loss": 0.6346,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7216777321108652,
+      "grad_norm": 0.28446105122566223,
+      "learning_rate": 4.3765208441038405e-06,
+      "loss": 0.6251,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7220992728422384,
+      "grad_norm": 0.2820131480693817,
+      "learning_rate": 4.375780989823698e-06,
+      "loss": 0.5968,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7225208135736115,
+      "grad_norm": 0.27935776114463806,
+      "learning_rate": 4.375040759444097e-06,
+      "loss": 0.5975,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7229423543049848,
+      "grad_norm": 0.27597203850746155,
+      "learning_rate": 4.374300153113457e-06,
+      "loss": 0.6069,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7233638950363579,
+      "grad_norm": 0.2791931927204132,
+      "learning_rate": 4.373559170980271e-06,
+      "loss": 0.6049,
+      "step": 1716
+    },
+    {
+      "epoch": 0.723785435767731,
+      "grad_norm": 0.27633246779441833,
+      "learning_rate": 4.3728178131931074e-06,
+      "loss": 0.6054,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7242069764991043,
+      "grad_norm": 0.28217223286628723,
+      "learning_rate": 4.372076079900611e-06,
+      "loss": 0.6028,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7246285172304774,
+      "grad_norm": 0.2778562009334564,
+      "learning_rate": 4.371333971251502e-06,
+      "loss": 0.6262,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7250500579618505,
+      "grad_norm": 0.2792932093143463,
+      "learning_rate": 4.370591487394576e-06,
+      "loss": 0.6153,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7254715986932238,
+      "grad_norm": 0.27217432856559753,
+      "learning_rate": 4.369848628478701e-06,
+      "loss": 0.6194,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7258931394245969,
+      "grad_norm": 0.28681138157844543,
+      "learning_rate": 4.369105394652824e-06,
+      "loss": 0.6196,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7263146801559701,
+      "grad_norm": 0.2745744585990906,
+      "learning_rate": 4.368361786065964e-06,
+      "loss": 0.5945,
+      "step": 1723
+    },
+    {
+      "epoch": 0.7267362208873432,
+      "grad_norm": 0.2741144001483917,
+      "learning_rate": 4.367617802867218e-06,
+      "loss": 0.6027,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7271577616187164,
+      "grad_norm": 0.2729848027229309,
+      "learning_rate": 4.366873445205757e-06,
+      "loss": 0.5874,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7275793023500896,
+      "grad_norm": 0.2810051739215851,
+      "learning_rate": 4.366128713230825e-06,
+      "loss": 0.6091,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7280008430814627,
+      "grad_norm": 0.2682962715625763,
+      "learning_rate": 4.365383607091745e-06,
+      "loss": 0.5964,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7284223838128359,
+      "grad_norm": 0.27417638897895813,
+      "learning_rate": 4.364638126937911e-06,
+      "loss": 0.5947,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7288439245442091,
+      "grad_norm": 0.2687387466430664,
+      "learning_rate": 4.363892272918796e-06,
+      "loss": 0.5869,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7292654652755822,
+      "grad_norm": 0.2701239585876465,
+      "learning_rate": 4.363146045183943e-06,
+      "loss": 0.6246,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7296870060069555,
+      "grad_norm": 0.2831512689590454,
+      "learning_rate": 4.3623994438829765e-06,
+      "loss": 0.6243,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7301085467383286,
+      "grad_norm": 0.27394115924835205,
+      "learning_rate": 4.361652469165589e-06,
+      "loss": 0.6137,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7305300874697017,
+      "grad_norm": 0.27634257078170776,
+      "learning_rate": 4.360905121181553e-06,
+      "loss": 0.6053,
+      "step": 1733
+    },
+    {
+      "epoch": 0.730951628201075,
+      "grad_norm": 0.2747725546360016,
+      "learning_rate": 4.360157400080714e-06,
+      "loss": 0.626,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7313731689324481,
+      "grad_norm": 0.27450236678123474,
+      "learning_rate": 4.359409306012989e-06,
+      "loss": 0.5906,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7317947096638213,
+      "grad_norm": 0.27480992674827576,
+      "learning_rate": 4.358660839128377e-06,
+      "loss": 0.5955,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7322162503951944,
+      "grad_norm": 0.27613264322280884,
+      "learning_rate": 4.3579119995769466e-06,
+      "loss": 0.6362,
+      "step": 1737
+    },
+    {
+      "epoch": 0.7326377911265676,
+      "grad_norm": 0.2745966613292694,
+      "learning_rate": 4.357162787508842e-06,
+      "loss": 0.654,
+      "step": 1738
+    },
+    {
+      "epoch": 0.7330593318579408,
+      "grad_norm": 0.2735876441001892,
+      "learning_rate": 4.3564132030742815e-06,
+      "loss": 0.6222,
+      "step": 1739
+    },
+    {
+      "epoch": 0.7334808725893139,
+      "grad_norm": 0.27759552001953125,
+      "learning_rate": 4.355663246423559e-06,
+      "loss": 0.6053,
+      "step": 1740
+    },
+    {
+      "epoch": 0.7339024133206871,
+      "grad_norm": 0.2813555598258972,
+      "learning_rate": 4.354912917707044e-06,
+      "loss": 0.622,
+      "step": 1741
+    },
+    {
+      "epoch": 0.7343239540520603,
+      "grad_norm": 0.26837989687919617,
+      "learning_rate": 4.35416221707518e-06,
+      "loss": 0.5973,
+      "step": 1742
+    },
+    {
+      "epoch": 0.7347454947834334,
+      "grad_norm": 0.26664796471595764,
+      "learning_rate": 4.353411144678483e-06,
+      "loss": 0.5853,
+      "step": 1743
+    },
+    {
+      "epoch": 0.7351670355148067,
+      "grad_norm": 0.282762736082077,
+      "learning_rate": 4.3526597006675464e-06,
+      "loss": 0.6278,
+      "step": 1744
+    },
+    {
+      "epoch": 0.7355885762461798,
+      "grad_norm": 0.2791050672531128,
+      "learning_rate": 4.351907885193037e-06,
+      "loss": 0.6286,
+      "step": 1745
+    },
+    {
+      "epoch": 0.7360101169775529,
+      "grad_norm": 0.2787439823150635,
+      "learning_rate": 4.351155698405695e-06,
+      "loss": 0.6359,
+      "step": 1746
+    },
+    {
+      "epoch": 0.7364316577089262,
+      "grad_norm": 0.26782116293907166,
+      "learning_rate": 4.3504031404563355e-06,
+      "loss": 0.609,
+      "step": 1747
+    },
+    {
+      "epoch": 0.7368531984402993,
+      "grad_norm": 0.2759721279144287,
+      "learning_rate": 4.3496502114958505e-06,
+      "loss": 0.6048,
+      "step": 1748
+    },
+    {
+      "epoch": 0.7372747391716724,
+      "grad_norm": 0.2736058533191681,
+      "learning_rate": 4.3488969116752025e-06,
+      "loss": 0.6036,
+      "step": 1749
+    },
+    {
+      "epoch": 0.7376962799030456,
+      "grad_norm": 0.2725280523300171,
+      "learning_rate": 4.348143241145432e-06,
+      "loss": 0.6181,
+      "step": 1750
+    },
+    {
+      "epoch": 0.7381178206344188,
+      "grad_norm": 0.2820308208465576,
+      "learning_rate": 4.34738920005765e-06,
+      "loss": 0.6303,
+      "step": 1751
+    },
+    {
+      "epoch": 0.738539361365792,
+      "grad_norm": 0.2874038517475128,
+      "learning_rate": 4.346634788563045e-06,
+      "loss": 0.6375,
+      "step": 1752
+    },
+    {
+      "epoch": 0.7389609020971651,
+      "grad_norm": 0.2736482322216034,
+      "learning_rate": 4.3458800068128804e-06,
+      "loss": 0.5882,
+      "step": 1753
+    },
+    {
+      "epoch": 0.7393824428285383,
+      "grad_norm": 0.2805015742778778,
+      "learning_rate": 4.345124854958489e-06,
+      "loss": 0.5936,
+      "step": 1754
+    },
+    {
+      "epoch": 0.7398039835599115,
+      "grad_norm": 0.2744179666042328,
+      "learning_rate": 4.344369333151283e-06,
+      "loss": 0.6198,
+      "step": 1755
+    },
+    {
+      "epoch": 0.7402255242912846,
+      "grad_norm": 0.2678615152835846,
+      "learning_rate": 4.343613441542747e-06,
+      "loss": 0.5818,
+      "step": 1756
+    },
+    {
+      "epoch": 0.7406470650226579,
+      "grad_norm": 0.28406351804733276,
+      "learning_rate": 4.342857180284437e-06,
+      "loss": 0.5996,
+      "step": 1757
+    },
+    {
+      "epoch": 0.741068605754031,
+      "grad_norm": 0.2839788794517517,
+      "learning_rate": 4.3421005495279875e-06,
+      "loss": 0.5981,
+      "step": 1758
+    },
+    {
+      "epoch": 0.7414901464854041,
+      "grad_norm": 0.2863001227378845,
+      "learning_rate": 4.341343549425104e-06,
+      "loss": 0.614,
+      "step": 1759
+    },
+    {
+      "epoch": 0.7419116872167774,
+      "grad_norm": 0.26885831356048584,
+      "learning_rate": 4.34058618012757e-06,
+      "loss": 0.5978,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7423332279481505,
+      "grad_norm": 0.2819503843784332,
+      "learning_rate": 4.339828441787236e-06,
+      "loss": 0.6105,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7427547686795236,
+      "grad_norm": 0.27759721875190735,
+      "learning_rate": 4.339070334556033e-06,
+      "loss": 0.595,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7431763094108969,
+      "grad_norm": 0.2791682481765747,
+      "learning_rate": 4.338311858585964e-06,
+      "loss": 0.6264,
+      "step": 1763
+    },
+    {
+      "epoch": 0.74359785014227,
+      "grad_norm": 0.27755993604660034,
+      "learning_rate": 4.337553014029105e-06,
+      "loss": 0.5891,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7440193908736432,
+      "grad_norm": 0.2726767957210541,
+      "learning_rate": 4.336793801037606e-06,
+      "loss": 0.6021,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7444409316050163,
+      "grad_norm": 0.26053258776664734,
+      "learning_rate": 4.336034219763693e-06,
+      "loss": 0.622,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7448624723363895,
+      "grad_norm": 0.2840675115585327,
+      "learning_rate": 4.335274270359662e-06,
+      "loss": 0.6147,
+      "step": 1767
+    },
+    {
+      "epoch": 0.7452840130677627,
+      "grad_norm": 0.2740550935268402,
+      "learning_rate": 4.334513952977887e-06,
+      "loss": 0.6122,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7457055537991358,
+      "grad_norm": 0.2909219563007355,
+      "learning_rate": 4.333753267770813e-06,
+      "loss": 0.6108,
+      "step": 1769
+    },
+    {
+      "epoch": 0.746127094530509,
+      "grad_norm": 0.2811336815357208,
+      "learning_rate": 4.332992214890959e-06,
+      "loss": 0.6112,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7465486352618822,
+      "grad_norm": 0.28469184041023254,
+      "learning_rate": 4.3322307944909194e-06,
+      "loss": 0.5861,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7469701759932553,
+      "grad_norm": 0.29819414019584656,
+      "learning_rate": 4.33146900672336e-06,
+      "loss": 0.5953,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7473917167246286,
+      "grad_norm": 0.28422629833221436,
+      "learning_rate": 4.330706851741022e-06,
+      "loss": 0.6029,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7478132574560017,
+      "grad_norm": 0.26283109188079834,
+      "learning_rate": 4.329944329696719e-06,
+      "loss": 0.615,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7482347981873748,
+      "grad_norm": 0.27696385979652405,
+      "learning_rate": 4.329181440743341e-06,
+      "loss": 0.6079,
+      "step": 1775
+    },
+    {
+      "epoch": 0.748656338918748,
+      "grad_norm": 0.2810826897621155,
+      "learning_rate": 4.3284181850338484e-06,
+      "loss": 0.6352,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7490778796501212,
+      "grad_norm": 0.3097665309906006,
+      "learning_rate": 4.327654562721274e-06,
+      "loss": 0.5866,
+      "step": 1777
+    },
+    {
+      "epoch": 0.7494994203814943,
+      "grad_norm": 0.27209892868995667,
+      "learning_rate": 4.326890573958729e-06,
+      "loss": 0.6078,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7499209611128675,
+      "grad_norm": 0.28652647137641907,
+      "learning_rate": 4.326126218899395e-06,
+      "loss": 0.6124,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7503425018442407,
+      "grad_norm": 0.2889411747455597,
+      "learning_rate": 4.325361497696525e-06,
+      "loss": 0.6165,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7507640425756139,
+      "grad_norm": 0.28042012453079224,
+      "learning_rate": 4.32459641050345e-06,
+      "loss": 0.6102,
+      "step": 1781
+    },
+    {
+      "epoch": 0.751185583306987,
+      "grad_norm": 0.27746862173080444,
+      "learning_rate": 4.323830957473572e-06,
+      "loss": 0.6151,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7516071240383602,
+      "grad_norm": 0.27821752429008484,
+      "learning_rate": 4.323065138760366e-06,
+      "loss": 0.6456,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7520286647697334,
+      "grad_norm": 0.2764889895915985,
+      "learning_rate": 4.322298954517381e-06,
+      "loss": 0.6216,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7524502055011065,
+      "grad_norm": 0.3045622706413269,
+      "learning_rate": 4.32153240489824e-06,
+      "loss": 0.6215,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7528717462324798,
+      "grad_norm": 0.27773186564445496,
+      "learning_rate": 4.320765490056636e-06,
+      "loss": 0.6023,
+      "step": 1786
+    },
+    {
+      "epoch": 0.7532932869638529,
+      "grad_norm": 0.2807536721229553,
+      "learning_rate": 4.31999821014634e-06,
+      "loss": 0.6153,
+      "step": 1787
+    },
+    {
+      "epoch": 0.753714827695226,
+      "grad_norm": 0.2860603630542755,
+      "learning_rate": 4.319230565321193e-06,
+      "loss": 0.5811,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7541363684265993,
+      "grad_norm": 0.2918761372566223,
+      "learning_rate": 4.31846255573511e-06,
+      "loss": 0.6347,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7545579091579724,
+      "grad_norm": 0.27817055583000183,
+      "learning_rate": 4.317694181542079e-06,
+      "loss": 0.6141,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7549794498893455,
+      "grad_norm": 0.29519447684288025,
+      "learning_rate": 4.316925442896162e-06,
+      "loss": 0.6077,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7554009906207187,
+      "grad_norm": 0.26630932092666626,
+      "learning_rate": 4.316156339951492e-06,
+      "loss": 0.6275,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7558225313520919,
+      "grad_norm": 0.2745026648044586,
+      "learning_rate": 4.315386872862276e-06,
+      "loss": 0.6112,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7562440720834651,
+      "grad_norm": 0.2789205014705658,
+      "learning_rate": 4.314617041782796e-06,
+      "loss": 0.6196,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7566656128148382,
+      "grad_norm": 0.28642550110816956,
+      "learning_rate": 4.313846846867406e-06,
+      "loss": 0.6191,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7570871535462114,
+      "grad_norm": 0.2764802873134613,
+      "learning_rate": 4.313076288270529e-06,
+      "loss": 0.5964,
+      "step": 1796
+    },
+    {
+      "epoch": 0.7575086942775846,
+      "grad_norm": 0.28983137011528015,
+      "learning_rate": 4.312305366146667e-06,
+      "loss": 0.5838,
+      "step": 1797
+    },
+    {
+      "epoch": 0.7579302350089577,
+      "grad_norm": 0.2779114246368408,
+      "learning_rate": 4.31153408065039e-06,
+      "loss": 0.6009,
+      "step": 1798
+    },
+    {
+      "epoch": 0.7583517757403309,
+      "grad_norm": 0.294474333524704,
+      "learning_rate": 4.310762431936346e-06,
+      "loss": 0.599,
+      "step": 1799
+    },
+    {
+      "epoch": 0.7587733164717041,
+      "grad_norm": 0.27603766322135925,
+      "learning_rate": 4.3099904201592505e-06,
+      "loss": 0.6098,
+      "step": 1800
+    },
+    {
+      "epoch": 0.7591948572030772,
+      "grad_norm": 0.2785513699054718,
+      "learning_rate": 4.309218045473894e-06,
+      "loss": 0.614,
+      "step": 1801
+    },
+    {
+      "epoch": 0.7596163979344505,
+      "grad_norm": 0.2740197479724884,
+      "learning_rate": 4.308445308035141e-06,
+      "loss": 0.6025,
+      "step": 1802
+    },
+    {
+      "epoch": 0.7600379386658236,
+      "grad_norm": 0.30990973114967346,
+      "learning_rate": 4.307672207997926e-06,
+      "loss": 0.5914,
+      "step": 1803
+    },
+    {
+      "epoch": 0.7604594793971967,
+      "grad_norm": 0.28241434693336487,
+      "learning_rate": 4.3068987455172595e-06,
+      "loss": 0.6049,
+      "step": 1804
+    },
+    {
+      "epoch": 0.76088102012857,
+      "grad_norm": 0.30772578716278076,
+      "learning_rate": 4.306124920748221e-06,
+      "loss": 0.5808,
+      "step": 1805
+    },
+    {
+      "epoch": 0.7613025608599431,
+      "grad_norm": 0.291250616312027,
+      "learning_rate": 4.305350733845967e-06,
+      "loss": 0.6223,
+      "step": 1806
+    },
+    {
+      "epoch": 0.7617241015913163,
+      "grad_norm": 0.28212836384773254,
+      "learning_rate": 4.304576184965723e-06,
+      "loss": 0.6136,
+      "step": 1807
+    },
+    {
+      "epoch": 0.7621456423226894,
+      "grad_norm": 0.2886107861995697,
+      "learning_rate": 4.303801274262788e-06,
+      "loss": 0.6272,
+      "step": 1808
+    },
+    {
+      "epoch": 0.7625671830540626,
+      "grad_norm": 0.3118770718574524,
+      "learning_rate": 4.303026001892533e-06,
+      "loss": 0.628,
+      "step": 1809
+    },
+    {
+      "epoch": 0.7629887237854358,
+      "grad_norm": 0.2872169315814972,
+      "learning_rate": 4.302250368010405e-06,
+      "loss": 0.5963,
+      "step": 1810
+    },
+    {
+      "epoch": 0.7634102645168089,
+      "grad_norm": 0.29370740056037903,
+      "learning_rate": 4.301474372771917e-06,
+      "loss": 0.6119,
+      "step": 1811
+    },
+    {
+      "epoch": 0.7638318052481821,
+      "grad_norm": 0.3142782747745514,
+      "learning_rate": 4.300698016332662e-06,
+      "loss": 0.622,
+      "step": 1812
+    },
+    {
+      "epoch": 0.7642533459795553,
+      "grad_norm": 0.31696972250938416,
+      "learning_rate": 4.299921298848299e-06,
+      "loss": 0.6128,
+      "step": 1813
+    },
+    {
+      "epoch": 0.7646748867109284,
+      "grad_norm": 0.2781507670879364,
+      "learning_rate": 4.299144220474564e-06,
+      "loss": 0.6284,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7650964274423017,
+      "grad_norm": 0.3658636212348938,
+      "learning_rate": 4.298366781367262e-06,
+      "loss": 0.5742,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7655179681736748,
+      "grad_norm": 0.3189171552658081,
+      "learning_rate": 4.297588981682274e-06,
+      "loss": 0.5901,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7659395089050479,
+      "grad_norm": 0.2833607494831085,
+      "learning_rate": 4.296810821575548e-06,
+      "loss": 0.6056,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7663610496364212,
+      "grad_norm": 0.3028174638748169,
+      "learning_rate": 4.296032301203108e-06,
+      "loss": 0.6137,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7667825903677943,
+      "grad_norm": 0.3240715563297272,
+      "learning_rate": 4.295253420721051e-06,
+      "loss": 0.6234,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7672041310991674,
+      "grad_norm": 0.30187803506851196,
+      "learning_rate": 4.294474180285545e-06,
+      "loss": 0.6113,
+      "step": 1820
+    },
+    {
+      "epoch": 0.7676256718305406,
+      "grad_norm": 0.28072071075439453,
+      "learning_rate": 4.293694580052828e-06,
+      "loss": 0.6214,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7680472125619138,
+      "grad_norm": 0.34691575169563293,
+      "learning_rate": 4.292914620179214e-06,
+      "loss": 0.6182,
+      "step": 1822
+    },
+    {
+      "epoch": 0.768468753293287,
+      "grad_norm": 0.30613550543785095,
+      "learning_rate": 4.292134300821086e-06,
+      "loss": 0.6091,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7688902940246601,
+      "grad_norm": 0.2682267427444458,
+      "learning_rate": 4.291353622134901e-06,
+      "loss": 0.6126,
+      "step": 1824
+    },
+    {
+      "epoch": 0.7693118347560333,
+      "grad_norm": 0.2843674123287201,
+      "learning_rate": 4.290572584277185e-06,
+      "loss": 0.6114,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7697333754874065,
+      "grad_norm": 0.2934292256832123,
+      "learning_rate": 4.2897911874045416e-06,
+      "loss": 0.6077,
+      "step": 1826
+    },
+    {
+      "epoch": 0.7701549162187796,
+      "grad_norm": 0.31129300594329834,
+      "learning_rate": 4.289009431673642e-06,
+      "loss": 0.6172,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7705764569501528,
+      "grad_norm": 0.2624717652797699,
+      "learning_rate": 4.288227317241229e-06,
+      "loss": 0.6077,
+      "step": 1828
+    },
+    {
+      "epoch": 0.770997997681526,
+      "grad_norm": 0.3125290274620056,
+      "learning_rate": 4.287444844264121e-06,
+      "loss": 0.6274,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7714195384128991,
+      "grad_norm": 0.3222668468952179,
+      "learning_rate": 4.286662012899206e-06,
+      "loss": 0.5822,
+      "step": 1830
+    },
+    {
+      "epoch": 0.7718410791442724,
+      "grad_norm": 0.2739278972148895,
+      "learning_rate": 4.285878823303441e-06,
+      "loss": 0.6115,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7722626198756455,
+      "grad_norm": 0.2804218530654907,
+      "learning_rate": 4.285095275633861e-06,
+      "loss": 0.5799,
+      "step": 1832
+    },
+    {
+      "epoch": 0.7726841606070186,
+      "grad_norm": 0.28704482316970825,
+      "learning_rate": 4.284311370047567e-06,
+      "loss": 0.602,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7731057013383918,
+      "grad_norm": 0.3058801591396332,
+      "learning_rate": 4.283527106701736e-06,
+      "loss": 0.6049,
+      "step": 1834
+    },
+    {
+      "epoch": 0.773527242069765,
+      "grad_norm": 0.2656762897968292,
+      "learning_rate": 4.282742485753615e-06,
+      "loss": 0.5868,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7739487828011382,
+      "grad_norm": 0.30584681034088135,
+      "learning_rate": 4.281957507360523e-06,
+      "loss": 0.6005,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7743703235325113,
+      "grad_norm": 0.27888980507850647,
+      "learning_rate": 4.281172171679848e-06,
+      "loss": 0.5838,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7747918642638845,
+      "grad_norm": 0.2767752408981323,
+      "learning_rate": 4.280386478869055e-06,
+      "loss": 0.6067,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7752134049952577,
+      "grad_norm": 0.2759985327720642,
+      "learning_rate": 4.279600429085676e-06,
+      "loss": 0.602,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7756349457266308,
+      "grad_norm": 0.2769717276096344,
+      "learning_rate": 4.278814022487317e-06,
+      "loss": 0.612,
+      "step": 1840
+    },
+    {
+      "epoch": 0.776056486458004,
+      "grad_norm": 0.2806585729122162,
+      "learning_rate": 4.278027259231655e-06,
+      "loss": 0.6131,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7764780271893772,
+      "grad_norm": 0.2769262194633484,
+      "learning_rate": 4.277240139476438e-06,
+      "loss": 0.6187,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7768995679207503,
+      "grad_norm": 0.2785145044326782,
+      "learning_rate": 4.276452663379485e-06,
+      "loss": 0.5968,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7773211086521236,
+      "grad_norm": 0.27963513135910034,
+      "learning_rate": 4.275664831098687e-06,
+      "loss": 0.6225,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7777426493834967,
+      "grad_norm": 0.2933167815208435,
+      "learning_rate": 4.274876642792009e-06,
+      "loss": 0.597,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7781641901148698,
+      "grad_norm": 0.2733224034309387,
+      "learning_rate": 4.274088098617481e-06,
+      "loss": 0.6133,
+      "step": 1846
+    },
+    {
+      "epoch": 0.778585730846243,
+      "grad_norm": 0.28585439920425415,
+      "learning_rate": 4.273299198733213e-06,
+      "loss": 0.5768,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7790072715776162,
+      "grad_norm": 0.2819828987121582,
+      "learning_rate": 4.272509943297378e-06,
+      "loss": 0.6234,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7794288123089893,
+      "grad_norm": 0.27826929092407227,
+      "learning_rate": 4.271720332468227e-06,
+      "loss": 0.6031,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7798503530403625,
+      "grad_norm": 0.27371424436569214,
+      "learning_rate": 4.2709303664040765e-06,
+      "loss": 0.6148,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7802718937717357,
+      "grad_norm": 0.28164538741111755,
+      "learning_rate": 4.270140045263318e-06,
+      "loss": 0.6033,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7806934345031089,
+      "grad_norm": 0.26457610726356506,
+      "learning_rate": 4.269349369204414e-06,
+      "loss": 0.6011,
+      "step": 1852
+    },
+    {
+      "epoch": 0.781114975234482,
+      "grad_norm": 0.2710423171520233,
+      "learning_rate": 4.268558338385897e-06,
+      "loss": 0.6144,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7815365159658552,
+      "grad_norm": 0.2840898334980011,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.6178,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7819580566972284,
+      "grad_norm": 0.28516873717308044,
+      "learning_rate": 4.2669752131045075e-06,
+      "loss": 0.6187,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7823795974286015,
+      "grad_norm": 0.27867427468299866,
+      "learning_rate": 4.266183118959057e-06,
+      "loss": 0.6023,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7828011381599747,
+      "grad_norm": 0.2843732237815857,
+      "learning_rate": 4.265390670688836e-06,
+      "loss": 0.5775,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7832226788913479,
+      "grad_norm": 0.2725246846675873,
+      "learning_rate": 4.2645978684527324e-06,
+      "loss": 0.6062,
+      "step": 1858
+    },
+    {
+      "epoch": 0.783644219622721,
+      "grad_norm": 0.2737867832183838,
+      "learning_rate": 4.263804712409704e-06,
+      "loss": 0.6143,
+      "step": 1859
+    },
+    {
+      "epoch": 0.7840657603540943,
+      "grad_norm": 0.2791579067707062,
+      "learning_rate": 4.263011202718782e-06,
+      "loss": 0.6044,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7844873010854674,
+      "grad_norm": 0.28584977984428406,
+      "learning_rate": 4.262217339539066e-06,
+      "loss": 0.6097,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7849088418168405,
+      "grad_norm": 0.2914561331272125,
+      "learning_rate": 4.2614231230297305e-06,
+      "loss": 0.6094,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7853303825482137,
+      "grad_norm": 0.27307790517807007,
+      "learning_rate": 4.260628553350017e-06,
+      "loss": 0.6083,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7857519232795869,
+      "grad_norm": 0.2828793525695801,
+      "learning_rate": 4.259833630659238e-06,
+      "loss": 0.5962,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7861734640109601,
+      "grad_norm": 0.2726515829563141,
+      "learning_rate": 4.259038355116778e-06,
+      "loss": 0.6413,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7865950047423332,
+      "grad_norm": 0.2797689735889435,
+      "learning_rate": 4.258242726882093e-06,
+      "loss": 0.62,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7870165454737064,
+      "grad_norm": 0.2748602628707886,
+      "learning_rate": 4.257446746114707e-06,
+      "loss": 0.5977,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7874380862050796,
+      "grad_norm": 0.28758981823921204,
+      "learning_rate": 4.256650412974218e-06,
+      "loss": 0.6291,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7878596269364527,
+      "grad_norm": 0.27436304092407227,
+      "learning_rate": 4.255853727620293e-06,
+      "loss": 0.6068,
+      "step": 1869
+    },
+    {
+      "epoch": 0.7882811676678259,
+      "grad_norm": 0.2714298963546753,
+      "learning_rate": 4.255056690212669e-06,
+      "loss": 0.6064,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7887027083991991,
+      "grad_norm": 0.27300503849983215,
+      "learning_rate": 4.254259300911155e-06,
+      "loss": 0.583,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7891242491305722,
+      "grad_norm": 0.27161818742752075,
+      "learning_rate": 4.2534615598756305e-06,
+      "loss": 0.6282,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7895457898619455,
+      "grad_norm": 0.26760098338127136,
+      "learning_rate": 4.2526634672660426e-06,
+      "loss": 0.5991,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7899673305933186,
+      "grad_norm": 0.2806521952152252,
+      "learning_rate": 4.251865023242412e-06,
+      "loss": 0.6195,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7903888713246917,
+      "grad_norm": 0.2772590219974518,
+      "learning_rate": 4.251066227964829e-06,
+      "loss": 0.622,
+      "step": 1875
+    },
+    {
+      "epoch": 0.790810412056065,
+      "grad_norm": 0.2802678048610687,
+      "learning_rate": 4.250267081593457e-06,
+      "loss": 0.6119,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7912319527874381,
+      "grad_norm": 0.27274298667907715,
+      "learning_rate": 4.249467584288523e-06,
+      "loss": 0.6229,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7916534935188112,
+      "grad_norm": 0.2765823304653168,
+      "learning_rate": 4.248667736210331e-06,
+      "loss": 0.6195,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7920750342501844,
+      "grad_norm": 0.2746501863002777,
+      "learning_rate": 4.247867537519252e-06,
+      "loss": 0.6048,
+      "step": 1879
+    },
+    {
+      "epoch": 0.7924965749815576,
+      "grad_norm": 0.27662572264671326,
+      "learning_rate": 4.247066988375728e-06,
+      "loss": 0.6194,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7929181157129308,
+      "grad_norm": 0.2744790017604828,
+      "learning_rate": 4.246266088940273e-06,
+      "loss": 0.6145,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7933396564443039,
+      "grad_norm": 0.2794719636440277,
+      "learning_rate": 4.2454648393734664e-06,
+      "loss": 0.6312,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7937611971756771,
+      "grad_norm": 0.28646135330200195,
+      "learning_rate": 4.244663239835964e-06,
+      "loss": 0.6086,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7941827379070503,
+      "grad_norm": 0.27879732847213745,
+      "learning_rate": 4.243861290488488e-06,
+      "loss": 0.6111,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7946042786384234,
+      "grad_norm": 0.2798142433166504,
+      "learning_rate": 4.2430589914918315e-06,
+      "loss": 0.605,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7950258193697967,
+      "grad_norm": 0.2951052486896515,
+      "learning_rate": 4.242256343006857e-06,
+      "loss": 0.6209,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7954473601011698,
+      "grad_norm": 0.26625362038612366,
+      "learning_rate": 4.241453345194499e-06,
+      "loss": 0.6067,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7958689008325429,
+      "grad_norm": 0.27114495635032654,
+      "learning_rate": 4.2406499982157605e-06,
+      "loss": 0.6133,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7962904415639162,
+      "grad_norm": 0.2642914652824402,
+      "learning_rate": 4.239846302231713e-06,
+      "loss": 0.6236,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7967119822952893,
+      "grad_norm": 0.2724757790565491,
+      "learning_rate": 4.239042257403503e-06,
+      "loss": 0.6114,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7971335230266624,
+      "grad_norm": 0.2783864438533783,
+      "learning_rate": 4.238237863892341e-06,
+      "loss": 0.611,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7975550637580356,
+      "grad_norm": 0.27949777245521545,
+      "learning_rate": 4.237433121859512e-06,
+      "loss": 0.6417,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7979766044894088,
+      "grad_norm": 0.2734021842479706,
+      "learning_rate": 4.236628031466369e-06,
+      "loss": 0.6233,
+      "step": 1893
+    },
+    {
+      "epoch": 0.798398145220782,
+      "grad_norm": 0.2807888388633728,
+      "learning_rate": 4.2358225928743345e-06,
+      "loss": 0.6237,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7988196859521551,
+      "grad_norm": 0.29392918944358826,
+      "learning_rate": 4.2350168062449e-06,
+      "loss": 0.6087,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7992412266835283,
+      "grad_norm": 0.27849727869033813,
+      "learning_rate": 4.23421067173963e-06,
+      "loss": 0.6134,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7996627674149015,
+      "grad_norm": 0.2796291410923004,
+      "learning_rate": 4.233404189520157e-06,
+      "loss": 0.6204,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8000843081462746,
+      "grad_norm": 0.2870771884918213,
+      "learning_rate": 4.232597359748181e-06,
+      "loss": 0.5965,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8005058488776478,
+      "grad_norm": 0.277759313583374,
+      "learning_rate": 4.231790182585474e-06,
+      "loss": 0.6161,
+      "step": 1899
+    },
+    {
+      "epoch": 0.800927389609021,
+      "grad_norm": 0.2767733633518219,
+      "learning_rate": 4.230982658193879e-06,
+      "loss": 0.6029,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8013489303403941,
+      "grad_norm": 0.2914954721927643,
+      "learning_rate": 4.230174786735306e-06,
+      "loss": 0.6172,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8017704710717674,
+      "grad_norm": 0.27767834067344666,
+      "learning_rate": 4.229366568371735e-06,
+      "loss": 0.5865,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8021920118031405,
+      "grad_norm": 0.27222493290901184,
+      "learning_rate": 4.228558003265217e-06,
+      "loss": 0.5893,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8026135525345136,
+      "grad_norm": 0.273752361536026,
+      "learning_rate": 4.227749091577871e-06,
+      "loss": 0.6052,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8030350932658868,
+      "grad_norm": 0.2827487885951996,
+      "learning_rate": 4.2269398334718855e-06,
+      "loss": 0.617,
+      "step": 1905
+    },
+    {
+      "epoch": 0.80345663399726,
+      "grad_norm": 0.27512606978416443,
+      "learning_rate": 4.226130229109522e-06,
+      "loss": 0.6044,
+      "step": 1906
+    },
+    {
+      "epoch": 0.8038781747286331,
+      "grad_norm": 0.28484004735946655,
+      "learning_rate": 4.225320278653105e-06,
+      "loss": 0.6148,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8042997154600063,
+      "grad_norm": 0.28284281492233276,
+      "learning_rate": 4.224509982265032e-06,
+      "loss": 0.5991,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8047212561913795,
+      "grad_norm": 0.2844797968864441,
+      "learning_rate": 4.223699340107773e-06,
+      "loss": 0.62,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8051427969227527,
+      "grad_norm": 0.29437506198883057,
+      "learning_rate": 4.2228883523438614e-06,
+      "loss": 0.5912,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8055643376541258,
+      "grad_norm": 0.2809103727340698,
+      "learning_rate": 4.222077019135904e-06,
+      "loss": 0.6136,
+      "step": 1911
+    },
+    {
+      "epoch": 0.805985878385499,
+      "grad_norm": 0.2818552255630493,
+      "learning_rate": 4.221265340646574e-06,
+      "loss": 0.6153,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8064074191168722,
+      "grad_norm": 0.2797611951828003,
+      "learning_rate": 4.220453317038615e-06,
+      "loss": 0.5874,
+      "step": 1913
+    },
+    {
+      "epoch": 0.8068289598482453,
+      "grad_norm": 0.2917400896549225,
+      "learning_rate": 4.2196409484748425e-06,
+      "loss": 0.6314,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8072505005796186,
+      "grad_norm": 0.27538177371025085,
+      "learning_rate": 4.218828235118137e-06,
+      "loss": 0.6155,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8076720413109917,
+      "grad_norm": 0.2839571535587311,
+      "learning_rate": 4.21801517713145e-06,
+      "loss": 0.611,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8080935820423648,
+      "grad_norm": 0.2872380316257477,
+      "learning_rate": 4.217201774677802e-06,
+      "loss": 0.5935,
+      "step": 1917
+    },
+    {
+      "epoch": 0.808515122773738,
+      "grad_norm": 0.27312177419662476,
+      "learning_rate": 4.216388027920282e-06,
+      "loss": 0.6241,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8089366635051112,
+      "grad_norm": 0.28297922015190125,
+      "learning_rate": 4.21557393702205e-06,
+      "loss": 0.5935,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8093582042364843,
+      "grad_norm": 0.2818053364753723,
+      "learning_rate": 4.214759502146331e-06,
+      "loss": 0.5713,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8097797449678575,
+      "grad_norm": 0.29005730152130127,
+      "learning_rate": 4.213944723456425e-06,
+      "loss": 0.5647,
+      "step": 1921
+    },
+    {
+      "epoch": 0.8102012856992307,
+      "grad_norm": 0.283902108669281,
+      "learning_rate": 4.213129601115695e-06,
+      "loss": 0.5922,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8106228264306039,
+      "grad_norm": 0.28247883915901184,
+      "learning_rate": 4.212314135287577e-06,
+      "loss": 0.5887,
+      "step": 1923
+    },
+    {
+      "epoch": 0.811044367161977,
+      "grad_norm": 0.2845180928707123,
+      "learning_rate": 4.211498326135571e-06,
+      "loss": 0.6055,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8114659078933502,
+      "grad_norm": 0.28263020515441895,
+      "learning_rate": 4.210682173823253e-06,
+      "loss": 0.6054,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8118874486247234,
+      "grad_norm": 0.2770913541316986,
+      "learning_rate": 4.209865678514262e-06,
+      "loss": 0.6232,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8123089893560965,
+      "grad_norm": 0.30108752846717834,
+      "learning_rate": 4.209048840372307e-06,
+      "loss": 0.6072,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8127305300874696,
+      "grad_norm": 0.26995745301246643,
+      "learning_rate": 4.208231659561167e-06,
+      "loss": 0.6127,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8131520708188429,
+      "grad_norm": 0.28315192461013794,
+      "learning_rate": 4.207414136244691e-06,
+      "loss": 0.6352,
+      "step": 1929
+    },
+    {
+      "epoch": 0.813573611550216,
+      "grad_norm": 0.3060400187969208,
+      "learning_rate": 4.206596270586792e-06,
+      "loss": 0.6106,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8139951522815893,
+      "grad_norm": 0.29358184337615967,
+      "learning_rate": 4.205778062751456e-06,
+      "loss": 0.6097,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8144166930129624,
+      "grad_norm": 0.2807903289794922,
+      "learning_rate": 4.204959512902736e-06,
+      "loss": 0.5994,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8148382337443355,
+      "grad_norm": 0.3359369933605194,
+      "learning_rate": 4.2041406212047516e-06,
+      "loss": 0.5951,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8152597744757087,
+      "grad_norm": 0.27375471591949463,
+      "learning_rate": 4.203321387821695e-06,
+      "loss": 0.6111,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8156813152070819,
+      "grad_norm": 0.2824607491493225,
+      "learning_rate": 4.202501812917824e-06,
+      "loss": 0.6155,
+      "step": 1935
+    },
+    {
+      "epoch": 0.8161028559384551,
+      "grad_norm": 0.33188438415527344,
+      "learning_rate": 4.2016818966574665e-06,
+      "loss": 0.5854,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8165243966698282,
+      "grad_norm": 0.2901972532272339,
+      "learning_rate": 4.200861639205018e-06,
+      "loss": 0.6122,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8169459374012014,
+      "grad_norm": 0.2772602438926697,
+      "learning_rate": 4.20004104072494e-06,
+      "loss": 0.6084,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8173674781325746,
+      "grad_norm": 0.3018006980419159,
+      "learning_rate": 4.199220101381769e-06,
+      "loss": 0.6172,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8177890188639477,
+      "grad_norm": 0.302866131067276,
+      "learning_rate": 4.1983988213401025e-06,
+      "loss": 0.5795,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8182105595953209,
+      "grad_norm": 0.2862773537635803,
+      "learning_rate": 4.197577200764611e-06,
+      "loss": 0.6226,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8186321003266941,
+      "grad_norm": 0.27157968282699585,
+      "learning_rate": 4.196755239820031e-06,
+      "loss": 0.5843,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8190536410580672,
+      "grad_norm": 0.32020455598831177,
+      "learning_rate": 4.195932938671169e-06,
+      "loss": 0.5782,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8194751817894405,
+      "grad_norm": 0.2892654836177826,
+      "learning_rate": 4.195110297482897e-06,
+      "loss": 0.5627,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8198967225208136,
+      "grad_norm": 0.2753625810146332,
+      "learning_rate": 4.194287316420158e-06,
+      "loss": 0.6216,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8203182632521867,
+      "grad_norm": 0.33662664890289307,
+      "learning_rate": 4.193463995647963e-06,
+      "loss": 0.6106,
+      "step": 1946
+    },
+    {
+      "epoch": 0.82073980398356,
+      "grad_norm": 0.2730756103992462,
+      "learning_rate": 4.192640335331388e-06,
+      "loss": 0.6125,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8211613447149331,
+      "grad_norm": 0.28652599453926086,
+      "learning_rate": 4.191816335635581e-06,
+      "loss": 0.6094,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8215828854463062,
+      "grad_norm": 0.3052879273891449,
+      "learning_rate": 4.190991996725756e-06,
+      "loss": 0.6199,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8220044261776794,
+      "grad_norm": 0.30441632866859436,
+      "learning_rate": 4.190167318767195e-06,
+      "loss": 0.626,
+      "step": 1950
+    },
+    {
+      "epoch": 0.8224259669090526,
+      "grad_norm": 0.2776713967323303,
+      "learning_rate": 4.189342301925249e-06,
+      "loss": 0.5914,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8228475076404258,
+      "grad_norm": 0.27965450286865234,
+      "learning_rate": 4.188516946365334e-06,
+      "loss": 0.6236,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8232690483717989,
+      "grad_norm": 0.28747665882110596,
+      "learning_rate": 4.187691252252939e-06,
+      "loss": 0.5975,
+      "step": 1953
+    },
+    {
+      "epoch": 0.823690589103172,
+      "grad_norm": 0.2900429368019104,
+      "learning_rate": 4.186865219753616e-06,
+      "loss": 0.6085,
+      "step": 1954
+    },
+    {
+      "epoch": 0.8241121298345453,
+      "grad_norm": 0.26926830410957336,
+      "learning_rate": 4.1860388490329886e-06,
+      "loss": 0.5945,
+      "step": 1955
+    },
+    {
+      "epoch": 0.8245336705659184,
+      "grad_norm": 0.28972309827804565,
+      "learning_rate": 4.185212140256745e-06,
+      "loss": 0.5942,
+      "step": 1956
+    },
+    {
+      "epoch": 0.8249552112972915,
+      "grad_norm": 0.26291900873184204,
+      "learning_rate": 4.184385093590643e-06,
+      "loss": 0.6262,
+      "step": 1957
+    },
+    {
+      "epoch": 0.8253767520286648,
+      "grad_norm": 0.26619863510131836,
+      "learning_rate": 4.183557709200508e-06,
+      "loss": 0.5852,
+      "step": 1958
+    },
+    {
+      "epoch": 0.8257982927600379,
+      "grad_norm": 0.27881160378456116,
+      "learning_rate": 4.182729987252233e-06,
+      "loss": 0.6214,
+      "step": 1959
+    },
+    {
+      "epoch": 0.8262198334914111,
+      "grad_norm": 0.2759825587272644,
+      "learning_rate": 4.181901927911779e-06,
+      "loss": 0.6289,
+      "step": 1960
+    },
+    {
+      "epoch": 0.8266413742227843,
+      "grad_norm": 0.2735176384449005,
+      "learning_rate": 4.181073531345174e-06,
+      "loss": 0.6258,
+      "step": 1961
+    },
+    {
+      "epoch": 0.8270629149541574,
+      "grad_norm": 0.28626999258995056,
+      "learning_rate": 4.1802447977185135e-06,
+      "loss": 0.6037,
+      "step": 1962
+    },
+    {
+      "epoch": 0.8274844556855306,
+      "grad_norm": 0.273320734500885,
+      "learning_rate": 4.179415727197961e-06,
+      "loss": 0.6285,
+      "step": 1963
+    },
+    {
+      "epoch": 0.8279059964169038,
+      "grad_norm": 0.2882641851902008,
+      "learning_rate": 4.1785863199497485e-06,
+      "loss": 0.605,
+      "step": 1964
+    },
+    {
+      "epoch": 0.828327537148277,
+      "grad_norm": 0.2794593572616577,
+      "learning_rate": 4.177756576140173e-06,
+      "loss": 0.621,
+      "step": 1965
+    },
+    {
+      "epoch": 0.8287490778796501,
+      "grad_norm": 0.28361931443214417,
+      "learning_rate": 4.176926495935601e-06,
+      "loss": 0.6359,
+      "step": 1966
+    },
+    {
+      "epoch": 0.8291706186110233,
+      "grad_norm": 0.29417088627815247,
+      "learning_rate": 4.176096079502467e-06,
+      "loss": 0.6037,
+      "step": 1967
+    },
+    {
+      "epoch": 0.8295921593423965,
+      "grad_norm": 0.29032111167907715,
+      "learning_rate": 4.175265327007271e-06,
+      "loss": 0.5849,
+      "step": 1968
+    },
+    {
+      "epoch": 0.8300137000737696,
+      "grad_norm": 0.2781698703765869,
+      "learning_rate": 4.174434238616582e-06,
+      "loss": 0.6146,
+      "step": 1969
+    },
+    {
+      "epoch": 0.8304352408051427,
+      "grad_norm": 0.2866978645324707,
+      "learning_rate": 4.173602814497033e-06,
+      "loss": 0.6018,
+      "step": 1970
+    },
+    {
+      "epoch": 0.830856781536516,
+      "grad_norm": 0.2820422351360321,
+      "learning_rate": 4.172771054815331e-06,
+      "loss": 0.6047,
+      "step": 1971
+    },
+    {
+      "epoch": 0.8312783222678891,
+      "grad_norm": 0.28404533863067627,
+      "learning_rate": 4.171938959738243e-06,
+      "loss": 0.6003,
+      "step": 1972
+    },
+    {
+      "epoch": 0.8316998629992624,
+      "grad_norm": 0.2743147611618042,
+      "learning_rate": 4.171106529432608e-06,
+      "loss": 0.624,
+      "step": 1973
+    },
+    {
+      "epoch": 0.8321214037306355,
+      "grad_norm": 0.26434770226478577,
+      "learning_rate": 4.17027376406533e-06,
+      "loss": 0.6231,
+      "step": 1974
+    },
+    {
+      "epoch": 0.8325429444620086,
+      "grad_norm": 0.2709536552429199,
+      "learning_rate": 4.16944066380338e-06,
+      "loss": 0.5906,
+      "step": 1975
+    },
+    {
+      "epoch": 0.8329644851933818,
+      "grad_norm": 0.27898532152175903,
+      "learning_rate": 4.168607228813799e-06,
+      "loss": 0.5955,
+      "step": 1976
+    },
+    {
+      "epoch": 0.833386025924755,
+      "grad_norm": 0.27570927143096924,
+      "learning_rate": 4.167773459263693e-06,
+      "loss": 0.6344,
+      "step": 1977
+    },
+    {
+      "epoch": 0.8338075666561281,
+      "grad_norm": 0.2857637107372284,
+      "learning_rate": 4.166939355320232e-06,
+      "loss": 0.5819,
+      "step": 1978
+    },
+    {
+      "epoch": 0.8342291073875013,
+      "grad_norm": 0.2833573520183563,
+      "learning_rate": 4.166104917150661e-06,
+      "loss": 0.5874,
+      "step": 1979
+    },
+    {
+      "epoch": 0.8346506481188745,
+      "grad_norm": 0.2838643491268158,
+      "learning_rate": 4.165270144922283e-06,
+      "loss": 0.6121,
+      "step": 1980
+    },
+    {
+      "epoch": 0.8350721888502477,
+      "grad_norm": 0.28722503781318665,
+      "learning_rate": 4.1644350388024745e-06,
+      "loss": 0.619,
+      "step": 1981
+    },
+    {
+      "epoch": 0.8354937295816208,
+      "grad_norm": 0.27415475249290466,
+      "learning_rate": 4.163599598958676e-06,
+      "loss": 0.5907,
+      "step": 1982
+    },
+    {
+      "epoch": 0.835915270312994,
+      "grad_norm": 0.2793228030204773,
+      "learning_rate": 4.162763825558396e-06,
+      "loss": 0.635,
+      "step": 1983
+    },
+    {
+      "epoch": 0.8363368110443672,
+      "grad_norm": 0.2753475308418274,
+      "learning_rate": 4.161927718769209e-06,
+      "loss": 0.6237,
+      "step": 1984
+    },
+    {
+      "epoch": 0.8367583517757403,
+      "grad_norm": 0.2738324701786041,
+      "learning_rate": 4.1610912787587575e-06,
+      "loss": 0.6119,
+      "step": 1985
+    },
+    {
+      "epoch": 0.8371798925071136,
+      "grad_norm": 0.2788059711456299,
+      "learning_rate": 4.160254505694749e-06,
+      "loss": 0.6041,
+      "step": 1986
+    },
+    {
+      "epoch": 0.8376014332384867,
+      "grad_norm": 0.2716056704521179,
+      "learning_rate": 4.15941739974496e-06,
+      "loss": 0.6232,
+      "step": 1987
+    },
+    {
+      "epoch": 0.8380229739698598,
+      "grad_norm": 0.2745053470134735,
+      "learning_rate": 4.1585799610772315e-06,
+      "loss": 0.6226,
+      "step": 1988
+    },
+    {
+      "epoch": 0.838444514701233,
+      "grad_norm": 0.2858763635158539,
+      "learning_rate": 4.157742189859473e-06,
+      "loss": 0.5855,
+      "step": 1989
+    },
+    {
+      "epoch": 0.8388660554326062,
+      "grad_norm": 0.2706812024116516,
+      "learning_rate": 4.156904086259661e-06,
+      "loss": 0.6023,
+      "step": 1990
+    },
+    {
+      "epoch": 0.8392875961639793,
+      "grad_norm": 0.2723211348056793,
+      "learning_rate": 4.156065650445835e-06,
+      "loss": 0.6248,
+      "step": 1991
+    },
+    {
+      "epoch": 0.8397091368953525,
+      "grad_norm": 0.28066903352737427,
+      "learning_rate": 4.155226882586107e-06,
+      "loss": 0.6029,
+      "step": 1992
+    },
+    {
+      "epoch": 0.8401306776267257,
+      "grad_norm": 0.27876657247543335,
+      "learning_rate": 4.15438778284865e-06,
+      "loss": 0.6016,
+      "step": 1993
+    },
+    {
+      "epoch": 0.8405522183580989,
+      "grad_norm": 0.2750357985496521,
+      "learning_rate": 4.153548351401708e-06,
+      "loss": 0.6097,
+      "step": 1994
+    },
+    {
+      "epoch": 0.840973759089472,
+      "grad_norm": 0.27342572808265686,
+      "learning_rate": 4.152708588413587e-06,
+      "loss": 0.592,
+      "step": 1995
+    },
+    {
+      "epoch": 0.8413952998208452,
+      "grad_norm": 0.27430519461631775,
+      "learning_rate": 4.151868494052663e-06,
+      "loss": 0.6127,
+      "step": 1996
+    },
+    {
+      "epoch": 0.8418168405522184,
+      "grad_norm": 0.2829979658126831,
+      "learning_rate": 4.151028068487378e-06,
+      "loss": 0.6187,
+      "step": 1997
+    },
+    {
+      "epoch": 0.8422383812835915,
+      "grad_norm": 0.28300151228904724,
+      "learning_rate": 4.150187311886238e-06,
+      "loss": 0.6044,
+      "step": 1998
+    },
+    {
+      "epoch": 0.8426599220149646,
+      "grad_norm": 0.2855989336967468,
+      "learning_rate": 4.149346224417819e-06,
+      "loss": 0.6189,
+      "step": 1999
+    },
+    {
+      "epoch": 0.8430814627463379,
+      "grad_norm": 0.294101744890213,
+      "learning_rate": 4.148504806250761e-06,
+      "loss": 0.6041,
+      "step": 2000
+    },
+    {
+      "epoch": 0.843503003477711,
+      "grad_norm": 0.2791568636894226,
+      "learning_rate": 4.147663057553769e-06,
+      "loss": 0.5911,
+      "step": 2001
+    },
+    {
+      "epoch": 0.8439245442090842,
+      "grad_norm": 0.2881386876106262,
+      "learning_rate": 4.146820978495619e-06,
+      "loss": 0.6099,
+      "step": 2002
+    },
+    {
+      "epoch": 0.8443460849404574,
+      "grad_norm": 0.3120289742946625,
+      "learning_rate": 4.145978569245148e-06,
+      "loss": 0.6001,
+      "step": 2003
+    },
+    {
+      "epoch": 0.8447676256718305,
+      "grad_norm": 0.2827363610267639,
+      "learning_rate": 4.145135829971262e-06,
+      "loss": 0.6054,
+      "step": 2004
+    },
+    {
+      "epoch": 0.8451891664032037,
+      "grad_norm": 0.27895763516426086,
+      "learning_rate": 4.144292760842932e-06,
+      "loss": 0.6192,
+      "step": 2005
+    },
+    {
+      "epoch": 0.8456107071345769,
+      "grad_norm": 0.2944113612174988,
+      "learning_rate": 4.143449362029198e-06,
+      "loss": 0.6098,
+      "step": 2006
+    },
+    {
+      "epoch": 0.84603224786595,
+      "grad_norm": 0.285103440284729,
+      "learning_rate": 4.142605633699162e-06,
+      "loss": 0.5932,
+      "step": 2007
+    },
+    {
+      "epoch": 0.8464537885973232,
+      "grad_norm": 0.2834998667240143,
+      "learning_rate": 4.1417615760219935e-06,
+      "loss": 0.6112,
+      "step": 2008
+    },
+    {
+      "epoch": 0.8468753293286964,
+      "grad_norm": 0.2918637990951538,
+      "learning_rate": 4.14091718916693e-06,
+      "loss": 0.6099,
+      "step": 2009
+    },
+    {
+      "epoch": 0.8472968700600696,
+      "grad_norm": 0.2774882912635803,
+      "learning_rate": 4.1400724733032725e-06,
+      "loss": 0.6121,
+      "step": 2010
+    },
+    {
+      "epoch": 0.8477184107914427,
+      "grad_norm": 0.2920991778373718,
+      "learning_rate": 4.13922742860039e-06,
+      "loss": 0.6247,
+      "step": 2011
+    },
+    {
+      "epoch": 0.8481399515228158,
+      "grad_norm": 0.29502829909324646,
+      "learning_rate": 4.138382055227715e-06,
+      "loss": 0.5536,
+      "step": 2012
+    },
+    {
+      "epoch": 0.8485614922541891,
+      "grad_norm": 0.2787350118160248,
+      "learning_rate": 4.137536353354748e-06,
+      "loss": 0.6079,
+      "step": 2013
+    },
+    {
+      "epoch": 0.8489830329855622,
+      "grad_norm": 0.2818499207496643,
+      "learning_rate": 4.136690323151055e-06,
+      "loss": 0.6054,
+      "step": 2014
+    },
+    {
+      "epoch": 0.8494045737169355,
+      "grad_norm": 0.2817799150943756,
+      "learning_rate": 4.135843964786267e-06,
+      "loss": 0.6166,
+      "step": 2015
+    },
+    {
+      "epoch": 0.8498261144483086,
+      "grad_norm": 0.28017786145210266,
+      "learning_rate": 4.134997278430081e-06,
+      "loss": 0.6161,
+      "step": 2016
+    },
+    {
+      "epoch": 0.8502476551796817,
+      "grad_norm": 0.2978954613208771,
+      "learning_rate": 4.13415026425226e-06,
+      "loss": 0.597,
+      "step": 2017
+    },
+    {
+      "epoch": 0.8506691959110549,
+      "grad_norm": 0.2903663218021393,
+      "learning_rate": 4.133302922422633e-06,
+      "loss": 0.6178,
+      "step": 2018
+    },
+    {
+      "epoch": 0.8510907366424281,
+      "grad_norm": 0.2889491617679596,
+      "learning_rate": 4.132455253111096e-06,
+      "loss": 0.6028,
+      "step": 2019
+    },
+    {
+      "epoch": 0.8515122773738012,
+      "grad_norm": 0.2871458828449249,
+      "learning_rate": 4.131607256487607e-06,
+      "loss": 0.5936,
+      "step": 2020
+    },
+    {
+      "epoch": 0.8519338181051744,
+      "grad_norm": 0.29041045904159546,
+      "learning_rate": 4.130758932722192e-06,
+      "loss": 0.6211,
+      "step": 2021
+    },
+    {
+      "epoch": 0.8523553588365476,
+      "grad_norm": 0.2993323504924774,
+      "learning_rate": 4.129910281984944e-06,
+      "loss": 0.6045,
+      "step": 2022
+    },
+    {
+      "epoch": 0.8527768995679208,
+      "grad_norm": 0.2895345687866211,
+      "learning_rate": 4.129061304446017e-06,
+      "loss": 0.6036,
+      "step": 2023
+    },
+    {
+      "epoch": 0.8531984402992939,
+      "grad_norm": 0.27798959612846375,
+      "learning_rate": 4.128212000275636e-06,
+      "loss": 0.5955,
+      "step": 2024
+    },
+    {
+      "epoch": 0.853619981030667,
+      "grad_norm": 0.2896115481853485,
+      "learning_rate": 4.127362369644089e-06,
+      "loss": 0.6151,
+      "step": 2025
+    },
+    {
+      "epoch": 0.8540415217620403,
+      "grad_norm": 0.2754329442977905,
+      "learning_rate": 4.126512412721727e-06,
+      "loss": 0.6007,
+      "step": 2026
+    },
+    {
+      "epoch": 0.8544630624934134,
+      "grad_norm": 0.2852628231048584,
+      "learning_rate": 4.125662129678971e-06,
+      "loss": 0.6061,
+      "step": 2027
+    },
+    {
+      "epoch": 0.8548846032247865,
+      "grad_norm": 0.2820385694503784,
+      "learning_rate": 4.1248115206863035e-06,
+      "loss": 0.6003,
+      "step": 2028
+    },
+    {
+      "epoch": 0.8553061439561598,
+      "grad_norm": 0.298114538192749,
+      "learning_rate": 4.123960585914276e-06,
+      "loss": 0.6213,
+      "step": 2029
+    },
+    {
+      "epoch": 0.8557276846875329,
+      "grad_norm": 0.2903365194797516,
+      "learning_rate": 4.123109325533502e-06,
+      "loss": 0.6313,
+      "step": 2030
+    },
+    {
+      "epoch": 0.8561492254189061,
+      "grad_norm": 0.2795485556125641,
+      "learning_rate": 4.1222577397146625e-06,
+      "loss": 0.5913,
+      "step": 2031
+    },
+    {
+      "epoch": 0.8565707661502793,
+      "grad_norm": 0.279148668050766,
+      "learning_rate": 4.121405828628501e-06,
+      "loss": 0.607,
+      "step": 2032
+    },
+    {
+      "epoch": 0.8569923068816524,
+      "grad_norm": 0.27291420102119446,
+      "learning_rate": 4.120553592445831e-06,
+      "loss": 0.6152,
+      "step": 2033
+    },
+    {
+      "epoch": 0.8574138476130256,
+      "grad_norm": 0.2836747169494629,
+      "learning_rate": 4.119701031337526e-06,
+      "loss": 0.6189,
+      "step": 2034
+    },
+    {
+      "epoch": 0.8578353883443988,
+      "grad_norm": 0.27607133984565735,
+      "learning_rate": 4.118848145474528e-06,
+      "loss": 0.6143,
+      "step": 2035
+    },
+    {
+      "epoch": 0.858256929075772,
+      "grad_norm": 0.2850489914417267,
+      "learning_rate": 4.117994935027842e-06,
+      "loss": 0.6025,
+      "step": 2036
+    },
+    {
+      "epoch": 0.8586784698071451,
+      "grad_norm": 0.2796688377857208,
+      "learning_rate": 4.11714140016854e-06,
+      "loss": 0.6299,
+      "step": 2037
+    },
+    {
+      "epoch": 0.8591000105385183,
+      "grad_norm": 0.2783965766429901,
+      "learning_rate": 4.116287541067759e-06,
+      "loss": 0.623,
+      "step": 2038
+    },
+    {
+      "epoch": 0.8595215512698915,
+      "grad_norm": 0.26927781105041504,
+      "learning_rate": 4.115433357896699e-06,
+      "loss": 0.6057,
+      "step": 2039
+    },
+    {
+      "epoch": 0.8599430920012646,
+      "grad_norm": 0.2853672206401825,
+      "learning_rate": 4.114578850826625e-06,
+      "loss": 0.5987,
+      "step": 2040
+    },
+    {
+      "epoch": 0.8603646327326377,
+      "grad_norm": 0.27144283056259155,
+      "learning_rate": 4.1137240200288715e-06,
+      "loss": 0.6373,
+      "step": 2041
+    },
+    {
+      "epoch": 0.860786173464011,
+      "grad_norm": 0.28986623883247375,
+      "learning_rate": 4.112868865674832e-06,
+      "loss": 0.6201,
+      "step": 2042
+    },
+    {
+      "epoch": 0.8612077141953841,
+      "grad_norm": 0.27960702776908875,
+      "learning_rate": 4.112013387935967e-06,
+      "loss": 0.6033,
+      "step": 2043
+    },
+    {
+      "epoch": 0.8616292549267573,
+      "grad_norm": 0.2714480757713318,
+      "learning_rate": 4.111157586983804e-06,
+      "loss": 0.6073,
+      "step": 2044
+    },
+    {
+      "epoch": 0.8620507956581305,
+      "grad_norm": 0.2804807722568512,
+      "learning_rate": 4.110301462989933e-06,
+      "loss": 0.6023,
+      "step": 2045
+    },
+    {
+      "epoch": 0.8624723363895036,
+      "grad_norm": 0.28676900267601013,
+      "learning_rate": 4.1094450161260095e-06,
+      "loss": 0.6373,
+      "step": 2046
+    },
+    {
+      "epoch": 0.8628938771208768,
+      "grad_norm": 0.2765231430530548,
+      "learning_rate": 4.108588246563752e-06,
+      "loss": 0.6094,
+      "step": 2047
+    },
+    {
+      "epoch": 0.86331541785225,
+      "grad_norm": 0.27119365334510803,
+      "learning_rate": 4.107731154474946e-06,
+      "loss": 0.5949,
+      "step": 2048
+    },
+    {
+      "epoch": 0.8637369585836231,
+      "grad_norm": 0.28397083282470703,
+      "learning_rate": 4.106873740031442e-06,
+      "loss": 0.6089,
+      "step": 2049
+    },
+    {
+      "epoch": 0.8641584993149963,
+      "grad_norm": 0.2776951193809509,
+      "learning_rate": 4.1060160034051535e-06,
+      "loss": 0.6175,
+      "step": 2050
+    },
+    {
+      "epoch": 0.8645800400463695,
+      "grad_norm": 0.2785867750644684,
+      "learning_rate": 4.105157944768059e-06,
+      "loss": 0.5928,
+      "step": 2051
+    },
+    {
+      "epoch": 0.8650015807777427,
+      "grad_norm": 0.278852641582489,
+      "learning_rate": 4.104299564292201e-06,
+      "loss": 0.6116,
+      "step": 2052
+    },
+    {
+      "epoch": 0.8654231215091158,
+      "grad_norm": 0.30016806721687317,
+      "learning_rate": 4.103440862149689e-06,
+      "loss": 0.6372,
+      "step": 2053
+    },
+    {
+      "epoch": 0.865844662240489,
+      "grad_norm": 0.26678794622421265,
+      "learning_rate": 4.102581838512694e-06,
+      "loss": 0.5757,
+      "step": 2054
+    },
+    {
+      "epoch": 0.8662662029718622,
+      "grad_norm": 0.341045081615448,
+      "learning_rate": 4.101722493553452e-06,
+      "loss": 0.6175,
+      "step": 2055
+    },
+    {
+      "epoch": 0.8666877437032353,
+      "grad_norm": 0.30285635590553284,
+      "learning_rate": 4.100862827444266e-06,
+      "loss": 0.6304,
+      "step": 2056
+    },
+    {
+      "epoch": 0.8671092844346084,
+      "grad_norm": 0.2856931984424591,
+      "learning_rate": 4.100002840357499e-06,
+      "loss": 0.6074,
+      "step": 2057
+    },
+    {
+      "epoch": 0.8675308251659817,
+      "grad_norm": 0.27615147829055786,
+      "learning_rate": 4.0991425324655835e-06,
+      "loss": 0.6198,
+      "step": 2058
+    },
+    {
+      "epoch": 0.8679523658973548,
+      "grad_norm": 0.27827200293540955,
+      "learning_rate": 4.098281903941012e-06,
+      "loss": 0.5936,
+      "step": 2059
+    },
+    {
+      "epoch": 0.868373906628728,
+      "grad_norm": 0.2723389267921448,
+      "learning_rate": 4.097420954956345e-06,
+      "loss": 0.593,
+      "step": 2060
+    },
+    {
+      "epoch": 0.8687954473601012,
+      "grad_norm": 0.28361982107162476,
+      "learning_rate": 4.096559685684203e-06,
+      "loss": 0.5853,
+      "step": 2061
+    },
+    {
+      "epoch": 0.8692169880914743,
+      "grad_norm": 0.28583288192749023,
+      "learning_rate": 4.095698096297273e-06,
+      "loss": 0.5833,
+      "step": 2062
+    },
+    {
+      "epoch": 0.8696385288228475,
+      "grad_norm": 0.27001211047172546,
+      "learning_rate": 4.094836186968308e-06,
+      "loss": 0.5809,
+      "step": 2063
+    },
+    {
+      "epoch": 0.8700600695542207,
+      "grad_norm": 0.283908873796463,
+      "learning_rate": 4.093973957870121e-06,
+      "loss": 0.5933,
+      "step": 2064
+    },
+    {
+      "epoch": 0.8704816102855939,
+      "grad_norm": 0.2776426076889038,
+      "learning_rate": 4.093111409175593e-06,
+      "loss": 0.5974,
+      "step": 2065
+    },
+    {
+      "epoch": 0.870903151016967,
+      "grad_norm": 0.3008350133895874,
+      "learning_rate": 4.092248541057667e-06,
+      "loss": 0.6128,
+      "step": 2066
+    },
+    {
+      "epoch": 0.8713246917483402,
+      "grad_norm": 0.2751711905002594,
+      "learning_rate": 4.091385353689351e-06,
+      "loss": 0.5836,
+      "step": 2067
+    },
+    {
+      "epoch": 0.8717462324797134,
+      "grad_norm": 0.27652910351753235,
+      "learning_rate": 4.090521847243716e-06,
+      "loss": 0.6078,
+      "step": 2068
+    },
+    {
+      "epoch": 0.8721677732110865,
+      "grad_norm": 0.2825339138507843,
+      "learning_rate": 4.089658021893898e-06,
+      "loss": 0.628,
+      "step": 2069
+    },
+    {
+      "epoch": 0.8725893139424596,
+      "grad_norm": 0.27583521604537964,
+      "learning_rate": 4.088793877813095e-06,
+      "loss": 0.6123,
+      "step": 2070
+    },
+    {
+      "epoch": 0.8730108546738329,
+      "grad_norm": 0.28151851892471313,
+      "learning_rate": 4.087929415174572e-06,
+      "loss": 0.6286,
+      "step": 2071
+    },
+    {
+      "epoch": 0.873432395405206,
+      "grad_norm": 0.2782251536846161,
+      "learning_rate": 4.087064634151657e-06,
+      "loss": 0.6048,
+      "step": 2072
+    },
+    {
+      "epoch": 0.8738539361365792,
+      "grad_norm": 0.27317818999290466,
+      "learning_rate": 4.0861995349177375e-06,
+      "loss": 0.597,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8742754768679524,
+      "grad_norm": 0.2704969346523285,
+      "learning_rate": 4.085334117646272e-06,
+      "loss": 0.622,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8746970175993255,
+      "grad_norm": 0.28758692741394043,
+      "learning_rate": 4.084468382510777e-06,
+      "loss": 0.5941,
+      "step": 2075
+    },
+    {
+      "epoch": 0.8751185583306987,
+      "grad_norm": 0.2900252938270569,
+      "learning_rate": 4.083602329684836e-06,
+      "loss": 0.6328,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8755400990620719,
+      "grad_norm": 0.2709287106990814,
+      "learning_rate": 4.082735959342094e-06,
+      "loss": 0.6142,
+      "step": 2077
+    },
+    {
+      "epoch": 0.875961639793445,
+      "grad_norm": 0.28346288204193115,
+      "learning_rate": 4.081869271656263e-06,
+      "loss": 0.6258,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8763831805248182,
+      "grad_norm": 0.2880302369594574,
+      "learning_rate": 4.081002266801114e-06,
+      "loss": 0.5716,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8768047212561914,
+      "grad_norm": 0.2749912142753601,
+      "learning_rate": 4.080134944950485e-06,
+      "loss": 0.5979,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8772262619875646,
+      "grad_norm": 0.29808875918388367,
+      "learning_rate": 4.079267306278276e-06,
+      "loss": 0.5893,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8776478027189377,
+      "grad_norm": 0.2836261987686157,
+      "learning_rate": 4.078399350958452e-06,
+      "loss": 0.6049,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8780693434503108,
+      "grad_norm": 0.27648887038230896,
+      "learning_rate": 4.077531079165039e-06,
+      "loss": 0.5956,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8784908841816841,
+      "grad_norm": 0.2816130518913269,
+      "learning_rate": 4.076662491072129e-06,
+      "loss": 0.6191,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8789124249130572,
+      "grad_norm": 0.2897788882255554,
+      "learning_rate": 4.075793586853876e-06,
+      "loss": 0.5803,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8793339656444303,
+      "grad_norm": 0.2893310487270355,
+      "learning_rate": 4.0749243666844986e-06,
+      "loss": 0.6012,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8797555063758036,
+      "grad_norm": 0.29058441519737244,
+      "learning_rate": 4.074054830738277e-06,
+      "loss": 0.6248,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8801770471071767,
+      "grad_norm": 0.297395795583725,
+      "learning_rate": 4.073184979189556e-06,
+      "loss": 0.5923,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8805985878385499,
+      "grad_norm": 0.2873823344707489,
+      "learning_rate": 4.072314812212744e-06,
+      "loss": 0.6042,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8810201285699231,
+      "grad_norm": 0.293666809797287,
+      "learning_rate": 4.07144432998231e-06,
+      "loss": 0.6173,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8814416693012962,
+      "grad_norm": 0.28066980838775635,
+      "learning_rate": 4.07057353267279e-06,
+      "loss": 0.5898,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8818632100326694,
+      "grad_norm": 0.28104695677757263,
+      "learning_rate": 4.069702420458781e-06,
+      "loss": 0.5983,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8822847507640426,
+      "grad_norm": 0.2850969433784485,
+      "learning_rate": 4.0688309935149435e-06,
+      "loss": 0.5849,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8827062914954158,
+      "grad_norm": 0.2807822823524475,
+      "learning_rate": 4.067959252016e-06,
+      "loss": 0.6008,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8831278322267889,
+      "grad_norm": 0.28492605686187744,
+      "learning_rate": 4.06708719613674e-06,
+      "loss": 0.5981,
+      "step": 2095
+    },
+    {
+      "epoch": 0.883549372958162,
+      "grad_norm": 0.28317704796791077,
+      "learning_rate": 4.066214826052009e-06,
+      "loss": 0.5959,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8839709136895353,
+      "grad_norm": 0.28549128770828247,
+      "learning_rate": 4.065342141936724e-06,
+      "loss": 0.613,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8843924544209084,
+      "grad_norm": 0.2889178991317749,
+      "learning_rate": 4.064469143965857e-06,
+      "loss": 0.6036,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8848139951522815,
+      "grad_norm": 0.30202168226242065,
+      "learning_rate": 4.063595832314449e-06,
+      "loss": 0.5883,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8852355358836548,
+      "grad_norm": 0.29604771733283997,
+      "learning_rate": 4.0627222071576e-06,
+      "loss": 0.5909,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8856570766150279,
+      "grad_norm": 0.277469664812088,
+      "learning_rate": 4.061848268670475e-06,
+      "loss": 0.5871,
+      "step": 2101
+    },
+    {
+      "epoch": 0.8860786173464011,
+      "grad_norm": 0.31157124042510986,
+      "learning_rate": 4.060974017028303e-06,
+      "loss": 0.6118,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8865001580777743,
+      "grad_norm": 0.2846732437610626,
+      "learning_rate": 4.0600994524063704e-06,
+      "loss": 0.6098,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8869216988091474,
+      "grad_norm": 0.28533172607421875,
+      "learning_rate": 4.059224574980033e-06,
+      "loss": 0.6356,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8873432395405206,
+      "grad_norm": 0.29689931869506836,
+      "learning_rate": 4.0583493849247035e-06,
+      "loss": 0.6021,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8877647802718938,
+      "grad_norm": 0.28279995918273926,
+      "learning_rate": 4.057473882415862e-06,
+      "loss": 0.6133,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8881863210032669,
+      "grad_norm": 0.2909523844718933,
+      "learning_rate": 4.056598067629048e-06,
+      "loss": 0.6006,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8886078617346401,
+      "grad_norm": 0.2784798741340637,
+      "learning_rate": 4.055721940739867e-06,
+      "loss": 0.5653,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8890294024660133,
+      "grad_norm": 0.2829459309577942,
+      "learning_rate": 4.054845501923983e-06,
+      "loss": 0.6157,
+      "step": 2109
+    },
+    {
+      "epoch": 0.8894509431973865,
+      "grad_norm": 0.28051742911338806,
+      "learning_rate": 4.053968751357125e-06,
+      "loss": 0.6012,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8898724839287596,
+      "grad_norm": 0.2818240821361542,
+      "learning_rate": 4.053091689215084e-06,
+      "loss": 0.6132,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8902940246601327,
+      "grad_norm": 0.28318852186203003,
+      "learning_rate": 4.0522143156737125e-06,
+      "loss": 0.5706,
+      "step": 2112
+    },
+    {
+      "epoch": 0.890715565391506,
+      "grad_norm": 0.2676553726196289,
+      "learning_rate": 4.051336630908929e-06,
+      "loss": 0.6234,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8911371061228791,
+      "grad_norm": 0.2781136929988861,
+      "learning_rate": 4.050458635096709e-06,
+      "loss": 0.5785,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8915586468542523,
+      "grad_norm": 0.2733917236328125,
+      "learning_rate": 4.0495803284130945e-06,
+      "loss": 0.6194,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8919801875856255,
+      "grad_norm": 0.27440735697746277,
+      "learning_rate": 4.048701711034189e-06,
+      "loss": 0.6064,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8924017283169986,
+      "grad_norm": 0.2773163318634033,
+      "learning_rate": 4.047822783136156e-06,
+      "loss": 0.6064,
+      "step": 2117
+    },
+    {
+      "epoch": 0.8928232690483718,
+      "grad_norm": 0.26787060499191284,
+      "learning_rate": 4.0469435448952255e-06,
+      "loss": 0.5858,
+      "step": 2118
+    },
+    {
+      "epoch": 0.893244809779745,
+      "grad_norm": 0.28501734137535095,
+      "learning_rate": 4.0460639964876855e-06,
+      "loss": 0.6108,
+      "step": 2119
+    },
+    {
+      "epoch": 0.8936663505111181,
+      "grad_norm": 0.2865848243236542,
+      "learning_rate": 4.0451841380898885e-06,
+      "loss": 0.6145,
+      "step": 2120
+    },
+    {
+      "epoch": 0.8940878912424913,
+      "grad_norm": 0.2869279980659485,
+      "learning_rate": 4.0443039698782485e-06,
+      "loss": 0.6076,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8945094319738645,
+      "grad_norm": 0.2846921384334564,
+      "learning_rate": 4.0434234920292415e-06,
+      "loss": 0.5921,
+      "step": 2122
+    },
+    {
+      "epoch": 0.8949309727052377,
+      "grad_norm": 0.2824147045612335,
+      "learning_rate": 4.042542704719407e-06,
+      "loss": 0.5961,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8953525134366108,
+      "grad_norm": 0.28017786145210266,
+      "learning_rate": 4.041661608125344e-06,
+      "loss": 0.5939,
+      "step": 2124
+    },
+    {
+      "epoch": 0.895774054167984,
+      "grad_norm": 0.28550079464912415,
+      "learning_rate": 4.040780202423715e-06,
+      "loss": 0.6183,
+      "step": 2125
+    },
+    {
+      "epoch": 0.8961955948993572,
+      "grad_norm": 0.27808085083961487,
+      "learning_rate": 4.039898487791246e-06,
+      "loss": 0.6218,
+      "step": 2126
+    },
+    {
+      "epoch": 0.8966171356307303,
+      "grad_norm": 0.2862350344657898,
+      "learning_rate": 4.039016464404721e-06,
+      "loss": 0.6342,
+      "step": 2127
+    },
+    {
+      "epoch": 0.8970386763621034,
+      "grad_norm": 0.28022876381874084,
+      "learning_rate": 4.038134132440991e-06,
+      "loss": 0.5981,
+      "step": 2128
+    },
+    {
+      "epoch": 0.8974602170934767,
+      "grad_norm": 0.28031426668167114,
+      "learning_rate": 4.037251492076962e-06,
+      "loss": 0.605,
+      "step": 2129
+    },
+    {
+      "epoch": 0.8978817578248498,
+      "grad_norm": 0.2737732231616974,
+      "learning_rate": 4.0363685434896095e-06,
+      "loss": 0.5851,
+      "step": 2130
+    },
+    {
+      "epoch": 0.898303298556223,
+      "grad_norm": 0.2849540710449219,
+      "learning_rate": 4.0354852868559665e-06,
+      "loss": 0.5969,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8987248392875962,
+      "grad_norm": 0.27989661693573,
+      "learning_rate": 4.0346017223531284e-06,
+      "loss": 0.6348,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8991463800189693,
+      "grad_norm": 0.2764364182949066,
+      "learning_rate": 4.0337178501582505e-06,
+      "loss": 0.5799,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8995679207503425,
+      "grad_norm": 0.30946192145347595,
+      "learning_rate": 4.032833670448555e-06,
+      "loss": 0.5944,
+      "step": 2134
+    },
+    {
+      "epoch": 0.8999894614817157,
+      "grad_norm": 0.27153879404067993,
+      "learning_rate": 4.031949183401319e-06,
+      "loss": 0.6277,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9004110022130888,
+      "grad_norm": 0.2839837074279785,
+      "learning_rate": 4.031064389193888e-06,
+      "loss": 0.5693,
+      "step": 2136
+    },
+    {
+      "epoch": 0.900832542944462,
+      "grad_norm": 0.30949097871780396,
+      "learning_rate": 4.030179288003663e-06,
+      "loss": 0.5885,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9012540836758351,
+      "grad_norm": 0.2840593159198761,
+      "learning_rate": 4.029293880008111e-06,
+      "loss": 0.6036,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9016756244072084,
+      "grad_norm": 0.2975214123725891,
+      "learning_rate": 4.028408165384758e-06,
+      "loss": 0.6035,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9020971651385815,
+      "grad_norm": 0.27985626459121704,
+      "learning_rate": 4.027522144311193e-06,
+      "loss": 0.6074,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9025187058699546,
+      "grad_norm": 0.2713305354118347,
+      "learning_rate": 4.026635816965065e-06,
+      "loss": 0.6016,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9029402466013279,
+      "grad_norm": 0.2814711034297943,
+      "learning_rate": 4.025749183524087e-06,
+      "loss": 0.6192,
+      "step": 2142
+    },
+    {
+      "epoch": 0.903361787332701,
+      "grad_norm": 0.2971091568470001,
+      "learning_rate": 4.024862244166029e-06,
+      "loss": 0.5867,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9037833280640742,
+      "grad_norm": 0.2876945436000824,
+      "learning_rate": 4.023974999068726e-06,
+      "loss": 0.618,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9042048687954474,
+      "grad_norm": 0.2947264015674591,
+      "learning_rate": 4.023087448410073e-06,
+      "loss": 0.6162,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9046264095268205,
+      "grad_norm": 0.3105650246143341,
+      "learning_rate": 4.022199592368027e-06,
+      "loss": 0.568,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9050479502581937,
+      "grad_norm": 0.2721710205078125,
+      "learning_rate": 4.021311431120604e-06,
+      "loss": 0.6261,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9054694909895669,
+      "grad_norm": 0.2941567301750183,
+      "learning_rate": 4.0204229648458864e-06,
+      "loss": 0.6028,
+      "step": 2148
+    },
+    {
+      "epoch": 0.90589103172094,
+      "grad_norm": 0.2927354872226715,
+      "learning_rate": 4.019534193722011e-06,
+      "loss": 0.6009,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9063125724523132,
+      "grad_norm": 0.2825859487056732,
+      "learning_rate": 4.01864511792718e-06,
+      "loss": 0.6094,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9067341131836864,
+      "grad_norm": 0.28243502974510193,
+      "learning_rate": 4.017755737639656e-06,
+      "loss": 0.605,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9071556539150596,
+      "grad_norm": 0.2718103229999542,
+      "learning_rate": 4.016866053037762e-06,
+      "loss": 0.5913,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9075771946464327,
+      "grad_norm": 0.2946881055831909,
+      "learning_rate": 4.015976064299882e-06,
+      "loss": 0.618,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9079987353778058,
+      "grad_norm": 0.27851560711860657,
+      "learning_rate": 4.0150857716044606e-06,
+      "loss": 0.5828,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9084202761091791,
+      "grad_norm": 0.2766689360141754,
+      "learning_rate": 4.014195175130005e-06,
+      "loss": 0.5989,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9088418168405522,
+      "grad_norm": 0.2926025688648224,
+      "learning_rate": 4.013304275055083e-06,
+      "loss": 0.5808,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9092633575719253,
+      "grad_norm": 0.27221032977104187,
+      "learning_rate": 4.012413071558322e-06,
+      "loss": 0.6068,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9096848983032986,
+      "grad_norm": 0.2939716577529907,
+      "learning_rate": 4.011521564818411e-06,
+      "loss": 0.6064,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9101064390346717,
+      "grad_norm": 0.2940266728401184,
+      "learning_rate": 4.010629755014099e-06,
+      "loss": 0.6018,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9105279797660449,
+      "grad_norm": 0.2799992859363556,
+      "learning_rate": 4.009737642324197e-06,
+      "loss": 0.6157,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9109495204974181,
+      "grad_norm": 0.30319494009017944,
+      "learning_rate": 4.008845226927575e-06,
+      "loss": 0.6128,
+      "step": 2161
+    },
+    {
+      "epoch": 0.9113710612287912,
+      "grad_norm": 0.29442089796066284,
+      "learning_rate": 4.007952509003167e-06,
+      "loss": 0.6005,
+      "step": 2162
+    },
+    {
+      "epoch": 0.9117926019601644,
+      "grad_norm": 0.28662458062171936,
+      "learning_rate": 4.007059488729964e-06,
+      "loss": 0.5993,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9122141426915376,
+      "grad_norm": 0.28034350275993347,
+      "learning_rate": 4.006166166287019e-06,
+      "loss": 0.6241,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9126356834229108,
+      "grad_norm": 0.2975638806819916,
+      "learning_rate": 4.0052725418534465e-06,
+      "loss": 0.5874,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9130572241542839,
+      "grad_norm": 0.2957809567451477,
+      "learning_rate": 4.0043786156084206e-06,
+      "loss": 0.5858,
+      "step": 2166
+    },
+    {
+      "epoch": 0.913478764885657,
+      "grad_norm": 0.28277072310447693,
+      "learning_rate": 4.003484387731176e-06,
+      "loss": 0.5811,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9139003056170303,
+      "grad_norm": 0.2862344980239868,
+      "learning_rate": 4.002589858401007e-06,
+      "loss": 0.5795,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9143218463484034,
+      "grad_norm": 0.2989932894706726,
+      "learning_rate": 4.001695027797271e-06,
+      "loss": 0.5955,
+      "step": 2169
+    },
+    {
+      "epoch": 0.9147433870797765,
+      "grad_norm": 0.28623783588409424,
+      "learning_rate": 4.000799896099382e-06,
+      "loss": 0.6076,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9151649278111498,
+      "grad_norm": 0.29774290323257446,
+      "learning_rate": 3.999904463486819e-06,
+      "loss": 0.5991,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9155864685425229,
+      "grad_norm": 0.294663667678833,
+      "learning_rate": 3.999008730139117e-06,
+      "loss": 0.6146,
+      "step": 2172
+    },
+    {
+      "epoch": 0.9160080092738961,
+      "grad_norm": 0.2871439456939697,
+      "learning_rate": 3.998112696235874e-06,
+      "loss": 0.6131,
+      "step": 2173
+    },
+    {
+      "epoch": 0.9164295500052693,
+      "grad_norm": 0.273807168006897,
+      "learning_rate": 3.997216361956746e-06,
+      "loss": 0.6248,
+      "step": 2174
+    },
+    {
+      "epoch": 0.9168510907366424,
+      "grad_norm": 0.29719939827919006,
+      "learning_rate": 3.996319727481452e-06,
+      "loss": 0.5881,
+      "step": 2175
+    },
+    {
+      "epoch": 0.9172726314680156,
+      "grad_norm": 0.2858426570892334,
+      "learning_rate": 3.995422792989769e-06,
+      "loss": 0.6049,
+      "step": 2176
+    },
+    {
+      "epoch": 0.9176941721993888,
+      "grad_norm": 0.2690010964870453,
+      "learning_rate": 3.994525558661536e-06,
+      "loss": 0.6066,
+      "step": 2177
+    },
+    {
+      "epoch": 0.9181157129307619,
+      "grad_norm": 0.2945442795753479,
+      "learning_rate": 3.99362802467665e-06,
+      "loss": 0.6094,
+      "step": 2178
+    },
+    {
+      "epoch": 0.9185372536621351,
+      "grad_norm": 0.2932618260383606,
+      "learning_rate": 3.992730191215069e-06,
+      "loss": 0.6055,
+      "step": 2179
+    },
+    {
+      "epoch": 0.9189587943935082,
+      "grad_norm": 0.26866838335990906,
+      "learning_rate": 3.991832058456812e-06,
+      "loss": 0.5828,
+      "step": 2180
+    },
+    {
+      "epoch": 0.9193803351248815,
+      "grad_norm": 0.3017159402370453,
+      "learning_rate": 3.9909336265819565e-06,
+      "loss": 0.6112,
+      "step": 2181
+    },
+    {
+      "epoch": 0.9198018758562546,
+      "grad_norm": 0.2945748269557953,
+      "learning_rate": 3.9900348957706416e-06,
+      "loss": 0.5996,
+      "step": 2182
+    },
+    {
+      "epoch": 0.9202234165876277,
+      "grad_norm": 0.2749302387237549,
+      "learning_rate": 3.989135866203064e-06,
+      "loss": 0.6118,
+      "step": 2183
+    },
+    {
+      "epoch": 0.920644957319001,
+      "grad_norm": 0.2886725068092346,
+      "learning_rate": 3.988236538059483e-06,
+      "loss": 0.6015,
+      "step": 2184
+    },
+    {
+      "epoch": 0.9210664980503741,
+      "grad_norm": 0.28767129778862,
+      "learning_rate": 3.987336911520215e-06,
+      "loss": 0.6178,
+      "step": 2185
+    },
+    {
+      "epoch": 0.9214880387817472,
+      "grad_norm": 0.27905577421188354,
+      "learning_rate": 3.98643698676564e-06,
+      "loss": 0.6294,
+      "step": 2186
+    },
+    {
+      "epoch": 0.9219095795131205,
+      "grad_norm": 0.28357744216918945,
+      "learning_rate": 3.9855367639761935e-06,
+      "loss": 0.6085,
+      "step": 2187
+    },
+    {
+      "epoch": 0.9223311202444936,
+      "grad_norm": 0.2827235460281372,
+      "learning_rate": 3.984636243332373e-06,
+      "loss": 0.605,
+      "step": 2188
+    },
+    {
+      "epoch": 0.9227526609758668,
+      "grad_norm": 0.2820284068584442,
+      "learning_rate": 3.983735425014736e-06,
+      "loss": 0.5978,
+      "step": 2189
+    },
+    {
+      "epoch": 0.92317420170724,
+      "grad_norm": 0.28684645891189575,
+      "learning_rate": 3.9828343092038995e-06,
+      "loss": 0.593,
+      "step": 2190
+    },
+    {
+      "epoch": 0.9235957424386131,
+      "grad_norm": 0.26792189478874207,
+      "learning_rate": 3.981932896080539e-06,
+      "loss": 0.6164,
+      "step": 2191
+    },
+    {
+      "epoch": 0.9240172831699863,
+      "grad_norm": 0.2801945209503174,
+      "learning_rate": 3.98103118582539e-06,
+      "loss": 0.5925,
+      "step": 2192
+    },
+    {
+      "epoch": 0.9244388239013595,
+      "grad_norm": 0.29323193430900574,
+      "learning_rate": 3.98012917861925e-06,
+      "loss": 0.6089,
+      "step": 2193
+    },
+    {
+      "epoch": 0.9248603646327327,
+      "grad_norm": 0.275608092546463,
+      "learning_rate": 3.9792268746429715e-06,
+      "loss": 0.6156,
+      "step": 2194
+    },
+    {
+      "epoch": 0.9252819053641058,
+      "grad_norm": 0.28546401858329773,
+      "learning_rate": 3.978324274077471e-06,
+      "loss": 0.5979,
+      "step": 2195
+    },
+    {
+      "epoch": 0.9257034460954789,
+      "grad_norm": 0.2746219336986542,
+      "learning_rate": 3.977421377103721e-06,
+      "loss": 0.6389,
+      "step": 2196
+    },
+    {
+      "epoch": 0.9261249868268522,
+      "grad_norm": 0.2820807993412018,
+      "learning_rate": 3.976518183902757e-06,
+      "loss": 0.598,
+      "step": 2197
+    },
+    {
+      "epoch": 0.9265465275582253,
+      "grad_norm": 0.28698429465293884,
+      "learning_rate": 3.9756146946556686e-06,
+      "loss": 0.6201,
+      "step": 2198
+    },
+    {
+      "epoch": 0.9269680682895984,
+      "grad_norm": 0.2853392958641052,
+      "learning_rate": 3.974710909543611e-06,
+      "loss": 0.6018,
+      "step": 2199
+    },
+    {
+      "epoch": 0.9273896090209717,
+      "grad_norm": 0.3022915720939636,
+      "learning_rate": 3.973806828747795e-06,
+      "loss": 0.6333,
+      "step": 2200
+    },
+    {
+      "epoch": 0.9278111497523448,
+      "grad_norm": 0.2829367220401764,
+      "learning_rate": 3.97290245244949e-06,
+      "loss": 0.6063,
+      "step": 2201
+    },
+    {
+      "epoch": 0.928232690483718,
+      "grad_norm": 0.2798102796077728,
+      "learning_rate": 3.971997780830027e-06,
+      "loss": 0.6347,
+      "step": 2202
+    },
+    {
+      "epoch": 0.9286542312150912,
+      "grad_norm": 0.279094934463501,
+      "learning_rate": 3.971092814070796e-06,
+      "loss": 0.5913,
+      "step": 2203
+    },
+    {
+      "epoch": 0.9290757719464643,
+      "grad_norm": 0.2940881550312042,
+      "learning_rate": 3.970187552353244e-06,
+      "loss": 0.6,
+      "step": 2204
+    },
+    {
+      "epoch": 0.9294973126778375,
+      "grad_norm": 0.2886712849140167,
+      "learning_rate": 3.96928199585888e-06,
+      "loss": 0.5928,
+      "step": 2205
+    },
+    {
+      "epoch": 0.9299188534092107,
+      "grad_norm": 0.2864512503147125,
+      "learning_rate": 3.968376144769269e-06,
+      "loss": 0.6013,
+      "step": 2206
+    },
+    {
+      "epoch": 0.9303403941405838,
+      "grad_norm": 0.26435351371765137,
+      "learning_rate": 3.9674699992660385e-06,
+      "loss": 0.5832,
+      "step": 2207
+    },
+    {
+      "epoch": 0.930761934871957,
+      "grad_norm": 0.29388976097106934,
+      "learning_rate": 3.966563559530872e-06,
+      "loss": 0.6015,
+      "step": 2208
+    },
+    {
+      "epoch": 0.9311834756033301,
+      "grad_norm": 0.28892943263053894,
+      "learning_rate": 3.965656825745513e-06,
+      "loss": 0.6005,
+      "step": 2209
+    },
+    {
+      "epoch": 0.9316050163347034,
+      "grad_norm": 0.2897278070449829,
+      "learning_rate": 3.964749798091766e-06,
+      "loss": 0.6364,
+      "step": 2210
+    },
+    {
+      "epoch": 0.9320265570660765,
+      "grad_norm": 0.29189011454582214,
+      "learning_rate": 3.963842476751491e-06,
+      "loss": 0.6212,
+      "step": 2211
+    },
+    {
+      "epoch": 0.9324480977974496,
+      "grad_norm": 0.28161856532096863,
+      "learning_rate": 3.962934861906609e-06,
+      "loss": 0.5859,
+      "step": 2212
+    },
+    {
+      "epoch": 0.9328696385288229,
+      "grad_norm": 0.28370827436447144,
+      "learning_rate": 3.962026953739099e-06,
+      "loss": 0.6085,
+      "step": 2213
+    },
+    {
+      "epoch": 0.933291179260196,
+      "grad_norm": 0.29442110657691956,
+      "learning_rate": 3.961118752431e-06,
+      "loss": 0.6222,
+      "step": 2214
+    },
+    {
+      "epoch": 0.9337127199915692,
+      "grad_norm": 0.2824901342391968,
+      "learning_rate": 3.960210258164409e-06,
+      "loss": 0.596,
+      "step": 2215
+    },
+    {
+      "epoch": 0.9341342607229424,
+      "grad_norm": 0.2776724398136139,
+      "learning_rate": 3.959301471121481e-06,
+      "loss": 0.6049,
+      "step": 2216
+    },
+    {
+      "epoch": 0.9345558014543155,
+      "grad_norm": 0.27715787291526794,
+      "learning_rate": 3.95839239148443e-06,
+      "loss": 0.578,
+      "step": 2217
+    },
+    {
+      "epoch": 0.9349773421856887,
+      "grad_norm": 0.28056320548057556,
+      "learning_rate": 3.957483019435531e-06,
+      "loss": 0.6166,
+      "step": 2218
+    },
+    {
+      "epoch": 0.9353988829170619,
+      "grad_norm": 0.30024999380111694,
+      "learning_rate": 3.956573355157113e-06,
+      "loss": 0.6303,
+      "step": 2219
+    },
+    {
+      "epoch": 0.935820423648435,
+      "grad_norm": 0.2794478237628937,
+      "learning_rate": 3.955663398831569e-06,
+      "loss": 0.5946,
+      "step": 2220
+    },
+    {
+      "epoch": 0.9362419643798082,
+      "grad_norm": 0.28899577260017395,
+      "learning_rate": 3.954753150641346e-06,
+      "loss": 0.6316,
+      "step": 2221
+    },
+    {
+      "epoch": 0.9366635051111813,
+      "grad_norm": 0.27376696467399597,
+      "learning_rate": 3.953842610768952e-06,
+      "loss": 0.6022,
+      "step": 2222
+    },
+    {
+      "epoch": 0.9370850458425546,
+      "grad_norm": 0.29236894845962524,
+      "learning_rate": 3.952931779396953e-06,
+      "loss": 0.6022,
+      "step": 2223
+    },
+    {
+      "epoch": 0.9375065865739277,
+      "grad_norm": 0.2876782715320587,
+      "learning_rate": 3.952020656707973e-06,
+      "loss": 0.6077,
+      "step": 2224
+    },
+    {
+      "epoch": 0.9379281273053008,
+      "grad_norm": 0.2863978147506714,
+      "learning_rate": 3.951109242884694e-06,
+      "loss": 0.6278,
+      "step": 2225
+    },
+    {
+      "epoch": 0.9383496680366741,
+      "grad_norm": 0.27618297934532166,
+      "learning_rate": 3.9501975381098585e-06,
+      "loss": 0.5925,
+      "step": 2226
+    },
+    {
+      "epoch": 0.9387712087680472,
+      "grad_norm": 0.2849017083644867,
+      "learning_rate": 3.949285542566266e-06,
+      "loss": 0.6074,
+      "step": 2227
+    },
+    {
+      "epoch": 0.9391927494994203,
+      "grad_norm": 0.3025689721107483,
+      "learning_rate": 3.948373256436773e-06,
+      "loss": 0.6055,
+      "step": 2228
+    },
+    {
+      "epoch": 0.9396142902307936,
+      "grad_norm": 0.2864554226398468,
+      "learning_rate": 3.947460679904297e-06,
+      "loss": 0.6045,
+      "step": 2229
+    },
+    {
+      "epoch": 0.9400358309621667,
+      "grad_norm": 0.28702396154403687,
+      "learning_rate": 3.946547813151809e-06,
+      "loss": 0.6175,
+      "step": 2230
+    },
+    {
+      "epoch": 0.9404573716935399,
+      "grad_norm": 0.28559887409210205,
+      "learning_rate": 3.945634656362346e-06,
+      "loss": 0.6241,
+      "step": 2231
+    },
+    {
+      "epoch": 0.9408789124249131,
+      "grad_norm": 0.29609599709510803,
+      "learning_rate": 3.9447212097189945e-06,
+      "loss": 0.6133,
+      "step": 2232
+    },
+    {
+      "epoch": 0.9413004531562862,
+      "grad_norm": 0.2630096971988678,
+      "learning_rate": 3.943807473404906e-06,
+      "loss": 0.5952,
+      "step": 2233
+    },
+    {
+      "epoch": 0.9417219938876594,
+      "grad_norm": 0.28403013944625854,
+      "learning_rate": 3.942893447603284e-06,
+      "loss": 0.5923,
+      "step": 2234
+    },
+    {
+      "epoch": 0.9421435346190326,
+      "grad_norm": 0.3007854223251343,
+      "learning_rate": 3.941979132497398e-06,
+      "loss": 0.6169,
+      "step": 2235
+    },
+    {
+      "epoch": 0.9425650753504057,
+      "grad_norm": 0.2730088233947754,
+      "learning_rate": 3.941064528270566e-06,
+      "loss": 0.616,
+      "step": 2236
+    },
+    {
+      "epoch": 0.9429866160817789,
+      "grad_norm": 0.28775206208229065,
+      "learning_rate": 3.9401496351061725e-06,
+      "loss": 0.6064,
+      "step": 2237
+    },
+    {
+      "epoch": 0.943408156813152,
+      "grad_norm": 0.3401184380054474,
+      "learning_rate": 3.939234453187653e-06,
+      "loss": 0.6055,
+      "step": 2238
+    },
+    {
+      "epoch": 0.9438296975445253,
+      "grad_norm": 0.27999570965766907,
+      "learning_rate": 3.938318982698507e-06,
+      "loss": 0.6076,
+      "step": 2239
+    },
+    {
+      "epoch": 0.9442512382758984,
+      "grad_norm": 0.2810767889022827,
+      "learning_rate": 3.937403223822288e-06,
+      "loss": 0.6112,
+      "step": 2240
+    },
+    {
+      "epoch": 0.9446727790072715,
+      "grad_norm": 0.285456120967865,
+      "learning_rate": 3.936487176742607e-06,
+      "loss": 0.5906,
+      "step": 2241
+    },
+    {
+      "epoch": 0.9450943197386448,
+      "grad_norm": 0.2654629349708557,
+      "learning_rate": 3.935570841643136e-06,
+      "loss": 0.5868,
+      "step": 2242
+    },
+    {
+      "epoch": 0.9455158604700179,
+      "grad_norm": 0.27784648537635803,
+      "learning_rate": 3.934654218707601e-06,
+      "loss": 0.5785,
+      "step": 2243
+    },
+    {
+      "epoch": 0.9459374012013911,
+      "grad_norm": 0.2784576714038849,
+      "learning_rate": 3.933737308119789e-06,
+      "loss": 0.5953,
+      "step": 2244
+    },
+    {
+      "epoch": 0.9463589419327643,
+      "grad_norm": 0.27703163027763367,
+      "learning_rate": 3.932820110063541e-06,
+      "loss": 0.6094,
+      "step": 2245
+    },
+    {
+      "epoch": 0.9467804826641374,
+      "grad_norm": 0.2766515910625458,
+      "learning_rate": 3.93190262472276e-06,
+      "loss": 0.587,
+      "step": 2246
+    },
+    {
+      "epoch": 0.9472020233955106,
+      "grad_norm": 0.2773672044277191,
+      "learning_rate": 3.9309848522814045e-06,
+      "loss": 0.6043,
+      "step": 2247
+    },
+    {
+      "epoch": 0.9476235641268838,
+      "grad_norm": 0.29011252522468567,
+      "learning_rate": 3.930066792923489e-06,
+      "loss": 0.6091,
+      "step": 2248
+    },
+    {
+      "epoch": 0.9480451048582569,
+      "grad_norm": 0.31273454427719116,
+      "learning_rate": 3.929148446833089e-06,
+      "loss": 0.5832,
+      "step": 2249
+    },
+    {
+      "epoch": 0.9484666455896301,
+      "grad_norm": 0.299483984708786,
+      "learning_rate": 3.928229814194335e-06,
+      "loss": 0.6033,
+      "step": 2250
+    },
+    {
+      "epoch": 0.9488881863210032,
+      "grad_norm": 0.3058757185935974,
+      "learning_rate": 3.927310895191413e-06,
+      "loss": 0.6114,
+      "step": 2251
+    },
+    {
+      "epoch": 0.9493097270523765,
+      "grad_norm": 0.32154780626296997,
+      "learning_rate": 3.926391690008571e-06,
+      "loss": 0.6026,
+      "step": 2252
+    },
+    {
+      "epoch": 0.9497312677837496,
+      "grad_norm": 0.27533838152885437,
+      "learning_rate": 3.925472198830112e-06,
+      "loss": 0.6083,
+      "step": 2253
+    },
+    {
+      "epoch": 0.9501528085151227,
+      "grad_norm": 0.30225443840026855,
+      "learning_rate": 3.924552421840397e-06,
+      "loss": 0.5971,
+      "step": 2254
+    },
+    {
+      "epoch": 0.950574349246496,
+      "grad_norm": 0.3585510849952698,
+      "learning_rate": 3.923632359223843e-06,
+      "loss": 0.6134,
+      "step": 2255
+    },
+    {
+      "epoch": 0.9509958899778691,
+      "grad_norm": 0.2780791223049164,
+      "learning_rate": 3.922712011164926e-06,
+      "loss": 0.5796,
+      "step": 2256
+    },
+    {
+      "epoch": 0.9514174307092422,
+      "grad_norm": 0.31672948598861694,
+      "learning_rate": 3.921791377848179e-06,
+      "loss": 0.6094,
+      "step": 2257
+    },
+    {
+      "epoch": 0.9518389714406155,
+      "grad_norm": 0.3601115942001343,
+      "learning_rate": 3.9208704594581895e-06,
+      "loss": 0.5935,
+      "step": 2258
+    },
+    {
+      "epoch": 0.9522605121719886,
+      "grad_norm": 0.27946773171424866,
+      "learning_rate": 3.919949256179607e-06,
+      "loss": 0.6065,
+      "step": 2259
+    },
+    {
+      "epoch": 0.9526820529033618,
+      "grad_norm": 0.3284471929073334,
+      "learning_rate": 3.9190277681971325e-06,
+      "loss": 0.6035,
+      "step": 2260
+    },
+    {
+      "epoch": 0.953103593634735,
+      "grad_norm": 0.3442308306694031,
+      "learning_rate": 3.918105995695529e-06,
+      "loss": 0.5955,
+      "step": 2261
+    },
+    {
+      "epoch": 0.9535251343661081,
+      "grad_norm": 0.3055817186832428,
+      "learning_rate": 3.917183938859614e-06,
+      "loss": 0.5905,
+      "step": 2262
+    },
+    {
+      "epoch": 0.9539466750974813,
+      "grad_norm": 0.29305440187454224,
+      "learning_rate": 3.916261597874262e-06,
+      "loss": 0.6027,
+      "step": 2263
+    },
+    {
+      "epoch": 0.9543682158288544,
+      "grad_norm": 0.3483583331108093,
+      "learning_rate": 3.915338972924405e-06,
+      "loss": 0.5893,
+      "step": 2264
+    },
+    {
+      "epoch": 0.9547897565602277,
+      "grad_norm": 0.2935795485973358,
+      "learning_rate": 3.9144160641950336e-06,
+      "loss": 0.602,
+      "step": 2265
+    },
+    {
+      "epoch": 0.9552112972916008,
+      "grad_norm": 0.27118217945098877,
+      "learning_rate": 3.913492871871192e-06,
+      "loss": 0.6086,
+      "step": 2266
+    },
+    {
+      "epoch": 0.9556328380229739,
+      "grad_norm": 0.2959030866622925,
+      "learning_rate": 3.912569396137984e-06,
+      "loss": 0.59,
+      "step": 2267
+    },
+    {
+      "epoch": 0.9560543787543472,
+      "grad_norm": 0.3291478753089905,
+      "learning_rate": 3.911645637180568e-06,
+      "loss": 0.6072,
+      "step": 2268
+    },
+    {
+      "epoch": 0.9564759194857203,
+      "grad_norm": 0.33360931277275085,
+      "learning_rate": 3.91072159518416e-06,
+      "loss": 0.5805,
+      "step": 2269
+    },
+    {
+      "epoch": 0.9568974602170934,
+      "grad_norm": 0.28666454553604126,
+      "learning_rate": 3.909797270334035e-06,
+      "loss": 0.5875,
+      "step": 2270
+    },
+    {
+      "epoch": 0.9573190009484667,
+      "grad_norm": 0.3093635141849518,
+      "learning_rate": 3.908872662815521e-06,
+      "loss": 0.5923,
+      "step": 2271
+    },
+    {
+      "epoch": 0.9577405416798398,
+      "grad_norm": 0.28461939096450806,
+      "learning_rate": 3.907947772814006e-06,
+      "loss": 0.6116,
+      "step": 2272
+    },
+    {
+      "epoch": 0.958162082411213,
+      "grad_norm": 0.282041996717453,
+      "learning_rate": 3.907022600514931e-06,
+      "loss": 0.6101,
+      "step": 2273
+    },
+    {
+      "epoch": 0.9585836231425862,
+      "grad_norm": 0.281760036945343,
+      "learning_rate": 3.9060971461037965e-06,
+      "loss": 0.6207,
+      "step": 2274
+    },
+    {
+      "epoch": 0.9590051638739593,
+      "grad_norm": 0.28164204955101013,
+      "learning_rate": 3.905171409766161e-06,
+      "loss": 0.6052,
+      "step": 2275
+    },
+    {
+      "epoch": 0.9594267046053325,
+      "grad_norm": 0.31009265780448914,
+      "learning_rate": 3.904245391687635e-06,
+      "loss": 0.6141,
+      "step": 2276
+    },
+    {
+      "epoch": 0.9598482453367057,
+      "grad_norm": 0.28976762294769287,
+      "learning_rate": 3.903319092053886e-06,
+      "loss": 0.6034,
+      "step": 2277
+    },
+    {
+      "epoch": 0.9602697860680788,
+      "grad_norm": 0.2753479480743408,
+      "learning_rate": 3.902392511050645e-06,
+      "loss": 0.6004,
+      "step": 2278
+    },
+    {
+      "epoch": 0.960691326799452,
+      "grad_norm": 0.2792729139328003,
+      "learning_rate": 3.90146564886369e-06,
+      "loss": 0.604,
+      "step": 2279
+    },
+    {
+      "epoch": 0.9611128675308251,
+      "grad_norm": 0.27799367904663086,
+      "learning_rate": 3.900538505678859e-06,
+      "loss": 0.6119,
+      "step": 2280
+    },
+    {
+      "epoch": 0.9615344082621984,
+      "grad_norm": 0.30479997396469116,
+      "learning_rate": 3.89961108168205e-06,
+      "loss": 0.6084,
+      "step": 2281
+    },
+    {
+      "epoch": 0.9619559489935715,
+      "grad_norm": 0.2862239480018616,
+      "learning_rate": 3.8986833770592125e-06,
+      "loss": 0.6349,
+      "step": 2282
+    },
+    {
+      "epoch": 0.9623774897249446,
+      "grad_norm": 0.2916741967201233,
+      "learning_rate": 3.897755391996354e-06,
+      "loss": 0.5836,
+      "step": 2283
+    },
+    {
+      "epoch": 0.9627990304563179,
+      "grad_norm": 0.30355191230773926,
+      "learning_rate": 3.896827126679537e-06,
+      "loss": 0.5871,
+      "step": 2284
+    },
+    {
+      "epoch": 0.963220571187691,
+      "grad_norm": 0.29013755917549133,
+      "learning_rate": 3.8958985812948844e-06,
+      "loss": 0.6115,
+      "step": 2285
+    },
+    {
+      "epoch": 0.9636421119190641,
+      "grad_norm": 0.2741113305091858,
+      "learning_rate": 3.8949697560285695e-06,
+      "loss": 0.6003,
+      "step": 2286
+    },
+    {
+      "epoch": 0.9640636526504374,
+      "grad_norm": 0.28248217701911926,
+      "learning_rate": 3.894040651066826e-06,
+      "loss": 0.6074,
+      "step": 2287
+    },
+    {
+      "epoch": 0.9644851933818105,
+      "grad_norm": 0.29614850878715515,
+      "learning_rate": 3.893111266595941e-06,
+      "loss": 0.6123,
+      "step": 2288
+    },
+    {
+      "epoch": 0.9649067341131837,
+      "grad_norm": 0.26676252484321594,
+      "learning_rate": 3.892181602802259e-06,
+      "loss": 0.6171,
+      "step": 2289
+    },
+    {
+      "epoch": 0.9653282748445569,
+      "grad_norm": 0.28357061743736267,
+      "learning_rate": 3.891251659872181e-06,
+      "loss": 0.5868,
+      "step": 2290
+    },
+    {
+      "epoch": 0.96574981557593,
+      "grad_norm": 0.2830214202404022,
+      "learning_rate": 3.8903214379921614e-06,
+      "loss": 0.6049,
+      "step": 2291
+    },
+    {
+      "epoch": 0.9661713563073032,
+      "grad_norm": 0.28438079357147217,
+      "learning_rate": 3.889390937348715e-06,
+      "loss": 0.616,
+      "step": 2292
+    },
+    {
+      "epoch": 0.9665928970386763,
+      "grad_norm": 0.2946721613407135,
+      "learning_rate": 3.8884601581284075e-06,
+      "loss": 0.5774,
+      "step": 2293
+    },
+    {
+      "epoch": 0.9670144377700496,
+      "grad_norm": 0.26881149411201477,
+      "learning_rate": 3.887529100517864e-06,
+      "loss": 0.6072,
+      "step": 2294
+    },
+    {
+      "epoch": 0.9674359785014227,
+      "grad_norm": 0.2788190543651581,
+      "learning_rate": 3.886597764703763e-06,
+      "loss": 0.6226,
+      "step": 2295
+    },
+    {
+      "epoch": 0.9678575192327958,
+      "grad_norm": 0.3008344769477844,
+      "learning_rate": 3.885666150872843e-06,
+      "loss": 0.5968,
+      "step": 2296
+    },
+    {
+      "epoch": 0.9682790599641691,
+      "grad_norm": 0.27828311920166016,
+      "learning_rate": 3.884734259211892e-06,
+      "loss": 0.5991,
+      "step": 2297
+    },
+    {
+      "epoch": 0.9687006006955422,
+      "grad_norm": 0.29506435990333557,
+      "learning_rate": 3.883802089907758e-06,
+      "loss": 0.6074,
+      "step": 2298
+    },
+    {
+      "epoch": 0.9691221414269153,
+      "grad_norm": 0.2771647870540619,
+      "learning_rate": 3.882869643147344e-06,
+      "loss": 0.6181,
+      "step": 2299
+    },
+    {
+      "epoch": 0.9695436821582886,
+      "grad_norm": 0.29433438181877136,
+      "learning_rate": 3.881936919117608e-06,
+      "loss": 0.6017,
+      "step": 2300
+    },
+    {
+      "epoch": 0.9699652228896617,
+      "grad_norm": 0.2772413194179535,
+      "learning_rate": 3.881003918005564e-06,
+      "loss": 0.5979,
+      "step": 2301
+    },
+    {
+      "epoch": 0.9703867636210349,
+      "grad_norm": 0.2829548120498657,
+      "learning_rate": 3.88007063999828e-06,
+      "loss": 0.62,
+      "step": 2302
+    },
+    {
+      "epoch": 0.9708083043524081,
+      "grad_norm": 0.2813149094581604,
+      "learning_rate": 3.879137085282883e-06,
+      "loss": 0.6179,
+      "step": 2303
+    },
+    {
+      "epoch": 0.9712298450837812,
+      "grad_norm": 0.27367615699768066,
+      "learning_rate": 3.878203254046553e-06,
+      "loss": 0.6103,
+      "step": 2304
+    },
+    {
+      "epoch": 0.9716513858151544,
+      "grad_norm": 0.28040611743927,
+      "learning_rate": 3.8772691464765244e-06,
+      "loss": 0.5811,
+      "step": 2305
+    },
+    {
+      "epoch": 0.9720729265465275,
+      "grad_norm": 0.2823663055896759,
+      "learning_rate": 3.87633476276009e-06,
+      "loss": 0.5909,
+      "step": 2306
+    },
+    {
+      "epoch": 0.9724944672779007,
+      "grad_norm": 0.29227545857429504,
+      "learning_rate": 3.875400103084596e-06,
+      "loss": 0.5947,
+      "step": 2307
+    },
+    {
+      "epoch": 0.9729160080092739,
+      "grad_norm": 0.2901232838630676,
+      "learning_rate": 3.874465167637443e-06,
+      "loss": 0.5743,
+      "step": 2308
+    },
+    {
+      "epoch": 0.973337548740647,
+      "grad_norm": 0.27965834736824036,
+      "learning_rate": 3.873529956606091e-06,
+      "loss": 0.615,
+      "step": 2309
+    },
+    {
+      "epoch": 0.9737590894720203,
+      "grad_norm": 0.2752435803413391,
+      "learning_rate": 3.87259447017805e-06,
+      "loss": 0.6172,
+      "step": 2310
+    },
+    {
+      "epoch": 0.9741806302033934,
+      "grad_norm": 0.2852821350097656,
+      "learning_rate": 3.871658708540888e-06,
+      "loss": 0.5948,
+      "step": 2311
+    },
+    {
+      "epoch": 0.9746021709347665,
+      "grad_norm": 0.29288598895072937,
+      "learning_rate": 3.8707226718822295e-06,
+      "loss": 0.6041,
+      "step": 2312
+    },
+    {
+      "epoch": 0.9750237116661398,
+      "grad_norm": 0.2894674837589264,
+      "learning_rate": 3.8697863603897515e-06,
+      "loss": 0.6066,
+      "step": 2313
+    },
+    {
+      "epoch": 0.9754452523975129,
+      "grad_norm": 0.28330737352371216,
+      "learning_rate": 3.868849774251186e-06,
+      "loss": 0.5978,
+      "step": 2314
+    },
+    {
+      "epoch": 0.975866793128886,
+      "grad_norm": 0.2776613235473633,
+      "learning_rate": 3.867912913654324e-06,
+      "loss": 0.6046,
+      "step": 2315
+    },
+    {
+      "epoch": 0.9762883338602593,
+      "grad_norm": 0.28327205777168274,
+      "learning_rate": 3.866975778787006e-06,
+      "loss": 0.5907,
+      "step": 2316
+    },
+    {
+      "epoch": 0.9767098745916324,
+      "grad_norm": 0.28314539790153503,
+      "learning_rate": 3.866038369837131e-06,
+      "loss": 0.6052,
+      "step": 2317
+    },
+    {
+      "epoch": 0.9771314153230056,
+      "grad_norm": 0.29251641035079956,
+      "learning_rate": 3.865100686992652e-06,
+      "loss": 0.5939,
+      "step": 2318
+    },
+    {
+      "epoch": 0.9775529560543788,
+      "grad_norm": 0.2855965793132782,
+      "learning_rate": 3.864162730441579e-06,
+      "loss": 0.6067,
+      "step": 2319
+    },
+    {
+      "epoch": 0.9779744967857519,
+      "grad_norm": 0.28611767292022705,
+      "learning_rate": 3.863224500371973e-06,
+      "loss": 0.6204,
+      "step": 2320
+    },
+    {
+      "epoch": 0.9783960375171251,
+      "grad_norm": 0.2908514440059662,
+      "learning_rate": 3.862285996971952e-06,
+      "loss": 0.6402,
+      "step": 2321
+    },
+    {
+      "epoch": 0.9788175782484982,
+      "grad_norm": 0.28100308775901794,
+      "learning_rate": 3.861347220429689e-06,
+      "loss": 0.6074,
+      "step": 2322
+    },
+    {
+      "epoch": 0.9792391189798715,
+      "grad_norm": 0.2851255536079407,
+      "learning_rate": 3.860408170933412e-06,
+      "loss": 0.6179,
+      "step": 2323
+    },
+    {
+      "epoch": 0.9796606597112446,
+      "grad_norm": 0.3015802800655365,
+      "learning_rate": 3.859468848671401e-06,
+      "loss": 0.6111,
+      "step": 2324
+    },
+    {
+      "epoch": 0.9800822004426177,
+      "grad_norm": 0.26925206184387207,
+      "learning_rate": 3.858529253831996e-06,
+      "loss": 0.574,
+      "step": 2325
+    },
+    {
+      "epoch": 0.980503741173991,
+      "grad_norm": 0.2821880578994751,
+      "learning_rate": 3.857589386603586e-06,
+      "loss": 0.5811,
+      "step": 2326
+    },
+    {
+      "epoch": 0.9809252819053641,
+      "grad_norm": 0.2831347584724426,
+      "learning_rate": 3.856649247174618e-06,
+      "loss": 0.6052,
+      "step": 2327
+    },
+    {
+      "epoch": 0.9813468226367372,
+      "grad_norm": 0.2848794460296631,
+      "learning_rate": 3.855708835733592e-06,
+      "loss": 0.5959,
+      "step": 2328
+    },
+    {
+      "epoch": 0.9817683633681105,
+      "grad_norm": 0.27481842041015625,
+      "learning_rate": 3.8547681524690636e-06,
+      "loss": 0.6096,
+      "step": 2329
+    },
+    {
+      "epoch": 0.9821899040994836,
+      "grad_norm": 0.2689361870288849,
+      "learning_rate": 3.853827197569643e-06,
+      "loss": 0.6068,
+      "step": 2330
+    },
+    {
+      "epoch": 0.9826114448308568,
+      "grad_norm": 0.2996506094932556,
+      "learning_rate": 3.852885971223994e-06,
+      "loss": 0.6057,
+      "step": 2331
+    },
+    {
+      "epoch": 0.98303298556223,
+      "grad_norm": 0.27831658720970154,
+      "learning_rate": 3.851944473620834e-06,
+      "loss": 0.6407,
+      "step": 2332
+    },
+    {
+      "epoch": 0.9834545262936031,
+      "grad_norm": 0.2729586958885193,
+      "learning_rate": 3.8510027049489375e-06,
+      "loss": 0.6112,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9838760670249763,
+      "grad_norm": 0.2839052975177765,
+      "learning_rate": 3.850060665397132e-06,
+      "loss": 0.6031,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9842976077563494,
+      "grad_norm": 0.277802437543869,
+      "learning_rate": 3.849118355154298e-06,
+      "loss": 0.6166,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9847191484877226,
+      "grad_norm": 0.29497143626213074,
+      "learning_rate": 3.848175774409371e-06,
+      "loss": 0.5952,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9851406892190958,
+      "grad_norm": 0.29097846150398254,
+      "learning_rate": 3.847232923351343e-06,
+      "loss": 0.6188,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9855622299504689,
+      "grad_norm": 0.28244611620903015,
+      "learning_rate": 3.846289802169257e-06,
+      "loss": 0.6027,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9859837706818422,
+      "grad_norm": 0.2995868921279907,
+      "learning_rate": 3.845346411052211e-06,
+      "loss": 0.6142,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9864053114132153,
+      "grad_norm": 0.2752271294593811,
+      "learning_rate": 3.84440275018936e-06,
+      "loss": 0.6136,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9868268521445884,
+      "grad_norm": 0.2759368419647217,
+      "learning_rate": 3.843458819769908e-06,
+      "loss": 0.604,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9872483928759617,
+      "grad_norm": 0.2889411747455597,
+      "learning_rate": 3.842514619983118e-06,
+      "loss": 0.6046,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9876699336073348,
+      "grad_norm": 0.29494187235832214,
+      "learning_rate": 3.8415701510183014e-06,
+      "loss": 0.5794,
+      "step": 2343
+    },
+    {
+      "epoch": 0.988091474338708,
+      "grad_norm": 0.27018147706985474,
+      "learning_rate": 3.840625413064831e-06,
+      "loss": 0.613,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9885130150700812,
+      "grad_norm": 0.27789413928985596,
+      "learning_rate": 3.8396804063121275e-06,
+      "loss": 0.5997,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9889345558014543,
+      "grad_norm": 0.30701005458831787,
+      "learning_rate": 3.8387351309496674e-06,
+      "loss": 0.5921,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9893560965328275,
+      "grad_norm": 0.2997194826602936,
+      "learning_rate": 3.837789587166981e-06,
+      "loss": 0.6125,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9897776372642006,
+      "grad_norm": 0.28274911642074585,
+      "learning_rate": 3.836843775153654e-06,
+      "loss": 0.6087,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9901991779955738,
+      "grad_norm": 0.29039493203163147,
+      "learning_rate": 3.835897695099322e-06,
+      "loss": 0.6195,
+      "step": 2349
+    },
+    {
+      "epoch": 0.990620718726947,
+      "grad_norm": 0.2735041081905365,
+      "learning_rate": 3.834951347193679e-06,
+      "loss": 0.5989,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9910422594583201,
+      "grad_norm": 0.2798038721084595,
+      "learning_rate": 3.8340047316264685e-06,
+      "loss": 0.6231,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9914638001896934,
+      "grad_norm": 0.2791386842727661,
+      "learning_rate": 3.83305784858749e-06,
+      "loss": 0.6082,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9918853409210665,
+      "grad_norm": 0.2851346433162689,
+      "learning_rate": 3.832110698266598e-06,
+      "loss": 0.6046,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9923068816524396,
+      "grad_norm": 0.28105953335762024,
+      "learning_rate": 3.831163280853698e-06,
+      "loss": 0.5793,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9927284223838129,
+      "grad_norm": 0.28375425934791565,
+      "learning_rate": 3.830215596538749e-06,
+      "loss": 0.6022,
+      "step": 2355
+    },
+    {
+      "epoch": 0.993149963115186,
+      "grad_norm": 0.29085788130760193,
+      "learning_rate": 3.8292676455117665e-06,
+      "loss": 0.6111,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9935715038465591,
+      "grad_norm": 0.28563249111175537,
+      "learning_rate": 3.8283194279628136e-06,
+      "loss": 0.5985,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9939930445779324,
+      "grad_norm": 0.2845872938632965,
+      "learning_rate": 3.827370944082014e-06,
+      "loss": 0.5824,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9944145853093055,
+      "grad_norm": 0.27961254119873047,
+      "learning_rate": 3.8264221940595415e-06,
+      "loss": 0.601,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9948361260406787,
+      "grad_norm": 0.2869557738304138,
+      "learning_rate": 3.825473178085621e-06,
+      "loss": 0.5925,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9952576667720519,
+      "grad_norm": 0.28052857518196106,
+      "learning_rate": 3.824523896350533e-06,
+      "loss": 0.6276,
+      "step": 2361
+    },
+    {
+      "epoch": 0.995679207503425,
+      "grad_norm": 0.28515782952308655,
+      "learning_rate": 3.823574349044614e-06,
+      "loss": 0.6044,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9961007482347982,
+      "grad_norm": 0.2769378125667572,
+      "learning_rate": 3.822624536358249e-06,
+      "loss": 0.6143,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9965222889661713,
+      "grad_norm": 0.27190500497817993,
+      "learning_rate": 3.8216744584818775e-06,
+      "loss": 0.5809,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9969438296975445,
+      "grad_norm": 0.28445911407470703,
+      "learning_rate": 3.8207241156059945e-06,
+      "loss": 0.6106,
+      "step": 2365
+    },
+    {
+      "epoch": 0.9973653704289177,
+      "grad_norm": 0.29215818643569946,
+      "learning_rate": 3.819773507921145e-06,
+      "loss": 0.606,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9977869111602908,
+      "grad_norm": 0.2882162630558014,
+      "learning_rate": 3.818822635617929e-06,
+      "loss": 0.5858,
+      "step": 2367
+    },
+    {
+      "epoch": 0.9982084518916641,
+      "grad_norm": 0.28893187642097473,
+      "learning_rate": 3.817871498887e-06,
+      "loss": 0.5993,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9986299926230372,
+      "grad_norm": 0.28756967186927795,
+      "learning_rate": 3.816920097919063e-06,
+      "loss": 0.582,
+      "step": 2369
+    },
+    {
+      "epoch": 0.9990515333544103,
+      "grad_norm": 0.2795482873916626,
+      "learning_rate": 3.815968432904877e-06,
+      "loss": 0.612,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9994730740857836,
+      "grad_norm": 0.2794437110424042,
+      "learning_rate": 3.815016504035252e-06,
+      "loss": 0.6052,
+      "step": 2371
+    },
+    {
+      "epoch": 0.9998946148171567,
+      "grad_norm": 0.26533517241477966,
+      "learning_rate": 3.8140643115010546e-06,
+      "loss": 0.5968,
+      "step": 2372
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7116,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 2372,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.125410346482008e+19,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}