diff --git "a/checkpoint-4744/trainer_state.json" "b/checkpoint-4744/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-4744/trainer_state.json"
@@ -0,0 +1,33241 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9994730740857836,
+  "eval_steps": 500,
+  "global_step": 4744,
+  "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
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.26533517241477966,
+      "learning_rate": 3.8131118554932e-06,
+      "loss": 0.6352,
+      "step": 2373
+    },
+    {
+      "epoch": 1.0004215407313732,
+      "grad_norm": 0.5903716683387756,
+      "learning_rate": 3.8121591362026604e-06,
+      "loss": 0.6096,
+      "step": 2374
+    },
+    {
+      "epoch": 1.0008430814627463,
+      "grad_norm": 0.27815574407577515,
+      "learning_rate": 3.8112061538204575e-06,
+      "loss": 0.6001,
+      "step": 2375
+    },
+    {
+      "epoch": 1.0012646221941195,
+      "grad_norm": 0.28100040555000305,
+      "learning_rate": 3.8102529085376673e-06,
+      "loss": 0.6063,
+      "step": 2376
+    },
+    {
+      "epoch": 1.0016861629254927,
+      "grad_norm": 0.27900347113609314,
+      "learning_rate": 3.8092994005454185e-06,
+      "loss": 0.5896,
+      "step": 2377
+    },
+    {
+      "epoch": 1.0021077036568657,
+      "grad_norm": 0.28961941599845886,
+      "learning_rate": 3.808345630034892e-06,
+      "loss": 0.6143,
+      "step": 2378
+    },
+    {
+      "epoch": 1.002529244388239,
+      "grad_norm": 0.2715349495410919,
+      "learning_rate": 3.8073915971973206e-06,
+      "loss": 0.6064,
+      "step": 2379
+    },
+    {
+      "epoch": 1.0029507851196122,
+      "grad_norm": 0.28668367862701416,
+      "learning_rate": 3.8064373022239925e-06,
+      "loss": 0.6137,
+      "step": 2380
+    },
+    {
+      "epoch": 1.0033723258509855,
+      "grad_norm": 0.2908131182193756,
+      "learning_rate": 3.805482745306245e-06,
+      "loss": 0.6147,
+      "step": 2381
+    },
+    {
+      "epoch": 1.0037938665823585,
+      "grad_norm": 0.2781730890274048,
+      "learning_rate": 3.8045279266354696e-06,
+      "loss": 0.6233,
+      "step": 2382
+    },
+    {
+      "epoch": 1.0042154073137317,
+      "grad_norm": 0.28361013531684875,
+      "learning_rate": 3.803572846403111e-06,
+      "loss": 0.6009,
+      "step": 2383
+    },
+    {
+      "epoch": 1.004636948045105,
+      "grad_norm": 0.29377511143684387,
+      "learning_rate": 3.8026175048006653e-06,
+      "loss": 0.5998,
+      "step": 2384
+    },
+    {
+      "epoch": 1.005058488776478,
+      "grad_norm": 0.2818347215652466,
+      "learning_rate": 3.801661902019681e-06,
+      "loss": 0.5866,
+      "step": 2385
+    },
+    {
+      "epoch": 1.0054800295078512,
+      "grad_norm": 0.2711334824562073,
+      "learning_rate": 3.8007060382517598e-06,
+      "loss": 0.5704,
+      "step": 2386
+    },
+    {
+      "epoch": 1.0059015702392244,
+      "grad_norm": 0.2865384817123413,
+      "learning_rate": 3.7997499136885543e-06,
+      "loss": 0.5957,
+      "step": 2387
+    },
+    {
+      "epoch": 1.0063231109705975,
+      "grad_norm": 0.27715036273002625,
+      "learning_rate": 3.79879352852177e-06,
+      "loss": 0.585,
+      "step": 2388
+    },
+    {
+      "epoch": 1.0067446517019707,
+      "grad_norm": 0.28890323638916016,
+      "learning_rate": 3.7978368829431657e-06,
+      "loss": 0.6108,
+      "step": 2389
+    },
+    {
+      "epoch": 1.007166192433344,
+      "grad_norm": 0.29252752661705017,
+      "learning_rate": 3.79687997714455e-06,
+      "loss": 0.606,
+      "step": 2390
+    },
+    {
+      "epoch": 1.007587733164717,
+      "grad_norm": 0.29419147968292236,
+      "learning_rate": 3.7959228113177883e-06,
+      "loss": 0.6145,
+      "step": 2391
+    },
+    {
+      "epoch": 1.0080092738960902,
+      "grad_norm": 0.29228246212005615,
+      "learning_rate": 3.794965385654792e-06,
+      "loss": 0.6061,
+      "step": 2392
+    },
+    {
+      "epoch": 1.0084308146274634,
+      "grad_norm": 0.2752278447151184,
+      "learning_rate": 3.7940077003475296e-06,
+      "loss": 0.5735,
+      "step": 2393
+    },
+    {
+      "epoch": 1.0088523553588364,
+      "grad_norm": 0.28382110595703125,
+      "learning_rate": 3.7930497555880184e-06,
+      "loss": 0.5891,
+      "step": 2394
+    },
+    {
+      "epoch": 1.0092738960902097,
+      "grad_norm": 0.28895801305770874,
+      "learning_rate": 3.7920915515683303e-06,
+      "loss": 0.5817,
+      "step": 2395
+    },
+    {
+      "epoch": 1.009695436821583,
+      "grad_norm": 0.2854900062084198,
+      "learning_rate": 3.7911330884805876e-06,
+      "loss": 0.5724,
+      "step": 2396
+    },
+    {
+      "epoch": 1.0101169775529562,
+      "grad_norm": 0.28303125500679016,
+      "learning_rate": 3.790174366516964e-06,
+      "loss": 0.597,
+      "step": 2397
+    },
+    {
+      "epoch": 1.0105385182843292,
+      "grad_norm": 0.2916780710220337,
+      "learning_rate": 3.7892153858696867e-06,
+      "loss": 0.6163,
+      "step": 2398
+    },
+    {
+      "epoch": 1.0109600590157024,
+      "grad_norm": 0.27739304304122925,
+      "learning_rate": 3.7882561467310345e-06,
+      "loss": 0.5901,
+      "step": 2399
+    },
+    {
+      "epoch": 1.0113815997470756,
+      "grad_norm": 0.292867511510849,
+      "learning_rate": 3.787296649293337e-06,
+      "loss": 0.5897,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0118031404784487,
+      "grad_norm": 0.27453479170799255,
+      "learning_rate": 3.7863368937489766e-06,
+      "loss": 0.5868,
+      "step": 2401
+    },
+    {
+      "epoch": 1.012224681209822,
+      "grad_norm": 0.2932438552379608,
+      "learning_rate": 3.785376880290387e-06,
+      "loss": 0.5807,
+      "step": 2402
+    },
+    {
+      "epoch": 1.0126462219411951,
+      "grad_norm": 0.28369542956352234,
+      "learning_rate": 3.784416609110052e-06,
+      "loss": 0.6002,
+      "step": 2403
+    },
+    {
+      "epoch": 1.0130677626725682,
+      "grad_norm": 0.28187182545661926,
+      "learning_rate": 3.7834560804005117e-06,
+      "loss": 0.5782,
+      "step": 2404
+    },
+    {
+      "epoch": 1.0134893034039414,
+      "grad_norm": 0.28517404198646545,
+      "learning_rate": 3.7824952943543523e-06,
+      "loss": 0.6027,
+      "step": 2405
+    },
+    {
+      "epoch": 1.0139108441353146,
+      "grad_norm": 0.2758466303348541,
+      "learning_rate": 3.781534251164216e-06,
+      "loss": 0.5952,
+      "step": 2406
+    },
+    {
+      "epoch": 1.0143323848666876,
+      "grad_norm": 0.27708351612091064,
+      "learning_rate": 3.7805729510227933e-06,
+      "loss": 0.5991,
+      "step": 2407
+    },
+    {
+      "epoch": 1.0147539255980609,
+      "grad_norm": 0.30474427342414856,
+      "learning_rate": 3.7796113941228285e-06,
+      "loss": 0.5809,
+      "step": 2408
+    },
+    {
+      "epoch": 1.0151754663294341,
+      "grad_norm": 0.2769697904586792,
+      "learning_rate": 3.778649580657116e-06,
+      "loss": 0.5955,
+      "step": 2409
+    },
+    {
+      "epoch": 1.0155970070608074,
+      "grad_norm": 0.290905237197876,
+      "learning_rate": 3.777687510818502e-06,
+      "loss": 0.6151,
+      "step": 2410
+    },
+    {
+      "epoch": 1.0160185477921804,
+      "grad_norm": 0.2769460380077362,
+      "learning_rate": 3.7767251847998846e-06,
+      "loss": 0.6034,
+      "step": 2411
+    },
+    {
+      "epoch": 1.0164400885235536,
+      "grad_norm": 0.277402400970459,
+      "learning_rate": 3.7757626027942137e-06,
+      "loss": 0.5964,
+      "step": 2412
+    },
+    {
+      "epoch": 1.0168616292549268,
+      "grad_norm": 0.29812806844711304,
+      "learning_rate": 3.7747997649944885e-06,
+      "loss": 0.62,
+      "step": 2413
+    },
+    {
+      "epoch": 1.0172831699862999,
+      "grad_norm": 0.2880840599536896,
+      "learning_rate": 3.7738366715937615e-06,
+      "loss": 0.5696,
+      "step": 2414
+    },
+    {
+      "epoch": 1.017704710717673,
+      "grad_norm": 0.2917765974998474,
+      "learning_rate": 3.7728733227851342e-06,
+      "loss": 0.6035,
+      "step": 2415
+    },
+    {
+      "epoch": 1.0181262514490463,
+      "grad_norm": 0.2949848473072052,
+      "learning_rate": 3.7719097187617636e-06,
+      "loss": 0.5963,
+      "step": 2416
+    },
+    {
+      "epoch": 1.0185477921804194,
+      "grad_norm": 0.29505565762519836,
+      "learning_rate": 3.7709458597168527e-06,
+      "loss": 0.5917,
+      "step": 2417
+    },
+    {
+      "epoch": 1.0189693329117926,
+      "grad_norm": 0.28490954637527466,
+      "learning_rate": 3.769981745843659e-06,
+      "loss": 0.6247,
+      "step": 2418
+    },
+    {
+      "epoch": 1.0193908736431658,
+      "grad_norm": 0.30833199620246887,
+      "learning_rate": 3.7690173773354895e-06,
+      "loss": 0.5906,
+      "step": 2419
+    },
+    {
+      "epoch": 1.0198124143745388,
+      "grad_norm": 0.2938544452190399,
+      "learning_rate": 3.768052754385703e-06,
+      "loss": 0.5893,
+      "step": 2420
+    },
+    {
+      "epoch": 1.020233955105912,
+      "grad_norm": 0.2891618311405182,
+      "learning_rate": 3.7670878771877096e-06,
+      "loss": 0.5859,
+      "step": 2421
+    },
+    {
+      "epoch": 1.0206554958372853,
+      "grad_norm": 0.3134137690067291,
+      "learning_rate": 3.766122745934969e-06,
+      "loss": 0.5904,
+      "step": 2422
+    },
+    {
+      "epoch": 1.0210770365686583,
+      "grad_norm": 0.2937788963317871,
+      "learning_rate": 3.765157360820995e-06,
+      "loss": 0.6053,
+      "step": 2423
+    },
+    {
+      "epoch": 1.0214985773000316,
+      "grad_norm": 0.2848683297634125,
+      "learning_rate": 3.7641917220393474e-06,
+      "loss": 0.6147,
+      "step": 2424
+    },
+    {
+      "epoch": 1.0219201180314048,
+      "grad_norm": 0.2970363199710846,
+      "learning_rate": 3.7632258297836417e-06,
+      "loss": 0.5987,
+      "step": 2425
+    },
+    {
+      "epoch": 1.022341658762778,
+      "grad_norm": 0.30204838514328003,
+      "learning_rate": 3.76225968424754e-06,
+      "loss": 0.5915,
+      "step": 2426
+    },
+    {
+      "epoch": 1.022763199494151,
+      "grad_norm": 0.28239697217941284,
+      "learning_rate": 3.761293285624759e-06,
+      "loss": 0.6072,
+      "step": 2427
+    },
+    {
+      "epoch": 1.0231847402255243,
+      "grad_norm": 0.2941496968269348,
+      "learning_rate": 3.760326634109064e-06,
+      "loss": 0.6074,
+      "step": 2428
+    },
+    {
+      "epoch": 1.0236062809568975,
+      "grad_norm": 0.2880486249923706,
+      "learning_rate": 3.7593597298942697e-06,
+      "loss": 0.6208,
+      "step": 2429
+    },
+    {
+      "epoch": 1.0240278216882706,
+      "grad_norm": 0.31154540181159973,
+      "learning_rate": 3.7583925731742456e-06,
+      "loss": 0.6177,
+      "step": 2430
+    },
+    {
+      "epoch": 1.0244493624196438,
+      "grad_norm": 0.30512574315071106,
+      "learning_rate": 3.757425164142907e-06,
+      "loss": 0.597,
+      "step": 2431
+    },
+    {
+      "epoch": 1.024870903151017,
+      "grad_norm": 0.3082779347896576,
+      "learning_rate": 3.756457502994225e-06,
+      "loss": 0.5899,
+      "step": 2432
+    },
+    {
+      "epoch": 1.02529244388239,
+      "grad_norm": 0.29537060856819153,
+      "learning_rate": 3.7554895899222166e-06,
+      "loss": 0.6169,
+      "step": 2433
+    },
+    {
+      "epoch": 1.0257139846137633,
+      "grad_norm": 0.29673340916633606,
+      "learning_rate": 3.7545214251209516e-06,
+      "loss": 0.5915,
+      "step": 2434
+    },
+    {
+      "epoch": 1.0261355253451365,
+      "grad_norm": 0.2857362926006317,
+      "learning_rate": 3.7535530087845496e-06,
+      "loss": 0.5807,
+      "step": 2435
+    },
+    {
+      "epoch": 1.0265570660765095,
+      "grad_norm": 0.2983939051628113,
+      "learning_rate": 3.7525843411071807e-06,
+      "loss": 0.5864,
+      "step": 2436
+    },
+    {
+      "epoch": 1.0269786068078828,
+      "grad_norm": 0.2947620153427124,
+      "learning_rate": 3.7516154222830657e-06,
+      "loss": 0.5888,
+      "step": 2437
+    },
+    {
+      "epoch": 1.027400147539256,
+      "grad_norm": 0.30119210481643677,
+      "learning_rate": 3.750646252506476e-06,
+      "loss": 0.6127,
+      "step": 2438
+    },
+    {
+      "epoch": 1.0278216882706293,
+      "grad_norm": 0.2999134361743927,
+      "learning_rate": 3.749676831971733e-06,
+      "loss": 0.5914,
+      "step": 2439
+    },
+    {
+      "epoch": 1.0282432290020023,
+      "grad_norm": 0.2842731475830078,
+      "learning_rate": 3.7487071608732075e-06,
+      "loss": 0.597,
+      "step": 2440
+    },
+    {
+      "epoch": 1.0286647697333755,
+      "grad_norm": 0.28657662868499756,
+      "learning_rate": 3.747737239405321e-06,
+      "loss": 0.5948,
+      "step": 2441
+    },
+    {
+      "epoch": 1.0290863104647487,
+      "grad_norm": 0.27576345205307007,
+      "learning_rate": 3.746767067762547e-06,
+      "loss": 0.6117,
+      "step": 2442
+    },
+    {
+      "epoch": 1.0295078511961218,
+      "grad_norm": 0.2808803617954254,
+      "learning_rate": 3.745796646139407e-06,
+      "loss": 0.5883,
+      "step": 2443
+    },
+    {
+      "epoch": 1.029929391927495,
+      "grad_norm": 0.2850990891456604,
+      "learning_rate": 3.744825974730473e-06,
+      "loss": 0.602,
+      "step": 2444
+    },
+    {
+      "epoch": 1.0303509326588682,
+      "grad_norm": 0.2878445088863373,
+      "learning_rate": 3.7438550537303675e-06,
+      "loss": 0.5901,
+      "step": 2445
+    },
+    {
+      "epoch": 1.0307724733902413,
+      "grad_norm": 0.28286924958229065,
+      "learning_rate": 3.742883883333764e-06,
+      "loss": 0.5775,
+      "step": 2446
+    },
+    {
+      "epoch": 1.0311940141216145,
+      "grad_norm": 0.2687399387359619,
+      "learning_rate": 3.741912463735383e-06,
+      "loss": 0.5791,
+      "step": 2447
+    },
+    {
+      "epoch": 1.0316155548529877,
+      "grad_norm": 0.27169176936149597,
+      "learning_rate": 3.7409407951299987e-06,
+      "loss": 0.5972,
+      "step": 2448
+    },
+    {
+      "epoch": 1.0320370955843607,
+      "grad_norm": 0.27999189496040344,
+      "learning_rate": 3.739968877712432e-06,
+      "loss": 0.5969,
+      "step": 2449
+    },
+    {
+      "epoch": 1.032458636315734,
+      "grad_norm": 0.2751108407974243,
+      "learning_rate": 3.738996711677556e-06,
+      "loss": 0.6011,
+      "step": 2450
+    },
+    {
+      "epoch": 1.0328801770471072,
+      "grad_norm": 0.2807179391384125,
+      "learning_rate": 3.738024297220293e-06,
+      "loss": 0.5744,
+      "step": 2451
+    },
+    {
+      "epoch": 1.0333017177784805,
+      "grad_norm": 0.2862537205219269,
+      "learning_rate": 3.7370516345356144e-06,
+      "loss": 0.5806,
+      "step": 2452
+    },
+    {
+      "epoch": 1.0337232585098535,
+      "grad_norm": 0.28018903732299805,
+      "learning_rate": 3.7360787238185414e-06,
+      "loss": 0.5773,
+      "step": 2453
+    },
+    {
+      "epoch": 1.0341447992412267,
+      "grad_norm": 0.27865031361579895,
+      "learning_rate": 3.735105565264146e-06,
+      "loss": 0.6021,
+      "step": 2454
+    },
+    {
+      "epoch": 1.0345663399726,
+      "grad_norm": 0.28188958764076233,
+      "learning_rate": 3.7341321590675494e-06,
+      "loss": 0.6239,
+      "step": 2455
+    },
+    {
+      "epoch": 1.034987880703973,
+      "grad_norm": 0.2811955511569977,
+      "learning_rate": 3.733158505423922e-06,
+      "loss": 0.6247,
+      "step": 2456
+    },
+    {
+      "epoch": 1.0354094214353462,
+      "grad_norm": 0.27360332012176514,
+      "learning_rate": 3.7321846045284847e-06,
+      "loss": 0.5901,
+      "step": 2457
+    },
+    {
+      "epoch": 1.0358309621667194,
+      "grad_norm": 0.26987841725349426,
+      "learning_rate": 3.731210456576507e-06,
+      "loss": 0.5904,
+      "step": 2458
+    },
+    {
+      "epoch": 1.0362525028980925,
+      "grad_norm": 0.29671669006347656,
+      "learning_rate": 3.7302360617633077e-06,
+      "loss": 0.5967,
+      "step": 2459
+    },
+    {
+      "epoch": 1.0366740436294657,
+      "grad_norm": 0.2803356647491455,
+      "learning_rate": 3.729261420284257e-06,
+      "loss": 0.583,
+      "step": 2460
+    },
+    {
+      "epoch": 1.037095584360839,
+      "grad_norm": 0.28673017024993896,
+      "learning_rate": 3.7282865323347718e-06,
+      "loss": 0.5726,
+      "step": 2461
+    },
+    {
+      "epoch": 1.037517125092212,
+      "grad_norm": 0.26936209201812744,
+      "learning_rate": 3.7273113981103213e-06,
+      "loss": 0.5995,
+      "step": 2462
+    },
+    {
+      "epoch": 1.0379386658235852,
+      "grad_norm": 0.2799616754055023,
+      "learning_rate": 3.7263360178064223e-06,
+      "loss": 0.5968,
+      "step": 2463
+    },
+    {
+      "epoch": 1.0383602065549584,
+      "grad_norm": 0.28504323959350586,
+      "learning_rate": 3.725360391618641e-06,
+      "loss": 0.5987,
+      "step": 2464
+    },
+    {
+      "epoch": 1.0387817472863314,
+      "grad_norm": 0.27934786677360535,
+      "learning_rate": 3.724384519742594e-06,
+      "loss": 0.6091,
+      "step": 2465
+    },
+    {
+      "epoch": 1.0392032880177047,
+      "grad_norm": 0.28981906175613403,
+      "learning_rate": 3.723408402373946e-06,
+      "loss": 0.5924,
+      "step": 2466
+    },
+    {
+      "epoch": 1.039624828749078,
+      "grad_norm": 0.2790740728378296,
+      "learning_rate": 3.72243203970841e-06,
+      "loss": 0.6075,
+      "step": 2467
+    },
+    {
+      "epoch": 1.0400463694804511,
+      "grad_norm": 0.2756091058254242,
+      "learning_rate": 3.7214554319417516e-06,
+      "loss": 0.5827,
+      "step": 2468
+    },
+    {
+      "epoch": 1.0404679102118242,
+      "grad_norm": 0.2847738564014435,
+      "learning_rate": 3.7204785792697826e-06,
+      "loss": 0.5995,
+      "step": 2469
+    },
+    {
+      "epoch": 1.0408894509431974,
+      "grad_norm": 0.29248878359794617,
+      "learning_rate": 3.719501481888364e-06,
+      "loss": 0.5887,
+      "step": 2470
+    },
+    {
+      "epoch": 1.0413109916745706,
+      "grad_norm": 0.2976021468639374,
+      "learning_rate": 3.7185241399934083e-06,
+      "loss": 0.6019,
+      "step": 2471
+    },
+    {
+      "epoch": 1.0417325324059437,
+      "grad_norm": 0.27759262919425964,
+      "learning_rate": 3.7175465537808744e-06,
+      "loss": 0.6095,
+      "step": 2472
+    },
+    {
+      "epoch": 1.042154073137317,
+      "grad_norm": 0.29298537969589233,
+      "learning_rate": 3.7165687234467708e-06,
+      "loss": 0.5689,
+      "step": 2473
+    },
+    {
+      "epoch": 1.0425756138686901,
+      "grad_norm": 0.2899170517921448,
+      "learning_rate": 3.7155906491871556e-06,
+      "loss": 0.6187,
+      "step": 2474
+    },
+    {
+      "epoch": 1.0429971546000631,
+      "grad_norm": 0.3062553405761719,
+      "learning_rate": 3.7146123311981357e-06,
+      "loss": 0.5885,
+      "step": 2475
+    },
+    {
+      "epoch": 1.0434186953314364,
+      "grad_norm": 0.2824786603450775,
+      "learning_rate": 3.7136337696758665e-06,
+      "loss": 0.5743,
+      "step": 2476
+    },
+    {
+      "epoch": 1.0438402360628096,
+      "grad_norm": 0.28722918033599854,
+      "learning_rate": 3.712654964816552e-06,
+      "loss": 0.6115,
+      "step": 2477
+    },
+    {
+      "epoch": 1.0442617767941826,
+      "grad_norm": 0.28087639808654785,
+      "learning_rate": 3.7116759168164464e-06,
+      "loss": 0.6185,
+      "step": 2478
+    },
+    {
+      "epoch": 1.0446833175255559,
+      "grad_norm": 0.27586498856544495,
+      "learning_rate": 3.71069662587185e-06,
+      "loss": 0.6109,
+      "step": 2479
+    },
+    {
+      "epoch": 1.0451048582569291,
+      "grad_norm": 0.278489351272583,
+      "learning_rate": 3.7097170921791153e-06,
+      "loss": 0.6041,
+      "step": 2480
+    },
+    {
+      "epoch": 1.0455263989883021,
+      "grad_norm": 0.29185420274734497,
+      "learning_rate": 3.70873731593464e-06,
+      "loss": 0.5871,
+      "step": 2481
+    },
+    {
+      "epoch": 1.0459479397196754,
+      "grad_norm": 0.27844512462615967,
+      "learning_rate": 3.7077572973348735e-06,
+      "loss": 0.5989,
+      "step": 2482
+    },
+    {
+      "epoch": 1.0463694804510486,
+      "grad_norm": 0.2835767865180969,
+      "learning_rate": 3.706777036576311e-06,
+      "loss": 0.6089,
+      "step": 2483
+    },
+    {
+      "epoch": 1.0467910211824218,
+      "grad_norm": 0.2888692021369934,
+      "learning_rate": 3.7057965338554984e-06,
+      "loss": 0.5876,
+      "step": 2484
+    },
+    {
+      "epoch": 1.0472125619137949,
+      "grad_norm": 0.2834422290325165,
+      "learning_rate": 3.7048157893690295e-06,
+      "loss": 0.6105,
+      "step": 2485
+    },
+    {
+      "epoch": 1.047634102645168,
+      "grad_norm": 0.27649667859077454,
+      "learning_rate": 3.703834803313545e-06,
+      "loss": 0.587,
+      "step": 2486
+    },
+    {
+      "epoch": 1.0480556433765413,
+      "grad_norm": 0.2771513760089874,
+      "learning_rate": 3.7028535758857377e-06,
+      "loss": 0.5796,
+      "step": 2487
+    },
+    {
+      "epoch": 1.0484771841079144,
+      "grad_norm": 0.2841707766056061,
+      "learning_rate": 3.701872107282344e-06,
+      "loss": 0.6292,
+      "step": 2488
+    },
+    {
+      "epoch": 1.0488987248392876,
+      "grad_norm": 0.2871394753456116,
+      "learning_rate": 3.7008903977001526e-06,
+      "loss": 0.5903,
+      "step": 2489
+    },
+    {
+      "epoch": 1.0493202655706608,
+      "grad_norm": 0.289627343416214,
+      "learning_rate": 3.6999084473359987e-06,
+      "loss": 0.5921,
+      "step": 2490
+    },
+    {
+      "epoch": 1.0497418063020338,
+      "grad_norm": 0.28536897897720337,
+      "learning_rate": 3.698926256386767e-06,
+      "loss": 0.5936,
+      "step": 2491
+    },
+    {
+      "epoch": 1.050163347033407,
+      "grad_norm": 0.2884984612464905,
+      "learning_rate": 3.697943825049388e-06,
+      "loss": 0.5916,
+      "step": 2492
+    },
+    {
+      "epoch": 1.0505848877647803,
+      "grad_norm": 0.2809816300868988,
+      "learning_rate": 3.6969611535208426e-06,
+      "loss": 0.5977,
+      "step": 2493
+    },
+    {
+      "epoch": 1.0510064284961533,
+      "grad_norm": 0.28425243496894836,
+      "learning_rate": 3.695978241998161e-06,
+      "loss": 0.595,
+      "step": 2494
+    },
+    {
+      "epoch": 1.0514279692275266,
+      "grad_norm": 0.28718262910842896,
+      "learning_rate": 3.694995090678417e-06,
+      "loss": 0.5852,
+      "step": 2495
+    },
+    {
+      "epoch": 1.0518495099588998,
+      "grad_norm": 0.27504515647888184,
+      "learning_rate": 3.6940116997587377e-06,
+      "loss": 0.5854,
+      "step": 2496
+    },
+    {
+      "epoch": 1.052271050690273,
+      "grad_norm": 0.2825590968132019,
+      "learning_rate": 3.6930280694362937e-06,
+      "loss": 0.6161,
+      "step": 2497
+    },
+    {
+      "epoch": 1.052692591421646,
+      "grad_norm": 0.27865704894065857,
+      "learning_rate": 3.6920441999083068e-06,
+      "loss": 0.5984,
+      "step": 2498
+    },
+    {
+      "epoch": 1.0531141321530193,
+      "grad_norm": 0.28494739532470703,
+      "learning_rate": 3.6910600913720464e-06,
+      "loss": 0.5814,
+      "step": 2499
+    },
+    {
+      "epoch": 1.0535356728843925,
+      "grad_norm": 0.2932209372520447,
+      "learning_rate": 3.6900757440248263e-06,
+      "loss": 0.5983,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0539572136157656,
+      "grad_norm": 0.29247331619262695,
+      "learning_rate": 3.6890911580640144e-06,
+      "loss": 0.5795,
+      "step": 2501
+    },
+    {
+      "epoch": 1.0543787543471388,
+      "grad_norm": 0.2868572473526001,
+      "learning_rate": 3.6881063336870203e-06,
+      "loss": 0.5657,
+      "step": 2502
+    },
+    {
+      "epoch": 1.054800295078512,
+      "grad_norm": 0.2859145402908325,
+      "learning_rate": 3.6871212710913068e-06,
+      "loss": 0.5832,
+      "step": 2503
+    },
+    {
+      "epoch": 1.055221835809885,
+      "grad_norm": 0.2849128842353821,
+      "learning_rate": 3.6861359704743795e-06,
+      "loss": 0.5964,
+      "step": 2504
+    },
+    {
+      "epoch": 1.0556433765412583,
+      "grad_norm": 0.2932277023792267,
+      "learning_rate": 3.6851504320337946e-06,
+      "loss": 0.5864,
+      "step": 2505
+    },
+    {
+      "epoch": 1.0560649172726315,
+      "grad_norm": 0.2887207567691803,
+      "learning_rate": 3.684164655967155e-06,
+      "loss": 0.5932,
+      "step": 2506
+    },
+    {
+      "epoch": 1.0564864580040045,
+      "grad_norm": 0.27733615040779114,
+      "learning_rate": 3.683178642472113e-06,
+      "loss": 0.6032,
+      "step": 2507
+    },
+    {
+      "epoch": 1.0569079987353778,
+      "grad_norm": 0.2703331410884857,
+      "learning_rate": 3.6821923917463653e-06,
+      "loss": 0.6053,
+      "step": 2508
+    },
+    {
+      "epoch": 1.057329539466751,
+      "grad_norm": 0.27762606739997864,
+      "learning_rate": 3.68120590398766e-06,
+      "loss": 0.5727,
+      "step": 2509
+    },
+    {
+      "epoch": 1.0577510801981242,
+      "grad_norm": 0.2731919288635254,
+      "learning_rate": 3.6802191793937887e-06,
+      "loss": 0.5733,
+      "step": 2510
+    },
+    {
+      "epoch": 1.0581726209294973,
+      "grad_norm": 0.2809663414955139,
+      "learning_rate": 3.679232218162594e-06,
+      "loss": 0.5821,
+      "step": 2511
+    },
+    {
+      "epoch": 1.0585941616608705,
+      "grad_norm": 0.2788139879703522,
+      "learning_rate": 3.678245020491964e-06,
+      "loss": 0.5811,
+      "step": 2512
+    },
+    {
+      "epoch": 1.0590157023922437,
+      "grad_norm": 0.2923920154571533,
+      "learning_rate": 3.6772575865798343e-06,
+      "loss": 0.6108,
+      "step": 2513
+    },
+    {
+      "epoch": 1.0594372431236168,
+      "grad_norm": 0.29640305042266846,
+      "learning_rate": 3.6762699166241883e-06,
+      "loss": 0.6023,
+      "step": 2514
+    },
+    {
+      "epoch": 1.05985878385499,
+      "grad_norm": 0.27571600675582886,
+      "learning_rate": 3.6752820108230576e-06,
+      "loss": 0.6093,
+      "step": 2515
+    },
+    {
+      "epoch": 1.0602803245863632,
+      "grad_norm": 0.29457738995552063,
+      "learning_rate": 3.6742938693745184e-06,
+      "loss": 0.6067,
+      "step": 2516
+    },
+    {
+      "epoch": 1.0607018653177362,
+      "grad_norm": 0.290279358625412,
+      "learning_rate": 3.6733054924766975e-06,
+      "loss": 0.6003,
+      "step": 2517
+    },
+    {
+      "epoch": 1.0611234060491095,
+      "grad_norm": 0.2917850613594055,
+      "learning_rate": 3.6723168803277664e-06,
+      "loss": 0.5966,
+      "step": 2518
+    },
+    {
+      "epoch": 1.0615449467804827,
+      "grad_norm": 0.3032371997833252,
+      "learning_rate": 3.671328033125946e-06,
+      "loss": 0.5701,
+      "step": 2519
+    },
+    {
+      "epoch": 1.0619664875118557,
+      "grad_norm": 0.30758729577064514,
+      "learning_rate": 3.6703389510695005e-06,
+      "loss": 0.5861,
+      "step": 2520
+    },
+    {
+      "epoch": 1.062388028243229,
+      "grad_norm": 0.28572311997413635,
+      "learning_rate": 3.6693496343567457e-06,
+      "loss": 0.5913,
+      "step": 2521
+    },
+    {
+      "epoch": 1.0628095689746022,
+      "grad_norm": 0.347613662481308,
+      "learning_rate": 3.6683600831860415e-06,
+      "loss": 0.5788,
+      "step": 2522
+    },
+    {
+      "epoch": 1.0632311097059755,
+      "grad_norm": 0.2872063219547272,
+      "learning_rate": 3.667370297755797e-06,
+      "loss": 0.5754,
+      "step": 2523
+    },
+    {
+      "epoch": 1.0636526504373485,
+      "grad_norm": 0.2859274446964264,
+      "learning_rate": 3.6663802782644653e-06,
+      "loss": 0.5996,
+      "step": 2524
+    },
+    {
+      "epoch": 1.0640741911687217,
+      "grad_norm": 0.3069130480289459,
+      "learning_rate": 3.6653900249105493e-06,
+      "loss": 0.5698,
+      "step": 2525
+    },
+    {
+      "epoch": 1.064495731900095,
+      "grad_norm": 0.2940823435783386,
+      "learning_rate": 3.6643995378925977e-06,
+      "loss": 0.5825,
+      "step": 2526
+    },
+    {
+      "epoch": 1.064917272631468,
+      "grad_norm": 0.31003430485725403,
+      "learning_rate": 3.6634088174092046e-06,
+      "loss": 0.6035,
+      "step": 2527
+    },
+    {
+      "epoch": 1.0653388133628412,
+      "grad_norm": 0.2796047329902649,
+      "learning_rate": 3.6624178636590136e-06,
+      "loss": 0.5954,
+      "step": 2528
+    },
+    {
+      "epoch": 1.0657603540942144,
+      "grad_norm": 0.27728259563446045,
+      "learning_rate": 3.6614266768407137e-06,
+      "loss": 0.5954,
+      "step": 2529
+    },
+    {
+      "epoch": 1.0661818948255875,
+      "grad_norm": 0.28616467118263245,
+      "learning_rate": 3.6604352571530407e-06,
+      "loss": 0.5867,
+      "step": 2530
+    },
+    {
+      "epoch": 1.0666034355569607,
+      "grad_norm": 0.2975550591945648,
+      "learning_rate": 3.6594436047947764e-06,
+      "loss": 0.62,
+      "step": 2531
+    },
+    {
+      "epoch": 1.067024976288334,
+      "grad_norm": 0.2735946476459503,
+      "learning_rate": 3.6584517199647503e-06,
+      "loss": 0.5977,
+      "step": 2532
+    },
+    {
+      "epoch": 1.067446517019707,
+      "grad_norm": 0.2838672995567322,
+      "learning_rate": 3.657459602861838e-06,
+      "loss": 0.6098,
+      "step": 2533
+    },
+    {
+      "epoch": 1.0678680577510802,
+      "grad_norm": 0.28769245743751526,
+      "learning_rate": 3.6564672536849623e-06,
+      "loss": 0.6034,
+      "step": 2534
+    },
+    {
+      "epoch": 1.0682895984824534,
+      "grad_norm": 0.2891589105129242,
+      "learning_rate": 3.655474672633092e-06,
+      "loss": 0.5966,
+      "step": 2535
+    },
+    {
+      "epoch": 1.0687111392138264,
+      "grad_norm": 0.2943994998931885,
+      "learning_rate": 3.6544818599052418e-06,
+      "loss": 0.586,
+      "step": 2536
+    },
+    {
+      "epoch": 1.0691326799451997,
+      "grad_norm": 0.27871251106262207,
+      "learning_rate": 3.653488815700474e-06,
+      "loss": 0.5723,
+      "step": 2537
+    },
+    {
+      "epoch": 1.069554220676573,
+      "grad_norm": 0.3064369261264801,
+      "learning_rate": 3.6524955402178968e-06,
+      "loss": 0.586,
+      "step": 2538
+    },
+    {
+      "epoch": 1.0699757614079461,
+      "grad_norm": 0.27515941858291626,
+      "learning_rate": 3.651502033656664e-06,
+      "loss": 0.6111,
+      "step": 2539
+    },
+    {
+      "epoch": 1.0703973021393192,
+      "grad_norm": 0.28543412685394287,
+      "learning_rate": 3.650508296215978e-06,
+      "loss": 0.6167,
+      "step": 2540
+    },
+    {
+      "epoch": 1.0708188428706924,
+      "grad_norm": 0.28744739294052124,
+      "learning_rate": 3.6495143280950852e-06,
+      "loss": 0.5887,
+      "step": 2541
+    },
+    {
+      "epoch": 1.0712403836020656,
+      "grad_norm": 0.28765735030174255,
+      "learning_rate": 3.6485201294932793e-06,
+      "loss": 0.5885,
+      "step": 2542
+    },
+    {
+      "epoch": 1.0716619243334387,
+      "grad_norm": 0.27654334902763367,
+      "learning_rate": 3.6475257006098987e-06,
+      "loss": 0.5868,
+      "step": 2543
+    },
+    {
+      "epoch": 1.072083465064812,
+      "grad_norm": 0.2989504635334015,
+      "learning_rate": 3.646531041644331e-06,
+      "loss": 0.5693,
+      "step": 2544
+    },
+    {
+      "epoch": 1.0725050057961851,
+      "grad_norm": 0.270542711019516,
+      "learning_rate": 3.6455361527960076e-06,
+      "loss": 0.5846,
+      "step": 2545
+    },
+    {
+      "epoch": 1.0729265465275581,
+      "grad_norm": 0.27963951230049133,
+      "learning_rate": 3.6445410342644056e-06,
+      "loss": 0.5973,
+      "step": 2546
+    },
+    {
+      "epoch": 1.0733480872589314,
+      "grad_norm": 0.2802349925041199,
+      "learning_rate": 3.6435456862490504e-06,
+      "loss": 0.5776,
+      "step": 2547
+    },
+    {
+      "epoch": 1.0737696279903046,
+      "grad_norm": 0.2750988006591797,
+      "learning_rate": 3.642550108949511e-06,
+      "loss": 0.5832,
+      "step": 2548
+    },
+    {
+      "epoch": 1.0741911687216776,
+      "grad_norm": 0.28265538811683655,
+      "learning_rate": 3.6415543025654047e-06,
+      "loss": 0.5988,
+      "step": 2549
+    },
+    {
+      "epoch": 1.0746127094530509,
+      "grad_norm": 0.27463632822036743,
+      "learning_rate": 3.6405582672963928e-06,
+      "loss": 0.566,
+      "step": 2550
+    },
+    {
+      "epoch": 1.0750342501844241,
+      "grad_norm": 0.28450167179107666,
+      "learning_rate": 3.639562003342184e-06,
+      "loss": 0.5851,
+      "step": 2551
+    },
+    {
+      "epoch": 1.0754557909157971,
+      "grad_norm": 0.2708694338798523,
+      "learning_rate": 3.6385655109025308e-06,
+      "loss": 0.6107,
+      "step": 2552
+    },
+    {
+      "epoch": 1.0758773316471704,
+      "grad_norm": 0.2803850769996643,
+      "learning_rate": 3.6375687901772334e-06,
+      "loss": 0.5822,
+      "step": 2553
+    },
+    {
+      "epoch": 1.0762988723785436,
+      "grad_norm": 0.28604578971862793,
+      "learning_rate": 3.636571841366138e-06,
+      "loss": 0.5787,
+      "step": 2554
+    },
+    {
+      "epoch": 1.0767204131099168,
+      "grad_norm": 0.29270198941230774,
+      "learning_rate": 3.6355746646691335e-06,
+      "loss": 0.6028,
+      "step": 2555
+    },
+    {
+      "epoch": 1.0771419538412899,
+      "grad_norm": 0.30059996247291565,
+      "learning_rate": 3.634577260286159e-06,
+      "loss": 0.5871,
+      "step": 2556
+    },
+    {
+      "epoch": 1.077563494572663,
+      "grad_norm": 0.2833390533924103,
+      "learning_rate": 3.6335796284171955e-06,
+      "loss": 0.6128,
+      "step": 2557
+    },
+    {
+      "epoch": 1.0779850353040363,
+      "grad_norm": 0.27905479073524475,
+      "learning_rate": 3.6325817692622723e-06,
+      "loss": 0.5994,
+      "step": 2558
+    },
+    {
+      "epoch": 1.0784065760354093,
+      "grad_norm": 0.30388033390045166,
+      "learning_rate": 3.631583683021461e-06,
+      "loss": 0.5861,
+      "step": 2559
+    },
+    {
+      "epoch": 1.0788281167667826,
+      "grad_norm": 0.3080260455608368,
+      "learning_rate": 3.6305853698948833e-06,
+      "loss": 0.6088,
+      "step": 2560
+    },
+    {
+      "epoch": 1.0792496574981558,
+      "grad_norm": 0.2924356758594513,
+      "learning_rate": 3.629586830082702e-06,
+      "loss": 0.5999,
+      "step": 2561
+    },
+    {
+      "epoch": 1.0796711982295288,
+      "grad_norm": 0.32406899333000183,
+      "learning_rate": 3.6285880637851274e-06,
+      "loss": 0.6043,
+      "step": 2562
+    },
+    {
+      "epoch": 1.080092738960902,
+      "grad_norm": 0.2764522433280945,
+      "learning_rate": 3.627589071202416e-06,
+      "loss": 0.5728,
+      "step": 2563
+    },
+    {
+      "epoch": 1.0805142796922753,
+      "grad_norm": 0.30847322940826416,
+      "learning_rate": 3.6265898525348676e-06,
+      "loss": 0.598,
+      "step": 2564
+    },
+    {
+      "epoch": 1.0809358204236483,
+      "grad_norm": 0.2982868552207947,
+      "learning_rate": 3.625590407982829e-06,
+      "loss": 0.5963,
+      "step": 2565
+    },
+    {
+      "epoch": 1.0813573611550216,
+      "grad_norm": 0.2983497083187103,
+      "learning_rate": 3.6245907377466915e-06,
+      "loss": 0.5995,
+      "step": 2566
+    },
+    {
+      "epoch": 1.0817789018863948,
+      "grad_norm": 0.2949068546295166,
+      "learning_rate": 3.6235908420268915e-06,
+      "loss": 0.6041,
+      "step": 2567
+    },
+    {
+      "epoch": 1.082200442617768,
+      "grad_norm": 0.2899754047393799,
+      "learning_rate": 3.622590721023912e-06,
+      "loss": 0.5981,
+      "step": 2568
+    },
+    {
+      "epoch": 1.082621983349141,
+      "grad_norm": 0.27796003222465515,
+      "learning_rate": 3.621590374938279e-06,
+      "loss": 0.5836,
+      "step": 2569
+    },
+    {
+      "epoch": 1.0830435240805143,
+      "grad_norm": 0.29543960094451904,
+      "learning_rate": 3.620589803970565e-06,
+      "loss": 0.5923,
+      "step": 2570
+    },
+    {
+      "epoch": 1.0834650648118875,
+      "grad_norm": 0.2981010675430298,
+      "learning_rate": 3.6195890083213876e-06,
+      "loss": 0.6175,
+      "step": 2571
+    },
+    {
+      "epoch": 1.0838866055432606,
+      "grad_norm": 0.28342947363853455,
+      "learning_rate": 3.61858798819141e-06,
+      "loss": 0.5956,
+      "step": 2572
+    },
+    {
+      "epoch": 1.0843081462746338,
+      "grad_norm": 0.2851843237876892,
+      "learning_rate": 3.6175867437813373e-06,
+      "loss": 0.6077,
+      "step": 2573
+    },
+    {
+      "epoch": 1.084729687006007,
+      "grad_norm": 0.28936031460762024,
+      "learning_rate": 3.6165852752919247e-06,
+      "loss": 0.6038,
+      "step": 2574
+    },
+    {
+      "epoch": 1.08515122773738,
+      "grad_norm": 0.29018160700798035,
+      "learning_rate": 3.6155835829239674e-06,
+      "loss": 0.5981,
+      "step": 2575
+    },
+    {
+      "epoch": 1.0855727684687533,
+      "grad_norm": 0.28621312975883484,
+      "learning_rate": 3.614581666878308e-06,
+      "loss": 0.6221,
+      "step": 2576
+    },
+    {
+      "epoch": 1.0859943092001265,
+      "grad_norm": 0.2818050980567932,
+      "learning_rate": 3.6135795273558343e-06,
+      "loss": 0.6106,
+      "step": 2577
+    },
+    {
+      "epoch": 1.0864158499314995,
+      "grad_norm": 0.288422167301178,
+      "learning_rate": 3.612577164557477e-06,
+      "loss": 0.5681,
+      "step": 2578
+    },
+    {
+      "epoch": 1.0868373906628728,
+      "grad_norm": 0.2895830571651459,
+      "learning_rate": 3.611574578684214e-06,
+      "loss": 0.6007,
+      "step": 2579
+    },
+    {
+      "epoch": 1.087258931394246,
+      "grad_norm": 0.2817233204841614,
+      "learning_rate": 3.6105717699370655e-06,
+      "loss": 0.5823,
+      "step": 2580
+    },
+    {
+      "epoch": 1.0876804721256192,
+      "grad_norm": 0.2812691330909729,
+      "learning_rate": 3.6095687385170984e-06,
+      "loss": 0.6182,
+      "step": 2581
+    },
+    {
+      "epoch": 1.0881020128569923,
+      "grad_norm": 0.2807496190071106,
+      "learning_rate": 3.608565484625423e-06,
+      "loss": 0.5805,
+      "step": 2582
+    },
+    {
+      "epoch": 1.0885235535883655,
+      "grad_norm": 0.28207871317863464,
+      "learning_rate": 3.607562008463195e-06,
+      "loss": 0.6116,
+      "step": 2583
+    },
+    {
+      "epoch": 1.0889450943197387,
+      "grad_norm": 0.29034751653671265,
+      "learning_rate": 3.606558310231613e-06,
+      "loss": 0.5724,
+      "step": 2584
+    },
+    {
+      "epoch": 1.0893666350511118,
+      "grad_norm": 0.29230085015296936,
+      "learning_rate": 3.605554390131922e-06,
+      "loss": 0.5955,
+      "step": 2585
+    },
+    {
+      "epoch": 1.089788175782485,
+      "grad_norm": 0.2773236930370331,
+      "learning_rate": 3.6045502483654127e-06,
+      "loss": 0.5983,
+      "step": 2586
+    },
+    {
+      "epoch": 1.0902097165138582,
+      "grad_norm": 0.29371052980422974,
+      "learning_rate": 3.6035458851334155e-06,
+      "loss": 0.5948,
+      "step": 2587
+    },
+    {
+      "epoch": 1.0906312572452312,
+      "grad_norm": 0.2887687087059021,
+      "learning_rate": 3.60254130063731e-06,
+      "loss": 0.5942,
+      "step": 2588
+    },
+    {
+      "epoch": 1.0910527979766045,
+      "grad_norm": 0.277793824672699,
+      "learning_rate": 3.601536495078516e-06,
+      "loss": 0.5779,
+      "step": 2589
+    },
+    {
+      "epoch": 1.0914743387079777,
+      "grad_norm": 0.2853635549545288,
+      "learning_rate": 3.600531468658503e-06,
+      "loss": 0.5774,
+      "step": 2590
+    },
+    {
+      "epoch": 1.0918958794393507,
+      "grad_norm": 0.3024192154407501,
+      "learning_rate": 3.5995262215787795e-06,
+      "loss": 0.5741,
+      "step": 2591
+    },
+    {
+      "epoch": 1.092317420170724,
+      "grad_norm": 0.28188440203666687,
+      "learning_rate": 3.5985207540409016e-06,
+      "loss": 0.6028,
+      "step": 2592
+    },
+    {
+      "epoch": 1.0927389609020972,
+      "grad_norm": 0.29133689403533936,
+      "learning_rate": 3.597515066246468e-06,
+      "loss": 0.5998,
+      "step": 2593
+    },
+    {
+      "epoch": 1.0931605016334704,
+      "grad_norm": 0.2798634469509125,
+      "learning_rate": 3.5965091583971214e-06,
+      "loss": 0.5766,
+      "step": 2594
+    },
+    {
+      "epoch": 1.0935820423648435,
+      "grad_norm": 0.2877105474472046,
+      "learning_rate": 3.5955030306945504e-06,
+      "loss": 0.5867,
+      "step": 2595
+    },
+    {
+      "epoch": 1.0940035830962167,
+      "grad_norm": 0.28240665793418884,
+      "learning_rate": 3.594496683340485e-06,
+      "loss": 0.5861,
+      "step": 2596
+    },
+    {
+      "epoch": 1.09442512382759,
+      "grad_norm": 0.29148098826408386,
+      "learning_rate": 3.593490116536702e-06,
+      "loss": 0.589,
+      "step": 2597
+    },
+    {
+      "epoch": 1.094846664558963,
+      "grad_norm": 0.2914937734603882,
+      "learning_rate": 3.5924833304850193e-06,
+      "loss": 0.6099,
+      "step": 2598
+    },
+    {
+      "epoch": 1.0952682052903362,
+      "grad_norm": 0.33105266094207764,
+      "learning_rate": 3.591476325387303e-06,
+      "loss": 0.5802,
+      "step": 2599
+    },
+    {
+      "epoch": 1.0956897460217094,
+      "grad_norm": 0.27701982855796814,
+      "learning_rate": 3.5904691014454587e-06,
+      "loss": 0.6027,
+      "step": 2600
+    },
+    {
+      "epoch": 1.0961112867530824,
+      "grad_norm": 0.307112455368042,
+      "learning_rate": 3.5894616588614373e-06,
+      "loss": 0.5955,
+      "step": 2601
+    },
+    {
+      "epoch": 1.0965328274844557,
+      "grad_norm": 0.2855505645275116,
+      "learning_rate": 3.5884539978372357e-06,
+      "loss": 0.5993,
+      "step": 2602
+    },
+    {
+      "epoch": 1.096954368215829,
+      "grad_norm": 0.2865561544895172,
+      "learning_rate": 3.587446118574891e-06,
+      "loss": 0.6045,
+      "step": 2603
+    },
+    {
+      "epoch": 1.097375908947202,
+      "grad_norm": 0.2936447560787201,
+      "learning_rate": 3.586438021276487e-06,
+      "loss": 0.6013,
+      "step": 2604
+    },
+    {
+      "epoch": 1.0977974496785752,
+      "grad_norm": 0.29071640968322754,
+      "learning_rate": 3.585429706144149e-06,
+      "loss": 0.578,
+      "step": 2605
+    },
+    {
+      "epoch": 1.0982189904099484,
+      "grad_norm": 0.3015688359737396,
+      "learning_rate": 3.5844211733800484e-06,
+      "loss": 0.5927,
+      "step": 2606
+    },
+    {
+      "epoch": 1.0986405311413214,
+      "grad_norm": 0.28990429639816284,
+      "learning_rate": 3.5834124231863986e-06,
+      "loss": 0.6145,
+      "step": 2607
+    },
+    {
+      "epoch": 1.0990620718726947,
+      "grad_norm": 0.2906133830547333,
+      "learning_rate": 3.5824034557654558e-06,
+      "loss": 0.5997,
+      "step": 2608
+    },
+    {
+      "epoch": 1.099483612604068,
+      "grad_norm": 0.32171374559402466,
+      "learning_rate": 3.5813942713195222e-06,
+      "loss": 0.6139,
+      "step": 2609
+    },
+    {
+      "epoch": 1.099905153335441,
+      "grad_norm": 0.2818268835544586,
+      "learning_rate": 3.5803848700509414e-06,
+      "loss": 0.5922,
+      "step": 2610
+    },
+    {
+      "epoch": 1.1003266940668142,
+      "grad_norm": 0.2853582203388214,
+      "learning_rate": 3.5793752521621018e-06,
+      "loss": 0.6138,
+      "step": 2611
+    },
+    {
+      "epoch": 1.1007482347981874,
+      "grad_norm": 0.29389527440071106,
+      "learning_rate": 3.578365417855434e-06,
+      "loss": 0.5919,
+      "step": 2612
+    },
+    {
+      "epoch": 1.1011697755295606,
+      "grad_norm": 0.3015442490577698,
+      "learning_rate": 3.577355367333413e-06,
+      "loss": 0.5869,
+      "step": 2613
+    },
+    {
+      "epoch": 1.1015913162609337,
+      "grad_norm": 0.2924076020717621,
+      "learning_rate": 3.5763451007985573e-06,
+      "loss": 0.6093,
+      "step": 2614
+    },
+    {
+      "epoch": 1.1020128569923069,
+      "grad_norm": 0.27183932065963745,
+      "learning_rate": 3.5753346184534272e-06,
+      "loss": 0.5847,
+      "step": 2615
+    },
+    {
+      "epoch": 1.1024343977236801,
+      "grad_norm": 0.3014562427997589,
+      "learning_rate": 3.5743239205006284e-06,
+      "loss": 0.5981,
+      "step": 2616
+    },
+    {
+      "epoch": 1.1028559384550531,
+      "grad_norm": 0.2923741340637207,
+      "learning_rate": 3.573313007142808e-06,
+      "loss": 0.6068,
+      "step": 2617
+    },
+    {
+      "epoch": 1.1032774791864264,
+      "grad_norm": 0.2920997440814972,
+      "learning_rate": 3.572301878582657e-06,
+      "loss": 0.5978,
+      "step": 2618
+    },
+    {
+      "epoch": 1.1036990199177996,
+      "grad_norm": 0.27717190980911255,
+      "learning_rate": 3.5712905350229094e-06,
+      "loss": 0.5867,
+      "step": 2619
+    },
+    {
+      "epoch": 1.1041205606491726,
+      "grad_norm": 0.29597896337509155,
+      "learning_rate": 3.5702789766663438e-06,
+      "loss": 0.5725,
+      "step": 2620
+    },
+    {
+      "epoch": 1.1045421013805459,
+      "grad_norm": 0.28755244612693787,
+      "learning_rate": 3.5692672037157793e-06,
+      "loss": 0.5965,
+      "step": 2621
+    },
+    {
+      "epoch": 1.104963642111919,
+      "grad_norm": 0.2905902862548828,
+      "learning_rate": 3.5682552163740798e-06,
+      "loss": 0.5674,
+      "step": 2622
+    },
+    {
+      "epoch": 1.1053851828432921,
+      "grad_norm": 0.2794736921787262,
+      "learning_rate": 3.5672430148441504e-06,
+      "loss": 0.5817,
+      "step": 2623
+    },
+    {
+      "epoch": 1.1058067235746654,
+      "grad_norm": 0.29186660051345825,
+      "learning_rate": 3.5662305993289427e-06,
+      "loss": 0.5917,
+      "step": 2624
+    },
+    {
+      "epoch": 1.1062282643060386,
+      "grad_norm": 0.2905406057834625,
+      "learning_rate": 3.5652179700314477e-06,
+      "loss": 0.5874,
+      "step": 2625
+    },
+    {
+      "epoch": 1.1066498050374118,
+      "grad_norm": 0.2877916991710663,
+      "learning_rate": 3.5642051271546997e-06,
+      "loss": 0.5809,
+      "step": 2626
+    },
+    {
+      "epoch": 1.1070713457687849,
+      "grad_norm": 0.2977445721626282,
+      "learning_rate": 3.5631920709017774e-06,
+      "loss": 0.5873,
+      "step": 2627
+    },
+    {
+      "epoch": 1.107492886500158,
+      "grad_norm": 0.33070480823516846,
+      "learning_rate": 3.5621788014758013e-06,
+      "loss": 0.5829,
+      "step": 2628
+    },
+    {
+      "epoch": 1.1079144272315313,
+      "grad_norm": 0.28070664405822754,
+      "learning_rate": 3.5611653190799357e-06,
+      "loss": 0.5761,
+      "step": 2629
+    },
+    {
+      "epoch": 1.1083359679629043,
+      "grad_norm": 0.28899332880973816,
+      "learning_rate": 3.560151623917385e-06,
+      "loss": 0.6162,
+      "step": 2630
+    },
+    {
+      "epoch": 1.1087575086942776,
+      "grad_norm": 0.35850659012794495,
+      "learning_rate": 3.5591377161913998e-06,
+      "loss": 0.5702,
+      "step": 2631
+    },
+    {
+      "epoch": 1.1091790494256508,
+      "grad_norm": 0.2796879708766937,
+      "learning_rate": 3.558123596105271e-06,
+      "loss": 0.5916,
+      "step": 2632
+    },
+    {
+      "epoch": 1.1096005901570238,
+      "grad_norm": 0.29282253980636597,
+      "learning_rate": 3.5571092638623304e-06,
+      "loss": 0.6057,
+      "step": 2633
+    },
+    {
+      "epoch": 1.110022130888397,
+      "grad_norm": 0.3051425516605377,
+      "learning_rate": 3.5560947196659578e-06,
+      "loss": 0.5931,
+      "step": 2634
+    },
+    {
+      "epoch": 1.1104436716197703,
+      "grad_norm": 0.30543747544288635,
+      "learning_rate": 3.5550799637195705e-06,
+      "loss": 0.6093,
+      "step": 2635
+    },
+    {
+      "epoch": 1.1108652123511433,
+      "grad_norm": 0.27268680930137634,
+      "learning_rate": 3.5540649962266305e-06,
+      "loss": 0.5965,
+      "step": 2636
+    },
+    {
+      "epoch": 1.1112867530825166,
+      "grad_norm": 0.3191748261451721,
+      "learning_rate": 3.553049817390641e-06,
+      "loss": 0.601,
+      "step": 2637
+    },
+    {
+      "epoch": 1.1117082938138898,
+      "grad_norm": 0.2780769467353821,
+      "learning_rate": 3.5520344274151487e-06,
+      "loss": 0.5836,
+      "step": 2638
+    },
+    {
+      "epoch": 1.112129834545263,
+      "grad_norm": 0.2842741310596466,
+      "learning_rate": 3.5510188265037414e-06,
+      "loss": 0.5977,
+      "step": 2639
+    },
+    {
+      "epoch": 1.112551375276636,
+      "grad_norm": 0.29763513803482056,
+      "learning_rate": 3.550003014860052e-06,
+      "loss": 0.6002,
+      "step": 2640
+    },
+    {
+      "epoch": 1.1129729160080093,
+      "grad_norm": 0.29912155866622925,
+      "learning_rate": 3.548986992687752e-06,
+      "loss": 0.6091,
+      "step": 2641
+    },
+    {
+      "epoch": 1.1133944567393825,
+      "grad_norm": 0.3004581928253174,
+      "learning_rate": 3.5479707601905574e-06,
+      "loss": 0.594,
+      "step": 2642
+    },
+    {
+      "epoch": 1.1138159974707555,
+      "grad_norm": 0.2813160717487335,
+      "learning_rate": 3.5469543175722253e-06,
+      "loss": 0.579,
+      "step": 2643
+    },
+    {
+      "epoch": 1.1142375382021288,
+      "grad_norm": 0.29291537404060364,
+      "learning_rate": 3.545937665036555e-06,
+      "loss": 0.6157,
+      "step": 2644
+    },
+    {
+      "epoch": 1.114659078933502,
+      "grad_norm": 0.288615345954895,
+      "learning_rate": 3.5449208027873896e-06,
+      "loss": 0.5994,
+      "step": 2645
+    },
+    {
+      "epoch": 1.115080619664875,
+      "grad_norm": 0.2769889831542969,
+      "learning_rate": 3.5439037310286117e-06,
+      "loss": 0.5974,
+      "step": 2646
+    },
+    {
+      "epoch": 1.1155021603962483,
+      "grad_norm": 0.2861781120300293,
+      "learning_rate": 3.5428864499641474e-06,
+      "loss": 0.5941,
+      "step": 2647
+    },
+    {
+      "epoch": 1.1159237011276215,
+      "grad_norm": 0.293282151222229,
+      "learning_rate": 3.5418689597979644e-06,
+      "loss": 0.6094,
+      "step": 2648
+    },
+    {
+      "epoch": 1.1163452418589945,
+      "grad_norm": 0.2921757102012634,
+      "learning_rate": 3.5408512607340727e-06,
+      "loss": 0.6027,
+      "step": 2649
+    },
+    {
+      "epoch": 1.1167667825903678,
+      "grad_norm": 0.2875699996948242,
+      "learning_rate": 3.5398333529765238e-06,
+      "loss": 0.6066,
+      "step": 2650
+    },
+    {
+      "epoch": 1.117188323321741,
+      "grad_norm": 0.28018566966056824,
+      "learning_rate": 3.5388152367294115e-06,
+      "loss": 0.6024,
+      "step": 2651
+    },
+    {
+      "epoch": 1.1176098640531142,
+      "grad_norm": 0.29060301184654236,
+      "learning_rate": 3.5377969121968704e-06,
+      "loss": 0.5827,
+      "step": 2652
+    },
+    {
+      "epoch": 1.1180314047844873,
+      "grad_norm": 0.2893514931201935,
+      "learning_rate": 3.536778379583077e-06,
+      "loss": 0.5541,
+      "step": 2653
+    },
+    {
+      "epoch": 1.1184529455158605,
+      "grad_norm": 0.2917473316192627,
+      "learning_rate": 3.5357596390922507e-06,
+      "loss": 0.6129,
+      "step": 2654
+    },
+    {
+      "epoch": 1.1188744862472337,
+      "grad_norm": 0.3088398575782776,
+      "learning_rate": 3.534740690928652e-06,
+      "loss": 0.5899,
+      "step": 2655
+    },
+    {
+      "epoch": 1.1192960269786068,
+      "grad_norm": 0.2942919135093689,
+      "learning_rate": 3.5337215352965827e-06,
+      "loss": 0.6137,
+      "step": 2656
+    },
+    {
+      "epoch": 1.11971756770998,
+      "grad_norm": 0.28957223892211914,
+      "learning_rate": 3.532702172400387e-06,
+      "loss": 0.5863,
+      "step": 2657
+    },
+    {
+      "epoch": 1.1201391084413532,
+      "grad_norm": 0.29068779945373535,
+      "learning_rate": 3.5316826024444486e-06,
+      "loss": 0.5897,
+      "step": 2658
+    },
+    {
+      "epoch": 1.1205606491727262,
+      "grad_norm": 0.2823987901210785,
+      "learning_rate": 3.530662825633196e-06,
+      "loss": 0.6062,
+      "step": 2659
+    },
+    {
+      "epoch": 1.1209821899040995,
+      "grad_norm": 0.29268357157707214,
+      "learning_rate": 3.529642842171096e-06,
+      "loss": 0.5835,
+      "step": 2660
+    },
+    {
+      "epoch": 1.1214037306354727,
+      "grad_norm": 0.30355116724967957,
+      "learning_rate": 3.528622652262659e-06,
+      "loss": 0.5985,
+      "step": 2661
+    },
+    {
+      "epoch": 1.1218252713668457,
+      "grad_norm": 0.28067952394485474,
+      "learning_rate": 3.5276022561124356e-06,
+      "loss": 0.5843,
+      "step": 2662
+    },
+    {
+      "epoch": 1.122246812098219,
+      "grad_norm": 0.27409207820892334,
+      "learning_rate": 3.526581653925018e-06,
+      "loss": 0.6017,
+      "step": 2663
+    },
+    {
+      "epoch": 1.1226683528295922,
+      "grad_norm": 0.3388080298900604,
+      "learning_rate": 3.5255608459050406e-06,
+      "loss": 0.5632,
+      "step": 2664
+    },
+    {
+      "epoch": 1.1230898935609652,
+      "grad_norm": 0.2872592508792877,
+      "learning_rate": 3.5245398322571774e-06,
+      "loss": 0.587,
+      "step": 2665
+    },
+    {
+      "epoch": 1.1235114342923385,
+      "grad_norm": 0.2952466905117035,
+      "learning_rate": 3.5235186131861455e-06,
+      "loss": 0.5968,
+      "step": 2666
+    },
+    {
+      "epoch": 1.1239329750237117,
+      "grad_norm": 0.29374104738235474,
+      "learning_rate": 3.522497188896701e-06,
+      "loss": 0.5737,
+      "step": 2667
+    },
+    {
+      "epoch": 1.124354515755085,
+      "grad_norm": 0.29709920287132263,
+      "learning_rate": 3.521475559593644e-06,
+      "loss": 0.6199,
+      "step": 2668
+    },
+    {
+      "epoch": 1.124776056486458,
+      "grad_norm": 0.2843749225139618,
+      "learning_rate": 3.520453725481813e-06,
+      "loss": 0.578,
+      "step": 2669
+    },
+    {
+      "epoch": 1.1251975972178312,
+      "grad_norm": 0.29528793692588806,
+      "learning_rate": 3.5194316867660885e-06,
+      "loss": 0.5903,
+      "step": 2670
+    },
+    {
+      "epoch": 1.1256191379492044,
+      "grad_norm": 0.30987870693206787,
+      "learning_rate": 3.5184094436513926e-06,
+      "loss": 0.5942,
+      "step": 2671
+    },
+    {
+      "epoch": 1.1260406786805774,
+      "grad_norm": 0.2854704260826111,
+      "learning_rate": 3.517386996342688e-06,
+      "loss": 0.5909,
+      "step": 2672
+    },
+    {
+      "epoch": 1.1264622194119507,
+      "grad_norm": 0.3042103350162506,
+      "learning_rate": 3.516364345044979e-06,
+      "loss": 0.6107,
+      "step": 2673
+    },
+    {
+      "epoch": 1.126883760143324,
+      "grad_norm": 0.2783954441547394,
+      "learning_rate": 3.5153414899633086e-06,
+      "loss": 0.5967,
+      "step": 2674
+    },
+    {
+      "epoch": 1.127305300874697,
+      "grad_norm": 0.2763589918613434,
+      "learning_rate": 3.514318431302764e-06,
+      "loss": 0.6079,
+      "step": 2675
+    },
+    {
+      "epoch": 1.1277268416060702,
+      "grad_norm": 0.2961251735687256,
+      "learning_rate": 3.5132951692684693e-06,
+      "loss": 0.6025,
+      "step": 2676
+    },
+    {
+      "epoch": 1.1281483823374434,
+      "grad_norm": 0.2741136848926544,
+      "learning_rate": 3.512271704065593e-06,
+      "loss": 0.5762,
+      "step": 2677
+    },
+    {
+      "epoch": 1.1285699230688164,
+      "grad_norm": 0.28271257877349854,
+      "learning_rate": 3.5112480358993417e-06,
+      "loss": 0.6021,
+      "step": 2678
+    },
+    {
+      "epoch": 1.1289914638001897,
+      "grad_norm": 0.2953920364379883,
+      "learning_rate": 3.5102241649749653e-06,
+      "loss": 0.5931,
+      "step": 2679
+    },
+    {
+      "epoch": 1.129413004531563,
+      "grad_norm": 0.2885507047176361,
+      "learning_rate": 3.509200091497752e-06,
+      "loss": 0.5656,
+      "step": 2680
+    },
+    {
+      "epoch": 1.129834545262936,
+      "grad_norm": 0.2877041697502136,
+      "learning_rate": 3.5081758156730305e-06,
+      "loss": 0.5829,
+      "step": 2681
+    },
+    {
+      "epoch": 1.1302560859943092,
+      "grad_norm": 0.2856444716453552,
+      "learning_rate": 3.5071513377061727e-06,
+      "loss": 0.5919,
+      "step": 2682
+    },
+    {
+      "epoch": 1.1306776267256824,
+      "grad_norm": 0.290411114692688,
+      "learning_rate": 3.506126657802589e-06,
+      "loss": 0.591,
+      "step": 2683
+    },
+    {
+      "epoch": 1.1310991674570556,
+      "grad_norm": 0.3116011917591095,
+      "learning_rate": 3.5051017761677292e-06,
+      "loss": 0.5838,
+      "step": 2684
+    },
+    {
+      "epoch": 1.1315207081884286,
+      "grad_norm": 0.284015953540802,
+      "learning_rate": 3.504076693007087e-06,
+      "loss": 0.593,
+      "step": 2685
+    },
+    {
+      "epoch": 1.1319422489198019,
+      "grad_norm": 0.3101438581943512,
+      "learning_rate": 3.5030514085261937e-06,
+      "loss": 0.6034,
+      "step": 2686
+    },
+    {
+      "epoch": 1.1323637896511751,
+      "grad_norm": 0.30127018690109253,
+      "learning_rate": 3.502025922930621e-06,
+      "loss": 0.6059,
+      "step": 2687
+    },
+    {
+      "epoch": 1.1327853303825481,
+      "grad_norm": 0.29326939582824707,
+      "learning_rate": 3.5010002364259822e-06,
+      "loss": 0.5842,
+      "step": 2688
+    },
+    {
+      "epoch": 1.1332068711139214,
+      "grad_norm": 0.30117979645729065,
+      "learning_rate": 3.499974349217931e-06,
+      "loss": 0.5926,
+      "step": 2689
+    },
+    {
+      "epoch": 1.1336284118452946,
+      "grad_norm": 0.29290971159935,
+      "learning_rate": 3.4989482615121607e-06,
+      "loss": 0.6003,
+      "step": 2690
+    },
+    {
+      "epoch": 1.1340499525766676,
+      "grad_norm": 0.27996838092803955,
+      "learning_rate": 3.4979219735144044e-06,
+      "loss": 0.6204,
+      "step": 2691
+    },
+    {
+      "epoch": 1.1344714933080409,
+      "grad_norm": 0.2856767475605011,
+      "learning_rate": 3.496895485430435e-06,
+      "loss": 0.5879,
+      "step": 2692
+    },
+    {
+      "epoch": 1.134893034039414,
+      "grad_norm": 0.2992691695690155,
+      "learning_rate": 3.4958687974660672e-06,
+      "loss": 0.5947,
+      "step": 2693
+    },
+    {
+      "epoch": 1.1353145747707871,
+      "grad_norm": 0.30726170539855957,
+      "learning_rate": 3.494841909827155e-06,
+      "loss": 0.6162,
+      "step": 2694
+    },
+    {
+      "epoch": 1.1357361155021604,
+      "grad_norm": 0.2814769744873047,
+      "learning_rate": 3.4938148227195923e-06,
+      "loss": 0.5782,
+      "step": 2695
+    },
+    {
+      "epoch": 1.1361576562335336,
+      "grad_norm": 0.281276673078537,
+      "learning_rate": 3.4927875363493125e-06,
+      "loss": 0.5924,
+      "step": 2696
+    },
+    {
+      "epoch": 1.1365791969649068,
+      "grad_norm": 0.28659602999687195,
+      "learning_rate": 3.491760050922289e-06,
+      "loss": 0.5931,
+      "step": 2697
+    },
+    {
+      "epoch": 1.1370007376962799,
+      "grad_norm": 0.2893781363964081,
+      "learning_rate": 3.490732366644538e-06,
+      "loss": 0.6063,
+      "step": 2698
+    },
+    {
+      "epoch": 1.137422278427653,
+      "grad_norm": 0.3095279037952423,
+      "learning_rate": 3.48970448372211e-06,
+      "loss": 0.5688,
+      "step": 2699
+    },
+    {
+      "epoch": 1.1378438191590263,
+      "grad_norm": 0.28217244148254395,
+      "learning_rate": 3.4886764023611006e-06,
+      "loss": 0.607,
+      "step": 2700
+    },
+    {
+      "epoch": 1.1382653598903993,
+      "grad_norm": 0.2839772403240204,
+      "learning_rate": 3.4876481227676424e-06,
+      "loss": 0.6036,
+      "step": 2701
+    },
+    {
+      "epoch": 1.1386869006217726,
+      "grad_norm": 0.3189666271209717,
+      "learning_rate": 3.486619645147908e-06,
+      "loss": 0.6001,
+      "step": 2702
+    },
+    {
+      "epoch": 1.1391084413531458,
+      "grad_norm": 0.3095225691795349,
+      "learning_rate": 3.485590969708111e-06,
+      "loss": 0.5794,
+      "step": 2703
+    },
+    {
+      "epoch": 1.1395299820845188,
+      "grad_norm": 0.2787955403327942,
+      "learning_rate": 3.4845620966545026e-06,
+      "loss": 0.5891,
+      "step": 2704
+    },
+    {
+      "epoch": 1.139951522815892,
+      "grad_norm": 0.30805930495262146,
+      "learning_rate": 3.4835330261933763e-06,
+      "loss": 0.5612,
+      "step": 2705
+    },
+    {
+      "epoch": 1.1403730635472653,
+      "grad_norm": 0.292636901140213,
+      "learning_rate": 3.4825037585310617e-06,
+      "loss": 0.5984,
+      "step": 2706
+    },
+    {
+      "epoch": 1.1407946042786383,
+      "grad_norm": 0.29500332474708557,
+      "learning_rate": 3.4814742938739316e-06,
+      "loss": 0.6044,
+      "step": 2707
+    },
+    {
+      "epoch": 1.1412161450100116,
+      "grad_norm": 0.28307783603668213,
+      "learning_rate": 3.4804446324283957e-06,
+      "loss": 0.5958,
+      "step": 2708
+    },
+    {
+      "epoch": 1.1416376857413848,
+      "grad_norm": 0.2885133922100067,
+      "learning_rate": 3.479414774400904e-06,
+      "loss": 0.5875,
+      "step": 2709
+    },
+    {
+      "epoch": 1.142059226472758,
+      "grad_norm": 0.2950177490711212,
+      "learning_rate": 3.478384719997948e-06,
+      "loss": 0.6031,
+      "step": 2710
+    },
+    {
+      "epoch": 1.142480767204131,
+      "grad_norm": 0.2877272367477417,
+      "learning_rate": 3.477354469426054e-06,
+      "loss": 0.5899,
+      "step": 2711
+    },
+    {
+      "epoch": 1.1429023079355043,
+      "grad_norm": 0.2992627024650574,
+      "learning_rate": 3.4763240228917916e-06,
+      "loss": 0.5617,
+      "step": 2712
+    },
+    {
+      "epoch": 1.1433238486668775,
+      "grad_norm": 0.30091431736946106,
+      "learning_rate": 3.475293380601767e-06,
+      "loss": 0.6179,
+      "step": 2713
+    },
+    {
+      "epoch": 1.1437453893982505,
+      "grad_norm": 0.29273703694343567,
+      "learning_rate": 3.4742625427626284e-06,
+      "loss": 0.5827,
+      "step": 2714
+    },
+    {
+      "epoch": 1.1441669301296238,
+      "grad_norm": 0.27961406111717224,
+      "learning_rate": 3.473231509581061e-06,
+      "loss": 0.5824,
+      "step": 2715
+    },
+    {
+      "epoch": 1.144588470860997,
+      "grad_norm": 0.29855257272720337,
+      "learning_rate": 3.472200281263789e-06,
+      "loss": 0.6071,
+      "step": 2716
+    },
+    {
+      "epoch": 1.14501001159237,
+      "grad_norm": 0.29434284567832947,
+      "learning_rate": 3.4711688580175783e-06,
+      "loss": 0.6045,
+      "step": 2717
+    },
+    {
+      "epoch": 1.1454315523237433,
+      "grad_norm": 0.3144681453704834,
+      "learning_rate": 3.4701372400492317e-06,
+      "loss": 0.5731,
+      "step": 2718
+    },
+    {
+      "epoch": 1.1458530930551165,
+      "grad_norm": 0.2905290424823761,
+      "learning_rate": 3.4691054275655913e-06,
+      "loss": 0.5902,
+      "step": 2719
+    },
+    {
+      "epoch": 1.1462746337864895,
+      "grad_norm": 0.2854931950569153,
+      "learning_rate": 3.4680734207735385e-06,
+      "loss": 0.5753,
+      "step": 2720
+    },
+    {
+      "epoch": 1.1466961745178628,
+      "grad_norm": 0.293403685092926,
+      "learning_rate": 3.467041219879994e-06,
+      "loss": 0.5882,
+      "step": 2721
+    },
+    {
+      "epoch": 1.147117715249236,
+      "grad_norm": 0.28555771708488464,
+      "learning_rate": 3.466008825091916e-06,
+      "loss": 0.5877,
+      "step": 2722
+    },
+    {
+      "epoch": 1.1475392559806092,
+      "grad_norm": 0.3087461292743683,
+      "learning_rate": 3.4649762366163038e-06,
+      "loss": 0.5838,
+      "step": 2723
+    },
+    {
+      "epoch": 1.1479607967119823,
+      "grad_norm": 0.2841496467590332,
+      "learning_rate": 3.4639434546601934e-06,
+      "loss": 0.5853,
+      "step": 2724
+    },
+    {
+      "epoch": 1.1483823374433555,
+      "grad_norm": 0.29895615577697754,
+      "learning_rate": 3.462910479430661e-06,
+      "loss": 0.5978,
+      "step": 2725
+    },
+    {
+      "epoch": 1.1488038781747285,
+      "grad_norm": 0.31959399580955505,
+      "learning_rate": 3.461877311134822e-06,
+      "loss": 0.5953,
+      "step": 2726
+    },
+    {
+      "epoch": 1.1492254189061017,
+      "grad_norm": 0.2783656716346741,
+      "learning_rate": 3.4608439499798274e-06,
+      "loss": 0.6011,
+      "step": 2727
+    },
+    {
+      "epoch": 1.149646959637475,
+      "grad_norm": 0.30042019486427307,
+      "learning_rate": 3.4598103961728713e-06,
+      "loss": 0.5971,
+      "step": 2728
+    },
+    {
+      "epoch": 1.1500685003688482,
+      "grad_norm": 0.29413869976997375,
+      "learning_rate": 3.4587766499211828e-06,
+      "loss": 0.6236,
+      "step": 2729
+    },
+    {
+      "epoch": 1.1504900411002212,
+      "grad_norm": 0.29102590680122375,
+      "learning_rate": 3.4577427114320317e-06,
+      "loss": 0.62,
+      "step": 2730
+    },
+    {
+      "epoch": 1.1509115818315945,
+      "grad_norm": 0.2757788598537445,
+      "learning_rate": 3.4567085809127247e-06,
+      "loss": 0.5908,
+      "step": 2731
+    },
+    {
+      "epoch": 1.1513331225629677,
+      "grad_norm": 0.2919961214065552,
+      "learning_rate": 3.4556742585706086e-06,
+      "loss": 0.6108,
+      "step": 2732
+    },
+    {
+      "epoch": 1.1517546632943407,
+      "grad_norm": 0.2971349358558655,
+      "learning_rate": 3.454639744613069e-06,
+      "loss": 0.6015,
+      "step": 2733
+    },
+    {
+      "epoch": 1.152176204025714,
+      "grad_norm": 0.2976561188697815,
+      "learning_rate": 3.453605039247527e-06,
+      "loss": 0.5914,
+      "step": 2734
+    },
+    {
+      "epoch": 1.1525977447570872,
+      "grad_norm": 0.29397323727607727,
+      "learning_rate": 3.4525701426814457e-06,
+      "loss": 0.6039,
+      "step": 2735
+    },
+    {
+      "epoch": 1.1530192854884604,
+      "grad_norm": 0.3003097176551819,
+      "learning_rate": 3.4515350551223227e-06,
+      "loss": 0.5882,
+      "step": 2736
+    },
+    {
+      "epoch": 1.1534408262198335,
+      "grad_norm": 0.28930890560150146,
+      "learning_rate": 3.450499776777698e-06,
+      "loss": 0.5755,
+      "step": 2737
+    },
+    {
+      "epoch": 1.1538623669512067,
+      "grad_norm": 0.2893696129322052,
+      "learning_rate": 3.449464307855147e-06,
+      "loss": 0.6019,
+      "step": 2738
+    },
+    {
+      "epoch": 1.1542839076825797,
+      "grad_norm": 0.29216113686561584,
+      "learning_rate": 3.4484286485622855e-06,
+      "loss": 0.5817,
+      "step": 2739
+    },
+    {
+      "epoch": 1.154705448413953,
+      "grad_norm": 0.3068190813064575,
+      "learning_rate": 3.4473927991067635e-06,
+      "loss": 0.5805,
+      "step": 2740
+    },
+    {
+      "epoch": 1.1551269891453262,
+      "grad_norm": 0.3021355867385864,
+      "learning_rate": 3.446356759696274e-06,
+      "loss": 0.6092,
+      "step": 2741
+    },
+    {
+      "epoch": 1.1555485298766994,
+      "grad_norm": 0.2902212142944336,
+      "learning_rate": 3.4453205305385452e-06,
+      "loss": 0.5813,
+      "step": 2742
+    },
+    {
+      "epoch": 1.1559700706080724,
+      "grad_norm": 0.28853633999824524,
+      "learning_rate": 3.4442841118413433e-06,
+      "loss": 0.6247,
+      "step": 2743
+    },
+    {
+      "epoch": 1.1563916113394457,
+      "grad_norm": 0.30047425627708435,
+      "learning_rate": 3.443247503812475e-06,
+      "loss": 0.5866,
+      "step": 2744
+    },
+    {
+      "epoch": 1.156813152070819,
+      "grad_norm": 0.29848119616508484,
+      "learning_rate": 3.442210706659781e-06,
+      "loss": 0.5979,
+      "step": 2745
+    },
+    {
+      "epoch": 1.157234692802192,
+      "grad_norm": 0.29118335247039795,
+      "learning_rate": 3.441173720591143e-06,
+      "loss": 0.5742,
+      "step": 2746
+    },
+    {
+      "epoch": 1.1576562335335652,
+      "grad_norm": 0.3058696389198303,
+      "learning_rate": 3.440136545814481e-06,
+      "loss": 0.5739,
+      "step": 2747
+    },
+    {
+      "epoch": 1.1580777742649384,
+      "grad_norm": 0.2745908796787262,
+      "learning_rate": 3.4390991825377495e-06,
+      "loss": 0.6234,
+      "step": 2748
+    },
+    {
+      "epoch": 1.1584993149963114,
+      "grad_norm": 0.28864726424217224,
+      "learning_rate": 3.438061630968944e-06,
+      "loss": 0.6096,
+      "step": 2749
+    },
+    {
+      "epoch": 1.1589208557276847,
+      "grad_norm": 0.28922584652900696,
+      "learning_rate": 3.4370238913160964e-06,
+      "loss": 0.5847,
+      "step": 2750
+    },
+    {
+      "epoch": 1.159342396459058,
+      "grad_norm": 0.298467218875885,
+      "learning_rate": 3.4359859637872763e-06,
+      "loss": 0.5815,
+      "step": 2751
+    },
+    {
+      "epoch": 1.159763937190431,
+      "grad_norm": 0.2921035885810852,
+      "learning_rate": 3.4349478485905906e-06,
+      "loss": 0.6099,
+      "step": 2752
+    },
+    {
+      "epoch": 1.1601854779218042,
+      "grad_norm": 0.28973084688186646,
+      "learning_rate": 3.4339095459341855e-06,
+      "loss": 0.5909,
+      "step": 2753
+    },
+    {
+      "epoch": 1.1606070186531774,
+      "grad_norm": 0.2750161290168762,
+      "learning_rate": 3.432871056026243e-06,
+      "loss": 0.5816,
+      "step": 2754
+    },
+    {
+      "epoch": 1.1610285593845506,
+      "grad_norm": 0.2851737141609192,
+      "learning_rate": 3.4318323790749825e-06,
+      "loss": 0.5802,
+      "step": 2755
+    },
+    {
+      "epoch": 1.1614501001159236,
+      "grad_norm": 0.29903727769851685,
+      "learning_rate": 3.430793515288663e-06,
+      "loss": 0.5967,
+      "step": 2756
+    },
+    {
+      "epoch": 1.1618716408472969,
+      "grad_norm": 0.2803456783294678,
+      "learning_rate": 3.4297544648755794e-06,
+      "loss": 0.5896,
+      "step": 2757
+    },
+    {
+      "epoch": 1.1622931815786701,
+      "grad_norm": 0.28222203254699707,
+      "learning_rate": 3.4287152280440644e-06,
+      "loss": 0.592,
+      "step": 2758
+    },
+    {
+      "epoch": 1.1627147223100431,
+      "grad_norm": 0.3007655143737793,
+      "learning_rate": 3.4276758050024873e-06,
+      "loss": 0.5946,
+      "step": 2759
+    },
+    {
+      "epoch": 1.1631362630414164,
+      "grad_norm": 0.28357571363449097,
+      "learning_rate": 3.426636195959256e-06,
+      "loss": 0.5813,
+      "step": 2760
+    },
+    {
+      "epoch": 1.1635578037727896,
+      "grad_norm": 0.2872273325920105,
+      "learning_rate": 3.4255964011228145e-06,
+      "loss": 0.5837,
+      "step": 2761
+    },
+    {
+      "epoch": 1.1639793445041626,
+      "grad_norm": 0.28393200039863586,
+      "learning_rate": 3.4245564207016445e-06,
+      "loss": 0.5942,
+      "step": 2762
+    },
+    {
+      "epoch": 1.1644008852355359,
+      "grad_norm": 0.28443461656570435,
+      "learning_rate": 3.4235162549042668e-06,
+      "loss": 0.5801,
+      "step": 2763
+    },
+    {
+      "epoch": 1.164822425966909,
+      "grad_norm": 0.2915363311767578,
+      "learning_rate": 3.4224759039392352e-06,
+      "loss": 0.6284,
+      "step": 2764
+    },
+    {
+      "epoch": 1.1652439666982821,
+      "grad_norm": 0.28612372279167175,
+      "learning_rate": 3.4214353680151457e-06,
+      "loss": 0.5669,
+      "step": 2765
+    },
+    {
+      "epoch": 1.1656655074296554,
+      "grad_norm": 0.30041825771331787,
+      "learning_rate": 3.420394647340625e-06,
+      "loss": 0.5975,
+      "step": 2766
+    },
+    {
+      "epoch": 1.1660870481610286,
+      "grad_norm": 0.28432849049568176,
+      "learning_rate": 3.4193537421243444e-06,
+      "loss": 0.5911,
+      "step": 2767
+    },
+    {
+      "epoch": 1.1665085888924018,
+      "grad_norm": 0.2819357216358185,
+      "learning_rate": 3.4183126525750067e-06,
+      "loss": 0.6013,
+      "step": 2768
+    },
+    {
+      "epoch": 1.1669301296237748,
+      "grad_norm": 0.2863505482673645,
+      "learning_rate": 3.4172713789013536e-06,
+      "loss": 0.6,
+      "step": 2769
+    },
+    {
+      "epoch": 1.167351670355148,
+      "grad_norm": 0.27280110120773315,
+      "learning_rate": 3.4162299213121626e-06,
+      "loss": 0.5808,
+      "step": 2770
+    },
+    {
+      "epoch": 1.1677732110865213,
+      "grad_norm": 0.29524919390678406,
+      "learning_rate": 3.4151882800162497e-06,
+      "loss": 0.5946,
+      "step": 2771
+    },
+    {
+      "epoch": 1.1681947518178943,
+      "grad_norm": 0.2856743633747101,
+      "learning_rate": 3.414146455222467e-06,
+      "loss": 0.5858,
+      "step": 2772
+    },
+    {
+      "epoch": 1.1686162925492676,
+      "grad_norm": 0.29249611496925354,
+      "learning_rate": 3.413104447139703e-06,
+      "loss": 0.5868,
+      "step": 2773
+    },
+    {
+      "epoch": 1.1690378332806408,
+      "grad_norm": 0.28621965646743774,
+      "learning_rate": 3.412062255976884e-06,
+      "loss": 0.5742,
+      "step": 2774
+    },
+    {
+      "epoch": 1.1694593740120138,
+      "grad_norm": 0.28955206274986267,
+      "learning_rate": 3.4110198819429713e-06,
+      "loss": 0.6141,
+      "step": 2775
+    },
+    {
+      "epoch": 1.169880914743387,
+      "grad_norm": 0.2940289080142975,
+      "learning_rate": 3.409977325246965e-06,
+      "loss": 0.6041,
+      "step": 2776
+    },
+    {
+      "epoch": 1.1703024554747603,
+      "grad_norm": 0.284508615732193,
+      "learning_rate": 3.4089345860978996e-06,
+      "loss": 0.6006,
+      "step": 2777
+    },
+    {
+      "epoch": 1.1707239962061333,
+      "grad_norm": 0.2838660478591919,
+      "learning_rate": 3.4078916647048477e-06,
+      "loss": 0.6066,
+      "step": 2778
+    },
+    {
+      "epoch": 1.1711455369375066,
+      "grad_norm": 0.2932637631893158,
+      "learning_rate": 3.406848561276919e-06,
+      "loss": 0.6047,
+      "step": 2779
+    },
+    {
+      "epoch": 1.1715670776688798,
+      "grad_norm": 0.3022191524505615,
+      "learning_rate": 3.4058052760232563e-06,
+      "loss": 0.5758,
+      "step": 2780
+    },
+    {
+      "epoch": 1.171988618400253,
+      "grad_norm": 0.2855169475078583,
+      "learning_rate": 3.4047618091530443e-06,
+      "loss": 0.5857,
+      "step": 2781
+    },
+    {
+      "epoch": 1.172410159131626,
+      "grad_norm": 0.283949077129364,
+      "learning_rate": 3.4037181608754987e-06,
+      "loss": 0.5827,
+      "step": 2782
+    },
+    {
+      "epoch": 1.1728316998629993,
+      "grad_norm": 0.2878352105617523,
+      "learning_rate": 3.4026743313998756e-06,
+      "loss": 0.5833,
+      "step": 2783
+    },
+    {
+      "epoch": 1.1732532405943725,
+      "grad_norm": 0.28269946575164795,
+      "learning_rate": 3.4016303209354645e-06,
+      "loss": 0.5985,
+      "step": 2784
+    },
+    {
+      "epoch": 1.1736747813257455,
+      "grad_norm": 0.29440540075302124,
+      "learning_rate": 3.400586129691593e-06,
+      "loss": 0.5805,
+      "step": 2785
+    },
+    {
+      "epoch": 1.1740963220571188,
+      "grad_norm": 0.30033645033836365,
+      "learning_rate": 3.399541757877625e-06,
+      "loss": 0.603,
+      "step": 2786
+    },
+    {
+      "epoch": 1.174517862788492,
+      "grad_norm": 0.3029514253139496,
+      "learning_rate": 3.398497205702959e-06,
+      "loss": 0.5745,
+      "step": 2787
+    },
+    {
+      "epoch": 1.174939403519865,
+      "grad_norm": 0.2957550287246704,
+      "learning_rate": 3.397452473377032e-06,
+      "loss": 0.6001,
+      "step": 2788
+    },
+    {
+      "epoch": 1.1753609442512383,
+      "grad_norm": 0.29505452513694763,
+      "learning_rate": 3.3964075611093143e-06,
+      "loss": 0.6054,
+      "step": 2789
+    },
+    {
+      "epoch": 1.1757824849826115,
+      "grad_norm": 0.31108254194259644,
+      "learning_rate": 3.395362469109315e-06,
+      "loss": 0.5867,
+      "step": 2790
+    },
+    {
+      "epoch": 1.1762040257139845,
+      "grad_norm": 0.3059183955192566,
+      "learning_rate": 3.3943171975865774e-06,
+      "loss": 0.6052,
+      "step": 2791
+    },
+    {
+      "epoch": 1.1766255664453578,
+      "grad_norm": 0.28976449370384216,
+      "learning_rate": 3.393271746750682e-06,
+      "loss": 0.5943,
+      "step": 2792
+    },
+    {
+      "epoch": 1.177047107176731,
+      "grad_norm": 0.2717047929763794,
+      "learning_rate": 3.3922261168112443e-06,
+      "loss": 0.5843,
+      "step": 2793
+    },
+    {
+      "epoch": 1.1774686479081042,
+      "grad_norm": 0.2864796221256256,
+      "learning_rate": 3.3911803079779153e-06,
+      "loss": 0.6006,
+      "step": 2794
+    },
+    {
+      "epoch": 1.1778901886394773,
+      "grad_norm": 0.3030761778354645,
+      "learning_rate": 3.3901343204603844e-06,
+      "loss": 0.6039,
+      "step": 2795
+    },
+    {
+      "epoch": 1.1783117293708505,
+      "grad_norm": 0.2941778600215912,
+      "learning_rate": 3.389088154468373e-06,
+      "loss": 0.6128,
+      "step": 2796
+    },
+    {
+      "epoch": 1.1787332701022235,
+      "grad_norm": 0.29698446393013,
+      "learning_rate": 3.388041810211643e-06,
+      "loss": 0.5898,
+      "step": 2797
+    },
+    {
+      "epoch": 1.1791548108335967,
+      "grad_norm": 0.2953915297985077,
+      "learning_rate": 3.3869952878999868e-06,
+      "loss": 0.6047,
+      "step": 2798
+    },
+    {
+      "epoch": 1.17957635156497,
+      "grad_norm": 0.2961310148239136,
+      "learning_rate": 3.3859485877432373e-06,
+      "loss": 0.6241,
+      "step": 2799
+    },
+    {
+      "epoch": 1.1799978922963432,
+      "grad_norm": 0.2875201106071472,
+      "learning_rate": 3.3849017099512592e-06,
+      "loss": 0.5966,
+      "step": 2800
+    },
+    {
+      "epoch": 1.1804194330277162,
+      "grad_norm": 0.2866179347038269,
+      "learning_rate": 3.3838546547339553e-06,
+      "loss": 0.5822,
+      "step": 2801
+    },
+    {
+      "epoch": 1.1808409737590895,
+      "grad_norm": 0.2947884798049927,
+      "learning_rate": 3.382807422301263e-06,
+      "loss": 0.5782,
+      "step": 2802
+    },
+    {
+      "epoch": 1.1812625144904627,
+      "grad_norm": 0.2855282425880432,
+      "learning_rate": 3.3817600128631555e-06,
+      "loss": 0.6005,
+      "step": 2803
+    },
+    {
+      "epoch": 1.1816840552218357,
+      "grad_norm": 0.28616204857826233,
+      "learning_rate": 3.3807124266296416e-06,
+      "loss": 0.5738,
+      "step": 2804
+    },
+    {
+      "epoch": 1.182105595953209,
+      "grad_norm": 0.2901262044906616,
+      "learning_rate": 3.3796646638107645e-06,
+      "loss": 0.5911,
+      "step": 2805
+    },
+    {
+      "epoch": 1.1825271366845822,
+      "grad_norm": 0.3059486448764801,
+      "learning_rate": 3.3786167246166047e-06,
+      "loss": 0.5652,
+      "step": 2806
+    },
+    {
+      "epoch": 1.1829486774159554,
+      "grad_norm": 0.30247431993484497,
+      "learning_rate": 3.3775686092572757e-06,
+      "loss": 0.5813,
+      "step": 2807
+    },
+    {
+      "epoch": 1.1833702181473285,
+      "grad_norm": 0.30039161443710327,
+      "learning_rate": 3.3765203179429295e-06,
+      "loss": 0.5717,
+      "step": 2808
+    },
+    {
+      "epoch": 1.1837917588787017,
+      "grad_norm": 0.2914503812789917,
+      "learning_rate": 3.3754718508837496e-06,
+      "loss": 0.5776,
+      "step": 2809
+    },
+    {
+      "epoch": 1.1842132996100747,
+      "grad_norm": 0.3075217306613922,
+      "learning_rate": 3.3744232082899573e-06,
+      "loss": 0.5647,
+      "step": 2810
+    },
+    {
+      "epoch": 1.184634840341448,
+      "grad_norm": 0.2824708819389343,
+      "learning_rate": 3.3733743903718096e-06,
+      "loss": 0.6093,
+      "step": 2811
+    },
+    {
+      "epoch": 1.1850563810728212,
+      "grad_norm": 0.30193227529525757,
+      "learning_rate": 3.3723253973395963e-06,
+      "loss": 0.5928,
+      "step": 2812
+    },
+    {
+      "epoch": 1.1854779218041944,
+      "grad_norm": 0.2963663339614868,
+      "learning_rate": 3.3712762294036437e-06,
+      "loss": 0.5996,
+      "step": 2813
+    },
+    {
+      "epoch": 1.1858994625355674,
+      "grad_norm": 0.31942465901374817,
+      "learning_rate": 3.3702268867743125e-06,
+      "loss": 0.5962,
+      "step": 2814
+    },
+    {
+      "epoch": 1.1863210032669407,
+      "grad_norm": 0.2889573574066162,
+      "learning_rate": 3.3691773696619997e-06,
+      "loss": 0.605,
+      "step": 2815
+    },
+    {
+      "epoch": 1.186742543998314,
+      "grad_norm": 0.282882958650589,
+      "learning_rate": 3.368127678277136e-06,
+      "loss": 0.6167,
+      "step": 2816
+    },
+    {
+      "epoch": 1.187164084729687,
+      "grad_norm": 0.31821882724761963,
+      "learning_rate": 3.3670778128301883e-06,
+      "loss": 0.5841,
+      "step": 2817
+    },
+    {
+      "epoch": 1.1875856254610602,
+      "grad_norm": 0.32916757464408875,
+      "learning_rate": 3.3660277735316566e-06,
+      "loss": 0.5895,
+      "step": 2818
+    },
+    {
+      "epoch": 1.1880071661924334,
+      "grad_norm": 0.2977168560028076,
+      "learning_rate": 3.364977560592077e-06,
+      "loss": 0.5826,
+      "step": 2819
+    },
+    {
+      "epoch": 1.1884287069238064,
+      "grad_norm": 0.3196712136268616,
+      "learning_rate": 3.3639271742220215e-06,
+      "loss": 0.5788,
+      "step": 2820
+    },
+    {
+      "epoch": 1.1888502476551797,
+      "grad_norm": 0.3064783215522766,
+      "learning_rate": 3.3628766146320934e-06,
+      "loss": 0.5919,
+      "step": 2821
+    },
+    {
+      "epoch": 1.189271788386553,
+      "grad_norm": 0.2900356352329254,
+      "learning_rate": 3.3618258820329346e-06,
+      "loss": 0.5936,
+      "step": 2822
+    },
+    {
+      "epoch": 1.189693329117926,
+      "grad_norm": 0.3159439265727997,
+      "learning_rate": 3.360774976635219e-06,
+      "loss": 0.5916,
+      "step": 2823
+    },
+    {
+      "epoch": 1.1901148698492992,
+      "grad_norm": 0.3000507354736328,
+      "learning_rate": 3.3597238986496568e-06,
+      "loss": 0.5805,
+      "step": 2824
+    },
+    {
+      "epoch": 1.1905364105806724,
+      "grad_norm": 0.28818172216415405,
+      "learning_rate": 3.3586726482869913e-06,
+      "loss": 0.6061,
+      "step": 2825
+    },
+    {
+      "epoch": 1.1909579513120456,
+      "grad_norm": 0.29824429750442505,
+      "learning_rate": 3.3576212257580027e-06,
+      "loss": 0.5857,
+      "step": 2826
+    },
+    {
+      "epoch": 1.1913794920434186,
+      "grad_norm": 0.29827600717544556,
+      "learning_rate": 3.356569631273503e-06,
+      "loss": 0.568,
+      "step": 2827
+    },
+    {
+      "epoch": 1.1918010327747919,
+      "grad_norm": 0.32512733340263367,
+      "learning_rate": 3.3555178650443403e-06,
+      "loss": 0.5869,
+      "step": 2828
+    },
+    {
+      "epoch": 1.1922225735061651,
+      "grad_norm": 0.28230491280555725,
+      "learning_rate": 3.3544659272813966e-06,
+      "loss": 0.6087,
+      "step": 2829
+    },
+    {
+      "epoch": 1.1926441142375381,
+      "grad_norm": 0.29187309741973877,
+      "learning_rate": 3.3534138181955887e-06,
+      "loss": 0.6072,
+      "step": 2830
+    },
+    {
+      "epoch": 1.1930656549689114,
+      "grad_norm": 0.2979649007320404,
+      "learning_rate": 3.3523615379978675e-06,
+      "loss": 0.5867,
+      "step": 2831
+    },
+    {
+      "epoch": 1.1934871957002846,
+      "grad_norm": 0.30297034978866577,
+      "learning_rate": 3.351309086899217e-06,
+      "loss": 0.5965,
+      "step": 2832
+    },
+    {
+      "epoch": 1.1939087364316576,
+      "grad_norm": 0.3009093105792999,
+      "learning_rate": 3.3502564651106584e-06,
+      "loss": 0.5886,
+      "step": 2833
+    },
+    {
+      "epoch": 1.1943302771630309,
+      "grad_norm": 0.2944256365299225,
+      "learning_rate": 3.349203672843245e-06,
+      "loss": 0.6158,
+      "step": 2834
+    },
+    {
+      "epoch": 1.194751817894404,
+      "grad_norm": 0.32348740100860596,
+      "learning_rate": 3.348150710308062e-06,
+      "loss": 0.5916,
+      "step": 2835
+    },
+    {
+      "epoch": 1.1951733586257771,
+      "grad_norm": 0.3070305585861206,
+      "learning_rate": 3.3470975777162363e-06,
+      "loss": 0.5912,
+      "step": 2836
+    },
+    {
+      "epoch": 1.1955948993571504,
+      "grad_norm": 0.28429922461509705,
+      "learning_rate": 3.3460442752789203e-06,
+      "loss": 0.5727,
+      "step": 2837
+    },
+    {
+      "epoch": 1.1960164400885236,
+      "grad_norm": 0.3068450093269348,
+      "learning_rate": 3.3449908032073057e-06,
+      "loss": 0.6004,
+      "step": 2838
+    },
+    {
+      "epoch": 1.1964379808198968,
+      "grad_norm": 0.38045230507850647,
+      "learning_rate": 3.343937161712616e-06,
+      "loss": 0.555,
+      "step": 2839
+    },
+    {
+      "epoch": 1.1968595215512698,
+      "grad_norm": 0.2899267375469208,
+      "learning_rate": 3.3428833510061093e-06,
+      "loss": 0.6063,
+      "step": 2840
+    },
+    {
+      "epoch": 1.197281062282643,
+      "grad_norm": 0.2976974844932556,
+      "learning_rate": 3.3418293712990783e-06,
+      "loss": 0.5476,
+      "step": 2841
+    },
+    {
+      "epoch": 1.1977026030140163,
+      "grad_norm": 0.34704163670539856,
+      "learning_rate": 3.3407752228028483e-06,
+      "loss": 0.6109,
+      "step": 2842
+    },
+    {
+      "epoch": 1.1981241437453893,
+      "grad_norm": 0.2866734564304352,
+      "learning_rate": 3.33972090572878e-06,
+      "loss": 0.59,
+      "step": 2843
+    },
+    {
+      "epoch": 1.1985456844767626,
+      "grad_norm": 0.2868588864803314,
+      "learning_rate": 3.3386664202882653e-06,
+      "loss": 0.5885,
+      "step": 2844
+    },
+    {
+      "epoch": 1.1989672252081358,
+      "grad_norm": 0.30863067507743835,
+      "learning_rate": 3.3376117666927325e-06,
+      "loss": 0.6057,
+      "step": 2845
+    },
+    {
+      "epoch": 1.1993887659395088,
+      "grad_norm": 0.299625962972641,
+      "learning_rate": 3.3365569451536432e-06,
+      "loss": 0.5838,
+      "step": 2846
+    },
+    {
+      "epoch": 1.199810306670882,
+      "grad_norm": 0.3054928779602051,
+      "learning_rate": 3.3355019558824924e-06,
+      "loss": 0.6128,
+      "step": 2847
+    },
+    {
+      "epoch": 1.2002318474022553,
+      "grad_norm": 0.2984982132911682,
+      "learning_rate": 3.334446799090807e-06,
+      "loss": 0.6057,
+      "step": 2848
+    },
+    {
+      "epoch": 1.2006533881336283,
+      "grad_norm": 0.3390073776245117,
+      "learning_rate": 3.3333914749901497e-06,
+      "loss": 0.5913,
+      "step": 2849
+    },
+    {
+      "epoch": 1.2010749288650016,
+      "grad_norm": 0.30187833309173584,
+      "learning_rate": 3.3323359837921164e-06,
+      "loss": 0.5941,
+      "step": 2850
+    },
+    {
+      "epoch": 1.2014964695963748,
+      "grad_norm": 0.304463654756546,
+      "learning_rate": 3.3312803257083354e-06,
+      "loss": 0.5685,
+      "step": 2851
+    },
+    {
+      "epoch": 1.201918010327748,
+      "grad_norm": 0.28608307242393494,
+      "learning_rate": 3.3302245009504695e-06,
+      "loss": 0.6203,
+      "step": 2852
+    },
+    {
+      "epoch": 1.202339551059121,
+      "grad_norm": 0.2900875210762024,
+      "learning_rate": 3.3291685097302147e-06,
+      "loss": 0.5655,
+      "step": 2853
+    },
+    {
+      "epoch": 1.2027610917904943,
+      "grad_norm": 0.2904196083545685,
+      "learning_rate": 3.3281123522593e-06,
+      "loss": 0.597,
+      "step": 2854
+    },
+    {
+      "epoch": 1.2031826325218673,
+      "grad_norm": 0.3009113669395447,
+      "learning_rate": 3.327056028749488e-06,
+      "loss": 0.6152,
+      "step": 2855
+    },
+    {
+      "epoch": 1.2036041732532405,
+      "grad_norm": 0.2790769040584564,
+      "learning_rate": 3.3259995394125755e-06,
+      "loss": 0.5797,
+      "step": 2856
+    },
+    {
+      "epoch": 1.2040257139846138,
+      "grad_norm": 0.2796967029571533,
+      "learning_rate": 3.3249428844603898e-06,
+      "loss": 0.6061,
+      "step": 2857
+    },
+    {
+      "epoch": 1.204447254715987,
+      "grad_norm": 0.2795465588569641,
+      "learning_rate": 3.323886064104795e-06,
+      "loss": 0.5826,
+      "step": 2858
+    },
+    {
+      "epoch": 1.20486879544736,
+      "grad_norm": 0.29257622361183167,
+      "learning_rate": 3.3228290785576856e-06,
+      "loss": 0.5953,
+      "step": 2859
+    },
+    {
+      "epoch": 1.2052903361787333,
+      "grad_norm": 0.2874898612499237,
+      "learning_rate": 3.3217719280309907e-06,
+      "loss": 0.594,
+      "step": 2860
+    },
+    {
+      "epoch": 1.2057118769101065,
+      "grad_norm": 0.30102795362472534,
+      "learning_rate": 3.3207146127366716e-06,
+      "loss": 0.5765,
+      "step": 2861
+    },
+    {
+      "epoch": 1.2061334176414795,
+      "grad_norm": 0.28700411319732666,
+      "learning_rate": 3.3196571328867237e-06,
+      "loss": 0.5919,
+      "step": 2862
+    },
+    {
+      "epoch": 1.2065549583728528,
+      "grad_norm": 0.2873523533344269,
+      "learning_rate": 3.3185994886931738e-06,
+      "loss": 0.5804,
+      "step": 2863
+    },
+    {
+      "epoch": 1.206976499104226,
+      "grad_norm": 0.29356878995895386,
+      "learning_rate": 3.3175416803680837e-06,
+      "loss": 0.5809,
+      "step": 2864
+    },
+    {
+      "epoch": 1.2073980398355992,
+      "grad_norm": 0.29523390531539917,
+      "learning_rate": 3.316483708123546e-06,
+      "loss": 0.5738,
+      "step": 2865
+    },
+    {
+      "epoch": 1.2078195805669723,
+      "grad_norm": 0.302283376455307,
+      "learning_rate": 3.3154255721716877e-06,
+      "loss": 0.6011,
+      "step": 2866
+    },
+    {
+      "epoch": 1.2082411212983455,
+      "grad_norm": 0.27887189388275146,
+      "learning_rate": 3.314367272724668e-06,
+      "loss": 0.5643,
+      "step": 2867
+    },
+    {
+      "epoch": 1.2086626620297185,
+      "grad_norm": 0.28625643253326416,
+      "learning_rate": 3.313308809994679e-06,
+      "loss": 0.5788,
+      "step": 2868
+    },
+    {
+      "epoch": 1.2090842027610917,
+      "grad_norm": 0.2839239239692688,
+      "learning_rate": 3.3122501841939453e-06,
+      "loss": 0.6006,
+      "step": 2869
+    },
+    {
+      "epoch": 1.209505743492465,
+      "grad_norm": 0.287297785282135,
+      "learning_rate": 3.311191395534725e-06,
+      "loss": 0.5793,
+      "step": 2870
+    },
+    {
+      "epoch": 1.2099272842238382,
+      "grad_norm": 0.2881029546260834,
+      "learning_rate": 3.310132444229307e-06,
+      "loss": 0.5821,
+      "step": 2871
+    },
+    {
+      "epoch": 1.2103488249552112,
+      "grad_norm": 0.28204676508903503,
+      "learning_rate": 3.309073330490015e-06,
+      "loss": 0.576,
+      "step": 2872
+    },
+    {
+      "epoch": 1.2107703656865845,
+      "grad_norm": 0.2957751452922821,
+      "learning_rate": 3.308014054529205e-06,
+      "loss": 0.5785,
+      "step": 2873
+    },
+    {
+      "epoch": 1.2111919064179577,
+      "grad_norm": 0.2971872389316559,
+      "learning_rate": 3.3069546165592624e-06,
+      "loss": 0.6129,
+      "step": 2874
+    },
+    {
+      "epoch": 1.2116134471493307,
+      "grad_norm": 0.28660765290260315,
+      "learning_rate": 3.305895016792611e-06,
+      "loss": 0.5981,
+      "step": 2875
+    },
+    {
+      "epoch": 1.212034987880704,
+      "grad_norm": 0.29435017704963684,
+      "learning_rate": 3.3048352554417008e-06,
+      "loss": 0.5885,
+      "step": 2876
+    },
+    {
+      "epoch": 1.2124565286120772,
+      "grad_norm": 0.29911547899246216,
+      "learning_rate": 3.303775332719019e-06,
+      "loss": 0.5965,
+      "step": 2877
+    },
+    {
+      "epoch": 1.2128780693434502,
+      "grad_norm": 0.2892564535140991,
+      "learning_rate": 3.302715248837081e-06,
+      "loss": 0.611,
+      "step": 2878
+    },
+    {
+      "epoch": 1.2132996100748235,
+      "grad_norm": 0.2960320711135864,
+      "learning_rate": 3.3016550040084385e-06,
+      "loss": 0.5928,
+      "step": 2879
+    },
+    {
+      "epoch": 1.2137211508061967,
+      "grad_norm": 0.3021013140678406,
+      "learning_rate": 3.3005945984456726e-06,
+      "loss": 0.6042,
+      "step": 2880
+    },
+    {
+      "epoch": 1.2141426915375697,
+      "grad_norm": 0.2932465076446533,
+      "learning_rate": 3.2995340323613977e-06,
+      "loss": 0.597,
+      "step": 2881
+    },
+    {
+      "epoch": 1.214564232268943,
+      "grad_norm": 0.28704750537872314,
+      "learning_rate": 3.298473305968261e-06,
+      "loss": 0.582,
+      "step": 2882
+    },
+    {
+      "epoch": 1.2149857730003162,
+      "grad_norm": 0.2947606146335602,
+      "learning_rate": 3.297412419478941e-06,
+      "loss": 0.5785,
+      "step": 2883
+    },
+    {
+      "epoch": 1.2154073137316894,
+      "grad_norm": 0.27768388390541077,
+      "learning_rate": 3.2963513731061476e-06,
+      "loss": 0.5777,
+      "step": 2884
+    },
+    {
+      "epoch": 1.2158288544630624,
+      "grad_norm": 0.3019585907459259,
+      "learning_rate": 3.295290167062624e-06,
+      "loss": 0.6056,
+      "step": 2885
+    },
+    {
+      "epoch": 1.2162503951944357,
+      "grad_norm": 0.3015257716178894,
+      "learning_rate": 3.294228801561146e-06,
+      "loss": 0.6069,
+      "step": 2886
+    },
+    {
+      "epoch": 1.216671935925809,
+      "grad_norm": 0.2867284119129181,
+      "learning_rate": 3.2931672768145195e-06,
+      "loss": 0.584,
+      "step": 2887
+    },
+    {
+      "epoch": 1.217093476657182,
+      "grad_norm": 0.2988928556442261,
+      "learning_rate": 3.292105593035584e-06,
+      "loss": 0.5982,
+      "step": 2888
+    },
+    {
+      "epoch": 1.2175150173885552,
+      "grad_norm": 0.29856210947036743,
+      "learning_rate": 3.2910437504372088e-06,
+      "loss": 0.5827,
+      "step": 2889
+    },
+    {
+      "epoch": 1.2179365581199284,
+      "grad_norm": 0.29539811611175537,
+      "learning_rate": 3.2899817492322975e-06,
+      "loss": 0.5797,
+      "step": 2890
+    },
+    {
+      "epoch": 1.2183580988513014,
+      "grad_norm": 0.3076742887496948,
+      "learning_rate": 3.288919589633785e-06,
+      "loss": 0.6158,
+      "step": 2891
+    },
+    {
+      "epoch": 1.2187796395826747,
+      "grad_norm": 0.3075838088989258,
+      "learning_rate": 3.2878572718546355e-06,
+      "loss": 0.5921,
+      "step": 2892
+    },
+    {
+      "epoch": 1.219201180314048,
+      "grad_norm": 0.2868872582912445,
+      "learning_rate": 3.2867947961078474e-06,
+      "loss": 0.5982,
+      "step": 2893
+    },
+    {
+      "epoch": 1.219622721045421,
+      "grad_norm": 0.3105333745479584,
+      "learning_rate": 3.2857321626064507e-06,
+      "loss": 0.5862,
+      "step": 2894
+    },
+    {
+      "epoch": 1.2200442617767941,
+      "grad_norm": 0.30114465951919556,
+      "learning_rate": 3.284669371563507e-06,
+      "loss": 0.5958,
+      "step": 2895
+    },
+    {
+      "epoch": 1.2204658025081674,
+      "grad_norm": 0.2860717177391052,
+      "learning_rate": 3.2836064231921074e-06,
+      "loss": 0.5866,
+      "step": 2896
+    },
+    {
+      "epoch": 1.2208873432395406,
+      "grad_norm": 0.29310521483421326,
+      "learning_rate": 3.2825433177053767e-06,
+      "loss": 0.5618,
+      "step": 2897
+    },
+    {
+      "epoch": 1.2213088839709136,
+      "grad_norm": 0.2963055372238159,
+      "learning_rate": 3.281480055316471e-06,
+      "loss": 0.6013,
+      "step": 2898
+    },
+    {
+      "epoch": 1.2217304247022869,
+      "grad_norm": 0.2909132242202759,
+      "learning_rate": 3.280416636238577e-06,
+      "loss": 0.6108,
+      "step": 2899
+    },
+    {
+      "epoch": 1.2221519654336601,
+      "grad_norm": 0.2926136553287506,
+      "learning_rate": 3.2793530606849137e-06,
+      "loss": 0.5854,
+      "step": 2900
+    },
+    {
+      "epoch": 1.2225735061650331,
+      "grad_norm": 0.2822209894657135,
+      "learning_rate": 3.27828932886873e-06,
+      "loss": 0.5807,
+      "step": 2901
+    },
+    {
+      "epoch": 1.2229950468964064,
+      "grad_norm": 0.2973901629447937,
+      "learning_rate": 3.277225441003309e-06,
+      "loss": 0.5842,
+      "step": 2902
+    },
+    {
+      "epoch": 1.2234165876277796,
+      "grad_norm": 0.2844250202178955,
+      "learning_rate": 3.276161397301962e-06,
+      "loss": 0.619,
+      "step": 2903
+    },
+    {
+      "epoch": 1.2238381283591526,
+      "grad_norm": 0.28823116421699524,
+      "learning_rate": 3.275097197978032e-06,
+      "loss": 0.5932,
+      "step": 2904
+    },
+    {
+      "epoch": 1.2242596690905259,
+      "grad_norm": 0.28294095396995544,
+      "learning_rate": 3.274032843244896e-06,
+      "loss": 0.5754,
+      "step": 2905
+    },
+    {
+      "epoch": 1.224681209821899,
+      "grad_norm": 0.29827770590782166,
+      "learning_rate": 3.272968333315959e-06,
+      "loss": 0.6035,
+      "step": 2906
+    },
+    {
+      "epoch": 1.2251027505532721,
+      "grad_norm": 0.2882911264896393,
+      "learning_rate": 3.2719036684046583e-06,
+      "loss": 0.6187,
+      "step": 2907
+    },
+    {
+      "epoch": 1.2255242912846454,
+      "grad_norm": 0.3004266023635864,
+      "learning_rate": 3.270838848724463e-06,
+      "loss": 0.5902,
+      "step": 2908
+    },
+    {
+      "epoch": 1.2259458320160186,
+      "grad_norm": 0.2958316504955292,
+      "learning_rate": 3.2697738744888717e-06,
+      "loss": 0.5871,
+      "step": 2909
+    },
+    {
+      "epoch": 1.2263673727473918,
+      "grad_norm": 0.28620994091033936,
+      "learning_rate": 3.2687087459114146e-06,
+      "loss": 0.5914,
+      "step": 2910
+    },
+    {
+      "epoch": 1.2267889134787648,
+      "grad_norm": 0.2892815172672272,
+      "learning_rate": 3.2676434632056537e-06,
+      "loss": 0.5928,
+      "step": 2911
+    },
+    {
+      "epoch": 1.227210454210138,
+      "grad_norm": 0.30785197019577026,
+      "learning_rate": 3.2665780265851815e-06,
+      "loss": 0.581,
+      "step": 2912
+    },
+    {
+      "epoch": 1.2276319949415113,
+      "grad_norm": 0.27877235412597656,
+      "learning_rate": 3.2655124362636196e-06,
+      "loss": 0.5969,
+      "step": 2913
+    },
+    {
+      "epoch": 1.2280535356728843,
+      "grad_norm": 0.2913534939289093,
+      "learning_rate": 3.2644466924546238e-06,
+      "loss": 0.5951,
+      "step": 2914
+    },
+    {
+      "epoch": 1.2284750764042576,
+      "grad_norm": 0.28130847215652466,
+      "learning_rate": 3.263380795371877e-06,
+      "loss": 0.5803,
+      "step": 2915
+    },
+    {
+      "epoch": 1.2288966171356308,
+      "grad_norm": 0.2923814654350281,
+      "learning_rate": 3.2623147452290967e-06,
+      "loss": 0.5678,
+      "step": 2916
+    },
+    {
+      "epoch": 1.2293181578670038,
+      "grad_norm": 0.2956162691116333,
+      "learning_rate": 3.2612485422400266e-06,
+      "loss": 0.6104,
+      "step": 2917
+    },
+    {
+      "epoch": 1.229739698598377,
+      "grad_norm": 0.2928110659122467,
+      "learning_rate": 3.2601821866184447e-06,
+      "loss": 0.5769,
+      "step": 2918
+    },
+    {
+      "epoch": 1.2301612393297503,
+      "grad_norm": 0.2939610183238983,
+      "learning_rate": 3.25911567857816e-06,
+      "loss": 0.5973,
+      "step": 2919
+    },
+    {
+      "epoch": 1.2305827800611233,
+      "grad_norm": 0.30369046330451965,
+      "learning_rate": 3.2580490183330065e-06,
+      "loss": 0.5961,
+      "step": 2920
+    },
+    {
+      "epoch": 1.2310043207924966,
+      "grad_norm": 0.2875109016895294,
+      "learning_rate": 3.2569822060968565e-06,
+      "loss": 0.6117,
+      "step": 2921
+    },
+    {
+      "epoch": 1.2314258615238698,
+      "grad_norm": 0.29523080587387085,
+      "learning_rate": 3.255915242083606e-06,
+      "loss": 0.5902,
+      "step": 2922
+    },
+    {
+      "epoch": 1.231847402255243,
+      "grad_norm": 0.28411832451820374,
+      "learning_rate": 3.254848126507187e-06,
+      "loss": 0.5933,
+      "step": 2923
+    },
+    {
+      "epoch": 1.232268942986616,
+      "grad_norm": 0.29212629795074463,
+      "learning_rate": 3.2537808595815557e-06,
+      "loss": 0.5933,
+      "step": 2924
+    },
+    {
+      "epoch": 1.2326904837179893,
+      "grad_norm": 0.30325162410736084,
+      "learning_rate": 3.252713441520706e-06,
+      "loss": 0.588,
+      "step": 2925
+    },
+    {
+      "epoch": 1.2331120244493623,
+      "grad_norm": 0.28501972556114197,
+      "learning_rate": 3.251645872538656e-06,
+      "loss": 0.6129,
+      "step": 2926
+    },
+    {
+      "epoch": 1.2335335651807355,
+      "grad_norm": 0.2803444564342499,
+      "learning_rate": 3.2505781528494574e-06,
+      "loss": 0.5523,
+      "step": 2927
+    },
+    {
+      "epoch": 1.2339551059121088,
+      "grad_norm": 0.3181038200855255,
+      "learning_rate": 3.2495102826671903e-06,
+      "loss": 0.6013,
+      "step": 2928
+    },
+    {
+      "epoch": 1.234376646643482,
+      "grad_norm": 0.2950741946697235,
+      "learning_rate": 3.2484422622059657e-06,
+      "loss": 0.5854,
+      "step": 2929
+    },
+    {
+      "epoch": 1.234798187374855,
+      "grad_norm": 0.302639901638031,
+      "learning_rate": 3.247374091679926e-06,
+      "loss": 0.5916,
+      "step": 2930
+    },
+    {
+      "epoch": 1.2352197281062283,
+      "grad_norm": 0.3046797215938568,
+      "learning_rate": 3.2463057713032404e-06,
+      "loss": 0.5509,
+      "step": 2931
+    },
+    {
+      "epoch": 1.2356412688376015,
+      "grad_norm": 0.300002783536911,
+      "learning_rate": 3.2452373012901117e-06,
+      "loss": 0.614,
+      "step": 2932
+    },
+    {
+      "epoch": 1.2360628095689745,
+      "grad_norm": 0.29890480637550354,
+      "learning_rate": 3.2441686818547706e-06,
+      "loss": 0.595,
+      "step": 2933
+    },
+    {
+      "epoch": 1.2364843503003478,
+      "grad_norm": 0.2952747941017151,
+      "learning_rate": 3.2430999132114786e-06,
+      "loss": 0.5786,
+      "step": 2934
+    },
+    {
+      "epoch": 1.236905891031721,
+      "grad_norm": 0.3025519847869873,
+      "learning_rate": 3.242030995574527e-06,
+      "loss": 0.5682,
+      "step": 2935
+    },
+    {
+      "epoch": 1.2373274317630942,
+      "grad_norm": 0.2985527813434601,
+      "learning_rate": 3.240961929158236e-06,
+      "loss": 0.5809,
+      "step": 2936
+    },
+    {
+      "epoch": 1.2377489724944672,
+      "grad_norm": 0.29586586356163025,
+      "learning_rate": 3.239892714176959e-06,
+      "loss": 0.5917,
+      "step": 2937
+    },
+    {
+      "epoch": 1.2381705132258405,
+      "grad_norm": 0.3063954710960388,
+      "learning_rate": 3.238823350845073e-06,
+      "loss": 0.589,
+      "step": 2938
+    },
+    {
+      "epoch": 1.2385920539572135,
+      "grad_norm": 0.29052525758743286,
+      "learning_rate": 3.2377538393769916e-06,
+      "loss": 0.5837,
+      "step": 2939
+    },
+    {
+      "epoch": 1.2390135946885867,
+      "grad_norm": 0.30704253911972046,
+      "learning_rate": 3.236684179987152e-06,
+      "loss": 0.5764,
+      "step": 2940
+    },
+    {
+      "epoch": 1.23943513541996,
+      "grad_norm": 0.2853536605834961,
+      "learning_rate": 3.235614372890026e-06,
+      "loss": 0.5929,
+      "step": 2941
+    },
+    {
+      "epoch": 1.2398566761513332,
+      "grad_norm": 0.29201391339302063,
+      "learning_rate": 3.2345444183001128e-06,
+      "loss": 0.6152,
+      "step": 2942
+    },
+    {
+      "epoch": 1.2402782168827062,
+      "grad_norm": 0.30878573656082153,
+      "learning_rate": 3.2334743164319403e-06,
+      "loss": 0.5747,
+      "step": 2943
+    },
+    {
+      "epoch": 1.2406997576140795,
+      "grad_norm": 0.3250546455383301,
+      "learning_rate": 3.232404067500069e-06,
+      "loss": 0.6014,
+      "step": 2944
+    },
+    {
+      "epoch": 1.2411212983454527,
+      "grad_norm": 0.2825911045074463,
+      "learning_rate": 3.2313336717190845e-06,
+      "loss": 0.6263,
+      "step": 2945
+    },
+    {
+      "epoch": 1.2415428390768257,
+      "grad_norm": 0.31028929352760315,
+      "learning_rate": 3.2302631293036065e-06,
+      "loss": 0.5811,
+      "step": 2946
+    },
+    {
+      "epoch": 1.241964379808199,
+      "grad_norm": 0.3401135504245758,
+      "learning_rate": 3.2291924404682797e-06,
+      "loss": 0.5562,
+      "step": 2947
+    },
+    {
+      "epoch": 1.2423859205395722,
+      "grad_norm": 0.2901981770992279,
+      "learning_rate": 3.228121605427782e-06,
+      "loss": 0.6025,
+      "step": 2948
+    },
+    {
+      "epoch": 1.2428074612709452,
+      "grad_norm": 0.36488083004951477,
+      "learning_rate": 3.2270506243968167e-06,
+      "loss": 0.5935,
+      "step": 2949
+    },
+    {
+      "epoch": 1.2432290020023185,
+      "grad_norm": 0.30141857266426086,
+      "learning_rate": 3.2259794975901217e-06,
+      "loss": 0.584,
+      "step": 2950
+    },
+    {
+      "epoch": 1.2436505427336917,
+      "grad_norm": 0.2890578508377075,
+      "learning_rate": 3.2249082252224583e-06,
+      "loss": 0.5999,
+      "step": 2951
+    },
+    {
+      "epoch": 1.2440720834650647,
+      "grad_norm": 0.31032517552375793,
+      "learning_rate": 3.2238368075086206e-06,
+      "loss": 0.598,
+      "step": 2952
+    },
+    {
+      "epoch": 1.244493624196438,
+      "grad_norm": 0.3465369641780853,
+      "learning_rate": 3.2227652446634305e-06,
+      "loss": 0.5634,
+      "step": 2953
+    },
+    {
+      "epoch": 1.2449151649278112,
+      "grad_norm": 0.27520596981048584,
+      "learning_rate": 3.2216935369017405e-06,
+      "loss": 0.5801,
+      "step": 2954
+    },
+    {
+      "epoch": 1.2453367056591844,
+      "grad_norm": 0.2888118028640747,
+      "learning_rate": 3.2206216844384306e-06,
+      "loss": 0.5708,
+      "step": 2955
+    },
+    {
+      "epoch": 1.2457582463905574,
+      "grad_norm": 0.34516483545303345,
+      "learning_rate": 3.2195496874884104e-06,
+      "loss": 0.5969,
+      "step": 2956
+    },
+    {
+      "epoch": 1.2461797871219307,
+      "grad_norm": 0.30031219124794006,
+      "learning_rate": 3.2184775462666174e-06,
+      "loss": 0.6015,
+      "step": 2957
+    },
+    {
+      "epoch": 1.246601327853304,
+      "grad_norm": 0.2976769506931305,
+      "learning_rate": 3.2174052609880202e-06,
+      "loss": 0.5848,
+      "step": 2958
+    },
+    {
+      "epoch": 1.247022868584677,
+      "grad_norm": 0.32738006114959717,
+      "learning_rate": 3.216332831867614e-06,
+      "loss": 0.5907,
+      "step": 2959
+    },
+    {
+      "epoch": 1.2474444093160502,
+      "grad_norm": 0.29337558150291443,
+      "learning_rate": 3.215260259120425e-06,
+      "loss": 0.5794,
+      "step": 2960
+    },
+    {
+      "epoch": 1.2478659500474234,
+      "grad_norm": 0.28369826078414917,
+      "learning_rate": 3.214187542961506e-06,
+      "loss": 0.5994,
+      "step": 2961
+    },
+    {
+      "epoch": 1.2482874907787964,
+      "grad_norm": 0.3199905753135681,
+      "learning_rate": 3.213114683605941e-06,
+      "loss": 0.5893,
+      "step": 2962
+    },
+    {
+      "epoch": 1.2487090315101697,
+      "grad_norm": 0.2889053225517273,
+      "learning_rate": 3.212041681268839e-06,
+      "loss": 0.5883,
+      "step": 2963
+    },
+    {
+      "epoch": 1.249130572241543,
+      "grad_norm": 0.29060226678848267,
+      "learning_rate": 3.210968536165343e-06,
+      "loss": 0.5979,
+      "step": 2964
+    },
+    {
+      "epoch": 1.249552112972916,
+      "grad_norm": 0.28468137979507446,
+      "learning_rate": 3.2098952485106203e-06,
+      "loss": 0.5913,
+      "step": 2965
+    },
+    {
+      "epoch": 1.2499736537042891,
+      "grad_norm": 0.29601263999938965,
+      "learning_rate": 3.2088218185198676e-06,
+      "loss": 0.5909,
+      "step": 2966
+    },
+    {
+      "epoch": 1.2503951944356624,
+      "grad_norm": 0.2911202013492584,
+      "learning_rate": 3.207748246408312e-06,
+      "loss": 0.5931,
+      "step": 2967
+    },
+    {
+      "epoch": 1.2508167351670356,
+      "grad_norm": 0.28125715255737305,
+      "learning_rate": 3.2066745323912065e-06,
+      "loss": 0.5771,
+      "step": 2968
+    },
+    {
+      "epoch": 1.2512382758984086,
+      "grad_norm": 0.2744985520839691,
+      "learning_rate": 3.2056006766838344e-06,
+      "loss": 0.5891,
+      "step": 2969
+    },
+    {
+      "epoch": 1.2516598166297819,
+      "grad_norm": 0.2888781726360321,
+      "learning_rate": 3.204526679501506e-06,
+      "loss": 0.5904,
+      "step": 2970
+    },
+    {
+      "epoch": 1.252081357361155,
+      "grad_norm": 0.2891712486743927,
+      "learning_rate": 3.203452541059563e-06,
+      "loss": 0.5907,
+      "step": 2971
+    },
+    {
+      "epoch": 1.2525028980925281,
+      "grad_norm": 0.2881668210029602,
+      "learning_rate": 3.2023782615733713e-06,
+      "loss": 0.5968,
+      "step": 2972
+    },
+    {
+      "epoch": 1.2529244388239014,
+      "grad_norm": 0.2839834988117218,
+      "learning_rate": 3.201303841258327e-06,
+      "loss": 0.5973,
+      "step": 2973
+    },
+    {
+      "epoch": 1.2533459795552746,
+      "grad_norm": 0.2921658456325531,
+      "learning_rate": 3.200229280329855e-06,
+      "loss": 0.5758,
+      "step": 2974
+    },
+    {
+      "epoch": 1.2537675202866476,
+      "grad_norm": 0.284833699464798,
+      "learning_rate": 3.1991545790034083e-06,
+      "loss": 0.615,
+      "step": 2975
+    },
+    {
+      "epoch": 1.2541890610180209,
+      "grad_norm": 0.29288944602012634,
+      "learning_rate": 3.1980797374944673e-06,
+      "loss": 0.5921,
+      "step": 2976
+    },
+    {
+      "epoch": 1.254610601749394,
+      "grad_norm": 0.2956557869911194,
+      "learning_rate": 3.1970047560185402e-06,
+      "loss": 0.5651,
+      "step": 2977
+    },
+    {
+      "epoch": 1.255032142480767,
+      "grad_norm": 0.29037681221961975,
+      "learning_rate": 3.195929634791165e-06,
+      "loss": 0.579,
+      "step": 2978
+    },
+    {
+      "epoch": 1.2554536832121403,
+      "grad_norm": 0.2828306555747986,
+      "learning_rate": 3.1948543740279055e-06,
+      "loss": 0.5945,
+      "step": 2979
+    },
+    {
+      "epoch": 1.2558752239435136,
+      "grad_norm": 0.2788093686103821,
+      "learning_rate": 3.193778973944354e-06,
+      "loss": 0.582,
+      "step": 2980
+    },
+    {
+      "epoch": 1.2562967646748868,
+      "grad_norm": 0.2912706732749939,
+      "learning_rate": 3.1927034347561335e-06,
+      "loss": 0.5773,
+      "step": 2981
+    },
+    {
+      "epoch": 1.2567183054062598,
+      "grad_norm": 0.29398930072784424,
+      "learning_rate": 3.191627756678891e-06,
+      "loss": 0.5899,
+      "step": 2982
+    },
+    {
+      "epoch": 1.257139846137633,
+      "grad_norm": 0.2964211404323578,
+      "learning_rate": 3.190551939928303e-06,
+      "loss": 0.6057,
+      "step": 2983
+    },
+    {
+      "epoch": 1.257561386869006,
+      "grad_norm": 0.2916293740272522,
+      "learning_rate": 3.1894759847200744e-06,
+      "loss": 0.6211,
+      "step": 2984
+    },
+    {
+      "epoch": 1.2579829276003793,
+      "grad_norm": 0.2902526259422302,
+      "learning_rate": 3.1883998912699384e-06,
+      "loss": 0.5923,
+      "step": 2985
+    },
+    {
+      "epoch": 1.2584044683317526,
+      "grad_norm": 0.28431254625320435,
+      "learning_rate": 3.187323659793652e-06,
+      "loss": 0.564,
+      "step": 2986
+    },
+    {
+      "epoch": 1.2588260090631258,
+      "grad_norm": 0.30226027965545654,
+      "learning_rate": 3.186247290507005e-06,
+      "loss": 0.5863,
+      "step": 2987
+    },
+    {
+      "epoch": 1.2592475497944988,
+      "grad_norm": 0.29238319396972656,
+      "learning_rate": 3.185170783625812e-06,
+      "loss": 0.594,
+      "step": 2988
+    },
+    {
+      "epoch": 1.259669090525872,
+      "grad_norm": 0.3000998795032501,
+      "learning_rate": 3.1840941393659154e-06,
+      "loss": 0.5962,
+      "step": 2989
+    },
+    {
+      "epoch": 1.2600906312572453,
+      "grad_norm": 0.30129823088645935,
+      "learning_rate": 3.1830173579431856e-06,
+      "loss": 0.5919,
+      "step": 2990
+    },
+    {
+      "epoch": 1.2605121719886183,
+      "grad_norm": 0.30526435375213623,
+      "learning_rate": 3.18194043957352e-06,
+      "loss": 0.6114,
+      "step": 2991
+    },
+    {
+      "epoch": 1.2609337127199916,
+      "grad_norm": 0.2874204218387604,
+      "learning_rate": 3.1808633844728448e-06,
+      "loss": 0.6008,
+      "step": 2992
+    },
+    {
+      "epoch": 1.2613552534513648,
+      "grad_norm": 0.316562682390213,
+      "learning_rate": 3.1797861928571104e-06,
+      "loss": 0.5671,
+      "step": 2993
+    },
+    {
+      "epoch": 1.261776794182738,
+      "grad_norm": 0.3072529137134552,
+      "learning_rate": 3.1787088649423006e-06,
+      "loss": 0.5906,
+      "step": 2994
+    },
+    {
+      "epoch": 1.262198334914111,
+      "grad_norm": 0.289010226726532,
+      "learning_rate": 3.177631400944419e-06,
+      "loss": 0.5932,
+      "step": 2995
+    },
+    {
+      "epoch": 1.2626198756454843,
+      "grad_norm": 0.28156962990760803,
+      "learning_rate": 3.1765538010795023e-06,
+      "loss": 0.5854,
+      "step": 2996
+    },
+    {
+      "epoch": 1.2630414163768573,
+      "grad_norm": 0.29249367117881775,
+      "learning_rate": 3.1754760655636117e-06,
+      "loss": 0.6043,
+      "step": 2997
+    },
+    {
+      "epoch": 1.2634629571082305,
+      "grad_norm": 0.31385326385498047,
+      "learning_rate": 3.1743981946128364e-06,
+      "loss": 0.6256,
+      "step": 2998
+    },
+    {
+      "epoch": 1.2638844978396038,
+      "grad_norm": 0.30105239152908325,
+      "learning_rate": 3.1733201884432927e-06,
+      "loss": 0.584,
+      "step": 2999
+    },
+    {
+      "epoch": 1.264306038570977,
+      "grad_norm": 0.28737711906433105,
+      "learning_rate": 3.1722420472711234e-06,
+      "loss": 0.5802,
+      "step": 3000
+    },
+    {
+      "epoch": 1.26472757930235,
+      "grad_norm": 0.31372836232185364,
+      "learning_rate": 3.1711637713125e-06,
+      "loss": 0.6136,
+      "step": 3001
+    },
+    {
+      "epoch": 1.2651491200337233,
+      "grad_norm": 0.30143654346466064,
+      "learning_rate": 3.170085360783618e-06,
+      "loss": 0.6099,
+      "step": 3002
+    },
+    {
+      "epoch": 1.2655706607650965,
+      "grad_norm": 0.2839278280735016,
+      "learning_rate": 3.1690068159007042e-06,
+      "loss": 0.6051,
+      "step": 3003
+    },
+    {
+      "epoch": 1.2659922014964695,
+      "grad_norm": 0.2975417971611023,
+      "learning_rate": 3.167928136880008e-06,
+      "loss": 0.5877,
+      "step": 3004
+    },
+    {
+      "epoch": 1.2664137422278428,
+      "grad_norm": 0.3214844763278961,
+      "learning_rate": 3.1668493239378084e-06,
+      "loss": 0.5765,
+      "step": 3005
+    },
+    {
+      "epoch": 1.266835282959216,
+      "grad_norm": 0.30230712890625,
+      "learning_rate": 3.1657703772904114e-06,
+      "loss": 0.5972,
+      "step": 3006
+    },
+    {
+      "epoch": 1.2672568236905892,
+      "grad_norm": 0.3001837432384491,
+      "learning_rate": 3.1646912971541465e-06,
+      "loss": 0.5798,
+      "step": 3007
+    },
+    {
+      "epoch": 1.2676783644219622,
+      "grad_norm": 0.2910992205142975,
+      "learning_rate": 3.163612083745375e-06,
+      "loss": 0.5771,
+      "step": 3008
+    },
+    {
+      "epoch": 1.2680999051533355,
+      "grad_norm": 0.3037284314632416,
+      "learning_rate": 3.1625327372804804e-06,
+      "loss": 0.605,
+      "step": 3009
+    },
+    {
+      "epoch": 1.2685214458847085,
+      "grad_norm": 0.3083994388580322,
+      "learning_rate": 3.161453257975875e-06,
+      "loss": 0.5878,
+      "step": 3010
+    },
+    {
+      "epoch": 1.2689429866160817,
+      "grad_norm": 0.2846064269542694,
+      "learning_rate": 3.160373646047999e-06,
+      "loss": 0.6185,
+      "step": 3011
+    },
+    {
+      "epoch": 1.269364527347455,
+      "grad_norm": 0.2929023504257202,
+      "learning_rate": 3.1592939017133163e-06,
+      "loss": 0.5738,
+      "step": 3012
+    },
+    {
+      "epoch": 1.2697860680788282,
+      "grad_norm": 0.32218071818351746,
+      "learning_rate": 3.158214025188318e-06,
+      "loss": 0.5859,
+      "step": 3013
+    },
+    {
+      "epoch": 1.2702076088102012,
+      "grad_norm": 0.3124587833881378,
+      "learning_rate": 3.157134016689524e-06,
+      "loss": 0.5915,
+      "step": 3014
+    },
+    {
+      "epoch": 1.2706291495415745,
+      "grad_norm": 0.29292675852775574,
+      "learning_rate": 3.156053876433479e-06,
+      "loss": 0.593,
+      "step": 3015
+    },
+    {
+      "epoch": 1.2710506902729477,
+      "grad_norm": 0.27880150079727173,
+      "learning_rate": 3.154973604636754e-06,
+      "loss": 0.5874,
+      "step": 3016
+    },
+    {
+      "epoch": 1.2714722310043207,
+      "grad_norm": 0.29758381843566895,
+      "learning_rate": 3.153893201515946e-06,
+      "loss": 0.5804,
+      "step": 3017
+    },
+    {
+      "epoch": 1.271893771735694,
+      "grad_norm": 0.31115978956222534,
+      "learning_rate": 3.152812667287679e-06,
+      "loss": 0.5835,
+      "step": 3018
+    },
+    {
+      "epoch": 1.2723153124670672,
+      "grad_norm": 0.2895902097225189,
+      "learning_rate": 3.1517320021686037e-06,
+      "loss": 0.5981,
+      "step": 3019
+    },
+    {
+      "epoch": 1.2727368531984404,
+      "grad_norm": 0.29717880487442017,
+      "learning_rate": 3.150651206375397e-06,
+      "loss": 0.5784,
+      "step": 3020
+    },
+    {
+      "epoch": 1.2731583939298134,
+      "grad_norm": 0.29164358973503113,
+      "learning_rate": 3.1495702801247596e-06,
+      "loss": 0.6118,
+      "step": 3021
+    },
+    {
+      "epoch": 1.2735799346611867,
+      "grad_norm": 0.3144710659980774,
+      "learning_rate": 3.148489223633422e-06,
+      "loss": 0.5939,
+      "step": 3022
+    },
+    {
+      "epoch": 1.2740014753925597,
+      "grad_norm": 0.29779142141342163,
+      "learning_rate": 3.147408037118139e-06,
+      "loss": 0.592,
+      "step": 3023
+    },
+    {
+      "epoch": 1.274423016123933,
+      "grad_norm": 0.30781176686286926,
+      "learning_rate": 3.1463267207956915e-06,
+      "loss": 0.6107,
+      "step": 3024
+    },
+    {
+      "epoch": 1.2748445568553062,
+      "grad_norm": 0.2844652235507965,
+      "learning_rate": 3.145245274882886e-06,
+      "loss": 0.6107,
+      "step": 3025
+    },
+    {
+      "epoch": 1.2752660975866794,
+      "grad_norm": 0.29495880007743835,
+      "learning_rate": 3.1441636995965564e-06,
+      "loss": 0.5858,
+      "step": 3026
+    },
+    {
+      "epoch": 1.2756876383180524,
+      "grad_norm": 0.3077981173992157,
+      "learning_rate": 3.1430819951535603e-06,
+      "loss": 0.5694,
+      "step": 3027
+    },
+    {
+      "epoch": 1.2761091790494257,
+      "grad_norm": 0.29625582695007324,
+      "learning_rate": 3.142000161770783e-06,
+      "loss": 0.5801,
+      "step": 3028
+    },
+    {
+      "epoch": 1.2765307197807987,
+      "grad_norm": 0.30319759249687195,
+      "learning_rate": 3.140918199665136e-06,
+      "loss": 0.5973,
+      "step": 3029
+    },
+    {
+      "epoch": 1.276952260512172,
+      "grad_norm": 0.2957063615322113,
+      "learning_rate": 3.1398361090535558e-06,
+      "loss": 0.6153,
+      "step": 3030
+    },
+    {
+      "epoch": 1.2773738012435452,
+      "grad_norm": 0.28444933891296387,
+      "learning_rate": 3.1387538901530035e-06,
+      "loss": 0.5783,
+      "step": 3031
+    },
+    {
+      "epoch": 1.2777953419749184,
+      "grad_norm": 0.3025360703468323,
+      "learning_rate": 3.137671543180467e-06,
+      "loss": 0.5645,
+      "step": 3032
+    },
+    {
+      "epoch": 1.2782168827062914,
+      "grad_norm": 0.2887069284915924,
+      "learning_rate": 3.1365890683529628e-06,
+      "loss": 0.594,
+      "step": 3033
+    },
+    {
+      "epoch": 1.2786384234376647,
+      "grad_norm": 0.30408304929733276,
+      "learning_rate": 3.1355064658875266e-06,
+      "loss": 0.5821,
+      "step": 3034
+    },
+    {
+      "epoch": 1.2790599641690379,
+      "grad_norm": 0.30999764800071716,
+      "learning_rate": 3.134423736001225e-06,
+      "loss": 0.5943,
+      "step": 3035
+    },
+    {
+      "epoch": 1.279481504900411,
+      "grad_norm": 0.3117923438549042,
+      "learning_rate": 3.1333408789111487e-06,
+      "loss": 0.5917,
+      "step": 3036
+    },
+    {
+      "epoch": 1.2799030456317841,
+      "grad_norm": 0.29884204268455505,
+      "learning_rate": 3.132257894834412e-06,
+      "loss": 0.5687,
+      "step": 3037
+    },
+    {
+      "epoch": 1.2803245863631574,
+      "grad_norm": 0.300371915102005,
+      "learning_rate": 3.131174783988159e-06,
+      "loss": 0.5984,
+      "step": 3038
+    },
+    {
+      "epoch": 1.2807461270945306,
+      "grad_norm": 0.2932301163673401,
+      "learning_rate": 3.1300915465895543e-06,
+      "loss": 0.5765,
+      "step": 3039
+    },
+    {
+      "epoch": 1.2811676678259036,
+      "grad_norm": 0.29098963737487793,
+      "learning_rate": 3.1290081828557915e-06,
+      "loss": 0.593,
+      "step": 3040
+    },
+    {
+      "epoch": 1.2815892085572769,
+      "grad_norm": 0.29830822348594666,
+      "learning_rate": 3.1279246930040863e-06,
+      "loss": 0.6184,
+      "step": 3041
+    },
+    {
+      "epoch": 1.2820107492886499,
+      "grad_norm": 0.2879214584827423,
+      "learning_rate": 3.1268410772516832e-06,
+      "loss": 0.607,
+      "step": 3042
+    },
+    {
+      "epoch": 1.2824322900200231,
+      "grad_norm": 0.2879657447338104,
+      "learning_rate": 3.12575733581585e-06,
+      "loss": 0.5948,
+      "step": 3043
+    },
+    {
+      "epoch": 1.2828538307513964,
+      "grad_norm": 0.29299023747444153,
+      "learning_rate": 3.1246734689138793e-06,
+      "loss": 0.5786,
+      "step": 3044
+    },
+    {
+      "epoch": 1.2832753714827696,
+      "grad_norm": 0.292179137468338,
+      "learning_rate": 3.1235894767630903e-06,
+      "loss": 0.5816,
+      "step": 3045
+    },
+    {
+      "epoch": 1.2836969122141426,
+      "grad_norm": 0.2995583713054657,
+      "learning_rate": 3.122505359580825e-06,
+      "loss": 0.6051,
+      "step": 3046
+    },
+    {
+      "epoch": 1.2841184529455159,
+      "grad_norm": 0.3072488605976105,
+      "learning_rate": 3.121421117584455e-06,
+      "loss": 0.5994,
+      "step": 3047
+    },
+    {
+      "epoch": 1.284539993676889,
+      "grad_norm": 0.28982973098754883,
+      "learning_rate": 3.120336750991371e-06,
+      "loss": 0.6169,
+      "step": 3048
+    },
+    {
+      "epoch": 1.284961534408262,
+      "grad_norm": 0.2878229022026062,
+      "learning_rate": 3.119252260018992e-06,
+      "loss": 0.6058,
+      "step": 3049
+    },
+    {
+      "epoch": 1.2853830751396353,
+      "grad_norm": 0.2940329909324646,
+      "learning_rate": 3.1181676448847636e-06,
+      "loss": 0.6003,
+      "step": 3050
+    },
+    {
+      "epoch": 1.2858046158710086,
+      "grad_norm": 0.3074142038822174,
+      "learning_rate": 3.1170829058061525e-06,
+      "loss": 0.5663,
+      "step": 3051
+    },
+    {
+      "epoch": 1.2862261566023818,
+      "grad_norm": 0.2887779176235199,
+      "learning_rate": 3.1159980430006515e-06,
+      "loss": 0.591,
+      "step": 3052
+    },
+    {
+      "epoch": 1.2866476973337548,
+      "grad_norm": 0.2836134731769562,
+      "learning_rate": 3.11491305668578e-06,
+      "loss": 0.6101,
+      "step": 3053
+    },
+    {
+      "epoch": 1.287069238065128,
+      "grad_norm": 0.30661657452583313,
+      "learning_rate": 3.113827947079081e-06,
+      "loss": 0.5989,
+      "step": 3054
+    },
+    {
+      "epoch": 1.287490778796501,
+      "grad_norm": 0.3115035891532898,
+      "learning_rate": 3.1127427143981203e-06,
+      "loss": 0.5968,
+      "step": 3055
+    },
+    {
+      "epoch": 1.2879123195278743,
+      "grad_norm": 0.28865671157836914,
+      "learning_rate": 3.1116573588604926e-06,
+      "loss": 0.5908,
+      "step": 3056
+    },
+    {
+      "epoch": 1.2883338602592476,
+      "grad_norm": 0.2864128351211548,
+      "learning_rate": 3.110571880683813e-06,
+      "loss": 0.5653,
+      "step": 3057
+    },
+    {
+      "epoch": 1.2887554009906208,
+      "grad_norm": 0.2976711094379425,
+      "learning_rate": 3.109486280085723e-06,
+      "loss": 0.6111,
+      "step": 3058
+    },
+    {
+      "epoch": 1.2891769417219938,
+      "grad_norm": 0.3057122230529785,
+      "learning_rate": 3.10840055728389e-06,
+      "loss": 0.588,
+      "step": 3059
+    },
+    {
+      "epoch": 1.289598482453367,
+      "grad_norm": 0.2989809215068817,
+      "learning_rate": 3.1073147124960033e-06,
+      "loss": 0.581,
+      "step": 3060
+    },
+    {
+      "epoch": 1.2900200231847403,
+      "grad_norm": 0.2891315817832947,
+      "learning_rate": 3.1062287459397784e-06,
+      "loss": 0.6316,
+      "step": 3061
+    },
+    {
+      "epoch": 1.2904415639161133,
+      "grad_norm": 0.2992073595523834,
+      "learning_rate": 3.1051426578329538e-06,
+      "loss": 0.6061,
+      "step": 3062
+    },
+    {
+      "epoch": 1.2908631046474865,
+      "grad_norm": 0.3035721778869629,
+      "learning_rate": 3.104056448393295e-06,
+      "loss": 0.5881,
+      "step": 3063
+    },
+    {
+      "epoch": 1.2912846453788598,
+      "grad_norm": 0.2956811189651489,
+      "learning_rate": 3.1029701178385886e-06,
+      "loss": 0.5976,
+      "step": 3064
+    },
+    {
+      "epoch": 1.291706186110233,
+      "grad_norm": 0.28393349051475525,
+      "learning_rate": 3.1018836663866482e-06,
+      "loss": 0.6024,
+      "step": 3065
+    },
+    {
+      "epoch": 1.292127726841606,
+      "grad_norm": 0.30275246500968933,
+      "learning_rate": 3.100797094255309e-06,
+      "loss": 0.5873,
+      "step": 3066
+    },
+    {
+      "epoch": 1.2925492675729793,
+      "grad_norm": 0.29854869842529297,
+      "learning_rate": 3.0997104016624324e-06,
+      "loss": 0.5804,
+      "step": 3067
+    },
+    {
+      "epoch": 1.2929708083043523,
+      "grad_norm": 0.29440420866012573,
+      "learning_rate": 3.098623588825904e-06,
+      "loss": 0.6026,
+      "step": 3068
+    },
+    {
+      "epoch": 1.2933923490357255,
+      "grad_norm": 0.3163248300552368,
+      "learning_rate": 3.0975366559636326e-06,
+      "loss": 0.5682,
+      "step": 3069
+    },
+    {
+      "epoch": 1.2938138897670988,
+      "grad_norm": 0.2915099263191223,
+      "learning_rate": 3.0964496032935513e-06,
+      "loss": 0.5914,
+      "step": 3070
+    },
+    {
+      "epoch": 1.294235430498472,
+      "grad_norm": 0.30830061435699463,
+      "learning_rate": 3.095362431033616e-06,
+      "loss": 0.6006,
+      "step": 3071
+    },
+    {
+      "epoch": 1.294656971229845,
+      "grad_norm": 0.2983812689781189,
+      "learning_rate": 3.0942751394018107e-06,
+      "loss": 0.6032,
+      "step": 3072
+    },
+    {
+      "epoch": 1.2950785119612183,
+      "grad_norm": 0.3005537986755371,
+      "learning_rate": 3.0931877286161375e-06,
+      "loss": 0.576,
+      "step": 3073
+    },
+    {
+      "epoch": 1.2955000526925915,
+      "grad_norm": 0.3035803437232971,
+      "learning_rate": 3.092100198894627e-06,
+      "loss": 0.5728,
+      "step": 3074
+    },
+    {
+      "epoch": 1.2959215934239645,
+      "grad_norm": 0.2781146168708801,
+      "learning_rate": 3.0910125504553324e-06,
+      "loss": 0.6068,
+      "step": 3075
+    },
+    {
+      "epoch": 1.2963431341553378,
+      "grad_norm": 0.2845194637775421,
+      "learning_rate": 3.089924783516329e-06,
+      "loss": 0.6127,
+      "step": 3076
+    },
+    {
+      "epoch": 1.296764674886711,
+      "grad_norm": 0.2918822169303894,
+      "learning_rate": 3.088836898295718e-06,
+      "loss": 0.6184,
+      "step": 3077
+    },
+    {
+      "epoch": 1.2971862156180842,
+      "grad_norm": 0.282073050737381,
+      "learning_rate": 3.087748895011623e-06,
+      "loss": 0.6056,
+      "step": 3078
+    },
+    {
+      "epoch": 1.2976077563494572,
+      "grad_norm": 0.2913799583911896,
+      "learning_rate": 3.0866607738821924e-06,
+      "loss": 0.597,
+      "step": 3079
+    },
+    {
+      "epoch": 1.2980292970808305,
+      "grad_norm": 0.300231009721756,
+      "learning_rate": 3.0855725351255974e-06,
+      "loss": 0.5771,
+      "step": 3080
+    },
+    {
+      "epoch": 1.2984508378122035,
+      "grad_norm": 0.28110629320144653,
+      "learning_rate": 3.0844841789600317e-06,
+      "loss": 0.5723,
+      "step": 3081
+    },
+    {
+      "epoch": 1.2988723785435767,
+      "grad_norm": 0.2900198996067047,
+      "learning_rate": 3.0833957056037155e-06,
+      "loss": 0.5858,
+      "step": 3082
+    },
+    {
+      "epoch": 1.29929391927495,
+      "grad_norm": 0.31364527344703674,
+      "learning_rate": 3.0823071152748908e-06,
+      "loss": 0.6205,
+      "step": 3083
+    },
+    {
+      "epoch": 1.2997154600063232,
+      "grad_norm": 0.3030346632003784,
+      "learning_rate": 3.081218408191823e-06,
+      "loss": 0.6014,
+      "step": 3084
+    },
+    {
+      "epoch": 1.3001370007376962,
+      "grad_norm": 0.3080947995185852,
+      "learning_rate": 3.0801295845728e-06,
+      "loss": 0.6142,
+      "step": 3085
+    },
+    {
+      "epoch": 1.3005585414690695,
+      "grad_norm": 0.2827150523662567,
+      "learning_rate": 3.079040644636134e-06,
+      "loss": 0.5848,
+      "step": 3086
+    },
+    {
+      "epoch": 1.3009800822004427,
+      "grad_norm": 0.3031114637851715,
+      "learning_rate": 3.077951588600162e-06,
+      "loss": 0.5945,
+      "step": 3087
+    },
+    {
+      "epoch": 1.3014016229318157,
+      "grad_norm": 0.2775058150291443,
+      "learning_rate": 3.0768624166832415e-06,
+      "loss": 0.5925,
+      "step": 3088
+    },
+    {
+      "epoch": 1.301823163663189,
+      "grad_norm": 0.2882087528705597,
+      "learning_rate": 3.075773129103756e-06,
+      "loss": 0.5851,
+      "step": 3089
+    },
+    {
+      "epoch": 1.3022447043945622,
+      "grad_norm": 0.29760000109672546,
+      "learning_rate": 3.0746837260801094e-06,
+      "loss": 0.5939,
+      "step": 3090
+    },
+    {
+      "epoch": 1.3026662451259354,
+      "grad_norm": 0.2907105088233948,
+      "learning_rate": 3.073594207830731e-06,
+      "loss": 0.5827,
+      "step": 3091
+    },
+    {
+      "epoch": 1.3030877858573084,
+      "grad_norm": 0.3009587526321411,
+      "learning_rate": 3.072504574574071e-06,
+      "loss": 0.5886,
+      "step": 3092
+    },
+    {
+      "epoch": 1.3035093265886817,
+      "grad_norm": 0.2903907001018524,
+      "learning_rate": 3.071414826528606e-06,
+      "loss": 0.5896,
+      "step": 3093
+    },
+    {
+      "epoch": 1.3039308673200547,
+      "grad_norm": 0.2904217541217804,
+      "learning_rate": 3.0703249639128323e-06,
+      "loss": 0.5857,
+      "step": 3094
+    },
+    {
+      "epoch": 1.304352408051428,
+      "grad_norm": 0.27806368470191956,
+      "learning_rate": 3.069234986945271e-06,
+      "loss": 0.5739,
+      "step": 3095
+    },
+    {
+      "epoch": 1.3047739487828012,
+      "grad_norm": 0.2908913791179657,
+      "learning_rate": 3.068144895844465e-06,
+      "loss": 0.5624,
+      "step": 3096
+    },
+    {
+      "epoch": 1.3051954895141744,
+      "grad_norm": 0.2911220192909241,
+      "learning_rate": 3.0670546908289806e-06,
+      "loss": 0.6225,
+      "step": 3097
+    },
+    {
+      "epoch": 1.3056170302455474,
+      "grad_norm": 0.28194209933280945,
+      "learning_rate": 3.065964372117408e-06,
+      "loss": 0.5786,
+      "step": 3098
+    },
+    {
+      "epoch": 1.3060385709769207,
+      "grad_norm": 0.28787535429000854,
+      "learning_rate": 3.0648739399283584e-06,
+      "loss": 0.6085,
+      "step": 3099
+    },
+    {
+      "epoch": 1.3064601117082937,
+      "grad_norm": 0.28823864459991455,
+      "learning_rate": 3.0637833944804675e-06,
+      "loss": 0.5905,
+      "step": 3100
+    },
+    {
+      "epoch": 1.306881652439667,
+      "grad_norm": 0.30594581365585327,
+      "learning_rate": 3.062692735992391e-06,
+      "loss": 0.5892,
+      "step": 3101
+    },
+    {
+      "epoch": 1.3073031931710402,
+      "grad_norm": 0.2931262254714966,
+      "learning_rate": 3.061601964682811e-06,
+      "loss": 0.6042,
+      "step": 3102
+    },
+    {
+      "epoch": 1.3077247339024134,
+      "grad_norm": 0.288114458322525,
+      "learning_rate": 3.060511080770429e-06,
+      "loss": 0.61,
+      "step": 3103
+    },
+    {
+      "epoch": 1.3081462746337864,
+      "grad_norm": 0.28504639863967896,
+      "learning_rate": 3.059420084473971e-06,
+      "loss": 0.5738,
+      "step": 3104
+    },
+    {
+      "epoch": 1.3085678153651596,
+      "grad_norm": 0.3045184016227722,
+      "learning_rate": 3.0583289760121847e-06,
+      "loss": 0.592,
+      "step": 3105
+    },
+    {
+      "epoch": 1.3089893560965329,
+      "grad_norm": 0.29840025305747986,
+      "learning_rate": 3.05723775560384e-06,
+      "loss": 0.6032,
+      "step": 3106
+    },
+    {
+      "epoch": 1.309410896827906,
+      "grad_norm": 0.2914488613605499,
+      "learning_rate": 3.0561464234677314e-06,
+      "loss": 0.604,
+      "step": 3107
+    },
+    {
+      "epoch": 1.3098324375592791,
+      "grad_norm": 0.29519954323768616,
+      "learning_rate": 3.055054979822672e-06,
+      "loss": 0.564,
+      "step": 3108
+    },
+    {
+      "epoch": 1.3102539782906524,
+      "grad_norm": 0.2988862991333008,
+      "learning_rate": 3.0539634248875006e-06,
+      "loss": 0.5813,
+      "step": 3109
+    },
+    {
+      "epoch": 1.3106755190220256,
+      "grad_norm": 0.2976135313510895,
+      "learning_rate": 3.0528717588810763e-06,
+      "loss": 0.5904,
+      "step": 3110
+    },
+    {
+      "epoch": 1.3110970597533986,
+      "grad_norm": 0.28965333104133606,
+      "learning_rate": 3.0517799820222822e-06,
+      "loss": 0.6083,
+      "step": 3111
+    },
+    {
+      "epoch": 1.3115186004847719,
+      "grad_norm": 0.2968409061431885,
+      "learning_rate": 3.0506880945300216e-06,
+      "loss": 0.5959,
+      "step": 3112
+    },
+    {
+      "epoch": 1.3119401412161449,
+      "grad_norm": 0.3077045977115631,
+      "learning_rate": 3.0495960966232217e-06,
+      "loss": 0.5992,
+      "step": 3113
+    },
+    {
+      "epoch": 1.3123616819475181,
+      "grad_norm": 0.2916930019855499,
+      "learning_rate": 3.048503988520832e-06,
+      "loss": 0.5831,
+      "step": 3114
+    },
+    {
+      "epoch": 1.3127832226788914,
+      "grad_norm": 0.2837620675563812,
+      "learning_rate": 3.0474117704418217e-06,
+      "loss": 0.6091,
+      "step": 3115
+    },
+    {
+      "epoch": 1.3132047634102646,
+      "grad_norm": 0.2942317724227905,
+      "learning_rate": 3.0463194426051847e-06,
+      "loss": 0.6133,
+      "step": 3116
+    },
+    {
+      "epoch": 1.3136263041416376,
+      "grad_norm": 0.305570513010025,
+      "learning_rate": 3.0452270052299352e-06,
+      "loss": 0.5744,
+      "step": 3117
+    },
+    {
+      "epoch": 1.3140478448730109,
+      "grad_norm": 0.30055367946624756,
+      "learning_rate": 3.0441344585351114e-06,
+      "loss": 0.6034,
+      "step": 3118
+    },
+    {
+      "epoch": 1.314469385604384,
+      "grad_norm": 0.2891502380371094,
+      "learning_rate": 3.04304180273977e-06,
+      "loss": 0.5852,
+      "step": 3119
+    },
+    {
+      "epoch": 1.314890926335757,
+      "grad_norm": 0.29979053139686584,
+      "learning_rate": 3.041949038062993e-06,
+      "loss": 0.5639,
+      "step": 3120
+    },
+    {
+      "epoch": 1.3153124670671303,
+      "grad_norm": 0.29319891333580017,
+      "learning_rate": 3.040856164723882e-06,
+      "loss": 0.5856,
+      "step": 3121
+    },
+    {
+      "epoch": 1.3157340077985036,
+      "grad_norm": 0.2964184284210205,
+      "learning_rate": 3.0397631829415624e-06,
+      "loss": 0.5873,
+      "step": 3122
+    },
+    {
+      "epoch": 1.3161555485298768,
+      "grad_norm": 0.30781280994415283,
+      "learning_rate": 3.0386700929351796e-06,
+      "loss": 0.58,
+      "step": 3123
+    },
+    {
+      "epoch": 1.3165770892612498,
+      "grad_norm": 0.2821938991546631,
+      "learning_rate": 3.0375768949239005e-06,
+      "loss": 0.5761,
+      "step": 3124
+    },
+    {
+      "epoch": 1.316998629992623,
+      "grad_norm": 0.28142300248146057,
+      "learning_rate": 3.036483589126916e-06,
+      "loss": 0.6007,
+      "step": 3125
+    },
+    {
+      "epoch": 1.317420170723996,
+      "grad_norm": 0.27912941575050354,
+      "learning_rate": 3.0353901757634357e-06,
+      "loss": 0.5871,
+      "step": 3126
+    },
+    {
+      "epoch": 1.3178417114553693,
+      "grad_norm": 0.300674706697464,
+      "learning_rate": 3.034296655052692e-06,
+      "loss": 0.5955,
+      "step": 3127
+    },
+    {
+      "epoch": 1.3182632521867426,
+      "grad_norm": 0.29524505138397217,
+      "learning_rate": 3.0332030272139407e-06,
+      "loss": 0.5794,
+      "step": 3128
+    },
+    {
+      "epoch": 1.3186847929181158,
+      "grad_norm": 0.28989532589912415,
+      "learning_rate": 3.032109292466455e-06,
+      "loss": 0.5702,
+      "step": 3129
+    },
+    {
+      "epoch": 1.3191063336494888,
+      "grad_norm": 0.3156152069568634,
+      "learning_rate": 3.031015451029534e-06,
+      "loss": 0.5597,
+      "step": 3130
+    },
+    {
+      "epoch": 1.319527874380862,
+      "grad_norm": 0.28968748450279236,
+      "learning_rate": 3.029921503122494e-06,
+      "loss": 0.5892,
+      "step": 3131
+    },
+    {
+      "epoch": 1.3199494151122353,
+      "grad_norm": 0.2941772937774658,
+      "learning_rate": 3.0288274489646762e-06,
+      "loss": 0.5781,
+      "step": 3132
+    },
+    {
+      "epoch": 1.3203709558436083,
+      "grad_norm": 0.29386624693870544,
+      "learning_rate": 3.0277332887754408e-06,
+      "loss": 0.581,
+      "step": 3133
+    },
+    {
+      "epoch": 1.3207924965749815,
+      "grad_norm": 0.29249513149261475,
+      "learning_rate": 3.026639022774171e-06,
+      "loss": 0.5544,
+      "step": 3134
+    },
+    {
+      "epoch": 1.3212140373063548,
+      "grad_norm": 0.29027602076530457,
+      "learning_rate": 3.025544651180269e-06,
+      "loss": 0.5586,
+      "step": 3135
+    },
+    {
+      "epoch": 1.321635578037728,
+      "grad_norm": 0.29468464851379395,
+      "learning_rate": 3.0244501742131595e-06,
+      "loss": 0.5805,
+      "step": 3136
+    },
+    {
+      "epoch": 1.322057118769101,
+      "grad_norm": 0.31095948815345764,
+      "learning_rate": 3.0233555920922897e-06,
+      "loss": 0.6056,
+      "step": 3137
+    },
+    {
+      "epoch": 1.3224786595004743,
+      "grad_norm": 0.29086071252822876,
+      "learning_rate": 3.0222609050371253e-06,
+      "loss": 0.5697,
+      "step": 3138
+    },
+    {
+      "epoch": 1.3229002002318473,
+      "grad_norm": 0.2896486520767212,
+      "learning_rate": 3.0211661132671544e-06,
+      "loss": 0.5767,
+      "step": 3139
+    },
+    {
+      "epoch": 1.3233217409632205,
+      "grad_norm": 0.3025626242160797,
+      "learning_rate": 3.0200712170018853e-06,
+      "loss": 0.6005,
+      "step": 3140
+    },
+    {
+      "epoch": 1.3237432816945938,
+      "grad_norm": 0.29715508222579956,
+      "learning_rate": 3.0189762164608494e-06,
+      "loss": 0.6061,
+      "step": 3141
+    },
+    {
+      "epoch": 1.324164822425967,
+      "grad_norm": 0.28710439801216125,
+      "learning_rate": 3.017881111863596e-06,
+      "loss": 0.6095,
+      "step": 3142
+    },
+    {
+      "epoch": 1.32458636315734,
+      "grad_norm": 0.2858271598815918,
+      "learning_rate": 3.016785903429697e-06,
+      "loss": 0.6231,
+      "step": 3143
+    },
+    {
+      "epoch": 1.3250079038887133,
+      "grad_norm": 0.3013313114643097,
+      "learning_rate": 3.015690591378746e-06,
+      "loss": 0.5918,
+      "step": 3144
+    },
+    {
+      "epoch": 1.3254294446200865,
+      "grad_norm": 0.2894471287727356,
+      "learning_rate": 3.014595175930354e-06,
+      "loss": 0.5993,
+      "step": 3145
+    },
+    {
+      "epoch": 1.3258509853514595,
+      "grad_norm": 0.2933000922203064,
+      "learning_rate": 3.0134996573041576e-06,
+      "loss": 0.5999,
+      "step": 3146
+    },
+    {
+      "epoch": 1.3262725260828327,
+      "grad_norm": 0.29087579250335693,
+      "learning_rate": 3.0124040357198087e-06,
+      "loss": 0.5858,
+      "step": 3147
+    },
+    {
+      "epoch": 1.326694066814206,
+      "grad_norm": 0.2906246483325958,
+      "learning_rate": 3.0113083113969845e-06,
+      "loss": 0.5714,
+      "step": 3148
+    },
+    {
+      "epoch": 1.3271156075455792,
+      "grad_norm": 0.3135024905204773,
+      "learning_rate": 3.01021248455538e-06,
+      "loss": 0.5848,
+      "step": 3149
+    },
+    {
+      "epoch": 1.3275371482769522,
+      "grad_norm": 0.2935161292552948,
+      "learning_rate": 3.0091165554147117e-06,
+      "loss": 0.5622,
+      "step": 3150
+    },
+    {
+      "epoch": 1.3279586890083255,
+      "grad_norm": 0.2927139401435852,
+      "learning_rate": 3.0080205241947165e-06,
+      "loss": 0.6042,
+      "step": 3151
+    },
+    {
+      "epoch": 1.3283802297396985,
+      "grad_norm": 0.2967769205570221,
+      "learning_rate": 3.006924391115152e-06,
+      "loss": 0.5891,
+      "step": 3152
+    },
+    {
+      "epoch": 1.3288017704710717,
+      "grad_norm": 0.3088895082473755,
+      "learning_rate": 3.005828156395796e-06,
+      "loss": 0.5605,
+      "step": 3153
+    },
+    {
+      "epoch": 1.329223311202445,
+      "grad_norm": 0.2915705740451813,
+      "learning_rate": 3.004731820256447e-06,
+      "loss": 0.5866,
+      "step": 3154
+    },
+    {
+      "epoch": 1.3296448519338182,
+      "grad_norm": 0.3022144138813019,
+      "learning_rate": 3.0036353829169235e-06,
+      "loss": 0.6108,
+      "step": 3155
+    },
+    {
+      "epoch": 1.3300663926651912,
+      "grad_norm": 0.29434481263160706,
+      "learning_rate": 3.0025388445970633e-06,
+      "loss": 0.6148,
+      "step": 3156
+    },
+    {
+      "epoch": 1.3304879333965645,
+      "grad_norm": 0.2921946048736572,
+      "learning_rate": 3.001442205516727e-06,
+      "loss": 0.6168,
+      "step": 3157
+    },
+    {
+      "epoch": 1.3309094741279375,
+      "grad_norm": 0.29331886768341064,
+      "learning_rate": 3.000345465895792e-06,
+      "loss": 0.6083,
+      "step": 3158
+    },
+    {
+      "epoch": 1.3313310148593107,
+      "grad_norm": 0.2883720397949219,
+      "learning_rate": 2.999248625954159e-06,
+      "loss": 0.6083,
+      "step": 3159
+    },
+    {
+      "epoch": 1.331752555590684,
+      "grad_norm": 0.29853686690330505,
+      "learning_rate": 2.998151685911748e-06,
+      "loss": 0.5995,
+      "step": 3160
+    },
+    {
+      "epoch": 1.3321740963220572,
+      "grad_norm": 0.2934533655643463,
+      "learning_rate": 2.9970546459884974e-06,
+      "loss": 0.5833,
+      "step": 3161
+    },
+    {
+      "epoch": 1.3325956370534304,
+      "grad_norm": 0.286330908536911,
+      "learning_rate": 2.9959575064043676e-06,
+      "loss": 0.5753,
+      "step": 3162
+    },
+    {
+      "epoch": 1.3330171777848034,
+      "grad_norm": 0.2999122738838196,
+      "learning_rate": 2.994860267379338e-06,
+      "loss": 0.5938,
+      "step": 3163
+    },
+    {
+      "epoch": 1.3334387185161767,
+      "grad_norm": 0.2805075943470001,
+      "learning_rate": 2.993762929133409e-06,
+      "loss": 0.6158,
+      "step": 3164
+    },
+    {
+      "epoch": 1.3338602592475497,
+      "grad_norm": 0.2918041944503784,
+      "learning_rate": 2.9926654918865983e-06,
+      "loss": 0.5796,
+      "step": 3165
+    },
+    {
+      "epoch": 1.334281799978923,
+      "grad_norm": 0.31967973709106445,
+      "learning_rate": 2.9915679558589467e-06,
+      "loss": 0.601,
+      "step": 3166
+    },
+    {
+      "epoch": 1.3347033407102962,
+      "grad_norm": 0.29790499806404114,
+      "learning_rate": 2.990470321270513e-06,
+      "loss": 0.5969,
+      "step": 3167
+    },
+    {
+      "epoch": 1.3351248814416694,
+      "grad_norm": 0.2858971655368805,
+      "learning_rate": 2.989372588341375e-06,
+      "loss": 0.5679,
+      "step": 3168
+    },
+    {
+      "epoch": 1.3355464221730424,
+      "grad_norm": 0.3054177463054657,
+      "learning_rate": 2.988274757291633e-06,
+      "loss": 0.5816,
+      "step": 3169
+    },
+    {
+      "epoch": 1.3359679629044157,
+      "grad_norm": 0.30209848284721375,
+      "learning_rate": 2.9871768283414036e-06,
+      "loss": 0.5705,
+      "step": 3170
+    },
+    {
+      "epoch": 1.3363895036357887,
+      "grad_norm": 0.2943930923938751,
+      "learning_rate": 2.986078801710827e-06,
+      "loss": 0.6064,
+      "step": 3171
+    },
+    {
+      "epoch": 1.336811044367162,
+      "grad_norm": 0.31612086296081543,
+      "learning_rate": 2.9849806776200572e-06,
+      "loss": 0.5745,
+      "step": 3172
+    },
+    {
+      "epoch": 1.3372325850985352,
+      "grad_norm": 0.29537567496299744,
+      "learning_rate": 2.9838824562892747e-06,
+      "loss": 0.6094,
+      "step": 3173
+    },
+    {
+      "epoch": 1.3376541258299084,
+      "grad_norm": 0.3031534254550934,
+      "learning_rate": 2.982784137938674e-06,
+      "loss": 0.5672,
+      "step": 3174
+    },
+    {
+      "epoch": 1.3380756665612814,
+      "grad_norm": 0.2966405153274536,
+      "learning_rate": 2.9816857227884723e-06,
+      "loss": 0.5702,
+      "step": 3175
+    },
+    {
+      "epoch": 1.3384972072926546,
+      "grad_norm": 0.30172649025917053,
+      "learning_rate": 2.980587211058904e-06,
+      "loss": 0.5893,
+      "step": 3176
+    },
+    {
+      "epoch": 1.3389187480240279,
+      "grad_norm": 0.30330729484558105,
+      "learning_rate": 2.9794886029702237e-06,
+      "loss": 0.589,
+      "step": 3177
+    },
+    {
+      "epoch": 1.339340288755401,
+      "grad_norm": 0.29037579894065857,
+      "learning_rate": 2.978389898742706e-06,
+      "loss": 0.6148,
+      "step": 3178
+    },
+    {
+      "epoch": 1.3397618294867741,
+      "grad_norm": 0.2832915484905243,
+      "learning_rate": 2.9772910985966445e-06,
+      "loss": 0.5904,
+      "step": 3179
+    },
+    {
+      "epoch": 1.3401833702181474,
+      "grad_norm": 0.29548880457878113,
+      "learning_rate": 2.9761922027523515e-06,
+      "loss": 0.5863,
+      "step": 3180
+    },
+    {
+      "epoch": 1.3406049109495206,
+      "grad_norm": 0.3028402328491211,
+      "learning_rate": 2.9750932114301582e-06,
+      "loss": 0.5896,
+      "step": 3181
+    },
+    {
+      "epoch": 1.3410264516808936,
+      "grad_norm": 0.29047226905822754,
+      "learning_rate": 2.973994124850416e-06,
+      "loss": 0.5855,
+      "step": 3182
+    },
+    {
+      "epoch": 1.3414479924122669,
+      "grad_norm": 0.2862197160720825,
+      "learning_rate": 2.972894943233496e-06,
+      "loss": 0.5958,
+      "step": 3183
+    },
+    {
+      "epoch": 1.3418695331436399,
+      "grad_norm": 0.2949672043323517,
+      "learning_rate": 2.9717956667997857e-06,
+      "loss": 0.5795,
+      "step": 3184
+    },
+    {
+      "epoch": 1.3422910738750131,
+      "grad_norm": 0.28680792450904846,
+      "learning_rate": 2.9706962957696933e-06,
+      "loss": 0.5668,
+      "step": 3185
+    },
+    {
+      "epoch": 1.3427126146063864,
+      "grad_norm": 0.29078683257102966,
+      "learning_rate": 2.969596830363646e-06,
+      "loss": 0.5933,
+      "step": 3186
+    },
+    {
+      "epoch": 1.3431341553377596,
+      "grad_norm": 0.29687416553497314,
+      "learning_rate": 2.9684972708020907e-06,
+      "loss": 0.5682,
+      "step": 3187
+    },
+    {
+      "epoch": 1.3435556960691326,
+      "grad_norm": 0.2884622812271118,
+      "learning_rate": 2.9673976173054912e-06,
+      "loss": 0.6053,
+      "step": 3188
+    },
+    {
+      "epoch": 1.3439772368005058,
+      "grad_norm": 0.2925822138786316,
+      "learning_rate": 2.966297870094331e-06,
+      "loss": 0.6003,
+      "step": 3189
+    },
+    {
+      "epoch": 1.344398777531879,
+      "grad_norm": 0.3059365153312683,
+      "learning_rate": 2.9651980293891138e-06,
+      "loss": 0.5739,
+      "step": 3190
+    },
+    {
+      "epoch": 1.344820318263252,
+      "grad_norm": 0.28787434101104736,
+      "learning_rate": 2.9640980954103588e-06,
+      "loss": 0.5697,
+      "step": 3191
+    },
+    {
+      "epoch": 1.3452418589946253,
+      "grad_norm": 0.301727831363678,
+      "learning_rate": 2.9629980683786085e-06,
+      "loss": 0.5846,
+      "step": 3192
+    },
+    {
+      "epoch": 1.3456633997259986,
+      "grad_norm": 0.28149595856666565,
+      "learning_rate": 2.9618979485144193e-06,
+      "loss": 0.5974,
+      "step": 3193
+    },
+    {
+      "epoch": 1.3460849404573718,
+      "grad_norm": 0.30097371339797974,
+      "learning_rate": 2.96079773603837e-06,
+      "loss": 0.5798,
+      "step": 3194
+    },
+    {
+      "epoch": 1.3465064811887448,
+      "grad_norm": 0.29800325632095337,
+      "learning_rate": 2.959697431171055e-06,
+      "loss": 0.6039,
+      "step": 3195
+    },
+    {
+      "epoch": 1.346928021920118,
+      "grad_norm": 0.28295162320137024,
+      "learning_rate": 2.9585970341330896e-06,
+      "loss": 0.5825,
+      "step": 3196
+    },
+    {
+      "epoch": 1.347349562651491,
+      "grad_norm": 0.29735344648361206,
+      "learning_rate": 2.9574965451451053e-06,
+      "loss": 0.5863,
+      "step": 3197
+    },
+    {
+      "epoch": 1.3477711033828643,
+      "grad_norm": 0.28632697463035583,
+      "learning_rate": 2.9563959644277546e-06,
+      "loss": 0.6005,
+      "step": 3198
+    },
+    {
+      "epoch": 1.3481926441142376,
+      "grad_norm": 0.30630141496658325,
+      "learning_rate": 2.9552952922017063e-06,
+      "loss": 0.5927,
+      "step": 3199
+    },
+    {
+      "epoch": 1.3486141848456108,
+      "grad_norm": 0.29010358452796936,
+      "learning_rate": 2.9541945286876485e-06,
+      "loss": 0.5924,
+      "step": 3200
+    },
+    {
+      "epoch": 1.3490357255769838,
+      "grad_norm": 0.3123135566711426,
+      "learning_rate": 2.9530936741062877e-06,
+      "loss": 0.59,
+      "step": 3201
+    },
+    {
+      "epoch": 1.349457266308357,
+      "grad_norm": 0.2904750108718872,
+      "learning_rate": 2.9519927286783486e-06,
+      "loss": 0.5952,
+      "step": 3202
+    },
+    {
+      "epoch": 1.3498788070397303,
+      "grad_norm": 0.3003297448158264,
+      "learning_rate": 2.9508916926245732e-06,
+      "loss": 0.5913,
+      "step": 3203
+    },
+    {
+      "epoch": 1.3503003477711033,
+      "grad_norm": 0.2907561957836151,
+      "learning_rate": 2.949790566165723e-06,
+      "loss": 0.5781,
+      "step": 3204
+    },
+    {
+      "epoch": 1.3507218885024765,
+      "grad_norm": 0.2939736247062683,
+      "learning_rate": 2.948689349522576e-06,
+      "loss": 0.5795,
+      "step": 3205
+    },
+    {
+      "epoch": 1.3511434292338498,
+      "grad_norm": 0.29564517736434937,
+      "learning_rate": 2.947588042915931e-06,
+      "loss": 0.5726,
+      "step": 3206
+    },
+    {
+      "epoch": 1.351564969965223,
+      "grad_norm": 0.2909988760948181,
+      "learning_rate": 2.9464866465666016e-06,
+      "loss": 0.5812,
+      "step": 3207
+    },
+    {
+      "epoch": 1.351986510696596,
+      "grad_norm": 0.29635369777679443,
+      "learning_rate": 2.9453851606954214e-06,
+      "loss": 0.5658,
+      "step": 3208
+    },
+    {
+      "epoch": 1.3524080514279693,
+      "grad_norm": 0.30680081248283386,
+      "learning_rate": 2.944283585523241e-06,
+      "loss": 0.6061,
+      "step": 3209
+    },
+    {
+      "epoch": 1.3528295921593423,
+      "grad_norm": 0.29627856612205505,
+      "learning_rate": 2.943181921270931e-06,
+      "loss": 0.5961,
+      "step": 3210
+    },
+    {
+      "epoch": 1.3532511328907155,
+      "grad_norm": 0.2780596613883972,
+      "learning_rate": 2.9420801681593763e-06,
+      "loss": 0.589,
+      "step": 3211
+    },
+    {
+      "epoch": 1.3536726736220888,
+      "grad_norm": 0.2940075099468231,
+      "learning_rate": 2.940978326409483e-06,
+      "loss": 0.5879,
+      "step": 3212
+    },
+    {
+      "epoch": 1.354094214353462,
+      "grad_norm": 0.29517731070518494,
+      "learning_rate": 2.939876396242173e-06,
+      "loss": 0.5916,
+      "step": 3213
+    },
+    {
+      "epoch": 1.354515755084835,
+      "grad_norm": 0.2923842668533325,
+      "learning_rate": 2.9387743778783857e-06,
+      "loss": 0.602,
+      "step": 3214
+    },
+    {
+      "epoch": 1.3549372958162083,
+      "grad_norm": 0.2976822555065155,
+      "learning_rate": 2.9376722715390803e-06,
+      "loss": 0.5712,
+      "step": 3215
+    },
+    {
+      "epoch": 1.3553588365475815,
+      "grad_norm": 0.293218195438385,
+      "learning_rate": 2.936570077445231e-06,
+      "loss": 0.5861,
+      "step": 3216
+    },
+    {
+      "epoch": 1.3557803772789545,
+      "grad_norm": 0.2899850606918335,
+      "learning_rate": 2.935467795817832e-06,
+      "loss": 0.5788,
+      "step": 3217
+    },
+    {
+      "epoch": 1.3562019180103277,
+      "grad_norm": 0.2954835295677185,
+      "learning_rate": 2.9343654268778933e-06,
+      "loss": 0.587,
+      "step": 3218
+    },
+    {
+      "epoch": 1.356623458741701,
+      "grad_norm": 0.3063293993473053,
+      "learning_rate": 2.9332629708464422e-06,
+      "loss": 0.5748,
+      "step": 3219
+    },
+    {
+      "epoch": 1.3570449994730742,
+      "grad_norm": 0.3030092716217041,
+      "learning_rate": 2.9321604279445258e-06,
+      "loss": 0.5797,
+      "step": 3220
+    },
+    {
+      "epoch": 1.3574665402044472,
+      "grad_norm": 0.2969261407852173,
+      "learning_rate": 2.9310577983932065e-06,
+      "loss": 0.5802,
+      "step": 3221
+    },
+    {
+      "epoch": 1.3578880809358205,
+      "grad_norm": 0.33054599165916443,
+      "learning_rate": 2.9299550824135653e-06,
+      "loss": 0.5929,
+      "step": 3222
+    },
+    {
+      "epoch": 1.3583096216671935,
+      "grad_norm": 0.29640910029411316,
+      "learning_rate": 2.9288522802266983e-06,
+      "loss": 0.551,
+      "step": 3223
+    },
+    {
+      "epoch": 1.3587311623985667,
+      "grad_norm": 0.28479498624801636,
+      "learning_rate": 2.9277493920537226e-06,
+      "loss": 0.5547,
+      "step": 3224
+    },
+    {
+      "epoch": 1.35915270312994,
+      "grad_norm": 0.3212515115737915,
+      "learning_rate": 2.9266464181157683e-06,
+      "loss": 0.5923,
+      "step": 3225
+    },
+    {
+      "epoch": 1.3595742438613132,
+      "grad_norm": 0.29704737663269043,
+      "learning_rate": 2.9255433586339865e-06,
+      "loss": 0.6,
+      "step": 3226
+    },
+    {
+      "epoch": 1.3599957845926862,
+      "grad_norm": 0.3032103478908539,
+      "learning_rate": 2.924440213829543e-06,
+      "loss": 0.5977,
+      "step": 3227
+    },
+    {
+      "epoch": 1.3604173253240595,
+      "grad_norm": 0.2971370816230774,
+      "learning_rate": 2.9233369839236213e-06,
+      "loss": 0.5798,
+      "step": 3228
+    },
+    {
+      "epoch": 1.3608388660554325,
+      "grad_norm": 0.30755865573883057,
+      "learning_rate": 2.9222336691374224e-06,
+      "loss": 0.5735,
+      "step": 3229
+    },
+    {
+      "epoch": 1.3612604067868057,
+      "grad_norm": 0.31366461515426636,
+      "learning_rate": 2.9211302696921647e-06,
+      "loss": 0.5926,
+      "step": 3230
+    },
+    {
+      "epoch": 1.361681947518179,
+      "grad_norm": 0.29211634397506714,
+      "learning_rate": 2.9200267858090825e-06,
+      "loss": 0.5998,
+      "step": 3231
+    },
+    {
+      "epoch": 1.3621034882495522,
+      "grad_norm": 0.295075923204422,
+      "learning_rate": 2.918923217709427e-06,
+      "loss": 0.5944,
+      "step": 3232
+    },
+    {
+      "epoch": 1.3625250289809252,
+      "grad_norm": 0.3224492073059082,
+      "learning_rate": 2.9178195656144677e-06,
+      "loss": 0.5994,
+      "step": 3233
+    },
+    {
+      "epoch": 1.3629465697122984,
+      "grad_norm": 0.3001447021961212,
+      "learning_rate": 2.916715829745488e-06,
+      "loss": 0.6108,
+      "step": 3234
+    },
+    {
+      "epoch": 1.3633681104436717,
+      "grad_norm": 0.3148466944694519,
+      "learning_rate": 2.915612010323793e-06,
+      "loss": 0.5872,
+      "step": 3235
+    },
+    {
+      "epoch": 1.3637896511750447,
+      "grad_norm": 0.30204513669013977,
+      "learning_rate": 2.9145081075707e-06,
+      "loss": 0.582,
+      "step": 3236
+    },
+    {
+      "epoch": 1.364211191906418,
+      "grad_norm": 0.3093981444835663,
+      "learning_rate": 2.9134041217075444e-06,
+      "loss": 0.5806,
+      "step": 3237
+    },
+    {
+      "epoch": 1.3646327326377912,
+      "grad_norm": 0.2945973873138428,
+      "learning_rate": 2.9123000529556796e-06,
+      "loss": 0.5878,
+      "step": 3238
+    },
+    {
+      "epoch": 1.3650542733691644,
+      "grad_norm": 0.3033713102340698,
+      "learning_rate": 2.9111959015364726e-06,
+      "loss": 0.5877,
+      "step": 3239
+    },
+    {
+      "epoch": 1.3654758141005374,
+      "grad_norm": 0.29255878925323486,
+      "learning_rate": 2.910091667671312e-06,
+      "loss": 0.5792,
+      "step": 3240
+    },
+    {
+      "epoch": 1.3658973548319107,
+      "grad_norm": 0.29490023851394653,
+      "learning_rate": 2.908987351581597e-06,
+      "loss": 0.606,
+      "step": 3241
+    },
+    {
+      "epoch": 1.3663188955632837,
+      "grad_norm": 0.2923828065395355,
+      "learning_rate": 2.907882953488748e-06,
+      "loss": 0.6092,
+      "step": 3242
+    },
+    {
+      "epoch": 1.366740436294657,
+      "grad_norm": 0.289171427488327,
+      "learning_rate": 2.9067784736141984e-06,
+      "loss": 0.5674,
+      "step": 3243
+    },
+    {
+      "epoch": 1.3671619770260302,
+      "grad_norm": 0.31211209297180176,
+      "learning_rate": 2.905673912179401e-06,
+      "loss": 0.5723,
+      "step": 3244
+    },
+    {
+      "epoch": 1.3675835177574034,
+      "grad_norm": 0.30536699295043945,
+      "learning_rate": 2.904569269405823e-06,
+      "loss": 0.5784,
+      "step": 3245
+    },
+    {
+      "epoch": 1.3680050584887764,
+      "grad_norm": 0.29537567496299744,
+      "learning_rate": 2.903464545514948e-06,
+      "loss": 0.5814,
+      "step": 3246
+    },
+    {
+      "epoch": 1.3684265992201496,
+      "grad_norm": 0.3208700716495514,
+      "learning_rate": 2.9023597407282772e-06,
+      "loss": 0.5986,
+      "step": 3247
+    },
+    {
+      "epoch": 1.3688481399515229,
+      "grad_norm": 0.31728222966194153,
+      "learning_rate": 2.901254855267326e-06,
+      "loss": 0.5767,
+      "step": 3248
+    },
+    {
+      "epoch": 1.369269680682896,
+      "grad_norm": 0.29879680275917053,
+      "learning_rate": 2.900149889353628e-06,
+      "loss": 0.5907,
+      "step": 3249
+    },
+    {
+      "epoch": 1.3696912214142691,
+      "grad_norm": 0.2861499488353729,
+      "learning_rate": 2.8990448432087308e-06,
+      "loss": 0.6041,
+      "step": 3250
+    },
+    {
+      "epoch": 1.3701127621456424,
+      "grad_norm": 0.2829798758029938,
+      "learning_rate": 2.897939717054202e-06,
+      "loss": 0.596,
+      "step": 3251
+    },
+    {
+      "epoch": 1.3705343028770156,
+      "grad_norm": 0.2924060821533203,
+      "learning_rate": 2.8968345111116193e-06,
+      "loss": 0.6134,
+      "step": 3252
+    },
+    {
+      "epoch": 1.3709558436083886,
+      "grad_norm": 0.28703057765960693,
+      "learning_rate": 2.8957292256025815e-06,
+      "loss": 0.5846,
+      "step": 3253
+    },
+    {
+      "epoch": 1.3713773843397619,
+      "grad_norm": 0.29625463485717773,
+      "learning_rate": 2.894623860748702e-06,
+      "loss": 0.55,
+      "step": 3254
+    },
+    {
+      "epoch": 1.3717989250711349,
+      "grad_norm": 0.30700868368148804,
+      "learning_rate": 2.893518416771608e-06,
+      "loss": 0.5761,
+      "step": 3255
+    },
+    {
+      "epoch": 1.3722204658025081,
+      "grad_norm": 0.2919626832008362,
+      "learning_rate": 2.8924128938929452e-06,
+      "loss": 0.6011,
+      "step": 3256
+    },
+    {
+      "epoch": 1.3726420065338814,
+      "grad_norm": 0.2915749251842499,
+      "learning_rate": 2.891307292334374e-06,
+      "loss": 0.5826,
+      "step": 3257
+    },
+    {
+      "epoch": 1.3730635472652546,
+      "grad_norm": 0.2913828492164612,
+      "learning_rate": 2.89020161231757e-06,
+      "loss": 0.5893,
+      "step": 3258
+    },
+    {
+      "epoch": 1.3734850879966276,
+      "grad_norm": 0.3077673017978668,
+      "learning_rate": 2.8890958540642254e-06,
+      "loss": 0.5977,
+      "step": 3259
+    },
+    {
+      "epoch": 1.3739066287280008,
+      "grad_norm": 0.3018398582935333,
+      "learning_rate": 2.887990017796049e-06,
+      "loss": 0.5659,
+      "step": 3260
+    },
+    {
+      "epoch": 1.374328169459374,
+      "grad_norm": 0.31572481989860535,
+      "learning_rate": 2.886884103734764e-06,
+      "loss": 0.5678,
+      "step": 3261
+    },
+    {
+      "epoch": 1.374749710190747,
+      "grad_norm": 0.2937295734882355,
+      "learning_rate": 2.885778112102108e-06,
+      "loss": 0.5821,
+      "step": 3262
+    },
+    {
+      "epoch": 1.3751712509221203,
+      "grad_norm": 0.30527257919311523,
+      "learning_rate": 2.884672043119836e-06,
+      "loss": 0.58,
+      "step": 3263
+    },
+    {
+      "epoch": 1.3755927916534936,
+      "grad_norm": 0.29499003291130066,
+      "learning_rate": 2.8835658970097185e-06,
+      "loss": 0.6036,
+      "step": 3264
+    },
+    {
+      "epoch": 1.3760143323848668,
+      "grad_norm": 0.31405261158943176,
+      "learning_rate": 2.8824596739935405e-06,
+      "loss": 0.5776,
+      "step": 3265
+    },
+    {
+      "epoch": 1.3764358731162398,
+      "grad_norm": 0.309256374835968,
+      "learning_rate": 2.881353374293103e-06,
+      "loss": 0.5874,
+      "step": 3266
+    },
+    {
+      "epoch": 1.376857413847613,
+      "grad_norm": 0.2924270033836365,
+      "learning_rate": 2.880246998130223e-06,
+      "loss": 0.5834,
+      "step": 3267
+    },
+    {
+      "epoch": 1.377278954578986,
+      "grad_norm": 0.29045042395591736,
+      "learning_rate": 2.8791405457267307e-06,
+      "loss": 0.6076,
+      "step": 3268
+    },
+    {
+      "epoch": 1.3777004953103593,
+      "grad_norm": 0.2996369004249573,
+      "learning_rate": 2.878034017304473e-06,
+      "loss": 0.6053,
+      "step": 3269
+    },
+    {
+      "epoch": 1.3781220360417326,
+      "grad_norm": 0.3043166697025299,
+      "learning_rate": 2.876927413085314e-06,
+      "loss": 0.5898,
+      "step": 3270
+    },
+    {
+      "epoch": 1.3785435767731058,
+      "grad_norm": 0.28621184825897217,
+      "learning_rate": 2.8758207332911292e-06,
+      "loss": 0.581,
+      "step": 3271
+    },
+    {
+      "epoch": 1.3789651175044788,
+      "grad_norm": 0.28504282236099243,
+      "learning_rate": 2.8747139781438117e-06,
+      "loss": 0.5798,
+      "step": 3272
+    },
+    {
+      "epoch": 1.379386658235852,
+      "grad_norm": 0.3045307695865631,
+      "learning_rate": 2.8736071478652684e-06,
+      "loss": 0.5903,
+      "step": 3273
+    },
+    {
+      "epoch": 1.3798081989672253,
+      "grad_norm": 0.2909851372241974,
+      "learning_rate": 2.872500242677423e-06,
+      "loss": 0.6107,
+      "step": 3274
+    },
+    {
+      "epoch": 1.3802297396985983,
+      "grad_norm": 0.32000166177749634,
+      "learning_rate": 2.8713932628022123e-06,
+      "loss": 0.5724,
+      "step": 3275
+    },
+    {
+      "epoch": 1.3806512804299715,
+      "grad_norm": 0.293891578912735,
+      "learning_rate": 2.8702862084615896e-06,
+      "loss": 0.6027,
+      "step": 3276
+    },
+    {
+      "epoch": 1.3810728211613448,
+      "grad_norm": 0.29472848773002625,
+      "learning_rate": 2.869179079877522e-06,
+      "loss": 0.564,
+      "step": 3277
+    },
+    {
+      "epoch": 1.381494361892718,
+      "grad_norm": 0.3020780384540558,
+      "learning_rate": 2.8680718772719917e-06,
+      "loss": 0.5945,
+      "step": 3278
+    },
+    {
+      "epoch": 1.381915902624091,
+      "grad_norm": 0.2836039364337921,
+      "learning_rate": 2.8669646008669973e-06,
+      "loss": 0.6107,
+      "step": 3279
+    },
+    {
+      "epoch": 1.3823374433554643,
+      "grad_norm": 0.3008824288845062,
+      "learning_rate": 2.865857250884549e-06,
+      "loss": 0.5733,
+      "step": 3280
+    },
+    {
+      "epoch": 1.3827589840868373,
+      "grad_norm": 0.3005940616130829,
+      "learning_rate": 2.864749827546675e-06,
+      "loss": 0.5871,
+      "step": 3281
+    },
+    {
+      "epoch": 1.3831805248182105,
+      "grad_norm": 0.3132604956626892,
+      "learning_rate": 2.8636423310754163e-06,
+      "loss": 0.5946,
+      "step": 3282
+    },
+    {
+      "epoch": 1.3836020655495838,
+      "grad_norm": 0.28387653827667236,
+      "learning_rate": 2.862534761692829e-06,
+      "loss": 0.5851,
+      "step": 3283
+    },
+    {
+      "epoch": 1.384023606280957,
+      "grad_norm": 0.29056882858276367,
+      "learning_rate": 2.8614271196209853e-06,
+      "loss": 0.6188,
+      "step": 3284
+    },
+    {
+      "epoch": 1.38444514701233,
+      "grad_norm": 0.3029514253139496,
+      "learning_rate": 2.860319405081968e-06,
+      "loss": 0.5874,
+      "step": 3285
+    },
+    {
+      "epoch": 1.3848666877437033,
+      "grad_norm": 0.3067701458930969,
+      "learning_rate": 2.85921161829788e-06,
+      "loss": 0.5831,
+      "step": 3286
+    },
+    {
+      "epoch": 1.3852882284750763,
+      "grad_norm": 0.2888418436050415,
+      "learning_rate": 2.8581037594908336e-06,
+      "loss": 0.613,
+      "step": 3287
+    },
+    {
+      "epoch": 1.3857097692064495,
+      "grad_norm": 0.29466885328292847,
+      "learning_rate": 2.856995828882958e-06,
+      "loss": 0.5624,
+      "step": 3288
+    },
+    {
+      "epoch": 1.3861313099378227,
+      "grad_norm": 0.3002772033214569,
+      "learning_rate": 2.8558878266963968e-06,
+      "loss": 0.579,
+      "step": 3289
+    },
+    {
+      "epoch": 1.386552850669196,
+      "grad_norm": 0.28085362911224365,
+      "learning_rate": 2.854779753153308e-06,
+      "loss": 0.5807,
+      "step": 3290
+    },
+    {
+      "epoch": 1.3869743914005692,
+      "grad_norm": 0.30366602540016174,
+      "learning_rate": 2.853671608475863e-06,
+      "loss": 0.5846,
+      "step": 3291
+    },
+    {
+      "epoch": 1.3873959321319422,
+      "grad_norm": 0.30100739002227783,
+      "learning_rate": 2.8525633928862482e-06,
+      "loss": 0.5582,
+      "step": 3292
+    },
+    {
+      "epoch": 1.3878174728633155,
+      "grad_norm": 0.28741583228111267,
+      "learning_rate": 2.851455106606664e-06,
+      "loss": 0.592,
+      "step": 3293
+    },
+    {
+      "epoch": 1.3882390135946885,
+      "grad_norm": 0.2952061593532562,
+      "learning_rate": 2.8503467498593248e-06,
+      "loss": 0.578,
+      "step": 3294
+    },
+    {
+      "epoch": 1.3886605543260617,
+      "grad_norm": 0.30367347598075867,
+      "learning_rate": 2.8492383228664593e-06,
+      "loss": 0.6067,
+      "step": 3295
+    },
+    {
+      "epoch": 1.389082095057435,
+      "grad_norm": 0.28744083642959595,
+      "learning_rate": 2.8481298258503097e-06,
+      "loss": 0.5839,
+      "step": 3296
+    },
+    {
+      "epoch": 1.3895036357888082,
+      "grad_norm": 0.28898563981056213,
+      "learning_rate": 2.8470212590331344e-06,
+      "loss": 0.5555,
+      "step": 3297
+    },
+    {
+      "epoch": 1.3899251765201812,
+      "grad_norm": 0.28731560707092285,
+      "learning_rate": 2.8459126226372025e-06,
+      "loss": 0.6015,
+      "step": 3298
+    },
+    {
+      "epoch": 1.3903467172515545,
+      "grad_norm": 0.30630233883857727,
+      "learning_rate": 2.8448039168847997e-06,
+      "loss": 0.5783,
+      "step": 3299
+    },
+    {
+      "epoch": 1.3907682579829275,
+      "grad_norm": 0.2931341528892517,
+      "learning_rate": 2.843695141998225e-06,
+      "loss": 0.5866,
+      "step": 3300
+    },
+    {
+      "epoch": 1.3911897987143007,
+      "grad_norm": 0.3029172420501709,
+      "learning_rate": 2.8425862981997903e-06,
+      "loss": 0.5901,
+      "step": 3301
+    },
+    {
+      "epoch": 1.391611339445674,
+      "grad_norm": 0.2813885807991028,
+      "learning_rate": 2.8414773857118226e-06,
+      "loss": 0.5918,
+      "step": 3302
+    },
+    {
+      "epoch": 1.3920328801770472,
+      "grad_norm": 0.30794450640678406,
+      "learning_rate": 2.840368404756661e-06,
+      "loss": 0.5972,
+      "step": 3303
+    },
+    {
+      "epoch": 1.3924544209084202,
+      "grad_norm": 0.31940972805023193,
+      "learning_rate": 2.83925935555666e-06,
+      "loss": 0.5702,
+      "step": 3304
+    },
+    {
+      "epoch": 1.3928759616397934,
+      "grad_norm": 0.2903609871864319,
+      "learning_rate": 2.838150238334187e-06,
+      "loss": 0.5729,
+      "step": 3305
+    },
+    {
+      "epoch": 1.3932975023711667,
+      "grad_norm": 0.2887631952762604,
+      "learning_rate": 2.8370410533116233e-06,
+      "loss": 0.5843,
+      "step": 3306
+    },
+    {
+      "epoch": 1.3937190431025397,
+      "grad_norm": 0.29673609137535095,
+      "learning_rate": 2.8359318007113632e-06,
+      "loss": 0.6032,
+      "step": 3307
+    },
+    {
+      "epoch": 1.394140583833913,
+      "grad_norm": 0.2946152985095978,
+      "learning_rate": 2.8348224807558157e-06,
+      "loss": 0.5974,
+      "step": 3308
+    },
+    {
+      "epoch": 1.3945621245652862,
+      "grad_norm": 0.3950257897377014,
+      "learning_rate": 2.8337130936674027e-06,
+      "loss": 0.5894,
+      "step": 3309
+    },
+    {
+      "epoch": 1.3949836652966594,
+      "grad_norm": 0.2880473732948303,
+      "learning_rate": 2.8326036396685584e-06,
+      "loss": 0.5732,
+      "step": 3310
+    },
+    {
+      "epoch": 1.3954052060280324,
+      "grad_norm": 0.34103360772132874,
+      "learning_rate": 2.8314941189817336e-06,
+      "loss": 0.6033,
+      "step": 3311
+    },
+    {
+      "epoch": 1.3958267467594057,
+      "grad_norm": 0.3171156346797943,
+      "learning_rate": 2.8303845318293875e-06,
+      "loss": 0.58,
+      "step": 3312
+    },
+    {
+      "epoch": 1.3962482874907787,
+      "grad_norm": 0.28944987058639526,
+      "learning_rate": 2.8292748784339975e-06,
+      "loss": 0.5988,
+      "step": 3313
+    },
+    {
+      "epoch": 1.396669828222152,
+      "grad_norm": 0.29095104336738586,
+      "learning_rate": 2.8281651590180526e-06,
+      "loss": 0.6176,
+      "step": 3314
+    },
+    {
+      "epoch": 1.3970913689535251,
+      "grad_norm": 0.2921895384788513,
+      "learning_rate": 2.827055373804053e-06,
+      "loss": 0.5649,
+      "step": 3315
+    },
+    {
+      "epoch": 1.3975129096848984,
+      "grad_norm": 0.4189492464065552,
+      "learning_rate": 2.8259455230145156e-06,
+      "loss": 0.5901,
+      "step": 3316
+    },
+    {
+      "epoch": 1.3979344504162714,
+      "grad_norm": 0.2776779234409332,
+      "learning_rate": 2.824835606871967e-06,
+      "loss": 0.5797,
+      "step": 3317
+    },
+    {
+      "epoch": 1.3983559911476446,
+      "grad_norm": 0.3262343406677246,
+      "learning_rate": 2.823725625598949e-06,
+      "loss": 0.5943,
+      "step": 3318
+    },
+    {
+      "epoch": 1.3987775318790179,
+      "grad_norm": 0.3268984854221344,
+      "learning_rate": 2.8226155794180173e-06,
+      "loss": 0.5822,
+      "step": 3319
+    },
+    {
+      "epoch": 1.399199072610391,
+      "grad_norm": 0.3165684938430786,
+      "learning_rate": 2.8215054685517377e-06,
+      "loss": 0.5611,
+      "step": 3320
+    },
+    {
+      "epoch": 1.3996206133417641,
+      "grad_norm": 0.28046083450317383,
+      "learning_rate": 2.820395293222692e-06,
+      "loss": 0.6041,
+      "step": 3321
+    },
+    {
+      "epoch": 1.4000421540731374,
+      "grad_norm": 0.2927754819393158,
+      "learning_rate": 2.819285053653472e-06,
+      "loss": 0.5861,
+      "step": 3322
+    },
+    {
+      "epoch": 1.4004636948045106,
+      "grad_norm": 0.3096882700920105,
+      "learning_rate": 2.8181747500666857e-06,
+      "loss": 0.6039,
+      "step": 3323
+    },
+    {
+      "epoch": 1.4008852355358836,
+      "grad_norm": 0.326020210981369,
+      "learning_rate": 2.817064382684951e-06,
+      "loss": 0.6031,
+      "step": 3324
+    },
+    {
+      "epoch": 1.4013067762672569,
+      "grad_norm": 0.3185347318649292,
+      "learning_rate": 2.8159539517309e-06,
+      "loss": 0.5987,
+      "step": 3325
+    },
+    {
+      "epoch": 1.4017283169986299,
+      "grad_norm": 0.3040182590484619,
+      "learning_rate": 2.8148434574271766e-06,
+      "loss": 0.5829,
+      "step": 3326
+    },
+    {
+      "epoch": 1.4021498577300031,
+      "grad_norm": 0.3296566605567932,
+      "learning_rate": 2.8137328999964393e-06,
+      "loss": 0.5969,
+      "step": 3327
+    },
+    {
+      "epoch": 1.4025713984613764,
+      "grad_norm": 0.3105715811252594,
+      "learning_rate": 2.8126222796613566e-06,
+      "loss": 0.5879,
+      "step": 3328
+    },
+    {
+      "epoch": 1.4029929391927496,
+      "grad_norm": 0.29562389850616455,
+      "learning_rate": 2.811511596644613e-06,
+      "loss": 0.5862,
+      "step": 3329
+    },
+    {
+      "epoch": 1.4034144799241226,
+      "grad_norm": 0.2920660078525543,
+      "learning_rate": 2.8104008511689022e-06,
+      "loss": 0.5994,
+      "step": 3330
+    },
+    {
+      "epoch": 1.4038360206554958,
+      "grad_norm": 0.3171636462211609,
+      "learning_rate": 2.809290043456932e-06,
+      "loss": 0.5939,
+      "step": 3331
+    },
+    {
+      "epoch": 1.404257561386869,
+      "grad_norm": 0.30199718475341797,
+      "learning_rate": 2.8081791737314233e-06,
+      "loss": 0.5691,
+      "step": 3332
+    },
+    {
+      "epoch": 1.404679102118242,
+      "grad_norm": 0.31090831756591797,
+      "learning_rate": 2.8070682422151074e-06,
+      "loss": 0.5722,
+      "step": 3333
+    },
+    {
+      "epoch": 1.4051006428496153,
+      "grad_norm": 0.2929885983467102,
+      "learning_rate": 2.8059572491307308e-06,
+      "loss": 0.5928,
+      "step": 3334
+    },
+    {
+      "epoch": 1.4055221835809886,
+      "grad_norm": 0.3145792484283447,
+      "learning_rate": 2.8048461947010484e-06,
+      "loss": 0.5489,
+      "step": 3335
+    },
+    {
+      "epoch": 1.4059437243123618,
+      "grad_norm": 0.3168901801109314,
+      "learning_rate": 2.803735079148833e-06,
+      "loss": 0.6022,
+      "step": 3336
+    },
+    {
+      "epoch": 1.4063652650437348,
+      "grad_norm": 0.3119531273841858,
+      "learning_rate": 2.802623902696864e-06,
+      "loss": 0.5851,
+      "step": 3337
+    },
+    {
+      "epoch": 1.406786805775108,
+      "grad_norm": 0.28591373562812805,
+      "learning_rate": 2.8015126655679355e-06,
+      "loss": 0.5666,
+      "step": 3338
+    },
+    {
+      "epoch": 1.407208346506481,
+      "grad_norm": 0.2941668927669525,
+      "learning_rate": 2.8004013679848553e-06,
+      "loss": 0.5907,
+      "step": 3339
+    },
+    {
+      "epoch": 1.4076298872378543,
+      "grad_norm": 0.29509755969047546,
+      "learning_rate": 2.799290010170441e-06,
+      "loss": 0.5742,
+      "step": 3340
+    },
+    {
+      "epoch": 1.4080514279692276,
+      "grad_norm": 0.3081119954586029,
+      "learning_rate": 2.7981785923475224e-06,
+      "loss": 0.5898,
+      "step": 3341
+    },
+    {
+      "epoch": 1.4084729687006008,
+      "grad_norm": 0.30175018310546875,
+      "learning_rate": 2.797067114738943e-06,
+      "loss": 0.5847,
+      "step": 3342
+    },
+    {
+      "epoch": 1.4088945094319738,
+      "grad_norm": 0.2850877046585083,
+      "learning_rate": 2.7959555775675566e-06,
+      "loss": 0.5788,
+      "step": 3343
+    },
+    {
+      "epoch": 1.409316050163347,
+      "grad_norm": 0.3132425546646118,
+      "learning_rate": 2.7948439810562295e-06,
+      "loss": 0.5939,
+      "step": 3344
+    },
+    {
+      "epoch": 1.4097375908947203,
+      "grad_norm": 0.31174319982528687,
+      "learning_rate": 2.7937323254278396e-06,
+      "loss": 0.621,
+      "step": 3345
+    },
+    {
+      "epoch": 1.4101591316260933,
+      "grad_norm": 0.30393752455711365,
+      "learning_rate": 2.792620610905279e-06,
+      "loss": 0.5856,
+      "step": 3346
+    },
+    {
+      "epoch": 1.4105806723574665,
+      "grad_norm": 0.2986343801021576,
+      "learning_rate": 2.7915088377114467e-06,
+      "loss": 0.6026,
+      "step": 3347
+    },
+    {
+      "epoch": 1.4110022130888398,
+      "grad_norm": 0.2806013226509094,
+      "learning_rate": 2.7903970060692593e-06,
+      "loss": 0.5715,
+      "step": 3348
+    },
+    {
+      "epoch": 1.411423753820213,
+      "grad_norm": 0.32553255558013916,
+      "learning_rate": 2.7892851162016404e-06,
+      "loss": 0.5952,
+      "step": 3349
+    },
+    {
+      "epoch": 1.411845294551586,
+      "grad_norm": 0.2910824120044708,
+      "learning_rate": 2.7881731683315278e-06,
+      "loss": 0.6011,
+      "step": 3350
+    },
+    {
+      "epoch": 1.4122668352829593,
+      "grad_norm": 0.292082816362381,
+      "learning_rate": 2.7870611626818695e-06,
+      "loss": 0.5837,
+      "step": 3351
+    },
+    {
+      "epoch": 1.4126883760143323,
+      "grad_norm": 0.2835635244846344,
+      "learning_rate": 2.7859490994756266e-06,
+      "loss": 0.6001,
+      "step": 3352
+    },
+    {
+      "epoch": 1.4131099167457055,
+      "grad_norm": 0.304046094417572,
+      "learning_rate": 2.784836978935771e-06,
+      "loss": 0.589,
+      "step": 3353
+    },
+    {
+      "epoch": 1.4135314574770788,
+      "grad_norm": 0.3100425899028778,
+      "learning_rate": 2.7837248012852856e-06,
+      "loss": 0.5999,
+      "step": 3354
+    },
+    {
+      "epoch": 1.413952998208452,
+      "grad_norm": 0.29196494817733765,
+      "learning_rate": 2.782612566747166e-06,
+      "loss": 0.596,
+      "step": 3355
+    },
+    {
+      "epoch": 1.414374538939825,
+      "grad_norm": 0.2883398234844208,
+      "learning_rate": 2.7815002755444167e-06,
+      "loss": 0.5694,
+      "step": 3356
+    },
+    {
+      "epoch": 1.4147960796711982,
+      "grad_norm": 0.2867739796638489,
+      "learning_rate": 2.7803879279000564e-06,
+      "loss": 0.5888,
+      "step": 3357
+    },
+    {
+      "epoch": 1.4152176204025713,
+      "grad_norm": 0.29258131980895996,
+      "learning_rate": 2.779275524037114e-06,
+      "loss": 0.5924,
+      "step": 3358
+    },
+    {
+      "epoch": 1.4156391611339445,
+      "grad_norm": 0.3025737702846527,
+      "learning_rate": 2.77816306417863e-06,
+      "loss": 0.5627,
+      "step": 3359
+    },
+    {
+      "epoch": 1.4160607018653177,
+      "grad_norm": 0.2961178719997406,
+      "learning_rate": 2.777050548547655e-06,
+      "loss": 0.5986,
+      "step": 3360
+    },
+    {
+      "epoch": 1.416482242596691,
+      "grad_norm": 0.28487756848335266,
+      "learning_rate": 2.7759379773672523e-06,
+      "loss": 0.5978,
+      "step": 3361
+    },
+    {
+      "epoch": 1.416903783328064,
+      "grad_norm": 0.3012565076351166,
+      "learning_rate": 2.7748253508604956e-06,
+      "loss": 0.6132,
+      "step": 3362
+    },
+    {
+      "epoch": 1.4173253240594372,
+      "grad_norm": 0.29162392020225525,
+      "learning_rate": 2.7737126692504684e-06,
+      "loss": 0.6173,
+      "step": 3363
+    },
+    {
+      "epoch": 1.4177468647908105,
+      "grad_norm": 0.32096660137176514,
+      "learning_rate": 2.772599932760268e-06,
+      "loss": 0.5966,
+      "step": 3364
+    },
+    {
+      "epoch": 1.4181684055221835,
+      "grad_norm": 0.29748719930648804,
+      "learning_rate": 2.7714871416130006e-06,
+      "loss": 0.59,
+      "step": 3365
+    },
+    {
+      "epoch": 1.4185899462535567,
+      "grad_norm": 0.29640376567840576,
+      "learning_rate": 2.770374296031784e-06,
+      "loss": 0.5852,
+      "step": 3366
+    },
+    {
+      "epoch": 1.41901148698493,
+      "grad_norm": 0.30361825227737427,
+      "learning_rate": 2.7692613962397476e-06,
+      "loss": 0.5811,
+      "step": 3367
+    },
+    {
+      "epoch": 1.4194330277163032,
+      "grad_norm": 0.32674944400787354,
+      "learning_rate": 2.7681484424600304e-06,
+      "loss": 0.5659,
+      "step": 3368
+    },
+    {
+      "epoch": 1.4198545684476762,
+      "grad_norm": 0.29427820444107056,
+      "learning_rate": 2.7670354349157823e-06,
+      "loss": 0.5876,
+      "step": 3369
+    },
+    {
+      "epoch": 1.4202761091790495,
+      "grad_norm": 0.2985539138317108,
+      "learning_rate": 2.7659223738301654e-06,
+      "loss": 0.5802,
+      "step": 3370
+    },
+    {
+      "epoch": 1.4206976499104225,
+      "grad_norm": 0.31124410033226013,
+      "learning_rate": 2.7648092594263515e-06,
+      "loss": 0.5731,
+      "step": 3371
+    },
+    {
+      "epoch": 1.4211191906417957,
+      "grad_norm": 0.29945337772369385,
+      "learning_rate": 2.763696091927523e-06,
+      "loss": 0.6045,
+      "step": 3372
+    },
+    {
+      "epoch": 1.421540731373169,
+      "grad_norm": 0.2963219881057739,
+      "learning_rate": 2.762582871556873e-06,
+      "loss": 0.5843,
+      "step": 3373
+    },
+    {
+      "epoch": 1.4219622721045422,
+      "grad_norm": 0.2961929738521576,
+      "learning_rate": 2.761469598537606e-06,
+      "loss": 0.5812,
+      "step": 3374
+    },
+    {
+      "epoch": 1.4223838128359152,
+      "grad_norm": 0.2981439232826233,
+      "learning_rate": 2.7603562730929356e-06,
+      "loss": 0.5993,
+      "step": 3375
+    },
+    {
+      "epoch": 1.4228053535672884,
+      "grad_norm": 0.3159082233905792,
+      "learning_rate": 2.7592428954460874e-06,
+      "loss": 0.5809,
+      "step": 3376
+    },
+    {
+      "epoch": 1.4232268942986617,
+      "grad_norm": 0.2877364158630371,
+      "learning_rate": 2.7581294658202955e-06,
+      "loss": 0.5854,
+      "step": 3377
+    },
+    {
+      "epoch": 1.4236484350300347,
+      "grad_norm": 0.29503753781318665,
+      "learning_rate": 2.7570159844388084e-06,
+      "loss": 0.5787,
+      "step": 3378
+    },
+    {
+      "epoch": 1.424069975761408,
+      "grad_norm": 0.30632492899894714,
+      "learning_rate": 2.75590245152488e-06,
+      "loss": 0.6043,
+      "step": 3379
+    },
+    {
+      "epoch": 1.4244915164927812,
+      "grad_norm": 0.29542362689971924,
+      "learning_rate": 2.7547888673017773e-06,
+      "loss": 0.6089,
+      "step": 3380
+    },
+    {
+      "epoch": 1.4249130572241544,
+      "grad_norm": 0.28478533029556274,
+      "learning_rate": 2.7536752319927777e-06,
+      "loss": 0.6077,
+      "step": 3381
+    },
+    {
+      "epoch": 1.4253345979555274,
+      "grad_norm": 0.2832476496696472,
+      "learning_rate": 2.7525615458211687e-06,
+      "loss": 0.5999,
+      "step": 3382
+    },
+    {
+      "epoch": 1.4257561386869007,
+      "grad_norm": 0.3003033399581909,
+      "learning_rate": 2.7514478090102454e-06,
+      "loss": 0.5729,
+      "step": 3383
+    },
+    {
+      "epoch": 1.4261776794182737,
+      "grad_norm": 0.2935294210910797,
+      "learning_rate": 2.750334021783317e-06,
+      "loss": 0.6154,
+      "step": 3384
+    },
+    {
+      "epoch": 1.426599220149647,
+      "grad_norm": 0.3114016354084015,
+      "learning_rate": 2.7492201843637014e-06,
+      "loss": 0.5796,
+      "step": 3385
+    },
+    {
+      "epoch": 1.4270207608810201,
+      "grad_norm": 0.2895793914794922,
+      "learning_rate": 2.748106296974724e-06,
+      "loss": 0.5936,
+      "step": 3386
+    },
+    {
+      "epoch": 1.4274423016123934,
+      "grad_norm": 0.2924756109714508,
+      "learning_rate": 2.746992359839725e-06,
+      "loss": 0.5916,
+      "step": 3387
+    },
+    {
+      "epoch": 1.4278638423437664,
+      "grad_norm": 0.3021862506866455,
+      "learning_rate": 2.74587837318205e-06,
+      "loss": 0.6037,
+      "step": 3388
+    },
+    {
+      "epoch": 1.4282853830751396,
+      "grad_norm": 0.308737188577652,
+      "learning_rate": 2.744764337225058e-06,
+      "loss": 0.5791,
+      "step": 3389
+    },
+    {
+      "epoch": 1.4287069238065129,
+      "grad_norm": 0.2955123782157898,
+      "learning_rate": 2.7436502521921156e-06,
+      "loss": 0.6107,
+      "step": 3390
+    },
+    {
+      "epoch": 1.429128464537886,
+      "grad_norm": 0.2995598018169403,
+      "learning_rate": 2.7425361183065996e-06,
+      "loss": 0.5842,
+      "step": 3391
+    },
+    {
+      "epoch": 1.4295500052692591,
+      "grad_norm": 0.2977871894836426,
+      "learning_rate": 2.7414219357918985e-06,
+      "loss": 0.5895,
+      "step": 3392
+    },
+    {
+      "epoch": 1.4299715460006324,
+      "grad_norm": 0.3206859529018402,
+      "learning_rate": 2.740307704871408e-06,
+      "loss": 0.5828,
+      "step": 3393
+    },
+    {
+      "epoch": 1.4303930867320056,
+      "grad_norm": 0.2925015389919281,
+      "learning_rate": 2.7391934257685343e-06,
+      "loss": 0.6093,
+      "step": 3394
+    },
+    {
+      "epoch": 1.4308146274633786,
+      "grad_norm": 0.29272252321243286,
+      "learning_rate": 2.7380790987066943e-06,
+      "loss": 0.591,
+      "step": 3395
+    },
+    {
+      "epoch": 1.4312361681947519,
+      "grad_norm": 0.2932765781879425,
+      "learning_rate": 2.736964723909314e-06,
+      "loss": 0.5765,
+      "step": 3396
+    },
+    {
+      "epoch": 1.4316577089261249,
+      "grad_norm": 0.29696348309516907,
+      "learning_rate": 2.7358503015998268e-06,
+      "loss": 0.6216,
+      "step": 3397
+    },
+    {
+      "epoch": 1.432079249657498,
+      "grad_norm": 0.309659481048584,
+      "learning_rate": 2.7347358320016803e-06,
+      "loss": 0.5788,
+      "step": 3398
+    },
+    {
+      "epoch": 1.4325007903888713,
+      "grad_norm": 0.29304036498069763,
+      "learning_rate": 2.733621315338327e-06,
+      "loss": 0.5837,
+      "step": 3399
+    },
+    {
+      "epoch": 1.4329223311202446,
+      "grad_norm": 0.27830079197883606,
+      "learning_rate": 2.732506751833231e-06,
+      "loss": 0.6073,
+      "step": 3400
+    },
+    {
+      "epoch": 1.4333438718516176,
+      "grad_norm": 0.28949394822120667,
+      "learning_rate": 2.731392141709866e-06,
+      "loss": 0.5966,
+      "step": 3401
+    },
+    {
+      "epoch": 1.4337654125829908,
+      "grad_norm": 0.29491496086120605,
+      "learning_rate": 2.730277485191714e-06,
+      "loss": 0.6048,
+      "step": 3402
+    },
+    {
+      "epoch": 1.434186953314364,
+      "grad_norm": 0.3171347975730896,
+      "learning_rate": 2.7291627825022676e-06,
+      "loss": 0.5688,
+      "step": 3403
+    },
+    {
+      "epoch": 1.434608494045737,
+      "grad_norm": 0.3027823865413666,
+      "learning_rate": 2.7280480338650274e-06,
+      "loss": 0.5947,
+      "step": 3404
+    },
+    {
+      "epoch": 1.4350300347771103,
+      "grad_norm": 0.305703341960907,
+      "learning_rate": 2.7269332395035027e-06,
+      "loss": 0.5897,
+      "step": 3405
+    },
+    {
+      "epoch": 1.4354515755084836,
+      "grad_norm": 0.294720321893692,
+      "learning_rate": 2.7258183996412143e-06,
+      "loss": 0.5947,
+      "step": 3406
+    },
+    {
+      "epoch": 1.4358731162398568,
+      "grad_norm": 0.2926400303840637,
+      "learning_rate": 2.72470351450169e-06,
+      "loss": 0.5854,
+      "step": 3407
+    },
+    {
+      "epoch": 1.4362946569712298,
+      "grad_norm": 0.299633264541626,
+      "learning_rate": 2.723588584308469e-06,
+      "loss": 0.593,
+      "step": 3408
+    },
+    {
+      "epoch": 1.436716197702603,
+      "grad_norm": 0.29350244998931885,
+      "learning_rate": 2.7224736092850953e-06,
+      "loss": 0.5933,
+      "step": 3409
+    },
+    {
+      "epoch": 1.437137738433976,
+      "grad_norm": 0.3039315938949585,
+      "learning_rate": 2.721358589655127e-06,
+      "loss": 0.5949,
+      "step": 3410
+    },
+    {
+      "epoch": 1.4375592791653493,
+      "grad_norm": 0.2951664626598358,
+      "learning_rate": 2.7202435256421272e-06,
+      "loss": 0.5755,
+      "step": 3411
+    },
+    {
+      "epoch": 1.4379808198967226,
+      "grad_norm": 0.2923935055732727,
+      "learning_rate": 2.719128417469671e-06,
+      "loss": 0.604,
+      "step": 3412
+    },
+    {
+      "epoch": 1.4384023606280958,
+      "grad_norm": 0.2960101366043091,
+      "learning_rate": 2.718013265361339e-06,
+      "loss": 0.5938,
+      "step": 3413
+    },
+    {
+      "epoch": 1.4388239013594688,
+      "grad_norm": 0.29657021164894104,
+      "learning_rate": 2.7168980695407232e-06,
+      "loss": 0.582,
+      "step": 3414
+    },
+    {
+      "epoch": 1.439245442090842,
+      "grad_norm": 0.2999946177005768,
+      "learning_rate": 2.7157828302314243e-06,
+      "loss": 0.57,
+      "step": 3415
+    },
+    {
+      "epoch": 1.439666982822215,
+      "grad_norm": 0.3005353510379791,
+      "learning_rate": 2.7146675476570495e-06,
+      "loss": 0.5536,
+      "step": 3416
+    },
+    {
+      "epoch": 1.4400885235535883,
+      "grad_norm": 0.2926855981349945,
+      "learning_rate": 2.7135522220412173e-06,
+      "loss": 0.5661,
+      "step": 3417
+    },
+    {
+      "epoch": 1.4405100642849615,
+      "grad_norm": 0.3019557297229767,
+      "learning_rate": 2.712436853607553e-06,
+      "loss": 0.5761,
+      "step": 3418
+    },
+    {
+      "epoch": 1.4409316050163348,
+      "grad_norm": 0.30189451575279236,
+      "learning_rate": 2.711321442579692e-06,
+      "loss": 0.5955,
+      "step": 3419
+    },
+    {
+      "epoch": 1.441353145747708,
+      "grad_norm": 0.29458242654800415,
+      "learning_rate": 2.710205989181276e-06,
+      "loss": 0.6083,
+      "step": 3420
+    },
+    {
+      "epoch": 1.441774686479081,
+      "grad_norm": 0.28012341260910034,
+      "learning_rate": 2.7090904936359584e-06,
+      "loss": 0.598,
+      "step": 3421
+    },
+    {
+      "epoch": 1.4421962272104543,
+      "grad_norm": 0.29635128378868103,
+      "learning_rate": 2.7079749561673974e-06,
+      "loss": 0.584,
+      "step": 3422
+    },
+    {
+      "epoch": 1.4426177679418273,
+      "grad_norm": 0.31176018714904785,
+      "learning_rate": 2.706859376999263e-06,
+      "loss": 0.5904,
+      "step": 3423
+    },
+    {
+      "epoch": 1.4430393086732005,
+      "grad_norm": 0.2949005961418152,
+      "learning_rate": 2.7057437563552313e-06,
+      "loss": 0.5836,
+      "step": 3424
+    },
+    {
+      "epoch": 1.4434608494045738,
+      "grad_norm": 0.3031291663646698,
+      "learning_rate": 2.7046280944589868e-06,
+      "loss": 0.5879,
+      "step": 3425
+    },
+    {
+      "epoch": 1.443882390135947,
+      "grad_norm": 0.28590258955955505,
+      "learning_rate": 2.7035123915342243e-06,
+      "loss": 0.5792,
+      "step": 3426
+    },
+    {
+      "epoch": 1.44430393086732,
+      "grad_norm": 0.30855387449264526,
+      "learning_rate": 2.7023966478046444e-06,
+      "loss": 0.5664,
+      "step": 3427
+    },
+    {
+      "epoch": 1.4447254715986932,
+      "grad_norm": 0.30922797322273254,
+      "learning_rate": 2.7012808634939574e-06,
+      "loss": 0.582,
+      "step": 3428
+    },
+    {
+      "epoch": 1.4451470123300663,
+      "grad_norm": 0.296665757894516,
+      "learning_rate": 2.700165038825881e-06,
+      "loss": 0.596,
+      "step": 3429
+    },
+    {
+      "epoch": 1.4455685530614395,
+      "grad_norm": 0.2930912673473358,
+      "learning_rate": 2.699049174024141e-06,
+      "loss": 0.58,
+      "step": 3430
+    },
+    {
+      "epoch": 1.4459900937928127,
+      "grad_norm": 0.31657347083091736,
+      "learning_rate": 2.6979332693124725e-06,
+      "loss": 0.589,
+      "step": 3431
+    },
+    {
+      "epoch": 1.446411634524186,
+      "grad_norm": 0.29511207342147827,
+      "learning_rate": 2.6968173249146163e-06,
+      "loss": 0.5893,
+      "step": 3432
+    },
+    {
+      "epoch": 1.446833175255559,
+      "grad_norm": 0.3076256811618805,
+      "learning_rate": 2.6957013410543236e-06,
+      "loss": 0.6113,
+      "step": 3433
+    },
+    {
+      "epoch": 1.4472547159869322,
+      "grad_norm": 0.2825218439102173,
+      "learning_rate": 2.6945853179553515e-06,
+      "loss": 0.595,
+      "step": 3434
+    },
+    {
+      "epoch": 1.4476762567183055,
+      "grad_norm": 0.30000385642051697,
+      "learning_rate": 2.6934692558414668e-06,
+      "loss": 0.6155,
+      "step": 3435
+    },
+    {
+      "epoch": 1.4480977974496785,
+      "grad_norm": 0.300182044506073,
+      "learning_rate": 2.6923531549364406e-06,
+      "loss": 0.6117,
+      "step": 3436
+    },
+    {
+      "epoch": 1.4485193381810517,
+      "grad_norm": 0.31268733739852905,
+      "learning_rate": 2.691237015464058e-06,
+      "loss": 0.5822,
+      "step": 3437
+    },
+    {
+      "epoch": 1.448940878912425,
+      "grad_norm": 0.30006104707717896,
+      "learning_rate": 2.690120837648106e-06,
+      "loss": 0.586,
+      "step": 3438
+    },
+    {
+      "epoch": 1.4493624196437982,
+      "grad_norm": 0.30494168400764465,
+      "learning_rate": 2.689004621712381e-06,
+      "loss": 0.5882,
+      "step": 3439
+    },
+    {
+      "epoch": 1.4497839603751712,
+      "grad_norm": 0.3195059895515442,
+      "learning_rate": 2.6878883678806893e-06,
+      "loss": 0.567,
+      "step": 3440
+    },
+    {
+      "epoch": 1.4502055011065444,
+      "grad_norm": 0.3034472167491913,
+      "learning_rate": 2.6867720763768413e-06,
+      "loss": 0.5844,
+      "step": 3441
+    },
+    {
+      "epoch": 1.4506270418379175,
+      "grad_norm": 0.29971474409103394,
+      "learning_rate": 2.685655747424658e-06,
+      "loss": 0.6168,
+      "step": 3442
+    },
+    {
+      "epoch": 1.4510485825692907,
+      "grad_norm": 0.3150702714920044,
+      "learning_rate": 2.6845393812479654e-06,
+      "loss": 0.6112,
+      "step": 3443
+    },
+    {
+      "epoch": 1.451470123300664,
+      "grad_norm": 0.3192824721336365,
+      "learning_rate": 2.683422978070599e-06,
+      "loss": 0.5926,
+      "step": 3444
+    },
+    {
+      "epoch": 1.4518916640320372,
+      "grad_norm": 0.30889713764190674,
+      "learning_rate": 2.6823065381164005e-06,
+      "loss": 0.5906,
+      "step": 3445
+    },
+    {
+      "epoch": 1.4523132047634102,
+      "grad_norm": 0.2992887794971466,
+      "learning_rate": 2.6811900616092184e-06,
+      "loss": 0.5997,
+      "step": 3446
+    },
+    {
+      "epoch": 1.4527347454947834,
+      "grad_norm": 0.32777637243270874,
+      "learning_rate": 2.6800735487729105e-06,
+      "loss": 0.5836,
+      "step": 3447
+    },
+    {
+      "epoch": 1.4531562862261567,
+      "grad_norm": 0.316920667886734,
+      "learning_rate": 2.678956999831341e-06,
+      "loss": 0.5912,
+      "step": 3448
+    },
+    {
+      "epoch": 1.4535778269575297,
+      "grad_norm": 0.29719284176826477,
+      "learning_rate": 2.6778404150083807e-06,
+      "loss": 0.5782,
+      "step": 3449
+    },
+    {
+      "epoch": 1.453999367688903,
+      "grad_norm": 0.3020935356616974,
+      "learning_rate": 2.6767237945279085e-06,
+      "loss": 0.5684,
+      "step": 3450
+    },
+    {
+      "epoch": 1.4544209084202762,
+      "grad_norm": 0.3000442385673523,
+      "learning_rate": 2.6756071386138093e-06,
+      "loss": 0.6064,
+      "step": 3451
+    },
+    {
+      "epoch": 1.4548424491516494,
+      "grad_norm": 0.30084696412086487,
+      "learning_rate": 2.6744904474899757e-06,
+      "loss": 0.5717,
+      "step": 3452
+    },
+    {
+      "epoch": 1.4552639898830224,
+      "grad_norm": 0.2928771376609802,
+      "learning_rate": 2.673373721380308e-06,
+      "loss": 0.5652,
+      "step": 3453
+    },
+    {
+      "epoch": 1.4556855306143957,
+      "grad_norm": 0.2998981773853302,
+      "learning_rate": 2.6722569605087135e-06,
+      "loss": 0.5966,
+      "step": 3454
+    },
+    {
+      "epoch": 1.4561070713457687,
+      "grad_norm": 0.3031393885612488,
+      "learning_rate": 2.6711401650991046e-06,
+      "loss": 0.5948,
+      "step": 3455
+    },
+    {
+      "epoch": 1.456528612077142,
+      "grad_norm": 0.3077911138534546,
+      "learning_rate": 2.6700233353754028e-06,
+      "loss": 0.5786,
+      "step": 3456
+    },
+    {
+      "epoch": 1.4569501528085151,
+      "grad_norm": 0.29332536458969116,
+      "learning_rate": 2.6689064715615352e-06,
+      "loss": 0.597,
+      "step": 3457
+    },
+    {
+      "epoch": 1.4573716935398884,
+      "grad_norm": 0.2992127239704132,
+      "learning_rate": 2.667789573881438e-06,
+      "loss": 0.5742,
+      "step": 3458
+    },
+    {
+      "epoch": 1.4577932342712614,
+      "grad_norm": 0.3119100332260132,
+      "learning_rate": 2.6666726425590493e-06,
+      "loss": 0.5825,
+      "step": 3459
+    },
+    {
+      "epoch": 1.4582147750026346,
+      "grad_norm": 0.30374351143836975,
+      "learning_rate": 2.6655556778183202e-06,
+      "loss": 0.5925,
+      "step": 3460
+    },
+    {
+      "epoch": 1.4586363157340079,
+      "grad_norm": 0.2960633337497711,
+      "learning_rate": 2.664438679883203e-06,
+      "loss": 0.5645,
+      "step": 3461
+    },
+    {
+      "epoch": 1.4590578564653809,
+      "grad_norm": 0.29791486263275146,
+      "learning_rate": 2.663321648977659e-06,
+      "loss": 0.5827,
+      "step": 3462
+    },
+    {
+      "epoch": 1.4594793971967541,
+      "grad_norm": 0.2991044521331787,
+      "learning_rate": 2.662204585325658e-06,
+      "loss": 0.5899,
+      "step": 3463
+    },
+    {
+      "epoch": 1.4599009379281274,
+      "grad_norm": 0.3043212294578552,
+      "learning_rate": 2.661087489151173e-06,
+      "loss": 0.5996,
+      "step": 3464
+    },
+    {
+      "epoch": 1.4603224786595006,
+      "grad_norm": 0.30340638756752014,
+      "learning_rate": 2.659970360678185e-06,
+      "loss": 0.5693,
+      "step": 3465
+    },
+    {
+      "epoch": 1.4607440193908736,
+      "grad_norm": 0.3011249005794525,
+      "learning_rate": 2.6588532001306816e-06,
+      "loss": 0.5973,
+      "step": 3466
+    },
+    {
+      "epoch": 1.4611655601222469,
+      "grad_norm": 0.2995767593383789,
+      "learning_rate": 2.6577360077326575e-06,
+      "loss": 0.6199,
+      "step": 3467
+    },
+    {
+      "epoch": 1.4615871008536199,
+      "grad_norm": 0.29743605852127075,
+      "learning_rate": 2.656618783708112e-06,
+      "loss": 0.5928,
+      "step": 3468
+    },
+    {
+      "epoch": 1.462008641584993,
+      "grad_norm": 0.311765193939209,
+      "learning_rate": 2.6555015282810513e-06,
+      "loss": 0.6051,
+      "step": 3469
+    },
+    {
+      "epoch": 1.4624301823163663,
+      "grad_norm": 0.29331085085868835,
+      "learning_rate": 2.6543842416754904e-06,
+      "loss": 0.5727,
+      "step": 3470
+    },
+    {
+      "epoch": 1.4628517230477396,
+      "grad_norm": 0.29077184200286865,
+      "learning_rate": 2.6532669241154463e-06,
+      "loss": 0.5838,
+      "step": 3471
+    },
+    {
+      "epoch": 1.4632732637791126,
+      "grad_norm": 0.308862566947937,
+      "learning_rate": 2.652149575824946e-06,
+      "loss": 0.5654,
+      "step": 3472
+    },
+    {
+      "epoch": 1.4636948045104858,
+      "grad_norm": 0.3015170097351074,
+      "learning_rate": 2.651032197028019e-06,
+      "loss": 0.5797,
+      "step": 3473
+    },
+    {
+      "epoch": 1.464116345241859,
+      "grad_norm": 0.29836374521255493,
+      "learning_rate": 2.6499147879487055e-06,
+      "loss": 0.5767,
+      "step": 3474
+    },
+    {
+      "epoch": 1.464537885973232,
+      "grad_norm": 0.30396518111228943,
+      "learning_rate": 2.6487973488110464e-06,
+      "loss": 0.5884,
+      "step": 3475
+    },
+    {
+      "epoch": 1.4649594267046053,
+      "grad_norm": 0.2966499626636505,
+      "learning_rate": 2.6476798798390933e-06,
+      "loss": 0.5985,
+      "step": 3476
+    },
+    {
+      "epoch": 1.4653809674359786,
+      "grad_norm": 0.3014944791793823,
+      "learning_rate": 2.6465623812569017e-06,
+      "loss": 0.6082,
+      "step": 3477
+    },
+    {
+      "epoch": 1.4658025081673518,
+      "grad_norm": 0.30070722103118896,
+      "learning_rate": 2.6454448532885334e-06,
+      "loss": 0.5892,
+      "step": 3478
+    },
+    {
+      "epoch": 1.4662240488987248,
+      "grad_norm": 0.28342294692993164,
+      "learning_rate": 2.6443272961580557e-06,
+      "loss": 0.5803,
+      "step": 3479
+    },
+    {
+      "epoch": 1.466645589630098,
+      "grad_norm": 0.3001711964607239,
+      "learning_rate": 2.643209710089542e-06,
+      "loss": 0.5655,
+      "step": 3480
+    },
+    {
+      "epoch": 1.467067130361471,
+      "grad_norm": 0.307221919298172,
+      "learning_rate": 2.6420920953070718e-06,
+      "loss": 0.5767,
+      "step": 3481
+    },
+    {
+      "epoch": 1.4674886710928443,
+      "grad_norm": 0.2992505133152008,
+      "learning_rate": 2.6409744520347286e-06,
+      "loss": 0.5943,
+      "step": 3482
+    },
+    {
+      "epoch": 1.4679102118242175,
+      "grad_norm": 0.30098411440849304,
+      "learning_rate": 2.6398567804966047e-06,
+      "loss": 0.5817,
+      "step": 3483
+    },
+    {
+      "epoch": 1.4683317525555908,
+      "grad_norm": 0.28887873888015747,
+      "learning_rate": 2.638739080916797e-06,
+      "loss": 0.5822,
+      "step": 3484
+    },
+    {
+      "epoch": 1.4687532932869638,
+      "grad_norm": 0.2855852246284485,
+      "learning_rate": 2.637621353519405e-06,
+      "loss": 0.5815,
+      "step": 3485
+    },
+    {
+      "epoch": 1.469174834018337,
+      "grad_norm": 0.28934967517852783,
+      "learning_rate": 2.636503598528538e-06,
+      "loss": 0.5793,
+      "step": 3486
+    },
+    {
+      "epoch": 1.46959637474971,
+      "grad_norm": 0.29756221175193787,
+      "learning_rate": 2.635385816168309e-06,
+      "loss": 0.5791,
+      "step": 3487
+    },
+    {
+      "epoch": 1.4700179154810833,
+      "grad_norm": 0.29399409890174866,
+      "learning_rate": 2.634268006662836e-06,
+      "loss": 0.6228,
+      "step": 3488
+    },
+    {
+      "epoch": 1.4704394562124565,
+      "grad_norm": 0.304025262594223,
+      "learning_rate": 2.6331501702362438e-06,
+      "loss": 0.5977,
+      "step": 3489
+    },
+    {
+      "epoch": 1.4708609969438298,
+      "grad_norm": 0.29822152853012085,
+      "learning_rate": 2.6320323071126613e-06,
+      "loss": 0.5592,
+      "step": 3490
+    },
+    {
+      "epoch": 1.4712825376752028,
+      "grad_norm": 0.30142873525619507,
+      "learning_rate": 2.6309144175162227e-06,
+      "loss": 0.5986,
+      "step": 3491
+    },
+    {
+      "epoch": 1.471704078406576,
+      "grad_norm": 0.291329950094223,
+      "learning_rate": 2.629796501671069e-06,
+      "loss": 0.5712,
+      "step": 3492
+    },
+    {
+      "epoch": 1.4721256191379493,
+      "grad_norm": 0.2994924783706665,
+      "learning_rate": 2.6286785598013454e-06,
+      "loss": 0.5669,
+      "step": 3493
+    },
+    {
+      "epoch": 1.4725471598693223,
+      "grad_norm": 0.2885320484638214,
+      "learning_rate": 2.627560592131202e-06,
+      "loss": 0.5843,
+      "step": 3494
+    },
+    {
+      "epoch": 1.4729687006006955,
+      "grad_norm": 0.3015397787094116,
+      "learning_rate": 2.6264425988847946e-06,
+      "loss": 0.584,
+      "step": 3495
+    },
+    {
+      "epoch": 1.4733902413320688,
+      "grad_norm": 0.3015003502368927,
+      "learning_rate": 2.6253245802862835e-06,
+      "loss": 0.5954,
+      "step": 3496
+    },
+    {
+      "epoch": 1.473811782063442,
+      "grad_norm": 0.3006497919559479,
+      "learning_rate": 2.624206536559837e-06,
+      "loss": 0.58,
+      "step": 3497
+    },
+    {
+      "epoch": 1.474233322794815,
+      "grad_norm": 0.29307278990745544,
+      "learning_rate": 2.623088467929623e-06,
+      "loss": 0.5837,
+      "step": 3498
+    },
+    {
+      "epoch": 1.4746548635261882,
+      "grad_norm": 0.29331204295158386,
+      "learning_rate": 2.62197037461982e-06,
+      "loss": 0.6178,
+      "step": 3499
+    },
+    {
+      "epoch": 1.4750764042575613,
+      "grad_norm": 0.3011537194252014,
+      "learning_rate": 2.6208522568546065e-06,
+      "loss": 0.5716,
+      "step": 3500
+    },
+    {
+      "epoch": 1.4754979449889345,
+      "grad_norm": 0.29503676295280457,
+      "learning_rate": 2.61973411485817e-06,
+      "loss": 0.6134,
+      "step": 3501
+    },
+    {
+      "epoch": 1.4759194857203077,
+      "grad_norm": 0.2993324100971222,
+      "learning_rate": 2.6186159488547008e-06,
+      "loss": 0.5961,
+      "step": 3502
+    },
+    {
+      "epoch": 1.476341026451681,
+      "grad_norm": 0.286316454410553,
+      "learning_rate": 2.617497759068394e-06,
+      "loss": 0.5892,
+      "step": 3503
+    },
+    {
+      "epoch": 1.476762567183054,
+      "grad_norm": 0.2889440357685089,
+      "learning_rate": 2.6163795457234507e-06,
+      "loss": 0.6149,
+      "step": 3504
+    },
+    {
+      "epoch": 1.4771841079144272,
+      "grad_norm": 0.29811346530914307,
+      "learning_rate": 2.615261309044075e-06,
+      "loss": 0.6027,
+      "step": 3505
+    },
+    {
+      "epoch": 1.4776056486458005,
+      "grad_norm": 0.2963462173938751,
+      "learning_rate": 2.6141430492544774e-06,
+      "loss": 0.581,
+      "step": 3506
+    },
+    {
+      "epoch": 1.4780271893771735,
+      "grad_norm": 0.2980509400367737,
+      "learning_rate": 2.6130247665788704e-06,
+      "loss": 0.5791,
+      "step": 3507
+    },
+    {
+      "epoch": 1.4784487301085467,
+      "grad_norm": 0.30712074041366577,
+      "learning_rate": 2.611906461241475e-06,
+      "loss": 0.5673,
+      "step": 3508
+    },
+    {
+      "epoch": 1.47887027083992,
+      "grad_norm": 0.2894923985004425,
+      "learning_rate": 2.610788133466514e-06,
+      "loss": 0.5984,
+      "step": 3509
+    },
+    {
+      "epoch": 1.4792918115712932,
+      "grad_norm": 0.2792559266090393,
+      "learning_rate": 2.609669783478215e-06,
+      "loss": 0.6079,
+      "step": 3510
+    },
+    {
+      "epoch": 1.4797133523026662,
+      "grad_norm": 0.30070579051971436,
+      "learning_rate": 2.608551411500811e-06,
+      "loss": 0.5711,
+      "step": 3511
+    },
+    {
+      "epoch": 1.4801348930340394,
+      "grad_norm": 0.28686878085136414,
+      "learning_rate": 2.607433017758537e-06,
+      "loss": 0.5974,
+      "step": 3512
+    },
+    {
+      "epoch": 1.4805564337654125,
+      "grad_norm": 0.29124149680137634,
+      "learning_rate": 2.6063146024756363e-06,
+      "loss": 0.5842,
+      "step": 3513
+    },
+    {
+      "epoch": 1.4809779744967857,
+      "grad_norm": 0.29352059960365295,
+      "learning_rate": 2.605196165876353e-06,
+      "loss": 0.5859,
+      "step": 3514
+    },
+    {
+      "epoch": 1.481399515228159,
+      "grad_norm": 0.30183735489845276,
+      "learning_rate": 2.6040777081849376e-06,
+      "loss": 0.612,
+      "step": 3515
+    },
+    {
+      "epoch": 1.4818210559595322,
+      "grad_norm": 0.30462905764579773,
+      "learning_rate": 2.6029592296256432e-06,
+      "loss": 0.5595,
+      "step": 3516
+    },
+    {
+      "epoch": 1.4822425966909052,
+      "grad_norm": 0.31815826892852783,
+      "learning_rate": 2.6018407304227287e-06,
+      "loss": 0.5893,
+      "step": 3517
+    },
+    {
+      "epoch": 1.4826641374222784,
+      "grad_norm": 0.29226306080818176,
+      "learning_rate": 2.600722210800457e-06,
+      "loss": 0.5706,
+      "step": 3518
+    },
+    {
+      "epoch": 1.4830856781536517,
+      "grad_norm": 0.29249241948127747,
+      "learning_rate": 2.5996036709830936e-06,
+      "loss": 0.5614,
+      "step": 3519
+    },
+    {
+      "epoch": 1.4835072188850247,
+      "grad_norm": 0.29787373542785645,
+      "learning_rate": 2.5984851111949087e-06,
+      "loss": 0.5693,
+      "step": 3520
+    },
+    {
+      "epoch": 1.483928759616398,
+      "grad_norm": 0.31063440442085266,
+      "learning_rate": 2.597366531660177e-06,
+      "loss": 0.5676,
+      "step": 3521
+    },
+    {
+      "epoch": 1.4843503003477712,
+      "grad_norm": 0.28566092252731323,
+      "learning_rate": 2.596247932603177e-06,
+      "loss": 0.6046,
+      "step": 3522
+    },
+    {
+      "epoch": 1.4847718410791444,
+      "grad_norm": 0.29877951741218567,
+      "learning_rate": 2.5951293142481914e-06,
+      "loss": 0.583,
+      "step": 3523
+    },
+    {
+      "epoch": 1.4851933818105174,
+      "grad_norm": 0.2936956286430359,
+      "learning_rate": 2.594010676819505e-06,
+      "loss": 0.6062,
+      "step": 3524
+    },
+    {
+      "epoch": 1.4856149225418906,
+      "grad_norm": 0.2911481559276581,
+      "learning_rate": 2.592892020541409e-06,
+      "loss": 0.5897,
+      "step": 3525
+    },
+    {
+      "epoch": 1.4860364632732637,
+      "grad_norm": 0.3079085052013397,
+      "learning_rate": 2.591773345638198e-06,
+      "loss": 0.5691,
+      "step": 3526
+    },
+    {
+      "epoch": 1.486458004004637,
+      "grad_norm": 0.3014970123767853,
+      "learning_rate": 2.590654652334168e-06,
+      "loss": 0.5738,
+      "step": 3527
+    },
+    {
+      "epoch": 1.4868795447360101,
+      "grad_norm": 0.30996766686439514,
+      "learning_rate": 2.58953594085362e-06,
+      "loss": 0.5861,
+      "step": 3528
+    },
+    {
+      "epoch": 1.4873010854673834,
+      "grad_norm": 0.3025328814983368,
+      "learning_rate": 2.588417211420861e-06,
+      "loss": 0.5842,
+      "step": 3529
+    },
+    {
+      "epoch": 1.4877226261987564,
+      "grad_norm": 0.2956964373588562,
+      "learning_rate": 2.587298464260197e-06,
+      "loss": 0.5886,
+      "step": 3530
+    },
+    {
+      "epoch": 1.4881441669301296,
+      "grad_norm": 0.29815295338630676,
+      "learning_rate": 2.586179699595941e-06,
+      "loss": 0.5987,
+      "step": 3531
+    },
+    {
+      "epoch": 1.4885657076615029,
+      "grad_norm": 0.28119924664497375,
+      "learning_rate": 2.585060917652409e-06,
+      "loss": 0.6164,
+      "step": 3532
+    },
+    {
+      "epoch": 1.4889872483928759,
+      "grad_norm": 0.3042287528514862,
+      "learning_rate": 2.583942118653919e-06,
+      "loss": 0.5847,
+      "step": 3533
+    },
+    {
+      "epoch": 1.4894087891242491,
+      "grad_norm": 0.292683869600296,
+      "learning_rate": 2.5828233028247946e-06,
+      "loss": 0.6013,
+      "step": 3534
+    },
+    {
+      "epoch": 1.4898303298556224,
+      "grad_norm": 0.28549301624298096,
+      "learning_rate": 2.5817044703893596e-06,
+      "loss": 0.5862,
+      "step": 3535
+    },
+    {
+      "epoch": 1.4902518705869956,
+      "grad_norm": 0.2980212867259979,
+      "learning_rate": 2.5805856215719456e-06,
+      "loss": 0.5665,
+      "step": 3536
+    },
+    {
+      "epoch": 1.4906734113183686,
+      "grad_norm": 0.2925409972667694,
+      "learning_rate": 2.5794667565968835e-06,
+      "loss": 0.5911,
+      "step": 3537
+    },
+    {
+      "epoch": 1.4910949520497419,
+      "grad_norm": 0.3037315309047699,
+      "learning_rate": 2.5783478756885095e-06,
+      "loss": 0.5632,
+      "step": 3538
+    },
+    {
+      "epoch": 1.4915164927811149,
+      "grad_norm": 0.2972307801246643,
+      "learning_rate": 2.577228979071161e-06,
+      "loss": 0.5833,
+      "step": 3539
+    },
+    {
+      "epoch": 1.491938033512488,
+      "grad_norm": 0.3054017126560211,
+      "learning_rate": 2.576110066969182e-06,
+      "loss": 0.5723,
+      "step": 3540
+    },
+    {
+      "epoch": 1.4923595742438613,
+      "grad_norm": 0.29576724767684937,
+      "learning_rate": 2.574991139606917e-06,
+      "loss": 0.5994,
+      "step": 3541
+    },
+    {
+      "epoch": 1.4927811149752346,
+      "grad_norm": 0.29567933082580566,
+      "learning_rate": 2.5738721972087135e-06,
+      "loss": 0.5869,
+      "step": 3542
+    },
+    {
+      "epoch": 1.4932026557066076,
+      "grad_norm": 0.29343125224113464,
+      "learning_rate": 2.572753239998923e-06,
+      "loss": 0.5773,
+      "step": 3543
+    },
+    {
+      "epoch": 1.4936241964379808,
+      "grad_norm": 0.3047654628753662,
+      "learning_rate": 2.571634268201899e-06,
+      "loss": 0.591,
+      "step": 3544
+    },
+    {
+      "epoch": 1.494045737169354,
+      "grad_norm": 0.3021463453769684,
+      "learning_rate": 2.5705152820419997e-06,
+      "loss": 0.5939,
+      "step": 3545
+    },
+    {
+      "epoch": 1.494467277900727,
+      "grad_norm": 0.2943010628223419,
+      "learning_rate": 2.5693962817435847e-06,
+      "loss": 0.5854,
+      "step": 3546
+    },
+    {
+      "epoch": 1.4948888186321003,
+      "grad_norm": 0.2834304869174957,
+      "learning_rate": 2.5682772675310162e-06,
+      "loss": 0.5782,
+      "step": 3547
+    },
+    {
+      "epoch": 1.4953103593634736,
+      "grad_norm": 0.29256150126457214,
+      "learning_rate": 2.5671582396286602e-06,
+      "loss": 0.5813,
+      "step": 3548
+    },
+    {
+      "epoch": 1.4957319000948468,
+      "grad_norm": 0.287106454372406,
+      "learning_rate": 2.566039198260885e-06,
+      "loss": 0.582,
+      "step": 3549
+    },
+    {
+      "epoch": 1.4961534408262198,
+      "grad_norm": 0.29320839047431946,
+      "learning_rate": 2.564920143652062e-06,
+      "loss": 0.5626,
+      "step": 3550
+    },
+    {
+      "epoch": 1.496574981557593,
+      "grad_norm": 0.3126447796821594,
+      "learning_rate": 2.5638010760265634e-06,
+      "loss": 0.5958,
+      "step": 3551
+    },
+    {
+      "epoch": 1.496996522288966,
+      "grad_norm": 0.28795498609542847,
+      "learning_rate": 2.5626819956087677e-06,
+      "loss": 0.5699,
+      "step": 3552
+    },
+    {
+      "epoch": 1.4974180630203393,
+      "grad_norm": 0.3038882613182068,
+      "learning_rate": 2.5615629026230517e-06,
+      "loss": 0.5649,
+      "step": 3553
+    },
+    {
+      "epoch": 1.4978396037517125,
+      "grad_norm": 0.2957231104373932,
+      "learning_rate": 2.560443797293797e-06,
+      "loss": 0.5827,
+      "step": 3554
+    },
+    {
+      "epoch": 1.4982611444830858,
+      "grad_norm": 0.29233047366142273,
+      "learning_rate": 2.559324679845388e-06,
+      "loss": 0.5757,
+      "step": 3555
+    },
+    {
+      "epoch": 1.4986826852144588,
+      "grad_norm": 0.297922283411026,
+      "learning_rate": 2.5582055505022113e-06,
+      "loss": 0.5787,
+      "step": 3556
+    },
+    {
+      "epoch": 1.499104225945832,
+      "grad_norm": 0.2945384979248047,
+      "learning_rate": 2.5570864094886556e-06,
+      "loss": 0.6121,
+      "step": 3557
+    },
+    {
+      "epoch": 1.499525766677205,
+      "grad_norm": 0.3055792450904846,
+      "learning_rate": 2.5559672570291107e-06,
+      "loss": 0.6044,
+      "step": 3558
+    },
+    {
+      "epoch": 1.4999473074085783,
+      "grad_norm": 0.3113223612308502,
+      "learning_rate": 2.554848093347971e-06,
+      "loss": 0.5937,
+      "step": 3559
+    },
+    {
+      "epoch": 1.5003688481399515,
+      "grad_norm": 0.29192689061164856,
+      "learning_rate": 2.5537289186696307e-06,
+      "loss": 0.5915,
+      "step": 3560
+    },
+    {
+      "epoch": 1.5007903888713248,
+      "grad_norm": 0.28963759541511536,
+      "learning_rate": 2.552609733218489e-06,
+      "loss": 0.5816,
+      "step": 3561
+    },
+    {
+      "epoch": 1.501211929602698,
+      "grad_norm": 0.28231024742126465,
+      "learning_rate": 2.5514905372189452e-06,
+      "loss": 0.5811,
+      "step": 3562
+    },
+    {
+      "epoch": 1.501633470334071,
+      "grad_norm": 0.2863914966583252,
+      "learning_rate": 2.5503713308954016e-06,
+      "loss": 0.5827,
+      "step": 3563
+    },
+    {
+      "epoch": 1.502055011065444,
+      "grad_norm": 0.29540544748306274,
+      "learning_rate": 2.549252114472261e-06,
+      "loss": 0.5478,
+      "step": 3564
+    },
+    {
+      "epoch": 1.5024765517968173,
+      "grad_norm": 0.29869765043258667,
+      "learning_rate": 2.548132888173931e-06,
+      "loss": 0.5593,
+      "step": 3565
+    },
+    {
+      "epoch": 1.5028980925281905,
+      "grad_norm": 0.3039306402206421,
+      "learning_rate": 2.5470136522248194e-06,
+      "loss": 0.5822,
+      "step": 3566
+    },
+    {
+      "epoch": 1.5033196332595637,
+      "grad_norm": 0.2989601492881775,
+      "learning_rate": 2.545894406849336e-06,
+      "loss": 0.5878,
+      "step": 3567
+    },
+    {
+      "epoch": 1.503741173990937,
+      "grad_norm": 0.3000372052192688,
+      "learning_rate": 2.5447751522718923e-06,
+      "loss": 0.5866,
+      "step": 3568
+    },
+    {
+      "epoch": 1.50416271472231,
+      "grad_norm": 0.2912314236164093,
+      "learning_rate": 2.543655888716902e-06,
+      "loss": 0.5856,
+      "step": 3569
+    },
+    {
+      "epoch": 1.5045842554536832,
+      "grad_norm": 0.29722198843955994,
+      "learning_rate": 2.5425366164087815e-06,
+      "loss": 0.5895,
+      "step": 3570
+    },
+    {
+      "epoch": 1.5050057961850563,
+      "grad_norm": 0.29338255524635315,
+      "learning_rate": 2.5414173355719486e-06,
+      "loss": 0.6006,
+      "step": 3571
+    },
+    {
+      "epoch": 1.5054273369164295,
+      "grad_norm": 0.3094235360622406,
+      "learning_rate": 2.5402980464308204e-06,
+      "loss": 0.5825,
+      "step": 3572
+    },
+    {
+      "epoch": 1.5058488776478027,
+      "grad_norm": 0.3042086958885193,
+      "learning_rate": 2.5391787492098186e-06,
+      "loss": 0.5896,
+      "step": 3573
+    },
+    {
+      "epoch": 1.506270418379176,
+      "grad_norm": 0.30986449122428894,
+      "learning_rate": 2.5380594441333646e-06,
+      "loss": 0.5987,
+      "step": 3574
+    },
+    {
+      "epoch": 1.5066919591105492,
+      "grad_norm": 0.30189040303230286,
+      "learning_rate": 2.536940131425884e-06,
+      "loss": 0.6195,
+      "step": 3575
+    },
+    {
+      "epoch": 1.5071134998419222,
+      "grad_norm": 0.3006190359592438,
+      "learning_rate": 2.535820811311801e-06,
+      "loss": 0.6016,
+      "step": 3576
+    },
+    {
+      "epoch": 1.5075350405732952,
+      "grad_norm": 0.30533310770988464,
+      "learning_rate": 2.534701484015544e-06,
+      "loss": 0.6018,
+      "step": 3577
+    },
+    {
+      "epoch": 1.5079565813046685,
+      "grad_norm": 0.29603710770606995,
+      "learning_rate": 2.5335821497615386e-06,
+      "loss": 0.5945,
+      "step": 3578
+    },
+    {
+      "epoch": 1.5083781220360417,
+      "grad_norm": 0.2892718017101288,
+      "learning_rate": 2.5324628087742166e-06,
+      "loss": 0.5968,
+      "step": 3579
+    },
+    {
+      "epoch": 1.508799662767415,
+      "grad_norm": 0.2917032539844513,
+      "learning_rate": 2.5313434612780087e-06,
+      "loss": 0.5761,
+      "step": 3580
+    },
+    {
+      "epoch": 1.5092212034987882,
+      "grad_norm": 0.31096839904785156,
+      "learning_rate": 2.5302241074973466e-06,
+      "loss": 0.5889,
+      "step": 3581
+    },
+    {
+      "epoch": 1.5096427442301612,
+      "grad_norm": 0.2867354154586792,
+      "learning_rate": 2.529104747656665e-06,
+      "loss": 0.5918,
+      "step": 3582
+    },
+    {
+      "epoch": 1.5100642849615344,
+      "grad_norm": 0.2880653440952301,
+      "learning_rate": 2.527985381980398e-06,
+      "loss": 0.5955,
+      "step": 3583
+    },
+    {
+      "epoch": 1.5104858256929075,
+      "grad_norm": 0.2995817959308624,
+      "learning_rate": 2.5268660106929814e-06,
+      "loss": 0.5857,
+      "step": 3584
+    },
+    {
+      "epoch": 1.5109073664242807,
+      "grad_norm": 0.29353612661361694,
+      "learning_rate": 2.525746634018853e-06,
+      "loss": 0.5917,
+      "step": 3585
+    },
+    {
+      "epoch": 1.511328907155654,
+      "grad_norm": 0.3013385832309723,
+      "learning_rate": 2.5246272521824506e-06,
+      "loss": 0.5748,
+      "step": 3586
+    },
+    {
+      "epoch": 1.5117504478870272,
+      "grad_norm": 0.3044423758983612,
+      "learning_rate": 2.523507865408214e-06,
+      "loss": 0.5615,
+      "step": 3587
+    },
+    {
+      "epoch": 1.5121719886184004,
+      "grad_norm": 0.30474796891212463,
+      "learning_rate": 2.5223884739205827e-06,
+      "loss": 0.6001,
+      "step": 3588
+    },
+    {
+      "epoch": 1.5125935293497734,
+      "grad_norm": 0.2932731807231903,
+      "learning_rate": 2.521269077943999e-06,
+      "loss": 0.5937,
+      "step": 3589
+    },
+    {
+      "epoch": 1.5130150700811464,
+      "grad_norm": 0.28825250267982483,
+      "learning_rate": 2.520149677702904e-06,
+      "loss": 0.5741,
+      "step": 3590
+    },
+    {
+      "epoch": 1.5134366108125197,
+      "grad_norm": 0.3133872449398041,
+      "learning_rate": 2.519030273421741e-06,
+      "loss": 0.5995,
+      "step": 3591
+    },
+    {
+      "epoch": 1.513858151543893,
+      "grad_norm": 0.31731486320495605,
+      "learning_rate": 2.5179108653249527e-06,
+      "loss": 0.6189,
+      "step": 3592
+    },
+    {
+      "epoch": 1.5142796922752662,
+      "grad_norm": 0.2924398183822632,
+      "learning_rate": 2.5167914536369853e-06,
+      "loss": 0.5926,
+      "step": 3593
+    },
+    {
+      "epoch": 1.5147012330066394,
+      "grad_norm": 0.2987661063671112,
+      "learning_rate": 2.5156720385822834e-06,
+      "loss": 0.5855,
+      "step": 3594
+    },
+    {
+      "epoch": 1.5151227737380124,
+      "grad_norm": 0.32465311884880066,
+      "learning_rate": 2.5145526203852925e-06,
+      "loss": 0.6041,
+      "step": 3595
+    },
+    {
+      "epoch": 1.5155443144693856,
+      "grad_norm": 0.2953962981700897,
+      "learning_rate": 2.5134331992704608e-06,
+      "loss": 0.5837,
+      "step": 3596
+    },
+    {
+      "epoch": 1.5159658552007587,
+      "grad_norm": 0.30803224444389343,
+      "learning_rate": 2.5123137754622332e-06,
+      "loss": 0.5945,
+      "step": 3597
+    },
+    {
+      "epoch": 1.516387395932132,
+      "grad_norm": 0.3063245415687561,
+      "learning_rate": 2.511194349185059e-06,
+      "loss": 0.5923,
+      "step": 3598
+    },
+    {
+      "epoch": 1.5168089366635051,
+      "grad_norm": 0.32613253593444824,
+      "learning_rate": 2.510074920663385e-06,
+      "loss": 0.6146,
+      "step": 3599
+    },
+    {
+      "epoch": 1.5172304773948784,
+      "grad_norm": 0.31052982807159424,
+      "learning_rate": 2.5089554901216608e-06,
+      "loss": 0.5901,
+      "step": 3600
+    },
+    {
+      "epoch": 1.5176520181262516,
+      "grad_norm": 0.2959611415863037,
+      "learning_rate": 2.5078360577843358e-06,
+      "loss": 0.5751,
+      "step": 3601
+    },
+    {
+      "epoch": 1.5180735588576246,
+      "grad_norm": 0.3180309236049652,
+      "learning_rate": 2.506716623875858e-06,
+      "loss": 0.5789,
+      "step": 3602
+    },
+    {
+      "epoch": 1.5184950995889976,
+      "grad_norm": 0.3029859662055969,
+      "learning_rate": 2.505597188620679e-06,
+      "loss": 0.5847,
+      "step": 3603
+    },
+    {
+      "epoch": 1.5189166403203709,
+      "grad_norm": 0.3114137351512909,
+      "learning_rate": 2.5044777522432455e-06,
+      "loss": 0.5747,
+      "step": 3604
+    },
+    {
+      "epoch": 1.5193381810517441,
+      "grad_norm": 0.3022953271865845,
+      "learning_rate": 2.503358314968012e-06,
+      "loss": 0.5906,
+      "step": 3605
+    },
+    {
+      "epoch": 1.5197597217831174,
+      "grad_norm": 0.308308869600296,
+      "learning_rate": 2.5022388770194257e-06,
+      "loss": 0.5786,
+      "step": 3606
+    },
+    {
+      "epoch": 1.5201812625144906,
+      "grad_norm": 0.295617014169693,
+      "learning_rate": 2.5011194386219382e-06,
+      "loss": 0.5796,
+      "step": 3607
+    },
+    {
+      "epoch": 1.5206028032458636,
+      "grad_norm": 0.2933404743671417,
+      "learning_rate": 2.5e-06,
+      "loss": 0.6007,
+      "step": 3608
+    },
+    {
+      "epoch": 1.5210243439772368,
+      "grad_norm": 0.29691824316978455,
+      "learning_rate": 2.498880561378062e-06,
+      "loss": 0.5717,
+      "step": 3609
+    },
+    {
+      "epoch": 1.5214458847086099,
+      "grad_norm": 0.2917468249797821,
+      "learning_rate": 2.497761122980575e-06,
+      "loss": 0.6058,
+      "step": 3610
+    },
+    {
+      "epoch": 1.521867425439983,
+      "grad_norm": 0.2927970290184021,
+      "learning_rate": 2.4966416850319885e-06,
+      "loss": 0.5737,
+      "step": 3611
+    },
+    {
+      "epoch": 1.5222889661713563,
+      "grad_norm": 0.30854225158691406,
+      "learning_rate": 2.495522247756755e-06,
+      "loss": 0.6057,
+      "step": 3612
+    },
+    {
+      "epoch": 1.5227105069027296,
+      "grad_norm": 0.300367534160614,
+      "learning_rate": 2.494402811379322e-06,
+      "loss": 0.5872,
+      "step": 3613
+    },
+    {
+      "epoch": 1.5231320476341026,
+      "grad_norm": 0.2959570586681366,
+      "learning_rate": 2.4932833761241426e-06,
+      "loss": 0.5828,
+      "step": 3614
+    },
+    {
+      "epoch": 1.5235535883654758,
+      "grad_norm": 0.3017304241657257,
+      "learning_rate": 2.492163942215665e-06,
+      "loss": 0.58,
+      "step": 3615
+    },
+    {
+      "epoch": 1.5239751290968488,
+      "grad_norm": 0.3164164423942566,
+      "learning_rate": 2.4910445098783397e-06,
+      "loss": 0.5743,
+      "step": 3616
+    },
+    {
+      "epoch": 1.524396669828222,
+      "grad_norm": 0.32071593403816223,
+      "learning_rate": 2.4899250793366157e-06,
+      "loss": 0.5966,
+      "step": 3617
+    },
+    {
+      "epoch": 1.5248182105595953,
+      "grad_norm": 0.2972114682197571,
+      "learning_rate": 2.4888056508149416e-06,
+      "loss": 0.569,
+      "step": 3618
+    },
+    {
+      "epoch": 1.5252397512909686,
+      "grad_norm": 0.3042148947715759,
+      "learning_rate": 2.4876862245377676e-06,
+      "loss": 0.6178,
+      "step": 3619
+    },
+    {
+      "epoch": 1.5256612920223418,
+      "grad_norm": 0.31152641773223877,
+      "learning_rate": 2.486566800729539e-06,
+      "loss": 0.5802,
+      "step": 3620
+    },
+    {
+      "epoch": 1.5260828327537148,
+      "grad_norm": 0.3259347379207611,
+      "learning_rate": 2.485447379614708e-06,
+      "loss": 0.5961,
+      "step": 3621
+    },
+    {
+      "epoch": 1.526504373485088,
+      "grad_norm": 0.3148970901966095,
+      "learning_rate": 2.4843279614177166e-06,
+      "loss": 0.5829,
+      "step": 3622
+    },
+    {
+      "epoch": 1.526925914216461,
+      "grad_norm": 0.3175041079521179,
+      "learning_rate": 2.483208546363015e-06,
+      "loss": 0.5853,
+      "step": 3623
+    },
+    {
+      "epoch": 1.5273474549478343,
+      "grad_norm": 0.30138248205184937,
+      "learning_rate": 2.4820891346750477e-06,
+      "loss": 0.5925,
+      "step": 3624
+    },
+    {
+      "epoch": 1.5277689956792075,
+      "grad_norm": 0.30485856533050537,
+      "learning_rate": 2.4809697265782606e-06,
+      "loss": 0.591,
+      "step": 3625
+    },
+    {
+      "epoch": 1.5281905364105808,
+      "grad_norm": 0.2888585329055786,
+      "learning_rate": 2.479850322297097e-06,
+      "loss": 0.5558,
+      "step": 3626
+    },
+    {
+      "epoch": 1.5286120771419538,
+      "grad_norm": 0.299938827753067,
+      "learning_rate": 2.4787309220560022e-06,
+      "loss": 0.5782,
+      "step": 3627
+    },
+    {
+      "epoch": 1.529033617873327,
+      "grad_norm": 0.29689034819602966,
+      "learning_rate": 2.4776115260794177e-06,
+      "loss": 0.5739,
+      "step": 3628
+    },
+    {
+      "epoch": 1.5294551586047,
+      "grad_norm": 0.29847270250320435,
+      "learning_rate": 2.4764921345917863e-06,
+      "loss": 0.5763,
+      "step": 3629
+    },
+    {
+      "epoch": 1.5298766993360733,
+      "grad_norm": 0.29785361886024475,
+      "learning_rate": 2.4753727478175502e-06,
+      "loss": 0.5883,
+      "step": 3630
+    },
+    {
+      "epoch": 1.5302982400674465,
+      "grad_norm": 0.2942366302013397,
+      "learning_rate": 2.4742533659811474e-06,
+      "loss": 0.5652,
+      "step": 3631
+    },
+    {
+      "epoch": 1.5307197807988198,
+      "grad_norm": 0.30152860283851624,
+      "learning_rate": 2.4731339893070194e-06,
+      "loss": 0.5883,
+      "step": 3632
+    },
+    {
+      "epoch": 1.531141321530193,
+      "grad_norm": 0.29714319109916687,
+      "learning_rate": 2.472014618019603e-06,
+      "loss": 0.5892,
+      "step": 3633
+    },
+    {
+      "epoch": 1.531562862261566,
+      "grad_norm": 0.302165150642395,
+      "learning_rate": 2.470895252343336e-06,
+      "loss": 0.5955,
+      "step": 3634
+    },
+    {
+      "epoch": 1.531984402992939,
+      "grad_norm": 0.29542696475982666,
+      "learning_rate": 2.469775892502654e-06,
+      "loss": 0.5927,
+      "step": 3635
+    },
+    {
+      "epoch": 1.5324059437243123,
+      "grad_norm": 0.30531051754951477,
+      "learning_rate": 2.4686565387219925e-06,
+      "loss": 0.5958,
+      "step": 3636
+    },
+    {
+      "epoch": 1.5328274844556855,
+      "grad_norm": 0.316754549741745,
+      "learning_rate": 2.4675371912257843e-06,
+      "loss": 0.5614,
+      "step": 3637
+    },
+    {
+      "epoch": 1.5332490251870587,
+      "grad_norm": 0.2903212308883667,
+      "learning_rate": 2.4664178502384613e-06,
+      "loss": 0.5748,
+      "step": 3638
+    },
+    {
+      "epoch": 1.533670565918432,
+      "grad_norm": 0.29809728264808655,
+      "learning_rate": 2.465298515984457e-06,
+      "loss": 0.5798,
+      "step": 3639
+    },
+    {
+      "epoch": 1.534092106649805,
+      "grad_norm": 0.299789160490036,
+      "learning_rate": 2.464179188688199e-06,
+      "loss": 0.5658,
+      "step": 3640
+    },
+    {
+      "epoch": 1.5345136473811782,
+      "grad_norm": 0.29003390669822693,
+      "learning_rate": 2.4630598685741163e-06,
+      "loss": 0.591,
+      "step": 3641
+    },
+    {
+      "epoch": 1.5349351881125513,
+      "grad_norm": 0.2950785160064697,
+      "learning_rate": 2.4619405558666358e-06,
+      "loss": 0.5987,
+      "step": 3642
+    },
+    {
+      "epoch": 1.5353567288439245,
+      "grad_norm": 0.28770002722740173,
+      "learning_rate": 2.4608212507901827e-06,
+      "loss": 0.5842,
+      "step": 3643
+    },
+    {
+      "epoch": 1.5357782695752977,
+      "grad_norm": 0.2997967600822449,
+      "learning_rate": 2.459701953569181e-06,
+      "loss": 0.5812,
+      "step": 3644
+    },
+    {
+      "epoch": 1.536199810306671,
+      "grad_norm": 0.2968835234642029,
+      "learning_rate": 2.458582664428053e-06,
+      "loss": 0.5741,
+      "step": 3645
+    },
+    {
+      "epoch": 1.5366213510380442,
+      "grad_norm": 0.3013487458229065,
+      "learning_rate": 2.457463383591219e-06,
+      "loss": 0.58,
+      "step": 3646
+    },
+    {
+      "epoch": 1.5370428917694172,
+      "grad_norm": 0.2940315306186676,
+      "learning_rate": 2.456344111283099e-06,
+      "loss": 0.5984,
+      "step": 3647
+    },
+    {
+      "epoch": 1.5374644325007902,
+      "grad_norm": 0.29682525992393494,
+      "learning_rate": 2.455224847728108e-06,
+      "loss": 0.5871,
+      "step": 3648
+    },
+    {
+      "epoch": 1.5378859732321635,
+      "grad_norm": 0.2898973524570465,
+      "learning_rate": 2.4541055931506645e-06,
+      "loss": 0.5664,
+      "step": 3649
+    },
+    {
+      "epoch": 1.5383075139635367,
+      "grad_norm": 0.3045501708984375,
+      "learning_rate": 2.452986347775181e-06,
+      "loss": 0.5937,
+      "step": 3650
+    },
+    {
+      "epoch": 1.53872905469491,
+      "grad_norm": 0.28188663721084595,
+      "learning_rate": 2.4518671118260693e-06,
+      "loss": 0.5777,
+      "step": 3651
+    },
+    {
+      "epoch": 1.5391505954262832,
+      "grad_norm": 0.2984975576400757,
+      "learning_rate": 2.4507478855277397e-06,
+      "loss": 0.5837,
+      "step": 3652
+    },
+    {
+      "epoch": 1.5395721361576562,
+      "grad_norm": 0.29442572593688965,
+      "learning_rate": 2.4496286691045996e-06,
+      "loss": 0.5769,
+      "step": 3653
+    },
+    {
+      "epoch": 1.5399936768890294,
+      "grad_norm": 0.3023555874824524,
+      "learning_rate": 2.448509462781056e-06,
+      "loss": 0.5932,
+      "step": 3654
+    },
+    {
+      "epoch": 1.5404152176204025,
+      "grad_norm": 0.2973801791667938,
+      "learning_rate": 2.447390266781512e-06,
+      "loss": 0.5852,
+      "step": 3655
+    },
+    {
+      "epoch": 1.5408367583517757,
+      "grad_norm": 0.294986754655838,
+      "learning_rate": 2.4462710813303706e-06,
+      "loss": 0.5673,
+      "step": 3656
+    },
+    {
+      "epoch": 1.541258299083149,
+      "grad_norm": 0.3115188479423523,
+      "learning_rate": 2.44515190665203e-06,
+      "loss": 0.5757,
+      "step": 3657
+    },
+    {
+      "epoch": 1.5416798398145222,
+      "grad_norm": 0.29074788093566895,
+      "learning_rate": 2.4440327429708897e-06,
+      "loss": 0.5998,
+      "step": 3658
+    },
+    {
+      "epoch": 1.5421013805458954,
+      "grad_norm": 0.3190818727016449,
+      "learning_rate": 2.4429135905113456e-06,
+      "loss": 0.5982,
+      "step": 3659
+    },
+    {
+      "epoch": 1.5425229212772684,
+      "grad_norm": 0.30750471353530884,
+      "learning_rate": 2.4417944494977887e-06,
+      "loss": 0.586,
+      "step": 3660
+    },
+    {
+      "epoch": 1.5429444620086414,
+      "grad_norm": 0.28924497961997986,
+      "learning_rate": 2.4406753201546123e-06,
+      "loss": 0.6034,
+      "step": 3661
+    },
+    {
+      "epoch": 1.5433660027400147,
+      "grad_norm": 0.29935264587402344,
+      "learning_rate": 2.4395562027062036e-06,
+      "loss": 0.5804,
+      "step": 3662
+    },
+    {
+      "epoch": 1.543787543471388,
+      "grad_norm": 0.33243483304977417,
+      "learning_rate": 2.4384370973769496e-06,
+      "loss": 0.5922,
+      "step": 3663
+    },
+    {
+      "epoch": 1.5442090842027612,
+      "grad_norm": 0.2968990206718445,
+      "learning_rate": 2.437318004391233e-06,
+      "loss": 0.6048,
+      "step": 3664
+    },
+    {
+      "epoch": 1.5446306249341344,
+      "grad_norm": 0.2964472472667694,
+      "learning_rate": 2.4361989239734374e-06,
+      "loss": 0.5821,
+      "step": 3665
+    },
+    {
+      "epoch": 1.5450521656655074,
+      "grad_norm": 0.30615106225013733,
+      "learning_rate": 2.4350798563479387e-06,
+      "loss": 0.5954,
+      "step": 3666
+    },
+    {
+      "epoch": 1.5454737063968806,
+      "grad_norm": 0.29245394468307495,
+      "learning_rate": 2.4339608017391158e-06,
+      "loss": 0.602,
+      "step": 3667
+    },
+    {
+      "epoch": 1.5458952471282537,
+      "grad_norm": 0.28930625319480896,
+      "learning_rate": 2.43284176037134e-06,
+      "loss": 0.5777,
+      "step": 3668
+    },
+    {
+      "epoch": 1.546316787859627,
+      "grad_norm": 0.29431644082069397,
+      "learning_rate": 2.431722732468984e-06,
+      "loss": 0.588,
+      "step": 3669
+    },
+    {
+      "epoch": 1.5467383285910001,
+      "grad_norm": 0.301534503698349,
+      "learning_rate": 2.430603718256416e-06,
+      "loss": 0.5731,
+      "step": 3670
+    },
+    {
+      "epoch": 1.5471598693223734,
+      "grad_norm": 0.2948233485221863,
+      "learning_rate": 2.4294847179580008e-06,
+      "loss": 0.5926,
+      "step": 3671
+    },
+    {
+      "epoch": 1.5475814100537466,
+      "grad_norm": 0.29022207856178284,
+      "learning_rate": 2.4283657317981018e-06,
+      "loss": 0.6021,
+      "step": 3672
+    },
+    {
+      "epoch": 1.5480029507851196,
+      "grad_norm": 0.3023090958595276,
+      "learning_rate": 2.427246760001078e-06,
+      "loss": 0.6108,
+      "step": 3673
+    },
+    {
+      "epoch": 1.5484244915164926,
+      "grad_norm": 0.3090320825576782,
+      "learning_rate": 2.426127802791288e-06,
+      "loss": 0.5874,
+      "step": 3674
+    },
+    {
+      "epoch": 1.5488460322478659,
+      "grad_norm": 0.2943379282951355,
+      "learning_rate": 2.425008860393084e-06,
+      "loss": 0.564,
+      "step": 3675
+    },
+    {
+      "epoch": 1.5492675729792391,
+      "grad_norm": 0.29769188165664673,
+      "learning_rate": 2.423889933030819e-06,
+      "loss": 0.61,
+      "step": 3676
+    },
+    {
+      "epoch": 1.5496891137106124,
+      "grad_norm": 0.29603758454322815,
+      "learning_rate": 2.422771020928839e-06,
+      "loss": 0.5785,
+      "step": 3677
+    },
+    {
+      "epoch": 1.5501106544419856,
+      "grad_norm": 0.3057269752025604,
+      "learning_rate": 2.4216521243114917e-06,
+      "loss": 0.5871,
+      "step": 3678
+    },
+    {
+      "epoch": 1.5505321951733586,
+      "grad_norm": 0.29309865832328796,
+      "learning_rate": 2.4205332434031177e-06,
+      "loss": 0.6009,
+      "step": 3679
+    },
+    {
+      "epoch": 1.5509537359047318,
+      "grad_norm": 0.29742783308029175,
+      "learning_rate": 2.4194143784280552e-06,
+      "loss": 0.6048,
+      "step": 3680
+    },
+    {
+      "epoch": 1.5513752766361049,
+      "grad_norm": 0.2976910471916199,
+      "learning_rate": 2.4182955296106413e-06,
+      "loss": 0.5837,
+      "step": 3681
+    },
+    {
+      "epoch": 1.551796817367478,
+      "grad_norm": 0.3002120554447174,
+      "learning_rate": 2.4171766971752066e-06,
+      "loss": 0.588,
+      "step": 3682
+    },
+    {
+      "epoch": 1.5522183580988513,
+      "grad_norm": 0.294275164604187,
+      "learning_rate": 2.4160578813460825e-06,
+      "loss": 0.5644,
+      "step": 3683
+    },
+    {
+      "epoch": 1.5526398988302246,
+      "grad_norm": 0.2953031659126282,
+      "learning_rate": 2.414939082347592e-06,
+      "loss": 0.5954,
+      "step": 3684
+    },
+    {
+      "epoch": 1.5530614395615976,
+      "grad_norm": 0.29205140471458435,
+      "learning_rate": 2.41382030040406e-06,
+      "loss": 0.5887,
+      "step": 3685
+    },
+    {
+      "epoch": 1.5534829802929708,
+      "grad_norm": 0.28612515330314636,
+      "learning_rate": 2.412701535739804e-06,
+      "loss": 0.5903,
+      "step": 3686
+    },
+    {
+      "epoch": 1.5539045210243438,
+      "grad_norm": 0.29986050724983215,
+      "learning_rate": 2.41158278857914e-06,
+      "loss": 0.6028,
+      "step": 3687
+    },
+    {
+      "epoch": 1.554326061755717,
+      "grad_norm": 0.29555845260620117,
+      "learning_rate": 2.4104640591463806e-06,
+      "loss": 0.571,
+      "step": 3688
+    },
+    {
+      "epoch": 1.5547476024870903,
+      "grad_norm": 0.30315718054771423,
+      "learning_rate": 2.409345347665832e-06,
+      "loss": 0.5977,
+      "step": 3689
+    },
+    {
+      "epoch": 1.5551691432184636,
+      "grad_norm": 0.2878809869289398,
+      "learning_rate": 2.408226654361803e-06,
+      "loss": 0.5809,
+      "step": 3690
+    },
+    {
+      "epoch": 1.5555906839498368,
+      "grad_norm": 0.32228919863700867,
+      "learning_rate": 2.4071079794585905e-06,
+      "loss": 0.5854,
+      "step": 3691
+    },
+    {
+      "epoch": 1.5560122246812098,
+      "grad_norm": 0.3006466329097748,
+      "learning_rate": 2.4059893231804954e-06,
+      "loss": 0.572,
+      "step": 3692
+    },
+    {
+      "epoch": 1.5564337654125828,
+      "grad_norm": 0.28226619958877563,
+      "learning_rate": 2.4048706857518094e-06,
+      "loss": 0.6005,
+      "step": 3693
+    },
+    {
+      "epoch": 1.556855306143956,
+      "grad_norm": 0.306768000125885,
+      "learning_rate": 2.403752067396824e-06,
+      "loss": 0.5947,
+      "step": 3694
+    },
+    {
+      "epoch": 1.5572768468753293,
+      "grad_norm": 0.29907432198524475,
+      "learning_rate": 2.4026334683398236e-06,
+      "loss": 0.5843,
+      "step": 3695
+    },
+    {
+      "epoch": 1.5576983876067025,
+      "grad_norm": 0.29742729663848877,
+      "learning_rate": 2.4015148888050925e-06,
+      "loss": 0.5814,
+      "step": 3696
+    },
+    {
+      "epoch": 1.5581199283380758,
+      "grad_norm": 0.3074621260166168,
+      "learning_rate": 2.4003963290169073e-06,
+      "loss": 0.5905,
+      "step": 3697
+    },
+    {
+      "epoch": 1.5585414690694488,
+      "grad_norm": 0.2903301417827606,
+      "learning_rate": 2.3992777891995426e-06,
+      "loss": 0.5789,
+      "step": 3698
+    },
+    {
+      "epoch": 1.558963009800822,
+      "grad_norm": 0.2947191596031189,
+      "learning_rate": 2.3981592695772717e-06,
+      "loss": 0.5904,
+      "step": 3699
+    },
+    {
+      "epoch": 1.559384550532195,
+      "grad_norm": 0.30625274777412415,
+      "learning_rate": 2.3970407703743568e-06,
+      "loss": 0.5799,
+      "step": 3700
+    },
+    {
+      "epoch": 1.5598060912635683,
+      "grad_norm": 0.2962552011013031,
+      "learning_rate": 2.3959222918150633e-06,
+      "loss": 0.5705,
+      "step": 3701
+    },
+    {
+      "epoch": 1.5602276319949415,
+      "grad_norm": 0.3254883885383606,
+      "learning_rate": 2.3948038341236474e-06,
+      "loss": 0.6213,
+      "step": 3702
+    },
+    {
+      "epoch": 1.5606491727263148,
+      "grad_norm": 0.28492075204849243,
+      "learning_rate": 2.393685397524365e-06,
+      "loss": 0.5832,
+      "step": 3703
+    },
+    {
+      "epoch": 1.561070713457688,
+      "grad_norm": 0.30280378460884094,
+      "learning_rate": 2.3925669822414633e-06,
+      "loss": 0.588,
+      "step": 3704
+    },
+    {
+      "epoch": 1.561492254189061,
+      "grad_norm": 0.3157472312450409,
+      "learning_rate": 2.3914485884991906e-06,
+      "loss": 0.5842,
+      "step": 3705
+    },
+    {
+      "epoch": 1.561913794920434,
+      "grad_norm": 0.31567853689193726,
+      "learning_rate": 2.3903302165217856e-06,
+      "loss": 0.5836,
+      "step": 3706
+    },
+    {
+      "epoch": 1.5623353356518073,
+      "grad_norm": 0.29277628660202026,
+      "learning_rate": 2.389211866533486e-06,
+      "loss": 0.5832,
+      "step": 3707
+    },
+    {
+      "epoch": 1.5627568763831805,
+      "grad_norm": 0.30863189697265625,
+      "learning_rate": 2.3880935387585253e-06,
+      "loss": 0.5973,
+      "step": 3708
+    },
+    {
+      "epoch": 1.5631784171145537,
+      "grad_norm": 0.31813809275627136,
+      "learning_rate": 2.3869752334211296e-06,
+      "loss": 0.5957,
+      "step": 3709
+    },
+    {
+      "epoch": 1.563599957845927,
+      "grad_norm": 0.2997576594352722,
+      "learning_rate": 2.385856950745524e-06,
+      "loss": 0.5759,
+      "step": 3710
+    },
+    {
+      "epoch": 1.5640214985773,
+      "grad_norm": 0.2897513508796692,
+      "learning_rate": 2.3847386909559256e-06,
+      "loss": 0.5786,
+      "step": 3711
+    },
+    {
+      "epoch": 1.5644430393086732,
+      "grad_norm": 0.30299249291419983,
+      "learning_rate": 2.38362045427655e-06,
+      "loss": 0.5873,
+      "step": 3712
+    },
+    {
+      "epoch": 1.5648645800400462,
+      "grad_norm": 0.2884225845336914,
+      "learning_rate": 2.3825022409316064e-06,
+      "loss": 0.5854,
+      "step": 3713
+    },
+    {
+      "epoch": 1.5652861207714195,
+      "grad_norm": 0.2946826219558716,
+      "learning_rate": 2.3813840511453005e-06,
+      "loss": 0.5703,
+      "step": 3714
+    },
+    {
+      "epoch": 1.5657076615027927,
+      "grad_norm": 0.3155680000782013,
+      "learning_rate": 2.3802658851418307e-06,
+      "loss": 0.5717,
+      "step": 3715
+    },
+    {
+      "epoch": 1.566129202234166,
+      "grad_norm": 0.3106139600276947,
+      "learning_rate": 2.3791477431453948e-06,
+      "loss": 0.5781,
+      "step": 3716
+    },
+    {
+      "epoch": 1.5665507429655392,
+      "grad_norm": 0.31259074807167053,
+      "learning_rate": 2.378029625380181e-06,
+      "loss": 0.5507,
+      "step": 3717
+    },
+    {
+      "epoch": 1.5669722836969122,
+      "grad_norm": 0.33078715205192566,
+      "learning_rate": 2.376911532070377e-06,
+      "loss": 0.5547,
+      "step": 3718
+    },
+    {
+      "epoch": 1.5673938244282852,
+      "grad_norm": 0.2811542749404907,
+      "learning_rate": 2.375793463440164e-06,
+      "loss": 0.5879,
+      "step": 3719
+    },
+    {
+      "epoch": 1.5678153651596585,
+      "grad_norm": 0.293082594871521,
+      "learning_rate": 2.374675419713716e-06,
+      "loss": 0.5951,
+      "step": 3720
+    },
+    {
+      "epoch": 1.5682369058910317,
+      "grad_norm": 0.3015831708908081,
+      "learning_rate": 2.373557401115206e-06,
+      "loss": 0.5615,
+      "step": 3721
+    },
+    {
+      "epoch": 1.568658446622405,
+      "grad_norm": 0.3106010854244232,
+      "learning_rate": 2.3724394078687986e-06,
+      "loss": 0.5731,
+      "step": 3722
+    },
+    {
+      "epoch": 1.5690799873537782,
+      "grad_norm": 0.2961007356643677,
+      "learning_rate": 2.371321440198656e-06,
+      "loss": 0.5778,
+      "step": 3723
+    },
+    {
+      "epoch": 1.5695015280851512,
+      "grad_norm": 0.28620338439941406,
+      "learning_rate": 2.3702034983289317e-06,
+      "loss": 0.5936,
+      "step": 3724
+    },
+    {
+      "epoch": 1.5699230688165244,
+      "grad_norm": 0.30899062752723694,
+      "learning_rate": 2.369085582483778e-06,
+      "loss": 0.5877,
+      "step": 3725
+    },
+    {
+      "epoch": 1.5703446095478975,
+      "grad_norm": 0.30135616660118103,
+      "learning_rate": 2.367967692887339e-06,
+      "loss": 0.5743,
+      "step": 3726
+    },
+    {
+      "epoch": 1.5707661502792707,
+      "grad_norm": 0.29407235980033875,
+      "learning_rate": 2.3668498297637566e-06,
+      "loss": 0.5981,
+      "step": 3727
+    },
+    {
+      "epoch": 1.571187691010644,
+      "grad_norm": 0.3008360266685486,
+      "learning_rate": 2.3657319933371644e-06,
+      "loss": 0.5734,
+      "step": 3728
+    },
+    {
+      "epoch": 1.5716092317420172,
+      "grad_norm": 0.3053247928619385,
+      "learning_rate": 2.364614183831691e-06,
+      "loss": 0.6119,
+      "step": 3729
+    },
+    {
+      "epoch": 1.5720307724733904,
+      "grad_norm": 0.3027466833591461,
+      "learning_rate": 2.363496401471463e-06,
+      "loss": 0.5827,
+      "step": 3730
+    },
+    {
+      "epoch": 1.5724523132047634,
+      "grad_norm": 0.29496440291404724,
+      "learning_rate": 2.3623786464805958e-06,
+      "loss": 0.5935,
+      "step": 3731
+    },
+    {
+      "epoch": 1.5728738539361364,
+      "grad_norm": 0.3020428419113159,
+      "learning_rate": 2.361260919083205e-06,
+      "loss": 0.5988,
+      "step": 3732
+    },
+    {
+      "epoch": 1.5732953946675097,
+      "grad_norm": 0.2953042685985565,
+      "learning_rate": 2.3601432195033957e-06,
+      "loss": 0.5667,
+      "step": 3733
+    },
+    {
+      "epoch": 1.573716935398883,
+      "grad_norm": 0.32435497641563416,
+      "learning_rate": 2.359025547965272e-06,
+      "loss": 0.5499,
+      "step": 3734
+    },
+    {
+      "epoch": 1.5741384761302561,
+      "grad_norm": 0.3127506971359253,
+      "learning_rate": 2.3579079046929295e-06,
+      "loss": 0.5762,
+      "step": 3735
+    },
+    {
+      "epoch": 1.5745600168616294,
+      "grad_norm": 0.28838980197906494,
+      "learning_rate": 2.3567902899104595e-06,
+      "loss": 0.5823,
+      "step": 3736
+    },
+    {
+      "epoch": 1.5749815575930024,
+      "grad_norm": 0.30400171875953674,
+      "learning_rate": 2.3556727038419447e-06,
+      "loss": 0.5965,
+      "step": 3737
+    },
+    {
+      "epoch": 1.5754030983243756,
+      "grad_norm": 0.31098872423171997,
+      "learning_rate": 2.3545551467114666e-06,
+      "loss": 0.5772,
+      "step": 3738
+    },
+    {
+      "epoch": 1.5758246390557487,
+      "grad_norm": 0.3118710219860077,
+      "learning_rate": 2.3534376187430987e-06,
+      "loss": 0.6008,
+      "step": 3739
+    },
+    {
+      "epoch": 1.576246179787122,
+      "grad_norm": 0.30422765016555786,
+      "learning_rate": 2.352320120160907e-06,
+      "loss": 0.585,
+      "step": 3740
+    },
+    {
+      "epoch": 1.5766677205184951,
+      "grad_norm": 0.29810458421707153,
+      "learning_rate": 2.3512026511889544e-06,
+      "loss": 0.6,
+      "step": 3741
+    },
+    {
+      "epoch": 1.5770892612498684,
+      "grad_norm": 0.2977440357208252,
+      "learning_rate": 2.3500852120512954e-06,
+      "loss": 0.5609,
+      "step": 3742
+    },
+    {
+      "epoch": 1.5775108019812414,
+      "grad_norm": 0.3116713762283325,
+      "learning_rate": 2.348967802971982e-06,
+      "loss": 0.6073,
+      "step": 3743
+    },
+    {
+      "epoch": 1.5779323427126146,
+      "grad_norm": 0.31190240383148193,
+      "learning_rate": 2.3478504241750548e-06,
+      "loss": 0.5971,
+      "step": 3744
+    },
+    {
+      "epoch": 1.5783538834439876,
+      "grad_norm": 0.28863367438316345,
+      "learning_rate": 2.3467330758845545e-06,
+      "loss": 0.6132,
+      "step": 3745
+    },
+    {
+      "epoch": 1.5787754241753609,
+      "grad_norm": 0.3153589367866516,
+      "learning_rate": 2.3456157583245104e-06,
+      "loss": 0.5874,
+      "step": 3746
+    },
+    {
+      "epoch": 1.5791969649067341,
+      "grad_norm": 0.2964811325073242,
+      "learning_rate": 2.3444984717189487e-06,
+      "loss": 0.5717,
+      "step": 3747
+    },
+    {
+      "epoch": 1.5796185056381074,
+      "grad_norm": 0.3026202619075775,
+      "learning_rate": 2.343381216291889e-06,
+      "loss": 0.5795,
+      "step": 3748
+    },
+    {
+      "epoch": 1.5800400463694806,
+      "grad_norm": 0.282535582780838,
+      "learning_rate": 2.3422639922673433e-06,
+      "loss": 0.5478,
+      "step": 3749
+    },
+    {
+      "epoch": 1.5804615871008536,
+      "grad_norm": 0.3062657415866852,
+      "learning_rate": 2.341146799869319e-06,
+      "loss": 0.5515,
+      "step": 3750
+    },
+    {
+      "epoch": 1.5808831278322268,
+      "grad_norm": 0.29218563437461853,
+      "learning_rate": 2.3400296393218154e-06,
+      "loss": 0.5795,
+      "step": 3751
+    },
+    {
+      "epoch": 1.5813046685635999,
+      "grad_norm": 0.29354360699653625,
+      "learning_rate": 2.3389125108488283e-06,
+      "loss": 0.5962,
+      "step": 3752
+    },
+    {
+      "epoch": 1.581726209294973,
+      "grad_norm": 0.2914448082447052,
+      "learning_rate": 2.3377954146743424e-06,
+      "loss": 0.5972,
+      "step": 3753
+    },
+    {
+      "epoch": 1.5821477500263463,
+      "grad_norm": 0.3026101291179657,
+      "learning_rate": 2.3366783510223413e-06,
+      "loss": 0.5799,
+      "step": 3754
+    },
+    {
+      "epoch": 1.5825692907577196,
+      "grad_norm": 0.29034942388534546,
+      "learning_rate": 2.3355613201167982e-06,
+      "loss": 0.5684,
+      "step": 3755
+    },
+    {
+      "epoch": 1.5829908314890926,
+      "grad_norm": 0.294954389333725,
+      "learning_rate": 2.3344443221816806e-06,
+      "loss": 0.5758,
+      "step": 3756
+    },
+    {
+      "epoch": 1.5834123722204658,
+      "grad_norm": 0.29021692276000977,
+      "learning_rate": 2.333327357440951e-06,
+      "loss": 0.5816,
+      "step": 3757
+    },
+    {
+      "epoch": 1.5838339129518388,
+      "grad_norm": 0.31312957406044006,
+      "learning_rate": 2.332210426118563e-06,
+      "loss": 0.5983,
+      "step": 3758
+    },
+    {
+      "epoch": 1.584255453683212,
+      "grad_norm": 0.31288203597068787,
+      "learning_rate": 2.331093528438465e-06,
+      "loss": 0.5683,
+      "step": 3759
+    },
+    {
+      "epoch": 1.5846769944145853,
+      "grad_norm": 0.28266799449920654,
+      "learning_rate": 2.329976664624597e-06,
+      "loss": 0.5976,
+      "step": 3760
+    },
+    {
+      "epoch": 1.5850985351459586,
+      "grad_norm": 0.2996273338794708,
+      "learning_rate": 2.3288598349008967e-06,
+      "loss": 0.5769,
+      "step": 3761
+    },
+    {
+      "epoch": 1.5855200758773318,
+      "grad_norm": 0.3022620975971222,
+      "learning_rate": 2.3277430394912873e-06,
+      "loss": 0.6079,
+      "step": 3762
+    },
+    {
+      "epoch": 1.5859416166087048,
+      "grad_norm": 0.2925678491592407,
+      "learning_rate": 2.3266262786196924e-06,
+      "loss": 0.6016,
+      "step": 3763
+    },
+    {
+      "epoch": 1.5863631573400778,
+      "grad_norm": 0.28642264008522034,
+      "learning_rate": 2.3255095525100247e-06,
+      "loss": 0.5974,
+      "step": 3764
+    },
+    {
+      "epoch": 1.586784698071451,
+      "grad_norm": 0.2943025529384613,
+      "learning_rate": 2.324392861386192e-06,
+      "loss": 0.581,
+      "step": 3765
+    },
+    {
+      "epoch": 1.5872062388028243,
+      "grad_norm": 0.295654296875,
+      "learning_rate": 2.3232762054720924e-06,
+      "loss": 0.6025,
+      "step": 3766
+    },
+    {
+      "epoch": 1.5876277795341975,
+      "grad_norm": 0.298612505197525,
+      "learning_rate": 2.322159584991619e-06,
+      "loss": 0.5714,
+      "step": 3767
+    },
+    {
+      "epoch": 1.5880493202655708,
+      "grad_norm": 0.3012296259403229,
+      "learning_rate": 2.3210430001686595e-06,
+      "loss": 0.6189,
+      "step": 3768
+    },
+    {
+      "epoch": 1.5884708609969438,
+      "grad_norm": 0.2962808907032013,
+      "learning_rate": 2.3199264512270895e-06,
+      "loss": 0.5812,
+      "step": 3769
+    },
+    {
+      "epoch": 1.588892401728317,
+      "grad_norm": 0.29117119312286377,
+      "learning_rate": 2.318809938390782e-06,
+      "loss": 0.5781,
+      "step": 3770
+    },
+    {
+      "epoch": 1.58931394245969,
+      "grad_norm": 0.30441710352897644,
+      "learning_rate": 2.3176934618836004e-06,
+      "loss": 0.5895,
+      "step": 3771
+    },
+    {
+      "epoch": 1.5897354831910633,
+      "grad_norm": 0.2934112846851349,
+      "learning_rate": 2.316577021929402e-06,
+      "loss": 0.5736,
+      "step": 3772
+    },
+    {
+      "epoch": 1.5901570239224365,
+      "grad_norm": 0.3020102083683014,
+      "learning_rate": 2.315460618752035e-06,
+      "loss": 0.604,
+      "step": 3773
+    },
+    {
+      "epoch": 1.5905785646538098,
+      "grad_norm": 0.29384422302246094,
+      "learning_rate": 2.3143442525753433e-06,
+      "loss": 0.5538,
+      "step": 3774
+    },
+    {
+      "epoch": 1.591000105385183,
+      "grad_norm": 0.2866637408733368,
+      "learning_rate": 2.313227923623159e-06,
+      "loss": 0.579,
+      "step": 3775
+    },
+    {
+      "epoch": 1.591421646116556,
+      "grad_norm": 0.30028003454208374,
+      "learning_rate": 2.312111632119311e-06,
+      "loss": 0.5902,
+      "step": 3776
+    },
+    {
+      "epoch": 1.591843186847929,
+      "grad_norm": 0.2902231812477112,
+      "learning_rate": 2.3109953782876197e-06,
+      "loss": 0.6204,
+      "step": 3777
+    },
+    {
+      "epoch": 1.5922647275793023,
+      "grad_norm": 0.2988373637199402,
+      "learning_rate": 2.3098791623518945e-06,
+      "loss": 0.585,
+      "step": 3778
+    },
+    {
+      "epoch": 1.5926862683106755,
+      "grad_norm": 0.3021155893802643,
+      "learning_rate": 2.3087629845359425e-06,
+      "loss": 0.5718,
+      "step": 3779
+    },
+    {
+      "epoch": 1.5931078090420487,
+      "grad_norm": 0.3021233081817627,
+      "learning_rate": 2.3076468450635594e-06,
+      "loss": 0.5824,
+      "step": 3780
+    },
+    {
+      "epoch": 1.593529349773422,
+      "grad_norm": 0.3061806261539459,
+      "learning_rate": 2.306530744158535e-06,
+      "loss": 0.6014,
+      "step": 3781
+    },
+    {
+      "epoch": 1.593950890504795,
+      "grad_norm": 0.28445765376091003,
+      "learning_rate": 2.3054146820446493e-06,
+      "loss": 0.5971,
+      "step": 3782
+    },
+    {
+      "epoch": 1.5943724312361682,
+      "grad_norm": 0.28464338183403015,
+      "learning_rate": 2.3042986589456777e-06,
+      "loss": 0.5905,
+      "step": 3783
+    },
+    {
+      "epoch": 1.5947939719675412,
+      "grad_norm": 0.2990339696407318,
+      "learning_rate": 2.3031826750853845e-06,
+      "loss": 0.5805,
+      "step": 3784
+    },
+    {
+      "epoch": 1.5952155126989145,
+      "grad_norm": 0.3097289800643921,
+      "learning_rate": 2.3020667306875288e-06,
+      "loss": 0.5925,
+      "step": 3785
+    },
+    {
+      "epoch": 1.5956370534302877,
+      "grad_norm": 0.29291418194770813,
+      "learning_rate": 2.3009508259758598e-06,
+      "loss": 0.6024,
+      "step": 3786
+    },
+    {
+      "epoch": 1.596058594161661,
+      "grad_norm": 0.3197702169418335,
+      "learning_rate": 2.2998349611741196e-06,
+      "loss": 0.5595,
+      "step": 3787
+    },
+    {
+      "epoch": 1.5964801348930342,
+      "grad_norm": 0.28722575306892395,
+      "learning_rate": 2.2987191365060434e-06,
+      "loss": 0.5999,
+      "step": 3788
+    },
+    {
+      "epoch": 1.5969016756244072,
+      "grad_norm": 0.305113285779953,
+      "learning_rate": 2.297603352195356e-06,
+      "loss": 0.5321,
+      "step": 3789
+    },
+    {
+      "epoch": 1.5973232163557802,
+      "grad_norm": 0.27913209795951843,
+      "learning_rate": 2.2964876084657765e-06,
+      "loss": 0.5738,
+      "step": 3790
+    },
+    {
+      "epoch": 1.5977447570871535,
+      "grad_norm": 0.30234357714653015,
+      "learning_rate": 2.2953719055410136e-06,
+      "loss": 0.5805,
+      "step": 3791
+    },
+    {
+      "epoch": 1.5981662978185267,
+      "grad_norm": 0.2877219319343567,
+      "learning_rate": 2.2942562436447704e-06,
+      "loss": 0.5769,
+      "step": 3792
+    },
+    {
+      "epoch": 1.5985878385499,
+      "grad_norm": 0.30419233441352844,
+      "learning_rate": 2.293140623000738e-06,
+      "loss": 0.5616,
+      "step": 3793
+    },
+    {
+      "epoch": 1.5990093792812732,
+      "grad_norm": 0.29149892926216125,
+      "learning_rate": 2.2920250438326035e-06,
+      "loss": 0.5925,
+      "step": 3794
+    },
+    {
+      "epoch": 1.5994309200126462,
+      "grad_norm": 0.3032289743423462,
+      "learning_rate": 2.2909095063640424e-06,
+      "loss": 0.5927,
+      "step": 3795
+    },
+    {
+      "epoch": 1.5998524607440194,
+      "grad_norm": 0.2908257842063904,
+      "learning_rate": 2.289794010818724e-06,
+      "loss": 0.5729,
+      "step": 3796
+    },
+    {
+      "epoch": 1.6002740014753924,
+      "grad_norm": 0.29688510298728943,
+      "learning_rate": 2.288678557420309e-06,
+      "loss": 0.5972,
+      "step": 3797
+    },
+    {
+      "epoch": 1.6006955422067657,
+      "grad_norm": 0.2891368269920349,
+      "learning_rate": 2.287563146392447e-06,
+      "loss": 0.5975,
+      "step": 3798
+    },
+    {
+      "epoch": 1.601117082938139,
+      "grad_norm": 0.29990509152412415,
+      "learning_rate": 2.2864477779587835e-06,
+      "loss": 0.5781,
+      "step": 3799
+    },
+    {
+      "epoch": 1.6015386236695122,
+      "grad_norm": 0.2963329553604126,
+      "learning_rate": 2.285332452342951e-06,
+      "loss": 0.5816,
+      "step": 3800
+    },
+    {
+      "epoch": 1.6019601644008854,
+      "grad_norm": 0.2944004535675049,
+      "learning_rate": 2.2842171697685765e-06,
+      "loss": 0.5904,
+      "step": 3801
+    },
+    {
+      "epoch": 1.6023817051322584,
+      "grad_norm": 0.3057761788368225,
+      "learning_rate": 2.283101930459277e-06,
+      "loss": 0.5873,
+      "step": 3802
+    },
+    {
+      "epoch": 1.6028032458636314,
+      "grad_norm": 0.29325196146965027,
+      "learning_rate": 2.281986734638662e-06,
+      "loss": 0.5814,
+      "step": 3803
+    },
+    {
+      "epoch": 1.6032247865950047,
+      "grad_norm": 0.2932264804840088,
+      "learning_rate": 2.2808715825303295e-06,
+      "loss": 0.6,
+      "step": 3804
+    },
+    {
+      "epoch": 1.603646327326378,
+      "grad_norm": 0.2938579022884369,
+      "learning_rate": 2.2797564743578736e-06,
+      "loss": 0.6014,
+      "step": 3805
+    },
+    {
+      "epoch": 1.6040678680577511,
+      "grad_norm": 0.3051855266094208,
+      "learning_rate": 2.278641410344874e-06,
+      "loss": 0.5847,
+      "step": 3806
+    },
+    {
+      "epoch": 1.6044894087891244,
+      "grad_norm": 0.2890417277812958,
+      "learning_rate": 2.2775263907149047e-06,
+      "loss": 0.5828,
+      "step": 3807
+    },
+    {
+      "epoch": 1.6049109495204974,
+      "grad_norm": 0.29760703444480896,
+      "learning_rate": 2.2764114156915325e-06,
+      "loss": 0.5987,
+      "step": 3808
+    },
+    {
+      "epoch": 1.6053324902518706,
+      "grad_norm": 0.2906395494937897,
+      "learning_rate": 2.2752964854983102e-06,
+      "loss": 0.6015,
+      "step": 3809
+    },
+    {
+      "epoch": 1.6057540309832437,
+      "grad_norm": 0.29845455288887024,
+      "learning_rate": 2.2741816003587866e-06,
+      "loss": 0.5799,
+      "step": 3810
+    },
+    {
+      "epoch": 1.606175571714617,
+      "grad_norm": 0.315492182970047,
+      "learning_rate": 2.273066760496498e-06,
+      "loss": 0.5838,
+      "step": 3811
+    },
+    {
+      "epoch": 1.6065971124459901,
+      "grad_norm": 0.29044076800346375,
+      "learning_rate": 2.2719519661349743e-06,
+      "loss": 0.568,
+      "step": 3812
+    },
+    {
+      "epoch": 1.6070186531773634,
+      "grad_norm": 0.29746198654174805,
+      "learning_rate": 2.270837217497733e-06,
+      "loss": 0.5885,
+      "step": 3813
+    },
+    {
+      "epoch": 1.6074401939087364,
+      "grad_norm": 0.29549309611320496,
+      "learning_rate": 2.2697225148082873e-06,
+      "loss": 0.5848,
+      "step": 3814
+    },
+    {
+      "epoch": 1.6078617346401096,
+      "grad_norm": 0.29666194319725037,
+      "learning_rate": 2.2686078582901343e-06,
+      "loss": 0.5785,
+      "step": 3815
+    },
+    {
+      "epoch": 1.6082832753714826,
+      "grad_norm": 0.2864087224006653,
+      "learning_rate": 2.2674932481667693e-06,
+      "loss": 0.5875,
+      "step": 3816
+    },
+    {
+      "epoch": 1.6087048161028559,
+      "grad_norm": 0.3055816888809204,
+      "learning_rate": 2.266378684661674e-06,
+      "loss": 0.5804,
+      "step": 3817
+    },
+    {
+      "epoch": 1.609126356834229,
+      "grad_norm": 0.3104031980037689,
+      "learning_rate": 2.2652641679983205e-06,
+      "loss": 0.5925,
+      "step": 3818
+    },
+    {
+      "epoch": 1.6095478975656023,
+      "grad_norm": 0.29787400364875793,
+      "learning_rate": 2.264149698400174e-06,
+      "loss": 0.5739,
+      "step": 3819
+    },
+    {
+      "epoch": 1.6099694382969756,
+      "grad_norm": 0.30618953704833984,
+      "learning_rate": 2.263035276090687e-06,
+      "loss": 0.5763,
+      "step": 3820
+    },
+    {
+      "epoch": 1.6103909790283486,
+      "grad_norm": 0.29939594864845276,
+      "learning_rate": 2.261920901293307e-06,
+      "loss": 0.5767,
+      "step": 3821
+    },
+    {
+      "epoch": 1.6108125197597216,
+      "grad_norm": 0.28965404629707336,
+      "learning_rate": 2.260806574231466e-06,
+      "loss": 0.5775,
+      "step": 3822
+    },
+    {
+      "epoch": 1.6112340604910949,
+      "grad_norm": 0.31564050912857056,
+      "learning_rate": 2.2596922951285934e-06,
+      "loss": 0.5937,
+      "step": 3823
+    },
+    {
+      "epoch": 1.611655601222468,
+      "grad_norm": 0.3066418170928955,
+      "learning_rate": 2.2585780642081023e-06,
+      "loss": 0.6035,
+      "step": 3824
+    },
+    {
+      "epoch": 1.6120771419538413,
+      "grad_norm": 0.29945212602615356,
+      "learning_rate": 2.2574638816934004e-06,
+      "loss": 0.5669,
+      "step": 3825
+    },
+    {
+      "epoch": 1.6124986826852146,
+      "grad_norm": 0.2994775176048279,
+      "learning_rate": 2.2563497478078852e-06,
+      "loss": 0.5995,
+      "step": 3826
+    },
+    {
+      "epoch": 1.6129202234165876,
+      "grad_norm": 0.29710012674331665,
+      "learning_rate": 2.2552356627749424e-06,
+      "loss": 0.6206,
+      "step": 3827
+    },
+    {
+      "epoch": 1.6133417641479608,
+      "grad_norm": 0.3029572665691376,
+      "learning_rate": 2.254121626817951e-06,
+      "loss": 0.5748,
+      "step": 3828
+    },
+    {
+      "epoch": 1.6137633048793338,
+      "grad_norm": 0.28740835189819336,
+      "learning_rate": 2.2530076401602756e-06,
+      "loss": 0.5915,
+      "step": 3829
+    },
+    {
+      "epoch": 1.614184845610707,
+      "grad_norm": 0.28898197412490845,
+      "learning_rate": 2.251893703025277e-06,
+      "loss": 0.5483,
+      "step": 3830
+    },
+    {
+      "epoch": 1.6146063863420803,
+      "grad_norm": 0.29065069556236267,
+      "learning_rate": 2.2507798156363e-06,
+      "loss": 0.5906,
+      "step": 3831
+    },
+    {
+      "epoch": 1.6150279270734536,
+      "grad_norm": 0.3085039258003235,
+      "learning_rate": 2.249665978216684e-06,
+      "loss": 0.5762,
+      "step": 3832
+    },
+    {
+      "epoch": 1.6154494678048268,
+      "grad_norm": 0.31288060545921326,
+      "learning_rate": 2.2485521909897554e-06,
+      "loss": 0.5716,
+      "step": 3833
+    },
+    {
+      "epoch": 1.6158710085361998,
+      "grad_norm": 0.308288037776947,
+      "learning_rate": 2.247438454178833e-06,
+      "loss": 0.5634,
+      "step": 3834
+    },
+    {
+      "epoch": 1.6162925492675728,
+      "grad_norm": 0.2829032242298126,
+      "learning_rate": 2.2463247680072227e-06,
+      "loss": 0.5924,
+      "step": 3835
+    },
+    {
+      "epoch": 1.616714089998946,
+      "grad_norm": 0.2934906780719757,
+      "learning_rate": 2.2452111326982222e-06,
+      "loss": 0.5673,
+      "step": 3836
+    },
+    {
+      "epoch": 1.6171356307303193,
+      "grad_norm": 0.2893316149711609,
+      "learning_rate": 2.244097548475121e-06,
+      "loss": 0.5775,
+      "step": 3837
+    },
+    {
+      "epoch": 1.6175571714616925,
+      "grad_norm": 0.2971586585044861,
+      "learning_rate": 2.2429840155611916e-06,
+      "loss": 0.5746,
+      "step": 3838
+    },
+    {
+      "epoch": 1.6179787121930658,
+      "grad_norm": 0.2980149984359741,
+      "learning_rate": 2.241870534179705e-06,
+      "loss": 0.5863,
+      "step": 3839
+    },
+    {
+      "epoch": 1.6184002529244388,
+      "grad_norm": 0.2978213429450989,
+      "learning_rate": 2.2407571045539134e-06,
+      "loss": 0.5908,
+      "step": 3840
+    },
+    {
+      "epoch": 1.618821793655812,
+      "grad_norm": 0.29787227511405945,
+      "learning_rate": 2.2396437269070653e-06,
+      "loss": 0.5917,
+      "step": 3841
+    },
+    {
+      "epoch": 1.619243334387185,
+      "grad_norm": 0.2959786355495453,
+      "learning_rate": 2.2385304014623947e-06,
+      "loss": 0.6067,
+      "step": 3842
+    },
+    {
+      "epoch": 1.6196648751185583,
+      "grad_norm": 0.2992161214351654,
+      "learning_rate": 2.2374171284431277e-06,
+      "loss": 0.6078,
+      "step": 3843
+    },
+    {
+      "epoch": 1.6200864158499315,
+      "grad_norm": 0.30526602268218994,
+      "learning_rate": 2.236303908072478e-06,
+      "loss": 0.5998,
+      "step": 3844
+    },
+    {
+      "epoch": 1.6205079565813048,
+      "grad_norm": 0.29370543360710144,
+      "learning_rate": 2.2351907405736485e-06,
+      "loss": 0.5864,
+      "step": 3845
+    },
+    {
+      "epoch": 1.620929497312678,
+      "grad_norm": 0.29793694615364075,
+      "learning_rate": 2.2340776261698355e-06,
+      "loss": 0.5778,
+      "step": 3846
+    },
+    {
+      "epoch": 1.621351038044051,
+      "grad_norm": 0.3073398470878601,
+      "learning_rate": 2.232964565084218e-06,
+      "loss": 0.5969,
+      "step": 3847
+    },
+    {
+      "epoch": 1.621772578775424,
+      "grad_norm": 0.28826257586479187,
+      "learning_rate": 2.2318515575399705e-06,
+      "loss": 0.5793,
+      "step": 3848
+    },
+    {
+      "epoch": 1.6221941195067973,
+      "grad_norm": 0.3048885464668274,
+      "learning_rate": 2.2307386037602528e-06,
+      "loss": 0.5509,
+      "step": 3849
+    },
+    {
+      "epoch": 1.6226156602381705,
+      "grad_norm": 0.300300270318985,
+      "learning_rate": 2.2296257039682166e-06,
+      "loss": 0.5854,
+      "step": 3850
+    },
+    {
+      "epoch": 1.6230372009695437,
+      "grad_norm": 0.3111071288585663,
+      "learning_rate": 2.228512858387e-06,
+      "loss": 0.5986,
+      "step": 3851
+    },
+    {
+      "epoch": 1.623458741700917,
+      "grad_norm": 0.2905791103839874,
+      "learning_rate": 2.227400067239733e-06,
+      "loss": 0.5786,
+      "step": 3852
+    },
+    {
+      "epoch": 1.62388028243229,
+      "grad_norm": 0.3035646378993988,
+      "learning_rate": 2.2262873307495324e-06,
+      "loss": 0.5741,
+      "step": 3853
+    },
+    {
+      "epoch": 1.6243018231636632,
+      "grad_norm": 0.3065578043460846,
+      "learning_rate": 2.225174649139506e-06,
+      "loss": 0.5865,
+      "step": 3854
+    },
+    {
+      "epoch": 1.6247233638950362,
+      "grad_norm": 0.29130521416664124,
+      "learning_rate": 2.2240620226327485e-06,
+      "loss": 0.5768,
+      "step": 3855
+    },
+    {
+      "epoch": 1.6251449046264095,
+      "grad_norm": 0.2965773940086365,
+      "learning_rate": 2.2229494514523447e-06,
+      "loss": 0.5851,
+      "step": 3856
+    },
+    {
+      "epoch": 1.6255664453577827,
+      "grad_norm": 0.30325639247894287,
+      "learning_rate": 2.2218369358213703e-06,
+      "loss": 0.593,
+      "step": 3857
+    },
+    {
+      "epoch": 1.625987986089156,
+      "grad_norm": 0.30288684368133545,
+      "learning_rate": 2.220724475962886e-06,
+      "loss": 0.5745,
+      "step": 3858
+    },
+    {
+      "epoch": 1.6264095268205292,
+      "grad_norm": 0.3047354817390442,
+      "learning_rate": 2.219612072099944e-06,
+      "loss": 0.5767,
+      "step": 3859
+    },
+    {
+      "epoch": 1.6268310675519022,
+      "grad_norm": 0.30205976963043213,
+      "learning_rate": 2.218499724455584e-06,
+      "loss": 0.5798,
+      "step": 3860
+    },
+    {
+      "epoch": 1.6272526082832752,
+      "grad_norm": 0.305972158908844,
+      "learning_rate": 2.2173874332528357e-06,
+      "loss": 0.6062,
+      "step": 3861
+    },
+    {
+      "epoch": 1.6276741490146485,
+      "grad_norm": 0.2932242751121521,
+      "learning_rate": 2.2162751987147153e-06,
+      "loss": 0.595,
+      "step": 3862
+    },
+    {
+      "epoch": 1.6280956897460217,
+      "grad_norm": 0.30090704560279846,
+      "learning_rate": 2.21516302106423e-06,
+      "loss": 0.5971,
+      "step": 3863
+    },
+    {
+      "epoch": 1.628517230477395,
+      "grad_norm": 0.30456215143203735,
+      "learning_rate": 2.214050900524374e-06,
+      "loss": 0.5817,
+      "step": 3864
+    },
+    {
+      "epoch": 1.6289387712087682,
+      "grad_norm": 0.27835866808891296,
+      "learning_rate": 2.2129388373181305e-06,
+      "loss": 0.6125,
+      "step": 3865
+    },
+    {
+      "epoch": 1.6293603119401412,
+      "grad_norm": 0.29025998711586,
+      "learning_rate": 2.211826831668473e-06,
+      "loss": 0.5719,
+      "step": 3866
+    },
+    {
+      "epoch": 1.6297818526715144,
+      "grad_norm": 0.31984061002731323,
+      "learning_rate": 2.21071488379836e-06,
+      "loss": 0.5822,
+      "step": 3867
+    },
+    {
+      "epoch": 1.6302033934028874,
+      "grad_norm": 0.29266807436943054,
+      "learning_rate": 2.2096029939307416e-06,
+      "loss": 0.5778,
+      "step": 3868
+    },
+    {
+      "epoch": 1.6306249341342607,
+      "grad_norm": 0.3130320906639099,
+      "learning_rate": 2.2084911622885532e-06,
+      "loss": 0.5789,
+      "step": 3869
+    },
+    {
+      "epoch": 1.631046474865634,
+      "grad_norm": 0.30349254608154297,
+      "learning_rate": 2.2073793890947223e-06,
+      "loss": 0.6035,
+      "step": 3870
+    },
+    {
+      "epoch": 1.6314680155970072,
+      "grad_norm": 0.3037419319152832,
+      "learning_rate": 2.206267674572161e-06,
+      "loss": 0.6093,
+      "step": 3871
+    },
+    {
+      "epoch": 1.6318895563283802,
+      "grad_norm": 0.29563069343566895,
+      "learning_rate": 2.2051560189437717e-06,
+      "loss": 0.6143,
+      "step": 3872
+    },
+    {
+      "epoch": 1.6323110970597534,
+      "grad_norm": 0.2982088327407837,
+      "learning_rate": 2.204044422432444e-06,
+      "loss": 0.5952,
+      "step": 3873
+    },
+    {
+      "epoch": 1.6327326377911264,
+      "grad_norm": 0.29970020055770874,
+      "learning_rate": 2.2029328852610584e-06,
+      "loss": 0.6268,
+      "step": 3874
+    },
+    {
+      "epoch": 1.6331541785224997,
+      "grad_norm": 0.292939156293869,
+      "learning_rate": 2.201821407652478e-06,
+      "loss": 0.6241,
+      "step": 3875
+    },
+    {
+      "epoch": 1.633575719253873,
+      "grad_norm": 0.3019394278526306,
+      "learning_rate": 2.2007099898295596e-06,
+      "loss": 0.5756,
+      "step": 3876
+    },
+    {
+      "epoch": 1.6339972599852461,
+      "grad_norm": 0.294025182723999,
+      "learning_rate": 2.1995986320151456e-06,
+      "loss": 0.5854,
+      "step": 3877
+    },
+    {
+      "epoch": 1.6344188007166194,
+      "grad_norm": 0.294416218996048,
+      "learning_rate": 2.198487334432065e-06,
+      "loss": 0.5869,
+      "step": 3878
+    },
+    {
+      "epoch": 1.6348403414479924,
+      "grad_norm": 0.30028459429740906,
+      "learning_rate": 2.1973760973031373e-06,
+      "loss": 0.6039,
+      "step": 3879
+    },
+    {
+      "epoch": 1.6352618821793656,
+      "grad_norm": 0.30984142422676086,
+      "learning_rate": 2.1962649208511684e-06,
+      "loss": 0.5775,
+      "step": 3880
+    },
+    {
+      "epoch": 1.6356834229107386,
+      "grad_norm": 0.30308935046195984,
+      "learning_rate": 2.1951538052989524e-06,
+      "loss": 0.5674,
+      "step": 3881
+    },
+    {
+      "epoch": 1.6361049636421119,
+      "grad_norm": 0.30183571577072144,
+      "learning_rate": 2.19404275086927e-06,
+      "loss": 0.582,
+      "step": 3882
+    },
+    {
+      "epoch": 1.6365265043734851,
+      "grad_norm": 0.30629250407218933,
+      "learning_rate": 2.192931757784894e-06,
+      "loss": 0.6118,
+      "step": 3883
+    },
+    {
+      "epoch": 1.6369480451048584,
+      "grad_norm": 0.29491934180259705,
+      "learning_rate": 2.1918208262685775e-06,
+      "loss": 0.6088,
+      "step": 3884
+    },
+    {
+      "epoch": 1.6373695858362314,
+      "grad_norm": 0.3085861802101135,
+      "learning_rate": 2.190709956543068e-06,
+      "loss": 0.6131,
+      "step": 3885
+    },
+    {
+      "epoch": 1.6377911265676046,
+      "grad_norm": 0.2829180359840393,
+      "learning_rate": 2.1895991488310986e-06,
+      "loss": 0.603,
+      "step": 3886
+    },
+    {
+      "epoch": 1.6382126672989776,
+      "grad_norm": 0.2952893078327179,
+      "learning_rate": 2.1884884033553873e-06,
+      "loss": 0.6079,
+      "step": 3887
+    },
+    {
+      "epoch": 1.6386342080303509,
+      "grad_norm": 0.28901663422584534,
+      "learning_rate": 2.187377720338644e-06,
+      "loss": 0.5784,
+      "step": 3888
+    },
+    {
+      "epoch": 1.639055748761724,
+      "grad_norm": 0.3053705394268036,
+      "learning_rate": 2.186267100003562e-06,
+      "loss": 0.5835,
+      "step": 3889
+    },
+    {
+      "epoch": 1.6394772894930973,
+      "grad_norm": 0.29146575927734375,
+      "learning_rate": 2.185156542572824e-06,
+      "loss": 0.5819,
+      "step": 3890
+    },
+    {
+      "epoch": 1.6398988302244706,
+      "grad_norm": 0.3044636845588684,
+      "learning_rate": 2.184046048269101e-06,
+      "loss": 0.5512,
+      "step": 3891
+    },
+    {
+      "epoch": 1.6403203709558436,
+      "grad_norm": 0.28794920444488525,
+      "learning_rate": 2.1829356173150503e-06,
+      "loss": 0.5735,
+      "step": 3892
+    },
+    {
+      "epoch": 1.6407419116872166,
+      "grad_norm": 0.30698829889297485,
+      "learning_rate": 2.1818252499333147e-06,
+      "loss": 0.5676,
+      "step": 3893
+    },
+    {
+      "epoch": 1.6411634524185899,
+      "grad_norm": 0.2976018786430359,
+      "learning_rate": 2.180714946346528e-06,
+      "loss": 0.5829,
+      "step": 3894
+    },
+    {
+      "epoch": 1.641584993149963,
+      "grad_norm": 0.30707380175590515,
+      "learning_rate": 2.1796047067773088e-06,
+      "loss": 0.6193,
+      "step": 3895
+    },
+    {
+      "epoch": 1.6420065338813363,
+      "grad_norm": 0.3050495684146881,
+      "learning_rate": 2.1784945314482627e-06,
+      "loss": 0.5874,
+      "step": 3896
+    },
+    {
+      "epoch": 1.6424280746127096,
+      "grad_norm": 0.2902907431125641,
+      "learning_rate": 2.177384420581984e-06,
+      "loss": 0.5842,
+      "step": 3897
+    },
+    {
+      "epoch": 1.6428496153440826,
+      "grad_norm": 0.31011393666267395,
+      "learning_rate": 2.176274374401051e-06,
+      "loss": 0.5813,
+      "step": 3898
+    },
+    {
+      "epoch": 1.6432711560754558,
+      "grad_norm": 0.2995331585407257,
+      "learning_rate": 2.1751643931280344e-06,
+      "loss": 0.5883,
+      "step": 3899
+    },
+    {
+      "epoch": 1.6436926968068288,
+      "grad_norm": 0.28671976923942566,
+      "learning_rate": 2.1740544769854852e-06,
+      "loss": 0.5727,
+      "step": 3900
+    },
+    {
+      "epoch": 1.644114237538202,
+      "grad_norm": 0.2962295114994049,
+      "learning_rate": 2.172944626195948e-06,
+      "loss": 0.595,
+      "step": 3901
+    },
+    {
+      "epoch": 1.6445357782695753,
+      "grad_norm": 0.30610278248786926,
+      "learning_rate": 2.1718348409819482e-06,
+      "loss": 0.5857,
+      "step": 3902
+    },
+    {
+      "epoch": 1.6449573190009485,
+      "grad_norm": 0.29738277196884155,
+      "learning_rate": 2.170725121566003e-06,
+      "loss": 0.589,
+      "step": 3903
+    },
+    {
+      "epoch": 1.6453788597323218,
+      "grad_norm": 0.2954496145248413,
+      "learning_rate": 2.169615468170613e-06,
+      "loss": 0.5991,
+      "step": 3904
+    },
+    {
+      "epoch": 1.6458004004636948,
+      "grad_norm": 0.2987905740737915,
+      "learning_rate": 2.1685058810182672e-06,
+      "loss": 0.6006,
+      "step": 3905
+    },
+    {
+      "epoch": 1.6462219411950678,
+      "grad_norm": 0.2876895070075989,
+      "learning_rate": 2.167396360331442e-06,
+      "loss": 0.5873,
+      "step": 3906
+    },
+    {
+      "epoch": 1.646643481926441,
+      "grad_norm": 0.28600555658340454,
+      "learning_rate": 2.1662869063325977e-06,
+      "loss": 0.5789,
+      "step": 3907
+    },
+    {
+      "epoch": 1.6470650226578143,
+      "grad_norm": 0.2975482940673828,
+      "learning_rate": 2.165177519244185e-06,
+      "loss": 0.5814,
+      "step": 3908
+    },
+    {
+      "epoch": 1.6474865633891875,
+      "grad_norm": 0.30580130219459534,
+      "learning_rate": 2.164068199288637e-06,
+      "loss": 0.5945,
+      "step": 3909
+    },
+    {
+      "epoch": 1.6479081041205608,
+      "grad_norm": 0.29474562406539917,
+      "learning_rate": 2.1629589466883776e-06,
+      "loss": 0.5868,
+      "step": 3910
+    },
+    {
+      "epoch": 1.6483296448519338,
+      "grad_norm": 0.29148367047309875,
+      "learning_rate": 2.161849761665814e-06,
+      "loss": 0.6034,
+      "step": 3911
+    },
+    {
+      "epoch": 1.648751185583307,
+      "grad_norm": 0.32984310388565063,
+      "learning_rate": 2.160740644443341e-06,
+      "loss": 0.5971,
+      "step": 3912
+    },
+    {
+      "epoch": 1.64917272631468,
+      "grad_norm": 0.3104466497898102,
+      "learning_rate": 2.15963159524334e-06,
+      "loss": 0.5891,
+      "step": 3913
+    },
+    {
+      "epoch": 1.6495942670460533,
+      "grad_norm": 0.3005678653717041,
+      "learning_rate": 2.158522614288178e-06,
+      "loss": 0.5947,
+      "step": 3914
+    },
+    {
+      "epoch": 1.6500158077774265,
+      "grad_norm": 0.3075987696647644,
+      "learning_rate": 2.15741370180021e-06,
+      "loss": 0.5879,
+      "step": 3915
+    },
+    {
+      "epoch": 1.6504373485087998,
+      "grad_norm": 0.329988569021225,
+      "learning_rate": 2.156304858001775e-06,
+      "loss": 0.5539,
+      "step": 3916
+    },
+    {
+      "epoch": 1.650858889240173,
+      "grad_norm": 0.30327606201171875,
+      "learning_rate": 2.1551960831152007e-06,
+      "loss": 0.5881,
+      "step": 3917
+    },
+    {
+      "epoch": 1.651280429971546,
+      "grad_norm": 0.2988946735858917,
+      "learning_rate": 2.154087377362798e-06,
+      "loss": 0.6159,
+      "step": 3918
+    },
+    {
+      "epoch": 1.651701970702919,
+      "grad_norm": 0.3363494575023651,
+      "learning_rate": 2.1529787409668664e-06,
+      "loss": 0.5877,
+      "step": 3919
+    },
+    {
+      "epoch": 1.6521235114342923,
+      "grad_norm": 0.2932268977165222,
+      "learning_rate": 2.1518701741496907e-06,
+      "loss": 0.5705,
+      "step": 3920
+    },
+    {
+      "epoch": 1.6525450521656655,
+      "grad_norm": 0.318867027759552,
+      "learning_rate": 2.150761677133542e-06,
+      "loss": 0.5704,
+      "step": 3921
+    },
+    {
+      "epoch": 1.6529665928970387,
+      "grad_norm": 0.2991884648799896,
+      "learning_rate": 2.149653250140676e-06,
+      "loss": 0.5623,
+      "step": 3922
+    },
+    {
+      "epoch": 1.653388133628412,
+      "grad_norm": 0.3008774518966675,
+      "learning_rate": 2.148544893393337e-06,
+      "loss": 0.6039,
+      "step": 3923
+    },
+    {
+      "epoch": 1.653809674359785,
+      "grad_norm": 0.3129323720932007,
+      "learning_rate": 2.1474366071137526e-06,
+      "loss": 0.5911,
+      "step": 3924
+    },
+    {
+      "epoch": 1.6542312150911582,
+      "grad_norm": 0.3394458591938019,
+      "learning_rate": 2.1463283915241366e-06,
+      "loss": 0.5923,
+      "step": 3925
+    },
+    {
+      "epoch": 1.6546527558225312,
+      "grad_norm": 0.30955809354782104,
+      "learning_rate": 2.145220246846692e-06,
+      "loss": 0.582,
+      "step": 3926
+    },
+    {
+      "epoch": 1.6550742965539045,
+      "grad_norm": 0.29566019773483276,
+      "learning_rate": 2.144112173303603e-06,
+      "loss": 0.5924,
+      "step": 3927
+    },
+    {
+      "epoch": 1.6554958372852777,
+      "grad_norm": 0.3119506537914276,
+      "learning_rate": 2.1430041711170422e-06,
+      "loss": 0.5962,
+      "step": 3928
+    },
+    {
+      "epoch": 1.655917378016651,
+      "grad_norm": 0.36199161410331726,
+      "learning_rate": 2.141896240509167e-06,
+      "loss": 0.5731,
+      "step": 3929
+    },
+    {
+      "epoch": 1.6563389187480242,
+      "grad_norm": 0.34967562556266785,
+      "learning_rate": 2.140788381702121e-06,
+      "loss": 0.5789,
+      "step": 3930
+    },
+    {
+      "epoch": 1.6567604594793972,
+      "grad_norm": 0.313031941652298,
+      "learning_rate": 2.1396805949180322e-06,
+      "loss": 0.5781,
+      "step": 3931
+    },
+    {
+      "epoch": 1.6571820002107702,
+      "grad_norm": 0.3069029748439789,
+      "learning_rate": 2.138572880379016e-06,
+      "loss": 0.593,
+      "step": 3932
+    },
+    {
+      "epoch": 1.6576035409421435,
+      "grad_norm": 0.34837937355041504,
+      "learning_rate": 2.1374652383071716e-06,
+      "loss": 0.5619,
+      "step": 3933
+    },
+    {
+      "epoch": 1.6580250816735167,
+      "grad_norm": 0.34781119227409363,
+      "learning_rate": 2.1363576689245836e-06,
+      "loss": 0.5703,
+      "step": 3934
+    },
+    {
+      "epoch": 1.65844662240489,
+      "grad_norm": 0.3035370707511902,
+      "learning_rate": 2.1352501724533255e-06,
+      "loss": 0.5641,
+      "step": 3935
+    },
+    {
+      "epoch": 1.6588681631362632,
+      "grad_norm": 0.303160697221756,
+      "learning_rate": 2.1341427491154514e-06,
+      "loss": 0.5878,
+      "step": 3936
+    },
+    {
+      "epoch": 1.6592897038676362,
+      "grad_norm": 0.2868005931377411,
+      "learning_rate": 2.133035399133004e-06,
+      "loss": 0.601,
+      "step": 3937
+    },
+    {
+      "epoch": 1.6597112445990094,
+      "grad_norm": 0.33757805824279785,
+      "learning_rate": 2.1319281227280087e-06,
+      "loss": 0.5704,
+      "step": 3938
+    },
+    {
+      "epoch": 1.6601327853303824,
+      "grad_norm": 0.322021484375,
+      "learning_rate": 2.1308209201224793e-06,
+      "loss": 0.6115,
+      "step": 3939
+    },
+    {
+      "epoch": 1.6605543260617557,
+      "grad_norm": 0.2929777204990387,
+      "learning_rate": 2.1297137915384113e-06,
+      "loss": 0.585,
+      "step": 3940
+    },
+    {
+      "epoch": 1.660975866793129,
+      "grad_norm": 0.31446224451065063,
+      "learning_rate": 2.1286067371977886e-06,
+      "loss": 0.5705,
+      "step": 3941
+    },
+    {
+      "epoch": 1.6613974075245022,
+      "grad_norm": 0.3344728350639343,
+      "learning_rate": 2.127499757322578e-06,
+      "loss": 0.5795,
+      "step": 3942
+    },
+    {
+      "epoch": 1.6618189482558752,
+      "grad_norm": 0.3101010024547577,
+      "learning_rate": 2.126392852134733e-06,
+      "loss": 0.5883,
+      "step": 3943
+    },
+    {
+      "epoch": 1.6622404889872484,
+      "grad_norm": 0.29196396470069885,
+      "learning_rate": 2.125286021856189e-06,
+      "loss": 0.5667,
+      "step": 3944
+    },
+    {
+      "epoch": 1.6626620297186214,
+      "grad_norm": 0.297136127948761,
+      "learning_rate": 2.124179266708871e-06,
+      "loss": 0.5865,
+      "step": 3945
+    },
+    {
+      "epoch": 1.6630835704499947,
+      "grad_norm": 0.31624647974967957,
+      "learning_rate": 2.1230725869146867e-06,
+      "loss": 0.5756,
+      "step": 3946
+    },
+    {
+      "epoch": 1.663505111181368,
+      "grad_norm": 0.3036721348762512,
+      "learning_rate": 2.121965982695527e-06,
+      "loss": 0.5929,
+      "step": 3947
+    },
+    {
+      "epoch": 1.6639266519127411,
+      "grad_norm": 0.30356845259666443,
+      "learning_rate": 2.1208594542732706e-06,
+      "loss": 0.5881,
+      "step": 3948
+    },
+    {
+      "epoch": 1.6643481926441144,
+      "grad_norm": 0.29719191789627075,
+      "learning_rate": 2.119753001869778e-06,
+      "loss": 0.5963,
+      "step": 3949
+    },
+    {
+      "epoch": 1.6647697333754874,
+      "grad_norm": 0.31534871459007263,
+      "learning_rate": 2.118646625706898e-06,
+      "loss": 0.5631,
+      "step": 3950
+    },
+    {
+      "epoch": 1.6651912741068606,
+      "grad_norm": 0.31381848454475403,
+      "learning_rate": 2.11754032600646e-06,
+      "loss": 0.5874,
+      "step": 3951
+    },
+    {
+      "epoch": 1.6656128148382336,
+      "grad_norm": 0.3132377862930298,
+      "learning_rate": 2.1164341029902827e-06,
+      "loss": 0.5886,
+      "step": 3952
+    },
+    {
+      "epoch": 1.6660343555696069,
+      "grad_norm": 0.28571105003356934,
+      "learning_rate": 2.1153279568801643e-06,
+      "loss": 0.5487,
+      "step": 3953
+    },
+    {
+      "epoch": 1.6664558963009801,
+      "grad_norm": 0.3048253357410431,
+      "learning_rate": 2.1142218878978925e-06,
+      "loss": 0.5683,
+      "step": 3954
+    },
+    {
+      "epoch": 1.6668774370323534,
+      "grad_norm": 0.31735092401504517,
+      "learning_rate": 2.113115896265237e-06,
+      "loss": 0.5639,
+      "step": 3955
+    },
+    {
+      "epoch": 1.6672989777637264,
+      "grad_norm": 0.30613574385643005,
+      "learning_rate": 2.112009982203951e-06,
+      "loss": 0.5705,
+      "step": 3956
+    },
+    {
+      "epoch": 1.6677205184950996,
+      "grad_norm": 0.28592216968536377,
+      "learning_rate": 2.110904145935775e-06,
+      "loss": 0.5984,
+      "step": 3957
+    },
+    {
+      "epoch": 1.6681420592264726,
+      "grad_norm": 0.301405131816864,
+      "learning_rate": 2.1097983876824306e-06,
+      "loss": 0.562,
+      "step": 3958
+    },
+    {
+      "epoch": 1.6685635999578459,
+      "grad_norm": 0.3194476068019867,
+      "learning_rate": 2.1086927076656274e-06,
+      "loss": 0.5468,
+      "step": 3959
+    },
+    {
+      "epoch": 1.668985140689219,
+      "grad_norm": 0.2987363934516907,
+      "learning_rate": 2.107587106107055e-06,
+      "loss": 0.5524,
+      "step": 3960
+    },
+    {
+      "epoch": 1.6694066814205923,
+      "grad_norm": 0.30755314230918884,
+      "learning_rate": 2.1064815832283934e-06,
+      "loss": 0.5784,
+      "step": 3961
+    },
+    {
+      "epoch": 1.6698282221519656,
+      "grad_norm": 0.297219455242157,
+      "learning_rate": 2.1053761392512987e-06,
+      "loss": 0.6126,
+      "step": 3962
+    },
+    {
+      "epoch": 1.6702497628833386,
+      "grad_norm": 0.29891371726989746,
+      "learning_rate": 2.1042707743974185e-06,
+      "loss": 0.572,
+      "step": 3963
+    },
+    {
+      "epoch": 1.6706713036147116,
+      "grad_norm": 0.29837003350257874,
+      "learning_rate": 2.103165488888381e-06,
+      "loss": 0.5843,
+      "step": 3964
+    },
+    {
+      "epoch": 1.6710928443460848,
+      "grad_norm": 0.2902466058731079,
+      "learning_rate": 2.102060282945799e-06,
+      "loss": 0.5944,
+      "step": 3965
+    },
+    {
+      "epoch": 1.671514385077458,
+      "grad_norm": 0.3082202970981598,
+      "learning_rate": 2.1009551567912696e-06,
+      "loss": 0.569,
+      "step": 3966
+    },
+    {
+      "epoch": 1.6719359258088313,
+      "grad_norm": 0.2978470027446747,
+      "learning_rate": 2.0998501106463726e-06,
+      "loss": 0.5551,
+      "step": 3967
+    },
+    {
+      "epoch": 1.6723574665402046,
+      "grad_norm": 0.28995707631111145,
+      "learning_rate": 2.0987451447326752e-06,
+      "loss": 0.5803,
+      "step": 3968
+    },
+    {
+      "epoch": 1.6727790072715776,
+      "grad_norm": 0.3268853425979614,
+      "learning_rate": 2.0976402592717236e-06,
+      "loss": 0.565,
+      "step": 3969
+    },
+    {
+      "epoch": 1.6732005480029508,
+      "grad_norm": 0.30212825536727905,
+      "learning_rate": 2.0965354544850532e-06,
+      "loss": 0.5728,
+      "step": 3970
+    },
+    {
+      "epoch": 1.6736220887343238,
+      "grad_norm": 0.2952486276626587,
+      "learning_rate": 2.0954307305941777e-06,
+      "loss": 0.6007,
+      "step": 3971
+    },
+    {
+      "epoch": 1.674043629465697,
+      "grad_norm": 0.2981153726577759,
+      "learning_rate": 2.0943260878206e-06,
+      "loss": 0.5733,
+      "step": 3972
+    },
+    {
+      "epoch": 1.6744651701970703,
+      "grad_norm": 0.3174111247062683,
+      "learning_rate": 2.093221526385802e-06,
+      "loss": 0.5739,
+      "step": 3973
+    },
+    {
+      "epoch": 1.6748867109284435,
+      "grad_norm": 0.30132320523262024,
+      "learning_rate": 2.0921170465112526e-06,
+      "loss": 0.5639,
+      "step": 3974
+    },
+    {
+      "epoch": 1.6753082516598168,
+      "grad_norm": 0.28493696451187134,
+      "learning_rate": 2.091012648418404e-06,
+      "loss": 0.5804,
+      "step": 3975
+    },
+    {
+      "epoch": 1.6757297923911898,
+      "grad_norm": 0.3002147376537323,
+      "learning_rate": 2.0899083323286884e-06,
+      "loss": 0.6014,
+      "step": 3976
+    },
+    {
+      "epoch": 1.6761513331225628,
+      "grad_norm": 0.3044129014015198,
+      "learning_rate": 2.088804098463528e-06,
+      "loss": 0.5806,
+      "step": 3977
+    },
+    {
+      "epoch": 1.676572873853936,
+      "grad_norm": 0.3006686866283417,
+      "learning_rate": 2.087699947044321e-06,
+      "loss": 0.6086,
+      "step": 3978
+    },
+    {
+      "epoch": 1.6769944145853093,
+      "grad_norm": 0.3045450747013092,
+      "learning_rate": 2.086595878292456e-06,
+      "loss": 0.5916,
+      "step": 3979
+    },
+    {
+      "epoch": 1.6774159553166825,
+      "grad_norm": 0.3037409484386444,
+      "learning_rate": 2.085491892429301e-06,
+      "loss": 0.5794,
+      "step": 3980
+    },
+    {
+      "epoch": 1.6778374960480558,
+      "grad_norm": 0.28063368797302246,
+      "learning_rate": 2.084387989676208e-06,
+      "loss": 0.58,
+      "step": 3981
+    },
+    {
+      "epoch": 1.6782590367794288,
+      "grad_norm": 0.30729636549949646,
+      "learning_rate": 2.0832841702545122e-06,
+      "loss": 0.5705,
+      "step": 3982
+    },
+    {
+      "epoch": 1.678680577510802,
+      "grad_norm": 0.30675259232521057,
+      "learning_rate": 2.0821804343855335e-06,
+      "loss": 0.5787,
+      "step": 3983
+    },
+    {
+      "epoch": 1.679102118242175,
+      "grad_norm": 0.30647966265678406,
+      "learning_rate": 2.081076782290574e-06,
+      "loss": 0.591,
+      "step": 3984
+    },
+    {
+      "epoch": 1.6795236589735483,
+      "grad_norm": 0.29446524381637573,
+      "learning_rate": 2.079973214190918e-06,
+      "loss": 0.5709,
+      "step": 3985
+    },
+    {
+      "epoch": 1.6799451997049215,
+      "grad_norm": 0.2951411306858063,
+      "learning_rate": 2.078869730307836e-06,
+      "loss": 0.6104,
+      "step": 3986
+    },
+    {
+      "epoch": 1.6803667404362947,
+      "grad_norm": 0.301805704832077,
+      "learning_rate": 2.0777663308625776e-06,
+      "loss": 0.5822,
+      "step": 3987
+    },
+    {
+      "epoch": 1.680788281167668,
+      "grad_norm": 0.31825023889541626,
+      "learning_rate": 2.076663016076379e-06,
+      "loss": 0.5752,
+      "step": 3988
+    },
+    {
+      "epoch": 1.681209821899041,
+      "grad_norm": 0.29153966903686523,
+      "learning_rate": 2.075559786170458e-06,
+      "loss": 0.5774,
+      "step": 3989
+    },
+    {
+      "epoch": 1.681631362630414,
+      "grad_norm": 0.29823431372642517,
+      "learning_rate": 2.0744566413660143e-06,
+      "loss": 0.562,
+      "step": 3990
+    },
+    {
+      "epoch": 1.6820529033617873,
+      "grad_norm": 0.31415385007858276,
+      "learning_rate": 2.0733535818842325e-06,
+      "loss": 0.594,
+      "step": 3991
+    },
+    {
+      "epoch": 1.6824744440931605,
+      "grad_norm": 0.2983478903770447,
+      "learning_rate": 2.0722506079462787e-06,
+      "loss": 0.6076,
+      "step": 3992
+    },
+    {
+      "epoch": 1.6828959848245337,
+      "grad_norm": 0.28752991557121277,
+      "learning_rate": 2.071147719773302e-06,
+      "loss": 0.5668,
+      "step": 3993
+    },
+    {
+      "epoch": 1.683317525555907,
+      "grad_norm": 0.2957659363746643,
+      "learning_rate": 2.070044917586435e-06,
+      "loss": 0.5909,
+      "step": 3994
+    },
+    {
+      "epoch": 1.68373906628728,
+      "grad_norm": 0.29023295640945435,
+      "learning_rate": 2.0689422016067944e-06,
+      "loss": 0.5571,
+      "step": 3995
+    },
+    {
+      "epoch": 1.6841606070186532,
+      "grad_norm": 0.30019110441207886,
+      "learning_rate": 2.067839572055474e-06,
+      "loss": 0.5987,
+      "step": 3996
+    },
+    {
+      "epoch": 1.6845821477500262,
+      "grad_norm": 0.30265113711357117,
+      "learning_rate": 2.066737029153558e-06,
+      "loss": 0.5538,
+      "step": 3997
+    },
+    {
+      "epoch": 1.6850036884813995,
+      "grad_norm": 0.2957308292388916,
+      "learning_rate": 2.065634573122108e-06,
+      "loss": 0.5904,
+      "step": 3998
+    },
+    {
+      "epoch": 1.6854252292127727,
+      "grad_norm": 0.29527825117111206,
+      "learning_rate": 2.064532204182169e-06,
+      "loss": 0.5901,
+      "step": 3999
+    },
+    {
+      "epoch": 1.685846769944146,
+      "grad_norm": 0.2964383065700531,
+      "learning_rate": 2.0634299225547697e-06,
+      "loss": 0.5892,
+      "step": 4000
+    },
+    {
+      "epoch": 1.686268310675519,
+      "grad_norm": 0.3028421401977539,
+      "learning_rate": 2.062327728460921e-06,
+      "loss": 0.5732,
+      "step": 4001
+    },
+    {
+      "epoch": 1.6866898514068922,
+      "grad_norm": 0.3145774304866791,
+      "learning_rate": 2.061225622121615e-06,
+      "loss": 0.5634,
+      "step": 4002
+    },
+    {
+      "epoch": 1.6871113921382652,
+      "grad_norm": 0.29063668847084045,
+      "learning_rate": 2.0601236037578274e-06,
+      "loss": 0.5832,
+      "step": 4003
+    },
+    {
+      "epoch": 1.6875329328696385,
+      "grad_norm": 0.2931377589702606,
+      "learning_rate": 2.0590216735905173e-06,
+      "loss": 0.5965,
+      "step": 4004
+    },
+    {
+      "epoch": 1.6879544736010117,
+      "grad_norm": 0.2969213128089905,
+      "learning_rate": 2.0579198318406237e-06,
+      "loss": 0.6014,
+      "step": 4005
+    },
+    {
+      "epoch": 1.688376014332385,
+      "grad_norm": 0.2928997874259949,
+      "learning_rate": 2.0568180787290697e-06,
+      "loss": 0.5726,
+      "step": 4006
+    },
+    {
+      "epoch": 1.6887975550637582,
+      "grad_norm": 0.3018783926963806,
+      "learning_rate": 2.055716414476759e-06,
+      "loss": 0.5568,
+      "step": 4007
+    },
+    {
+      "epoch": 1.6892190957951312,
+      "grad_norm": 0.28472328186035156,
+      "learning_rate": 2.05461483930458e-06,
+      "loss": 0.596,
+      "step": 4008
+    },
+    {
+      "epoch": 1.6896406365265044,
+      "grad_norm": 0.2979946434497833,
+      "learning_rate": 2.0535133534333996e-06,
+      "loss": 0.5674,
+      "step": 4009
+    },
+    {
+      "epoch": 1.6900621772578774,
+      "grad_norm": 0.2890856862068176,
+      "learning_rate": 2.0524119570840702e-06,
+      "loss": 0.5495,
+      "step": 4010
+    },
+    {
+      "epoch": 1.6904837179892507,
+      "grad_norm": 0.2945740520954132,
+      "learning_rate": 2.0513106504774247e-06,
+      "loss": 0.5623,
+      "step": 4011
+    },
+    {
+      "epoch": 1.690905258720624,
+      "grad_norm": 0.31134259700775146,
+      "learning_rate": 2.0502094338342784e-06,
+      "loss": 0.5832,
+      "step": 4012
+    },
+    {
+      "epoch": 1.6913267994519972,
+      "grad_norm": 0.2912856638431549,
+      "learning_rate": 2.049108307375427e-06,
+      "loss": 0.592,
+      "step": 4013
+    },
+    {
+      "epoch": 1.6917483401833702,
+      "grad_norm": 0.2881820797920227,
+      "learning_rate": 2.048007271321652e-06,
+      "loss": 0.5782,
+      "step": 4014
+    },
+    {
+      "epoch": 1.6921698809147434,
+      "grad_norm": 0.31992286443710327,
+      "learning_rate": 2.0469063258937127e-06,
+      "loss": 0.5647,
+      "step": 4015
+    },
+    {
+      "epoch": 1.6925914216461164,
+      "grad_norm": 0.2918142080307007,
+      "learning_rate": 2.045805471312352e-06,
+      "loss": 0.5972,
+      "step": 4016
+    },
+    {
+      "epoch": 1.6930129623774897,
+      "grad_norm": 0.28819340467453003,
+      "learning_rate": 2.0447047077982946e-06,
+      "loss": 0.6007,
+      "step": 4017
+    },
+    {
+      "epoch": 1.693434503108863,
+      "grad_norm": 0.31619593501091003,
+      "learning_rate": 2.0436040355722463e-06,
+      "loss": 0.6094,
+      "step": 4018
+    },
+    {
+      "epoch": 1.6938560438402361,
+      "grad_norm": 0.2969091236591339,
+      "learning_rate": 2.042503454854896e-06,
+      "loss": 0.6064,
+      "step": 4019
+    },
+    {
+      "epoch": 1.6942775845716094,
+      "grad_norm": 0.3207230567932129,
+      "learning_rate": 2.0414029658669112e-06,
+      "loss": 0.6018,
+      "step": 4020
+    },
+    {
+      "epoch": 1.6946991253029824,
+      "grad_norm": 0.3042408525943756,
+      "learning_rate": 2.0403025688289464e-06,
+      "loss": 0.5901,
+      "step": 4021
+    },
+    {
+      "epoch": 1.6951206660343554,
+      "grad_norm": 0.28905197978019714,
+      "learning_rate": 2.0392022639616307e-06,
+      "loss": 0.5763,
+      "step": 4022
+    },
+    {
+      "epoch": 1.6955422067657286,
+      "grad_norm": 0.2964577376842499,
+      "learning_rate": 2.0381020514855806e-06,
+      "loss": 0.5787,
+      "step": 4023
+    },
+    {
+      "epoch": 1.6959637474971019,
+      "grad_norm": 0.33316096663475037,
+      "learning_rate": 2.037001931621392e-06,
+      "loss": 0.5798,
+      "step": 4024
+    },
+    {
+      "epoch": 1.6963852882284751,
+      "grad_norm": 0.30232948064804077,
+      "learning_rate": 2.0359019045896412e-06,
+      "loss": 0.6021,
+      "step": 4025
+    },
+    {
+      "epoch": 1.6968068289598484,
+      "grad_norm": 0.30385878682136536,
+      "learning_rate": 2.0348019706108875e-06,
+      "loss": 0.5857,
+      "step": 4026
+    },
+    {
+      "epoch": 1.6972283696912214,
+      "grad_norm": 0.29342302680015564,
+      "learning_rate": 2.0337021299056698e-06,
+      "loss": 0.5954,
+      "step": 4027
+    },
+    {
+      "epoch": 1.6976499104225946,
+      "grad_norm": 0.30073973536491394,
+      "learning_rate": 2.03260238269451e-06,
+      "loss": 0.601,
+      "step": 4028
+    },
+    {
+      "epoch": 1.6980714511539676,
+      "grad_norm": 0.3110145926475525,
+      "learning_rate": 2.0315027291979097e-06,
+      "loss": 0.5688,
+      "step": 4029
+    },
+    {
+      "epoch": 1.6984929918853409,
+      "grad_norm": 0.3007817268371582,
+      "learning_rate": 2.030403169636355e-06,
+      "loss": 0.5795,
+      "step": 4030
+    },
+    {
+      "epoch": 1.698914532616714,
+      "grad_norm": 0.3038915693759918,
+      "learning_rate": 2.0293037042303075e-06,
+      "loss": 0.5643,
+      "step": 4031
+    },
+    {
+      "epoch": 1.6993360733480873,
+      "grad_norm": 0.2968815863132477,
+      "learning_rate": 2.0282043332002147e-06,
+      "loss": 0.5698,
+      "step": 4032
+    },
+    {
+      "epoch": 1.6997576140794606,
+      "grad_norm": 0.3076591193675995,
+      "learning_rate": 2.0271050567665044e-06,
+      "loss": 0.5854,
+      "step": 4033
+    },
+    {
+      "epoch": 1.7001791548108336,
+      "grad_norm": 0.2927919626235962,
+      "learning_rate": 2.026005875149584e-06,
+      "loss": 0.5785,
+      "step": 4034
+    },
+    {
+      "epoch": 1.7006006955422066,
+      "grad_norm": 0.31080588698387146,
+      "learning_rate": 2.0249067885698426e-06,
+      "loss": 0.5703,
+      "step": 4035
+    },
+    {
+      "epoch": 1.7010222362735798,
+      "grad_norm": 0.3031671941280365,
+      "learning_rate": 2.0238077972476494e-06,
+      "loss": 0.6004,
+      "step": 4036
+    },
+    {
+      "epoch": 1.701443777004953,
+      "grad_norm": 0.295777827501297,
+      "learning_rate": 2.0227089014033564e-06,
+      "loss": 0.5604,
+      "step": 4037
+    },
+    {
+      "epoch": 1.7018653177363263,
+      "grad_norm": 0.2996889352798462,
+      "learning_rate": 2.0216101012572943e-06,
+      "loss": 0.5844,
+      "step": 4038
+    },
+    {
+      "epoch": 1.7022868584676996,
+      "grad_norm": 0.30648064613342285,
+      "learning_rate": 2.0205113970297776e-06,
+      "loss": 0.5863,
+      "step": 4039
+    },
+    {
+      "epoch": 1.7027083991990726,
+      "grad_norm": 0.2966659665107727,
+      "learning_rate": 2.019412788941097e-06,
+      "loss": 0.5645,
+      "step": 4040
+    },
+    {
+      "epoch": 1.7031299399304458,
+      "grad_norm": 0.3037954568862915,
+      "learning_rate": 2.018314277211529e-06,
+      "loss": 0.5941,
+      "step": 4041
+    },
+    {
+      "epoch": 1.7035514806618188,
+      "grad_norm": 0.2954508066177368,
+      "learning_rate": 2.0172158620613265e-06,
+      "loss": 0.5865,
+      "step": 4042
+    },
+    {
+      "epoch": 1.703973021393192,
+      "grad_norm": 0.2999778985977173,
+      "learning_rate": 2.0161175437107257e-06,
+      "loss": 0.5926,
+      "step": 4043
+    },
+    {
+      "epoch": 1.7043945621245653,
+      "grad_norm": 0.3074519634246826,
+      "learning_rate": 2.015019322379943e-06,
+      "loss": 0.6061,
+      "step": 4044
+    },
+    {
+      "epoch": 1.7048161028559385,
+      "grad_norm": 0.30139267444610596,
+      "learning_rate": 2.0139211982891736e-06,
+      "loss": 0.5972,
+      "step": 4045
+    },
+    {
+      "epoch": 1.7052376435873118,
+      "grad_norm": 0.30451667308807373,
+      "learning_rate": 2.0128231716585972e-06,
+      "loss": 0.6078,
+      "step": 4046
+    },
+    {
+      "epoch": 1.7056591843186848,
+      "grad_norm": 0.31132176518440247,
+      "learning_rate": 2.0117252427083673e-06,
+      "loss": 0.5852,
+      "step": 4047
+    },
+    {
+      "epoch": 1.7060807250500578,
+      "grad_norm": 0.31585150957107544,
+      "learning_rate": 2.010627411658626e-06,
+      "loss": 0.5697,
+      "step": 4048
+    },
+    {
+      "epoch": 1.706502265781431,
+      "grad_norm": 0.2985701858997345,
+      "learning_rate": 2.009529678729488e-06,
+      "loss": 0.5589,
+      "step": 4049
+    },
+    {
+      "epoch": 1.7069238065128043,
+      "grad_norm": 0.2996378242969513,
+      "learning_rate": 2.008432044141054e-06,
+      "loss": 0.5634,
+      "step": 4050
+    },
+    {
+      "epoch": 1.7073453472441775,
+      "grad_norm": 0.28197774291038513,
+      "learning_rate": 2.007334508113402e-06,
+      "loss": 0.5998,
+      "step": 4051
+    },
+    {
+      "epoch": 1.7077668879755508,
+      "grad_norm": 0.31098195910453796,
+      "learning_rate": 2.0062370708665917e-06,
+      "loss": 0.5876,
+      "step": 4052
+    },
+    {
+      "epoch": 1.7081884287069238,
+      "grad_norm": 0.2862898111343384,
+      "learning_rate": 2.0051397326206623e-06,
+      "loss": 0.5605,
+      "step": 4053
+    },
+    {
+      "epoch": 1.708609969438297,
+      "grad_norm": 0.3031028211116791,
+      "learning_rate": 2.0040424935956324e-06,
+      "loss": 0.585,
+      "step": 4054
+    },
+    {
+      "epoch": 1.70903151016967,
+      "grad_norm": 0.3033275306224823,
+      "learning_rate": 2.0029453540115035e-06,
+      "loss": 0.5952,
+      "step": 4055
+    },
+    {
+      "epoch": 1.7094530509010433,
+      "grad_norm": 0.30156210064888,
+      "learning_rate": 2.0018483140882525e-06,
+      "loss": 0.5503,
+      "step": 4056
+    },
+    {
+      "epoch": 1.7098745916324165,
+      "grad_norm": 0.3011897802352905,
+      "learning_rate": 2.0007513740458414e-06,
+      "loss": 0.5846,
+      "step": 4057
+    },
+    {
+      "epoch": 1.7102961323637897,
+      "grad_norm": 0.3047075867652893,
+      "learning_rate": 1.9996545341042085e-06,
+      "loss": 0.5705,
+      "step": 4058
+    },
+    {
+      "epoch": 1.710717673095163,
+      "grad_norm": 0.29714131355285645,
+      "learning_rate": 1.9985577944832745e-06,
+      "loss": 0.5841,
+      "step": 4059
+    },
+    {
+      "epoch": 1.711139213826536,
+      "grad_norm": 0.29091018438339233,
+      "learning_rate": 1.9974611554029375e-06,
+      "loss": 0.5909,
+      "step": 4060
+    },
+    {
+      "epoch": 1.711560754557909,
+      "grad_norm": 0.2976052761077881,
+      "learning_rate": 1.996364617083078e-06,
+      "loss": 0.5555,
+      "step": 4061
+    },
+    {
+      "epoch": 1.7119822952892823,
+      "grad_norm": 0.30129939317703247,
+      "learning_rate": 1.995268179743554e-06,
+      "loss": 0.571,
+      "step": 4062
+    },
+    {
+      "epoch": 1.7124038360206555,
+      "grad_norm": 0.2905229926109314,
+      "learning_rate": 1.994171843604204e-06,
+      "loss": 0.5808,
+      "step": 4063
+    },
+    {
+      "epoch": 1.7128253767520287,
+      "grad_norm": 0.30385708808898926,
+      "learning_rate": 1.9930756088848492e-06,
+      "loss": 0.607,
+      "step": 4064
+    },
+    {
+      "epoch": 1.713246917483402,
+      "grad_norm": 0.31346943974494934,
+      "learning_rate": 1.991979475805284e-06,
+      "loss": 0.5883,
+      "step": 4065
+    },
+    {
+      "epoch": 1.713668458214775,
+      "grad_norm": 0.29924046993255615,
+      "learning_rate": 1.990883444585289e-06,
+      "loss": 0.5892,
+      "step": 4066
+    },
+    {
+      "epoch": 1.7140899989461482,
+      "grad_norm": 0.3044843375682831,
+      "learning_rate": 1.989787515444621e-06,
+      "loss": 0.6072,
+      "step": 4067
+    },
+    {
+      "epoch": 1.7145115396775212,
+      "grad_norm": 0.31793534755706787,
+      "learning_rate": 1.9886916886030168e-06,
+      "loss": 0.5817,
+      "step": 4068
+    },
+    {
+      "epoch": 1.7149330804088945,
+      "grad_norm": 0.2940060794353485,
+      "learning_rate": 1.987595964280192e-06,
+      "loss": 0.5916,
+      "step": 4069
+    },
+    {
+      "epoch": 1.7153546211402677,
+      "grad_norm": 0.29678165912628174,
+      "learning_rate": 1.986500342695844e-06,
+      "loss": 0.5778,
+      "step": 4070
+    },
+    {
+      "epoch": 1.715776161871641,
+      "grad_norm": 0.29435351490974426,
+      "learning_rate": 1.9854048240696467e-06,
+      "loss": 0.6188,
+      "step": 4071
+    },
+    {
+      "epoch": 1.716197702603014,
+      "grad_norm": 0.3026663064956665,
+      "learning_rate": 1.9843094086212542e-06,
+      "loss": 0.5598,
+      "step": 4072
+    },
+    {
+      "epoch": 1.7166192433343872,
+      "grad_norm": 0.28752267360687256,
+      "learning_rate": 1.9832140965703033e-06,
+      "loss": 0.5652,
+      "step": 4073
+    },
+    {
+      "epoch": 1.7170407840657602,
+      "grad_norm": 0.2928142249584198,
+      "learning_rate": 1.982118888136404e-06,
+      "loss": 0.5817,
+      "step": 4074
+    },
+    {
+      "epoch": 1.7174623247971335,
+      "grad_norm": 0.299465537071228,
+      "learning_rate": 1.9810237835391515e-06,
+      "loss": 0.5859,
+      "step": 4075
+    },
+    {
+      "epoch": 1.7178838655285067,
+      "grad_norm": 0.29007890820503235,
+      "learning_rate": 1.9799287829981147e-06,
+      "loss": 0.5728,
+      "step": 4076
+    },
+    {
+      "epoch": 1.71830540625988,
+      "grad_norm": 0.2987871766090393,
+      "learning_rate": 1.9788338867328465e-06,
+      "loss": 0.5854,
+      "step": 4077
+    },
+    {
+      "epoch": 1.7187269469912532,
+      "grad_norm": 0.29531267285346985,
+      "learning_rate": 1.9777390949628756e-06,
+      "loss": 0.5744,
+      "step": 4078
+    },
+    {
+      "epoch": 1.7191484877226262,
+      "grad_norm": 0.3068120777606964,
+      "learning_rate": 1.976644407907711e-06,
+      "loss": 0.5755,
+      "step": 4079
+    },
+    {
+      "epoch": 1.7195700284539994,
+      "grad_norm": 0.30491113662719727,
+      "learning_rate": 1.975549825786841e-06,
+      "loss": 0.5832,
+      "step": 4080
+    },
+    {
+      "epoch": 1.7199915691853724,
+      "grad_norm": 0.2921272814273834,
+      "learning_rate": 1.9744553488197315e-06,
+      "loss": 0.5841,
+      "step": 4081
+    },
+    {
+      "epoch": 1.7204131099167457,
+      "grad_norm": 0.2958795428276062,
+      "learning_rate": 1.97336097722583e-06,
+      "loss": 0.5679,
+      "step": 4082
+    },
+    {
+      "epoch": 1.720834650648119,
+      "grad_norm": 0.2992113530635834,
+      "learning_rate": 1.9722667112245592e-06,
+      "loss": 0.5915,
+      "step": 4083
+    },
+    {
+      "epoch": 1.7212561913794922,
+      "grad_norm": 0.31725895404815674,
+      "learning_rate": 1.9711725510353246e-06,
+      "loss": 0.5691,
+      "step": 4084
+    },
+    {
+      "epoch": 1.7216777321108652,
+      "grad_norm": 0.31537750363349915,
+      "learning_rate": 1.9700784968775066e-06,
+      "loss": 0.5949,
+      "step": 4085
+    },
+    {
+      "epoch": 1.7220992728422384,
+      "grad_norm": 0.28967055678367615,
+      "learning_rate": 1.9689845489704672e-06,
+      "loss": 0.6036,
+      "step": 4086
+    },
+    {
+      "epoch": 1.7225208135736114,
+      "grad_norm": 0.3024856448173523,
+      "learning_rate": 1.9678907075335453e-06,
+      "loss": 0.579,
+      "step": 4087
+    },
+    {
+      "epoch": 1.7229423543049847,
+      "grad_norm": 0.32561060786247253,
+      "learning_rate": 1.9667969727860605e-06,
+      "loss": 0.5767,
+      "step": 4088
+    },
+    {
+      "epoch": 1.723363895036358,
+      "grad_norm": 0.3189857602119446,
+      "learning_rate": 1.9657033449473085e-06,
+      "loss": 0.57,
+      "step": 4089
+    },
+    {
+      "epoch": 1.7237854357677311,
+      "grad_norm": 0.30002301931381226,
+      "learning_rate": 1.9646098242365656e-06,
+      "loss": 0.5821,
+      "step": 4090
+    },
+    {
+      "epoch": 1.7242069764991044,
+      "grad_norm": 0.30201593041419983,
+      "learning_rate": 1.9635164108730847e-06,
+      "loss": 0.5845,
+      "step": 4091
+    },
+    {
+      "epoch": 1.7246285172304774,
+      "grad_norm": 0.3130504786968231,
+      "learning_rate": 1.9624231050760995e-06,
+      "loss": 0.5708,
+      "step": 4092
+    },
+    {
+      "epoch": 1.7250500579618504,
+      "grad_norm": 0.31365135312080383,
+      "learning_rate": 1.9613299070648212e-06,
+      "loss": 0.5812,
+      "step": 4093
+    },
+    {
+      "epoch": 1.7254715986932236,
+      "grad_norm": 0.32586532831192017,
+      "learning_rate": 1.960236817058438e-06,
+      "loss": 0.5772,
+      "step": 4094
+    },
+    {
+      "epoch": 1.7258931394245969,
+      "grad_norm": 0.300680547952652,
+      "learning_rate": 1.9591438352761187e-06,
+      "loss": 0.5705,
+      "step": 4095
+    },
+    {
+      "epoch": 1.7263146801559701,
+      "grad_norm": 0.3051530718803406,
+      "learning_rate": 1.9580509619370076e-06,
+      "loss": 0.5821,
+      "step": 4096
+    },
+    {
+      "epoch": 1.7267362208873434,
+      "grad_norm": 0.3080066442489624,
+      "learning_rate": 1.956958197260231e-06,
+      "loss": 0.5717,
+      "step": 4097
+    },
+    {
+      "epoch": 1.7271577616187164,
+      "grad_norm": 0.3209170401096344,
+      "learning_rate": 1.9558655414648894e-06,
+      "loss": 0.5737,
+      "step": 4098
+    },
+    {
+      "epoch": 1.7275793023500896,
+      "grad_norm": 0.31204280257225037,
+      "learning_rate": 1.954772994770066e-06,
+      "loss": 0.5924,
+      "step": 4099
+    },
+    {
+      "epoch": 1.7280008430814626,
+      "grad_norm": 0.2955874800682068,
+      "learning_rate": 1.9536805573948157e-06,
+      "loss": 0.5939,
+      "step": 4100
+    },
+    {
+      "epoch": 1.7284223838128359,
+      "grad_norm": 0.2937082052230835,
+      "learning_rate": 1.9525882295581787e-06,
+      "loss": 0.5911,
+      "step": 4101
+    },
+    {
+      "epoch": 1.728843924544209,
+      "grad_norm": 0.3060196340084076,
+      "learning_rate": 1.951496011479169e-06,
+      "loss": 0.5683,
+      "step": 4102
+    },
+    {
+      "epoch": 1.7292654652755823,
+      "grad_norm": 0.29985731840133667,
+      "learning_rate": 1.9504039033767787e-06,
+      "loss": 0.603,
+      "step": 4103
+    },
+    {
+      "epoch": 1.7296870060069556,
+      "grad_norm": 0.2952657639980316,
+      "learning_rate": 1.949311905469979e-06,
+      "loss": 0.5747,
+      "step": 4104
+    },
+    {
+      "epoch": 1.7301085467383286,
+      "grad_norm": 0.30361613631248474,
+      "learning_rate": 1.948220017977719e-06,
+      "loss": 0.5804,
+      "step": 4105
+    },
+    {
+      "epoch": 1.7305300874697016,
+      "grad_norm": 0.2859334647655487,
+      "learning_rate": 1.9471282411189245e-06,
+      "loss": 0.5704,
+      "step": 4106
+    },
+    {
+      "epoch": 1.7309516282010748,
+      "grad_norm": 0.2964695692062378,
+      "learning_rate": 1.9460365751125e-06,
+      "loss": 0.5817,
+      "step": 4107
+    },
+    {
+      "epoch": 1.731373168932448,
+      "grad_norm": 0.3150608241558075,
+      "learning_rate": 1.9449450201773294e-06,
+      "loss": 0.5626,
+      "step": 4108
+    },
+    {
+      "epoch": 1.7317947096638213,
+      "grad_norm": 0.28346341848373413,
+      "learning_rate": 1.943853576532269e-06,
+      "loss": 0.5773,
+      "step": 4109
+    },
+    {
+      "epoch": 1.7322162503951946,
+      "grad_norm": 0.29554420709609985,
+      "learning_rate": 1.9427622443961604e-06,
+      "loss": 0.593,
+      "step": 4110
+    },
+    {
+      "epoch": 1.7326377911265676,
+      "grad_norm": 0.3021925091743469,
+      "learning_rate": 1.9416710239878158e-06,
+      "loss": 0.5992,
+      "step": 4111
+    },
+    {
+      "epoch": 1.7330593318579408,
+      "grad_norm": 0.31216999888420105,
+      "learning_rate": 1.9405799155260293e-06,
+      "loss": 0.5859,
+      "step": 4112
+    },
+    {
+      "epoch": 1.7334808725893138,
+      "grad_norm": 0.30480310320854187,
+      "learning_rate": 1.9394889192295717e-06,
+      "loss": 0.5726,
+      "step": 4113
+    },
+    {
+      "epoch": 1.733902413320687,
+      "grad_norm": 0.30492696166038513,
+      "learning_rate": 1.93839803531719e-06,
+      "loss": 0.5964,
+      "step": 4114
+    },
+    {
+      "epoch": 1.7343239540520603,
+      "grad_norm": 0.2989857494831085,
+      "learning_rate": 1.93730726400761e-06,
+      "loss": 0.5959,
+      "step": 4115
+    },
+    {
+      "epoch": 1.7347454947834335,
+      "grad_norm": 0.30628320574760437,
+      "learning_rate": 1.9362166055195334e-06,
+      "loss": 0.5438,
+      "step": 4116
+    },
+    {
+      "epoch": 1.7351670355148068,
+      "grad_norm": 0.3354431688785553,
+      "learning_rate": 1.935126060071643e-06,
+      "loss": 0.5704,
+      "step": 4117
+    },
+    {
+      "epoch": 1.7355885762461798,
+      "grad_norm": 0.30182376503944397,
+      "learning_rate": 1.9340356278825924e-06,
+      "loss": 0.5943,
+      "step": 4118
+    },
+    {
+      "epoch": 1.7360101169775528,
+      "grad_norm": 0.30262264609336853,
+      "learning_rate": 1.9329453091710203e-06,
+      "loss": 0.5828,
+      "step": 4119
+    },
+    {
+      "epoch": 1.736431657708926,
+      "grad_norm": 0.28704753518104553,
+      "learning_rate": 1.9318551041555363e-06,
+      "loss": 0.5646,
+      "step": 4120
+    },
+    {
+      "epoch": 1.7368531984402993,
+      "grad_norm": 0.3103243112564087,
+      "learning_rate": 1.9307650130547296e-06,
+      "loss": 0.5769,
+      "step": 4121
+    },
+    {
+      "epoch": 1.7372747391716725,
+      "grad_norm": 0.2982974052429199,
+      "learning_rate": 1.9296750360871686e-06,
+      "loss": 0.5626,
+      "step": 4122
+    },
+    {
+      "epoch": 1.7376962799030458,
+      "grad_norm": 0.2965680956840515,
+      "learning_rate": 1.928585173471394e-06,
+      "loss": 0.5747,
+      "step": 4123
+    },
+    {
+      "epoch": 1.7381178206344188,
+      "grad_norm": 0.3240896463394165,
+      "learning_rate": 1.9274954254259297e-06,
+      "loss": 0.5783,
+      "step": 4124
+    },
+    {
+      "epoch": 1.738539361365792,
+      "grad_norm": 0.3016936182975769,
+      "learning_rate": 1.9264057921692695e-06,
+      "loss": 0.5913,
+      "step": 4125
+    },
+    {
+      "epoch": 1.738960902097165,
+      "grad_norm": 0.2911202907562256,
+      "learning_rate": 1.925316273919892e-06,
+      "loss": 0.5681,
+      "step": 4126
+    },
+    {
+      "epoch": 1.7393824428285383,
+      "grad_norm": 0.30953624844551086,
+      "learning_rate": 1.9242268708962443e-06,
+      "loss": 0.556,
+      "step": 4127
+    },
+    {
+      "epoch": 1.7398039835599115,
+      "grad_norm": 0.30381453037261963,
+      "learning_rate": 1.923137583316759e-06,
+      "loss": 0.5759,
+      "step": 4128
+    },
+    {
+      "epoch": 1.7402255242912847,
+      "grad_norm": 0.29236534237861633,
+      "learning_rate": 1.9220484113998385e-06,
+      "loss": 0.5885,
+      "step": 4129
+    },
+    {
+      "epoch": 1.740647065022658,
+      "grad_norm": 0.2987006902694702,
+      "learning_rate": 1.9209593553638667e-06,
+      "loss": 0.5932,
+      "step": 4130
+    },
+    {
+      "epoch": 1.741068605754031,
+      "grad_norm": 0.2981664538383484,
+      "learning_rate": 1.9198704154272015e-06,
+      "loss": 0.5977,
+      "step": 4131
+    },
+    {
+      "epoch": 1.741490146485404,
+      "grad_norm": 0.29997801780700684,
+      "learning_rate": 1.9187815918081774e-06,
+      "loss": 0.6123,
+      "step": 4132
+    },
+    {
+      "epoch": 1.7419116872167772,
+      "grad_norm": 0.2940407991409302,
+      "learning_rate": 1.91769288472511e-06,
+      "loss": 0.5831,
+      "step": 4133
+    },
+    {
+      "epoch": 1.7423332279481505,
+      "grad_norm": 0.2993693947792053,
+      "learning_rate": 1.916604294396284e-06,
+      "loss": 0.5783,
+      "step": 4134
+    },
+    {
+      "epoch": 1.7427547686795237,
+      "grad_norm": 0.3028712570667267,
+      "learning_rate": 1.9155158210399688e-06,
+      "loss": 0.5955,
+      "step": 4135
+    },
+    {
+      "epoch": 1.743176309410897,
+      "grad_norm": 0.29553648829460144,
+      "learning_rate": 1.914427464874404e-06,
+      "loss": 0.5709,
+      "step": 4136
+    },
+    {
+      "epoch": 1.74359785014227,
+      "grad_norm": 0.29845544695854187,
+      "learning_rate": 1.913339226117809e-06,
+      "loss": 0.5752,
+      "step": 4137
+    },
+    {
+      "epoch": 1.7440193908736432,
+      "grad_norm": 0.29652220010757446,
+      "learning_rate": 1.912251104988378e-06,
+      "loss": 0.5772,
+      "step": 4138
+    },
+    {
+      "epoch": 1.7444409316050162,
+      "grad_norm": 0.29238373041152954,
+      "learning_rate": 1.9111631017042832e-06,
+      "loss": 0.5693,
+      "step": 4139
+    },
+    {
+      "epoch": 1.7448624723363895,
+      "grad_norm": 0.3151078224182129,
+      "learning_rate": 1.910075216483672e-06,
+      "loss": 0.5839,
+      "step": 4140
+    },
+    {
+      "epoch": 1.7452840130677627,
+      "grad_norm": 0.29077666997909546,
+      "learning_rate": 1.908987449544668e-06,
+      "loss": 0.5963,
+      "step": 4141
+    },
+    {
+      "epoch": 1.745705553799136,
+      "grad_norm": 0.29950374364852905,
+      "learning_rate": 1.907899801105374e-06,
+      "loss": 0.5679,
+      "step": 4142
+    },
+    {
+      "epoch": 1.746127094530509,
+      "grad_norm": 0.2949780523777008,
+      "learning_rate": 1.9068122713838627e-06,
+      "loss": 0.5592,
+      "step": 4143
+    },
+    {
+      "epoch": 1.7465486352618822,
+      "grad_norm": 0.2971726655960083,
+      "learning_rate": 1.9057248605981901e-06,
+      "loss": 0.5964,
+      "step": 4144
+    },
+    {
+      "epoch": 1.7469701759932552,
+      "grad_norm": 0.295952171087265,
+      "learning_rate": 1.9046375689663838e-06,
+      "loss": 0.572,
+      "step": 4145
+    },
+    {
+      "epoch": 1.7473917167246285,
+      "grad_norm": 0.3132985830307007,
+      "learning_rate": 1.9035503967064498e-06,
+      "loss": 0.5678,
+      "step": 4146
+    },
+    {
+      "epoch": 1.7478132574560017,
+      "grad_norm": 0.2829049825668335,
+      "learning_rate": 1.902463344036368e-06,
+      "loss": 0.5931,
+      "step": 4147
+    },
+    {
+      "epoch": 1.748234798187375,
+      "grad_norm": 0.3019522726535797,
+      "learning_rate": 1.9013764111740968e-06,
+      "loss": 0.6002,
+      "step": 4148
+    },
+    {
+      "epoch": 1.7486563389187482,
+      "grad_norm": 0.2908456027507782,
+      "learning_rate": 1.900289598337568e-06,
+      "loss": 0.5801,
+      "step": 4149
+    },
+    {
+      "epoch": 1.7490778796501212,
+      "grad_norm": 0.297460675239563,
+      "learning_rate": 1.8992029057446913e-06,
+      "loss": 0.5642,
+      "step": 4150
+    },
+    {
+      "epoch": 1.7494994203814942,
+      "grad_norm": 0.298662394285202,
+      "learning_rate": 1.8981163336133526e-06,
+      "loss": 0.5848,
+      "step": 4151
+    },
+    {
+      "epoch": 1.7499209611128674,
+      "grad_norm": 0.2891845107078552,
+      "learning_rate": 1.8970298821614116e-06,
+      "loss": 0.588,
+      "step": 4152
+    },
+    {
+      "epoch": 1.7503425018442407,
+      "grad_norm": 0.3198443353176117,
+      "learning_rate": 1.8959435516067056e-06,
+      "loss": 0.5924,
+      "step": 4153
+    },
+    {
+      "epoch": 1.750764042575614,
+      "grad_norm": 0.29900237917900085,
+      "learning_rate": 1.8948573421670462e-06,
+      "loss": 0.58,
+      "step": 4154
+    },
+    {
+      "epoch": 1.7511855833069871,
+      "grad_norm": 0.3053187131881714,
+      "learning_rate": 1.8937712540602227e-06,
+      "loss": 0.5896,
+      "step": 4155
+    },
+    {
+      "epoch": 1.7516071240383602,
+      "grad_norm": 0.3004865050315857,
+      "learning_rate": 1.8926852875039975e-06,
+      "loss": 0.5716,
+      "step": 4156
+    },
+    {
+      "epoch": 1.7520286647697334,
+      "grad_norm": 0.29301679134368896,
+      "learning_rate": 1.891599442716111e-06,
+      "loss": 0.6032,
+      "step": 4157
+    },
+    {
+      "epoch": 1.7524502055011064,
+      "grad_norm": 0.31102532148361206,
+      "learning_rate": 1.8905137199142777e-06,
+      "loss": 0.6054,
+      "step": 4158
+    },
+    {
+      "epoch": 1.7528717462324797,
+      "grad_norm": 0.304212361574173,
+      "learning_rate": 1.8894281193161884e-06,
+      "loss": 0.5879,
+      "step": 4159
+    },
+    {
+      "epoch": 1.753293286963853,
+      "grad_norm": 0.28799712657928467,
+      "learning_rate": 1.888342641139508e-06,
+      "loss": 0.6008,
+      "step": 4160
+    },
+    {
+      "epoch": 1.7537148276952261,
+      "grad_norm": 0.301728755235672,
+      "learning_rate": 1.8872572856018795e-06,
+      "loss": 0.5961,
+      "step": 4161
+    },
+    {
+      "epoch": 1.7541363684265994,
+      "grad_norm": 0.3023589551448822,
+      "learning_rate": 1.88617205292092e-06,
+      "loss": 0.5815,
+      "step": 4162
+    },
+    {
+      "epoch": 1.7545579091579724,
+      "grad_norm": 0.2949874997138977,
+      "learning_rate": 1.8850869433142201e-06,
+      "loss": 0.607,
+      "step": 4163
+    },
+    {
+      "epoch": 1.7549794498893454,
+      "grad_norm": 0.2942228317260742,
+      "learning_rate": 1.8840019569993491e-06,
+      "loss": 0.569,
+      "step": 4164
+    },
+    {
+      "epoch": 1.7554009906207186,
+      "grad_norm": 0.2987922728061676,
+      "learning_rate": 1.8829170941938485e-06,
+      "loss": 0.6034,
+      "step": 4165
+    },
+    {
+      "epoch": 1.7558225313520919,
+      "grad_norm": 0.2996947765350342,
+      "learning_rate": 1.8818323551152376e-06,
+      "loss": 0.5492,
+      "step": 4166
+    },
+    {
+      "epoch": 1.7562440720834651,
+      "grad_norm": 0.3158491551876068,
+      "learning_rate": 1.880747739981008e-06,
+      "loss": 0.5773,
+      "step": 4167
+    },
+    {
+      "epoch": 1.7566656128148384,
+      "grad_norm": 0.30243033170700073,
+      "learning_rate": 1.8796632490086304e-06,
+      "loss": 0.5806,
+      "step": 4168
+    },
+    {
+      "epoch": 1.7570871535462114,
+      "grad_norm": 0.2954489588737488,
+      "learning_rate": 1.8785788824155456e-06,
+      "loss": 0.5883,
+      "step": 4169
+    },
+    {
+      "epoch": 1.7575086942775846,
+      "grad_norm": 0.30286651849746704,
+      "learning_rate": 1.8774946404191746e-06,
+      "loss": 0.5951,
+      "step": 4170
+    },
+    {
+      "epoch": 1.7579302350089576,
+      "grad_norm": 0.2889292240142822,
+      "learning_rate": 1.8764105232369106e-06,
+      "loss": 0.6018,
+      "step": 4171
+    },
+    {
+      "epoch": 1.7583517757403309,
+      "grad_norm": 0.2986442744731903,
+      "learning_rate": 1.8753265310861213e-06,
+      "loss": 0.5845,
+      "step": 4172
+    },
+    {
+      "epoch": 1.758773316471704,
+      "grad_norm": 0.30298787355422974,
+      "learning_rate": 1.874242664184151e-06,
+      "loss": 0.5821,
+      "step": 4173
+    },
+    {
+      "epoch": 1.7591948572030773,
+      "grad_norm": 0.2963033616542816,
+      "learning_rate": 1.8731589227483174e-06,
+      "loss": 0.5846,
+      "step": 4174
+    },
+    {
+      "epoch": 1.7596163979344506,
+      "grad_norm": 0.3034197688102722,
+      "learning_rate": 1.8720753069959147e-06,
+      "loss": 0.5812,
+      "step": 4175
+    },
+    {
+      "epoch": 1.7600379386658236,
+      "grad_norm": 0.31029969453811646,
+      "learning_rate": 1.8709918171442093e-06,
+      "loss": 0.6078,
+      "step": 4176
+    },
+    {
+      "epoch": 1.7604594793971966,
+      "grad_norm": 0.3179110586643219,
+      "learning_rate": 1.8699084534104467e-06,
+      "loss": 0.5693,
+      "step": 4177
+    },
+    {
+      "epoch": 1.7608810201285698,
+      "grad_norm": 0.29548123478889465,
+      "learning_rate": 1.8688252160118415e-06,
+      "loss": 0.5837,
+      "step": 4178
+    },
+    {
+      "epoch": 1.761302560859943,
+      "grad_norm": 0.29736748337745667,
+      "learning_rate": 1.8677421051655887e-06,
+      "loss": 0.5802,
+      "step": 4179
+    },
+    {
+      "epoch": 1.7617241015913163,
+      "grad_norm": 0.2952311933040619,
+      "learning_rate": 1.8666591210888524e-06,
+      "loss": 0.6018,
+      "step": 4180
+    },
+    {
+      "epoch": 1.7621456423226896,
+      "grad_norm": 0.3209156394004822,
+      "learning_rate": 1.8655762639987755e-06,
+      "loss": 0.5807,
+      "step": 4181
+    },
+    {
+      "epoch": 1.7625671830540626,
+      "grad_norm": 0.3088265657424927,
+      "learning_rate": 1.8644935341124743e-06,
+      "loss": 0.5729,
+      "step": 4182
+    },
+    {
+      "epoch": 1.7629887237854358,
+      "grad_norm": 0.2934715449810028,
+      "learning_rate": 1.8634109316470383e-06,
+      "loss": 0.577,
+      "step": 4183
+    },
+    {
+      "epoch": 1.7634102645168088,
+      "grad_norm": 0.3044416904449463,
+      "learning_rate": 1.8623284568195332e-06,
+      "loss": 0.5707,
+      "step": 4184
+    },
+    {
+      "epoch": 1.763831805248182,
+      "grad_norm": 0.29599013924598694,
+      "learning_rate": 1.8612461098469969e-06,
+      "loss": 0.592,
+      "step": 4185
+    },
+    {
+      "epoch": 1.7642533459795553,
+      "grad_norm": 0.29519131779670715,
+      "learning_rate": 1.8601638909464457e-06,
+      "loss": 0.585,
+      "step": 4186
+    },
+    {
+      "epoch": 1.7646748867109285,
+      "grad_norm": 0.28404030203819275,
+      "learning_rate": 1.8590818003348643e-06,
+      "loss": 0.5995,
+      "step": 4187
+    },
+    {
+      "epoch": 1.7650964274423018,
+      "grad_norm": 0.2950480282306671,
+      "learning_rate": 1.8579998382292176e-06,
+      "loss": 0.5695,
+      "step": 4188
+    },
+    {
+      "epoch": 1.7655179681736748,
+      "grad_norm": 0.28425708413124084,
+      "learning_rate": 1.8569180048464407e-06,
+      "loss": 0.5999,
+      "step": 4189
+    },
+    {
+      "epoch": 1.7659395089050478,
+      "grad_norm": 0.30628857016563416,
+      "learning_rate": 1.8558363004034446e-06,
+      "loss": 0.5883,
+      "step": 4190
+    },
+    {
+      "epoch": 1.766361049636421,
+      "grad_norm": 0.29682689905166626,
+      "learning_rate": 1.854754725117115e-06,
+      "loss": 0.5748,
+      "step": 4191
+    },
+    {
+      "epoch": 1.7667825903677943,
+      "grad_norm": 0.30673763155937195,
+      "learning_rate": 1.8536732792043093e-06,
+      "loss": 0.5652,
+      "step": 4192
+    },
+    {
+      "epoch": 1.7672041310991675,
+      "grad_norm": 0.29609861969947815,
+      "learning_rate": 1.8525919628818618e-06,
+      "loss": 0.5703,
+      "step": 4193
+    },
+    {
+      "epoch": 1.7676256718305408,
+      "grad_norm": 0.29188257455825806,
+      "learning_rate": 1.851510776366578e-06,
+      "loss": 0.5634,
+      "step": 4194
+    },
+    {
+      "epoch": 1.7680472125619138,
+      "grad_norm": 0.29071682691574097,
+      "learning_rate": 1.8504297198752413e-06,
+      "loss": 0.5744,
+      "step": 4195
+    },
+    {
+      "epoch": 1.768468753293287,
+      "grad_norm": 0.3230101466178894,
+      "learning_rate": 1.849348793624604e-06,
+      "loss": 0.5803,
+      "step": 4196
+    },
+    {
+      "epoch": 1.76889029402466,
+      "grad_norm": 0.28750747442245483,
+      "learning_rate": 1.848267997831397e-06,
+      "loss": 0.5589,
+      "step": 4197
+    },
+    {
+      "epoch": 1.7693118347560333,
+      "grad_norm": 0.30123937129974365,
+      "learning_rate": 1.8471873327123213e-06,
+      "loss": 0.5722,
+      "step": 4198
+    },
+    {
+      "epoch": 1.7697333754874065,
+      "grad_norm": 0.29799604415893555,
+      "learning_rate": 1.846106798484055e-06,
+      "loss": 0.5839,
+      "step": 4199
+    },
+    {
+      "epoch": 1.7701549162187797,
+      "grad_norm": 0.3089456260204315,
+      "learning_rate": 1.845026395363247e-06,
+      "loss": 0.5678,
+      "step": 4200
+    },
+    {
+      "epoch": 1.7705764569501528,
+      "grad_norm": 0.30584681034088135,
+      "learning_rate": 1.8439461235665208e-06,
+      "loss": 0.57,
+      "step": 4201
+    },
+    {
+      "epoch": 1.770997997681526,
+      "grad_norm": 0.3027849495410919,
+      "learning_rate": 1.8428659833104762e-06,
+      "loss": 0.5936,
+      "step": 4202
+    },
+    {
+      "epoch": 1.771419538412899,
+      "grad_norm": 0.29791751503944397,
+      "learning_rate": 1.8417859748116819e-06,
+      "loss": 0.5823,
+      "step": 4203
+    },
+    {
+      "epoch": 1.7718410791442722,
+      "grad_norm": 0.2928924858570099,
+      "learning_rate": 1.8407060982866848e-06,
+      "loss": 0.563,
+      "step": 4204
+    },
+    {
+      "epoch": 1.7722626198756455,
+      "grad_norm": 0.3059251010417938,
+      "learning_rate": 1.8396263539520014e-06,
+      "loss": 0.5803,
+      "step": 4205
+    },
+    {
+      "epoch": 1.7726841606070187,
+      "grad_norm": 0.2992413640022278,
+      "learning_rate": 1.8385467420241254e-06,
+      "loss": 0.5821,
+      "step": 4206
+    },
+    {
+      "epoch": 1.773105701338392,
+      "grad_norm": 0.29566678404808044,
+      "learning_rate": 1.8374672627195206e-06,
+      "loss": 0.5872,
+      "step": 4207
+    },
+    {
+      "epoch": 1.773527242069765,
+      "grad_norm": 0.28036025166511536,
+      "learning_rate": 1.8363879162546265e-06,
+      "loss": 0.5551,
+      "step": 4208
+    },
+    {
+      "epoch": 1.7739487828011382,
+      "grad_norm": 0.3077899217605591,
+      "learning_rate": 1.8353087028458541e-06,
+      "loss": 0.5701,
+      "step": 4209
+    },
+    {
+      "epoch": 1.7743703235325112,
+      "grad_norm": 0.2974056601524353,
+      "learning_rate": 1.834229622709589e-06,
+      "loss": 0.5829,
+      "step": 4210
+    },
+    {
+      "epoch": 1.7747918642638845,
+      "grad_norm": 0.300565242767334,
+      "learning_rate": 1.8331506760621922e-06,
+      "loss": 0.5808,
+      "step": 4211
+    },
+    {
+      "epoch": 1.7752134049952577,
+      "grad_norm": 0.2977977991104126,
+      "learning_rate": 1.832071863119992e-06,
+      "loss": 0.5867,
+      "step": 4212
+    },
+    {
+      "epoch": 1.775634945726631,
+      "grad_norm": 0.30195748805999756,
+      "learning_rate": 1.8309931840992964e-06,
+      "loss": 0.5847,
+      "step": 4213
+    },
+    {
+      "epoch": 1.776056486458004,
+      "grad_norm": 0.2990635335445404,
+      "learning_rate": 1.8299146392163819e-06,
+      "loss": 0.5743,
+      "step": 4214
+    },
+    {
+      "epoch": 1.7764780271893772,
+      "grad_norm": 0.29503145813941956,
+      "learning_rate": 1.8288362286875012e-06,
+      "loss": 0.5866,
+      "step": 4215
+    },
+    {
+      "epoch": 1.7768995679207502,
+      "grad_norm": 0.29583925008773804,
+      "learning_rate": 1.827757952728877e-06,
+      "loss": 0.6202,
+      "step": 4216
+    },
+    {
+      "epoch": 1.7773211086521234,
+      "grad_norm": 0.30492499470710754,
+      "learning_rate": 1.8266798115567083e-06,
+      "loss": 0.5824,
+      "step": 4217
+    },
+    {
+      "epoch": 1.7777426493834967,
+      "grad_norm": 0.3046671450138092,
+      "learning_rate": 1.8256018053871644e-06,
+      "loss": 0.5823,
+      "step": 4218
+    },
+    {
+      "epoch": 1.77816419011487,
+      "grad_norm": 0.29521623253822327,
+      "learning_rate": 1.8245239344363883e-06,
+      "loss": 0.5939,
+      "step": 4219
+    },
+    {
+      "epoch": 1.7785857308462432,
+      "grad_norm": 0.29788312315940857,
+      "learning_rate": 1.8234461989204985e-06,
+      "loss": 0.5986,
+      "step": 4220
+    },
+    {
+      "epoch": 1.7790072715776162,
+      "grad_norm": 0.3062983751296997,
+      "learning_rate": 1.822368599055581e-06,
+      "loss": 0.5813,
+      "step": 4221
+    },
+    {
+      "epoch": 1.7794288123089892,
+      "grad_norm": 0.30412042140960693,
+      "learning_rate": 1.8212911350577003e-06,
+      "loss": 0.5849,
+      "step": 4222
+    },
+    {
+      "epoch": 1.7798503530403624,
+      "grad_norm": 0.2860157787799835,
+      "learning_rate": 1.8202138071428893e-06,
+      "loss": 0.5702,
+      "step": 4223
+    },
+    {
+      "epoch": 1.7802718937717357,
+      "grad_norm": 0.29933735728263855,
+      "learning_rate": 1.8191366155271565e-06,
+      "loss": 0.6051,
+      "step": 4224
+    },
+    {
+      "epoch": 1.780693434503109,
+      "grad_norm": 0.3141981065273285,
+      "learning_rate": 1.8180595604264805e-06,
+      "loss": 0.581,
+      "step": 4225
+    },
+    {
+      "epoch": 1.7811149752344821,
+      "grad_norm": 0.294624924659729,
+      "learning_rate": 1.8169826420568156e-06,
+      "loss": 0.5727,
+      "step": 4226
+    },
+    {
+      "epoch": 1.7815365159658552,
+      "grad_norm": 0.30425429344177246,
+      "learning_rate": 1.8159058606340856e-06,
+      "loss": 0.5814,
+      "step": 4227
+    },
+    {
+      "epoch": 1.7819580566972284,
+      "grad_norm": 0.3096727430820465,
+      "learning_rate": 1.8148292163741894e-06,
+      "loss": 0.5756,
+      "step": 4228
+    },
+    {
+      "epoch": 1.7823795974286014,
+      "grad_norm": 0.29234400391578674,
+      "learning_rate": 1.8137527094929952e-06,
+      "loss": 0.5882,
+      "step": 4229
+    },
+    {
+      "epoch": 1.7828011381599747,
+      "grad_norm": 0.2903161346912384,
+      "learning_rate": 1.812676340206348e-06,
+      "loss": 0.5775,
+      "step": 4230
+    },
+    {
+      "epoch": 1.783222678891348,
+      "grad_norm": 0.3051566779613495,
+      "learning_rate": 1.8116001087300627e-06,
+      "loss": 0.5768,
+      "step": 4231
+    },
+    {
+      "epoch": 1.7836442196227211,
+      "grad_norm": 0.2977510392665863,
+      "learning_rate": 1.8105240152799256e-06,
+      "loss": 0.5829,
+      "step": 4232
+    },
+    {
+      "epoch": 1.7840657603540944,
+      "grad_norm": 0.30955231189727783,
+      "learning_rate": 1.8094480600716974e-06,
+      "loss": 0.5754,
+      "step": 4233
+    },
+    {
+      "epoch": 1.7844873010854674,
+      "grad_norm": 0.29226750135421753,
+      "learning_rate": 1.80837224332111e-06,
+      "loss": 0.5723,
+      "step": 4234
+    },
+    {
+      "epoch": 1.7849088418168404,
+      "grad_norm": 0.3077859878540039,
+      "learning_rate": 1.8072965652438673e-06,
+      "loss": 0.5753,
+      "step": 4235
+    },
+    {
+      "epoch": 1.7853303825482136,
+      "grad_norm": 0.3062484860420227,
+      "learning_rate": 1.8062210260556463e-06,
+      "loss": 0.5873,
+      "step": 4236
+    },
+    {
+      "epoch": 1.7857519232795869,
+      "grad_norm": 0.30027657747268677,
+      "learning_rate": 1.805145625972096e-06,
+      "loss": 0.5766,
+      "step": 4237
+    },
+    {
+      "epoch": 1.78617346401096,
+      "grad_norm": 0.2945031225681305,
+      "learning_rate": 1.8040703652088357e-06,
+      "loss": 0.5836,
+      "step": 4238
+    },
+    {
+      "epoch": 1.7865950047423333,
+      "grad_norm": 0.302629679441452,
+      "learning_rate": 1.8029952439814596e-06,
+      "loss": 0.5759,
+      "step": 4239
+    },
+    {
+      "epoch": 1.7870165454737064,
+      "grad_norm": 0.30523306131362915,
+      "learning_rate": 1.8019202625055333e-06,
+      "loss": 0.5877,
+      "step": 4240
+    },
+    {
+      "epoch": 1.7874380862050796,
+      "grad_norm": 0.30120614171028137,
+      "learning_rate": 1.8008454209965916e-06,
+      "loss": 0.6003,
+      "step": 4241
+    },
+    {
+      "epoch": 1.7878596269364526,
+      "grad_norm": 0.30872443318367004,
+      "learning_rate": 1.7997707196701452e-06,
+      "loss": 0.5882,
+      "step": 4242
+    },
+    {
+      "epoch": 1.7882811676678259,
+      "grad_norm": 0.30629709362983704,
+      "learning_rate": 1.7986961587416736e-06,
+      "loss": 0.5512,
+      "step": 4243
+    },
+    {
+      "epoch": 1.788702708399199,
+      "grad_norm": 0.29853981733322144,
+      "learning_rate": 1.79762173842663e-06,
+      "loss": 0.5737,
+      "step": 4244
+    },
+    {
+      "epoch": 1.7891242491305723,
+      "grad_norm": 0.30877846479415894,
+      "learning_rate": 1.796547458940438e-06,
+      "loss": 0.585,
+      "step": 4245
+    },
+    {
+      "epoch": 1.7895457898619456,
+      "grad_norm": 0.28830742835998535,
+      "learning_rate": 1.7954733204984948e-06,
+      "loss": 0.598,
+      "step": 4246
+    },
+    {
+      "epoch": 1.7899673305933186,
+      "grad_norm": 0.30494818091392517,
+      "learning_rate": 1.7943993233161664e-06,
+      "loss": 0.575,
+      "step": 4247
+    },
+    {
+      "epoch": 1.7903888713246916,
+      "grad_norm": 0.2860972583293915,
+      "learning_rate": 1.7933254676087952e-06,
+      "loss": 0.591,
+      "step": 4248
+    },
+    {
+      "epoch": 1.7908104120560648,
+      "grad_norm": 0.2914313077926636,
+      "learning_rate": 1.792251753591689e-06,
+      "loss": 0.5921,
+      "step": 4249
+    },
+    {
+      "epoch": 1.791231952787438,
+      "grad_norm": 0.29147520661354065,
+      "learning_rate": 1.7911781814801326e-06,
+      "loss": 0.6335,
+      "step": 4250
+    },
+    {
+      "epoch": 1.7916534935188113,
+      "grad_norm": 0.2939983606338501,
+      "learning_rate": 1.7901047514893805e-06,
+      "loss": 0.6148,
+      "step": 4251
+    },
+    {
+      "epoch": 1.7920750342501846,
+      "grad_norm": 0.288116455078125,
+      "learning_rate": 1.7890314638346574e-06,
+      "loss": 0.5859,
+      "step": 4252
+    },
+    {
+      "epoch": 1.7924965749815576,
+      "grad_norm": 0.3101334273815155,
+      "learning_rate": 1.7879583187311615e-06,
+      "loss": 0.5747,
+      "step": 4253
+    },
+    {
+      "epoch": 1.7929181157129308,
+      "grad_norm": 0.29647988080978394,
+      "learning_rate": 1.7868853163940597e-06,
+      "loss": 0.5655,
+      "step": 4254
+    },
+    {
+      "epoch": 1.7933396564443038,
+      "grad_norm": 0.28765782713890076,
+      "learning_rate": 1.7858124570384952e-06,
+      "loss": 0.59,
+      "step": 4255
+    },
+    {
+      "epoch": 1.793761197175677,
+      "grad_norm": 0.29500651359558105,
+      "learning_rate": 1.7847397408795757e-06,
+      "loss": 0.5908,
+      "step": 4256
+    },
+    {
+      "epoch": 1.7941827379070503,
+      "grad_norm": 0.29030075669288635,
+      "learning_rate": 1.7836671681323875e-06,
+      "loss": 0.5514,
+      "step": 4257
+    },
+    {
+      "epoch": 1.7946042786384235,
+      "grad_norm": 0.30086421966552734,
+      "learning_rate": 1.7825947390119808e-06,
+      "loss": 0.5968,
+      "step": 4258
+    },
+    {
+      "epoch": 1.7950258193697968,
+      "grad_norm": 0.30056172609329224,
+      "learning_rate": 1.781522453733383e-06,
+      "loss": 0.5754,
+      "step": 4259
+    },
+    {
+      "epoch": 1.7954473601011698,
+      "grad_norm": 0.29563242197036743,
+      "learning_rate": 1.7804503125115906e-06,
+      "loss": 0.5862,
+      "step": 4260
+    },
+    {
+      "epoch": 1.7958689008325428,
+      "grad_norm": 0.28768831491470337,
+      "learning_rate": 1.7793783155615698e-06,
+      "loss": 0.6093,
+      "step": 4261
+    },
+    {
+      "epoch": 1.796290441563916,
+      "grad_norm": 0.30151471495628357,
+      "learning_rate": 1.7783064630982599e-06,
+      "loss": 0.5981,
+      "step": 4262
+    },
+    {
+      "epoch": 1.7967119822952893,
+      "grad_norm": 0.295257031917572,
+      "learning_rate": 1.7772347553365693e-06,
+      "loss": 0.5789,
+      "step": 4263
+    },
+    {
+      "epoch": 1.7971335230266625,
+      "grad_norm": 0.30238810181617737,
+      "learning_rate": 1.7761631924913809e-06,
+      "loss": 0.5837,
+      "step": 4264
+    },
+    {
+      "epoch": 1.7975550637580358,
+      "grad_norm": 0.2979957163333893,
+      "learning_rate": 1.7750917747775425e-06,
+      "loss": 0.5795,
+      "step": 4265
+    },
+    {
+      "epoch": 1.7979766044894088,
+      "grad_norm": 0.31823426485061646,
+      "learning_rate": 1.7740205024098794e-06,
+      "loss": 0.5564,
+      "step": 4266
+    },
+    {
+      "epoch": 1.798398145220782,
+      "grad_norm": 0.2958109676837921,
+      "learning_rate": 1.7729493756031835e-06,
+      "loss": 0.5898,
+      "step": 4267
+    },
+    {
+      "epoch": 1.798819685952155,
+      "grad_norm": 0.2876497209072113,
+      "learning_rate": 1.7718783945722196e-06,
+      "loss": 0.5617,
+      "step": 4268
+    },
+    {
+      "epoch": 1.7992412266835283,
+      "grad_norm": 0.3032357394695282,
+      "learning_rate": 1.770807559531721e-06,
+      "loss": 0.556,
+      "step": 4269
+    },
+    {
+      "epoch": 1.7996627674149015,
+      "grad_norm": 0.28998830914497375,
+      "learning_rate": 1.7697368706963942e-06,
+      "loss": 0.588,
+      "step": 4270
+    },
+    {
+      "epoch": 1.8000843081462747,
+      "grad_norm": 0.30257388949394226,
+      "learning_rate": 1.7686663282809159e-06,
+      "loss": 0.5797,
+      "step": 4271
+    },
+    {
+      "epoch": 1.8005058488776478,
+      "grad_norm": 0.30849310755729675,
+      "learning_rate": 1.7675959324999312e-06,
+      "loss": 0.5834,
+      "step": 4272
+    },
+    {
+      "epoch": 1.800927389609021,
+      "grad_norm": 0.3058018684387207,
+      "learning_rate": 1.76652568356806e-06,
+      "loss": 0.5853,
+      "step": 4273
+    },
+    {
+      "epoch": 1.801348930340394,
+      "grad_norm": 0.2851511538028717,
+      "learning_rate": 1.7654555816998874e-06,
+      "loss": 0.5565,
+      "step": 4274
+    },
+    {
+      "epoch": 1.8017704710717672,
+      "grad_norm": 0.31438177824020386,
+      "learning_rate": 1.7643856271099746e-06,
+      "loss": 0.5954,
+      "step": 4275
+    },
+    {
+      "epoch": 1.8021920118031405,
+      "grad_norm": 0.302577406167984,
+      "learning_rate": 1.7633158200128486e-06,
+      "loss": 0.5707,
+      "step": 4276
+    },
+    {
+      "epoch": 1.8026135525345137,
+      "grad_norm": 0.2851966917514801,
+      "learning_rate": 1.76224616062301e-06,
+      "loss": 0.5647,
+      "step": 4277
+    },
+    {
+      "epoch": 1.803035093265887,
+      "grad_norm": 0.32691043615341187,
+      "learning_rate": 1.7611766491549277e-06,
+      "loss": 0.5713,
+      "step": 4278
+    },
+    {
+      "epoch": 1.80345663399726,
+      "grad_norm": 0.3069470524787903,
+      "learning_rate": 1.7601072858230417e-06,
+      "loss": 0.6043,
+      "step": 4279
+    },
+    {
+      "epoch": 1.803878174728633,
+      "grad_norm": 0.2942518889904022,
+      "learning_rate": 1.7590380708417642e-06,
+      "loss": 0.5836,
+      "step": 4280
+    },
+    {
+      "epoch": 1.8042997154600062,
+      "grad_norm": 0.30035078525543213,
+      "learning_rate": 1.7579690044254733e-06,
+      "loss": 0.6074,
+      "step": 4281
+    },
+    {
+      "epoch": 1.8047212561913795,
+      "grad_norm": 0.30913764238357544,
+      "learning_rate": 1.7569000867885216e-06,
+      "loss": 0.5779,
+      "step": 4282
+    },
+    {
+      "epoch": 1.8051427969227527,
+      "grad_norm": 0.310979425907135,
+      "learning_rate": 1.7558313181452296e-06,
+      "loss": 0.6027,
+      "step": 4283
+    },
+    {
+      "epoch": 1.805564337654126,
+      "grad_norm": 0.30673518776893616,
+      "learning_rate": 1.7547626987098893e-06,
+      "loss": 0.5791,
+      "step": 4284
+    },
+    {
+      "epoch": 1.805985878385499,
+      "grad_norm": 0.284819096326828,
+      "learning_rate": 1.7536942286967605e-06,
+      "loss": 0.5851,
+      "step": 4285
+    },
+    {
+      "epoch": 1.8064074191168722,
+      "grad_norm": 0.2829708456993103,
+      "learning_rate": 1.7526259083200757e-06,
+      "loss": 0.5764,
+      "step": 4286
+    },
+    {
+      "epoch": 1.8068289598482452,
+      "grad_norm": 0.30966663360595703,
+      "learning_rate": 1.751557737794035e-06,
+      "loss": 0.5864,
+      "step": 4287
+    },
+    {
+      "epoch": 1.8072505005796184,
+      "grad_norm": 0.2869812548160553,
+      "learning_rate": 1.7504897173328101e-06,
+      "loss": 0.5882,
+      "step": 4288
+    },
+    {
+      "epoch": 1.8076720413109917,
+      "grad_norm": 0.29545900225639343,
+      "learning_rate": 1.7494218471505437e-06,
+      "loss": 0.6119,
+      "step": 4289
+    },
+    {
+      "epoch": 1.808093582042365,
+      "grad_norm": 0.30397456884384155,
+      "learning_rate": 1.7483541274613441e-06,
+      "loss": 0.5693,
+      "step": 4290
+    },
+    {
+      "epoch": 1.8085151227737382,
+      "grad_norm": 0.31019827723503113,
+      "learning_rate": 1.7472865584792944e-06,
+      "loss": 0.5801,
+      "step": 4291
+    },
+    {
+      "epoch": 1.8089366635051112,
+      "grad_norm": 0.3037862181663513,
+      "learning_rate": 1.746219140418444e-06,
+      "loss": 0.5786,
+      "step": 4292
+    },
+    {
+      "epoch": 1.8093582042364842,
+      "grad_norm": 0.3005855679512024,
+      "learning_rate": 1.7451518734928146e-06,
+      "loss": 0.6116,
+      "step": 4293
+    },
+    {
+      "epoch": 1.8097797449678574,
+      "grad_norm": 0.2913413345813751,
+      "learning_rate": 1.7440847579163944e-06,
+      "loss": 0.587,
+      "step": 4294
+    },
+    {
+      "epoch": 1.8102012856992307,
+      "grad_norm": 0.2954038083553314,
+      "learning_rate": 1.743017793903145e-06,
+      "loss": 0.6054,
+      "step": 4295
+    },
+    {
+      "epoch": 1.810622826430604,
+      "grad_norm": 0.2997165620326996,
+      "learning_rate": 1.7419509816669942e-06,
+      "loss": 0.5966,
+      "step": 4296
+    },
+    {
+      "epoch": 1.8110443671619771,
+      "grad_norm": 0.3062772750854492,
+      "learning_rate": 1.7408843214218418e-06,
+      "loss": 0.5949,
+      "step": 4297
+    },
+    {
+      "epoch": 1.8114659078933502,
+      "grad_norm": 0.29827800393104553,
+      "learning_rate": 1.7398178133815557e-06,
+      "loss": 0.577,
+      "step": 4298
+    },
+    {
+      "epoch": 1.8118874486247234,
+      "grad_norm": 0.30910423398017883,
+      "learning_rate": 1.7387514577599734e-06,
+      "loss": 0.593,
+      "step": 4299
+    },
+    {
+      "epoch": 1.8123089893560964,
+      "grad_norm": 0.28573939204216003,
+      "learning_rate": 1.7376852547709044e-06,
+      "loss": 0.5662,
+      "step": 4300
+    },
+    {
+      "epoch": 1.8127305300874696,
+      "grad_norm": 0.2885774075984955,
+      "learning_rate": 1.7366192046281232e-06,
+      "loss": 0.5816,
+      "step": 4301
+    },
+    {
+      "epoch": 1.8131520708188429,
+      "grad_norm": 0.30526044964790344,
+      "learning_rate": 1.7355533075453773e-06,
+      "loss": 0.586,
+      "step": 4302
+    },
+    {
+      "epoch": 1.8135736115502161,
+      "grad_norm": 0.3103809654712677,
+      "learning_rate": 1.734487563736381e-06,
+      "loss": 0.5791,
+      "step": 4303
+    },
+    {
+      "epoch": 1.8139951522815894,
+      "grad_norm": 0.31333956122398376,
+      "learning_rate": 1.73342197341482e-06,
+      "loss": 0.5931,
+      "step": 4304
+    },
+    {
+      "epoch": 1.8144166930129624,
+      "grad_norm": 0.29798024892807007,
+      "learning_rate": 1.7323565367943469e-06,
+      "loss": 0.5894,
+      "step": 4305
+    },
+    {
+      "epoch": 1.8148382337443354,
+      "grad_norm": 0.31670835614204407,
+      "learning_rate": 1.7312912540885866e-06,
+      "loss": 0.5789,
+      "step": 4306
+    },
+    {
+      "epoch": 1.8152597744757086,
+      "grad_norm": 0.304760605096817,
+      "learning_rate": 1.730226125511129e-06,
+      "loss": 0.5986,
+      "step": 4307
+    },
+    {
+      "epoch": 1.8156813152070819,
+      "grad_norm": 0.30336910486221313,
+      "learning_rate": 1.7291611512755374e-06,
+      "loss": 0.6004,
+      "step": 4308
+    },
+    {
+      "epoch": 1.816102855938455,
+      "grad_norm": 0.29163795709609985,
+      "learning_rate": 1.728096331595342e-06,
+      "loss": 0.6172,
+      "step": 4309
+    },
+    {
+      "epoch": 1.8165243966698283,
+      "grad_norm": 0.3027088940143585,
+      "learning_rate": 1.7270316666840414e-06,
+      "loss": 0.5919,
+      "step": 4310
+    },
+    {
+      "epoch": 1.8169459374012014,
+      "grad_norm": 0.30588147044181824,
+      "learning_rate": 1.7259671567551046e-06,
+      "loss": 0.5997,
+      "step": 4311
+    },
+    {
+      "epoch": 1.8173674781325746,
+      "grad_norm": 0.2876339852809906,
+      "learning_rate": 1.7249028020219682e-06,
+      "loss": 0.5831,
+      "step": 4312
+    },
+    {
+      "epoch": 1.8177890188639476,
+      "grad_norm": 0.3059397339820862,
+      "learning_rate": 1.7238386026980394e-06,
+      "loss": 0.5636,
+      "step": 4313
+    },
+    {
+      "epoch": 1.8182105595953209,
+      "grad_norm": 0.2923504114151001,
+      "learning_rate": 1.7227745589966921e-06,
+      "loss": 0.586,
+      "step": 4314
+    },
+    {
+      "epoch": 1.818632100326694,
+      "grad_norm": 0.29322677850723267,
+      "learning_rate": 1.7217106711312709e-06,
+      "loss": 0.5865,
+      "step": 4315
+    },
+    {
+      "epoch": 1.8190536410580673,
+      "grad_norm": 0.3132327198982239,
+      "learning_rate": 1.7206469393150872e-06,
+      "loss": 0.5955,
+      "step": 4316
+    },
+    {
+      "epoch": 1.8194751817894406,
+      "grad_norm": 0.28935447335243225,
+      "learning_rate": 1.7195833637614245e-06,
+      "loss": 0.5863,
+      "step": 4317
+    },
+    {
+      "epoch": 1.8198967225208136,
+      "grad_norm": 0.3162103295326233,
+      "learning_rate": 1.7185199446835298e-06,
+      "loss": 0.5765,
+      "step": 4318
+    },
+    {
+      "epoch": 1.8203182632521866,
+      "grad_norm": 0.30961042642593384,
+      "learning_rate": 1.7174566822946237e-06,
+      "loss": 0.5584,
+      "step": 4319
+    },
+    {
+      "epoch": 1.8207398039835598,
+      "grad_norm": 0.2992355227470398,
+      "learning_rate": 1.7163935768078937e-06,
+      "loss": 0.6087,
+      "step": 4320
+    },
+    {
+      "epoch": 1.821161344714933,
+      "grad_norm": 0.29378727078437805,
+      "learning_rate": 1.715330628436494e-06,
+      "loss": 0.5789,
+      "step": 4321
+    },
+    {
+      "epoch": 1.8215828854463063,
+      "grad_norm": 0.321458101272583,
+      "learning_rate": 1.71426783739355e-06,
+      "loss": 0.587,
+      "step": 4322
+    },
+    {
+      "epoch": 1.8220044261776795,
+      "grad_norm": 0.3069184124469757,
+      "learning_rate": 1.7132052038921532e-06,
+      "loss": 0.595,
+      "step": 4323
+    },
+    {
+      "epoch": 1.8224259669090526,
+      "grad_norm": 0.3073382079601288,
+      "learning_rate": 1.712142728145366e-06,
+      "loss": 0.5948,
+      "step": 4324
+    },
+    {
+      "epoch": 1.8228475076404258,
+      "grad_norm": 0.2956736385822296,
+      "learning_rate": 1.711080410366216e-06,
+      "loss": 0.5913,
+      "step": 4325
+    },
+    {
+      "epoch": 1.8232690483717988,
+      "grad_norm": 0.3103010058403015,
+      "learning_rate": 1.7100182507677035e-06,
+      "loss": 0.5989,
+      "step": 4326
+    },
+    {
+      "epoch": 1.823690589103172,
+      "grad_norm": 0.3246273100376129,
+      "learning_rate": 1.7089562495627914e-06,
+      "loss": 0.5696,
+      "step": 4327
+    },
+    {
+      "epoch": 1.8241121298345453,
+      "grad_norm": 0.3084535598754883,
+      "learning_rate": 1.7078944069644165e-06,
+      "loss": 0.588,
+      "step": 4328
+    },
+    {
+      "epoch": 1.8245336705659185,
+      "grad_norm": 0.2881917655467987,
+      "learning_rate": 1.7068327231854809e-06,
+      "loss": 0.5889,
+      "step": 4329
+    },
+    {
+      "epoch": 1.8249552112972915,
+      "grad_norm": 0.3049517869949341,
+      "learning_rate": 1.7057711984388545e-06,
+      "loss": 0.5866,
+      "step": 4330
+    },
+    {
+      "epoch": 1.8253767520286648,
+      "grad_norm": 0.29577553272247314,
+      "learning_rate": 1.704709832937377e-06,
+      "loss": 0.5646,
+      "step": 4331
+    },
+    {
+      "epoch": 1.8257982927600378,
+      "grad_norm": 0.3923010230064392,
+      "learning_rate": 1.703648626893853e-06,
+      "loss": 0.5838,
+      "step": 4332
+    },
+    {
+      "epoch": 1.826219833491411,
+      "grad_norm": 0.28638988733291626,
+      "learning_rate": 1.7025875805210606e-06,
+      "loss": 0.5593,
+      "step": 4333
+    },
+    {
+      "epoch": 1.8266413742227843,
+      "grad_norm": 0.3096811771392822,
+      "learning_rate": 1.7015266940317394e-06,
+      "loss": 0.5913,
+      "step": 4334
+    },
+    {
+      "epoch": 1.8270629149541575,
+      "grad_norm": 0.3181338310241699,
+      "learning_rate": 1.7004659676386029e-06,
+      "loss": 0.5809,
+      "step": 4335
+    },
+    {
+      "epoch": 1.8274844556855308,
+      "grad_norm": 0.3296385109424591,
+      "learning_rate": 1.6994054015543282e-06,
+      "loss": 0.5724,
+      "step": 4336
+    },
+    {
+      "epoch": 1.8279059964169038,
+      "grad_norm": 0.3349359333515167,
+      "learning_rate": 1.6983449959915626e-06,
+      "loss": 0.575,
+      "step": 4337
+    },
+    {
+      "epoch": 1.828327537148277,
+      "grad_norm": 0.3077762722969055,
+      "learning_rate": 1.6972847511629195e-06,
+      "loss": 0.5911,
+      "step": 4338
+    },
+    {
+      "epoch": 1.82874907787965,
+      "grad_norm": 0.31147363781929016,
+      "learning_rate": 1.6962246672809818e-06,
+      "loss": 0.5789,
+      "step": 4339
+    },
+    {
+      "epoch": 1.8291706186110233,
+      "grad_norm": 0.33471378684043884,
+      "learning_rate": 1.6951647445582996e-06,
+      "loss": 0.581,
+      "step": 4340
+    },
+    {
+      "epoch": 1.8295921593423965,
+      "grad_norm": 0.3052794635295868,
+      "learning_rate": 1.6941049832073892e-06,
+      "loss": 0.5771,
+      "step": 4341
+    },
+    {
+      "epoch": 1.8300137000737697,
+      "grad_norm": 0.3080030381679535,
+      "learning_rate": 1.6930453834407378e-06,
+      "loss": 0.5767,
+      "step": 4342
+    },
+    {
+      "epoch": 1.8304352408051427,
+      "grad_norm": 0.2931380271911621,
+      "learning_rate": 1.6919859454707957e-06,
+      "loss": 0.5611,
+      "step": 4343
+    },
+    {
+      "epoch": 1.830856781536516,
+      "grad_norm": 0.30189087986946106,
+      "learning_rate": 1.6909266695099855e-06,
+      "loss": 0.5661,
+      "step": 4344
+    },
+    {
+      "epoch": 1.831278322267889,
+      "grad_norm": 0.31320425868034363,
+      "learning_rate": 1.6898675557706933e-06,
+      "loss": 0.5903,
+      "step": 4345
+    },
+    {
+      "epoch": 1.8316998629992622,
+      "grad_norm": 0.3208194971084595,
+      "learning_rate": 1.6888086044652764e-06,
+      "loss": 0.5547,
+      "step": 4346
+    },
+    {
+      "epoch": 1.8321214037306355,
+      "grad_norm": 0.30140161514282227,
+      "learning_rate": 1.6877498158060553e-06,
+      "loss": 0.5795,
+      "step": 4347
+    },
+    {
+      "epoch": 1.8325429444620087,
+      "grad_norm": 0.29885879158973694,
+      "learning_rate": 1.6866911900053218e-06,
+      "loss": 0.5844,
+      "step": 4348
+    },
+    {
+      "epoch": 1.832964485193382,
+      "grad_norm": 0.29837149381637573,
+      "learning_rate": 1.6856327272753328e-06,
+      "loss": 0.5726,
+      "step": 4349
+    },
+    {
+      "epoch": 1.833386025924755,
+      "grad_norm": 0.3210144639015198,
+      "learning_rate": 1.6845744278283125e-06,
+      "loss": 0.5737,
+      "step": 4350
+    },
+    {
+      "epoch": 1.833807566656128,
+      "grad_norm": 0.31543806195259094,
+      "learning_rate": 1.683516291876455e-06,
+      "loss": 0.5909,
+      "step": 4351
+    },
+    {
+      "epoch": 1.8342291073875012,
+      "grad_norm": 0.2972790598869324,
+      "learning_rate": 1.6824583196319167e-06,
+      "loss": 0.5876,
+      "step": 4352
+    },
+    {
+      "epoch": 1.8346506481188745,
+      "grad_norm": 0.29509395360946655,
+      "learning_rate": 1.6814005113068266e-06,
+      "loss": 0.567,
+      "step": 4353
+    },
+    {
+      "epoch": 1.8350721888502477,
+      "grad_norm": 0.29364708065986633,
+      "learning_rate": 1.680342867113277e-06,
+      "loss": 0.5643,
+      "step": 4354
+    },
+    {
+      "epoch": 1.835493729581621,
+      "grad_norm": 0.29872432351112366,
+      "learning_rate": 1.6792853872633292e-06,
+      "loss": 0.5946,
+      "step": 4355
+    },
+    {
+      "epoch": 1.835915270312994,
+      "grad_norm": 0.3029753863811493,
+      "learning_rate": 1.6782280719690103e-06,
+      "loss": 0.5851,
+      "step": 4356
+    },
+    {
+      "epoch": 1.8363368110443672,
+      "grad_norm": 0.3099585175514221,
+      "learning_rate": 1.6771709214423143e-06,
+      "loss": 0.5776,
+      "step": 4357
+    },
+    {
+      "epoch": 1.8367583517757402,
+      "grad_norm": 0.2916717827320099,
+      "learning_rate": 1.676113935895206e-06,
+      "loss": 0.5802,
+      "step": 4358
+    },
+    {
+      "epoch": 1.8371798925071134,
+      "grad_norm": 0.3061559200286865,
+      "learning_rate": 1.6750571155396102e-06,
+      "loss": 0.5812,
+      "step": 4359
+    },
+    {
+      "epoch": 1.8376014332384867,
+      "grad_norm": 0.31335458159446716,
+      "learning_rate": 1.6740004605874256e-06,
+      "loss": 0.5868,
+      "step": 4360
+    },
+    {
+      "epoch": 1.83802297396986,
+      "grad_norm": 0.3194263279438019,
+      "learning_rate": 1.6729439712505124e-06,
+      "loss": 0.597,
+      "step": 4361
+    },
+    {
+      "epoch": 1.8384445147012332,
+      "grad_norm": 0.2984003722667694,
+      "learning_rate": 1.6718876477407009e-06,
+      "loss": 0.5788,
+      "step": 4362
+    },
+    {
+      "epoch": 1.8388660554326062,
+      "grad_norm": 0.29671311378479004,
+      "learning_rate": 1.6708314902697858e-06,
+      "loss": 0.5909,
+      "step": 4363
+    },
+    {
+      "epoch": 1.8392875961639792,
+      "grad_norm": 0.30245769023895264,
+      "learning_rate": 1.6697754990495316e-06,
+      "loss": 0.5947,
+      "step": 4364
+    },
+    {
+      "epoch": 1.8397091368953524,
+      "grad_norm": 0.28124964237213135,
+      "learning_rate": 1.6687196742916656e-06,
+      "loss": 0.5858,
+      "step": 4365
+    },
+    {
+      "epoch": 1.8401306776267257,
+      "grad_norm": 0.3168337345123291,
+      "learning_rate": 1.667664016207885e-06,
+      "loss": 0.5627,
+      "step": 4366
+    },
+    {
+      "epoch": 1.840552218358099,
+      "grad_norm": 0.3076010048389435,
+      "learning_rate": 1.6666085250098512e-06,
+      "loss": 0.568,
+      "step": 4367
+    },
+    {
+      "epoch": 1.8409737590894721,
+      "grad_norm": 0.2968995273113251,
+      "learning_rate": 1.6655532009091936e-06,
+      "loss": 0.592,
+      "step": 4368
+    },
+    {
+      "epoch": 1.8413952998208452,
+      "grad_norm": 0.29465004801750183,
+      "learning_rate": 1.6644980441175085e-06,
+      "loss": 0.5813,
+      "step": 4369
+    },
+    {
+      "epoch": 1.8418168405522184,
+      "grad_norm": 0.2923990488052368,
+      "learning_rate": 1.6634430548463567e-06,
+      "loss": 0.5707,
+      "step": 4370
+    },
+    {
+      "epoch": 1.8422383812835914,
+      "grad_norm": 0.30503353476524353,
+      "learning_rate": 1.6623882333072677e-06,
+      "loss": 0.5807,
+      "step": 4371
+    },
+    {
+      "epoch": 1.8426599220149646,
+      "grad_norm": 0.3090004026889801,
+      "learning_rate": 1.6613335797117353e-06,
+      "loss": 0.5645,
+      "step": 4372
+    },
+    {
+      "epoch": 1.8430814627463379,
+      "grad_norm": 0.30269956588745117,
+      "learning_rate": 1.6602790942712216e-06,
+      "loss": 0.5668,
+      "step": 4373
+    },
+    {
+      "epoch": 1.8435030034777111,
+      "grad_norm": 0.3002566993236542,
+      "learning_rate": 1.6592247771971523e-06,
+      "loss": 0.5647,
+      "step": 4374
+    },
+    {
+      "epoch": 1.8439245442090844,
+      "grad_norm": 0.30495119094848633,
+      "learning_rate": 1.6581706287009228e-06,
+      "loss": 0.5706,
+      "step": 4375
+    },
+    {
+      "epoch": 1.8443460849404574,
+      "grad_norm": 0.29376599192619324,
+      "learning_rate": 1.6571166489938913e-06,
+      "loss": 0.5706,
+      "step": 4376
+    },
+    {
+      "epoch": 1.8447676256718304,
+      "grad_norm": 0.2991569936275482,
+      "learning_rate": 1.656062838287384e-06,
+      "loss": 0.5698,
+      "step": 4377
+    },
+    {
+      "epoch": 1.8451891664032036,
+      "grad_norm": 0.29785484075546265,
+      "learning_rate": 1.6550091967926947e-06,
+      "loss": 0.5541,
+      "step": 4378
+    },
+    {
+      "epoch": 1.8456107071345769,
+      "grad_norm": 0.30279120802879333,
+      "learning_rate": 1.6539557247210796e-06,
+      "loss": 0.5725,
+      "step": 4379
+    },
+    {
+      "epoch": 1.84603224786595,
+      "grad_norm": 0.295825719833374,
+      "learning_rate": 1.6529024222837643e-06,
+      "loss": 0.5759,
+      "step": 4380
+    },
+    {
+      "epoch": 1.8464537885973233,
+      "grad_norm": 0.31239810585975647,
+      "learning_rate": 1.6518492896919374e-06,
+      "loss": 0.5686,
+      "step": 4381
+    },
+    {
+      "epoch": 1.8468753293286964,
+      "grad_norm": 0.29998430609703064,
+      "learning_rate": 1.6507963271567568e-06,
+      "loss": 0.5757,
+      "step": 4382
+    },
+    {
+      "epoch": 1.8472968700600696,
+      "grad_norm": 0.3046194612979889,
+      "learning_rate": 1.6497435348893427e-06,
+      "loss": 0.589,
+      "step": 4383
+    },
+    {
+      "epoch": 1.8477184107914426,
+      "grad_norm": 0.29391366243362427,
+      "learning_rate": 1.648690913100784e-06,
+      "loss": 0.5745,
+      "step": 4384
+    },
+    {
+      "epoch": 1.8481399515228158,
+      "grad_norm": 0.29473206400871277,
+      "learning_rate": 1.6476384620021336e-06,
+      "loss": 0.567,
+      "step": 4385
+    },
+    {
+      "epoch": 1.848561492254189,
+      "grad_norm": 0.28942492604255676,
+      "learning_rate": 1.6465861818044126e-06,
+      "loss": 0.5678,
+      "step": 4386
+    },
+    {
+      "epoch": 1.8489830329855623,
+      "grad_norm": 0.3142037093639374,
+      "learning_rate": 1.6455340727186036e-06,
+      "loss": 0.601,
+      "step": 4387
+    },
+    {
+      "epoch": 1.8494045737169356,
+      "grad_norm": 0.29193004965782166,
+      "learning_rate": 1.64448213495566e-06,
+      "loss": 0.6002,
+      "step": 4388
+    },
+    {
+      "epoch": 1.8498261144483086,
+      "grad_norm": 0.29999685287475586,
+      "learning_rate": 1.6434303687264975e-06,
+      "loss": 0.5925,
+      "step": 4389
+    },
+    {
+      "epoch": 1.8502476551796816,
+      "grad_norm": 0.3111407160758972,
+      "learning_rate": 1.6423787742419978e-06,
+      "loss": 0.5873,
+      "step": 4390
+    },
+    {
+      "epoch": 1.8506691959110548,
+      "grad_norm": 0.2911715507507324,
+      "learning_rate": 1.6413273517130091e-06,
+      "loss": 0.5736,
+      "step": 4391
+    },
+    {
+      "epoch": 1.851090736642428,
+      "grad_norm": 0.31390896439552307,
+      "learning_rate": 1.6402761013503439e-06,
+      "loss": 0.6087,
+      "step": 4392
+    },
+    {
+      "epoch": 1.8515122773738013,
+      "grad_norm": 0.3135656714439392,
+      "learning_rate": 1.639225023364782e-06,
+      "loss": 0.5739,
+      "step": 4393
+    },
+    {
+      "epoch": 1.8519338181051745,
+      "grad_norm": 0.29544106125831604,
+      "learning_rate": 1.638174117967066e-06,
+      "loss": 0.5743,
+      "step": 4394
+    },
+    {
+      "epoch": 1.8523553588365476,
+      "grad_norm": 0.31050094962120056,
+      "learning_rate": 1.6371233853679081e-06,
+      "loss": 0.5709,
+      "step": 4395
+    },
+    {
+      "epoch": 1.8527768995679208,
+      "grad_norm": 0.293112188577652,
+      "learning_rate": 1.6360728257779795e-06,
+      "loss": 0.6045,
+      "step": 4396
+    },
+    {
+      "epoch": 1.8531984402992938,
+      "grad_norm": 0.2984670400619507,
+      "learning_rate": 1.635022439407923e-06,
+      "loss": 0.5276,
+      "step": 4397
+    },
+    {
+      "epoch": 1.853619981030667,
+      "grad_norm": 0.3196445107460022,
+      "learning_rate": 1.6339722264683442e-06,
+      "loss": 0.597,
+      "step": 4398
+    },
+    {
+      "epoch": 1.8540415217620403,
+      "grad_norm": 0.3151151239871979,
+      "learning_rate": 1.6329221871698125e-06,
+      "loss": 0.6137,
+      "step": 4399
+    },
+    {
+      "epoch": 1.8544630624934135,
+      "grad_norm": 0.3098038136959076,
+      "learning_rate": 1.6318723217228646e-06,
+      "loss": 0.5636,
+      "step": 4400
+    },
+    {
+      "epoch": 1.8548846032247865,
+      "grad_norm": 0.3179657757282257,
+      "learning_rate": 1.630822630338001e-06,
+      "loss": 0.565,
+      "step": 4401
+    },
+    {
+      "epoch": 1.8553061439561598,
+      "grad_norm": 0.30420148372650146,
+      "learning_rate": 1.6297731132256886e-06,
+      "loss": 0.5849,
+      "step": 4402
+    },
+    {
+      "epoch": 1.8557276846875328,
+      "grad_norm": 0.294290155172348,
+      "learning_rate": 1.6287237705963572e-06,
+      "loss": 0.5823,
+      "step": 4403
+    },
+    {
+      "epoch": 1.856149225418906,
+      "grad_norm": 0.2946355640888214,
+      "learning_rate": 1.6276746026604052e-06,
+      "loss": 0.5881,
+      "step": 4404
+    },
+    {
+      "epoch": 1.8565707661502793,
+      "grad_norm": 0.29911452531814575,
+      "learning_rate": 1.6266256096281906e-06,
+      "loss": 0.569,
+      "step": 4405
+    },
+    {
+      "epoch": 1.8569923068816525,
+      "grad_norm": 0.28612715005874634,
+      "learning_rate": 1.6255767917100429e-06,
+      "loss": 0.5912,
+      "step": 4406
+    },
+    {
+      "epoch": 1.8574138476130257,
+      "grad_norm": 0.29932069778442383,
+      "learning_rate": 1.6245281491162513e-06,
+      "loss": 0.575,
+      "step": 4407
+    },
+    {
+      "epoch": 1.8578353883443988,
+      "grad_norm": 0.3026377260684967,
+      "learning_rate": 1.6234796820570714e-06,
+      "loss": 0.5757,
+      "step": 4408
+    },
+    {
+      "epoch": 1.858256929075772,
+      "grad_norm": 0.29513537883758545,
+      "learning_rate": 1.622431390742725e-06,
+      "loss": 0.5788,
+      "step": 4409
+    },
+    {
+      "epoch": 1.858678469807145,
+      "grad_norm": 0.2892386317253113,
+      "learning_rate": 1.6213832753833958e-06,
+      "loss": 0.6059,
+      "step": 4410
+    },
+    {
+      "epoch": 1.8591000105385183,
+      "grad_norm": 0.3020263612270355,
+      "learning_rate": 1.6203353361892365e-06,
+      "loss": 0.5603,
+      "step": 4411
+    },
+    {
+      "epoch": 1.8595215512698915,
+      "grad_norm": 0.30972155928611755,
+      "learning_rate": 1.6192875733703588e-06,
+      "loss": 0.5782,
+      "step": 4412
+    },
+    {
+      "epoch": 1.8599430920012647,
+      "grad_norm": 0.28544992208480835,
+      "learning_rate": 1.618239987136845e-06,
+      "loss": 0.5841,
+      "step": 4413
+    },
+    {
+      "epoch": 1.8603646327326377,
+      "grad_norm": 0.29876527190208435,
+      "learning_rate": 1.6171925776987374e-06,
+      "loss": 0.5697,
+      "step": 4414
+    },
+    {
+      "epoch": 1.860786173464011,
+      "grad_norm": 0.2975681722164154,
+      "learning_rate": 1.6161453452660456e-06,
+      "loss": 0.5814,
+      "step": 4415
+    },
+    {
+      "epoch": 1.861207714195384,
+      "grad_norm": 0.2949362099170685,
+      "learning_rate": 1.6150982900487416e-06,
+      "loss": 0.5563,
+      "step": 4416
+    },
+    {
+      "epoch": 1.8616292549267572,
+      "grad_norm": 0.29798629879951477,
+      "learning_rate": 1.6140514122567633e-06,
+      "loss": 0.594,
+      "step": 4417
+    },
+    {
+      "epoch": 1.8620507956581305,
+      "grad_norm": 0.2998580038547516,
+      "learning_rate": 1.6130047121000136e-06,
+      "loss": 0.5763,
+      "step": 4418
+    },
+    {
+      "epoch": 1.8624723363895037,
+      "grad_norm": 0.29672154784202576,
+      "learning_rate": 1.6119581897883573e-06,
+      "loss": 0.5645,
+      "step": 4419
+    },
+    {
+      "epoch": 1.862893877120877,
+      "grad_norm": 0.2981659173965454,
+      "learning_rate": 1.6109118455316275e-06,
+      "loss": 0.6006,
+      "step": 4420
+    },
+    {
+      "epoch": 1.86331541785225,
+      "grad_norm": 0.29635435342788696,
+      "learning_rate": 1.609865679539616e-06,
+      "loss": 0.5748,
+      "step": 4421
+    },
+    {
+      "epoch": 1.863736958583623,
+      "grad_norm": 0.28521671891212463,
+      "learning_rate": 1.6088196920220853e-06,
+      "loss": 0.5571,
+      "step": 4422
+    },
+    {
+      "epoch": 1.8641584993149962,
+      "grad_norm": 0.3039669990539551,
+      "learning_rate": 1.6077738831887567e-06,
+      "loss": 0.5874,
+      "step": 4423
+    },
+    {
+      "epoch": 1.8645800400463695,
+      "grad_norm": 0.30347737669944763,
+      "learning_rate": 1.606728253249319e-06,
+      "loss": 0.5645,
+      "step": 4424
+    },
+    {
+      "epoch": 1.8650015807777427,
+      "grad_norm": 0.30112120509147644,
+      "learning_rate": 1.605682802413423e-06,
+      "loss": 0.5608,
+      "step": 4425
+    },
+    {
+      "epoch": 1.865423121509116,
+      "grad_norm": 0.29494374990463257,
+      "learning_rate": 1.6046375308906853e-06,
+      "loss": 0.6011,
+      "step": 4426
+    },
+    {
+      "epoch": 1.865844662240489,
+      "grad_norm": 0.3104344606399536,
+      "learning_rate": 1.6035924388906866e-06,
+      "loss": 0.5622,
+      "step": 4427
+    },
+    {
+      "epoch": 1.8662662029718622,
+      "grad_norm": 0.29932528734207153,
+      "learning_rate": 1.6025475266229685e-06,
+      "loss": 0.5614,
+      "step": 4428
+    },
+    {
+      "epoch": 1.8666877437032352,
+      "grad_norm": 0.2958607077598572,
+      "learning_rate": 1.6015027942970418e-06,
+      "loss": 0.5661,
+      "step": 4429
+    },
+    {
+      "epoch": 1.8671092844346084,
+      "grad_norm": 0.29685842990875244,
+      "learning_rate": 1.6004582421223752e-06,
+      "loss": 0.5884,
+      "step": 4430
+    },
+    {
+      "epoch": 1.8675308251659817,
+      "grad_norm": 0.30494019389152527,
+      "learning_rate": 1.5994138703084075e-06,
+      "loss": 0.5552,
+      "step": 4431
+    },
+    {
+      "epoch": 1.867952365897355,
+      "grad_norm": 0.29227104783058167,
+      "learning_rate": 1.598369679064536e-06,
+      "loss": 0.5853,
+      "step": 4432
+    },
+    {
+      "epoch": 1.8683739066287282,
+      "grad_norm": 0.3018190264701843,
+      "learning_rate": 1.5973256686001255e-06,
+      "loss": 0.5653,
+      "step": 4433
+    },
+    {
+      "epoch": 1.8687954473601012,
+      "grad_norm": 0.2987942397594452,
+      "learning_rate": 1.596281839124502e-06,
+      "loss": 0.5778,
+      "step": 4434
+    },
+    {
+      "epoch": 1.8692169880914742,
+      "grad_norm": 0.2950928807258606,
+      "learning_rate": 1.595238190846957e-06,
+      "loss": 0.5766,
+      "step": 4435
+    },
+    {
+      "epoch": 1.8696385288228474,
+      "grad_norm": 0.28716373443603516,
+      "learning_rate": 1.5941947239767441e-06,
+      "loss": 0.5683,
+      "step": 4436
+    },
+    {
+      "epoch": 1.8700600695542207,
+      "grad_norm": 0.29570573568344116,
+      "learning_rate": 1.5931514387230818e-06,
+      "loss": 0.5891,
+      "step": 4437
+    },
+    {
+      "epoch": 1.870481610285594,
+      "grad_norm": 0.2882426083087921,
+      "learning_rate": 1.5921083352951527e-06,
+      "loss": 0.5913,
+      "step": 4438
+    },
+    {
+      "epoch": 1.8709031510169671,
+      "grad_norm": 0.3001704514026642,
+      "learning_rate": 1.591065413902101e-06,
+      "loss": 0.5753,
+      "step": 4439
+    },
+    {
+      "epoch": 1.8713246917483402,
+      "grad_norm": 0.2810992896556854,
+      "learning_rate": 1.5900226747530361e-06,
+      "loss": 0.5999,
+      "step": 4440
+    },
+    {
+      "epoch": 1.8717462324797134,
+      "grad_norm": 0.31385543942451477,
+      "learning_rate": 1.588980118057029e-06,
+      "loss": 0.5796,
+      "step": 4441
+    },
+    {
+      "epoch": 1.8721677732110864,
+      "grad_norm": 0.30682188272476196,
+      "learning_rate": 1.587937744023117e-06,
+      "loss": 0.5782,
+      "step": 4442
+    },
+    {
+      "epoch": 1.8725893139424596,
+      "grad_norm": 0.3030414879322052,
+      "learning_rate": 1.5868955528602974e-06,
+      "loss": 0.5806,
+      "step": 4443
+    },
+    {
+      "epoch": 1.8730108546738329,
+      "grad_norm": 0.2980286777019501,
+      "learning_rate": 1.585853544777534e-06,
+      "loss": 0.5682,
+      "step": 4444
+    },
+    {
+      "epoch": 1.8734323954052061,
+      "grad_norm": 0.2962467670440674,
+      "learning_rate": 1.5848117199837512e-06,
+      "loss": 0.6116,
+      "step": 4445
+    },
+    {
+      "epoch": 1.8738539361365794,
+      "grad_norm": 0.3035505414009094,
+      "learning_rate": 1.583770078687838e-06,
+      "loss": 0.5793,
+      "step": 4446
+    },
+    {
+      "epoch": 1.8742754768679524,
+      "grad_norm": 0.29458993673324585,
+      "learning_rate": 1.5827286210986476e-06,
+      "loss": 0.5856,
+      "step": 4447
+    },
+    {
+      "epoch": 1.8746970175993254,
+      "grad_norm": 0.30196481943130493,
+      "learning_rate": 1.5816873474249939e-06,
+      "loss": 0.5593,
+      "step": 4448
+    },
+    {
+      "epoch": 1.8751185583306986,
+      "grad_norm": 0.2990199029445648,
+      "learning_rate": 1.580646257875656e-06,
+      "loss": 0.5638,
+      "step": 4449
+    },
+    {
+      "epoch": 1.8755400990620719,
+      "grad_norm": 0.30997008085250854,
+      "learning_rate": 1.579605352659375e-06,
+      "loss": 0.5805,
+      "step": 4450
+    },
+    {
+      "epoch": 1.875961639793445,
+      "grad_norm": 0.30490726232528687,
+      "learning_rate": 1.5785646319848557e-06,
+      "loss": 0.5978,
+      "step": 4451
+    },
+    {
+      "epoch": 1.8763831805248183,
+      "grad_norm": 0.31692323088645935,
+      "learning_rate": 1.5775240960607652e-06,
+      "loss": 0.5809,
+      "step": 4452
+    },
+    {
+      "epoch": 1.8768047212561914,
+      "grad_norm": 0.3026685118675232,
+      "learning_rate": 1.5764837450957343e-06,
+      "loss": 0.5839,
+      "step": 4453
+    },
+    {
+      "epoch": 1.8772262619875646,
+      "grad_norm": 0.2955408990383148,
+      "learning_rate": 1.5754435792983557e-06,
+      "loss": 0.5772,
+      "step": 4454
+    },
+    {
+      "epoch": 1.8776478027189376,
+      "grad_norm": 0.30411508679389954,
+      "learning_rate": 1.5744035988771867e-06,
+      "loss": 0.5848,
+      "step": 4455
+    },
+    {
+      "epoch": 1.8780693434503108,
+      "grad_norm": 0.28744637966156006,
+      "learning_rate": 1.5733638040407449e-06,
+      "loss": 0.5771,
+      "step": 4456
+    },
+    {
+      "epoch": 1.878490884181684,
+      "grad_norm": 0.29902106523513794,
+      "learning_rate": 1.5723241949975133e-06,
+      "loss": 0.5903,
+      "step": 4457
+    },
+    {
+      "epoch": 1.8789124249130573,
+      "grad_norm": 0.3046569228172302,
+      "learning_rate": 1.5712847719559366e-06,
+      "loss": 0.5857,
+      "step": 4458
+    },
+    {
+      "epoch": 1.8793339656444303,
+      "grad_norm": 0.283235639333725,
+      "learning_rate": 1.5702455351244208e-06,
+      "loss": 0.5735,
+      "step": 4459
+    },
+    {
+      "epoch": 1.8797555063758036,
+      "grad_norm": 0.30671319365501404,
+      "learning_rate": 1.5692064847113376e-06,
+      "loss": 0.6287,
+      "step": 4460
+    },
+    {
+      "epoch": 1.8801770471071766,
+      "grad_norm": 0.30714914202690125,
+      "learning_rate": 1.568167620925018e-06,
+      "loss": 0.5963,
+      "step": 4461
+    },
+    {
+      "epoch": 1.8805985878385498,
+      "grad_norm": 0.3014618158340454,
+      "learning_rate": 1.5671289439737587e-06,
+      "loss": 0.5795,
+      "step": 4462
+    },
+    {
+      "epoch": 1.881020128569923,
+      "grad_norm": 0.3004067540168762,
+      "learning_rate": 1.5660904540658151e-06,
+      "loss": 0.5652,
+      "step": 4463
+    },
+    {
+      "epoch": 1.8814416693012963,
+      "grad_norm": 0.30033108592033386,
+      "learning_rate": 1.5650521514094104e-06,
+      "loss": 0.5398,
+      "step": 4464
+    },
+    {
+      "epoch": 1.8818632100326695,
+      "grad_norm": 0.2952183485031128,
+      "learning_rate": 1.5640140362127243e-06,
+      "loss": 0.5782,
+      "step": 4465
+    },
+    {
+      "epoch": 1.8822847507640426,
+      "grad_norm": 0.2917748689651489,
+      "learning_rate": 1.562976108683904e-06,
+      "loss": 0.5807,
+      "step": 4466
+    },
+    {
+      "epoch": 1.8827062914954158,
+      "grad_norm": 0.29075175523757935,
+      "learning_rate": 1.5619383690310562e-06,
+      "loss": 0.5862,
+      "step": 4467
+    },
+    {
+      "epoch": 1.8831278322267888,
+      "grad_norm": 0.2971288561820984,
+      "learning_rate": 1.560900817462251e-06,
+      "loss": 0.5601,
+      "step": 4468
+    },
+    {
+      "epoch": 1.883549372958162,
+      "grad_norm": 0.3000071942806244,
+      "learning_rate": 1.5598634541855198e-06,
+      "loss": 0.5943,
+      "step": 4469
+    },
+    {
+      "epoch": 1.8839709136895353,
+      "grad_norm": 0.32184478640556335,
+      "learning_rate": 1.558826279408857e-06,
+      "loss": 0.5569,
+      "step": 4470
+    },
+    {
+      "epoch": 1.8843924544209085,
+      "grad_norm": 0.2952207922935486,
+      "learning_rate": 1.55778929334022e-06,
+      "loss": 0.588,
+      "step": 4471
+    },
+    {
+      "epoch": 1.8848139951522815,
+      "grad_norm": 0.29422956705093384,
+      "learning_rate": 1.5567524961875258e-06,
+      "loss": 0.5908,
+      "step": 4472
+    },
+    {
+      "epoch": 1.8852355358836548,
+      "grad_norm": 0.3135353922843933,
+      "learning_rate": 1.5557158881586576e-06,
+      "loss": 0.5903,
+      "step": 4473
+    },
+    {
+      "epoch": 1.8856570766150278,
+      "grad_norm": 0.30992406606674194,
+      "learning_rate": 1.5546794694614558e-06,
+      "loss": 0.6083,
+      "step": 4474
+    },
+    {
+      "epoch": 1.886078617346401,
+      "grad_norm": 0.286925733089447,
+      "learning_rate": 1.5536432403037271e-06,
+      "loss": 0.5897,
+      "step": 4475
+    },
+    {
+      "epoch": 1.8865001580777743,
+      "grad_norm": 0.2950926125049591,
+      "learning_rate": 1.5526072008932372e-06,
+      "loss": 0.5549,
+      "step": 4476
+    },
+    {
+      "epoch": 1.8869216988091475,
+      "grad_norm": 0.30131587386131287,
+      "learning_rate": 1.5515713514377155e-06,
+      "loss": 0.5747,
+      "step": 4477
+    },
+    {
+      "epoch": 1.8873432395405207,
+      "grad_norm": 0.3032747507095337,
+      "learning_rate": 1.5505356921448533e-06,
+      "loss": 0.5808,
+      "step": 4478
+    },
+    {
+      "epoch": 1.8877647802718938,
+      "grad_norm": 0.299258291721344,
+      "learning_rate": 1.5495002232223025e-06,
+      "loss": 0.5761,
+      "step": 4479
+    },
+    {
+      "epoch": 1.8881863210032668,
+      "grad_norm": 0.29161763191223145,
+      "learning_rate": 1.5484649448776781e-06,
+      "loss": 0.5827,
+      "step": 4480
+    },
+    {
+      "epoch": 1.88860786173464,
+      "grad_norm": 0.29021507501602173,
+      "learning_rate": 1.5474298573185551e-06,
+      "loss": 0.5552,
+      "step": 4481
+    },
+    {
+      "epoch": 1.8890294024660133,
+      "grad_norm": 0.2948141098022461,
+      "learning_rate": 1.5463949607524743e-06,
+      "loss": 0.5929,
+      "step": 4482
+    },
+    {
+      "epoch": 1.8894509431973865,
+      "grad_norm": 0.29191285371780396,
+      "learning_rate": 1.5453602553869318e-06,
+      "loss": 0.6026,
+      "step": 4483
+    },
+    {
+      "epoch": 1.8898724839287597,
+      "grad_norm": 0.30300724506378174,
+      "learning_rate": 1.5443257414293916e-06,
+      "loss": 0.6016,
+      "step": 4484
+    },
+    {
+      "epoch": 1.8902940246601327,
+      "grad_norm": 0.30841079354286194,
+      "learning_rate": 1.5432914190872757e-06,
+      "loss": 0.5965,
+      "step": 4485
+    },
+    {
+      "epoch": 1.890715565391506,
+      "grad_norm": 0.29503944516181946,
+      "learning_rate": 1.5422572885679694e-06,
+      "loss": 0.6008,
+      "step": 4486
+    },
+    {
+      "epoch": 1.891137106122879,
+      "grad_norm": 0.2919255197048187,
+      "learning_rate": 1.5412233500788183e-06,
+      "loss": 0.593,
+      "step": 4487
+    },
+    {
+      "epoch": 1.8915586468542522,
+      "grad_norm": 0.30569201707839966,
+      "learning_rate": 1.5401896038271291e-06,
+      "loss": 0.5939,
+      "step": 4488
+    },
+    {
+      "epoch": 1.8919801875856255,
+      "grad_norm": 0.3056679368019104,
+      "learning_rate": 1.5391560500201736e-06,
+      "loss": 0.5814,
+      "step": 4489
+    },
+    {
+      "epoch": 1.8924017283169987,
+      "grad_norm": 0.30033791065216064,
+      "learning_rate": 1.5381226888651788e-06,
+      "loss": 0.5764,
+      "step": 4490
+    },
+    {
+      "epoch": 1.892823269048372,
+      "grad_norm": 0.30453258752822876,
+      "learning_rate": 1.5370895205693393e-06,
+      "loss": 0.577,
+      "step": 4491
+    },
+    {
+      "epoch": 1.893244809779745,
+      "grad_norm": 0.3116928040981293,
+      "learning_rate": 1.5360565453398068e-06,
+      "loss": 0.5874,
+      "step": 4492
+    },
+    {
+      "epoch": 1.893666350511118,
+      "grad_norm": 0.28875428438186646,
+      "learning_rate": 1.5350237633836973e-06,
+      "loss": 0.5862,
+      "step": 4493
+    },
+    {
+      "epoch": 1.8940878912424912,
+      "grad_norm": 0.3176228106021881,
+      "learning_rate": 1.5339911749080846e-06,
+      "loss": 0.5724,
+      "step": 4494
+    },
+    {
+      "epoch": 1.8945094319738645,
+      "grad_norm": 0.29423555731773376,
+      "learning_rate": 1.532958780120007e-06,
+      "loss": 0.5797,
+      "step": 4495
+    },
+    {
+      "epoch": 1.8949309727052377,
+      "grad_norm": 0.2845533490180969,
+      "learning_rate": 1.5319265792264624e-06,
+      "loss": 0.5827,
+      "step": 4496
+    },
+    {
+      "epoch": 1.895352513436611,
+      "grad_norm": 0.29710453748703003,
+      "learning_rate": 1.5308945724344087e-06,
+      "loss": 0.5948,
+      "step": 4497
+    },
+    {
+      "epoch": 1.895774054167984,
+      "grad_norm": 0.29972243309020996,
+      "learning_rate": 1.5298627599507694e-06,
+      "loss": 0.5688,
+      "step": 4498
+    },
+    {
+      "epoch": 1.8961955948993572,
+      "grad_norm": 0.3140467703342438,
+      "learning_rate": 1.528831141982422e-06,
+      "loss": 0.5905,
+      "step": 4499
+    },
+    {
+      "epoch": 1.8966171356307302,
+      "grad_norm": 0.31190359592437744,
+      "learning_rate": 1.5277997187362114e-06,
+      "loss": 0.6062,
+      "step": 4500
+    },
+    {
+      "epoch": 1.8970386763621034,
+      "grad_norm": 0.30720430612564087,
+      "learning_rate": 1.52676849041894e-06,
+      "loss": 0.5769,
+      "step": 4501
+    },
+    {
+      "epoch": 1.8974602170934767,
+      "grad_norm": 0.3055114448070526,
+      "learning_rate": 1.5257374572373728e-06,
+      "loss": 0.57,
+      "step": 4502
+    },
+    {
+      "epoch": 1.89788175782485,
+      "grad_norm": 0.3037607967853546,
+      "learning_rate": 1.5247066193982338e-06,
+      "loss": 0.5973,
+      "step": 4503
+    },
+    {
+      "epoch": 1.8983032985562232,
+      "grad_norm": 0.3058910667896271,
+      "learning_rate": 1.5236759771082098e-06,
+      "loss": 0.5555,
+      "step": 4504
+    },
+    {
+      "epoch": 1.8987248392875962,
+      "grad_norm": 0.30457252264022827,
+      "learning_rate": 1.5226455305739468e-06,
+      "loss": 0.5811,
+      "step": 4505
+    },
+    {
+      "epoch": 1.8991463800189692,
+      "grad_norm": 0.3105520009994507,
+      "learning_rate": 1.5216152800020523e-06,
+      "loss": 0.5947,
+      "step": 4506
+    },
+    {
+      "epoch": 1.8995679207503424,
+      "grad_norm": 0.30968064069747925,
+      "learning_rate": 1.520585225599096e-06,
+      "loss": 0.5646,
+      "step": 4507
+    },
+    {
+      "epoch": 1.8999894614817157,
+      "grad_norm": 0.30218222737312317,
+      "learning_rate": 1.5195553675716043e-06,
+      "loss": 0.6014,
+      "step": 4508
+    },
+    {
+      "epoch": 1.900411002213089,
+      "grad_norm": 0.31163984537124634,
+      "learning_rate": 1.5185257061260692e-06,
+      "loss": 0.5704,
+      "step": 4509
+    },
+    {
+      "epoch": 1.9008325429444621,
+      "grad_norm": 0.2980576753616333,
+      "learning_rate": 1.5174962414689387e-06,
+      "loss": 0.5981,
+      "step": 4510
+    },
+    {
+      "epoch": 1.9012540836758351,
+      "grad_norm": 0.29494360089302063,
+      "learning_rate": 1.5164669738066252e-06,
+      "loss": 0.5892,
+      "step": 4511
+    },
+    {
+      "epoch": 1.9016756244072084,
+      "grad_norm": 0.29181554913520813,
+      "learning_rate": 1.515437903345498e-06,
+      "loss": 0.5952,
+      "step": 4512
+    },
+    {
+      "epoch": 1.9020971651385814,
+      "grad_norm": 0.2940387427806854,
+      "learning_rate": 1.51440903029189e-06,
+      "loss": 0.5917,
+      "step": 4513
+    },
+    {
+      "epoch": 1.9025187058699546,
+      "grad_norm": 0.2979465425014496,
+      "learning_rate": 1.5133803548520926e-06,
+      "loss": 0.565,
+      "step": 4514
+    },
+    {
+      "epoch": 1.9029402466013279,
+      "grad_norm": 0.293889582157135,
+      "learning_rate": 1.5123518772323578e-06,
+      "loss": 0.6051,
+      "step": 4515
+    },
+    {
+      "epoch": 1.9033617873327011,
+      "grad_norm": 0.30419933795928955,
+      "learning_rate": 1.5113235976388996e-06,
+      "loss": 0.5998,
+      "step": 4516
+    },
+    {
+      "epoch": 1.9037833280640744,
+      "grad_norm": 0.30472204089164734,
+      "learning_rate": 1.51029551627789e-06,
+      "loss": 0.5995,
+      "step": 4517
+    },
+    {
+      "epoch": 1.9042048687954474,
+      "grad_norm": 0.3197183310985565,
+      "learning_rate": 1.509267633355463e-06,
+      "loss": 0.5788,
+      "step": 4518
+    },
+    {
+      "epoch": 1.9046264095268204,
+      "grad_norm": 0.30708879232406616,
+      "learning_rate": 1.5082399490777108e-06,
+      "loss": 0.5764,
+      "step": 4519
+    },
+    {
+      "epoch": 1.9050479502581936,
+      "grad_norm": 0.2969115376472473,
+      "learning_rate": 1.5072124636506885e-06,
+      "loss": 0.591,
+      "step": 4520
+    },
+    {
+      "epoch": 1.9054694909895669,
+      "grad_norm": 0.3024512529373169,
+      "learning_rate": 1.5061851772804088e-06,
+      "loss": 0.5996,
+      "step": 4521
+    },
+    {
+      "epoch": 1.90589103172094,
+      "grad_norm": 0.2955225706100464,
+      "learning_rate": 1.5051580901728458e-06,
+      "loss": 0.5756,
+      "step": 4522
+    },
+    {
+      "epoch": 1.9063125724523133,
+      "grad_norm": 0.29697927832603455,
+      "learning_rate": 1.5041312025339336e-06,
+      "loss": 0.5907,
+      "step": 4523
+    },
+    {
+      "epoch": 1.9067341131836864,
+      "grad_norm": 0.30706870555877686,
+      "learning_rate": 1.5031045145695663e-06,
+      "loss": 0.5875,
+      "step": 4524
+    },
+    {
+      "epoch": 1.9071556539150596,
+      "grad_norm": 0.29781344532966614,
+      "learning_rate": 1.5020780264855967e-06,
+      "loss": 0.5912,
+      "step": 4525
+    },
+    {
+      "epoch": 1.9075771946464326,
+      "grad_norm": 0.29279589653015137,
+      "learning_rate": 1.50105173848784e-06,
+      "loss": 0.5662,
+      "step": 4526
+    },
+    {
+      "epoch": 1.9079987353778058,
+      "grad_norm": 0.30588385462760925,
+      "learning_rate": 1.5000256507820692e-06,
+      "loss": 0.582,
+      "step": 4527
+    },
+    {
+      "epoch": 1.908420276109179,
+      "grad_norm": 0.3080850839614868,
+      "learning_rate": 1.498999763574018e-06,
+      "loss": 0.5823,
+      "step": 4528
+    },
+    {
+      "epoch": 1.9088418168405523,
+      "grad_norm": 0.31531137228012085,
+      "learning_rate": 1.4979740770693802e-06,
+      "loss": 0.567,
+      "step": 4529
+    },
+    {
+      "epoch": 1.9092633575719253,
+      "grad_norm": 0.2887154817581177,
+      "learning_rate": 1.4969485914738074e-06,
+      "loss": 0.5929,
+      "step": 4530
+    },
+    {
+      "epoch": 1.9096848983032986,
+      "grad_norm": 0.29872772097587585,
+      "learning_rate": 1.495923306992914e-06,
+      "loss": 0.597,
+      "step": 4531
+    },
+    {
+      "epoch": 1.9101064390346716,
+      "grad_norm": 0.29551443457603455,
+      "learning_rate": 1.4948982238322712e-06,
+      "loss": 0.605,
+      "step": 4532
+    },
+    {
+      "epoch": 1.9105279797660448,
+      "grad_norm": 0.2895394265651703,
+      "learning_rate": 1.4938733421974125e-06,
+      "loss": 0.5775,
+      "step": 4533
+    },
+    {
+      "epoch": 1.910949520497418,
+      "grad_norm": 0.2920834720134735,
+      "learning_rate": 1.4928486622938277e-06,
+      "loss": 0.581,
+      "step": 4534
+    },
+    {
+      "epoch": 1.9113710612287913,
+      "grad_norm": 0.3090733289718628,
+      "learning_rate": 1.4918241843269697e-06,
+      "loss": 0.5774,
+      "step": 4535
+    },
+    {
+      "epoch": 1.9117926019601645,
+      "grad_norm": 0.3010360896587372,
+      "learning_rate": 1.4907999085022489e-06,
+      "loss": 0.589,
+      "step": 4536
+    },
+    {
+      "epoch": 1.9122141426915376,
+      "grad_norm": 0.2988319396972656,
+      "learning_rate": 1.4897758350250351e-06,
+      "loss": 0.5975,
+      "step": 4537
+    },
+    {
+      "epoch": 1.9126356834229108,
+      "grad_norm": 0.29945817589759827,
+      "learning_rate": 1.4887519641006587e-06,
+      "loss": 0.6079,
+      "step": 4538
+    },
+    {
+      "epoch": 1.9130572241542838,
+      "grad_norm": 0.3017979860305786,
+      "learning_rate": 1.4877282959344078e-06,
+      "loss": 0.5614,
+      "step": 4539
+    },
+    {
+      "epoch": 1.913478764885657,
+      "grad_norm": 0.3044842779636383,
+      "learning_rate": 1.4867048307315318e-06,
+      "loss": 0.5798,
+      "step": 4540
+    },
+    {
+      "epoch": 1.9139003056170303,
+      "grad_norm": 0.295697420835495,
+      "learning_rate": 1.4856815686972367e-06,
+      "loss": 0.582,
+      "step": 4541
+    },
+    {
+      "epoch": 1.9143218463484035,
+      "grad_norm": 0.3095501959323883,
+      "learning_rate": 1.4846585100366922e-06,
+      "loss": 0.5857,
+      "step": 4542
+    },
+    {
+      "epoch": 1.9147433870797765,
+      "grad_norm": 0.29692405462265015,
+      "learning_rate": 1.4836356549550216e-06,
+      "loss": 0.5566,
+      "step": 4543
+    },
+    {
+      "epoch": 1.9151649278111498,
+      "grad_norm": 0.3118845224380493,
+      "learning_rate": 1.4826130036573122e-06,
+      "loss": 0.5898,
+      "step": 4544
+    },
+    {
+      "epoch": 1.9155864685425228,
+      "grad_norm": 0.2969605028629303,
+      "learning_rate": 1.4815905563486076e-06,
+      "loss": 0.5632,
+      "step": 4545
+    },
+    {
+      "epoch": 1.916008009273896,
+      "grad_norm": 0.305317759513855,
+      "learning_rate": 1.4805683132339122e-06,
+      "loss": 0.5913,
+      "step": 4546
+    },
+    {
+      "epoch": 1.9164295500052693,
+      "grad_norm": 0.318040668964386,
+      "learning_rate": 1.4795462745181882e-06,
+      "loss": 0.5704,
+      "step": 4547
+    },
+    {
+      "epoch": 1.9168510907366425,
+      "grad_norm": 0.3117920458316803,
+      "learning_rate": 1.4785244404063571e-06,
+      "loss": 0.6001,
+      "step": 4548
+    },
+    {
+      "epoch": 1.9172726314680157,
+      "grad_norm": 0.29056215286254883,
+      "learning_rate": 1.4775028111032999e-06,
+      "loss": 0.5668,
+      "step": 4549
+    },
+    {
+      "epoch": 1.9176941721993888,
+      "grad_norm": 0.2960626184940338,
+      "learning_rate": 1.476481386813855e-06,
+      "loss": 0.5709,
+      "step": 4550
+    },
+    {
+      "epoch": 1.9181157129307618,
+      "grad_norm": 0.2983936667442322,
+      "learning_rate": 1.4754601677428237e-06,
+      "loss": 0.5677,
+      "step": 4551
+    },
+    {
+      "epoch": 1.918537253662135,
+      "grad_norm": 0.3054211437702179,
+      "learning_rate": 1.4744391540949602e-06,
+      "loss": 0.5864,
+      "step": 4552
+    },
+    {
+      "epoch": 1.9189587943935082,
+      "grad_norm": 0.2943073511123657,
+      "learning_rate": 1.473418346074983e-06,
+      "loss": 0.5907,
+      "step": 4553
+    },
+    {
+      "epoch": 1.9193803351248815,
+      "grad_norm": 0.29889440536499023,
+      "learning_rate": 1.4723977438875652e-06,
+      "loss": 0.5845,
+      "step": 4554
+    },
+    {
+      "epoch": 1.9198018758562547,
+      "grad_norm": 0.30065566301345825,
+      "learning_rate": 1.4713773477373416e-06,
+      "loss": 0.5715,
+      "step": 4555
+    },
+    {
+      "epoch": 1.9202234165876277,
+      "grad_norm": 0.29881367087364197,
+      "learning_rate": 1.470357157828905e-06,
+      "loss": 0.5716,
+      "step": 4556
+    },
+    {
+      "epoch": 1.920644957319001,
+      "grad_norm": 0.2953413426876068,
+      "learning_rate": 1.469337174366805e-06,
+      "loss": 0.5592,
+      "step": 4557
+    },
+    {
+      "epoch": 1.921066498050374,
+      "grad_norm": 0.3009633421897888,
+      "learning_rate": 1.4683173975555523e-06,
+      "loss": 0.5965,
+      "step": 4558
+    },
+    {
+      "epoch": 1.9214880387817472,
+      "grad_norm": 0.31027278304100037,
+      "learning_rate": 1.4672978275996135e-06,
+      "loss": 0.5694,
+      "step": 4559
+    },
+    {
+      "epoch": 1.9219095795131205,
+      "grad_norm": 0.30285608768463135,
+      "learning_rate": 1.4662784647034184e-06,
+      "loss": 0.579,
+      "step": 4560
+    },
+    {
+      "epoch": 1.9223311202444937,
+      "grad_norm": 0.30260223150253296,
+      "learning_rate": 1.4652593090713484e-06,
+      "loss": 0.5945,
+      "step": 4561
+    },
+    {
+      "epoch": 1.922752660975867,
+      "grad_norm": 0.3089960217475891,
+      "learning_rate": 1.4642403609077501e-06,
+      "loss": 0.5942,
+      "step": 4562
+    },
+    {
+      "epoch": 1.92317420170724,
+      "grad_norm": 0.2936738431453705,
+      "learning_rate": 1.463221620416924e-06,
+      "loss": 0.5809,
+      "step": 4563
+    },
+    {
+      "epoch": 1.923595742438613,
+      "grad_norm": 0.2952802777290344,
+      "learning_rate": 1.4622030878031306e-06,
+      "loss": 0.5957,
+      "step": 4564
+    },
+    {
+      "epoch": 1.9240172831699862,
+      "grad_norm": 0.29464957118034363,
+      "learning_rate": 1.4611847632705895e-06,
+      "loss": 0.5632,
+      "step": 4565
+    },
+    {
+      "epoch": 1.9244388239013595,
+      "grad_norm": 0.2993890345096588,
+      "learning_rate": 1.460166647023476e-06,
+      "loss": 0.58,
+      "step": 4566
+    },
+    {
+      "epoch": 1.9248603646327327,
+      "grad_norm": 0.296705037355423,
+      "learning_rate": 1.4591487392659281e-06,
+      "loss": 0.5745,
+      "step": 4567
+    },
+    {
+      "epoch": 1.925281905364106,
+      "grad_norm": 0.300998330116272,
+      "learning_rate": 1.4581310402020358e-06,
+      "loss": 0.5814,
+      "step": 4568
+    },
+    {
+      "epoch": 1.925703446095479,
+      "grad_norm": 0.3061726987361908,
+      "learning_rate": 1.4571135500358532e-06,
+      "loss": 0.5899,
+      "step": 4569
+    },
+    {
+      "epoch": 1.9261249868268522,
+      "grad_norm": 0.32992810010910034,
+      "learning_rate": 1.4560962689713892e-06,
+      "loss": 0.5406,
+      "step": 4570
+    },
+    {
+      "epoch": 1.9265465275582252,
+      "grad_norm": 0.2998550236225128,
+      "learning_rate": 1.455079197212612e-06,
+      "loss": 0.5771,
+      "step": 4571
+    },
+    {
+      "epoch": 1.9269680682895984,
+      "grad_norm": 0.31893542408943176,
+      "learning_rate": 1.4540623349634453e-06,
+      "loss": 0.5693,
+      "step": 4572
+    },
+    {
+      "epoch": 1.9273896090209717,
+      "grad_norm": 0.30457383394241333,
+      "learning_rate": 1.453045682427776e-06,
+      "loss": 0.585,
+      "step": 4573
+    },
+    {
+      "epoch": 1.927811149752345,
+      "grad_norm": 0.3062846064567566,
+      "learning_rate": 1.4520292398094432e-06,
+      "loss": 0.5886,
+      "step": 4574
+    },
+    {
+      "epoch": 1.9282326904837181,
+      "grad_norm": 0.30481889843940735,
+      "learning_rate": 1.4510130073122486e-06,
+      "loss": 0.5824,
+      "step": 4575
+    },
+    {
+      "epoch": 1.9286542312150912,
+      "grad_norm": 0.28911569714546204,
+      "learning_rate": 1.4499969851399488e-06,
+      "loss": 0.5888,
+      "step": 4576
+    },
+    {
+      "epoch": 1.9290757719464642,
+      "grad_norm": 0.30919134616851807,
+      "learning_rate": 1.4489811734962584e-06,
+      "loss": 0.597,
+      "step": 4577
+    },
+    {
+      "epoch": 1.9294973126778374,
+      "grad_norm": 0.3152669370174408,
+      "learning_rate": 1.4479655725848523e-06,
+      "loss": 0.5832,
+      "step": 4578
+    },
+    {
+      "epoch": 1.9299188534092107,
+      "grad_norm": 0.28585779666900635,
+      "learning_rate": 1.4469501826093602e-06,
+      "loss": 0.5771,
+      "step": 4579
+    },
+    {
+      "epoch": 1.930340394140584,
+      "grad_norm": 0.3024516701698303,
+      "learning_rate": 1.445935003773371e-06,
+      "loss": 0.5743,
+      "step": 4580
+    },
+    {
+      "epoch": 1.9307619348719571,
+      "grad_norm": 0.29903778433799744,
+      "learning_rate": 1.44492003628043e-06,
+      "loss": 0.5793,
+      "step": 4581
+    },
+    {
+      "epoch": 1.9311834756033301,
+      "grad_norm": 0.2927221357822418,
+      "learning_rate": 1.443905280334043e-06,
+      "loss": 0.5912,
+      "step": 4582
+    },
+    {
+      "epoch": 1.9316050163347034,
+      "grad_norm": 0.3054066300392151,
+      "learning_rate": 1.4428907361376694e-06,
+      "loss": 0.5623,
+      "step": 4583
+    },
+    {
+      "epoch": 1.9320265570660764,
+      "grad_norm": 0.2923785150051117,
+      "learning_rate": 1.4418764038947302e-06,
+      "loss": 0.5727,
+      "step": 4584
+    },
+    {
+      "epoch": 1.9324480977974496,
+      "grad_norm": 0.29538094997406006,
+      "learning_rate": 1.440862283808601e-06,
+      "loss": 0.589,
+      "step": 4585
+    },
+    {
+      "epoch": 1.9328696385288229,
+      "grad_norm": 0.30141472816467285,
+      "learning_rate": 1.4398483760826148e-06,
+      "loss": 0.5938,
+      "step": 4586
+    },
+    {
+      "epoch": 1.9332911792601961,
+      "grad_norm": 0.3017902970314026,
+      "learning_rate": 1.4388346809200649e-06,
+      "loss": 0.5524,
+      "step": 4587
+    },
+    {
+      "epoch": 1.9337127199915694,
+      "grad_norm": 0.28731709718704224,
+      "learning_rate": 1.4378211985241982e-06,
+      "loss": 0.5933,
+      "step": 4588
+    },
+    {
+      "epoch": 1.9341342607229424,
+      "grad_norm": 0.30299219489097595,
+      "learning_rate": 1.436807929098224e-06,
+      "loss": 0.5714,
+      "step": 4589
+    },
+    {
+      "epoch": 1.9345558014543154,
+      "grad_norm": 0.29303184151649475,
+      "learning_rate": 1.435794872845301e-06,
+      "loss": 0.5941,
+      "step": 4590
+    },
+    {
+      "epoch": 1.9349773421856886,
+      "grad_norm": 0.2993137240409851,
+      "learning_rate": 1.434782029968554e-06,
+      "loss": 0.6081,
+      "step": 4591
+    },
+    {
+      "epoch": 1.9353988829170619,
+      "grad_norm": 0.2940230667591095,
+      "learning_rate": 1.4337694006710575e-06,
+      "loss": 0.5608,
+      "step": 4592
+    },
+    {
+      "epoch": 1.935820423648435,
+      "grad_norm": 0.2894679009914398,
+      "learning_rate": 1.4327569851558498e-06,
+      "loss": 0.5827,
+      "step": 4593
+    },
+    {
+      "epoch": 1.9362419643798083,
+      "grad_norm": 0.31070685386657715,
+      "learning_rate": 1.4317447836259215e-06,
+      "loss": 0.5629,
+      "step": 4594
+    },
+    {
+      "epoch": 1.9366635051111813,
+      "grad_norm": 0.29463085532188416,
+      "learning_rate": 1.430732796284221e-06,
+      "loss": 0.5998,
+      "step": 4595
+    },
+    {
+      "epoch": 1.9370850458425546,
+      "grad_norm": 0.30977028608322144,
+      "learning_rate": 1.4297210233336566e-06,
+      "loss": 0.6126,
+      "step": 4596
+    },
+    {
+      "epoch": 1.9375065865739276,
+      "grad_norm": 0.30875885486602783,
+      "learning_rate": 1.4287094649770902e-06,
+      "loss": 0.5882,
+      "step": 4597
+    },
+    {
+      "epoch": 1.9379281273053008,
+      "grad_norm": 0.3065561056137085,
+      "learning_rate": 1.4276981214173447e-06,
+      "loss": 0.5804,
+      "step": 4598
+    },
+    {
+      "epoch": 1.938349668036674,
+      "grad_norm": 0.32208576798439026,
+      "learning_rate": 1.426686992857193e-06,
+      "loss": 0.5644,
+      "step": 4599
+    },
+    {
+      "epoch": 1.9387712087680473,
+      "grad_norm": 0.3125014305114746,
+      "learning_rate": 1.4256760794993728e-06,
+      "loss": 0.571,
+      "step": 4600
+    },
+    {
+      "epoch": 1.9391927494994203,
+      "grad_norm": 0.3041897714138031,
+      "learning_rate": 1.424665381546573e-06,
+      "loss": 0.5698,
+      "step": 4601
+    },
+    {
+      "epoch": 1.9396142902307936,
+      "grad_norm": 0.2938082218170166,
+      "learning_rate": 1.4236548992014437e-06,
+      "loss": 0.5701,
+      "step": 4602
+    },
+    {
+      "epoch": 1.9400358309621666,
+      "grad_norm": 0.30883046984672546,
+      "learning_rate": 1.4226446326665877e-06,
+      "loss": 0.5822,
+      "step": 4603
+    },
+    {
+      "epoch": 1.9404573716935398,
+      "grad_norm": 0.31547343730926514,
+      "learning_rate": 1.4216345821445665e-06,
+      "loss": 0.5923,
+      "step": 4604
+    },
+    {
+      "epoch": 1.940878912424913,
+      "grad_norm": 0.29558488726615906,
+      "learning_rate": 1.420624747837899e-06,
+      "loss": 0.5949,
+      "step": 4605
+    },
+    {
+      "epoch": 1.9413004531562863,
+      "grad_norm": 0.30665937066078186,
+      "learning_rate": 1.4196151299490586e-06,
+      "loss": 0.5668,
+      "step": 4606
+    },
+    {
+      "epoch": 1.9417219938876595,
+      "grad_norm": 0.29589587450027466,
+      "learning_rate": 1.418605728680479e-06,
+      "loss": 0.592,
+      "step": 4607
+    },
+    {
+      "epoch": 1.9421435346190326,
+      "grad_norm": 0.30573490262031555,
+      "learning_rate": 1.4175965442345446e-06,
+      "loss": 0.5607,
+      "step": 4608
+    },
+    {
+      "epoch": 1.9425650753504056,
+      "grad_norm": 0.2964244484901428,
+      "learning_rate": 1.4165875768136028e-06,
+      "loss": 0.6027,
+      "step": 4609
+    },
+    {
+      "epoch": 1.9429866160817788,
+      "grad_norm": 0.2971614897251129,
+      "learning_rate": 1.4155788266199516e-06,
+      "loss": 0.5679,
+      "step": 4610
+    },
+    {
+      "epoch": 1.943408156813152,
+      "grad_norm": 0.29606181383132935,
+      "learning_rate": 1.4145702938558515e-06,
+      "loss": 0.5928,
+      "step": 4611
+    },
+    {
+      "epoch": 1.9438296975445253,
+      "grad_norm": 0.2927257716655731,
+      "learning_rate": 1.4135619787235144e-06,
+      "loss": 0.5627,
+      "step": 4612
+    },
+    {
+      "epoch": 1.9442512382758985,
+      "grad_norm": 0.2961556911468506,
+      "learning_rate": 1.4125538814251105e-06,
+      "loss": 0.5947,
+      "step": 4613
+    },
+    {
+      "epoch": 1.9446727790072715,
+      "grad_norm": 0.29371100664138794,
+      "learning_rate": 1.4115460021627652e-06,
+      "loss": 0.589,
+      "step": 4614
+    },
+    {
+      "epoch": 1.9450943197386448,
+      "grad_norm": 0.2966255843639374,
+      "learning_rate": 1.4105383411385627e-06,
+      "loss": 0.5357,
+      "step": 4615
+    },
+    {
+      "epoch": 1.9455158604700178,
+      "grad_norm": 0.2974972724914551,
+      "learning_rate": 1.409530898554542e-06,
+      "loss": 0.5664,
+      "step": 4616
+    },
+    {
+      "epoch": 1.945937401201391,
+      "grad_norm": 0.30105337500572205,
+      "learning_rate": 1.4085236746126975e-06,
+      "loss": 0.6023,
+      "step": 4617
+    },
+    {
+      "epoch": 1.9463589419327643,
+      "grad_norm": 0.30153408646583557,
+      "learning_rate": 1.407516669514981e-06,
+      "loss": 0.5916,
+      "step": 4618
+    },
+    {
+      "epoch": 1.9467804826641375,
+      "grad_norm": 0.2971515953540802,
+      "learning_rate": 1.4065098834632986e-06,
+      "loss": 0.5549,
+      "step": 4619
+    },
+    {
+      "epoch": 1.9472020233955107,
+      "grad_norm": 0.30676528811454773,
+      "learning_rate": 1.4055033166595158e-06,
+      "loss": 0.5938,
+      "step": 4620
+    },
+    {
+      "epoch": 1.9476235641268838,
+      "grad_norm": 0.2994089424610138,
+      "learning_rate": 1.404496969305451e-06,
+      "loss": 0.5945,
+      "step": 4621
+    },
+    {
+      "epoch": 1.9480451048582568,
+      "grad_norm": 0.2935415208339691,
+      "learning_rate": 1.4034908416028798e-06,
+      "loss": 0.5792,
+      "step": 4622
+    },
+    {
+      "epoch": 1.94846664558963,
+      "grad_norm": 0.3015369474887848,
+      "learning_rate": 1.402484933753533e-06,
+      "loss": 0.5928,
+      "step": 4623
+    },
+    {
+      "epoch": 1.9488881863210032,
+      "grad_norm": 0.2903369069099426,
+      "learning_rate": 1.4014792459590982e-06,
+      "loss": 0.577,
+      "step": 4624
+    },
+    {
+      "epoch": 1.9493097270523765,
+      "grad_norm": 0.3168697953224182,
+      "learning_rate": 1.4004737784212205e-06,
+      "loss": 0.5428,
+      "step": 4625
+    },
+    {
+      "epoch": 1.9497312677837497,
+      "grad_norm": 0.2955181300640106,
+      "learning_rate": 1.3994685313414974e-06,
+      "loss": 0.5724,
+      "step": 4626
+    },
+    {
+      "epoch": 1.9501528085151227,
+      "grad_norm": 0.3027075529098511,
+      "learning_rate": 1.3984635049214846e-06,
+      "loss": 0.5904,
+      "step": 4627
+    },
+    {
+      "epoch": 1.950574349246496,
+      "grad_norm": 0.29810813069343567,
+      "learning_rate": 1.397458699362691e-06,
+      "loss": 0.5544,
+      "step": 4628
+    },
+    {
+      "epoch": 1.950995889977869,
+      "grad_norm": 0.29686400294303894,
+      "learning_rate": 1.3964541148665853e-06,
+      "loss": 0.5914,
+      "step": 4629
+    },
+    {
+      "epoch": 1.9514174307092422,
+      "grad_norm": 0.3061332702636719,
+      "learning_rate": 1.3954497516345886e-06,
+      "loss": 0.5868,
+      "step": 4630
+    },
+    {
+      "epoch": 1.9518389714406155,
+      "grad_norm": 0.2901374101638794,
+      "learning_rate": 1.3944456098680786e-06,
+      "loss": 0.5767,
+      "step": 4631
+    },
+    {
+      "epoch": 1.9522605121719887,
+      "grad_norm": 0.29686108231544495,
+      "learning_rate": 1.3934416897683872e-06,
+      "loss": 0.5845,
+      "step": 4632
+    },
+    {
+      "epoch": 1.952682052903362,
+      "grad_norm": 0.28853079676628113,
+      "learning_rate": 1.3924379915368053e-06,
+      "loss": 0.5821,
+      "step": 4633
+    },
+    {
+      "epoch": 1.953103593634735,
+      "grad_norm": 0.3024919927120209,
+      "learning_rate": 1.3914345153745771e-06,
+      "loss": 0.5897,
+      "step": 4634
+    },
+    {
+      "epoch": 1.953525134366108,
+      "grad_norm": 0.2967340350151062,
+      "learning_rate": 1.3904312614829018e-06,
+      "loss": 0.5897,
+      "step": 4635
+    },
+    {
+      "epoch": 1.9539466750974812,
+      "grad_norm": 0.2953436076641083,
+      "learning_rate": 1.3894282300629353e-06,
+      "loss": 0.6137,
+      "step": 4636
+    },
+    {
+      "epoch": 1.9543682158288544,
+      "grad_norm": 0.2965579926967621,
+      "learning_rate": 1.3884254213157863e-06,
+      "loss": 0.572,
+      "step": 4637
+    },
+    {
+      "epoch": 1.9547897565602277,
+      "grad_norm": 0.29487472772598267,
+      "learning_rate": 1.3874228354425234e-06,
+      "loss": 0.554,
+      "step": 4638
+    },
+    {
+      "epoch": 1.955211297291601,
+      "grad_norm": 0.29605230689048767,
+      "learning_rate": 1.3864204726441672e-06,
+      "loss": 0.578,
+      "step": 4639
+    },
+    {
+      "epoch": 1.955632838022974,
+      "grad_norm": 0.29806187748908997,
+      "learning_rate": 1.3854183331216932e-06,
+      "loss": 0.5645,
+      "step": 4640
+    },
+    {
+      "epoch": 1.9560543787543472,
+      "grad_norm": 0.2935843765735626,
+      "learning_rate": 1.3844164170760335e-06,
+      "loss": 0.5773,
+      "step": 4641
+    },
+    {
+      "epoch": 1.9564759194857202,
+      "grad_norm": 0.3013414144515991,
+      "learning_rate": 1.3834147247080766e-06,
+      "loss": 0.5539,
+      "step": 4642
+    },
+    {
+      "epoch": 1.9568974602170934,
+      "grad_norm": 0.3085986077785492,
+      "learning_rate": 1.3824132562186627e-06,
+      "loss": 0.571,
+      "step": 4643
+    },
+    {
+      "epoch": 1.9573190009484667,
+      "grad_norm": 0.30693432688713074,
+      "learning_rate": 1.381412011808591e-06,
+      "loss": 0.592,
+      "step": 4644
+    },
+    {
+      "epoch": 1.95774054167984,
+      "grad_norm": 0.29129108786582947,
+      "learning_rate": 1.3804109916786128e-06,
+      "loss": 0.5779,
+      "step": 4645
+    },
+    {
+      "epoch": 1.9581620824112131,
+      "grad_norm": 0.3013078570365906,
+      "learning_rate": 1.379410196029435e-06,
+      "loss": 0.5917,
+      "step": 4646
+    },
+    {
+      "epoch": 1.9585836231425862,
+      "grad_norm": 0.29583704471588135,
+      "learning_rate": 1.378409625061722e-06,
+      "loss": 0.6015,
+      "step": 4647
+    },
+    {
+      "epoch": 1.9590051638739592,
+      "grad_norm": 0.29593995213508606,
+      "learning_rate": 1.377409278976089e-06,
+      "loss": 0.5807,
+      "step": 4648
+    },
+    {
+      "epoch": 1.9594267046053324,
+      "grad_norm": 0.2986476421356201,
+      "learning_rate": 1.3764091579731095e-06,
+      "loss": 0.5805,
+      "step": 4649
+    },
+    {
+      "epoch": 1.9598482453367057,
+      "grad_norm": 0.3030495345592499,
+      "learning_rate": 1.3754092622533091e-06,
+      "loss": 0.6078,
+      "step": 4650
+    },
+    {
+      "epoch": 1.960269786068079,
+      "grad_norm": 0.3079785406589508,
+      "learning_rate": 1.3744095920171719e-06,
+      "loss": 0.5743,
+      "step": 4651
+    },
+    {
+      "epoch": 1.9606913267994521,
+      "grad_norm": 0.2965410649776459,
+      "learning_rate": 1.3734101474651326e-06,
+      "loss": 0.5968,
+      "step": 4652
+    },
+    {
+      "epoch": 1.9611128675308251,
+      "grad_norm": 0.3046989142894745,
+      "learning_rate": 1.3724109287975845e-06,
+      "loss": 0.5636,
+      "step": 4653
+    },
+    {
+      "epoch": 1.9615344082621984,
+      "grad_norm": 0.3089590072631836,
+      "learning_rate": 1.3714119362148732e-06,
+      "loss": 0.572,
+      "step": 4654
+    },
+    {
+      "epoch": 1.9619559489935714,
+      "grad_norm": 0.3016713559627533,
+      "learning_rate": 1.3704131699172984e-06,
+      "loss": 0.5524,
+      "step": 4655
+    },
+    {
+      "epoch": 1.9623774897249446,
+      "grad_norm": 0.2958734929561615,
+      "learning_rate": 1.3694146301051175e-06,
+      "loss": 0.5733,
+      "step": 4656
+    },
+    {
+      "epoch": 1.9627990304563179,
+      "grad_norm": 0.30459874868392944,
+      "learning_rate": 1.3684163169785397e-06,
+      "loss": 0.5974,
+      "step": 4657
+    },
+    {
+      "epoch": 1.963220571187691,
+      "grad_norm": 0.2985385060310364,
+      "learning_rate": 1.3674182307377294e-06,
+      "loss": 0.5817,
+      "step": 4658
+    },
+    {
+      "epoch": 1.9636421119190641,
+      "grad_norm": 0.3125889003276825,
+      "learning_rate": 1.366420371582805e-06,
+      "loss": 0.5793,
+      "step": 4659
+    },
+    {
+      "epoch": 1.9640636526504374,
+      "grad_norm": 0.3037717640399933,
+      "learning_rate": 1.365422739713842e-06,
+      "loss": 0.6018,
+      "step": 4660
+    },
+    {
+      "epoch": 1.9644851933818104,
+      "grad_norm": 0.32527971267700195,
+      "learning_rate": 1.3644253353308666e-06,
+      "loss": 0.57,
+      "step": 4661
+    },
+    {
+      "epoch": 1.9649067341131836,
+      "grad_norm": 0.3246605396270752,
+      "learning_rate": 1.3634281586338631e-06,
+      "loss": 0.5606,
+      "step": 4662
+    },
+    {
+      "epoch": 1.9653282748445569,
+      "grad_norm": 0.29228079319000244,
+      "learning_rate": 1.3624312098227677e-06,
+      "loss": 0.5974,
+      "step": 4663
+    },
+    {
+      "epoch": 1.96574981557593,
+      "grad_norm": 0.30279770493507385,
+      "learning_rate": 1.3614344890974696e-06,
+      "loss": 0.5823,
+      "step": 4664
+    },
+    {
+      "epoch": 1.9661713563073033,
+      "grad_norm": 0.3035810589790344,
+      "learning_rate": 1.360437996657817e-06,
+      "loss": 0.5622,
+      "step": 4665
+    },
+    {
+      "epoch": 1.9665928970386763,
+      "grad_norm": 0.3064010739326477,
+      "learning_rate": 1.359441732703607e-06,
+      "loss": 0.5667,
+      "step": 4666
+    },
+    {
+      "epoch": 1.9670144377700496,
+      "grad_norm": 0.29581373929977417,
+      "learning_rate": 1.3584456974345961e-06,
+      "loss": 0.5763,
+      "step": 4667
+    },
+    {
+      "epoch": 1.9674359785014226,
+      "grad_norm": 0.305389404296875,
+      "learning_rate": 1.357449891050489e-06,
+      "loss": 0.5686,
+      "step": 4668
+    },
+    {
+      "epoch": 1.9678575192327958,
+      "grad_norm": 0.3082119822502136,
+      "learning_rate": 1.3564543137509506e-06,
+      "loss": 0.5678,
+      "step": 4669
+    },
+    {
+      "epoch": 1.968279059964169,
+      "grad_norm": 0.3004262149333954,
+      "learning_rate": 1.3554589657355949e-06,
+      "loss": 0.5782,
+      "step": 4670
+    },
+    {
+      "epoch": 1.9687006006955423,
+      "grad_norm": 0.3016479015350342,
+      "learning_rate": 1.3544638472039934e-06,
+      "loss": 0.5747,
+      "step": 4671
+    },
+    {
+      "epoch": 1.9691221414269153,
+      "grad_norm": 0.30888670682907104,
+      "learning_rate": 1.3534689583556699e-06,
+      "loss": 0.5804,
+      "step": 4672
+    },
+    {
+      "epoch": 1.9695436821582886,
+      "grad_norm": 0.30024394392967224,
+      "learning_rate": 1.3524742993901013e-06,
+      "loss": 0.5792,
+      "step": 4673
+    },
+    {
+      "epoch": 1.9699652228896616,
+      "grad_norm": 0.3005749583244324,
+      "learning_rate": 1.3514798705067217e-06,
+      "loss": 0.58,
+      "step": 4674
+    },
+    {
+      "epoch": 1.9703867636210348,
+      "grad_norm": 0.2902109920978546,
+      "learning_rate": 1.3504856719049148e-06,
+      "loss": 0.5749,
+      "step": 4675
+    },
+    {
+      "epoch": 1.970808304352408,
+      "grad_norm": 0.2963609993457794,
+      "learning_rate": 1.3494917037840226e-06,
+      "loss": 0.5735,
+      "step": 4676
+    },
+    {
+      "epoch": 1.9712298450837813,
+      "grad_norm": 0.30344250798225403,
+      "learning_rate": 1.3484979663433362e-06,
+      "loss": 0.5858,
+      "step": 4677
+    },
+    {
+      "epoch": 1.9716513858151545,
+      "grad_norm": 0.2965548038482666,
+      "learning_rate": 1.347504459782104e-06,
+      "loss": 0.5734,
+      "step": 4678
+    },
+    {
+      "epoch": 1.9720729265465275,
+      "grad_norm": 0.2932372987270355,
+      "learning_rate": 1.3465111842995262e-06,
+      "loss": 0.588,
+      "step": 4679
+    },
+    {
+      "epoch": 1.9724944672779006,
+      "grad_norm": 0.29379570484161377,
+      "learning_rate": 1.3455181400947587e-06,
+      "loss": 0.5899,
+      "step": 4680
+    },
+    {
+      "epoch": 1.9729160080092738,
+      "grad_norm": 0.3104860484600067,
+      "learning_rate": 1.344525327366909e-06,
+      "loss": 0.5764,
+      "step": 4681
+    },
+    {
+      "epoch": 1.973337548740647,
+      "grad_norm": 0.30391260981559753,
+      "learning_rate": 1.3435327463150388e-06,
+      "loss": 0.5684,
+      "step": 4682
+    },
+    {
+      "epoch": 1.9737590894720203,
+      "grad_norm": 0.29841527342796326,
+      "learning_rate": 1.3425403971381623e-06,
+      "loss": 0.5794,
+      "step": 4683
+    },
+    {
+      "epoch": 1.9741806302033935,
+      "grad_norm": 0.30236920714378357,
+      "learning_rate": 1.3415482800352497e-06,
+      "loss": 0.589,
+      "step": 4684
+    },
+    {
+      "epoch": 1.9746021709347665,
+      "grad_norm": 0.29592373967170715,
+      "learning_rate": 1.340556395205225e-06,
+      "loss": 0.5792,
+      "step": 4685
+    },
+    {
+      "epoch": 1.9750237116661398,
+      "grad_norm": 0.2936176061630249,
+      "learning_rate": 1.33956474284696e-06,
+      "loss": 0.5832,
+      "step": 4686
+    },
+    {
+      "epoch": 1.9754452523975128,
+      "grad_norm": 0.30533790588378906,
+      "learning_rate": 1.338573323159287e-06,
+      "loss": 0.5968,
+      "step": 4687
+    },
+    {
+      "epoch": 1.975866793128886,
+      "grad_norm": 0.3105771541595459,
+      "learning_rate": 1.3375821363409864e-06,
+      "loss": 0.5702,
+      "step": 4688
+    },
+    {
+      "epoch": 1.9762883338602593,
+      "grad_norm": 0.29880794882774353,
+      "learning_rate": 1.3365911825907958e-06,
+      "loss": 0.5819,
+      "step": 4689
+    },
+    {
+      "epoch": 1.9767098745916325,
+      "grad_norm": 0.29673486948013306,
+      "learning_rate": 1.3356004621074037e-06,
+      "loss": 0.5913,
+      "step": 4690
+    },
+    {
+      "epoch": 1.9771314153230057,
+      "grad_norm": 0.29334282875061035,
+      "learning_rate": 1.334609975089452e-06,
+      "loss": 0.6048,
+      "step": 4691
+    },
+    {
+      "epoch": 1.9775529560543788,
+      "grad_norm": 0.2981925904750824,
+      "learning_rate": 1.3336197217355351e-06,
+      "loss": 0.5675,
+      "step": 4692
+    },
+    {
+      "epoch": 1.9779744967857518,
+      "grad_norm": 0.2953258454799652,
+      "learning_rate": 1.3326297022442031e-06,
+      "loss": 0.6045,
+      "step": 4693
+    },
+    {
+      "epoch": 1.978396037517125,
+      "grad_norm": 0.297554075717926,
+      "learning_rate": 1.3316399168139583e-06,
+      "loss": 0.6071,
+      "step": 4694
+    },
+    {
+      "epoch": 1.9788175782484982,
+      "grad_norm": 0.31925079226493835,
+      "learning_rate": 1.3306503656432549e-06,
+      "loss": 0.5939,
+      "step": 4695
+    },
+    {
+      "epoch": 1.9792391189798715,
+      "grad_norm": 0.2995094060897827,
+      "learning_rate": 1.3296610489305005e-06,
+      "loss": 0.5778,
+      "step": 4696
+    },
+    {
+      "epoch": 1.9796606597112447,
+      "grad_norm": 0.3042725920677185,
+      "learning_rate": 1.328671966874055e-06,
+      "loss": 0.5766,
+      "step": 4697
+    },
+    {
+      "epoch": 1.9800822004426177,
+      "grad_norm": 0.28731945157051086,
+      "learning_rate": 1.327683119672234e-06,
+      "loss": 0.5527,
+      "step": 4698
+    },
+    {
+      "epoch": 1.980503741173991,
+      "grad_norm": 0.3179102838039398,
+      "learning_rate": 1.3266945075233033e-06,
+      "loss": 0.5779,
+      "step": 4699
+    },
+    {
+      "epoch": 1.980925281905364,
+      "grad_norm": 0.29832780361175537,
+      "learning_rate": 1.3257061306254827e-06,
+      "loss": 0.5925,
+      "step": 4700
+    },
+    {
+      "epoch": 1.9813468226367372,
+      "grad_norm": 0.29283973574638367,
+      "learning_rate": 1.3247179891769434e-06,
+      "loss": 0.5613,
+      "step": 4701
+    },
+    {
+      "epoch": 1.9817683633681105,
+      "grad_norm": 0.3045477569103241,
+      "learning_rate": 1.3237300833758115e-06,
+      "loss": 0.5922,
+      "step": 4702
+    },
+    {
+      "epoch": 1.9821899040994837,
+      "grad_norm": 0.3027907907962799,
+      "learning_rate": 1.322742413420166e-06,
+      "loss": 0.5833,
+      "step": 4703
+    },
+    {
+      "epoch": 1.982611444830857,
+      "grad_norm": 0.29397523403167725,
+      "learning_rate": 1.321754979508037e-06,
+      "loss": 0.5825,
+      "step": 4704
+    },
+    {
+      "epoch": 1.98303298556223,
+      "grad_norm": 0.2943994700908661,
+      "learning_rate": 1.320767781837407e-06,
+      "loss": 0.5863,
+      "step": 4705
+    },
+    {
+      "epoch": 1.983454526293603,
+      "grad_norm": 0.2917468845844269,
+      "learning_rate": 1.3197808206062115e-06,
+      "loss": 0.5835,
+      "step": 4706
+    },
+    {
+      "epoch": 1.9838760670249762,
+      "grad_norm": 0.29843994975090027,
+      "learning_rate": 1.318794096012341e-06,
+      "loss": 0.5793,
+      "step": 4707
+    },
+    {
+      "epoch": 1.9842976077563494,
+      "grad_norm": 0.30398741364479065,
+      "learning_rate": 1.3178076082536357e-06,
+      "loss": 0.5777,
+      "step": 4708
+    },
+    {
+      "epoch": 1.9847191484877227,
+      "grad_norm": 0.2920643985271454,
+      "learning_rate": 1.3168213575278888e-06,
+      "loss": 0.565,
+      "step": 4709
+    },
+    {
+      "epoch": 1.985140689219096,
+      "grad_norm": 0.31287071108818054,
+      "learning_rate": 1.3158353440328457e-06,
+      "loss": 0.5927,
+      "step": 4710
+    },
+    {
+      "epoch": 1.985562229950469,
+      "grad_norm": 0.29759126901626587,
+      "learning_rate": 1.314849567966207e-06,
+      "loss": 0.5844,
+      "step": 4711
+    },
+    {
+      "epoch": 1.9859837706818422,
+      "grad_norm": 0.3121117353439331,
+      "learning_rate": 1.313864029525621e-06,
+      "loss": 0.6064,
+      "step": 4712
+    },
+    {
+      "epoch": 1.9864053114132152,
+      "grad_norm": 0.30579298734664917,
+      "learning_rate": 1.312878728908694e-06,
+      "loss": 0.5822,
+      "step": 4713
+    },
+    {
+      "epoch": 1.9868268521445884,
+      "grad_norm": 0.2999986410140991,
+      "learning_rate": 1.31189366631298e-06,
+      "loss": 0.5795,
+      "step": 4714
+    },
+    {
+      "epoch": 1.9872483928759617,
+      "grad_norm": 0.2856597900390625,
+      "learning_rate": 1.310908841935986e-06,
+      "loss": 0.5946,
+      "step": 4715
+    },
+    {
+      "epoch": 1.987669933607335,
+      "grad_norm": 0.296792209148407,
+      "learning_rate": 1.309924255975174e-06,
+      "loss": 0.5707,
+      "step": 4716
+    },
+    {
+      "epoch": 1.9880914743387081,
+      "grad_norm": 0.30547645688056946,
+      "learning_rate": 1.3089399086279553e-06,
+      "loss": 0.5895,
+      "step": 4717
+    },
+    {
+      "epoch": 1.9885130150700812,
+      "grad_norm": 0.30077433586120605,
+      "learning_rate": 1.3079558000916943e-06,
+      "loss": 0.5855,
+      "step": 4718
+    },
+    {
+      "epoch": 1.9889345558014542,
+      "grad_norm": 0.31542882323265076,
+      "learning_rate": 1.306971930563707e-06,
+      "loss": 0.5801,
+      "step": 4719
+    },
+    {
+      "epoch": 1.9893560965328274,
+      "grad_norm": 0.2930469512939453,
+      "learning_rate": 1.3059883002412638e-06,
+      "loss": 0.5809,
+      "step": 4720
+    },
+    {
+      "epoch": 1.9897776372642006,
+      "grad_norm": 0.3118014633655548,
+      "learning_rate": 1.3050049093215827e-06,
+      "loss": 0.5958,
+      "step": 4721
+    },
+    {
+      "epoch": 1.9901991779955739,
+      "grad_norm": 0.29340726137161255,
+      "learning_rate": 1.30402175800184e-06,
+      "loss": 0.5726,
+      "step": 4722
+    },
+    {
+      "epoch": 1.9906207187269471,
+      "grad_norm": 0.29217982292175293,
+      "learning_rate": 1.3030388464791576e-06,
+      "loss": 0.5733,
+      "step": 4723
+    },
+    {
+      "epoch": 1.9910422594583201,
+      "grad_norm": 0.300712913274765,
+      "learning_rate": 1.3020561749506122e-06,
+      "loss": 0.5591,
+      "step": 4724
+    },
+    {
+      "epoch": 1.9914638001896934,
+      "grad_norm": 0.3036811947822571,
+      "learning_rate": 1.301073743613234e-06,
+      "loss": 0.5916,
+      "step": 4725
+    },
+    {
+      "epoch": 1.9918853409210664,
+      "grad_norm": 0.30233949422836304,
+      "learning_rate": 1.3000915526640022e-06,
+      "loss": 0.5761,
+      "step": 4726
+    },
+    {
+      "epoch": 1.9923068816524396,
+      "grad_norm": 0.29689234495162964,
+      "learning_rate": 1.2991096022998487e-06,
+      "loss": 0.5644,
+      "step": 4727
+    },
+    {
+      "epoch": 1.9927284223838129,
+      "grad_norm": 0.30044257640838623,
+      "learning_rate": 1.2981278927176566e-06,
+      "loss": 0.5964,
+      "step": 4728
+    },
+    {
+      "epoch": 1.993149963115186,
+      "grad_norm": 0.29428786039352417,
+      "learning_rate": 1.2971464241142636e-06,
+      "loss": 0.5598,
+      "step": 4729
+    },
+    {
+      "epoch": 1.9935715038465591,
+      "grad_norm": 0.3051070272922516,
+      "learning_rate": 1.296165196686455e-06,
+      "loss": 0.5866,
+      "step": 4730
+    },
+    {
+      "epoch": 1.9939930445779324,
+      "grad_norm": 0.29763248562812805,
+      "learning_rate": 1.2951842106309714e-06,
+      "loss": 0.5925,
+      "step": 4731
+    },
+    {
+      "epoch": 1.9944145853093054,
+      "grad_norm": 0.3007625341415405,
+      "learning_rate": 1.2942034661445025e-06,
+      "loss": 0.5652,
+      "step": 4732
+    },
+    {
+      "epoch": 1.9948361260406786,
+      "grad_norm": 0.29817333817481995,
+      "learning_rate": 1.293222963423689e-06,
+      "loss": 0.5717,
+      "step": 4733
+    },
+    {
+      "epoch": 1.9952576667720519,
+      "grad_norm": 0.31926408410072327,
+      "learning_rate": 1.292242702665127e-06,
+      "loss": 0.5887,
+      "step": 4734
+    },
+    {
+      "epoch": 1.995679207503425,
+      "grad_norm": 0.3112740218639374,
+      "learning_rate": 1.2912626840653602e-06,
+      "loss": 0.5784,
+      "step": 4735
+    },
+    {
+      "epoch": 1.9961007482347983,
+      "grad_norm": 0.30629029870033264,
+      "learning_rate": 1.290282907820886e-06,
+      "loss": 0.5974,
+      "step": 4736
+    },
+    {
+      "epoch": 1.9965222889661713,
+      "grad_norm": 0.3124987483024597,
+      "learning_rate": 1.2893033741281502e-06,
+      "loss": 0.542,
+      "step": 4737
+    },
+    {
+      "epoch": 1.9969438296975444,
+      "grad_norm": 0.29427480697631836,
+      "learning_rate": 1.2883240831835548e-06,
+      "loss": 0.5786,
+      "step": 4738
+    },
+    {
+      "epoch": 1.9973653704289176,
+      "grad_norm": 0.3125250041484833,
+      "learning_rate": 1.287345035183448e-06,
+      "loss": 0.5899,
+      "step": 4739
+    },
+    {
+      "epoch": 1.9977869111602908,
+      "grad_norm": 0.29705965518951416,
+      "learning_rate": 1.2863662303241343e-06,
+      "loss": 0.576,
+      "step": 4740
+    },
+    {
+      "epoch": 1.998208451891664,
+      "grad_norm": 0.3088743984699249,
+      "learning_rate": 1.285387668801865e-06,
+      "loss": 0.5726,
+      "step": 4741
+    },
+    {
+      "epoch": 1.9986299926230373,
+      "grad_norm": 0.29193341732025146,
+      "learning_rate": 1.2844093508128443e-06,
+      "loss": 0.6017,
+      "step": 4742
+    },
+    {
+      "epoch": 1.9990515333544103,
+      "grad_norm": 0.2962130010128021,
+      "learning_rate": 1.2834312765532296e-06,
+      "loss": 0.5556,
+      "step": 4743
+    },
+    {
+      "epoch": 1.9994730740857836,
+      "grad_norm": 0.304010272026062,
+      "learning_rate": 1.2824534462191256e-06,
+      "loss": 0.5859,
+      "step": 4744
+    }
+  ],
+  "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": 4.250073978620399e+19,
+  "train_batch_size": 12,
+  "trial_name": null,
+  "trial_params": null
+}