diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,30415 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4339,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00023046784973496196,
+      "grad_norm": 49.091495513916016,
+      "learning_rate": 7.633587786259542e-07,
+      "loss": 2.2176,
+      "step": 1
+    },
+    {
+      "epoch": 0.0004609356994699239,
+      "grad_norm": 46.45875930786133,
+      "learning_rate": 1.5267175572519084e-06,
+      "loss": 2.1992,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006914035492048859,
+      "grad_norm": 47.82147979736328,
+      "learning_rate": 2.2900763358778625e-06,
+      "loss": 2.2094,
+      "step": 3
+    },
+    {
+      "epoch": 0.0009218713989398478,
+      "grad_norm": 44.642494201660156,
+      "learning_rate": 3.053435114503817e-06,
+      "loss": 2.1912,
+      "step": 4
+    },
+    {
+      "epoch": 0.00115233924867481,
+      "grad_norm": 35.80204772949219,
+      "learning_rate": 3.816793893129772e-06,
+      "loss": 2.0999,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013828070984097719,
+      "grad_norm": 30.77144432067871,
+      "learning_rate": 4.580152671755725e-06,
+      "loss": 2.0554,
+      "step": 6
+    },
+    {
+      "epoch": 0.0016132749481447338,
+      "grad_norm": 16.0904483795166,
+      "learning_rate": 5.343511450381679e-06,
+      "loss": 1.9752,
+      "step": 7
+    },
+    {
+      "epoch": 0.0018437427978796957,
+      "grad_norm": 21.500225067138672,
+      "learning_rate": 6.106870229007634e-06,
+      "loss": 1.8799,
+      "step": 8
+    },
+    {
+      "epoch": 0.0020742106476146576,
+      "grad_norm": 25.848834991455078,
+      "learning_rate": 6.870229007633589e-06,
+      "loss": 2.0303,
+      "step": 9
+    },
+    {
+      "epoch": 0.00230467849734962,
+      "grad_norm": 21.812026977539062,
+      "learning_rate": 7.633587786259543e-06,
+      "loss": 1.9578,
+      "step": 10
+    },
+    {
+      "epoch": 0.002535146347084582,
+      "grad_norm": 29.04991912841797,
+      "learning_rate": 8.396946564885497e-06,
+      "loss": 1.7629,
+      "step": 11
+    },
+    {
+      "epoch": 0.0027656141968195437,
+      "grad_norm": 6.3423566818237305,
+      "learning_rate": 9.16030534351145e-06,
+      "loss": 1.6577,
+      "step": 12
+    },
+    {
+      "epoch": 0.0029960820465545056,
+      "grad_norm": 37.8600959777832,
+      "learning_rate": 9.923664122137405e-06,
+      "loss": 1.7216,
+      "step": 13
+    },
+    {
+      "epoch": 0.0032265498962894676,
+      "grad_norm": 13.13255500793457,
+      "learning_rate": 1.0687022900763359e-05,
+      "loss": 1.6066,
+      "step": 14
+    },
+    {
+      "epoch": 0.0034570177460244295,
+      "grad_norm": 5.612797260284424,
+      "learning_rate": 1.1450381679389314e-05,
+      "loss": 1.5293,
+      "step": 15
+    },
+    {
+      "epoch": 0.0036874855957593914,
+      "grad_norm": 4.661881923675537,
+      "learning_rate": 1.2213740458015267e-05,
+      "loss": 1.5133,
+      "step": 16
+    },
+    {
+      "epoch": 0.003917953445494353,
+      "grad_norm": 2.878075361251831,
+      "learning_rate": 1.2977099236641221e-05,
+      "loss": 1.4992,
+      "step": 17
+    },
+    {
+      "epoch": 0.004148421295229315,
+      "grad_norm": 2.4321236610412598,
+      "learning_rate": 1.3740458015267178e-05,
+      "loss": 1.4797,
+      "step": 18
+    },
+    {
+      "epoch": 0.004378889144964277,
+      "grad_norm": 2.1403255462646484,
+      "learning_rate": 1.450381679389313e-05,
+      "loss": 1.4561,
+      "step": 19
+    },
+    {
+      "epoch": 0.00460935699469924,
+      "grad_norm": 1.831888198852539,
+      "learning_rate": 1.5267175572519086e-05,
+      "loss": 1.4504,
+      "step": 20
+    },
+    {
+      "epoch": 0.004839824844434202,
+      "grad_norm": 2.384835958480835,
+      "learning_rate": 1.6030534351145038e-05,
+      "loss": 1.4364,
+      "step": 21
+    },
+    {
+      "epoch": 0.005070292694169164,
+      "grad_norm": 1.8255668878555298,
+      "learning_rate": 1.6793893129770993e-05,
+      "loss": 1.4241,
+      "step": 22
+    },
+    {
+      "epoch": 0.005300760543904126,
+      "grad_norm": 20.282546997070312,
+      "learning_rate": 1.7557251908396945e-05,
+      "loss": 1.4204,
+      "step": 23
+    },
+    {
+      "epoch": 0.0055312283936390875,
+      "grad_norm": 1.2769057750701904,
+      "learning_rate": 1.83206106870229e-05,
+      "loss": 1.4212,
+      "step": 24
+    },
+    {
+      "epoch": 0.005761696243374049,
+      "grad_norm": 1.2663878202438354,
+      "learning_rate": 1.9083969465648855e-05,
+      "loss": 1.3945,
+      "step": 25
+    },
+    {
+      "epoch": 0.005992164093109011,
+      "grad_norm": 1.3147037029266357,
+      "learning_rate": 1.984732824427481e-05,
+      "loss": 1.3976,
+      "step": 26
+    },
+    {
+      "epoch": 0.006222631942843973,
+      "grad_norm": 1.3301242589950562,
+      "learning_rate": 2.0610687022900766e-05,
+      "loss": 1.3872,
+      "step": 27
+    },
+    {
+      "epoch": 0.006453099792578935,
+      "grad_norm": 1.579620599746704,
+      "learning_rate": 2.1374045801526718e-05,
+      "loss": 1.3758,
+      "step": 28
+    },
+    {
+      "epoch": 0.006683567642313897,
+      "grad_norm": 1.2853105068206787,
+      "learning_rate": 2.2137404580152673e-05,
+      "loss": 1.3719,
+      "step": 29
+    },
+    {
+      "epoch": 0.006914035492048859,
+      "grad_norm": 1.264007329940796,
+      "learning_rate": 2.2900763358778628e-05,
+      "loss": 1.3545,
+      "step": 30
+    },
+    {
+      "epoch": 0.007144503341783821,
+      "grad_norm": 1.1497324705123901,
+      "learning_rate": 2.3664122137404583e-05,
+      "loss": 1.3492,
+      "step": 31
+    },
+    {
+      "epoch": 0.007374971191518783,
+      "grad_norm": 0.9906431436538696,
+      "learning_rate": 2.4427480916030535e-05,
+      "loss": 1.3537,
+      "step": 32
+    },
+    {
+      "epoch": 0.0076054390412537455,
+      "grad_norm": 0.7753263115882874,
+      "learning_rate": 2.5190839694656487e-05,
+      "loss": 1.3452,
+      "step": 33
+    },
+    {
+      "epoch": 0.007835906890988707,
+      "grad_norm": 0.7904003858566284,
+      "learning_rate": 2.5954198473282442e-05,
+      "loss": 1.3312,
+      "step": 34
+    },
+    {
+      "epoch": 0.00806637474072367,
+      "grad_norm": 0.826572597026825,
+      "learning_rate": 2.6717557251908397e-05,
+      "loss": 1.3327,
+      "step": 35
+    },
+    {
+      "epoch": 0.00829684259045863,
+      "grad_norm": 0.7653324007987976,
+      "learning_rate": 2.7480916030534355e-05,
+      "loss": 1.3333,
+      "step": 36
+    },
+    {
+      "epoch": 0.008527310440193593,
+      "grad_norm": 0.6942451000213623,
+      "learning_rate": 2.824427480916031e-05,
+      "loss": 1.3149,
+      "step": 37
+    },
+    {
+      "epoch": 0.008757778289928554,
+      "grad_norm": 0.6542969942092896,
+      "learning_rate": 2.900763358778626e-05,
+      "loss": 1.3153,
+      "step": 38
+    },
+    {
+      "epoch": 0.008988246139663517,
+      "grad_norm": 0.619317889213562,
+      "learning_rate": 2.9770992366412214e-05,
+      "loss": 1.3035,
+      "step": 39
+    },
+    {
+      "epoch": 0.00921871398939848,
+      "grad_norm": 0.5745211839675903,
+      "learning_rate": 3.053435114503817e-05,
+      "loss": 1.303,
+      "step": 40
+    },
+    {
+      "epoch": 0.00944918183913344,
+      "grad_norm": 0.5202656388282776,
+      "learning_rate": 3.129770992366413e-05,
+      "loss": 1.2884,
+      "step": 41
+    },
+    {
+      "epoch": 0.009679649688868404,
+      "grad_norm": 0.5099622011184692,
+      "learning_rate": 3.2061068702290076e-05,
+      "loss": 1.2864,
+      "step": 42
+    },
+    {
+      "epoch": 0.009910117538603365,
+      "grad_norm": 0.5038639307022095,
+      "learning_rate": 3.282442748091603e-05,
+      "loss": 1.2713,
+      "step": 43
+    },
+    {
+      "epoch": 0.010140585388338327,
+      "grad_norm": 0.5165843963623047,
+      "learning_rate": 3.358778625954199e-05,
+      "loss": 1.2843,
+      "step": 44
+    },
+    {
+      "epoch": 0.010371053238073288,
+      "grad_norm": 0.4482625424861908,
+      "learning_rate": 3.435114503816794e-05,
+      "loss": 1.2722,
+      "step": 45
+    },
+    {
+      "epoch": 0.010601521087808251,
+      "grad_norm": 0.4263209402561188,
+      "learning_rate": 3.511450381679389e-05,
+      "loss": 1.2686,
+      "step": 46
+    },
+    {
+      "epoch": 0.010831988937543212,
+      "grad_norm": 0.4388543367385864,
+      "learning_rate": 3.5877862595419845e-05,
+      "loss": 1.2519,
+      "step": 47
+    },
+    {
+      "epoch": 0.011062456787278175,
+      "grad_norm": 0.40465718507766724,
+      "learning_rate": 3.66412213740458e-05,
+      "loss": 1.2509,
+      "step": 48
+    },
+    {
+      "epoch": 0.011292924637013136,
+      "grad_norm": 0.4082145094871521,
+      "learning_rate": 3.7404580152671756e-05,
+      "loss": 1.2527,
+      "step": 49
+    },
+    {
+      "epoch": 0.011523392486748099,
+      "grad_norm": 0.40573209524154663,
+      "learning_rate": 3.816793893129771e-05,
+      "loss": 1.2557,
+      "step": 50
+    },
+    {
+      "epoch": 0.01175386033648306,
+      "grad_norm": 0.39484360814094543,
+      "learning_rate": 3.8931297709923666e-05,
+      "loss": 1.2488,
+      "step": 51
+    },
+    {
+      "epoch": 0.011984328186218023,
+      "grad_norm": 0.32370519638061523,
+      "learning_rate": 3.969465648854962e-05,
+      "loss": 1.2396,
+      "step": 52
+    },
+    {
+      "epoch": 0.012214796035952985,
+      "grad_norm": 0.36196169257164,
+      "learning_rate": 4.0458015267175576e-05,
+      "loss": 1.2454,
+      "step": 53
+    },
+    {
+      "epoch": 0.012445263885687946,
+      "grad_norm": 0.36752983927726746,
+      "learning_rate": 4.122137404580153e-05,
+      "loss": 1.2354,
+      "step": 54
+    },
+    {
+      "epoch": 0.01267573173542291,
+      "grad_norm": 0.312123566865921,
+      "learning_rate": 4.198473282442748e-05,
+      "loss": 1.2354,
+      "step": 55
+    },
+    {
+      "epoch": 0.01290619958515787,
+      "grad_norm": 0.30579739809036255,
+      "learning_rate": 4.2748091603053435e-05,
+      "loss": 1.2225,
+      "step": 56
+    },
+    {
+      "epoch": 0.013136667434892833,
+      "grad_norm": 0.3238474428653717,
+      "learning_rate": 4.351145038167939e-05,
+      "loss": 1.233,
+      "step": 57
+    },
+    {
+      "epoch": 0.013367135284627794,
+      "grad_norm": 0.32208338379859924,
+      "learning_rate": 4.4274809160305345e-05,
+      "loss": 1.2306,
+      "step": 58
+    },
+    {
+      "epoch": 0.013597603134362757,
+      "grad_norm": 0.3156214654445648,
+      "learning_rate": 4.5038167938931294e-05,
+      "loss": 1.22,
+      "step": 59
+    },
+    {
+      "epoch": 0.013828070984097718,
+      "grad_norm": 0.27707698941230774,
+      "learning_rate": 4.5801526717557256e-05,
+      "loss": 1.2219,
+      "step": 60
+    },
+    {
+      "epoch": 0.01405853883383268,
+      "grad_norm": 0.29033875465393066,
+      "learning_rate": 4.656488549618321e-05,
+      "loss": 1.214,
+      "step": 61
+    },
+    {
+      "epoch": 0.014289006683567642,
+      "grad_norm": 0.316847562789917,
+      "learning_rate": 4.7328244274809166e-05,
+      "loss": 1.2227,
+      "step": 62
+    },
+    {
+      "epoch": 0.014519474533302604,
+      "grad_norm": 0.2887651324272156,
+      "learning_rate": 4.809160305343512e-05,
+      "loss": 1.2157,
+      "step": 63
+    },
+    {
+      "epoch": 0.014749942383037565,
+      "grad_norm": 0.36876314878463745,
+      "learning_rate": 4.885496183206107e-05,
+      "loss": 1.1988,
+      "step": 64
+    },
+    {
+      "epoch": 0.014980410232772528,
+      "grad_norm": 0.2880174517631531,
+      "learning_rate": 4.9618320610687025e-05,
+      "loss": 1.2203,
+      "step": 65
+    },
+    {
+      "epoch": 0.015210878082507491,
+      "grad_norm": 0.2820625603199005,
+      "learning_rate": 5.038167938931297e-05,
+      "loss": 1.2109,
+      "step": 66
+    },
+    {
+      "epoch": 0.015441345932242452,
+      "grad_norm": 0.27386295795440674,
+      "learning_rate": 5.114503816793893e-05,
+      "loss": 1.2097,
+      "step": 67
+    },
+    {
+      "epoch": 0.015671813781977413,
+      "grad_norm": 0.26974838972091675,
+      "learning_rate": 5.1908396946564884e-05,
+      "loss": 1.205,
+      "step": 68
+    },
+    {
+      "epoch": 0.015902281631712378,
+      "grad_norm": 0.2931689918041229,
+      "learning_rate": 5.267175572519084e-05,
+      "loss": 1.2097,
+      "step": 69
+    },
+    {
+      "epoch": 0.01613274948144734,
+      "grad_norm": 0.2843622863292694,
+      "learning_rate": 5.3435114503816794e-05,
+      "loss": 1.1955,
+      "step": 70
+    },
+    {
+      "epoch": 0.0163632173311823,
+      "grad_norm": 0.28659501671791077,
+      "learning_rate": 5.419847328244275e-05,
+      "loss": 1.1836,
+      "step": 71
+    },
+    {
+      "epoch": 0.01659368518091726,
+      "grad_norm": 0.2758232355117798,
+      "learning_rate": 5.496183206106871e-05,
+      "loss": 1.2005,
+      "step": 72
+    },
+    {
+      "epoch": 0.016824153030652225,
+      "grad_norm": 0.37796103954315186,
+      "learning_rate": 5.5725190839694666e-05,
+      "loss": 1.1996,
+      "step": 73
+    },
+    {
+      "epoch": 0.017054620880387186,
+      "grad_norm": 0.6489807963371277,
+      "learning_rate": 5.648854961832062e-05,
+      "loss": 1.1854,
+      "step": 74
+    },
+    {
+      "epoch": 0.017285088730122147,
+      "grad_norm": 0.9581599831581116,
+      "learning_rate": 5.725190839694656e-05,
+      "loss": 1.1948,
+      "step": 75
+    },
+    {
+      "epoch": 0.01751555657985711,
+      "grad_norm": 0.6904142498970032,
+      "learning_rate": 5.801526717557252e-05,
+      "loss": 1.1835,
+      "step": 76
+    },
+    {
+      "epoch": 0.017746024429592073,
+      "grad_norm": 0.546022355556488,
+      "learning_rate": 5.877862595419847e-05,
+      "loss": 1.1789,
+      "step": 77
+    },
+    {
+      "epoch": 0.017976492279327034,
+      "grad_norm": 0.7557573914527893,
+      "learning_rate": 5.954198473282443e-05,
+      "loss": 1.1842,
+      "step": 78
+    },
+    {
+      "epoch": 0.018206960129061995,
+      "grad_norm": 0.40329957008361816,
+      "learning_rate": 6.0305343511450384e-05,
+      "loss": 1.179,
+      "step": 79
+    },
+    {
+      "epoch": 0.01843742797879696,
+      "grad_norm": 0.5815610885620117,
+      "learning_rate": 6.106870229007635e-05,
+      "loss": 1.173,
+      "step": 80
+    },
+    {
+      "epoch": 0.01866789582853192,
+      "grad_norm": 0.5922554135322571,
+      "learning_rate": 6.18320610687023e-05,
+      "loss": 1.1773,
+      "step": 81
+    },
+    {
+      "epoch": 0.01889836367826688,
+      "grad_norm": 0.347842276096344,
+      "learning_rate": 6.259541984732826e-05,
+      "loss": 1.1756,
+      "step": 82
+    },
+    {
+      "epoch": 0.019128831528001843,
+      "grad_norm": 0.8533177971839905,
+      "learning_rate": 6.33587786259542e-05,
+      "loss": 1.1734,
+      "step": 83
+    },
+    {
+      "epoch": 0.019359299377736807,
+      "grad_norm": 0.9222707152366638,
+      "learning_rate": 6.412213740458015e-05,
+      "loss": 1.175,
+      "step": 84
+    },
+    {
+      "epoch": 0.019589767227471768,
+      "grad_norm": 0.4554316997528076,
+      "learning_rate": 6.488549618320611e-05,
+      "loss": 1.1612,
+      "step": 85
+    },
+    {
+      "epoch": 0.01982023507720673,
+      "grad_norm": 0.6344988346099854,
+      "learning_rate": 6.564885496183206e-05,
+      "loss": 1.1587,
+      "step": 86
+    },
+    {
+      "epoch": 0.02005070292694169,
+      "grad_norm": 0.8068343997001648,
+      "learning_rate": 6.641221374045802e-05,
+      "loss": 1.1671,
+      "step": 87
+    },
+    {
+      "epoch": 0.020281170776676655,
+      "grad_norm": 0.4366961121559143,
+      "learning_rate": 6.717557251908397e-05,
+      "loss": 1.169,
+      "step": 88
+    },
+    {
+      "epoch": 0.020511638626411616,
+      "grad_norm": 0.666015088558197,
+      "learning_rate": 6.793893129770993e-05,
+      "loss": 1.1451,
+      "step": 89
+    },
+    {
+      "epoch": 0.020742106476146577,
+      "grad_norm": 0.6575168371200562,
+      "learning_rate": 6.870229007633588e-05,
+      "loss": 1.1572,
+      "step": 90
+    },
+    {
+      "epoch": 0.020972574325881538,
+      "grad_norm": 0.4764537513256073,
+      "learning_rate": 6.946564885496184e-05,
+      "loss": 1.1506,
+      "step": 91
+    },
+    {
+      "epoch": 0.021203042175616502,
+      "grad_norm": 0.6614679098129272,
+      "learning_rate": 7.022900763358778e-05,
+      "loss": 1.1677,
+      "step": 92
+    },
+    {
+      "epoch": 0.021433510025351463,
+      "grad_norm": 0.6387923955917358,
+      "learning_rate": 7.099236641221374e-05,
+      "loss": 1.1541,
+      "step": 93
+    },
+    {
+      "epoch": 0.021663977875086424,
+      "grad_norm": 0.5480379462242126,
+      "learning_rate": 7.175572519083969e-05,
+      "loss": 1.1419,
+      "step": 94
+    },
+    {
+      "epoch": 0.02189444572482139,
+      "grad_norm": 0.7831230759620667,
+      "learning_rate": 7.251908396946565e-05,
+      "loss": 1.1501,
+      "step": 95
+    },
+    {
+      "epoch": 0.02212491357455635,
+      "grad_norm": 0.6071659326553345,
+      "learning_rate": 7.32824427480916e-05,
+      "loss": 1.1424,
+      "step": 96
+    },
+    {
+      "epoch": 0.02235538142429131,
+      "grad_norm": 0.4921586513519287,
+      "learning_rate": 7.404580152671756e-05,
+      "loss": 1.1442,
+      "step": 97
+    },
+    {
+      "epoch": 0.022585849274026272,
+      "grad_norm": 0.9020814299583435,
+      "learning_rate": 7.480916030534351e-05,
+      "loss": 1.1541,
+      "step": 98
+    },
+    {
+      "epoch": 0.022816317123761237,
+      "grad_norm": 0.9073303937911987,
+      "learning_rate": 7.557251908396947e-05,
+      "loss": 1.154,
+      "step": 99
+    },
+    {
+      "epoch": 0.023046784973496198,
+      "grad_norm": 0.5485337376594543,
+      "learning_rate": 7.633587786259542e-05,
+      "loss": 1.1372,
+      "step": 100
+    },
+    {
+      "epoch": 0.02327725282323116,
+      "grad_norm": 1.202399492263794,
+      "learning_rate": 7.709923664122138e-05,
+      "loss": 1.1339,
+      "step": 101
+    },
+    {
+      "epoch": 0.02350772067296612,
+      "grad_norm": 1.6502957344055176,
+      "learning_rate": 7.786259541984733e-05,
+      "loss": 1.1466,
+      "step": 102
+    },
+    {
+      "epoch": 0.023738188522701084,
+      "grad_norm": 0.9507301449775696,
+      "learning_rate": 7.862595419847329e-05,
+      "loss": 1.1418,
+      "step": 103
+    },
+    {
+      "epoch": 0.023968656372436045,
+      "grad_norm": 1.0010653734207153,
+      "learning_rate": 7.938931297709924e-05,
+      "loss": 1.1503,
+      "step": 104
+    },
+    {
+      "epoch": 0.024199124222171006,
+      "grad_norm": 0.6749522089958191,
+      "learning_rate": 8.01526717557252e-05,
+      "loss": 1.1335,
+      "step": 105
+    },
+    {
+      "epoch": 0.02442959207190597,
+      "grad_norm": 0.7454925775527954,
+      "learning_rate": 8.091603053435115e-05,
+      "loss": 1.1353,
+      "step": 106
+    },
+    {
+      "epoch": 0.024660059921640932,
+      "grad_norm": 0.6050741672515869,
+      "learning_rate": 8.167938931297711e-05,
+      "loss": 1.1396,
+      "step": 107
+    },
+    {
+      "epoch": 0.024890527771375893,
+      "grad_norm": 0.8601208925247192,
+      "learning_rate": 8.244274809160306e-05,
+      "loss": 1.135,
+      "step": 108
+    },
+    {
+      "epoch": 0.025120995621110854,
+      "grad_norm": 0.6721903085708618,
+      "learning_rate": 8.320610687022902e-05,
+      "loss": 1.1259,
+      "step": 109
+    },
+    {
+      "epoch": 0.02535146347084582,
+      "grad_norm": 0.5092166662216187,
+      "learning_rate": 8.396946564885496e-05,
+      "loss": 1.1276,
+      "step": 110
+    },
+    {
+      "epoch": 0.02558193132058078,
+      "grad_norm": 0.5667235851287842,
+      "learning_rate": 8.473282442748092e-05,
+      "loss": 1.1262,
+      "step": 111
+    },
+    {
+      "epoch": 0.02581239917031574,
+      "grad_norm": 0.6341401934623718,
+      "learning_rate": 8.549618320610687e-05,
+      "loss": 1.1269,
+      "step": 112
+    },
+    {
+      "epoch": 0.0260428670200507,
+      "grad_norm": 0.5853081941604614,
+      "learning_rate": 8.625954198473283e-05,
+      "loss": 1.1321,
+      "step": 113
+    },
+    {
+      "epoch": 0.026273334869785666,
+      "grad_norm": 0.514251708984375,
+      "learning_rate": 8.702290076335878e-05,
+      "loss": 1.1322,
+      "step": 114
+    },
+    {
+      "epoch": 0.026503802719520627,
+      "grad_norm": 0.8147274255752563,
+      "learning_rate": 8.778625954198474e-05,
+      "loss": 1.1153,
+      "step": 115
+    },
+    {
+      "epoch": 0.026734270569255588,
+      "grad_norm": 1.1456818580627441,
+      "learning_rate": 8.854961832061069e-05,
+      "loss": 1.1361,
+      "step": 116
+    },
+    {
+      "epoch": 0.02696473841899055,
+      "grad_norm": 1.8368473052978516,
+      "learning_rate": 8.931297709923665e-05,
+      "loss": 1.1387,
+      "step": 117
+    },
+    {
+      "epoch": 0.027195206268725514,
+      "grad_norm": 2.4498660564422607,
+      "learning_rate": 9.007633587786259e-05,
+      "loss": 1.1505,
+      "step": 118
+    },
+    {
+      "epoch": 0.027425674118460475,
+      "grad_norm": 2.034926652908325,
+      "learning_rate": 9.083969465648856e-05,
+      "loss": 1.1443,
+      "step": 119
+    },
+    {
+      "epoch": 0.027656141968195436,
+      "grad_norm": 1.2848972082138062,
+      "learning_rate": 9.160305343511451e-05,
+      "loss": 1.1466,
+      "step": 120
+    },
+    {
+      "epoch": 0.0278866098179304,
+      "grad_norm": 1.265265703201294,
+      "learning_rate": 9.236641221374047e-05,
+      "loss": 1.1345,
+      "step": 121
+    },
+    {
+      "epoch": 0.02811707766766536,
+      "grad_norm": 0.8716845512390137,
+      "learning_rate": 9.312977099236642e-05,
+      "loss": 1.1412,
+      "step": 122
+    },
+    {
+      "epoch": 0.028347545517400322,
+      "grad_norm": 0.9397414922714233,
+      "learning_rate": 9.389312977099238e-05,
+      "loss": 1.1358,
+      "step": 123
+    },
+    {
+      "epoch": 0.028578013367135283,
+      "grad_norm": 0.9730684757232666,
+      "learning_rate": 9.465648854961833e-05,
+      "loss": 1.1232,
+      "step": 124
+    },
+    {
+      "epoch": 0.028808481216870248,
+      "grad_norm": 0.8047937154769897,
+      "learning_rate": 9.541984732824429e-05,
+      "loss": 1.1329,
+      "step": 125
+    },
+    {
+      "epoch": 0.02903894906660521,
+      "grad_norm": 0.7371867895126343,
+      "learning_rate": 9.618320610687024e-05,
+      "loss": 1.1292,
+      "step": 126
+    },
+    {
+      "epoch": 0.02926941691634017,
+      "grad_norm": 0.9497559666633606,
+      "learning_rate": 9.694656488549618e-05,
+      "loss": 1.1213,
+      "step": 127
+    },
+    {
+      "epoch": 0.02949988476607513,
+      "grad_norm": 0.6228593587875366,
+      "learning_rate": 9.770992366412214e-05,
+      "loss": 1.1112,
+      "step": 128
+    },
+    {
+      "epoch": 0.029730352615810095,
+      "grad_norm": 0.8247762322425842,
+      "learning_rate": 9.84732824427481e-05,
+      "loss": 1.1187,
+      "step": 129
+    },
+    {
+      "epoch": 0.029960820465545056,
+      "grad_norm": 0.6637692451477051,
+      "learning_rate": 9.923664122137405e-05,
+      "loss": 1.1046,
+      "step": 130
+    },
+    {
+      "epoch": 0.030191288315280018,
+      "grad_norm": 0.8043597936630249,
+      "learning_rate": 0.0001,
+      "loss": 1.1031,
+      "step": 131
+    },
+    {
+      "epoch": 0.030421756165014982,
+      "grad_norm": 0.6135521531105042,
+      "learning_rate": 9.999998606560007e-05,
+      "loss": 1.1086,
+      "step": 132
+    },
+    {
+      "epoch": 0.030652224014749943,
+      "grad_norm": 0.5676527619361877,
+      "learning_rate": 9.999994426240797e-05,
+      "loss": 1.0931,
+      "step": 133
+    },
+    {
+      "epoch": 0.030882691864484904,
+      "grad_norm": 0.6945146918296814,
+      "learning_rate": 9.999987459044706e-05,
+      "loss": 1.1038,
+      "step": 134
+    },
+    {
+      "epoch": 0.031113159714219865,
+      "grad_norm": 0.6486102938652039,
+      "learning_rate": 9.999977704975617e-05,
+      "loss": 1.1031,
+      "step": 135
+    },
+    {
+      "epoch": 0.031343627563954826,
+      "grad_norm": 0.7887089848518372,
+      "learning_rate": 9.999965164038963e-05,
+      "loss": 1.1089,
+      "step": 136
+    },
+    {
+      "epoch": 0.03157409541368979,
+      "grad_norm": 1.3936797380447388,
+      "learning_rate": 9.999949836241736e-05,
+      "loss": 1.1004,
+      "step": 137
+    },
+    {
+      "epoch": 0.031804563263424755,
+      "grad_norm": 2.2348039150238037,
+      "learning_rate": 9.999931721592481e-05,
+      "loss": 1.1181,
+      "step": 138
+    },
+    {
+      "epoch": 0.032035031113159716,
+      "grad_norm": 1.54178786277771,
+      "learning_rate": 9.999910820101293e-05,
+      "loss": 1.1246,
+      "step": 139
+    },
+    {
+      "epoch": 0.03226549896289468,
+      "grad_norm": 0.7922399640083313,
+      "learning_rate": 9.99988713177982e-05,
+      "loss": 1.1064,
+      "step": 140
+    },
+    {
+      "epoch": 0.03249596681262964,
+      "grad_norm": 0.8318815231323242,
+      "learning_rate": 9.999860656641268e-05,
+      "loss": 1.1061,
+      "step": 141
+    },
+    {
+      "epoch": 0.0327264346623646,
+      "grad_norm": 0.800605297088623,
+      "learning_rate": 9.999831394700395e-05,
+      "loss": 1.1053,
+      "step": 142
+    },
+    {
+      "epoch": 0.03295690251209956,
+      "grad_norm": 0.6274577379226685,
+      "learning_rate": 9.999799345973506e-05,
+      "loss": 1.0969,
+      "step": 143
+    },
+    {
+      "epoch": 0.03318737036183452,
+      "grad_norm": 0.7087227702140808,
+      "learning_rate": 9.99976451047847e-05,
+      "loss": 1.0991,
+      "step": 144
+    },
+    {
+      "epoch": 0.03341783821156949,
+      "grad_norm": 0.68040931224823,
+      "learning_rate": 9.999726888234698e-05,
+      "loss": 1.0971,
+      "step": 145
+    },
+    {
+      "epoch": 0.03364830606130445,
+      "grad_norm": 0.6233592629432678,
+      "learning_rate": 9.999686479263164e-05,
+      "loss": 1.0923,
+      "step": 146
+    },
+    {
+      "epoch": 0.03387877391103941,
+      "grad_norm": 0.8178934454917908,
+      "learning_rate": 9.999643283586388e-05,
+      "loss": 1.0971,
+      "step": 147
+    },
+    {
+      "epoch": 0.03410924176077437,
+      "grad_norm": 0.6832984089851379,
+      "learning_rate": 9.999597301228448e-05,
+      "loss": 1.0957,
+      "step": 148
+    },
+    {
+      "epoch": 0.034339709610509334,
+      "grad_norm": 0.49561241269111633,
+      "learning_rate": 9.999548532214973e-05,
+      "loss": 1.0898,
+      "step": 149
+    },
+    {
+      "epoch": 0.034570177460244295,
+      "grad_norm": 0.6173511743545532,
+      "learning_rate": 9.999496976573145e-05,
+      "loss": 1.0912,
+      "step": 150
+    },
+    {
+      "epoch": 0.034800645309979256,
+      "grad_norm": 0.5704337358474731,
+      "learning_rate": 9.999442634331703e-05,
+      "loss": 1.0805,
+      "step": 151
+    },
+    {
+      "epoch": 0.03503111315971422,
+      "grad_norm": 0.4808043837547302,
+      "learning_rate": 9.999385505520931e-05,
+      "loss": 1.0877,
+      "step": 152
+    },
+    {
+      "epoch": 0.035261581009449185,
+      "grad_norm": 0.5703374743461609,
+      "learning_rate": 9.999325590172675e-05,
+      "loss": 1.0754,
+      "step": 153
+    },
+    {
+      "epoch": 0.035492048859184146,
+      "grad_norm": 0.7379648089408875,
+      "learning_rate": 9.999262888320329e-05,
+      "loss": 1.0987,
+      "step": 154
+    },
+    {
+      "epoch": 0.03572251670891911,
+      "grad_norm": 0.7264831066131592,
+      "learning_rate": 9.999197399998841e-05,
+      "loss": 1.0787,
+      "step": 155
+    },
+    {
+      "epoch": 0.03595298455865407,
+      "grad_norm": 0.5585530400276184,
+      "learning_rate": 9.999129125244714e-05,
+      "loss": 1.0797,
+      "step": 156
+    },
+    {
+      "epoch": 0.03618345240838903,
+      "grad_norm": 0.8825198411941528,
+      "learning_rate": 9.999058064096002e-05,
+      "loss": 1.0704,
+      "step": 157
+    },
+    {
+      "epoch": 0.03641392025812399,
+      "grad_norm": 1.1649545431137085,
+      "learning_rate": 9.998984216592313e-05,
+      "loss": 1.0844,
+      "step": 158
+    },
+    {
+      "epoch": 0.03664438810785895,
+      "grad_norm": 1.1949596405029297,
+      "learning_rate": 9.998907582774807e-05,
+      "loss": 1.071,
+      "step": 159
+    },
+    {
+      "epoch": 0.03687485595759392,
+      "grad_norm": 1.635244607925415,
+      "learning_rate": 9.998828162686197e-05,
+      "loss": 1.0777,
+      "step": 160
+    },
+    {
+      "epoch": 0.03710532380732888,
+      "grad_norm": 1.7061165571212769,
+      "learning_rate": 9.998745956370754e-05,
+      "loss": 1.0744,
+      "step": 161
+    },
+    {
+      "epoch": 0.03733579165706384,
+      "grad_norm": 1.4206252098083496,
+      "learning_rate": 9.998660963874294e-05,
+      "loss": 1.0777,
+      "step": 162
+    },
+    {
+      "epoch": 0.0375662595067988,
+      "grad_norm": 1.5643310546875,
+      "learning_rate": 9.998573185244192e-05,
+      "loss": 1.0783,
+      "step": 163
+    },
+    {
+      "epoch": 0.03779672735653376,
+      "grad_norm": 1.2039772272109985,
+      "learning_rate": 9.998482620529371e-05,
+      "loss": 1.0766,
+      "step": 164
+    },
+    {
+      "epoch": 0.038027195206268724,
+      "grad_norm": 0.8104932904243469,
+      "learning_rate": 9.998389269780312e-05,
+      "loss": 1.0693,
+      "step": 165
+    },
+    {
+      "epoch": 0.038257663056003685,
+      "grad_norm": 0.8776142597198486,
+      "learning_rate": 9.998293133049046e-05,
+      "loss": 1.0668,
+      "step": 166
+    },
+    {
+      "epoch": 0.038488130905738646,
+      "grad_norm": 0.837092936038971,
+      "learning_rate": 9.998194210389157e-05,
+      "loss": 1.061,
+      "step": 167
+    },
+    {
+      "epoch": 0.038718598755473614,
+      "grad_norm": 0.8989808559417725,
+      "learning_rate": 9.998092501855782e-05,
+      "loss": 1.0699,
+      "step": 168
+    },
+    {
+      "epoch": 0.038949066605208575,
+      "grad_norm": 0.97697913646698,
+      "learning_rate": 9.99798800750561e-05,
+      "loss": 1.0568,
+      "step": 169
+    },
+    {
+      "epoch": 0.039179534454943536,
+      "grad_norm": 0.8058460354804993,
+      "learning_rate": 9.997880727396886e-05,
+      "loss": 1.0501,
+      "step": 170
+    },
+    {
+      "epoch": 0.0394100023046785,
+      "grad_norm": 0.8557106852531433,
+      "learning_rate": 9.997770661589403e-05,
+      "loss": 1.0567,
+      "step": 171
+    },
+    {
+      "epoch": 0.03964047015441346,
+      "grad_norm": 0.794967532157898,
+      "learning_rate": 9.997657810144511e-05,
+      "loss": 1.0555,
+      "step": 172
+    },
+    {
+      "epoch": 0.03987093800414842,
+      "grad_norm": 0.7823812961578369,
+      "learning_rate": 9.99754217312511e-05,
+      "loss": 1.0373,
+      "step": 173
+    },
+    {
+      "epoch": 0.04010140585388338,
+      "grad_norm": 1.165642261505127,
+      "learning_rate": 9.997423750595651e-05,
+      "loss": 1.0486,
+      "step": 174
+    },
+    {
+      "epoch": 0.04033187370361835,
+      "grad_norm": 1.893602967262268,
+      "learning_rate": 9.997302542622144e-05,
+      "loss": 1.0738,
+      "step": 175
+    },
+    {
+      "epoch": 0.04056234155335331,
+      "grad_norm": 1.9267841577529907,
+      "learning_rate": 9.997178549272145e-05,
+      "loss": 1.0595,
+      "step": 176
+    },
+    {
+      "epoch": 0.04079280940308827,
+      "grad_norm": 1.4017184972763062,
+      "learning_rate": 9.997051770614765e-05,
+      "loss": 1.0589,
+      "step": 177
+    },
+    {
+      "epoch": 0.04102327725282323,
+      "grad_norm": 0.9901034235954285,
+      "learning_rate": 9.996922206720667e-05,
+      "loss": 1.0584,
+      "step": 178
+    },
+    {
+      "epoch": 0.04125374510255819,
+      "grad_norm": 0.9640342593193054,
+      "learning_rate": 9.996789857662068e-05,
+      "loss": 1.0513,
+      "step": 179
+    },
+    {
+      "epoch": 0.041484212952293154,
+      "grad_norm": 0.9942569136619568,
+      "learning_rate": 9.996654723512736e-05,
+      "loss": 1.0478,
+      "step": 180
+    },
+    {
+      "epoch": 0.041714680802028115,
+      "grad_norm": 0.9683095812797546,
+      "learning_rate": 9.996516804347991e-05,
+      "loss": 1.0514,
+      "step": 181
+    },
+    {
+      "epoch": 0.041945148651763076,
+      "grad_norm": 1.039068341255188,
+      "learning_rate": 9.996376100244704e-05,
+      "loss": 1.0368,
+      "step": 182
+    },
+    {
+      "epoch": 0.042175616501498044,
+      "grad_norm": 0.8390737175941467,
+      "learning_rate": 9.996232611281304e-05,
+      "loss": 1.033,
+      "step": 183
+    },
+    {
+      "epoch": 0.042406084351233005,
+      "grad_norm": 0.7644326090812683,
+      "learning_rate": 9.996086337537767e-05,
+      "loss": 1.0359,
+      "step": 184
+    },
+    {
+      "epoch": 0.042636552200967966,
+      "grad_norm": 0.7521729469299316,
+      "learning_rate": 9.995937279095621e-05,
+      "loss": 1.0259,
+      "step": 185
+    },
+    {
+      "epoch": 0.04286702005070293,
+      "grad_norm": 0.7316980361938477,
+      "learning_rate": 9.995785436037947e-05,
+      "loss": 1.0388,
+      "step": 186
+    },
+    {
+      "epoch": 0.04309748790043789,
+      "grad_norm": 0.7523071765899658,
+      "learning_rate": 9.995630808449383e-05,
+      "loss": 1.0286,
+      "step": 187
+    },
+    {
+      "epoch": 0.04332795575017285,
+      "grad_norm": 0.8001862168312073,
+      "learning_rate": 9.995473396416111e-05,
+      "loss": 1.0319,
+      "step": 188
+    },
+    {
+      "epoch": 0.04355842359990781,
+      "grad_norm": 0.859374463558197,
+      "learning_rate": 9.995313200025869e-05,
+      "loss": 1.0244,
+      "step": 189
+    },
+    {
+      "epoch": 0.04378889144964278,
+      "grad_norm": 0.9133301973342896,
+      "learning_rate": 9.995150219367946e-05,
+      "loss": 1.0106,
+      "step": 190
+    },
+    {
+      "epoch": 0.04401935929937774,
+      "grad_norm": 0.8615682125091553,
+      "learning_rate": 9.994984454533185e-05,
+      "loss": 1.0147,
+      "step": 191
+    },
+    {
+      "epoch": 0.0442498271491127,
+      "grad_norm": 0.7992554903030396,
+      "learning_rate": 9.994815905613981e-05,
+      "loss": 1.0163,
+      "step": 192
+    },
+    {
+      "epoch": 0.04448029499884766,
+      "grad_norm": 0.8807731866836548,
+      "learning_rate": 9.994644572704275e-05,
+      "loss": 1.0208,
+      "step": 193
+    },
+    {
+      "epoch": 0.04471076284858262,
+      "grad_norm": 1.1343860626220703,
+      "learning_rate": 9.994470455899568e-05,
+      "loss": 1.0152,
+      "step": 194
+    },
+    {
+      "epoch": 0.04494123069831758,
+      "grad_norm": 1.4376226663589478,
+      "learning_rate": 9.994293555296904e-05,
+      "loss": 1.0261,
+      "step": 195
+    },
+    {
+      "epoch": 0.045171698548052544,
+      "grad_norm": 1.5723721981048584,
+      "learning_rate": 9.994113870994888e-05,
+      "loss": 1.0137,
+      "step": 196
+    },
+    {
+      "epoch": 0.04540216639778751,
+      "grad_norm": 1.2862370014190674,
+      "learning_rate": 9.993931403093668e-05,
+      "loss": 1.0159,
+      "step": 197
+    },
+    {
+      "epoch": 0.04563263424752247,
+      "grad_norm": 1.0813417434692383,
+      "learning_rate": 9.99374615169495e-05,
+      "loss": 0.9976,
+      "step": 198
+    },
+    {
+      "epoch": 0.045863102097257434,
+      "grad_norm": 0.7794204354286194,
+      "learning_rate": 9.993558116901985e-05,
+      "loss": 1.0036,
+      "step": 199
+    },
+    {
+      "epoch": 0.046093569946992395,
+      "grad_norm": 1.225102186203003,
+      "learning_rate": 9.993367298819583e-05,
+      "loss": 0.9909,
+      "step": 200
+    },
+    {
+      "epoch": 0.046324037796727356,
+      "grad_norm": 1.1289259195327759,
+      "learning_rate": 9.9931736975541e-05,
+      "loss": 1.0123,
+      "step": 201
+    },
+    {
+      "epoch": 0.04655450564646232,
+      "grad_norm": 0.755088746547699,
+      "learning_rate": 9.992977313213443e-05,
+      "loss": 0.9916,
+      "step": 202
+    },
+    {
+      "epoch": 0.04678497349619728,
+      "grad_norm": 0.8576580286026001,
+      "learning_rate": 9.992778145907073e-05,
+      "loss": 0.9914,
+      "step": 203
+    },
+    {
+      "epoch": 0.04701544134593224,
+      "grad_norm": 0.8619646430015564,
+      "learning_rate": 9.992576195746003e-05,
+      "loss": 0.9963,
+      "step": 204
+    },
+    {
+      "epoch": 0.04724590919566721,
+      "grad_norm": 0.8409770727157593,
+      "learning_rate": 9.992371462842794e-05,
+      "loss": 0.9923,
+      "step": 205
+    },
+    {
+      "epoch": 0.04747637704540217,
+      "grad_norm": 0.7922968864440918,
+      "learning_rate": 9.992163947311557e-05,
+      "loss": 0.9859,
+      "step": 206
+    },
+    {
+      "epoch": 0.04770684489513713,
+      "grad_norm": 0.7922948002815247,
+      "learning_rate": 9.99195364926796e-05,
+      "loss": 0.9798,
+      "step": 207
+    },
+    {
+      "epoch": 0.04793731274487209,
+      "grad_norm": 0.8570684790611267,
+      "learning_rate": 9.991740568829215e-05,
+      "loss": 0.9644,
+      "step": 208
+    },
+    {
+      "epoch": 0.04816778059460705,
+      "grad_norm": 0.9860967993736267,
+      "learning_rate": 9.99152470611409e-05,
+      "loss": 0.976,
+      "step": 209
+    },
+    {
+      "epoch": 0.04839824844434201,
+      "grad_norm": 1.1492034196853638,
+      "learning_rate": 9.991306061242899e-05,
+      "loss": 0.9746,
+      "step": 210
+    },
+    {
+      "epoch": 0.048628716294076973,
+      "grad_norm": 1.3347015380859375,
+      "learning_rate": 9.991084634337511e-05,
+      "loss": 0.973,
+      "step": 211
+    },
+    {
+      "epoch": 0.04885918414381194,
+      "grad_norm": 1.097365140914917,
+      "learning_rate": 9.990860425521347e-05,
+      "loss": 0.9814,
+      "step": 212
+    },
+    {
+      "epoch": 0.0490896519935469,
+      "grad_norm": 1.006355881690979,
+      "learning_rate": 9.990633434919369e-05,
+      "loss": 0.9726,
+      "step": 213
+    },
+    {
+      "epoch": 0.049320119843281864,
+      "grad_norm": 0.8860179781913757,
+      "learning_rate": 9.990403662658104e-05,
+      "loss": 0.973,
+      "step": 214
+    },
+    {
+      "epoch": 0.049550587693016825,
+      "grad_norm": 0.8287469148635864,
+      "learning_rate": 9.990171108865614e-05,
+      "loss": 0.9673,
+      "step": 215
+    },
+    {
+      "epoch": 0.049781055542751786,
+      "grad_norm": 0.8757283687591553,
+      "learning_rate": 9.989935773671525e-05,
+      "loss": 0.9571,
+      "step": 216
+    },
+    {
+      "epoch": 0.05001152339248675,
+      "grad_norm": 0.8432666063308716,
+      "learning_rate": 9.989697657207002e-05,
+      "loss": 0.9543,
+      "step": 217
+    },
+    {
+      "epoch": 0.05024199124222171,
+      "grad_norm": 0.8379467725753784,
+      "learning_rate": 9.98945675960477e-05,
+      "loss": 0.9643,
+      "step": 218
+    },
+    {
+      "epoch": 0.05047245909195667,
+      "grad_norm": 0.7548062801361084,
+      "learning_rate": 9.989213080999097e-05,
+      "loss": 0.9637,
+      "step": 219
+    },
+    {
+      "epoch": 0.05070292694169164,
+      "grad_norm": 0.6869426965713501,
+      "learning_rate": 9.988966621525804e-05,
+      "loss": 0.9411,
+      "step": 220
+    },
+    {
+      "epoch": 0.0509333947914266,
+      "grad_norm": 0.8082894086837769,
+      "learning_rate": 9.988717381322262e-05,
+      "loss": 0.9441,
+      "step": 221
+    },
+    {
+      "epoch": 0.05116386264116156,
+      "grad_norm": 0.8997180461883545,
+      "learning_rate": 9.988465360527389e-05,
+      "loss": 0.9436,
+      "step": 222
+    },
+    {
+      "epoch": 0.05139433049089652,
+      "grad_norm": 0.9218918681144714,
+      "learning_rate": 9.988210559281658e-05,
+      "loss": 0.9616,
+      "step": 223
+    },
+    {
+      "epoch": 0.05162479834063148,
+      "grad_norm": 0.8721861839294434,
+      "learning_rate": 9.98795297772709e-05,
+      "loss": 0.9312,
+      "step": 224
+    },
+    {
+      "epoch": 0.05185526619036644,
+      "grad_norm": 0.8172184824943542,
+      "learning_rate": 9.987692616007253e-05,
+      "loss": 0.945,
+      "step": 225
+    },
+    {
+      "epoch": 0.0520857340401014,
+      "grad_norm": 0.8974547386169434,
+      "learning_rate": 9.987429474267268e-05,
+      "loss": 0.9419,
+      "step": 226
+    },
+    {
+      "epoch": 0.05231620188983637,
+      "grad_norm": 1.0265862941741943,
+      "learning_rate": 9.987163552653802e-05,
+      "loss": 0.9443,
+      "step": 227
+    },
+    {
+      "epoch": 0.05254666973957133,
+      "grad_norm": 1.184733510017395,
+      "learning_rate": 9.986894851315074e-05,
+      "loss": 0.9374,
+      "step": 228
+    },
+    {
+      "epoch": 0.05277713758930629,
+      "grad_norm": 1.3877384662628174,
+      "learning_rate": 9.98662337040085e-05,
+      "loss": 0.9444,
+      "step": 229
+    },
+    {
+      "epoch": 0.053007605439041254,
+      "grad_norm": 1.0500129461288452,
+      "learning_rate": 9.98634911006245e-05,
+      "loss": 0.9305,
+      "step": 230
+    },
+    {
+      "epoch": 0.053238073288776215,
+      "grad_norm": 0.6418678164482117,
+      "learning_rate": 9.986072070452738e-05,
+      "loss": 0.9261,
+      "step": 231
+    },
+    {
+      "epoch": 0.053468541138511176,
+      "grad_norm": 0.9703414440155029,
+      "learning_rate": 9.985792251726131e-05,
+      "loss": 0.9323,
+      "step": 232
+    },
+    {
+      "epoch": 0.05369900898824614,
+      "grad_norm": 0.8372150659561157,
+      "learning_rate": 9.985509654038591e-05,
+      "loss": 0.9335,
+      "step": 233
+    },
+    {
+      "epoch": 0.0539294768379811,
+      "grad_norm": 0.7052909135818481,
+      "learning_rate": 9.985224277547634e-05,
+      "loss": 0.9088,
+      "step": 234
+    },
+    {
+      "epoch": 0.054159944687716066,
+      "grad_norm": 0.7531741261482239,
+      "learning_rate": 9.984936122412319e-05,
+      "loss": 0.9182,
+      "step": 235
+    },
+    {
+      "epoch": 0.05439041253745103,
+      "grad_norm": 0.633127748966217,
+      "learning_rate": 9.98464518879326e-05,
+      "loss": 0.923,
+      "step": 236
+    },
+    {
+      "epoch": 0.05462088038718599,
+      "grad_norm": 0.5811619758605957,
+      "learning_rate": 9.984351476852613e-05,
+      "loss": 0.9184,
+      "step": 237
+    },
+    {
+      "epoch": 0.05485134823692095,
+      "grad_norm": 0.6498633027076721,
+      "learning_rate": 9.984054986754088e-05,
+      "loss": 0.919,
+      "step": 238
+    },
+    {
+      "epoch": 0.05508181608665591,
+      "grad_norm": 0.5947354435920715,
+      "learning_rate": 9.98375571866294e-05,
+      "loss": 0.9233,
+      "step": 239
+    },
+    {
+      "epoch": 0.05531228393639087,
+      "grad_norm": 0.5955505967140198,
+      "learning_rate": 9.983453672745975e-05,
+      "loss": 0.9212,
+      "step": 240
+    },
+    {
+      "epoch": 0.05554275178612583,
+      "grad_norm": 0.571624755859375,
+      "learning_rate": 9.983148849171546e-05,
+      "loss": 0.9166,
+      "step": 241
+    },
+    {
+      "epoch": 0.0557732196358608,
+      "grad_norm": 0.6101223230361938,
+      "learning_rate": 9.982841248109555e-05,
+      "loss": 0.912,
+      "step": 242
+    },
+    {
+      "epoch": 0.05600368748559576,
+      "grad_norm": 0.6820085644721985,
+      "learning_rate": 9.982530869731451e-05,
+      "loss": 0.9112,
+      "step": 243
+    },
+    {
+      "epoch": 0.05623415533533072,
+      "grad_norm": 0.9035933613777161,
+      "learning_rate": 9.982217714210232e-05,
+      "loss": 0.9129,
+      "step": 244
+    },
+    {
+      "epoch": 0.056464623185065684,
+      "grad_norm": 1.0833593606948853,
+      "learning_rate": 9.98190178172044e-05,
+      "loss": 0.9062,
+      "step": 245
+    },
+    {
+      "epoch": 0.056695091034800645,
+      "grad_norm": 1.1479918956756592,
+      "learning_rate": 9.981583072438173e-05,
+      "loss": 0.9166,
+      "step": 246
+    },
+    {
+      "epoch": 0.056925558884535606,
+      "grad_norm": 0.9499214291572571,
+      "learning_rate": 9.981261586541068e-05,
+      "loss": 0.9116,
+      "step": 247
+    },
+    {
+      "epoch": 0.05715602673427057,
+      "grad_norm": 0.8736324906349182,
+      "learning_rate": 9.980937324208317e-05,
+      "loss": 0.9076,
+      "step": 248
+    },
+    {
+      "epoch": 0.057386494584005535,
+      "grad_norm": 0.8064135313034058,
+      "learning_rate": 9.980610285620654e-05,
+      "loss": 0.9052,
+      "step": 249
+    },
+    {
+      "epoch": 0.057616962433740496,
+      "grad_norm": 0.7814821600914001,
+      "learning_rate": 9.980280470960363e-05,
+      "loss": 0.8956,
+      "step": 250
+    },
+    {
+      "epoch": 0.05784743028347546,
+      "grad_norm": 0.7428434491157532,
+      "learning_rate": 9.979947880411273e-05,
+      "loss": 0.8941,
+      "step": 251
+    },
+    {
+      "epoch": 0.05807789813321042,
+      "grad_norm": 0.6821320652961731,
+      "learning_rate": 9.979612514158765e-05,
+      "loss": 0.8985,
+      "step": 252
+    },
+    {
+      "epoch": 0.05830836598294538,
+      "grad_norm": 0.6651344895362854,
+      "learning_rate": 9.979274372389762e-05,
+      "loss": 0.8986,
+      "step": 253
+    },
+    {
+      "epoch": 0.05853883383268034,
+      "grad_norm": 0.7630136013031006,
+      "learning_rate": 9.978933455292736e-05,
+      "loss": 0.8916,
+      "step": 254
+    },
+    {
+      "epoch": 0.0587693016824153,
+      "grad_norm": 0.801031768321991,
+      "learning_rate": 9.978589763057708e-05,
+      "loss": 0.8947,
+      "step": 255
+    },
+    {
+      "epoch": 0.05899976953215026,
+      "grad_norm": 0.7151498794555664,
+      "learning_rate": 9.978243295876242e-05,
+      "loss": 0.8945,
+      "step": 256
+    },
+    {
+      "epoch": 0.05923023738188523,
+      "grad_norm": 0.7015817761421204,
+      "learning_rate": 9.97789405394145e-05,
+      "loss": 0.8903,
+      "step": 257
+    },
+    {
+      "epoch": 0.05946070523162019,
+      "grad_norm": 0.7685847282409668,
+      "learning_rate": 9.977542037447994e-05,
+      "loss": 0.8959,
+      "step": 258
+    },
+    {
+      "epoch": 0.05969117308135515,
+      "grad_norm": 0.8099703192710876,
+      "learning_rate": 9.977187246592076e-05,
+      "loss": 0.8984,
+      "step": 259
+    },
+    {
+      "epoch": 0.05992164093109011,
+      "grad_norm": 0.7858747243881226,
+      "learning_rate": 9.97682968157145e-05,
+      "loss": 0.8845,
+      "step": 260
+    },
+    {
+      "epoch": 0.060152108780825074,
+      "grad_norm": 0.8646757006645203,
+      "learning_rate": 9.976469342585413e-05,
+      "loss": 0.8776,
+      "step": 261
+    },
+    {
+      "epoch": 0.060382576630560035,
+      "grad_norm": 0.8440235257148743,
+      "learning_rate": 9.976106229834812e-05,
+      "loss": 0.8711,
+      "step": 262
+    },
+    {
+      "epoch": 0.060613044480294996,
+      "grad_norm": 0.8207287192344666,
+      "learning_rate": 9.975740343522033e-05,
+      "loss": 0.8806,
+      "step": 263
+    },
+    {
+      "epoch": 0.060843512330029964,
+      "grad_norm": 0.8601641058921814,
+      "learning_rate": 9.975371683851016e-05,
+      "loss": 0.8794,
+      "step": 264
+    },
+    {
+      "epoch": 0.061073980179764925,
+      "grad_norm": 0.8005819916725159,
+      "learning_rate": 9.975000251027242e-05,
+      "loss": 0.889,
+      "step": 265
+    },
+    {
+      "epoch": 0.061304448029499886,
+      "grad_norm": 0.7524363994598389,
+      "learning_rate": 9.974626045257738e-05,
+      "loss": 0.8734,
+      "step": 266
+    },
+    {
+      "epoch": 0.06153491587923485,
+      "grad_norm": 0.7972283959388733,
+      "learning_rate": 9.974249066751077e-05,
+      "loss": 0.8856,
+      "step": 267
+    },
+    {
+      "epoch": 0.06176538372896981,
+      "grad_norm": 0.925719678401947,
+      "learning_rate": 9.973869315717379e-05,
+      "loss": 0.8745,
+      "step": 268
+    },
+    {
+      "epoch": 0.06199585157870477,
+      "grad_norm": 0.8908208608627319,
+      "learning_rate": 9.973486792368307e-05,
+      "loss": 0.8676,
+      "step": 269
+    },
+    {
+      "epoch": 0.06222631942843973,
+      "grad_norm": 0.832204282283783,
+      "learning_rate": 9.973101496917072e-05,
+      "loss": 0.874,
+      "step": 270
+    },
+    {
+      "epoch": 0.06245678727817469,
+      "grad_norm": 0.8755982518196106,
+      "learning_rate": 9.972713429578427e-05,
+      "loss": 0.8564,
+      "step": 271
+    },
+    {
+      "epoch": 0.06268725512790965,
+      "grad_norm": 0.8395307660102844,
+      "learning_rate": 9.97232259056867e-05,
+      "loss": 0.8624,
+      "step": 272
+    },
+    {
+      "epoch": 0.06291772297764461,
+      "grad_norm": 0.7844563722610474,
+      "learning_rate": 9.97192898010565e-05,
+      "loss": 0.8637,
+      "step": 273
+    },
+    {
+      "epoch": 0.06314819082737957,
+      "grad_norm": 0.6658696532249451,
+      "learning_rate": 9.97153259840875e-05,
+      "loss": 0.8695,
+      "step": 274
+    },
+    {
+      "epoch": 0.06337865867711455,
+      "grad_norm": 0.7139994502067566,
+      "learning_rate": 9.971133445698908e-05,
+      "loss": 0.8486,
+      "step": 275
+    },
+    {
+      "epoch": 0.06360912652684951,
+      "grad_norm": 0.6378993391990662,
+      "learning_rate": 9.970731522198602e-05,
+      "loss": 0.8639,
+      "step": 276
+    },
+    {
+      "epoch": 0.06383959437658447,
+      "grad_norm": 0.6299728155136108,
+      "learning_rate": 9.970326828131852e-05,
+      "loss": 0.8541,
+      "step": 277
+    },
+    {
+      "epoch": 0.06407006222631943,
+      "grad_norm": 0.619757890701294,
+      "learning_rate": 9.969919363724226e-05,
+      "loss": 0.8506,
+      "step": 278
+    },
+    {
+      "epoch": 0.0643005300760544,
+      "grad_norm": 0.575935423374176,
+      "learning_rate": 9.969509129202837e-05,
+      "loss": 0.8599,
+      "step": 279
+    },
+    {
+      "epoch": 0.06453099792578935,
+      "grad_norm": 0.5778565406799316,
+      "learning_rate": 9.969096124796335e-05,
+      "loss": 0.8495,
+      "step": 280
+    },
+    {
+      "epoch": 0.06476146577552432,
+      "grad_norm": 0.6680006980895996,
+      "learning_rate": 9.968680350734922e-05,
+      "loss": 0.8468,
+      "step": 281
+    },
+    {
+      "epoch": 0.06499193362525928,
+      "grad_norm": 0.5861308574676514,
+      "learning_rate": 9.968261807250341e-05,
+      "loss": 0.8444,
+      "step": 282
+    },
+    {
+      "epoch": 0.06522240147499424,
+      "grad_norm": 0.5511894226074219,
+      "learning_rate": 9.967840494575879e-05,
+      "loss": 0.852,
+      "step": 283
+    },
+    {
+      "epoch": 0.0654528693247292,
+      "grad_norm": 0.6007016897201538,
+      "learning_rate": 9.967416412946362e-05,
+      "loss": 0.8503,
+      "step": 284
+    },
+    {
+      "epoch": 0.06568333717446416,
+      "grad_norm": 0.5989518761634827,
+      "learning_rate": 9.966989562598163e-05,
+      "loss": 0.8463,
+      "step": 285
+    },
+    {
+      "epoch": 0.06591380502419912,
+      "grad_norm": 0.623166561126709,
+      "learning_rate": 9.966559943769203e-05,
+      "loss": 0.8389,
+      "step": 286
+    },
+    {
+      "epoch": 0.06614427287393408,
+      "grad_norm": 0.5255123376846313,
+      "learning_rate": 9.966127556698936e-05,
+      "loss": 0.8486,
+      "step": 287
+    },
+    {
+      "epoch": 0.06637474072366904,
+      "grad_norm": 0.6101188063621521,
+      "learning_rate": 9.965692401628368e-05,
+      "loss": 0.8508,
+      "step": 288
+    },
+    {
+      "epoch": 0.066605208573404,
+      "grad_norm": 0.6269940733909607,
+      "learning_rate": 9.96525447880004e-05,
+      "loss": 0.8407,
+      "step": 289
+    },
+    {
+      "epoch": 0.06683567642313898,
+      "grad_norm": 0.7046296000480652,
+      "learning_rate": 9.964813788458043e-05,
+      "loss": 0.8329,
+      "step": 290
+    },
+    {
+      "epoch": 0.06706614427287394,
+      "grad_norm": 0.8176979422569275,
+      "learning_rate": 9.964370330848005e-05,
+      "loss": 0.8347,
+      "step": 291
+    },
+    {
+      "epoch": 0.0672966121226089,
+      "grad_norm": 0.8342549800872803,
+      "learning_rate": 9.963924106217102e-05,
+      "loss": 0.836,
+      "step": 292
+    },
+    {
+      "epoch": 0.06752707997234386,
+      "grad_norm": 0.8378840684890747,
+      "learning_rate": 9.963475114814045e-05,
+      "loss": 0.8339,
+      "step": 293
+    },
+    {
+      "epoch": 0.06775754782207882,
+      "grad_norm": 0.7933089733123779,
+      "learning_rate": 9.963023356889093e-05,
+      "loss": 0.8321,
+      "step": 294
+    },
+    {
+      "epoch": 0.06798801567181378,
+      "grad_norm": 0.7538984417915344,
+      "learning_rate": 9.962568832694044e-05,
+      "loss": 0.8395,
+      "step": 295
+    },
+    {
+      "epoch": 0.06821848352154875,
+      "grad_norm": 0.6032546162605286,
+      "learning_rate": 9.962111542482241e-05,
+      "loss": 0.8314,
+      "step": 296
+    },
+    {
+      "epoch": 0.0684489513712837,
+      "grad_norm": 0.694686770439148,
+      "learning_rate": 9.961651486508564e-05,
+      "loss": 0.826,
+      "step": 297
+    },
+    {
+      "epoch": 0.06867941922101867,
+      "grad_norm": 0.7310389876365662,
+      "learning_rate": 9.96118866502944e-05,
+      "loss": 0.8221,
+      "step": 298
+    },
+    {
+      "epoch": 0.06890988707075363,
+      "grad_norm": 0.6587698459625244,
+      "learning_rate": 9.960723078302832e-05,
+      "loss": 0.8275,
+      "step": 299
+    },
+    {
+      "epoch": 0.06914035492048859,
+      "grad_norm": 0.6249524354934692,
+      "learning_rate": 9.960254726588246e-05,
+      "loss": 0.8329,
+      "step": 300
+    },
+    {
+      "epoch": 0.06937082277022355,
+      "grad_norm": 0.6110243797302246,
+      "learning_rate": 9.959783610146733e-05,
+      "loss": 0.8296,
+      "step": 301
+    },
+    {
+      "epoch": 0.06960129061995851,
+      "grad_norm": 0.528419017791748,
+      "learning_rate": 9.959309729240882e-05,
+      "loss": 0.825,
+      "step": 302
+    },
+    {
+      "epoch": 0.06983175846969347,
+      "grad_norm": 0.47809135913848877,
+      "learning_rate": 9.95883308413482e-05,
+      "loss": 0.8163,
+      "step": 303
+    },
+    {
+      "epoch": 0.07006222631942843,
+      "grad_norm": 0.4991075098514557,
+      "learning_rate": 9.95835367509422e-05,
+      "loss": 0.8368,
+      "step": 304
+    },
+    {
+      "epoch": 0.07029269416916341,
+      "grad_norm": 0.5650983452796936,
+      "learning_rate": 9.957871502386291e-05,
+      "loss": 0.8194,
+      "step": 305
+    },
+    {
+      "epoch": 0.07052316201889837,
+      "grad_norm": 0.4641514718532562,
+      "learning_rate": 9.957386566279788e-05,
+      "loss": 0.8259,
+      "step": 306
+    },
+    {
+      "epoch": 0.07075362986863333,
+      "grad_norm": 0.4920820891857147,
+      "learning_rate": 9.956898867044999e-05,
+      "loss": 0.8188,
+      "step": 307
+    },
+    {
+      "epoch": 0.07098409771836829,
+      "grad_norm": 0.525382936000824,
+      "learning_rate": 9.956408404953756e-05,
+      "loss": 0.8156,
+      "step": 308
+    },
+    {
+      "epoch": 0.07121456556810325,
+      "grad_norm": 0.5120217800140381,
+      "learning_rate": 9.955915180279433e-05,
+      "loss": 0.8213,
+      "step": 309
+    },
+    {
+      "epoch": 0.07144503341783821,
+      "grad_norm": 0.448485791683197,
+      "learning_rate": 9.955419193296943e-05,
+      "loss": 0.8191,
+      "step": 310
+    },
+    {
+      "epoch": 0.07167550126757317,
+      "grad_norm": 0.5565616488456726,
+      "learning_rate": 9.954920444282732e-05,
+      "loss": 0.8244,
+      "step": 311
+    },
+    {
+      "epoch": 0.07190596911730814,
+      "grad_norm": 0.5660645365715027,
+      "learning_rate": 9.954418933514795e-05,
+      "loss": 0.8114,
+      "step": 312
+    },
+    {
+      "epoch": 0.0721364369670431,
+      "grad_norm": 0.5123659372329712,
+      "learning_rate": 9.953914661272661e-05,
+      "loss": 0.8135,
+      "step": 313
+    },
+    {
+      "epoch": 0.07236690481677806,
+      "grad_norm": 0.47135382890701294,
+      "learning_rate": 9.953407627837398e-05,
+      "loss": 0.813,
+      "step": 314
+    },
+    {
+      "epoch": 0.07259737266651302,
+      "grad_norm": 0.5098188519477844,
+      "learning_rate": 9.952897833491617e-05,
+      "loss": 0.8116,
+      "step": 315
+    },
+    {
+      "epoch": 0.07282784051624798,
+      "grad_norm": 0.5356022715568542,
+      "learning_rate": 9.952385278519462e-05,
+      "loss": 0.8152,
+      "step": 316
+    },
+    {
+      "epoch": 0.07305830836598294,
+      "grad_norm": 0.5658923983573914,
+      "learning_rate": 9.951869963206622e-05,
+      "loss": 0.8085,
+      "step": 317
+    },
+    {
+      "epoch": 0.0732887762157179,
+      "grad_norm": 0.6336301565170288,
+      "learning_rate": 9.951351887840317e-05,
+      "loss": 0.8137,
+      "step": 318
+    },
+    {
+      "epoch": 0.07351924406545286,
+      "grad_norm": 0.7044976949691772,
+      "learning_rate": 9.950831052709314e-05,
+      "loss": 0.809,
+      "step": 319
+    },
+    {
+      "epoch": 0.07374971191518784,
+      "grad_norm": 0.6830228567123413,
+      "learning_rate": 9.950307458103911e-05,
+      "loss": 0.8091,
+      "step": 320
+    },
+    {
+      "epoch": 0.0739801797649228,
+      "grad_norm": 0.6315388083457947,
+      "learning_rate": 9.949781104315951e-05,
+      "loss": 0.811,
+      "step": 321
+    },
+    {
+      "epoch": 0.07421064761465776,
+      "grad_norm": 0.44383642077445984,
+      "learning_rate": 9.949251991638806e-05,
+      "loss": 0.8062,
+      "step": 322
+    },
+    {
+      "epoch": 0.07444111546439272,
+      "grad_norm": 0.48055797815322876,
+      "learning_rate": 9.948720120367394e-05,
+      "loss": 0.8063,
+      "step": 323
+    },
+    {
+      "epoch": 0.07467158331412768,
+      "grad_norm": 0.452512264251709,
+      "learning_rate": 9.948185490798168e-05,
+      "loss": 0.8024,
+      "step": 324
+    },
+    {
+      "epoch": 0.07490205116386264,
+      "grad_norm": 0.46535223722457886,
+      "learning_rate": 9.947648103229113e-05,
+      "loss": 0.8065,
+      "step": 325
+    },
+    {
+      "epoch": 0.0751325190135976,
+      "grad_norm": 0.6422431468963623,
+      "learning_rate": 9.94710795795976e-05,
+      "loss": 0.8096,
+      "step": 326
+    },
+    {
+      "epoch": 0.07536298686333257,
+      "grad_norm": 0.6489291191101074,
+      "learning_rate": 9.946565055291174e-05,
+      "loss": 0.8136,
+      "step": 327
+    },
+    {
+      "epoch": 0.07559345471306753,
+      "grad_norm": 0.7198114991188049,
+      "learning_rate": 9.946019395525951e-05,
+      "loss": 0.8052,
+      "step": 328
+    },
+    {
+      "epoch": 0.07582392256280249,
+      "grad_norm": 0.6340928077697754,
+      "learning_rate": 9.945470978968234e-05,
+      "loss": 0.8,
+      "step": 329
+    },
+    {
+      "epoch": 0.07605439041253745,
+      "grad_norm": 0.6357820630073547,
+      "learning_rate": 9.944919805923694e-05,
+      "loss": 0.8031,
+      "step": 330
+    },
+    {
+      "epoch": 0.07628485826227241,
+      "grad_norm": 0.5786786675453186,
+      "learning_rate": 9.944365876699544e-05,
+      "loss": 0.8049,
+      "step": 331
+    },
+    {
+      "epoch": 0.07651532611200737,
+      "grad_norm": 0.4964509904384613,
+      "learning_rate": 9.943809191604527e-05,
+      "loss": 0.8092,
+      "step": 332
+    },
+    {
+      "epoch": 0.07674579396174233,
+      "grad_norm": 0.431325763463974,
+      "learning_rate": 9.943249750948929e-05,
+      "loss": 0.798,
+      "step": 333
+    },
+    {
+      "epoch": 0.07697626181147729,
+      "grad_norm": 0.41866233944892883,
+      "learning_rate": 9.942687555044568e-05,
+      "loss": 0.8017,
+      "step": 334
+    },
+    {
+      "epoch": 0.07720672966121227,
+      "grad_norm": 0.4726838767528534,
+      "learning_rate": 9.9421226042048e-05,
+      "loss": 0.8015,
+      "step": 335
+    },
+    {
+      "epoch": 0.07743719751094723,
+      "grad_norm": 0.473954439163208,
+      "learning_rate": 9.941554898744511e-05,
+      "loss": 0.7963,
+      "step": 336
+    },
+    {
+      "epoch": 0.07766766536068219,
+      "grad_norm": 0.49290353059768677,
+      "learning_rate": 9.940984438980131e-05,
+      "loss": 0.807,
+      "step": 337
+    },
+    {
+      "epoch": 0.07789813321041715,
+      "grad_norm": 0.4410178065299988,
+      "learning_rate": 9.940411225229618e-05,
+      "loss": 0.7958,
+      "step": 338
+    },
+    {
+      "epoch": 0.07812860106015211,
+      "grad_norm": 0.41889092326164246,
+      "learning_rate": 9.939835257812468e-05,
+      "loss": 0.7984,
+      "step": 339
+    },
+    {
+      "epoch": 0.07835906890988707,
+      "grad_norm": 0.4519578218460083,
+      "learning_rate": 9.939256537049711e-05,
+      "loss": 0.8046,
+      "step": 340
+    },
+    {
+      "epoch": 0.07858953675962203,
+      "grad_norm": 0.5191729664802551,
+      "learning_rate": 9.938675063263914e-05,
+      "loss": 0.8045,
+      "step": 341
+    },
+    {
+      "epoch": 0.078820004609357,
+      "grad_norm": 0.5066516399383545,
+      "learning_rate": 9.938090836779174e-05,
+      "loss": 0.7918,
+      "step": 342
+    },
+    {
+      "epoch": 0.07905047245909196,
+      "grad_norm": 0.4769982695579529,
+      "learning_rate": 9.937503857921125e-05,
+      "loss": 0.8122,
+      "step": 343
+    },
+    {
+      "epoch": 0.07928094030882692,
+      "grad_norm": 0.4449038505554199,
+      "learning_rate": 9.936914127016938e-05,
+      "loss": 0.8004,
+      "step": 344
+    },
+    {
+      "epoch": 0.07951140815856188,
+      "grad_norm": 0.38136690855026245,
+      "learning_rate": 9.936321644395312e-05,
+      "loss": 0.8063,
+      "step": 345
+    },
+    {
+      "epoch": 0.07974187600829684,
+      "grad_norm": 0.5003980994224548,
+      "learning_rate": 9.935726410386484e-05,
+      "loss": 0.7955,
+      "step": 346
+    },
+    {
+      "epoch": 0.0799723438580318,
+      "grad_norm": 0.47362205386161804,
+      "learning_rate": 9.93512842532222e-05,
+      "loss": 0.7994,
+      "step": 347
+    },
+    {
+      "epoch": 0.08020281170776676,
+      "grad_norm": 0.4894074499607086,
+      "learning_rate": 9.934527689535826e-05,
+      "loss": 0.8096,
+      "step": 348
+    },
+    {
+      "epoch": 0.08043327955750172,
+      "grad_norm": 0.4951101839542389,
+      "learning_rate": 9.933924203362138e-05,
+      "loss": 0.7788,
+      "step": 349
+    },
+    {
+      "epoch": 0.0806637474072367,
+      "grad_norm": 0.46733275055885315,
+      "learning_rate": 9.933317967137524e-05,
+      "loss": 0.7885,
+      "step": 350
+    },
+    {
+      "epoch": 0.08089421525697166,
+      "grad_norm": 0.48279932141304016,
+      "learning_rate": 9.932708981199883e-05,
+      "loss": 0.7924,
+      "step": 351
+    },
+    {
+      "epoch": 0.08112468310670662,
+      "grad_norm": 0.48102623224258423,
+      "learning_rate": 9.932097245888652e-05,
+      "loss": 0.7921,
+      "step": 352
+    },
+    {
+      "epoch": 0.08135515095644158,
+      "grad_norm": 0.4802628457546234,
+      "learning_rate": 9.931482761544797e-05,
+      "loss": 0.8044,
+      "step": 353
+    },
+    {
+      "epoch": 0.08158561880617654,
+      "grad_norm": 0.4573599398136139,
+      "learning_rate": 9.930865528510815e-05,
+      "loss": 0.7928,
+      "step": 354
+    },
+    {
+      "epoch": 0.0818160866559115,
+      "grad_norm": 0.5361927151679993,
+      "learning_rate": 9.93024554713074e-05,
+      "loss": 0.7967,
+      "step": 355
+    },
+    {
+      "epoch": 0.08204655450564646,
+      "grad_norm": 0.56016606092453,
+      "learning_rate": 9.929622817750133e-05,
+      "loss": 0.8019,
+      "step": 356
+    },
+    {
+      "epoch": 0.08227702235538142,
+      "grad_norm": 0.5165932774543762,
+      "learning_rate": 9.928997340716087e-05,
+      "loss": 0.7981,
+      "step": 357
+    },
+    {
+      "epoch": 0.08250749020511638,
+      "grad_norm": 0.4367765188217163,
+      "learning_rate": 9.928369116377231e-05,
+      "loss": 0.7892,
+      "step": 358
+    },
+    {
+      "epoch": 0.08273795805485135,
+      "grad_norm": 0.6103436946868896,
+      "learning_rate": 9.92773814508372e-05,
+      "loss": 0.7887,
+      "step": 359
+    },
+    {
+      "epoch": 0.08296842590458631,
+      "grad_norm": 0.6656332015991211,
+      "learning_rate": 9.927104427187243e-05,
+      "loss": 0.7924,
+      "step": 360
+    },
+    {
+      "epoch": 0.08319889375432127,
+      "grad_norm": 0.6310869455337524,
+      "learning_rate": 9.926467963041018e-05,
+      "loss": 0.7973,
+      "step": 361
+    },
+    {
+      "epoch": 0.08342936160405623,
+      "grad_norm": 0.6289921998977661,
+      "learning_rate": 9.925828752999797e-05,
+      "loss": 0.7897,
+      "step": 362
+    },
+    {
+      "epoch": 0.08365982945379119,
+      "grad_norm": 0.501392126083374,
+      "learning_rate": 9.925186797419858e-05,
+      "loss": 0.7915,
+      "step": 363
+    },
+    {
+      "epoch": 0.08389029730352615,
+      "grad_norm": 0.4528777301311493,
+      "learning_rate": 9.924542096659015e-05,
+      "loss": 0.7925,
+      "step": 364
+    },
+    {
+      "epoch": 0.08412076515326113,
+      "grad_norm": 0.5350551605224609,
+      "learning_rate": 9.923894651076605e-05,
+      "loss": 0.7941,
+      "step": 365
+    },
+    {
+      "epoch": 0.08435123300299609,
+      "grad_norm": 0.4819077253341675,
+      "learning_rate": 9.9232444610335e-05,
+      "loss": 0.7899,
+      "step": 366
+    },
+    {
+      "epoch": 0.08458170085273105,
+      "grad_norm": 0.47890499234199524,
+      "learning_rate": 9.9225915268921e-05,
+      "loss": 0.7899,
+      "step": 367
+    },
+    {
+      "epoch": 0.08481216870246601,
+      "grad_norm": 0.6021353602409363,
+      "learning_rate": 9.921935849016338e-05,
+      "loss": 0.7949,
+      "step": 368
+    },
+    {
+      "epoch": 0.08504263655220097,
+      "grad_norm": 0.624280571937561,
+      "learning_rate": 9.921277427771667e-05,
+      "loss": 0.7977,
+      "step": 369
+    },
+    {
+      "epoch": 0.08527310440193593,
+      "grad_norm": 0.48100805282592773,
+      "learning_rate": 9.92061626352508e-05,
+      "loss": 0.7847,
+      "step": 370
+    },
+    {
+      "epoch": 0.08550357225167089,
+      "grad_norm": 0.4303872287273407,
+      "learning_rate": 9.919952356645092e-05,
+      "loss": 0.7925,
+      "step": 371
+    },
+    {
+      "epoch": 0.08573404010140585,
+      "grad_norm": 0.4995540678501129,
+      "learning_rate": 9.91928570750175e-05,
+      "loss": 0.7931,
+      "step": 372
+    },
+    {
+      "epoch": 0.08596450795114081,
+      "grad_norm": 0.49925413727760315,
+      "learning_rate": 9.918616316466628e-05,
+      "loss": 0.7875,
+      "step": 373
+    },
+    {
+      "epoch": 0.08619497580087578,
+      "grad_norm": 0.4929007887840271,
+      "learning_rate": 9.917944183912828e-05,
+      "loss": 0.7821,
+      "step": 374
+    },
+    {
+      "epoch": 0.08642544365061074,
+      "grad_norm": 0.4669882357120514,
+      "learning_rate": 9.91726931021498e-05,
+      "loss": 0.786,
+      "step": 375
+    },
+    {
+      "epoch": 0.0866559115003457,
+      "grad_norm": 0.4014081358909607,
+      "learning_rate": 9.916591695749244e-05,
+      "loss": 0.7833,
+      "step": 376
+    },
+    {
+      "epoch": 0.08688637935008066,
+      "grad_norm": 0.38111427426338196,
+      "learning_rate": 9.915911340893305e-05,
+      "loss": 0.7894,
+      "step": 377
+    },
+    {
+      "epoch": 0.08711684719981562,
+      "grad_norm": 0.457692950963974,
+      "learning_rate": 9.915228246026376e-05,
+      "loss": 0.7912,
+      "step": 378
+    },
+    {
+      "epoch": 0.0873473150495506,
+      "grad_norm": 0.467342734336853,
+      "learning_rate": 9.9145424115292e-05,
+      "loss": 0.7843,
+      "step": 379
+    },
+    {
+      "epoch": 0.08757778289928556,
+      "grad_norm": 0.44467970728874207,
+      "learning_rate": 9.913853837784042e-05,
+      "loss": 0.7939,
+      "step": 380
+    },
+    {
+      "epoch": 0.08780825074902052,
+      "grad_norm": 0.4361025094985962,
+      "learning_rate": 9.913162525174697e-05,
+      "loss": 0.7803,
+      "step": 381
+    },
+    {
+      "epoch": 0.08803871859875548,
+      "grad_norm": 0.44507619738578796,
+      "learning_rate": 9.912468474086486e-05,
+      "loss": 0.7765,
+      "step": 382
+    },
+    {
+      "epoch": 0.08826918644849044,
+      "grad_norm": 0.5151132941246033,
+      "learning_rate": 9.911771684906257e-05,
+      "loss": 0.7824,
+      "step": 383
+    },
+    {
+      "epoch": 0.0884996542982254,
+      "grad_norm": 0.45923879742622375,
+      "learning_rate": 9.911072158022385e-05,
+      "loss": 0.7881,
+      "step": 384
+    },
+    {
+      "epoch": 0.08873012214796036,
+      "grad_norm": 0.41044628620147705,
+      "learning_rate": 9.910369893824767e-05,
+      "loss": 0.7841,
+      "step": 385
+    },
+    {
+      "epoch": 0.08896058999769532,
+      "grad_norm": 0.41591617465019226,
+      "learning_rate": 9.90966489270483e-05,
+      "loss": 0.7754,
+      "step": 386
+    },
+    {
+      "epoch": 0.08919105784743028,
+      "grad_norm": 0.5302747488021851,
+      "learning_rate": 9.908957155055523e-05,
+      "loss": 0.7784,
+      "step": 387
+    },
+    {
+      "epoch": 0.08942152569716524,
+      "grad_norm": 0.41914960741996765,
+      "learning_rate": 9.908246681271322e-05,
+      "loss": 0.7817,
+      "step": 388
+    },
+    {
+      "epoch": 0.0896519935469002,
+      "grad_norm": 0.38282233476638794,
+      "learning_rate": 9.907533471748231e-05,
+      "loss": 0.779,
+      "step": 389
+    },
+    {
+      "epoch": 0.08988246139663517,
+      "grad_norm": 0.4711924195289612,
+      "learning_rate": 9.906817526883774e-05,
+      "loss": 0.7853,
+      "step": 390
+    },
+    {
+      "epoch": 0.09011292924637013,
+      "grad_norm": 0.45906805992126465,
+      "learning_rate": 9.906098847076999e-05,
+      "loss": 0.7711,
+      "step": 391
+    },
+    {
+      "epoch": 0.09034339709610509,
+      "grad_norm": 0.44320812821388245,
+      "learning_rate": 9.905377432728484e-05,
+      "loss": 0.7792,
+      "step": 392
+    },
+    {
+      "epoch": 0.09057386494584005,
+      "grad_norm": 0.41927963495254517,
+      "learning_rate": 9.904653284240328e-05,
+      "loss": 0.784,
+      "step": 393
+    },
+    {
+      "epoch": 0.09080433279557502,
+      "grad_norm": 0.5191970467567444,
+      "learning_rate": 9.903926402016153e-05,
+      "loss": 0.7834,
+      "step": 394
+    },
+    {
+      "epoch": 0.09103480064530999,
+      "grad_norm": 0.6078465580940247,
+      "learning_rate": 9.903196786461106e-05,
+      "loss": 0.7784,
+      "step": 395
+    },
+    {
+      "epoch": 0.09126526849504495,
+      "grad_norm": 0.6223181486129761,
+      "learning_rate": 9.902464437981855e-05,
+      "loss": 0.7913,
+      "step": 396
+    },
+    {
+      "epoch": 0.09149573634477991,
+      "grad_norm": 0.6932902336120605,
+      "learning_rate": 9.901729356986597e-05,
+      "loss": 0.7817,
+      "step": 397
+    },
+    {
+      "epoch": 0.09172620419451487,
+      "grad_norm": 0.6165162920951843,
+      "learning_rate": 9.900991543885048e-05,
+      "loss": 0.775,
+      "step": 398
+    },
+    {
+      "epoch": 0.09195667204424983,
+      "grad_norm": 0.5054426193237305,
+      "learning_rate": 9.900250999088447e-05,
+      "loss": 0.7824,
+      "step": 399
+    },
+    {
+      "epoch": 0.09218713989398479,
+      "grad_norm": 0.5113014578819275,
+      "learning_rate": 9.899507723009554e-05,
+      "loss": 0.7827,
+      "step": 400
+    },
+    {
+      "epoch": 0.09241760774371975,
+      "grad_norm": 0.5064506530761719,
+      "learning_rate": 9.898761716062654e-05,
+      "loss": 0.7848,
+      "step": 401
+    },
+    {
+      "epoch": 0.09264807559345471,
+      "grad_norm": 0.4201597571372986,
+      "learning_rate": 9.898012978663553e-05,
+      "loss": 0.7804,
+      "step": 402
+    },
+    {
+      "epoch": 0.09287854344318967,
+      "grad_norm": 0.47890499234199524,
+      "learning_rate": 9.897261511229583e-05,
+      "loss": 0.7775,
+      "step": 403
+    },
+    {
+      "epoch": 0.09310901129292463,
+      "grad_norm": 0.5132191777229309,
+      "learning_rate": 9.896507314179588e-05,
+      "loss": 0.7739,
+      "step": 404
+    },
+    {
+      "epoch": 0.0933394791426596,
+      "grad_norm": 0.5053228735923767,
+      "learning_rate": 9.895750387933944e-05,
+      "loss": 0.7725,
+      "step": 405
+    },
+    {
+      "epoch": 0.09356994699239456,
+      "grad_norm": 0.5740991830825806,
+      "learning_rate": 9.894990732914541e-05,
+      "loss": 0.7784,
+      "step": 406
+    },
+    {
+      "epoch": 0.09380041484212952,
+      "grad_norm": 0.5145570635795593,
+      "learning_rate": 9.894228349544796e-05,
+      "loss": 0.7783,
+      "step": 407
+    },
+    {
+      "epoch": 0.09403088269186448,
+      "grad_norm": 0.5113042593002319,
+      "learning_rate": 9.893463238249638e-05,
+      "loss": 0.7835,
+      "step": 408
+    },
+    {
+      "epoch": 0.09426135054159945,
+      "grad_norm": 0.5256659984588623,
+      "learning_rate": 9.892695399455525e-05,
+      "loss": 0.7746,
+      "step": 409
+    },
+    {
+      "epoch": 0.09449181839133441,
+      "grad_norm": 0.5014989972114563,
+      "learning_rate": 9.891924833590431e-05,
+      "loss": 0.7715,
+      "step": 410
+    },
+    {
+      "epoch": 0.09472228624106938,
+      "grad_norm": 0.45504307746887207,
+      "learning_rate": 9.891151541083852e-05,
+      "loss": 0.7749,
+      "step": 411
+    },
+    {
+      "epoch": 0.09495275409080434,
+      "grad_norm": 0.4456666111946106,
+      "learning_rate": 9.8903755223668e-05,
+      "loss": 0.7685,
+      "step": 412
+    },
+    {
+      "epoch": 0.0951832219405393,
+      "grad_norm": 0.45063498616218567,
+      "learning_rate": 9.889596777871812e-05,
+      "loss": 0.7675,
+      "step": 413
+    },
+    {
+      "epoch": 0.09541368979027426,
+      "grad_norm": 0.48387956619262695,
+      "learning_rate": 9.88881530803294e-05,
+      "loss": 0.7759,
+      "step": 414
+    },
+    {
+      "epoch": 0.09564415764000922,
+      "grad_norm": 0.4509301483631134,
+      "learning_rate": 9.888031113285757e-05,
+      "loss": 0.7717,
+      "step": 415
+    },
+    {
+      "epoch": 0.09587462548974418,
+      "grad_norm": 0.453188419342041,
+      "learning_rate": 9.887244194067355e-05,
+      "loss": 0.7707,
+      "step": 416
+    },
+    {
+      "epoch": 0.09610509333947914,
+      "grad_norm": 0.4492356777191162,
+      "learning_rate": 9.886454550816342e-05,
+      "loss": 0.7736,
+      "step": 417
+    },
+    {
+      "epoch": 0.0963355611892141,
+      "grad_norm": 0.3863151967525482,
+      "learning_rate": 9.885662183972848e-05,
+      "loss": 0.7812,
+      "step": 418
+    },
+    {
+      "epoch": 0.09656602903894906,
+      "grad_norm": 0.4794161021709442,
+      "learning_rate": 9.884867093978519e-05,
+      "loss": 0.7704,
+      "step": 419
+    },
+    {
+      "epoch": 0.09679649688868402,
+      "grad_norm": 0.44275277853012085,
+      "learning_rate": 9.884069281276517e-05,
+      "loss": 0.7693,
+      "step": 420
+    },
+    {
+      "epoch": 0.09702696473841899,
+      "grad_norm": 0.4233629107475281,
+      "learning_rate": 9.883268746311528e-05,
+      "loss": 0.7742,
+      "step": 421
+    },
+    {
+      "epoch": 0.09725743258815395,
+      "grad_norm": 0.34396764636039734,
+      "learning_rate": 9.882465489529747e-05,
+      "loss": 0.772,
+      "step": 422
+    },
+    {
+      "epoch": 0.09748790043788891,
+      "grad_norm": 0.4056984484195709,
+      "learning_rate": 9.881659511378892e-05,
+      "loss": 0.7751,
+      "step": 423
+    },
+    {
+      "epoch": 0.09771836828762388,
+      "grad_norm": 0.4664100408554077,
+      "learning_rate": 9.880850812308196e-05,
+      "loss": 0.7706,
+      "step": 424
+    },
+    {
+      "epoch": 0.09794883613735884,
+      "grad_norm": 0.5346910357475281,
+      "learning_rate": 9.880039392768405e-05,
+      "loss": 0.7661,
+      "step": 425
+    },
+    {
+      "epoch": 0.0981793039870938,
+      "grad_norm": 0.4764558970928192,
+      "learning_rate": 9.87922525321179e-05,
+      "loss": 0.7794,
+      "step": 426
+    },
+    {
+      "epoch": 0.09840977183682877,
+      "grad_norm": 0.39610040187835693,
+      "learning_rate": 9.878408394092129e-05,
+      "loss": 0.7711,
+      "step": 427
+    },
+    {
+      "epoch": 0.09864023968656373,
+      "grad_norm": 0.42657968401908875,
+      "learning_rate": 9.877588815864722e-05,
+      "loss": 0.7739,
+      "step": 428
+    },
+    {
+      "epoch": 0.09887070753629869,
+      "grad_norm": 0.46280384063720703,
+      "learning_rate": 9.87676651898638e-05,
+      "loss": 0.7811,
+      "step": 429
+    },
+    {
+      "epoch": 0.09910117538603365,
+      "grad_norm": 0.44775912165641785,
+      "learning_rate": 9.875941503915434e-05,
+      "loss": 0.7733,
+      "step": 430
+    },
+    {
+      "epoch": 0.09933164323576861,
+      "grad_norm": 0.4808795154094696,
+      "learning_rate": 9.875113771111725e-05,
+      "loss": 0.7734,
+      "step": 431
+    },
+    {
+      "epoch": 0.09956211108550357,
+      "grad_norm": 0.42084455490112305,
+      "learning_rate": 9.874283321036615e-05,
+      "loss": 0.7679,
+      "step": 432
+    },
+    {
+      "epoch": 0.09979257893523853,
+      "grad_norm": 0.4245053231716156,
+      "learning_rate": 9.873450154152972e-05,
+      "loss": 0.7755,
+      "step": 433
+    },
+    {
+      "epoch": 0.1000230467849735,
+      "grad_norm": 0.4106716811656952,
+      "learning_rate": 9.872614270925188e-05,
+      "loss": 0.7757,
+      "step": 434
+    },
+    {
+      "epoch": 0.10025351463470845,
+      "grad_norm": 0.38926827907562256,
+      "learning_rate": 9.871775671819162e-05,
+      "loss": 0.7761,
+      "step": 435
+    },
+    {
+      "epoch": 0.10048398248444342,
+      "grad_norm": 0.3636447489261627,
+      "learning_rate": 9.870934357302308e-05,
+      "loss": 0.7722,
+      "step": 436
+    },
+    {
+      "epoch": 0.10071445033417838,
+      "grad_norm": 0.3969525992870331,
+      "learning_rate": 9.870090327843557e-05,
+      "loss": 0.7722,
+      "step": 437
+    },
+    {
+      "epoch": 0.10094491818391334,
+      "grad_norm": 0.36944982409477234,
+      "learning_rate": 9.869243583913348e-05,
+      "loss": 0.7716,
+      "step": 438
+    },
+    {
+      "epoch": 0.10117538603364831,
+      "grad_norm": 0.44781213998794556,
+      "learning_rate": 9.86839412598364e-05,
+      "loss": 0.7757,
+      "step": 439
+    },
+    {
+      "epoch": 0.10140585388338327,
+      "grad_norm": 0.5165262818336487,
+      "learning_rate": 9.867541954527894e-05,
+      "loss": 0.7678,
+      "step": 440
+    },
+    {
+      "epoch": 0.10163632173311823,
+      "grad_norm": 0.5470614433288574,
+      "learning_rate": 9.866687070021097e-05,
+      "loss": 0.7684,
+      "step": 441
+    },
+    {
+      "epoch": 0.1018667895828532,
+      "grad_norm": 0.4986436665058136,
+      "learning_rate": 9.865829472939736e-05,
+      "loss": 0.7682,
+      "step": 442
+    },
+    {
+      "epoch": 0.10209725743258816,
+      "grad_norm": 0.4134550988674164,
+      "learning_rate": 9.864969163761817e-05,
+      "loss": 0.7734,
+      "step": 443
+    },
+    {
+      "epoch": 0.10232772528232312,
+      "grad_norm": 0.5219729542732239,
+      "learning_rate": 9.864106142966856e-05,
+      "loss": 0.7744,
+      "step": 444
+    },
+    {
+      "epoch": 0.10255819313205808,
+      "grad_norm": 0.539275586605072,
+      "learning_rate": 9.863240411035878e-05,
+      "loss": 0.7721,
+      "step": 445
+    },
+    {
+      "epoch": 0.10278866098179304,
+      "grad_norm": 0.4756945073604584,
+      "learning_rate": 9.862371968451423e-05,
+      "loss": 0.7645,
+      "step": 446
+    },
+    {
+      "epoch": 0.103019128831528,
+      "grad_norm": 0.43748557567596436,
+      "learning_rate": 9.861500815697541e-05,
+      "loss": 0.7693,
+      "step": 447
+    },
+    {
+      "epoch": 0.10324959668126296,
+      "grad_norm": 0.356668084859848,
+      "learning_rate": 9.860626953259791e-05,
+      "loss": 0.7645,
+      "step": 448
+    },
+    {
+      "epoch": 0.10348006453099792,
+      "grad_norm": 0.31597620248794556,
+      "learning_rate": 9.859750381625241e-05,
+      "loss": 0.7626,
+      "step": 449
+    },
+    {
+      "epoch": 0.10371053238073288,
+      "grad_norm": 0.3589065968990326,
+      "learning_rate": 9.858871101282472e-05,
+      "loss": 0.782,
+      "step": 450
+    },
+    {
+      "epoch": 0.10394100023046784,
+      "grad_norm": 0.3751053810119629,
+      "learning_rate": 9.857989112721574e-05,
+      "loss": 0.7701,
+      "step": 451
+    },
+    {
+      "epoch": 0.1041714680802028,
+      "grad_norm": 0.38888126611709595,
+      "learning_rate": 9.857104416434147e-05,
+      "loss": 0.769,
+      "step": 452
+    },
+    {
+      "epoch": 0.10440193592993777,
+      "grad_norm": 0.3789352774620056,
+      "learning_rate": 9.856217012913299e-05,
+      "loss": 0.7725,
+      "step": 453
+    },
+    {
+      "epoch": 0.10463240377967274,
+      "grad_norm": 0.36712217330932617,
+      "learning_rate": 9.855326902653647e-05,
+      "loss": 0.7763,
+      "step": 454
+    },
+    {
+      "epoch": 0.1048628716294077,
+      "grad_norm": 0.38460057973861694,
+      "learning_rate": 9.854434086151318e-05,
+      "loss": 0.7603,
+      "step": 455
+    },
+    {
+      "epoch": 0.10509333947914266,
+      "grad_norm": 0.41386526823043823,
+      "learning_rate": 9.853538563903945e-05,
+      "loss": 0.7683,
+      "step": 456
+    },
+    {
+      "epoch": 0.10532380732887762,
+      "grad_norm": 0.42844244837760925,
+      "learning_rate": 9.852640336410671e-05,
+      "loss": 0.7587,
+      "step": 457
+    },
+    {
+      "epoch": 0.10555427517861259,
+      "grad_norm": 0.41544997692108154,
+      "learning_rate": 9.851739404172147e-05,
+      "loss": 0.7622,
+      "step": 458
+    },
+    {
+      "epoch": 0.10578474302834755,
+      "grad_norm": 0.42414671182632446,
+      "learning_rate": 9.850835767690532e-05,
+      "loss": 0.7644,
+      "step": 459
+    },
+    {
+      "epoch": 0.10601521087808251,
+      "grad_norm": 0.4521081745624542,
+      "learning_rate": 9.849929427469488e-05,
+      "loss": 0.7621,
+      "step": 460
+    },
+    {
+      "epoch": 0.10624567872781747,
+      "grad_norm": 0.4264926314353943,
+      "learning_rate": 9.849020384014192e-05,
+      "loss": 0.7661,
+      "step": 461
+    },
+    {
+      "epoch": 0.10647614657755243,
+      "grad_norm": 0.4072660207748413,
+      "learning_rate": 9.848108637831319e-05,
+      "loss": 0.7686,
+      "step": 462
+    },
+    {
+      "epoch": 0.10670661442728739,
+      "grad_norm": 0.38507211208343506,
+      "learning_rate": 9.847194189429058e-05,
+      "loss": 0.7705,
+      "step": 463
+    },
+    {
+      "epoch": 0.10693708227702235,
+      "grad_norm": 0.4261547923088074,
+      "learning_rate": 9.846277039317095e-05,
+      "loss": 0.7827,
+      "step": 464
+    },
+    {
+      "epoch": 0.10716755012675731,
+      "grad_norm": 0.5000432133674622,
+      "learning_rate": 9.845357188006635e-05,
+      "loss": 0.7627,
+      "step": 465
+    },
+    {
+      "epoch": 0.10739801797649227,
+      "grad_norm": 0.5209843516349792,
+      "learning_rate": 9.844434636010373e-05,
+      "loss": 0.7754,
+      "step": 466
+    },
+    {
+      "epoch": 0.10762848582622724,
+      "grad_norm": 0.4857335090637207,
+      "learning_rate": 9.843509383842525e-05,
+      "loss": 0.7687,
+      "step": 467
+    },
+    {
+      "epoch": 0.1078589536759622,
+      "grad_norm": 0.5346872806549072,
+      "learning_rate": 9.842581432018798e-05,
+      "loss": 0.772,
+      "step": 468
+    },
+    {
+      "epoch": 0.10808942152569717,
+      "grad_norm": 0.6131623387336731,
+      "learning_rate": 9.841650781056413e-05,
+      "loss": 0.7808,
+      "step": 469
+    },
+    {
+      "epoch": 0.10831988937543213,
+      "grad_norm": 0.5165646076202393,
+      "learning_rate": 9.840717431474094e-05,
+      "loss": 0.772,
+      "step": 470
+    },
+    {
+      "epoch": 0.1085503572251671,
+      "grad_norm": 0.39740481972694397,
+      "learning_rate": 9.839781383792064e-05,
+      "loss": 0.7671,
+      "step": 471
+    },
+    {
+      "epoch": 0.10878082507490205,
+      "grad_norm": 0.42045560479164124,
+      "learning_rate": 9.838842638532056e-05,
+      "loss": 0.7667,
+      "step": 472
+    },
+    {
+      "epoch": 0.10901129292463702,
+      "grad_norm": 0.4750736355781555,
+      "learning_rate": 9.837901196217303e-05,
+      "loss": 0.7565,
+      "step": 473
+    },
+    {
+      "epoch": 0.10924176077437198,
+      "grad_norm": 0.3689127266407013,
+      "learning_rate": 9.836957057372544e-05,
+      "loss": 0.7652,
+      "step": 474
+    },
+    {
+      "epoch": 0.10947222862410694,
+      "grad_norm": 0.3930782675743103,
+      "learning_rate": 9.836010222524018e-05,
+      "loss": 0.7538,
+      "step": 475
+    },
+    {
+      "epoch": 0.1097026964738419,
+      "grad_norm": 0.45599618554115295,
+      "learning_rate": 9.835060692199468e-05,
+      "loss": 0.7724,
+      "step": 476
+    },
+    {
+      "epoch": 0.10993316432357686,
+      "grad_norm": 0.5148651003837585,
+      "learning_rate": 9.83410846692814e-05,
+      "loss": 0.7612,
+      "step": 477
+    },
+    {
+      "epoch": 0.11016363217331182,
+      "grad_norm": 0.4592342674732208,
+      "learning_rate": 9.83315354724078e-05,
+      "loss": 0.7667,
+      "step": 478
+    },
+    {
+      "epoch": 0.11039410002304678,
+      "grad_norm": 0.40691468119621277,
+      "learning_rate": 9.83219593366964e-05,
+      "loss": 0.7676,
+      "step": 479
+    },
+    {
+      "epoch": 0.11062456787278174,
+      "grad_norm": 0.4532166123390198,
+      "learning_rate": 9.831235626748467e-05,
+      "loss": 0.7683,
+      "step": 480
+    },
+    {
+      "epoch": 0.1108550357225167,
+      "grad_norm": 0.42121079564094543,
+      "learning_rate": 9.830272627012518e-05,
+      "loss": 0.7691,
+      "step": 481
+    },
+    {
+      "epoch": 0.11108550357225166,
+      "grad_norm": 0.4073924422264099,
+      "learning_rate": 9.82930693499854e-05,
+      "loss": 0.7705,
+      "step": 482
+    },
+    {
+      "epoch": 0.11131597142198664,
+      "grad_norm": 0.39546170830726624,
+      "learning_rate": 9.828338551244794e-05,
+      "loss": 0.7684,
+      "step": 483
+    },
+    {
+      "epoch": 0.1115464392717216,
+      "grad_norm": 0.38143351674079895,
+      "learning_rate": 9.827367476291027e-05,
+      "loss": 0.7554,
+      "step": 484
+    },
+    {
+      "epoch": 0.11177690712145656,
+      "grad_norm": 0.3818657100200653,
+      "learning_rate": 9.826393710678497e-05,
+      "loss": 0.7632,
+      "step": 485
+    },
+    {
+      "epoch": 0.11200737497119152,
+      "grad_norm": 0.40947970747947693,
+      "learning_rate": 9.825417254949953e-05,
+      "loss": 0.7628,
+      "step": 486
+    },
+    {
+      "epoch": 0.11223784282092648,
+      "grad_norm": 0.40739309787750244,
+      "learning_rate": 9.824438109649654e-05,
+      "loss": 0.7659,
+      "step": 487
+    },
+    {
+      "epoch": 0.11246831067066144,
+      "grad_norm": 0.4805825650691986,
+      "learning_rate": 9.823456275323348e-05,
+      "loss": 0.7682,
+      "step": 488
+    },
+    {
+      "epoch": 0.1126987785203964,
+      "grad_norm": 0.48237094283103943,
+      "learning_rate": 9.822471752518288e-05,
+      "loss": 0.7658,
+      "step": 489
+    },
+    {
+      "epoch": 0.11292924637013137,
+      "grad_norm": 0.44534748792648315,
+      "learning_rate": 9.821484541783221e-05,
+      "loss": 0.769,
+      "step": 490
+    },
+    {
+      "epoch": 0.11315971421986633,
+      "grad_norm": 0.4775294363498688,
+      "learning_rate": 9.820494643668396e-05,
+      "loss": 0.7654,
+      "step": 491
+    },
+    {
+      "epoch": 0.11339018206960129,
+      "grad_norm": 0.4316689968109131,
+      "learning_rate": 9.81950205872556e-05,
+      "loss": 0.7689,
+      "step": 492
+    },
+    {
+      "epoch": 0.11362064991933625,
+      "grad_norm": 0.4406385123729706,
+      "learning_rate": 9.818506787507952e-05,
+      "loss": 0.7615,
+      "step": 493
+    },
+    {
+      "epoch": 0.11385111776907121,
+      "grad_norm": 0.37954282760620117,
+      "learning_rate": 9.817508830570318e-05,
+      "loss": 0.7581,
+      "step": 494
+    },
+    {
+      "epoch": 0.11408158561880617,
+      "grad_norm": 0.37608101963996887,
+      "learning_rate": 9.81650818846889e-05,
+      "loss": 0.7619,
+      "step": 495
+    },
+    {
+      "epoch": 0.11431205346854113,
+      "grad_norm": 0.35068264603614807,
+      "learning_rate": 9.815504861761404e-05,
+      "loss": 0.763,
+      "step": 496
+    },
+    {
+      "epoch": 0.1145425213182761,
+      "grad_norm": 0.3810550570487976,
+      "learning_rate": 9.81449885100709e-05,
+      "loss": 0.7551,
+      "step": 497
+    },
+    {
+      "epoch": 0.11477298916801107,
+      "grad_norm": 0.3873169422149658,
+      "learning_rate": 9.813490156766676e-05,
+      "loss": 0.7606,
+      "step": 498
+    },
+    {
+      "epoch": 0.11500345701774603,
+      "grad_norm": 0.39904019236564636,
+      "learning_rate": 9.812478779602381e-05,
+      "loss": 0.7562,
+      "step": 499
+    },
+    {
+      "epoch": 0.11523392486748099,
+      "grad_norm": 0.379290908575058,
+      "learning_rate": 9.811464720077923e-05,
+      "loss": 0.7544,
+      "step": 500
+    },
+    {
+      "epoch": 0.11546439271721595,
+      "grad_norm": 0.3641830384731293,
+      "learning_rate": 9.810447978758517e-05,
+      "loss": 0.7587,
+      "step": 501
+    },
+    {
+      "epoch": 0.11569486056695091,
+      "grad_norm": 0.3822626769542694,
+      "learning_rate": 9.809428556210867e-05,
+      "loss": 0.7577,
+      "step": 502
+    },
+    {
+      "epoch": 0.11592532841668587,
+      "grad_norm": 0.42176035046577454,
+      "learning_rate": 9.808406453003175e-05,
+      "loss": 0.7602,
+      "step": 503
+    },
+    {
+      "epoch": 0.11615579626642084,
+      "grad_norm": 0.3986082971096039,
+      "learning_rate": 9.80738166970514e-05,
+      "loss": 0.7504,
+      "step": 504
+    },
+    {
+      "epoch": 0.1163862641161558,
+      "grad_norm": 0.3723549246788025,
+      "learning_rate": 9.806354206887949e-05,
+      "loss": 0.764,
+      "step": 505
+    },
+    {
+      "epoch": 0.11661673196589076,
+      "grad_norm": 0.3810165524482727,
+      "learning_rate": 9.805324065124283e-05,
+      "loss": 0.7632,
+      "step": 506
+    },
+    {
+      "epoch": 0.11684719981562572,
+      "grad_norm": 0.3603302538394928,
+      "learning_rate": 9.804291244988324e-05,
+      "loss": 0.7572,
+      "step": 507
+    },
+    {
+      "epoch": 0.11707766766536068,
+      "grad_norm": 0.3849985897541046,
+      "learning_rate": 9.803255747055737e-05,
+      "loss": 0.7562,
+      "step": 508
+    },
+    {
+      "epoch": 0.11730813551509564,
+      "grad_norm": 0.3381344974040985,
+      "learning_rate": 9.802217571903685e-05,
+      "loss": 0.7643,
+      "step": 509
+    },
+    {
+      "epoch": 0.1175386033648306,
+      "grad_norm": 0.35623160004615784,
+      "learning_rate": 9.80117672011082e-05,
+      "loss": 0.752,
+      "step": 510
+    },
+    {
+      "epoch": 0.11776907121456556,
+      "grad_norm": 0.36817556619644165,
+      "learning_rate": 9.800133192257291e-05,
+      "loss": 0.7709,
+      "step": 511
+    },
+    {
+      "epoch": 0.11799953906430052,
+      "grad_norm": 0.5149376392364502,
+      "learning_rate": 9.799086988924733e-05,
+      "loss": 0.7566,
+      "step": 512
+    },
+    {
+      "epoch": 0.1182300069140355,
+      "grad_norm": 0.4210844039916992,
+      "learning_rate": 9.798038110696275e-05,
+      "loss": 0.7523,
+      "step": 513
+    },
+    {
+      "epoch": 0.11846047476377046,
+      "grad_norm": 0.4472396671772003,
+      "learning_rate": 9.796986558156538e-05,
+      "loss": 0.7586,
+      "step": 514
+    },
+    {
+      "epoch": 0.11869094261350542,
+      "grad_norm": 0.45051324367523193,
+      "learning_rate": 9.79593233189163e-05,
+      "loss": 0.7561,
+      "step": 515
+    },
+    {
+      "epoch": 0.11892141046324038,
+      "grad_norm": 0.4668648838996887,
+      "learning_rate": 9.794875432489152e-05,
+      "loss": 0.7548,
+      "step": 516
+    },
+    {
+      "epoch": 0.11915187831297534,
+      "grad_norm": 0.3892306685447693,
+      "learning_rate": 9.793815860538197e-05,
+      "loss": 0.7536,
+      "step": 517
+    },
+    {
+      "epoch": 0.1193823461627103,
+      "grad_norm": 0.36023855209350586,
+      "learning_rate": 9.792753616629342e-05,
+      "loss": 0.7595,
+      "step": 518
+    },
+    {
+      "epoch": 0.11961281401244526,
+      "grad_norm": 0.40682411193847656,
+      "learning_rate": 9.791688701354656e-05,
+      "loss": 0.7583,
+      "step": 519
+    },
+    {
+      "epoch": 0.11984328186218023,
+      "grad_norm": 0.464751660823822,
+      "learning_rate": 9.790621115307699e-05,
+      "loss": 0.7654,
+      "step": 520
+    },
+    {
+      "epoch": 0.12007374971191519,
+      "grad_norm": 0.4933690130710602,
+      "learning_rate": 9.789550859083517e-05,
+      "loss": 0.7567,
+      "step": 521
+    },
+    {
+      "epoch": 0.12030421756165015,
+      "grad_norm": 0.459985613822937,
+      "learning_rate": 9.788477933278646e-05,
+      "loss": 0.7464,
+      "step": 522
+    },
+    {
+      "epoch": 0.12053468541138511,
+      "grad_norm": 0.43470343947410583,
+      "learning_rate": 9.787402338491108e-05,
+      "loss": 0.7539,
+      "step": 523
+    },
+    {
+      "epoch": 0.12076515326112007,
+      "grad_norm": 0.3725113868713379,
+      "learning_rate": 9.786324075320414e-05,
+      "loss": 0.7536,
+      "step": 524
+    },
+    {
+      "epoch": 0.12099562111085503,
+      "grad_norm": 0.3641432225704193,
+      "learning_rate": 9.785243144367562e-05,
+      "loss": 0.7546,
+      "step": 525
+    },
+    {
+      "epoch": 0.12122608896058999,
+      "grad_norm": 0.3699088990688324,
+      "learning_rate": 9.784159546235037e-05,
+      "loss": 0.7544,
+      "step": 526
+    },
+    {
+      "epoch": 0.12145655681032495,
+      "grad_norm": 0.3829246461391449,
+      "learning_rate": 9.78307328152681e-05,
+      "loss": 0.7613,
+      "step": 527
+    },
+    {
+      "epoch": 0.12168702466005993,
+      "grad_norm": 0.42919662594795227,
+      "learning_rate": 9.78198435084834e-05,
+      "loss": 0.7546,
+      "step": 528
+    },
+    {
+      "epoch": 0.12191749250979489,
+      "grad_norm": 0.446468323469162,
+      "learning_rate": 9.780892754806571e-05,
+      "loss": 0.7496,
+      "step": 529
+    },
+    {
+      "epoch": 0.12214796035952985,
+      "grad_norm": 0.459417462348938,
+      "learning_rate": 9.779798494009931e-05,
+      "loss": 0.7559,
+      "step": 530
+    },
+    {
+      "epoch": 0.12237842820926481,
+      "grad_norm": 0.4683552086353302,
+      "learning_rate": 9.778701569068336e-05,
+      "loss": 0.7681,
+      "step": 531
+    },
+    {
+      "epoch": 0.12260889605899977,
+      "grad_norm": 0.38343241810798645,
+      "learning_rate": 9.777601980593184e-05,
+      "loss": 0.7649,
+      "step": 532
+    },
+    {
+      "epoch": 0.12283936390873473,
+      "grad_norm": 0.4223579466342926,
+      "learning_rate": 9.776499729197362e-05,
+      "loss": 0.75,
+      "step": 533
+    },
+    {
+      "epoch": 0.1230698317584697,
+      "grad_norm": 0.4097927510738373,
+      "learning_rate": 9.775394815495236e-05,
+      "loss": 0.7476,
+      "step": 534
+    },
+    {
+      "epoch": 0.12330029960820466,
+      "grad_norm": 0.4324532151222229,
+      "learning_rate": 9.77428724010266e-05,
+      "loss": 0.7568,
+      "step": 535
+    },
+    {
+      "epoch": 0.12353076745793962,
+      "grad_norm": 0.4441714584827423,
+      "learning_rate": 9.773177003636969e-05,
+      "loss": 0.7574,
+      "step": 536
+    },
+    {
+      "epoch": 0.12376123530767458,
+      "grad_norm": 0.49834537506103516,
+      "learning_rate": 9.77206410671698e-05,
+      "loss": 0.7559,
+      "step": 537
+    },
+    {
+      "epoch": 0.12399170315740954,
+      "grad_norm": 0.5015109181404114,
+      "learning_rate": 9.770948549963e-05,
+      "loss": 0.7566,
+      "step": 538
+    },
+    {
+      "epoch": 0.1242221710071445,
+      "grad_norm": 0.3993091881275177,
+      "learning_rate": 9.76983033399681e-05,
+      "loss": 0.7574,
+      "step": 539
+    },
+    {
+      "epoch": 0.12445263885687946,
+      "grad_norm": 0.37793558835983276,
+      "learning_rate": 9.768709459441676e-05,
+      "loss": 0.7433,
+      "step": 540
+    },
+    {
+      "epoch": 0.12468310670661442,
+      "grad_norm": 0.48581400513648987,
+      "learning_rate": 9.76758592692235e-05,
+      "loss": 0.7547,
+      "step": 541
+    },
+    {
+      "epoch": 0.12491357455634938,
+      "grad_norm": 0.5075335502624512,
+      "learning_rate": 9.76645973706506e-05,
+      "loss": 0.7625,
+      "step": 542
+    },
+    {
+      "epoch": 0.12514404240608434,
+      "grad_norm": 0.5207210779190063,
+      "learning_rate": 9.765330890497518e-05,
+      "loss": 0.7551,
+      "step": 543
+    },
+    {
+      "epoch": 0.1253745102558193,
+      "grad_norm": 0.42284563183784485,
+      "learning_rate": 9.764199387848915e-05,
+      "loss": 0.7599,
+      "step": 544
+    },
+    {
+      "epoch": 0.12560497810555427,
+      "grad_norm": 0.3669079542160034,
+      "learning_rate": 9.763065229749923e-05,
+      "loss": 0.7433,
+      "step": 545
+    },
+    {
+      "epoch": 0.12583544595528923,
+      "grad_norm": 0.3668546676635742,
+      "learning_rate": 9.761928416832695e-05,
+      "loss": 0.7589,
+      "step": 546
+    },
+    {
+      "epoch": 0.1260659138050242,
+      "grad_norm": 0.520101010799408,
+      "learning_rate": 9.760788949730866e-05,
+      "loss": 0.7503,
+      "step": 547
+    },
+    {
+      "epoch": 0.12629638165475915,
+      "grad_norm": 0.5019710659980774,
+      "learning_rate": 9.759646829079543e-05,
+      "loss": 0.7548,
+      "step": 548
+    },
+    {
+      "epoch": 0.1265268495044941,
+      "grad_norm": 0.4565122425556183,
+      "learning_rate": 9.75850205551532e-05,
+      "loss": 0.7479,
+      "step": 549
+    },
+    {
+      "epoch": 0.1267573173542291,
+      "grad_norm": 0.4054650664329529,
+      "learning_rate": 9.757354629676265e-05,
+      "loss": 0.7455,
+      "step": 550
+    },
+    {
+      "epoch": 0.12698778520396406,
+      "grad_norm": 0.5145794153213501,
+      "learning_rate": 9.756204552201926e-05,
+      "loss": 0.7514,
+      "step": 551
+    },
+    {
+      "epoch": 0.12721825305369902,
+      "grad_norm": 0.4342021644115448,
+      "learning_rate": 9.755051823733328e-05,
+      "loss": 0.7457,
+      "step": 552
+    },
+    {
+      "epoch": 0.12744872090343398,
+      "grad_norm": 0.3916120231151581,
+      "learning_rate": 9.753896444912973e-05,
+      "loss": 0.7557,
+      "step": 553
+    },
+    {
+      "epoch": 0.12767918875316894,
+      "grad_norm": 0.3801400363445282,
+      "learning_rate": 9.752738416384844e-05,
+      "loss": 0.7486,
+      "step": 554
+    },
+    {
+      "epoch": 0.1279096566029039,
+      "grad_norm": 0.4345585107803345,
+      "learning_rate": 9.751577738794398e-05,
+      "loss": 0.7561,
+      "step": 555
+    },
+    {
+      "epoch": 0.12814012445263886,
+      "grad_norm": 0.3579169809818268,
+      "learning_rate": 9.750414412788567e-05,
+      "loss": 0.7552,
+      "step": 556
+    },
+    {
+      "epoch": 0.12837059230237383,
+      "grad_norm": 0.3728508949279785,
+      "learning_rate": 9.749248439015763e-05,
+      "loss": 0.7605,
+      "step": 557
+    },
+    {
+      "epoch": 0.1286010601521088,
+      "grad_norm": 0.4141934812068939,
+      "learning_rate": 9.74807981812587e-05,
+      "loss": 0.7514,
+      "step": 558
+    },
+    {
+      "epoch": 0.12883152800184375,
+      "grad_norm": 0.34202003479003906,
+      "learning_rate": 9.746908550770252e-05,
+      "loss": 0.742,
+      "step": 559
+    },
+    {
+      "epoch": 0.1290619958515787,
+      "grad_norm": 0.39449524879455566,
+      "learning_rate": 9.745734637601743e-05,
+      "loss": 0.745,
+      "step": 560
+    },
+    {
+      "epoch": 0.12929246370131367,
+      "grad_norm": 0.3760475218296051,
+      "learning_rate": 9.744558079274652e-05,
+      "loss": 0.7586,
+      "step": 561
+    },
+    {
+      "epoch": 0.12952293155104863,
+      "grad_norm": 0.3431721329689026,
+      "learning_rate": 9.743378876444769e-05,
+      "loss": 0.7523,
+      "step": 562
+    },
+    {
+      "epoch": 0.1297533994007836,
+      "grad_norm": 0.3754710555076599,
+      "learning_rate": 9.74219702976935e-05,
+      "loss": 0.7374,
+      "step": 563
+    },
+    {
+      "epoch": 0.12998386725051855,
+      "grad_norm": 0.45317164063453674,
+      "learning_rate": 9.741012539907131e-05,
+      "loss": 0.7599,
+      "step": 564
+    },
+    {
+      "epoch": 0.13021433510025351,
+      "grad_norm": 0.4640779495239258,
+      "learning_rate": 9.739825407518314e-05,
+      "loss": 0.7408,
+      "step": 565
+    },
+    {
+      "epoch": 0.13044480294998848,
+      "grad_norm": 0.49257683753967285,
+      "learning_rate": 9.738635633264581e-05,
+      "loss": 0.752,
+      "step": 566
+    },
+    {
+      "epoch": 0.13067527079972344,
+      "grad_norm": 0.4536359906196594,
+      "learning_rate": 9.737443217809083e-05,
+      "loss": 0.7566,
+      "step": 567
+    },
+    {
+      "epoch": 0.1309057386494584,
+      "grad_norm": 0.4063389003276825,
+      "learning_rate": 9.736248161816446e-05,
+      "loss": 0.7556,
+      "step": 568
+    },
+    {
+      "epoch": 0.13113620649919336,
+      "grad_norm": 0.33868643641471863,
+      "learning_rate": 9.735050465952761e-05,
+      "loss": 0.751,
+      "step": 569
+    },
+    {
+      "epoch": 0.13136667434892832,
+      "grad_norm": 0.3786289095878601,
+      "learning_rate": 9.733850130885598e-05,
+      "loss": 0.7568,
+      "step": 570
+    },
+    {
+      "epoch": 0.13159714219866328,
+      "grad_norm": 0.43096908926963806,
+      "learning_rate": 9.732647157283994e-05,
+      "loss": 0.7508,
+      "step": 571
+    },
+    {
+      "epoch": 0.13182761004839824,
+      "grad_norm": 0.39179953932762146,
+      "learning_rate": 9.731441545818459e-05,
+      "loss": 0.749,
+      "step": 572
+    },
+    {
+      "epoch": 0.1320580778981332,
+      "grad_norm": 0.3335493803024292,
+      "learning_rate": 9.730233297160969e-05,
+      "loss": 0.7491,
+      "step": 573
+    },
+    {
+      "epoch": 0.13228854574786816,
+      "grad_norm": 0.4363969564437866,
+      "learning_rate": 9.729022411984975e-05,
+      "loss": 0.7496,
+      "step": 574
+    },
+    {
+      "epoch": 0.13251901359760312,
+      "grad_norm": 0.3922342360019684,
+      "learning_rate": 9.727808890965396e-05,
+      "loss": 0.7481,
+      "step": 575
+    },
+    {
+      "epoch": 0.13274948144733809,
+      "grad_norm": 0.3694200813770294,
+      "learning_rate": 9.726592734778617e-05,
+      "loss": 0.7561,
+      "step": 576
+    },
+    {
+      "epoch": 0.13297994929707305,
+      "grad_norm": 0.40044036507606506,
+      "learning_rate": 9.725373944102496e-05,
+      "loss": 0.7463,
+      "step": 577
+    },
+    {
+      "epoch": 0.133210417146808,
+      "grad_norm": 0.39088964462280273,
+      "learning_rate": 9.724152519616358e-05,
+      "loss": 0.7528,
+      "step": 578
+    },
+    {
+      "epoch": 0.13344088499654297,
+      "grad_norm": 0.35432836413383484,
+      "learning_rate": 9.722928462000995e-05,
+      "loss": 0.7393,
+      "step": 579
+    },
+    {
+      "epoch": 0.13367135284627796,
+      "grad_norm": 0.36177772283554077,
+      "learning_rate": 9.721701771938666e-05,
+      "loss": 0.7511,
+      "step": 580
+    },
+    {
+      "epoch": 0.13390182069601292,
+      "grad_norm": 0.3491220474243164,
+      "learning_rate": 9.7204724501131e-05,
+      "loss": 0.7516,
+      "step": 581
+    },
+    {
+      "epoch": 0.13413228854574788,
+      "grad_norm": 0.3720824420452118,
+      "learning_rate": 9.719240497209493e-05,
+      "loss": 0.7416,
+      "step": 582
+    },
+    {
+      "epoch": 0.13436275639548284,
+      "grad_norm": 0.3904234766960144,
+      "learning_rate": 9.718005913914503e-05,
+      "loss": 0.7429,
+      "step": 583
+    },
+    {
+      "epoch": 0.1345932242452178,
+      "grad_norm": 0.3584286868572235,
+      "learning_rate": 9.716768700916258e-05,
+      "loss": 0.7443,
+      "step": 584
+    },
+    {
+      "epoch": 0.13482369209495276,
+      "grad_norm": 0.36260986328125,
+      "learning_rate": 9.715528858904353e-05,
+      "loss": 0.7491,
+      "step": 585
+    },
+    {
+      "epoch": 0.13505415994468772,
+      "grad_norm": 0.41411370038986206,
+      "learning_rate": 9.714286388569845e-05,
+      "loss": 0.7493,
+      "step": 586
+    },
+    {
+      "epoch": 0.13528462779442268,
+      "grad_norm": 0.4128418266773224,
+      "learning_rate": 9.713041290605254e-05,
+      "loss": 0.7531,
+      "step": 587
+    },
+    {
+      "epoch": 0.13551509564415765,
+      "grad_norm": 0.37664923071861267,
+      "learning_rate": 9.711793565704572e-05,
+      "loss": 0.7491,
+      "step": 588
+    },
+    {
+      "epoch": 0.1357455634938926,
+      "grad_norm": 0.3493359386920929,
+      "learning_rate": 9.71054321456325e-05,
+      "loss": 0.743,
+      "step": 589
+    },
+    {
+      "epoch": 0.13597603134362757,
+      "grad_norm": 0.4216603934764862,
+      "learning_rate": 9.709290237878202e-05,
+      "loss": 0.7446,
+      "step": 590
+    },
+    {
+      "epoch": 0.13620649919336253,
+      "grad_norm": 0.4375728964805603,
+      "learning_rate": 9.708034636347807e-05,
+      "loss": 0.7533,
+      "step": 591
+    },
+    {
+      "epoch": 0.1364369670430975,
+      "grad_norm": 0.3984435796737671,
+      "learning_rate": 9.70677641067191e-05,
+      "loss": 0.7379,
+      "step": 592
+    },
+    {
+      "epoch": 0.13666743489283245,
+      "grad_norm": 0.3685891330242157,
+      "learning_rate": 9.705515561551814e-05,
+      "loss": 0.7462,
+      "step": 593
+    },
+    {
+      "epoch": 0.1368979027425674,
+      "grad_norm": 0.35260993242263794,
+      "learning_rate": 9.704252089690284e-05,
+      "loss": 0.7491,
+      "step": 594
+    },
+    {
+      "epoch": 0.13712837059230237,
+      "grad_norm": 0.33761847019195557,
+      "learning_rate": 9.702985995791554e-05,
+      "loss": 0.7403,
+      "step": 595
+    },
+    {
+      "epoch": 0.13735883844203733,
+      "grad_norm": 0.4064388573169708,
+      "learning_rate": 9.701717280561309e-05,
+      "loss": 0.7511,
+      "step": 596
+    },
+    {
+      "epoch": 0.1375893062917723,
+      "grad_norm": 0.4629051089286804,
+      "learning_rate": 9.700445944706704e-05,
+      "loss": 0.746,
+      "step": 597
+    },
+    {
+      "epoch": 0.13781977414150726,
+      "grad_norm": 0.45266193151474,
+      "learning_rate": 9.699171988936349e-05,
+      "loss": 0.7394,
+      "step": 598
+    },
+    {
+      "epoch": 0.13805024199124222,
+      "grad_norm": 0.3218851387500763,
+      "learning_rate": 9.697895413960319e-05,
+      "loss": 0.7477,
+      "step": 599
+    },
+    {
+      "epoch": 0.13828070984097718,
+      "grad_norm": 0.4335274398326874,
+      "learning_rate": 9.696616220490143e-05,
+      "loss": 0.7356,
+      "step": 600
+    },
+    {
+      "epoch": 0.13851117769071214,
+      "grad_norm": 0.45627596974372864,
+      "learning_rate": 9.695334409238813e-05,
+      "loss": 0.746,
+      "step": 601
+    },
+    {
+      "epoch": 0.1387416455404471,
+      "grad_norm": 0.35524532198905945,
+      "learning_rate": 9.694049980920783e-05,
+      "loss": 0.7494,
+      "step": 602
+    },
+    {
+      "epoch": 0.13897211339018206,
+      "grad_norm": 0.40399858355522156,
+      "learning_rate": 9.69276293625196e-05,
+      "loss": 0.7468,
+      "step": 603
+    },
+    {
+      "epoch": 0.13920258123991702,
+      "grad_norm": 0.4082375168800354,
+      "learning_rate": 9.691473275949712e-05,
+      "loss": 0.7518,
+      "step": 604
+    },
+    {
+      "epoch": 0.13943304908965198,
+      "grad_norm": 0.3440902829170227,
+      "learning_rate": 9.690181000732864e-05,
+      "loss": 0.7421,
+      "step": 605
+    },
+    {
+      "epoch": 0.13966351693938694,
+      "grad_norm": 0.481870174407959,
+      "learning_rate": 9.688886111321703e-05,
+      "loss": 0.7558,
+      "step": 606
+    },
+    {
+      "epoch": 0.1398939847891219,
+      "grad_norm": 0.3942829370498657,
+      "learning_rate": 9.687588608437963e-05,
+      "loss": 0.7387,
+      "step": 607
+    },
+    {
+      "epoch": 0.14012445263885687,
+      "grad_norm": 0.3642619550228119,
+      "learning_rate": 9.686288492804846e-05,
+      "loss": 0.7369,
+      "step": 608
+    },
+    {
+      "epoch": 0.14035492048859183,
+      "grad_norm": 0.4085538685321808,
+      "learning_rate": 9.684985765147006e-05,
+      "loss": 0.7455,
+      "step": 609
+    },
+    {
+      "epoch": 0.14058538833832682,
+      "grad_norm": 0.36684495210647583,
+      "learning_rate": 9.683680426190547e-05,
+      "loss": 0.7427,
+      "step": 610
+    },
+    {
+      "epoch": 0.14081585618806178,
+      "grad_norm": 0.3928658068180084,
+      "learning_rate": 9.682372476663037e-05,
+      "loss": 0.7503,
+      "step": 611
+    },
+    {
+      "epoch": 0.14104632403779674,
+      "grad_norm": 0.40327298641204834,
+      "learning_rate": 9.681061917293497e-05,
+      "loss": 0.7497,
+      "step": 612
+    },
+    {
+      "epoch": 0.1412767918875317,
+      "grad_norm": 0.45683008432388306,
+      "learning_rate": 9.679748748812397e-05,
+      "loss": 0.7443,
+      "step": 613
+    },
+    {
+      "epoch": 0.14150725973726666,
+      "grad_norm": 0.3893783986568451,
+      "learning_rate": 9.678432971951669e-05,
+      "loss": 0.7487,
+      "step": 614
+    },
+    {
+      "epoch": 0.14173772758700162,
+      "grad_norm": 0.36293652653694153,
+      "learning_rate": 9.677114587444696e-05,
+      "loss": 0.7477,
+      "step": 615
+    },
+    {
+      "epoch": 0.14196819543673658,
+      "grad_norm": 0.3704957365989685,
+      "learning_rate": 9.67579359602631e-05,
+      "loss": 0.7442,
+      "step": 616
+    },
+    {
+      "epoch": 0.14219866328647154,
+      "grad_norm": 0.3900201618671417,
+      "learning_rate": 9.674469998432802e-05,
+      "loss": 0.7408,
+      "step": 617
+    },
+    {
+      "epoch": 0.1424291311362065,
+      "grad_norm": 0.35476595163345337,
+      "learning_rate": 9.673143795401915e-05,
+      "loss": 0.7369,
+      "step": 618
+    },
+    {
+      "epoch": 0.14265959898594147,
+      "grad_norm": 0.39594173431396484,
+      "learning_rate": 9.671814987672842e-05,
+      "loss": 0.7495,
+      "step": 619
+    },
+    {
+      "epoch": 0.14289006683567643,
+      "grad_norm": 0.3954264521598816,
+      "learning_rate": 9.670483575986229e-05,
+      "loss": 0.7536,
+      "step": 620
+    },
+    {
+      "epoch": 0.1431205346854114,
+      "grad_norm": 0.329731285572052,
+      "learning_rate": 9.66914956108417e-05,
+      "loss": 0.7432,
+      "step": 621
+    },
+    {
+      "epoch": 0.14335100253514635,
+      "grad_norm": 0.33711501955986023,
+      "learning_rate": 9.667812943710215e-05,
+      "loss": 0.733,
+      "step": 622
+    },
+    {
+      "epoch": 0.1435814703848813,
+      "grad_norm": 0.3300781548023224,
+      "learning_rate": 9.666473724609364e-05,
+      "loss": 0.7366,
+      "step": 623
+    },
+    {
+      "epoch": 0.14381193823461627,
+      "grad_norm": 0.32129067182540894,
+      "learning_rate": 9.665131904528063e-05,
+      "loss": 0.7457,
+      "step": 624
+    },
+    {
+      "epoch": 0.14404240608435123,
+      "grad_norm": 0.3335452973842621,
+      "learning_rate": 9.66378748421421e-05,
+      "loss": 0.7453,
+      "step": 625
+    },
+    {
+      "epoch": 0.1442728739340862,
+      "grad_norm": 0.31821250915527344,
+      "learning_rate": 9.662440464417155e-05,
+      "loss": 0.7381,
+      "step": 626
+    },
+    {
+      "epoch": 0.14450334178382115,
+      "grad_norm": 0.6273105144500732,
+      "learning_rate": 9.661090845887693e-05,
+      "loss": 0.7383,
+      "step": 627
+    },
+    {
+      "epoch": 0.14473380963355612,
+      "grad_norm": 0.32164305448532104,
+      "learning_rate": 9.65973862937807e-05,
+      "loss": 0.7398,
+      "step": 628
+    },
+    {
+      "epoch": 0.14496427748329108,
+      "grad_norm": 0.3327626585960388,
+      "learning_rate": 9.658383815641978e-05,
+      "loss": 0.7357,
+      "step": 629
+    },
+    {
+      "epoch": 0.14519474533302604,
+      "grad_norm": 0.30448755621910095,
+      "learning_rate": 9.657026405434557e-05,
+      "loss": 0.7309,
+      "step": 630
+    },
+    {
+      "epoch": 0.145425213182761,
+      "grad_norm": 0.30915525555610657,
+      "learning_rate": 9.655666399512397e-05,
+      "loss": 0.7407,
+      "step": 631
+    },
+    {
+      "epoch": 0.14565568103249596,
+      "grad_norm": 0.3181747794151306,
+      "learning_rate": 9.654303798633532e-05,
+      "loss": 0.7443,
+      "step": 632
+    },
+    {
+      "epoch": 0.14588614888223092,
+      "grad_norm": 0.3349412977695465,
+      "learning_rate": 9.652938603557442e-05,
+      "loss": 0.7459,
+      "step": 633
+    },
+    {
+      "epoch": 0.14611661673196588,
+      "grad_norm": 0.32056349515914917,
+      "learning_rate": 9.651570815045054e-05,
+      "loss": 0.7415,
+      "step": 634
+    },
+    {
+      "epoch": 0.14634708458170084,
+      "grad_norm": 0.34287700057029724,
+      "learning_rate": 9.650200433858741e-05,
+      "loss": 0.7412,
+      "step": 635
+    },
+    {
+      "epoch": 0.1465775524314358,
+      "grad_norm": 0.3507481813430786,
+      "learning_rate": 9.648827460762322e-05,
+      "loss": 0.7487,
+      "step": 636
+    },
+    {
+      "epoch": 0.14680802028117076,
+      "grad_norm": 0.3356325328350067,
+      "learning_rate": 9.647451896521055e-05,
+      "loss": 0.7423,
+      "step": 637
+    },
+    {
+      "epoch": 0.14703848813090573,
+      "grad_norm": 0.37254631519317627,
+      "learning_rate": 9.646073741901652e-05,
+      "loss": 0.743,
+      "step": 638
+    },
+    {
+      "epoch": 0.14726895598064071,
+      "grad_norm": 0.3373221755027771,
+      "learning_rate": 9.64469299767226e-05,
+      "loss": 0.7406,
+      "step": 639
+    },
+    {
+      "epoch": 0.14749942383037568,
+      "grad_norm": 0.3344241976737976,
+      "learning_rate": 9.643309664602474e-05,
+      "loss": 0.7377,
+      "step": 640
+    },
+    {
+      "epoch": 0.14772989168011064,
+      "grad_norm": 0.34580978751182556,
+      "learning_rate": 9.641923743463327e-05,
+      "loss": 0.7454,
+      "step": 641
+    },
+    {
+      "epoch": 0.1479603595298456,
+      "grad_norm": 0.31924277544021606,
+      "learning_rate": 9.640535235027303e-05,
+      "loss": 0.7369,
+      "step": 642
+    },
+    {
+      "epoch": 0.14819082737958056,
+      "grad_norm": 0.33480551838874817,
+      "learning_rate": 9.639144140068324e-05,
+      "loss": 0.7582,
+      "step": 643
+    },
+    {
+      "epoch": 0.14842129522931552,
+      "grad_norm": 0.3792799413204193,
+      "learning_rate": 9.637750459361748e-05,
+      "loss": 0.7495,
+      "step": 644
+    },
+    {
+      "epoch": 0.14865176307905048,
+      "grad_norm": 0.42829740047454834,
+      "learning_rate": 9.63635419368438e-05,
+      "loss": 0.7406,
+      "step": 645
+    },
+    {
+      "epoch": 0.14888223092878544,
+      "grad_norm": 0.48589882254600525,
+      "learning_rate": 9.634955343814469e-05,
+      "loss": 0.7468,
+      "step": 646
+    },
+    {
+      "epoch": 0.1491126987785204,
+      "grad_norm": 0.43215352296829224,
+      "learning_rate": 9.633553910531697e-05,
+      "loss": 0.7444,
+      "step": 647
+    },
+    {
+      "epoch": 0.14934316662825536,
+      "grad_norm": 0.38566887378692627,
+      "learning_rate": 9.632149894617191e-05,
+      "loss": 0.7447,
+      "step": 648
+    },
+    {
+      "epoch": 0.14957363447799032,
+      "grad_norm": 0.31605324149131775,
+      "learning_rate": 9.630743296853513e-05,
+      "loss": 0.7462,
+      "step": 649
+    },
+    {
+      "epoch": 0.14980410232772529,
+      "grad_norm": 0.3389308452606201,
+      "learning_rate": 9.629334118024669e-05,
+      "loss": 0.7345,
+      "step": 650
+    },
+    {
+      "epoch": 0.15003457017746025,
+      "grad_norm": 0.3665103614330292,
+      "learning_rate": 9.627922358916102e-05,
+      "loss": 0.7393,
+      "step": 651
+    },
+    {
+      "epoch": 0.1502650380271952,
+      "grad_norm": 0.33083248138427734,
+      "learning_rate": 9.626508020314693e-05,
+      "loss": 0.7439,
+      "step": 652
+    },
+    {
+      "epoch": 0.15049550587693017,
+      "grad_norm": 0.3108620345592499,
+      "learning_rate": 9.625091103008757e-05,
+      "loss": 0.7347,
+      "step": 653
+    },
+    {
+      "epoch": 0.15072597372666513,
+      "grad_norm": 0.32993149757385254,
+      "learning_rate": 9.623671607788054e-05,
+      "loss": 0.7429,
+      "step": 654
+    },
+    {
+      "epoch": 0.1509564415764001,
+      "grad_norm": 0.35567355155944824,
+      "learning_rate": 9.622249535443773e-05,
+      "loss": 0.7365,
+      "step": 655
+    },
+    {
+      "epoch": 0.15118690942613505,
+      "grad_norm": 0.3489120602607727,
+      "learning_rate": 9.620824886768545e-05,
+      "loss": 0.7314,
+      "step": 656
+    },
+    {
+      "epoch": 0.15141737727587,
+      "grad_norm": 0.34390199184417725,
+      "learning_rate": 9.619397662556435e-05,
+      "loss": 0.7468,
+      "step": 657
+    },
+    {
+      "epoch": 0.15164784512560497,
+      "grad_norm": 0.3184950351715088,
+      "learning_rate": 9.617967863602941e-05,
+      "loss": 0.7474,
+      "step": 658
+    },
+    {
+      "epoch": 0.15187831297533994,
+      "grad_norm": 0.3136204779148102,
+      "learning_rate": 9.616535490705004e-05,
+      "loss": 0.7423,
+      "step": 659
+    },
+    {
+      "epoch": 0.1521087808250749,
+      "grad_norm": 0.29085636138916016,
+      "learning_rate": 9.615100544660987e-05,
+      "loss": 0.7374,
+      "step": 660
+    },
+    {
+      "epoch": 0.15233924867480986,
+      "grad_norm": 0.35054919123649597,
+      "learning_rate": 9.6136630262707e-05,
+      "loss": 0.7436,
+      "step": 661
+    },
+    {
+      "epoch": 0.15256971652454482,
+      "grad_norm": 0.32511407136917114,
+      "learning_rate": 9.61222293633538e-05,
+      "loss": 0.7401,
+      "step": 662
+    },
+    {
+      "epoch": 0.15280018437427978,
+      "grad_norm": 0.3348987400531769,
+      "learning_rate": 9.610780275657698e-05,
+      "loss": 0.7387,
+      "step": 663
+    },
+    {
+      "epoch": 0.15303065222401474,
+      "grad_norm": 0.3278294801712036,
+      "learning_rate": 9.609335045041759e-05,
+      "loss": 0.7402,
+      "step": 664
+    },
+    {
+      "epoch": 0.1532611200737497,
+      "grad_norm": 0.3211444020271301,
+      "learning_rate": 9.6078872452931e-05,
+      "loss": 0.7448,
+      "step": 665
+    },
+    {
+      "epoch": 0.15349158792348466,
+      "grad_norm": 0.341467946767807,
+      "learning_rate": 9.606436877218688e-05,
+      "loss": 0.7395,
+      "step": 666
+    },
+    {
+      "epoch": 0.15372205577321962,
+      "grad_norm": 0.3915879726409912,
+      "learning_rate": 9.604983941626924e-05,
+      "loss": 0.7405,
+      "step": 667
+    },
+    {
+      "epoch": 0.15395252362295458,
+      "grad_norm": 0.41400331258773804,
+      "learning_rate": 9.603528439327642e-05,
+      "loss": 0.7456,
+      "step": 668
+    },
+    {
+      "epoch": 0.15418299147268957,
+      "grad_norm": 0.35899117588996887,
+      "learning_rate": 9.602070371132102e-05,
+      "loss": 0.7282,
+      "step": 669
+    },
+    {
+      "epoch": 0.15441345932242453,
+      "grad_norm": 0.273082971572876,
+      "learning_rate": 9.600609737852995e-05,
+      "loss": 0.7323,
+      "step": 670
+    },
+    {
+      "epoch": 0.1546439271721595,
+      "grad_norm": 0.34714290499687195,
+      "learning_rate": 9.599146540304444e-05,
+      "loss": 0.7426,
+      "step": 671
+    },
+    {
+      "epoch": 0.15487439502189446,
+      "grad_norm": 0.3830714523792267,
+      "learning_rate": 9.597680779302e-05,
+      "loss": 0.7362,
+      "step": 672
+    },
+    {
+      "epoch": 0.15510486287162942,
+      "grad_norm": 0.30310484766960144,
+      "learning_rate": 9.596212455662645e-05,
+      "loss": 0.7368,
+      "step": 673
+    },
+    {
+      "epoch": 0.15533533072136438,
+      "grad_norm": 0.3948887586593628,
+      "learning_rate": 9.594741570204787e-05,
+      "loss": 0.7377,
+      "step": 674
+    },
+    {
+      "epoch": 0.15556579857109934,
+      "grad_norm": 0.42148318886756897,
+      "learning_rate": 9.593268123748259e-05,
+      "loss": 0.7415,
+      "step": 675
+    },
+    {
+      "epoch": 0.1557962664208343,
+      "grad_norm": 0.5412166118621826,
+      "learning_rate": 9.591792117114328e-05,
+      "loss": 0.7277,
+      "step": 676
+    },
+    {
+      "epoch": 0.15602673427056926,
+      "grad_norm": 0.6122573614120483,
+      "learning_rate": 9.590313551125683e-05,
+      "loss": 0.7376,
+      "step": 677
+    },
+    {
+      "epoch": 0.15625720212030422,
+      "grad_norm": 0.4985668361186981,
+      "learning_rate": 9.58883242660644e-05,
+      "loss": 0.742,
+      "step": 678
+    },
+    {
+      "epoch": 0.15648766997003918,
+      "grad_norm": 0.356157124042511,
+      "learning_rate": 9.587348744382145e-05,
+      "loss": 0.73,
+      "step": 679
+    },
+    {
+      "epoch": 0.15671813781977414,
+      "grad_norm": 0.3795771300792694,
+      "learning_rate": 9.585862505279766e-05,
+      "loss": 0.735,
+      "step": 680
+    },
+    {
+      "epoch": 0.1569486056695091,
+      "grad_norm": 0.4135006070137024,
+      "learning_rate": 9.584373710127697e-05,
+      "loss": 0.7361,
+      "step": 681
+    },
+    {
+      "epoch": 0.15717907351924407,
+      "grad_norm": 0.3758498728275299,
+      "learning_rate": 9.582882359755756e-05,
+      "loss": 0.7335,
+      "step": 682
+    },
+    {
+      "epoch": 0.15740954136897903,
+      "grad_norm": 0.35648760199546814,
+      "learning_rate": 9.581388454995187e-05,
+      "loss": 0.7387,
+      "step": 683
+    },
+    {
+      "epoch": 0.157640009218714,
+      "grad_norm": 0.3326440453529358,
+      "learning_rate": 9.579891996678655e-05,
+      "loss": 0.7326,
+      "step": 684
+    },
+    {
+      "epoch": 0.15787047706844895,
+      "grad_norm": 0.3023831248283386,
+      "learning_rate": 9.578392985640252e-05,
+      "loss": 0.7302,
+      "step": 685
+    },
+    {
+      "epoch": 0.1581009449181839,
+      "grad_norm": 0.3539700508117676,
+      "learning_rate": 9.576891422715489e-05,
+      "loss": 0.729,
+      "step": 686
+    },
+    {
+      "epoch": 0.15833141276791887,
+      "grad_norm": 0.38474544882774353,
+      "learning_rate": 9.575387308741301e-05,
+      "loss": 0.7299,
+      "step": 687
+    },
+    {
+      "epoch": 0.15856188061765383,
+      "grad_norm": 0.3182377517223358,
+      "learning_rate": 9.573880644556047e-05,
+      "loss": 0.7424,
+      "step": 688
+    },
+    {
+      "epoch": 0.1587923484673888,
+      "grad_norm": 0.3241504430770874,
+      "learning_rate": 9.572371430999506e-05,
+      "loss": 0.7363,
+      "step": 689
+    },
+    {
+      "epoch": 0.15902281631712376,
+      "grad_norm": 0.3680818974971771,
+      "learning_rate": 9.570859668912873e-05,
+      "loss": 0.7336,
+      "step": 690
+    },
+    {
+      "epoch": 0.15925328416685872,
+      "grad_norm": 0.3773689270019531,
+      "learning_rate": 9.569345359138771e-05,
+      "loss": 0.7435,
+      "step": 691
+    },
+    {
+      "epoch": 0.15948375201659368,
+      "grad_norm": 0.34049540758132935,
+      "learning_rate": 9.567828502521239e-05,
+      "loss": 0.7327,
+      "step": 692
+    },
+    {
+      "epoch": 0.15971421986632864,
+      "grad_norm": 0.34470391273498535,
+      "learning_rate": 9.566309099905739e-05,
+      "loss": 0.7308,
+      "step": 693
+    },
+    {
+      "epoch": 0.1599446877160636,
+      "grad_norm": 0.3475648760795593,
+      "learning_rate": 9.564787152139146e-05,
+      "loss": 0.7393,
+      "step": 694
+    },
+    {
+      "epoch": 0.16017515556579856,
+      "grad_norm": 0.304973840713501,
+      "learning_rate": 9.563262660069759e-05,
+      "loss": 0.7398,
+      "step": 695
+    },
+    {
+      "epoch": 0.16040562341553352,
+      "grad_norm": 0.33955731987953186,
+      "learning_rate": 9.561735624547294e-05,
+      "loss": 0.7388,
+      "step": 696
+    },
+    {
+      "epoch": 0.16063609126526848,
+      "grad_norm": 0.31284675002098083,
+      "learning_rate": 9.560206046422881e-05,
+      "loss": 0.7468,
+      "step": 697
+    },
+    {
+      "epoch": 0.16086655911500344,
+      "grad_norm": 0.3111793100833893,
+      "learning_rate": 9.558673926549075e-05,
+      "loss": 0.7445,
+      "step": 698
+    },
+    {
+      "epoch": 0.16109702696473843,
+      "grad_norm": 0.32062941789627075,
+      "learning_rate": 9.557139265779838e-05,
+      "loss": 0.7373,
+      "step": 699
+    },
+    {
+      "epoch": 0.1613274948144734,
+      "grad_norm": 0.32092171907424927,
+      "learning_rate": 9.555602064970554e-05,
+      "loss": 0.7315,
+      "step": 700
+    },
+    {
+      "epoch": 0.16155796266420835,
+      "grad_norm": 0.29449769854545593,
+      "learning_rate": 9.554062324978025e-05,
+      "loss": 0.7327,
+      "step": 701
+    },
+    {
+      "epoch": 0.16178843051394332,
+      "grad_norm": 0.339097261428833,
+      "learning_rate": 9.552520046660462e-05,
+      "loss": 0.7433,
+      "step": 702
+    },
+    {
+      "epoch": 0.16201889836367828,
+      "grad_norm": 0.3251134157180786,
+      "learning_rate": 9.550975230877495e-05,
+      "loss": 0.742,
+      "step": 703
+    },
+    {
+      "epoch": 0.16224936621341324,
+      "grad_norm": 0.2941744923591614,
+      "learning_rate": 9.549427878490168e-05,
+      "loss": 0.7321,
+      "step": 704
+    },
+    {
+      "epoch": 0.1624798340631482,
+      "grad_norm": 0.31333979964256287,
+      "learning_rate": 9.547877990360935e-05,
+      "loss": 0.7374,
+      "step": 705
+    },
+    {
+      "epoch": 0.16271030191288316,
+      "grad_norm": 0.3424711525440216,
+      "learning_rate": 9.546325567353671e-05,
+      "loss": 0.7381,
+      "step": 706
+    },
+    {
+      "epoch": 0.16294076976261812,
+      "grad_norm": 0.37201154232025146,
+      "learning_rate": 9.544770610333655e-05,
+      "loss": 0.7333,
+      "step": 707
+    },
+    {
+      "epoch": 0.16317123761235308,
+      "grad_norm": 0.29893067479133606,
+      "learning_rate": 9.543213120167586e-05,
+      "loss": 0.7397,
+      "step": 708
+    },
+    {
+      "epoch": 0.16340170546208804,
+      "grad_norm": 0.2980719804763794,
+      "learning_rate": 9.54165309772357e-05,
+      "loss": 0.7293,
+      "step": 709
+    },
+    {
+      "epoch": 0.163632173311823,
+      "grad_norm": 0.3534228503704071,
+      "learning_rate": 9.540090543871126e-05,
+      "loss": 0.7227,
+      "step": 710
+    },
+    {
+      "epoch": 0.16386264116155796,
+      "grad_norm": 0.42632773518562317,
+      "learning_rate": 9.538525459481185e-05,
+      "loss": 0.7304,
+      "step": 711
+    },
+    {
+      "epoch": 0.16409310901129293,
+      "grad_norm": 0.3898342549800873,
+      "learning_rate": 9.536957845426086e-05,
+      "loss": 0.7268,
+      "step": 712
+    },
+    {
+      "epoch": 0.1643235768610279,
+      "grad_norm": 0.290547251701355,
+      "learning_rate": 9.535387702579581e-05,
+      "loss": 0.7349,
+      "step": 713
+    },
+    {
+      "epoch": 0.16455404471076285,
+      "grad_norm": 0.2991439402103424,
+      "learning_rate": 9.53381503181683e-05,
+      "loss": 0.7406,
+      "step": 714
+    },
+    {
+      "epoch": 0.1647845125604978,
+      "grad_norm": 0.30085012316703796,
+      "learning_rate": 9.5322398340144e-05,
+      "loss": 0.7353,
+      "step": 715
+    },
+    {
+      "epoch": 0.16501498041023277,
+      "grad_norm": 0.39195868372917175,
+      "learning_rate": 9.53066211005027e-05,
+      "loss": 0.7314,
+      "step": 716
+    },
+    {
+      "epoch": 0.16524544825996773,
+      "grad_norm": 0.3485901951789856,
+      "learning_rate": 9.529081860803825e-05,
+      "loss": 0.7291,
+      "step": 717
+    },
+    {
+      "epoch": 0.1654759161097027,
+      "grad_norm": 0.27913734316825867,
+      "learning_rate": 9.527499087155857e-05,
+      "loss": 0.7291,
+      "step": 718
+    },
+    {
+      "epoch": 0.16570638395943765,
+      "grad_norm": 0.3681480288505554,
+      "learning_rate": 9.52591378998857e-05,
+      "loss": 0.7409,
+      "step": 719
+    },
+    {
+      "epoch": 0.16593685180917261,
+      "grad_norm": 0.37066853046417236,
+      "learning_rate": 9.524325970185565e-05,
+      "loss": 0.7291,
+      "step": 720
+    },
+    {
+      "epoch": 0.16616731965890758,
+      "grad_norm": 0.2994784712791443,
+      "learning_rate": 9.52273562863186e-05,
+      "loss": 0.7274,
+      "step": 721
+    },
+    {
+      "epoch": 0.16639778750864254,
+      "grad_norm": 0.3262505531311035,
+      "learning_rate": 9.521142766213869e-05,
+      "loss": 0.7387,
+      "step": 722
+    },
+    {
+      "epoch": 0.1666282553583775,
+      "grad_norm": 0.37113088369369507,
+      "learning_rate": 9.519547383819416e-05,
+      "loss": 0.7303,
+      "step": 723
+    },
+    {
+      "epoch": 0.16685872320811246,
+      "grad_norm": 0.35534918308258057,
+      "learning_rate": 9.517949482337732e-05,
+      "loss": 0.7383,
+      "step": 724
+    },
+    {
+      "epoch": 0.16708919105784742,
+      "grad_norm": 0.3642905056476593,
+      "learning_rate": 9.516349062659444e-05,
+      "loss": 0.7299,
+      "step": 725
+    },
+    {
+      "epoch": 0.16731965890758238,
+      "grad_norm": 0.3798482418060303,
+      "learning_rate": 9.514746125676593e-05,
+      "loss": 0.7279,
+      "step": 726
+    },
+    {
+      "epoch": 0.16755012675731734,
+      "grad_norm": 0.38027462363243103,
+      "learning_rate": 9.513140672282612e-05,
+      "loss": 0.7324,
+      "step": 727
+    },
+    {
+      "epoch": 0.1677805946070523,
+      "grad_norm": 0.34657275676727295,
+      "learning_rate": 9.511532703372348e-05,
+      "loss": 0.7394,
+      "step": 728
+    },
+    {
+      "epoch": 0.1680110624567873,
+      "grad_norm": 0.35259348154067993,
+      "learning_rate": 9.50992221984204e-05,
+      "loss": 0.7282,
+      "step": 729
+    },
+    {
+      "epoch": 0.16824153030652225,
+      "grad_norm": 0.3222431242465973,
+      "learning_rate": 9.508309222589333e-05,
+      "loss": 0.7242,
+      "step": 730
+    },
+    {
+      "epoch": 0.1684719981562572,
+      "grad_norm": 0.32721391320228577,
+      "learning_rate": 9.506693712513274e-05,
+      "loss": 0.7302,
+      "step": 731
+    },
+    {
+      "epoch": 0.16870246600599217,
+      "grad_norm": 0.3368002474308014,
+      "learning_rate": 9.505075690514312e-05,
+      "loss": 0.7265,
+      "step": 732
+    },
+    {
+      "epoch": 0.16893293385572714,
+      "grad_norm": 0.28599828481674194,
+      "learning_rate": 9.503455157494289e-05,
+      "loss": 0.7267,
+      "step": 733
+    },
+    {
+      "epoch": 0.1691634017054621,
+      "grad_norm": 0.30026113986968994,
+      "learning_rate": 9.501832114356453e-05,
+      "loss": 0.7375,
+      "step": 734
+    },
+    {
+      "epoch": 0.16939386955519706,
+      "grad_norm": 0.3616209030151367,
+      "learning_rate": 9.500206562005451e-05,
+      "loss": 0.736,
+      "step": 735
+    },
+    {
+      "epoch": 0.16962433740493202,
+      "grad_norm": 0.34600475430488586,
+      "learning_rate": 9.498578501347327e-05,
+      "loss": 0.7412,
+      "step": 736
+    },
+    {
+      "epoch": 0.16985480525466698,
+      "grad_norm": 0.33249667286872864,
+      "learning_rate": 9.49694793328952e-05,
+      "loss": 0.7355,
+      "step": 737
+    },
+    {
+      "epoch": 0.17008527310440194,
+      "grad_norm": 0.31944364309310913,
+      "learning_rate": 9.495314858740869e-05,
+      "loss": 0.7263,
+      "step": 738
+    },
+    {
+      "epoch": 0.1703157409541369,
+      "grad_norm": 0.29929810762405396,
+      "learning_rate": 9.493679278611616e-05,
+      "loss": 0.7244,
+      "step": 739
+    },
+    {
+      "epoch": 0.17054620880387186,
+      "grad_norm": 0.3145100474357605,
+      "learning_rate": 9.492041193813388e-05,
+      "loss": 0.7316,
+      "step": 740
+    },
+    {
+      "epoch": 0.17077667665360682,
+      "grad_norm": 0.34142231941223145,
+      "learning_rate": 9.490400605259218e-05,
+      "loss": 0.7281,
+      "step": 741
+    },
+    {
+      "epoch": 0.17100714450334178,
+      "grad_norm": 0.39629876613616943,
+      "learning_rate": 9.488757513863531e-05,
+      "loss": 0.7353,
+      "step": 742
+    },
+    {
+      "epoch": 0.17123761235307675,
+      "grad_norm": 0.43608033657073975,
+      "learning_rate": 9.487111920542143e-05,
+      "loss": 0.7323,
+      "step": 743
+    },
+    {
+      "epoch": 0.1714680802028117,
+      "grad_norm": 0.4607144296169281,
+      "learning_rate": 9.48546382621227e-05,
+      "loss": 0.7307,
+      "step": 744
+    },
+    {
+      "epoch": 0.17169854805254667,
+      "grad_norm": 0.4801574945449829,
+      "learning_rate": 9.483813231792523e-05,
+      "loss": 0.7299,
+      "step": 745
+    },
+    {
+      "epoch": 0.17192901590228163,
+      "grad_norm": 0.4828302264213562,
+      "learning_rate": 9.4821601382029e-05,
+      "loss": 0.734,
+      "step": 746
+    },
+    {
+      "epoch": 0.1721594837520166,
+      "grad_norm": 0.35751068592071533,
+      "learning_rate": 9.4805045463648e-05,
+      "loss": 0.7269,
+      "step": 747
+    },
+    {
+      "epoch": 0.17238995160175155,
+      "grad_norm": 0.35991883277893066,
+      "learning_rate": 9.478846457201003e-05,
+      "loss": 0.7308,
+      "step": 748
+    },
+    {
+      "epoch": 0.1726204194514865,
+      "grad_norm": 0.38963592052459717,
+      "learning_rate": 9.477185871635694e-05,
+      "loss": 0.7394,
+      "step": 749
+    },
+    {
+      "epoch": 0.17285088730122147,
+      "grad_norm": 0.3973761796951294,
+      "learning_rate": 9.475522790594443e-05,
+      "loss": 0.7279,
+      "step": 750
+    },
+    {
+      "epoch": 0.17308135515095643,
+      "grad_norm": 0.35446715354919434,
+      "learning_rate": 9.473857215004208e-05,
+      "loss": 0.7328,
+      "step": 751
+    },
+    {
+      "epoch": 0.1733118230006914,
+      "grad_norm": 0.32000207901000977,
+      "learning_rate": 9.472189145793345e-05,
+      "loss": 0.7304,
+      "step": 752
+    },
+    {
+      "epoch": 0.17354229085042636,
+      "grad_norm": 0.3462921679019928,
+      "learning_rate": 9.470518583891592e-05,
+      "loss": 0.7266,
+      "step": 753
+    },
+    {
+      "epoch": 0.17377275870016132,
+      "grad_norm": 0.37639400362968445,
+      "learning_rate": 9.468845530230084e-05,
+      "loss": 0.7261,
+      "step": 754
+    },
+    {
+      "epoch": 0.17400322654989628,
+      "grad_norm": 0.3057329058647156,
+      "learning_rate": 9.467169985741337e-05,
+      "loss": 0.7332,
+      "step": 755
+    },
+    {
+      "epoch": 0.17423369439963124,
+      "grad_norm": 0.33508437871932983,
+      "learning_rate": 9.465491951359265e-05,
+      "loss": 0.7242,
+      "step": 756
+    },
+    {
+      "epoch": 0.1744641622493662,
+      "grad_norm": 0.41482967138290405,
+      "learning_rate": 9.463811428019156e-05,
+      "loss": 0.7284,
+      "step": 757
+    },
+    {
+      "epoch": 0.1746946300991012,
+      "grad_norm": 0.3846109211444855,
+      "learning_rate": 9.4621284166577e-05,
+      "loss": 0.7358,
+      "step": 758
+    },
+    {
+      "epoch": 0.17492509794883615,
+      "grad_norm": 0.35918307304382324,
+      "learning_rate": 9.460442918212965e-05,
+      "loss": 0.7249,
+      "step": 759
+    },
+    {
+      "epoch": 0.1751555657985711,
+      "grad_norm": 0.38481009006500244,
+      "learning_rate": 9.458754933624406e-05,
+      "loss": 0.736,
+      "step": 760
+    },
+    {
+      "epoch": 0.17538603364830607,
+      "grad_norm": 0.34131038188934326,
+      "learning_rate": 9.457064463832868e-05,
+      "loss": 0.7291,
+      "step": 761
+    },
+    {
+      "epoch": 0.17561650149804103,
+      "grad_norm": 0.3763889670372009,
+      "learning_rate": 9.455371509780575e-05,
+      "loss": 0.7322,
+      "step": 762
+    },
+    {
+      "epoch": 0.175846969347776,
+      "grad_norm": 0.36019256711006165,
+      "learning_rate": 9.453676072411142e-05,
+      "loss": 0.7206,
+      "step": 763
+    },
+    {
+      "epoch": 0.17607743719751096,
+      "grad_norm": 0.324862003326416,
+      "learning_rate": 9.451978152669563e-05,
+      "loss": 0.7277,
+      "step": 764
+    },
+    {
+      "epoch": 0.17630790504724592,
+      "grad_norm": 0.3834820091724396,
+      "learning_rate": 9.450277751502218e-05,
+      "loss": 0.7416,
+      "step": 765
+    },
+    {
+      "epoch": 0.17653837289698088,
+      "grad_norm": 0.37926480174064636,
+      "learning_rate": 9.44857486985687e-05,
+      "loss": 0.7293,
+      "step": 766
+    },
+    {
+      "epoch": 0.17676884074671584,
+      "grad_norm": 0.3776107430458069,
+      "learning_rate": 9.446869508682666e-05,
+      "loss": 0.7384,
+      "step": 767
+    },
+    {
+      "epoch": 0.1769993085964508,
+      "grad_norm": 0.2972352206707001,
+      "learning_rate": 9.445161668930129e-05,
+      "loss": 0.7351,
+      "step": 768
+    },
+    {
+      "epoch": 0.17722977644618576,
+      "grad_norm": 0.2998862862586975,
+      "learning_rate": 9.443451351551174e-05,
+      "loss": 0.7326,
+      "step": 769
+    },
+    {
+      "epoch": 0.17746024429592072,
+      "grad_norm": 0.31847459077835083,
+      "learning_rate": 9.441738557499087e-05,
+      "loss": 0.7267,
+      "step": 770
+    },
+    {
+      "epoch": 0.17769071214565568,
+      "grad_norm": 0.3504515588283539,
+      "learning_rate": 9.440023287728537e-05,
+      "loss": 0.732,
+      "step": 771
+    },
+    {
+      "epoch": 0.17792117999539064,
+      "grad_norm": 0.2948881983757019,
+      "learning_rate": 9.438305543195579e-05,
+      "loss": 0.735,
+      "step": 772
+    },
+    {
+      "epoch": 0.1781516478451256,
+      "grad_norm": 0.3130739629268646,
+      "learning_rate": 9.436585324857636e-05,
+      "loss": 0.7325,
+      "step": 773
+    },
+    {
+      "epoch": 0.17838211569486057,
+      "grad_norm": 0.30023884773254395,
+      "learning_rate": 9.434862633673523e-05,
+      "loss": 0.7274,
+      "step": 774
+    },
+    {
+      "epoch": 0.17861258354459553,
+      "grad_norm": 0.36194613575935364,
+      "learning_rate": 9.433137470603424e-05,
+      "loss": 0.7273,
+      "step": 775
+    },
+    {
+      "epoch": 0.1788430513943305,
+      "grad_norm": 0.330105721950531,
+      "learning_rate": 9.431409836608902e-05,
+      "loss": 0.7236,
+      "step": 776
+    },
+    {
+      "epoch": 0.17907351924406545,
+      "grad_norm": 0.3341142535209656,
+      "learning_rate": 9.429679732652901e-05,
+      "loss": 0.7262,
+      "step": 777
+    },
+    {
+      "epoch": 0.1793039870938004,
+      "grad_norm": 0.32330140471458435,
+      "learning_rate": 9.427947159699738e-05,
+      "loss": 0.7299,
+      "step": 778
+    },
+    {
+      "epoch": 0.17953445494353537,
+      "grad_norm": 0.3477638363838196,
+      "learning_rate": 9.426212118715108e-05,
+      "loss": 0.7379,
+      "step": 779
+    },
+    {
+      "epoch": 0.17976492279327033,
+      "grad_norm": 0.36434227228164673,
+      "learning_rate": 9.424474610666082e-05,
+      "loss": 0.7247,
+      "step": 780
+    },
+    {
+      "epoch": 0.1799953906430053,
+      "grad_norm": 0.3425809144973755,
+      "learning_rate": 9.422734636521104e-05,
+      "loss": 0.7276,
+      "step": 781
+    },
+    {
+      "epoch": 0.18022585849274025,
+      "grad_norm": 0.33053991198539734,
+      "learning_rate": 9.420992197249994e-05,
+      "loss": 0.7255,
+      "step": 782
+    },
+    {
+      "epoch": 0.18045632634247522,
+      "grad_norm": 0.3649672567844391,
+      "learning_rate": 9.419247293823947e-05,
+      "loss": 0.7249,
+      "step": 783
+    },
+    {
+      "epoch": 0.18068679419221018,
+      "grad_norm": 0.4266059696674347,
+      "learning_rate": 9.417499927215528e-05,
+      "loss": 0.7247,
+      "step": 784
+    },
+    {
+      "epoch": 0.18091726204194514,
+      "grad_norm": 0.360500693321228,
+      "learning_rate": 9.41575009839868e-05,
+      "loss": 0.7343,
+      "step": 785
+    },
+    {
+      "epoch": 0.1811477298916801,
+      "grad_norm": 0.37427565455436707,
+      "learning_rate": 9.413997808348714e-05,
+      "loss": 0.7195,
+      "step": 786
+    },
+    {
+      "epoch": 0.18137819774141506,
+      "grad_norm": 0.42820265889167786,
+      "learning_rate": 9.412243058042315e-05,
+      "loss": 0.7426,
+      "step": 787
+    },
+    {
+      "epoch": 0.18160866559115005,
+      "grad_norm": 0.36763229966163635,
+      "learning_rate": 9.410485848457537e-05,
+      "loss": 0.7335,
+      "step": 788
+    },
+    {
+      "epoch": 0.181839133440885,
+      "grad_norm": 0.3920682668685913,
+      "learning_rate": 9.408726180573811e-05,
+      "loss": 0.7293,
+      "step": 789
+    },
+    {
+      "epoch": 0.18206960129061997,
+      "grad_norm": 0.41110143065452576,
+      "learning_rate": 9.406964055371928e-05,
+      "loss": 0.7226,
+      "step": 790
+    },
+    {
+      "epoch": 0.18230006914035493,
+      "grad_norm": 0.31723544001579285,
+      "learning_rate": 9.405199473834057e-05,
+      "loss": 0.7195,
+      "step": 791
+    },
+    {
+      "epoch": 0.1825305369900899,
+      "grad_norm": 0.3424417972564697,
+      "learning_rate": 9.403432436943733e-05,
+      "loss": 0.727,
+      "step": 792
+    },
+    {
+      "epoch": 0.18276100483982485,
+      "grad_norm": 0.4001381993293762,
+      "learning_rate": 9.40166294568586e-05,
+      "loss": 0.7267,
+      "step": 793
+    },
+    {
+      "epoch": 0.18299147268955981,
+      "grad_norm": 0.4231438636779785,
+      "learning_rate": 9.399891001046712e-05,
+      "loss": 0.7256,
+      "step": 794
+    },
+    {
+      "epoch": 0.18322194053929478,
+      "grad_norm": 0.3733879029750824,
+      "learning_rate": 9.398116604013925e-05,
+      "loss": 0.7341,
+      "step": 795
+    },
+    {
+      "epoch": 0.18345240838902974,
+      "grad_norm": 0.28780287504196167,
+      "learning_rate": 9.396339755576506e-05,
+      "loss": 0.7255,
+      "step": 796
+    },
+    {
+      "epoch": 0.1836828762387647,
+      "grad_norm": 0.3469153046607971,
+      "learning_rate": 9.39456045672483e-05,
+      "loss": 0.716,
+      "step": 797
+    },
+    {
+      "epoch": 0.18391334408849966,
+      "grad_norm": 0.30638718605041504,
+      "learning_rate": 9.392778708450633e-05,
+      "loss": 0.7111,
+      "step": 798
+    },
+    {
+      "epoch": 0.18414381193823462,
+      "grad_norm": 0.3063083291053772,
+      "learning_rate": 9.39099451174702e-05,
+      "loss": 0.7271,
+      "step": 799
+    },
+    {
+      "epoch": 0.18437427978796958,
+      "grad_norm": 0.32252591848373413,
+      "learning_rate": 9.389207867608461e-05,
+      "loss": 0.7266,
+      "step": 800
+    },
+    {
+      "epoch": 0.18460474763770454,
+      "grad_norm": 0.3048442304134369,
+      "learning_rate": 9.387418777030784e-05,
+      "loss": 0.7214,
+      "step": 801
+    },
+    {
+      "epoch": 0.1848352154874395,
+      "grad_norm": 0.3063707649707794,
+      "learning_rate": 9.385627241011187e-05,
+      "loss": 0.7223,
+      "step": 802
+    },
+    {
+      "epoch": 0.18506568333717446,
+      "grad_norm": 0.32035815715789795,
+      "learning_rate": 9.383833260548233e-05,
+      "loss": 0.724,
+      "step": 803
+    },
+    {
+      "epoch": 0.18529615118690942,
+      "grad_norm": 0.3516310453414917,
+      "learning_rate": 9.382036836641839e-05,
+      "loss": 0.7246,
+      "step": 804
+    },
+    {
+      "epoch": 0.18552661903664439,
+      "grad_norm": 0.34494486451148987,
+      "learning_rate": 9.380237970293291e-05,
+      "loss": 0.7116,
+      "step": 805
+    },
+    {
+      "epoch": 0.18575708688637935,
+      "grad_norm": 0.34283608198165894,
+      "learning_rate": 9.378436662505232e-05,
+      "loss": 0.7211,
+      "step": 806
+    },
+    {
+      "epoch": 0.1859875547361143,
+      "grad_norm": 0.30393049120903015,
+      "learning_rate": 9.376632914281669e-05,
+      "loss": 0.724,
+      "step": 807
+    },
+    {
+      "epoch": 0.18621802258584927,
+      "grad_norm": 0.27396926283836365,
+      "learning_rate": 9.374826726627969e-05,
+      "loss": 0.7202,
+      "step": 808
+    },
+    {
+      "epoch": 0.18644849043558423,
+      "grad_norm": 0.3302895724773407,
+      "learning_rate": 9.373018100550855e-05,
+      "loss": 0.7303,
+      "step": 809
+    },
+    {
+      "epoch": 0.1866789582853192,
+      "grad_norm": 0.34311676025390625,
+      "learning_rate": 9.371207037058414e-05,
+      "loss": 0.7205,
+      "step": 810
+    },
+    {
+      "epoch": 0.18690942613505415,
+      "grad_norm": 0.3147958219051361,
+      "learning_rate": 9.369393537160089e-05,
+      "loss": 0.7234,
+      "step": 811
+    },
+    {
+      "epoch": 0.1871398939847891,
+      "grad_norm": 0.29841968417167664,
+      "learning_rate": 9.36757760186668e-05,
+      "loss": 0.7242,
+      "step": 812
+    },
+    {
+      "epoch": 0.18737036183452407,
+      "grad_norm": 0.3176330029964447,
+      "learning_rate": 9.365759232190348e-05,
+      "loss": 0.7222,
+      "step": 813
+    },
+    {
+      "epoch": 0.18760082968425904,
+      "grad_norm": 0.3466809093952179,
+      "learning_rate": 9.363938429144605e-05,
+      "loss": 0.7221,
+      "step": 814
+    },
+    {
+      "epoch": 0.187831297533994,
+      "grad_norm": 0.35498663783073425,
+      "learning_rate": 9.362115193744328e-05,
+      "loss": 0.7221,
+      "step": 815
+    },
+    {
+      "epoch": 0.18806176538372896,
+      "grad_norm": 0.4183621108531952,
+      "learning_rate": 9.360289527005739e-05,
+      "loss": 0.7296,
+      "step": 816
+    },
+    {
+      "epoch": 0.18829223323346392,
+      "grad_norm": 0.41364750266075134,
+      "learning_rate": 9.358461429946425e-05,
+      "loss": 0.7226,
+      "step": 817
+    },
+    {
+      "epoch": 0.1885227010831989,
+      "grad_norm": 0.3849059045314789,
+      "learning_rate": 9.356630903585321e-05,
+      "loss": 0.7242,
+      "step": 818
+    },
+    {
+      "epoch": 0.18875316893293387,
+      "grad_norm": 0.3003298342227936,
+      "learning_rate": 9.354797948942719e-05,
+      "loss": 0.7267,
+      "step": 819
+    },
+    {
+      "epoch": 0.18898363678266883,
+      "grad_norm": 0.294464647769928,
+      "learning_rate": 9.352962567040266e-05,
+      "loss": 0.7208,
+      "step": 820
+    },
+    {
+      "epoch": 0.1892141046324038,
+      "grad_norm": 0.3354666531085968,
+      "learning_rate": 9.35112475890096e-05,
+      "loss": 0.7165,
+      "step": 821
+    },
+    {
+      "epoch": 0.18944457248213875,
+      "grad_norm": 0.3083663582801819,
+      "learning_rate": 9.349284525549147e-05,
+      "loss": 0.7148,
+      "step": 822
+    },
+    {
+      "epoch": 0.1896750403318737,
+      "grad_norm": 0.28768306970596313,
+      "learning_rate": 9.347441868010531e-05,
+      "loss": 0.7136,
+      "step": 823
+    },
+    {
+      "epoch": 0.18990550818160867,
+      "grad_norm": 0.2938915193080902,
+      "learning_rate": 9.345596787312165e-05,
+      "loss": 0.7293,
+      "step": 824
+    },
+    {
+      "epoch": 0.19013597603134363,
+      "grad_norm": 0.3318259119987488,
+      "learning_rate": 9.343749284482454e-05,
+      "loss": 0.7088,
+      "step": 825
+    },
+    {
+      "epoch": 0.1903664438810786,
+      "grad_norm": 0.3234262764453888,
+      "learning_rate": 9.341899360551151e-05,
+      "loss": 0.7196,
+      "step": 826
+    },
+    {
+      "epoch": 0.19059691173081356,
+      "grad_norm": 0.32499444484710693,
+      "learning_rate": 9.340047016549358e-05,
+      "loss": 0.7169,
+      "step": 827
+    },
+    {
+      "epoch": 0.19082737958054852,
+      "grad_norm": 0.28975191712379456,
+      "learning_rate": 9.33819225350953e-05,
+      "loss": 0.7216,
+      "step": 828
+    },
+    {
+      "epoch": 0.19105784743028348,
+      "grad_norm": 0.28860896825790405,
+      "learning_rate": 9.336335072465465e-05,
+      "loss": 0.7174,
+      "step": 829
+    },
+    {
+      "epoch": 0.19128831528001844,
+      "grad_norm": 0.3400057554244995,
+      "learning_rate": 9.334475474452308e-05,
+      "loss": 0.7273,
+      "step": 830
+    },
+    {
+      "epoch": 0.1915187831297534,
+      "grad_norm": 0.3378770649433136,
+      "learning_rate": 9.332613460506563e-05,
+      "loss": 0.7221,
+      "step": 831
+    },
+    {
+      "epoch": 0.19174925097948836,
+      "grad_norm": 0.2862728536128998,
+      "learning_rate": 9.330749031666064e-05,
+      "loss": 0.7144,
+      "step": 832
+    },
+    {
+      "epoch": 0.19197971882922332,
+      "grad_norm": 0.3653225004673004,
+      "learning_rate": 9.328882188970003e-05,
+      "loss": 0.7235,
+      "step": 833
+    },
+    {
+      "epoch": 0.19221018667895828,
+      "grad_norm": 0.40798598527908325,
+      "learning_rate": 9.327012933458909e-05,
+      "loss": 0.7232,
+      "step": 834
+    },
+    {
+      "epoch": 0.19244065452869324,
+      "grad_norm": 0.3767978250980377,
+      "learning_rate": 9.325141266174666e-05,
+      "loss": 0.7354,
+      "step": 835
+    },
+    {
+      "epoch": 0.1926711223784282,
+      "grad_norm": 0.3316027522087097,
+      "learning_rate": 9.323267188160494e-05,
+      "loss": 0.7245,
+      "step": 836
+    },
+    {
+      "epoch": 0.19290159022816317,
+      "grad_norm": 0.2936481535434723,
+      "learning_rate": 9.321390700460956e-05,
+      "loss": 0.7238,
+      "step": 837
+    },
+    {
+      "epoch": 0.19313205807789813,
+      "grad_norm": 0.29495328664779663,
+      "learning_rate": 9.319511804121967e-05,
+      "loss": 0.7136,
+      "step": 838
+    },
+    {
+      "epoch": 0.1933625259276331,
+      "grad_norm": 0.3090824484825134,
+      "learning_rate": 9.317630500190774e-05,
+      "loss": 0.7112,
+      "step": 839
+    },
+    {
+      "epoch": 0.19359299377736805,
+      "grad_norm": 0.288487046957016,
+      "learning_rate": 9.315746789715973e-05,
+      "loss": 0.722,
+      "step": 840
+    },
+    {
+      "epoch": 0.193823461627103,
+      "grad_norm": 0.31703829765319824,
+      "learning_rate": 9.313860673747496e-05,
+      "loss": 0.7235,
+      "step": 841
+    },
+    {
+      "epoch": 0.19405392947683797,
+      "grad_norm": 0.36949485540390015,
+      "learning_rate": 9.311972153336623e-05,
+      "loss": 0.7211,
+      "step": 842
+    },
+    {
+      "epoch": 0.19428439732657293,
+      "grad_norm": 0.3008967936038971,
+      "learning_rate": 9.310081229535968e-05,
+      "loss": 0.722,
+      "step": 843
+    },
+    {
+      "epoch": 0.1945148651763079,
+      "grad_norm": 0.28252121806144714,
+      "learning_rate": 9.308187903399486e-05,
+      "loss": 0.7231,
+      "step": 844
+    },
+    {
+      "epoch": 0.19474533302604286,
+      "grad_norm": 0.32899102568626404,
+      "learning_rate": 9.306292175982472e-05,
+      "loss": 0.7103,
+      "step": 845
+    },
+    {
+      "epoch": 0.19497580087577782,
+      "grad_norm": 0.331802099943161,
+      "learning_rate": 9.304394048341559e-05,
+      "loss": 0.7263,
+      "step": 846
+    },
+    {
+      "epoch": 0.1952062687255128,
+      "grad_norm": 0.2927376925945282,
+      "learning_rate": 9.302493521534719e-05,
+      "loss": 0.7177,
+      "step": 847
+    },
+    {
+      "epoch": 0.19543673657524777,
+      "grad_norm": 0.28728216886520386,
+      "learning_rate": 9.300590596621257e-05,
+      "loss": 0.7118,
+      "step": 848
+    },
+    {
+      "epoch": 0.19566720442498273,
+      "grad_norm": 0.2990799844264984,
+      "learning_rate": 9.29868527466182e-05,
+      "loss": 0.7141,
+      "step": 849
+    },
+    {
+      "epoch": 0.1958976722747177,
+      "grad_norm": 0.2656897306442261,
+      "learning_rate": 9.296777556718387e-05,
+      "loss": 0.728,
+      "step": 850
+    },
+    {
+      "epoch": 0.19612814012445265,
+      "grad_norm": 0.30828338861465454,
+      "learning_rate": 9.294867443854278e-05,
+      "loss": 0.7262,
+      "step": 851
+    },
+    {
+      "epoch": 0.1963586079741876,
+      "grad_norm": 0.29851728677749634,
+      "learning_rate": 9.292954937134142e-05,
+      "loss": 0.7277,
+      "step": 852
+    },
+    {
+      "epoch": 0.19658907582392257,
+      "grad_norm": 0.3060734272003174,
+      "learning_rate": 9.291040037623961e-05,
+      "loss": 0.7239,
+      "step": 853
+    },
+    {
+      "epoch": 0.19681954367365753,
+      "grad_norm": 0.3331679105758667,
+      "learning_rate": 9.28912274639106e-05,
+      "loss": 0.7227,
+      "step": 854
+    },
+    {
+      "epoch": 0.1970500115233925,
+      "grad_norm": 0.28434881567955017,
+      "learning_rate": 9.287203064504084e-05,
+      "loss": 0.7143,
+      "step": 855
+    },
+    {
+      "epoch": 0.19728047937312745,
+      "grad_norm": 0.2975858449935913,
+      "learning_rate": 9.285280993033024e-05,
+      "loss": 0.7194,
+      "step": 856
+    },
+    {
+      "epoch": 0.19751094722286242,
+      "grad_norm": 0.26783379912376404,
+      "learning_rate": 9.283356533049194e-05,
+      "loss": 0.7158,
+      "step": 857
+    },
+    {
+      "epoch": 0.19774141507259738,
+      "grad_norm": 0.31870049238204956,
+      "learning_rate": 9.281429685625243e-05,
+      "loss": 0.7282,
+      "step": 858
+    },
+    {
+      "epoch": 0.19797188292233234,
+      "grad_norm": 0.3032819330692291,
+      "learning_rate": 9.279500451835145e-05,
+      "loss": 0.7217,
+      "step": 859
+    },
+    {
+      "epoch": 0.1982023507720673,
+      "grad_norm": 0.2742626667022705,
+      "learning_rate": 9.277568832754216e-05,
+      "loss": 0.7199,
+      "step": 860
+    },
+    {
+      "epoch": 0.19843281862180226,
+      "grad_norm": 0.2852528989315033,
+      "learning_rate": 9.275634829459087e-05,
+      "loss": 0.7242,
+      "step": 861
+    },
+    {
+      "epoch": 0.19866328647153722,
+      "grad_norm": 0.26641765236854553,
+      "learning_rate": 9.27369844302773e-05,
+      "loss": 0.7202,
+      "step": 862
+    },
+    {
+      "epoch": 0.19889375432127218,
+      "grad_norm": 0.27515074610710144,
+      "learning_rate": 9.271759674539438e-05,
+      "loss": 0.7143,
+      "step": 863
+    },
+    {
+      "epoch": 0.19912422217100714,
+      "grad_norm": 0.3010948598384857,
+      "learning_rate": 9.269818525074833e-05,
+      "loss": 0.7182,
+      "step": 864
+    },
+    {
+      "epoch": 0.1993546900207421,
+      "grad_norm": 0.28209689259529114,
+      "learning_rate": 9.267874995715868e-05,
+      "loss": 0.724,
+      "step": 865
+    },
+    {
+      "epoch": 0.19958515787047706,
+      "grad_norm": 0.29856470227241516,
+      "learning_rate": 9.26592908754582e-05,
+      "loss": 0.7151,
+      "step": 866
+    },
+    {
+      "epoch": 0.19981562572021203,
+      "grad_norm": 0.3309277594089508,
+      "learning_rate": 9.263980801649286e-05,
+      "loss": 0.7171,
+      "step": 867
+    },
+    {
+      "epoch": 0.200046093569947,
+      "grad_norm": 0.264207661151886,
+      "learning_rate": 9.262030139112198e-05,
+      "loss": 0.723,
+      "step": 868
+    },
+    {
+      "epoch": 0.20027656141968195,
+      "grad_norm": 0.31711748242378235,
+      "learning_rate": 9.260077101021811e-05,
+      "loss": 0.7258,
+      "step": 869
+    },
+    {
+      "epoch": 0.2005070292694169,
+      "grad_norm": 0.3416388928890228,
+      "learning_rate": 9.258121688466696e-05,
+      "loss": 0.7271,
+      "step": 870
+    },
+    {
+      "epoch": 0.20073749711915187,
+      "grad_norm": 0.37464427947998047,
+      "learning_rate": 9.256163902536756e-05,
+      "loss": 0.7173,
+      "step": 871
+    },
+    {
+      "epoch": 0.20096796496888683,
+      "grad_norm": 0.36626744270324707,
+      "learning_rate": 9.254203744323216e-05,
+      "loss": 0.7294,
+      "step": 872
+    },
+    {
+      "epoch": 0.2011984328186218,
+      "grad_norm": 0.27999523282051086,
+      "learning_rate": 9.252241214918615e-05,
+      "loss": 0.7204,
+      "step": 873
+    },
+    {
+      "epoch": 0.20142890066835675,
+      "grad_norm": 0.4048886001110077,
+      "learning_rate": 9.250276315416825e-05,
+      "loss": 0.7151,
+      "step": 874
+    },
+    {
+      "epoch": 0.20165936851809171,
+      "grad_norm": 0.4276858866214752,
+      "learning_rate": 9.248309046913032e-05,
+      "loss": 0.7311,
+      "step": 875
+    },
+    {
+      "epoch": 0.20188983636782667,
+      "grad_norm": 0.3747239410877228,
+      "learning_rate": 9.246339410503745e-05,
+      "loss": 0.7223,
+      "step": 876
+    },
+    {
+      "epoch": 0.20212030421756166,
+      "grad_norm": 0.2862272560596466,
+      "learning_rate": 9.24436740728679e-05,
+      "loss": 0.7135,
+      "step": 877
+    },
+    {
+      "epoch": 0.20235077206729662,
+      "grad_norm": 0.3424084484577179,
+      "learning_rate": 9.242393038361316e-05,
+      "loss": 0.7224,
+      "step": 878
+    },
+    {
+      "epoch": 0.20258123991703159,
+      "grad_norm": 0.3472643792629242,
+      "learning_rate": 9.24041630482779e-05,
+      "loss": 0.7144,
+      "step": 879
+    },
+    {
+      "epoch": 0.20281170776676655,
+      "grad_norm": 0.31430667638778687,
+      "learning_rate": 9.238437207787994e-05,
+      "loss": 0.7234,
+      "step": 880
+    },
+    {
+      "epoch": 0.2030421756165015,
+      "grad_norm": 0.2940575182437897,
+      "learning_rate": 9.23645574834503e-05,
+      "loss": 0.7105,
+      "step": 881
+    },
+    {
+      "epoch": 0.20327264346623647,
+      "grad_norm": 0.3699895739555359,
+      "learning_rate": 9.234471927603314e-05,
+      "loss": 0.7107,
+      "step": 882
+    },
+    {
+      "epoch": 0.20350311131597143,
+      "grad_norm": 0.3853522837162018,
+      "learning_rate": 9.232485746668584e-05,
+      "loss": 0.721,
+      "step": 883
+    },
+    {
+      "epoch": 0.2037335791657064,
+      "grad_norm": 0.37054017186164856,
+      "learning_rate": 9.230497206647885e-05,
+      "loss": 0.7209,
+      "step": 884
+    },
+    {
+      "epoch": 0.20396404701544135,
+      "grad_norm": 0.2768975496292114,
+      "learning_rate": 9.228506308649585e-05,
+      "loss": 0.7226,
+      "step": 885
+    },
+    {
+      "epoch": 0.2041945148651763,
+      "grad_norm": 0.3139522671699524,
+      "learning_rate": 9.22651305378336e-05,
+      "loss": 0.7189,
+      "step": 886
+    },
+    {
+      "epoch": 0.20442498271491127,
+      "grad_norm": 0.33646050095558167,
+      "learning_rate": 9.224517443160205e-05,
+      "loss": 0.7163,
+      "step": 887
+    },
+    {
+      "epoch": 0.20465545056464624,
+      "grad_norm": 0.28335535526275635,
+      "learning_rate": 9.222519477892425e-05,
+      "loss": 0.7233,
+      "step": 888
+    },
+    {
+      "epoch": 0.2048859184143812,
+      "grad_norm": 0.31536251306533813,
+      "learning_rate": 9.220519159093637e-05,
+      "loss": 0.7152,
+      "step": 889
+    },
+    {
+      "epoch": 0.20511638626411616,
+      "grad_norm": 0.2616060972213745,
+      "learning_rate": 9.21851648787877e-05,
+      "loss": 0.7144,
+      "step": 890
+    },
+    {
+      "epoch": 0.20534685411385112,
+      "grad_norm": 0.30261725187301636,
+      "learning_rate": 9.216511465364066e-05,
+      "loss": 0.717,
+      "step": 891
+    },
+    {
+      "epoch": 0.20557732196358608,
+      "grad_norm": 0.3098640739917755,
+      "learning_rate": 9.214504092667075e-05,
+      "loss": 0.7182,
+      "step": 892
+    },
+    {
+      "epoch": 0.20580778981332104,
+      "grad_norm": 0.32920923829078674,
+      "learning_rate": 9.212494370906661e-05,
+      "loss": 0.7217,
+      "step": 893
+    },
+    {
+      "epoch": 0.206038257663056,
+      "grad_norm": 0.31871381402015686,
+      "learning_rate": 9.210482301202994e-05,
+      "loss": 0.7135,
+      "step": 894
+    },
+    {
+      "epoch": 0.20626872551279096,
+      "grad_norm": 0.2803013324737549,
+      "learning_rate": 9.208467884677551e-05,
+      "loss": 0.7218,
+      "step": 895
+    },
+    {
+      "epoch": 0.20649919336252592,
+      "grad_norm": 0.31041911244392395,
+      "learning_rate": 9.206451122453122e-05,
+      "loss": 0.7207,
+      "step": 896
+    },
+    {
+      "epoch": 0.20672966121226088,
+      "grad_norm": 0.3201538324356079,
+      "learning_rate": 9.204432015653801e-05,
+      "loss": 0.7163,
+      "step": 897
+    },
+    {
+      "epoch": 0.20696012906199585,
+      "grad_norm": 0.26762691140174866,
+      "learning_rate": 9.202410565404988e-05,
+      "loss": 0.7103,
+      "step": 898
+    },
+    {
+      "epoch": 0.2071905969117308,
+      "grad_norm": 0.3141172528266907,
+      "learning_rate": 9.200386772833394e-05,
+      "loss": 0.7251,
+      "step": 899
+    },
+    {
+      "epoch": 0.20742106476146577,
+      "grad_norm": 0.3220638632774353,
+      "learning_rate": 9.19836063906703e-05,
+      "loss": 0.716,
+      "step": 900
+    },
+    {
+      "epoch": 0.20765153261120073,
+      "grad_norm": 0.3352587819099426,
+      "learning_rate": 9.196332165235215e-05,
+      "loss": 0.7235,
+      "step": 901
+    },
+    {
+      "epoch": 0.2078820004609357,
+      "grad_norm": 0.3438172936439514,
+      "learning_rate": 9.194301352468572e-05,
+      "loss": 0.713,
+      "step": 902
+    },
+    {
+      "epoch": 0.20811246831067065,
+      "grad_norm": 0.3374473452568054,
+      "learning_rate": 9.192268201899028e-05,
+      "loss": 0.7203,
+      "step": 903
+    },
+    {
+      "epoch": 0.2083429361604056,
+      "grad_norm": 0.32671868801116943,
+      "learning_rate": 9.19023271465981e-05,
+      "loss": 0.7264,
+      "step": 904
+    },
+    {
+      "epoch": 0.20857340401014057,
+      "grad_norm": 0.27988868951797485,
+      "learning_rate": 9.188194891885453e-05,
+      "loss": 0.7129,
+      "step": 905
+    },
+    {
+      "epoch": 0.20880387185987553,
+      "grad_norm": 0.2564987540245056,
+      "learning_rate": 9.186154734711786e-05,
+      "loss": 0.7251,
+      "step": 906
+    },
+    {
+      "epoch": 0.20903433970961052,
+      "grad_norm": 0.30010056495666504,
+      "learning_rate": 9.184112244275948e-05,
+      "loss": 0.7137,
+      "step": 907
+    },
+    {
+      "epoch": 0.20926480755934548,
+      "grad_norm": 0.2974473536014557,
+      "learning_rate": 9.182067421716372e-05,
+      "loss": 0.7259,
+      "step": 908
+    },
+    {
+      "epoch": 0.20949527540908044,
+      "grad_norm": 0.31991463899612427,
+      "learning_rate": 9.180020268172794e-05,
+      "loss": 0.7124,
+      "step": 909
+    },
+    {
+      "epoch": 0.2097257432588154,
+      "grad_norm": 0.32082507014274597,
+      "learning_rate": 9.177970784786245e-05,
+      "loss": 0.7184,
+      "step": 910
+    },
+    {
+      "epoch": 0.20995621110855037,
+      "grad_norm": 0.32305994629859924,
+      "learning_rate": 9.175918972699063e-05,
+      "loss": 0.7147,
+      "step": 911
+    },
+    {
+      "epoch": 0.21018667895828533,
+      "grad_norm": 0.3100352883338928,
+      "learning_rate": 9.173864833054875e-05,
+      "loss": 0.7199,
+      "step": 912
+    },
+    {
+      "epoch": 0.2104171468080203,
+      "grad_norm": 0.30127498507499695,
+      "learning_rate": 9.17180836699861e-05,
+      "loss": 0.7188,
+      "step": 913
+    },
+    {
+      "epoch": 0.21064761465775525,
+      "grad_norm": 0.27540695667266846,
+      "learning_rate": 9.169749575676496e-05,
+      "loss": 0.7133,
+      "step": 914
+    },
+    {
+      "epoch": 0.2108780825074902,
+      "grad_norm": 0.3038792610168457,
+      "learning_rate": 9.167688460236049e-05,
+      "loss": 0.716,
+      "step": 915
+    },
+    {
+      "epoch": 0.21110855035722517,
+      "grad_norm": 0.3074706792831421,
+      "learning_rate": 9.165625021826087e-05,
+      "loss": 0.7136,
+      "step": 916
+    },
+    {
+      "epoch": 0.21133901820696013,
+      "grad_norm": 0.31850558519363403,
+      "learning_rate": 9.163559261596723e-05,
+      "loss": 0.7147,
+      "step": 917
+    },
+    {
+      "epoch": 0.2115694860566951,
+      "grad_norm": 0.326848566532135,
+      "learning_rate": 9.161491180699359e-05,
+      "loss": 0.7126,
+      "step": 918
+    },
+    {
+      "epoch": 0.21179995390643006,
+      "grad_norm": 0.278940349817276,
+      "learning_rate": 9.159420780286699e-05,
+      "loss": 0.7173,
+      "step": 919
+    },
+    {
+      "epoch": 0.21203042175616502,
+      "grad_norm": 0.3120673894882202,
+      "learning_rate": 9.157348061512727e-05,
+      "loss": 0.7157,
+      "step": 920
+    },
+    {
+      "epoch": 0.21226088960589998,
+      "grad_norm": 0.2851211428642273,
+      "learning_rate": 9.155273025532731e-05,
+      "loss": 0.7153,
+      "step": 921
+    },
+    {
+      "epoch": 0.21249135745563494,
+      "grad_norm": 0.2742621600627899,
+      "learning_rate": 9.153195673503286e-05,
+      "loss": 0.7158,
+      "step": 922
+    },
+    {
+      "epoch": 0.2127218253053699,
+      "grad_norm": 0.29812031984329224,
+      "learning_rate": 9.151116006582259e-05,
+      "loss": 0.7083,
+      "step": 923
+    },
+    {
+      "epoch": 0.21295229315510486,
+      "grad_norm": 0.33117279410362244,
+      "learning_rate": 9.149034025928804e-05,
+      "loss": 0.7196,
+      "step": 924
+    },
+    {
+      "epoch": 0.21318276100483982,
+      "grad_norm": 0.2820033133029938,
+      "learning_rate": 9.146949732703368e-05,
+      "loss": 0.7102,
+      "step": 925
+    },
+    {
+      "epoch": 0.21341322885457478,
+      "grad_norm": 0.2883508801460266,
+      "learning_rate": 9.144863128067687e-05,
+      "loss": 0.7246,
+      "step": 926
+    },
+    {
+      "epoch": 0.21364369670430974,
+      "grad_norm": 0.34615033864974976,
+      "learning_rate": 9.142774213184784e-05,
+      "loss": 0.7164,
+      "step": 927
+    },
+    {
+      "epoch": 0.2138741645540447,
+      "grad_norm": 0.33778396248817444,
+      "learning_rate": 9.140682989218969e-05,
+      "loss": 0.7152,
+      "step": 928
+    },
+    {
+      "epoch": 0.21410463240377967,
+      "grad_norm": 0.37993401288986206,
+      "learning_rate": 9.138589457335842e-05,
+      "loss": 0.715,
+      "step": 929
+    },
+    {
+      "epoch": 0.21433510025351463,
+      "grad_norm": 0.32394763827323914,
+      "learning_rate": 9.136493618702284e-05,
+      "loss": 0.7263,
+      "step": 930
+    },
+    {
+      "epoch": 0.2145655681032496,
+      "grad_norm": 0.2940785884857178,
+      "learning_rate": 9.13439547448647e-05,
+      "loss": 0.7267,
+      "step": 931
+    },
+    {
+      "epoch": 0.21479603595298455,
+      "grad_norm": 0.32346901297569275,
+      "learning_rate": 9.132295025857851e-05,
+      "loss": 0.719,
+      "step": 932
+    },
+    {
+      "epoch": 0.2150265038027195,
+      "grad_norm": 0.40225303173065186,
+      "learning_rate": 9.130192273987168e-05,
+      "loss": 0.7186,
+      "step": 933
+    },
+    {
+      "epoch": 0.21525697165245447,
+      "grad_norm": 0.28617575764656067,
+      "learning_rate": 9.128087220046445e-05,
+      "loss": 0.7148,
+      "step": 934
+    },
+    {
+      "epoch": 0.21548743950218943,
+      "grad_norm": 0.3431677222251892,
+      "learning_rate": 9.125979865208988e-05,
+      "loss": 0.7135,
+      "step": 935
+    },
+    {
+      "epoch": 0.2157179073519244,
+      "grad_norm": 0.4484764039516449,
+      "learning_rate": 9.123870210649387e-05,
+      "loss": 0.7096,
+      "step": 936
+    },
+    {
+      "epoch": 0.21594837520165938,
+      "grad_norm": 0.39285802841186523,
+      "learning_rate": 9.121758257543511e-05,
+      "loss": 0.7138,
+      "step": 937
+    },
+    {
+      "epoch": 0.21617884305139434,
+      "grad_norm": 0.2899276912212372,
+      "learning_rate": 9.119644007068513e-05,
+      "loss": 0.7072,
+      "step": 938
+    },
+    {
+      "epoch": 0.2164093109011293,
+      "grad_norm": 0.39116156101226807,
+      "learning_rate": 9.117527460402826e-05,
+      "loss": 0.7193,
+      "step": 939
+    },
+    {
+      "epoch": 0.21663977875086426,
+      "grad_norm": 0.3671886622905731,
+      "learning_rate": 9.115408618726162e-05,
+      "loss": 0.7171,
+      "step": 940
+    },
+    {
+      "epoch": 0.21687024660059923,
+      "grad_norm": 0.35713937878608704,
+      "learning_rate": 9.113287483219511e-05,
+      "loss": 0.7105,
+      "step": 941
+    },
+    {
+      "epoch": 0.2171007144503342,
+      "grad_norm": 0.3286116123199463,
+      "learning_rate": 9.111164055065145e-05,
+      "loss": 0.7185,
+      "step": 942
+    },
+    {
+      "epoch": 0.21733118230006915,
+      "grad_norm": 0.31977954506874084,
+      "learning_rate": 9.109038335446612e-05,
+      "loss": 0.7177,
+      "step": 943
+    },
+    {
+      "epoch": 0.2175616501498041,
+      "grad_norm": 0.3611249625682831,
+      "learning_rate": 9.106910325548734e-05,
+      "loss": 0.7169,
+      "step": 944
+    },
+    {
+      "epoch": 0.21779211799953907,
+      "grad_norm": 0.352547824382782,
+      "learning_rate": 9.104780026557618e-05,
+      "loss": 0.7129,
+      "step": 945
+    },
+    {
+      "epoch": 0.21802258584927403,
+      "grad_norm": 0.33489990234375,
+      "learning_rate": 9.102647439660637e-05,
+      "loss": 0.7123,
+      "step": 946
+    },
+    {
+      "epoch": 0.218253053699009,
+      "grad_norm": 0.37026363611221313,
+      "learning_rate": 9.100512566046444e-05,
+      "loss": 0.7225,
+      "step": 947
+    },
+    {
+      "epoch": 0.21848352154874395,
+      "grad_norm": 0.3738853335380554,
+      "learning_rate": 9.098375406904968e-05,
+      "loss": 0.7086,
+      "step": 948
+    },
+    {
+      "epoch": 0.21871398939847891,
+      "grad_norm": 0.2526826560497284,
+      "learning_rate": 9.09623596342741e-05,
+      "loss": 0.7142,
+      "step": 949
+    },
+    {
+      "epoch": 0.21894445724821388,
+      "grad_norm": 0.3173041045665741,
+      "learning_rate": 9.094094236806244e-05,
+      "loss": 0.7133,
+      "step": 950
+    },
+    {
+      "epoch": 0.21917492509794884,
+      "grad_norm": 0.35794541239738464,
+      "learning_rate": 9.091950228235217e-05,
+      "loss": 0.7211,
+      "step": 951
+    },
+    {
+      "epoch": 0.2194053929476838,
+      "grad_norm": 0.36241787672042847,
+      "learning_rate": 9.089803938909349e-05,
+      "loss": 0.7164,
+      "step": 952
+    },
+    {
+      "epoch": 0.21963586079741876,
+      "grad_norm": 0.28174248337745667,
+      "learning_rate": 9.087655370024928e-05,
+      "loss": 0.7085,
+      "step": 953
+    },
+    {
+      "epoch": 0.21986632864715372,
+      "grad_norm": 0.3472788631916046,
+      "learning_rate": 9.085504522779517e-05,
+      "loss": 0.7166,
+      "step": 954
+    },
+    {
+      "epoch": 0.22009679649688868,
+      "grad_norm": 0.39354413747787476,
+      "learning_rate": 9.083351398371944e-05,
+      "loss": 0.7086,
+      "step": 955
+    },
+    {
+      "epoch": 0.22032726434662364,
+      "grad_norm": 0.4095928966999054,
+      "learning_rate": 9.081195998002312e-05,
+      "loss": 0.7136,
+      "step": 956
+    },
+    {
+      "epoch": 0.2205577321963586,
+      "grad_norm": 0.3270823061466217,
+      "learning_rate": 9.079038322871987e-05,
+      "loss": 0.7218,
+      "step": 957
+    },
+    {
+      "epoch": 0.22078820004609356,
+      "grad_norm": 0.30970245599746704,
+      "learning_rate": 9.076878374183606e-05,
+      "loss": 0.7144,
+      "step": 958
+    },
+    {
+      "epoch": 0.22101866789582852,
+      "grad_norm": 0.33222824335098267,
+      "learning_rate": 9.074716153141074e-05,
+      "loss": 0.7069,
+      "step": 959
+    },
+    {
+      "epoch": 0.22124913574556349,
+      "grad_norm": 0.29064905643463135,
+      "learning_rate": 9.072551660949558e-05,
+      "loss": 0.7171,
+      "step": 960
+    },
+    {
+      "epoch": 0.22147960359529845,
+      "grad_norm": 0.2900223135948181,
+      "learning_rate": 9.070384898815497e-05,
+      "loss": 0.7182,
+      "step": 961
+    },
+    {
+      "epoch": 0.2217100714450334,
+      "grad_norm": 0.33210277557373047,
+      "learning_rate": 9.068215867946591e-05,
+      "loss": 0.7184,
+      "step": 962
+    },
+    {
+      "epoch": 0.22194053929476837,
+      "grad_norm": 0.35366907715797424,
+      "learning_rate": 9.066044569551807e-05,
+      "loss": 0.7209,
+      "step": 963
+    },
+    {
+      "epoch": 0.22217100714450333,
+      "grad_norm": 0.29984578490257263,
+      "learning_rate": 9.063871004841371e-05,
+      "loss": 0.7083,
+      "step": 964
+    },
+    {
+      "epoch": 0.2224014749942383,
+      "grad_norm": 0.2960159182548523,
+      "learning_rate": 9.061695175026779e-05,
+      "loss": 0.7146,
+      "step": 965
+    },
+    {
+      "epoch": 0.22263194284397328,
+      "grad_norm": 0.3301765024662018,
+      "learning_rate": 9.059517081320787e-05,
+      "loss": 0.7001,
+      "step": 966
+    },
+    {
+      "epoch": 0.22286241069370824,
+      "grad_norm": 0.3346138894557953,
+      "learning_rate": 9.057336724937409e-05,
+      "loss": 0.7132,
+      "step": 967
+    },
+    {
+      "epoch": 0.2230928785434432,
+      "grad_norm": 0.32978126406669617,
+      "learning_rate": 9.055154107091925e-05,
+      "loss": 0.7125,
+      "step": 968
+    },
+    {
+      "epoch": 0.22332334639317816,
+      "grad_norm": 0.3108513653278351,
+      "learning_rate": 9.052969229000874e-05,
+      "loss": 0.7251,
+      "step": 969
+    },
+    {
+      "epoch": 0.22355381424291312,
+      "grad_norm": 0.36438095569610596,
+      "learning_rate": 9.050782091882056e-05,
+      "loss": 0.7118,
+      "step": 970
+    },
+    {
+      "epoch": 0.22378428209264808,
+      "grad_norm": 0.3605962097644806,
+      "learning_rate": 9.048592696954524e-05,
+      "loss": 0.7099,
+      "step": 971
+    },
+    {
+      "epoch": 0.22401474994238305,
+      "grad_norm": 0.2907812297344208,
+      "learning_rate": 9.0464010454386e-05,
+      "loss": 0.7062,
+      "step": 972
+    },
+    {
+      "epoch": 0.224245217792118,
+      "grad_norm": 0.30365508794784546,
+      "learning_rate": 9.044207138555854e-05,
+      "loss": 0.7057,
+      "step": 973
+    },
+    {
+      "epoch": 0.22447568564185297,
+      "grad_norm": 0.3420591950416565,
+      "learning_rate": 9.042010977529118e-05,
+      "loss": 0.7136,
+      "step": 974
+    },
+    {
+      "epoch": 0.22470615349158793,
+      "grad_norm": 0.27449890971183777,
+      "learning_rate": 9.039812563582482e-05,
+      "loss": 0.7185,
+      "step": 975
+    },
+    {
+      "epoch": 0.2249366213413229,
+      "grad_norm": 0.3264601230621338,
+      "learning_rate": 9.037611897941283e-05,
+      "loss": 0.7205,
+      "step": 976
+    },
+    {
+      "epoch": 0.22516708919105785,
+      "grad_norm": 0.3168565332889557,
+      "learning_rate": 9.035408981832126e-05,
+      "loss": 0.723,
+      "step": 977
+    },
+    {
+      "epoch": 0.2253975570407928,
+      "grad_norm": 0.34214240312576294,
+      "learning_rate": 9.03320381648286e-05,
+      "loss": 0.7068,
+      "step": 978
+    },
+    {
+      "epoch": 0.22562802489052777,
+      "grad_norm": 0.3359646797180176,
+      "learning_rate": 9.030996403122592e-05,
+      "loss": 0.7159,
+      "step": 979
+    },
+    {
+      "epoch": 0.22585849274026273,
+      "grad_norm": 0.32919085025787354,
+      "learning_rate": 9.02878674298168e-05,
+      "loss": 0.7126,
+      "step": 980
+    },
+    {
+      "epoch": 0.2260889605899977,
+      "grad_norm": 0.2955667972564697,
+      "learning_rate": 9.026574837291739e-05,
+      "loss": 0.7104,
+      "step": 981
+    },
+    {
+      "epoch": 0.22631942843973266,
+      "grad_norm": 0.307659775018692,
+      "learning_rate": 9.024360687285629e-05,
+      "loss": 0.7065,
+      "step": 982
+    },
+    {
+      "epoch": 0.22654989628946762,
+      "grad_norm": 0.31468522548675537,
+      "learning_rate": 9.022144294197466e-05,
+      "loss": 0.7091,
+      "step": 983
+    },
+    {
+      "epoch": 0.22678036413920258,
+      "grad_norm": 0.2780282497406006,
+      "learning_rate": 9.019925659262612e-05,
+      "loss": 0.7115,
+      "step": 984
+    },
+    {
+      "epoch": 0.22701083198893754,
+      "grad_norm": 0.32122835516929626,
+      "learning_rate": 9.01770478371768e-05,
+      "loss": 0.7219,
+      "step": 985
+    },
+    {
+      "epoch": 0.2272412998386725,
+      "grad_norm": 0.33038368821144104,
+      "learning_rate": 9.015481668800537e-05,
+      "loss": 0.7051,
+      "step": 986
+    },
+    {
+      "epoch": 0.22747176768840746,
+      "grad_norm": 0.3402576744556427,
+      "learning_rate": 9.013256315750291e-05,
+      "loss": 0.7195,
+      "step": 987
+    },
+    {
+      "epoch": 0.22770223553814242,
+      "grad_norm": 0.37470388412475586,
+      "learning_rate": 9.011028725807302e-05,
+      "loss": 0.7078,
+      "step": 988
+    },
+    {
+      "epoch": 0.22793270338787738,
+      "grad_norm": 0.3758593797683716,
+      "learning_rate": 9.008798900213173e-05,
+      "loss": 0.7092,
+      "step": 989
+    },
+    {
+      "epoch": 0.22816317123761234,
+      "grad_norm": 0.31909018754959106,
+      "learning_rate": 9.006566840210757e-05,
+      "loss": 0.7129,
+      "step": 990
+    },
+    {
+      "epoch": 0.2283936390873473,
+      "grad_norm": 0.34421950578689575,
+      "learning_rate": 9.00433254704415e-05,
+      "loss": 0.7181,
+      "step": 991
+    },
+    {
+      "epoch": 0.22862410693708227,
+      "grad_norm": 0.3339763879776001,
+      "learning_rate": 9.002096021958693e-05,
+      "loss": 0.7227,
+      "step": 992
+    },
+    {
+      "epoch": 0.22885457478681723,
+      "grad_norm": 0.29700416326522827,
+      "learning_rate": 8.999857266200973e-05,
+      "loss": 0.712,
+      "step": 993
+    },
+    {
+      "epoch": 0.2290850426365522,
+      "grad_norm": 0.3107481896877289,
+      "learning_rate": 8.997616281018816e-05,
+      "loss": 0.7106,
+      "step": 994
+    },
+    {
+      "epoch": 0.22931551048628715,
+      "grad_norm": 0.3676799535751343,
+      "learning_rate": 8.995373067661296e-05,
+      "loss": 0.7113,
+      "step": 995
+    },
+    {
+      "epoch": 0.22954597833602214,
+      "grad_norm": 0.2843577563762665,
+      "learning_rate": 8.993127627378727e-05,
+      "loss": 0.7127,
+      "step": 996
+    },
+    {
+      "epoch": 0.2297764461857571,
+      "grad_norm": 0.30162549018859863,
+      "learning_rate": 8.99087996142266e-05,
+      "loss": 0.7159,
+      "step": 997
+    },
+    {
+      "epoch": 0.23000691403549206,
+      "grad_norm": 0.4627409279346466,
+      "learning_rate": 8.988630071045892e-05,
+      "loss": 0.714,
+      "step": 998
+    },
+    {
+      "epoch": 0.23023738188522702,
+      "grad_norm": 0.4146185517311096,
+      "learning_rate": 8.986377957502459e-05,
+      "loss": 0.7239,
+      "step": 999
+    },
+    {
+      "epoch": 0.23046784973496198,
+      "grad_norm": 0.29513731598854065,
+      "learning_rate": 8.984123622047632e-05,
+      "loss": 0.7084,
+      "step": 1000
+    },
+    {
+      "epoch": 0.23069831758469694,
+      "grad_norm": 0.29365894198417664,
+      "learning_rate": 8.981867065937925e-05,
+      "loss": 0.7183,
+      "step": 1001
+    },
+    {
+      "epoch": 0.2309287854344319,
+      "grad_norm": 0.2846957743167877,
+      "learning_rate": 8.97960829043109e-05,
+      "loss": 0.7087,
+      "step": 1002
+    },
+    {
+      "epoch": 0.23115925328416687,
+      "grad_norm": 0.3023827373981476,
+      "learning_rate": 8.977347296786113e-05,
+      "loss": 0.7142,
+      "step": 1003
+    },
+    {
+      "epoch": 0.23138972113390183,
+      "grad_norm": 0.2838916480541229,
+      "learning_rate": 8.975084086263217e-05,
+      "loss": 0.7173,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2316201889836368,
+      "grad_norm": 0.3648954927921295,
+      "learning_rate": 8.97281866012386e-05,
+      "loss": 0.7182,
+      "step": 1005
+    },
+    {
+      "epoch": 0.23185065683337175,
+      "grad_norm": 0.29563209414482117,
+      "learning_rate": 8.97055101963074e-05,
+      "loss": 0.7105,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2320811246831067,
+      "grad_norm": 0.3027130365371704,
+      "learning_rate": 8.968281166047781e-05,
+      "loss": 0.7092,
+      "step": 1007
+    },
+    {
+      "epoch": 0.23231159253284167,
+      "grad_norm": 0.3136472702026367,
+      "learning_rate": 8.966009100640148e-05,
+      "loss": 0.7097,
+      "step": 1008
+    },
+    {
+      "epoch": 0.23254206038257663,
+      "grad_norm": 0.2663586735725403,
+      "learning_rate": 8.963734824674235e-05,
+      "loss": 0.716,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2327725282323116,
+      "grad_norm": 0.27872374653816223,
+      "learning_rate": 8.961458339417669e-05,
+      "loss": 0.7119,
+      "step": 1010
+    },
+    {
+      "epoch": 0.23300299608204655,
+      "grad_norm": 0.32824763655662537,
+      "learning_rate": 8.959179646139306e-05,
+      "loss": 0.7013,
+      "step": 1011
+    },
+    {
+      "epoch": 0.23323346393178152,
+      "grad_norm": 0.2956524193286896,
+      "learning_rate": 8.95689874610924e-05,
+      "loss": 0.7092,
+      "step": 1012
+    },
+    {
+      "epoch": 0.23346393178151648,
+      "grad_norm": 0.2657110095024109,
+      "learning_rate": 8.954615640598782e-05,
+      "loss": 0.7061,
+      "step": 1013
+    },
+    {
+      "epoch": 0.23369439963125144,
+      "grad_norm": 0.31145045161247253,
+      "learning_rate": 8.952330330880487e-05,
+      "loss": 0.7105,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2339248674809864,
+      "grad_norm": 0.31407973170280457,
+      "learning_rate": 8.95004281822813e-05,
+      "loss": 0.7045,
+      "step": 1015
+    },
+    {
+      "epoch": 0.23415533533072136,
+      "grad_norm": 0.24322229623794556,
+      "learning_rate": 8.947753103916716e-05,
+      "loss": 0.7169,
+      "step": 1016
+    },
+    {
+      "epoch": 0.23438580318045632,
+      "grad_norm": 0.34636759757995605,
+      "learning_rate": 8.945461189222474e-05,
+      "loss": 0.7111,
+      "step": 1017
+    },
+    {
+      "epoch": 0.23461627103019128,
+      "grad_norm": 0.3195311725139618,
+      "learning_rate": 8.943167075422867e-05,
+      "loss": 0.7136,
+      "step": 1018
+    },
+    {
+      "epoch": 0.23484673887992624,
+      "grad_norm": 0.30927303433418274,
+      "learning_rate": 8.940870763796574e-05,
+      "loss": 0.7086,
+      "step": 1019
+    },
+    {
+      "epoch": 0.2350772067296612,
+      "grad_norm": 0.29347729682922363,
+      "learning_rate": 8.938572255623509e-05,
+      "loss": 0.7046,
+      "step": 1020
+    },
+    {
+      "epoch": 0.23530767457939616,
+      "grad_norm": 0.28396493196487427,
+      "learning_rate": 8.9362715521848e-05,
+      "loss": 0.7064,
+      "step": 1021
+    },
+    {
+      "epoch": 0.23553814242913113,
+      "grad_norm": 0.26140013337135315,
+      "learning_rate": 8.933968654762808e-05,
+      "loss": 0.7112,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2357686102788661,
+      "grad_norm": 0.2797393500804901,
+      "learning_rate": 8.93166356464111e-05,
+      "loss": 0.7127,
+      "step": 1023
+    },
+    {
+      "epoch": 0.23599907812860105,
+      "grad_norm": 0.289408802986145,
+      "learning_rate": 8.92935628310451e-05,
+      "loss": 0.7079,
+      "step": 1024
+    },
+    {
+      "epoch": 0.236229545978336,
+      "grad_norm": 0.31490978598594666,
+      "learning_rate": 8.92704681143903e-05,
+      "loss": 0.7066,
+      "step": 1025
+    },
+    {
+      "epoch": 0.236460013828071,
+      "grad_norm": 0.30412665009498596,
+      "learning_rate": 8.924735150931915e-05,
+      "loss": 0.7068,
+      "step": 1026
+    },
+    {
+      "epoch": 0.23669048167780596,
+      "grad_norm": 0.3499104380607605,
+      "learning_rate": 8.922421302871629e-05,
+      "loss": 0.7169,
+      "step": 1027
+    },
+    {
+      "epoch": 0.23692094952754092,
+      "grad_norm": 0.3591751158237457,
+      "learning_rate": 8.920105268547854e-05,
+      "loss": 0.7108,
+      "step": 1028
+    },
+    {
+      "epoch": 0.23715141737727588,
+      "grad_norm": 0.35666418075561523,
+      "learning_rate": 8.917787049251493e-05,
+      "loss": 0.7055,
+      "step": 1029
+    },
+    {
+      "epoch": 0.23738188522701084,
+      "grad_norm": 0.35615837574005127,
+      "learning_rate": 8.915466646274666e-05,
+      "loss": 0.7131,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2376123530767458,
+      "grad_norm": 0.37170687317848206,
+      "learning_rate": 8.91314406091071e-05,
+      "loss": 0.7194,
+      "step": 1031
+    },
+    {
+      "epoch": 0.23784282092648076,
+      "grad_norm": 0.35245972871780396,
+      "learning_rate": 8.910819294454177e-05,
+      "loss": 0.7031,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23807328877621572,
+      "grad_norm": 0.2725777328014374,
+      "learning_rate": 8.908492348200838e-05,
+      "loss": 0.7062,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23830375662595069,
+      "grad_norm": 0.3370811343193054,
+      "learning_rate": 8.906163223447676e-05,
+      "loss": 0.7078,
+      "step": 1034
+    },
+    {
+      "epoch": 0.23853422447568565,
+      "grad_norm": 0.35790640115737915,
+      "learning_rate": 8.903831921492889e-05,
+      "loss": 0.7107,
+      "step": 1035
+    },
+    {
+      "epoch": 0.2387646923254206,
+      "grad_norm": 0.2598463296890259,
+      "learning_rate": 8.90149844363589e-05,
+      "loss": 0.7191,
+      "step": 1036
+    },
+    {
+      "epoch": 0.23899516017515557,
+      "grad_norm": 0.2851402461528778,
+      "learning_rate": 8.899162791177301e-05,
+      "loss": 0.7103,
+      "step": 1037
+    },
+    {
+      "epoch": 0.23922562802489053,
+      "grad_norm": 0.34928953647613525,
+      "learning_rate": 8.896824965418961e-05,
+      "loss": 0.7202,
+      "step": 1038
+    },
+    {
+      "epoch": 0.2394560958746255,
+      "grad_norm": 0.3223908245563507,
+      "learning_rate": 8.894484967663917e-05,
+      "loss": 0.7086,
+      "step": 1039
+    },
+    {
+      "epoch": 0.23968656372436045,
+      "grad_norm": 0.31179115176200867,
+      "learning_rate": 8.892142799216428e-05,
+      "loss": 0.7122,
+      "step": 1040
+    },
+    {
+      "epoch": 0.2399170315740954,
+      "grad_norm": 0.3074735403060913,
+      "learning_rate": 8.889798461381961e-05,
+      "loss": 0.7091,
+      "step": 1041
+    },
+    {
+      "epoch": 0.24014749942383037,
+      "grad_norm": 0.2974226474761963,
+      "learning_rate": 8.887451955467195e-05,
+      "loss": 0.7096,
+      "step": 1042
+    },
+    {
+      "epoch": 0.24037796727356533,
+      "grad_norm": 0.3330132067203522,
+      "learning_rate": 8.885103282780016e-05,
+      "loss": 0.7104,
+      "step": 1043
+    },
+    {
+      "epoch": 0.2406084351233003,
+      "grad_norm": 0.28745895624160767,
+      "learning_rate": 8.882752444629518e-05,
+      "loss": 0.7174,
+      "step": 1044
+    },
+    {
+      "epoch": 0.24083890297303526,
+      "grad_norm": 0.2735026180744171,
+      "learning_rate": 8.880399442326001e-05,
+      "loss": 0.7211,
+      "step": 1045
+    },
+    {
+      "epoch": 0.24106937082277022,
+      "grad_norm": 0.2821330726146698,
+      "learning_rate": 8.878044277180975e-05,
+      "loss": 0.7153,
+      "step": 1046
+    },
+    {
+      "epoch": 0.24129983867250518,
+      "grad_norm": 0.30076584219932556,
+      "learning_rate": 8.875686950507147e-05,
+      "loss": 0.7149,
+      "step": 1047
+    },
+    {
+      "epoch": 0.24153030652224014,
+      "grad_norm": 0.24282529950141907,
+      "learning_rate": 8.873327463618438e-05,
+      "loss": 0.7039,
+      "step": 1048
+    },
+    {
+      "epoch": 0.2417607743719751,
+      "grad_norm": 0.26797229051589966,
+      "learning_rate": 8.87096581782997e-05,
+      "loss": 0.7115,
+      "step": 1049
+    },
+    {
+      "epoch": 0.24199124222171006,
+      "grad_norm": 0.27635279297828674,
+      "learning_rate": 8.868602014458065e-05,
+      "loss": 0.7142,
+      "step": 1050
+    },
+    {
+      "epoch": 0.24222171007144502,
+      "grad_norm": 0.6189936399459839,
+      "learning_rate": 8.866236054820251e-05,
+      "loss": 0.7268,
+      "step": 1051
+    },
+    {
+      "epoch": 0.24245217792117998,
+      "grad_norm": 0.25462353229522705,
+      "learning_rate": 8.86386794023526e-05,
+      "loss": 0.7054,
+      "step": 1052
+    },
+    {
+      "epoch": 0.24268264577091495,
+      "grad_norm": 0.2922193706035614,
+      "learning_rate": 8.861497672023018e-05,
+      "loss": 0.712,
+      "step": 1053
+    },
+    {
+      "epoch": 0.2429131136206499,
+      "grad_norm": 0.28528186678886414,
+      "learning_rate": 8.859125251504657e-05,
+      "loss": 0.7158,
+      "step": 1054
+    },
+    {
+      "epoch": 0.24314358147038487,
+      "grad_norm": 0.26813432574272156,
+      "learning_rate": 8.85675068000251e-05,
+      "loss": 0.7048,
+      "step": 1055
+    },
+    {
+      "epoch": 0.24337404932011986,
+      "grad_norm": 0.27089470624923706,
+      "learning_rate": 8.854373958840102e-05,
+      "loss": 0.6999,
+      "step": 1056
+    },
+    {
+      "epoch": 0.24360451716985482,
+      "grad_norm": 0.27272212505340576,
+      "learning_rate": 8.851995089342163e-05,
+      "loss": 0.7162,
+      "step": 1057
+    },
+    {
+      "epoch": 0.24383498501958978,
+      "grad_norm": 0.30942168831825256,
+      "learning_rate": 8.849614072834617e-05,
+      "loss": 0.7128,
+      "step": 1058
+    },
+    {
+      "epoch": 0.24406545286932474,
+      "grad_norm": 0.27839475870132446,
+      "learning_rate": 8.847230910644586e-05,
+      "loss": 0.7007,
+      "step": 1059
+    },
+    {
+      "epoch": 0.2442959207190597,
+      "grad_norm": 0.28866031765937805,
+      "learning_rate": 8.844845604100387e-05,
+      "loss": 0.707,
+      "step": 1060
+    },
+    {
+      "epoch": 0.24452638856879466,
+      "grad_norm": 0.30991750955581665,
+      "learning_rate": 8.842458154531533e-05,
+      "loss": 0.7081,
+      "step": 1061
+    },
+    {
+      "epoch": 0.24475685641852962,
+      "grad_norm": 0.2961871325969696,
+      "learning_rate": 8.840068563268728e-05,
+      "loss": 0.7093,
+      "step": 1062
+    },
+    {
+      "epoch": 0.24498732426826458,
+      "grad_norm": 0.26729321479797363,
+      "learning_rate": 8.837676831643877e-05,
+      "loss": 0.7093,
+      "step": 1063
+    },
+    {
+      "epoch": 0.24521779211799954,
+      "grad_norm": 0.32758355140686035,
+      "learning_rate": 8.835282960990073e-05,
+      "loss": 0.7103,
+      "step": 1064
+    },
+    {
+      "epoch": 0.2454482599677345,
+      "grad_norm": 0.2944065034389496,
+      "learning_rate": 8.8328869526416e-05,
+      "loss": 0.7166,
+      "step": 1065
+    },
+    {
+      "epoch": 0.24567872781746947,
+      "grad_norm": 0.2752934992313385,
+      "learning_rate": 8.830488807933937e-05,
+      "loss": 0.7063,
+      "step": 1066
+    },
+    {
+      "epoch": 0.24590919566720443,
+      "grad_norm": 0.29886919260025024,
+      "learning_rate": 8.828088528203753e-05,
+      "loss": 0.7187,
+      "step": 1067
+    },
+    {
+      "epoch": 0.2461396635169394,
+      "grad_norm": 0.32834017276763916,
+      "learning_rate": 8.825686114788904e-05,
+      "loss": 0.7117,
+      "step": 1068
+    },
+    {
+      "epoch": 0.24637013136667435,
+      "grad_norm": 0.3167296051979065,
+      "learning_rate": 8.823281569028439e-05,
+      "loss": 0.7019,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2466005992164093,
+      "grad_norm": 0.29267629981040955,
+      "learning_rate": 8.820874892262596e-05,
+      "loss": 0.7158,
+      "step": 1070
+    },
+    {
+      "epoch": 0.24683106706614427,
+      "grad_norm": 0.29060524702072144,
+      "learning_rate": 8.818466085832796e-05,
+      "loss": 0.7046,
+      "step": 1071
+    },
+    {
+      "epoch": 0.24706153491587923,
+      "grad_norm": 0.29323479533195496,
+      "learning_rate": 8.81605515108165e-05,
+      "loss": 0.7078,
+      "step": 1072
+    },
+    {
+      "epoch": 0.2472920027656142,
+      "grad_norm": 0.27811145782470703,
+      "learning_rate": 8.813642089352957e-05,
+      "loss": 0.7155,
+      "step": 1073
+    },
+    {
+      "epoch": 0.24752247061534915,
+      "grad_norm": 0.31902894377708435,
+      "learning_rate": 8.811226901991698e-05,
+      "loss": 0.7104,
+      "step": 1074
+    },
+    {
+      "epoch": 0.24775293846508412,
+      "grad_norm": 0.2658187747001648,
+      "learning_rate": 8.808809590344042e-05,
+      "loss": 0.6963,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24798340631481908,
+      "grad_norm": 0.2968449592590332,
+      "learning_rate": 8.806390155757339e-05,
+      "loss": 0.703,
+      "step": 1076
+    },
+    {
+      "epoch": 0.24821387416455404,
+      "grad_norm": 0.29377731680870056,
+      "learning_rate": 8.803968599580125e-05,
+      "loss": 0.7063,
+      "step": 1077
+    },
+    {
+      "epoch": 0.248444342014289,
+      "grad_norm": 0.2845487594604492,
+      "learning_rate": 8.801544923162116e-05,
+      "loss": 0.7152,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24867480986402396,
+      "grad_norm": 0.2773647904396057,
+      "learning_rate": 8.799119127854212e-05,
+      "loss": 0.7128,
+      "step": 1079
+    },
+    {
+      "epoch": 0.24890527771375892,
+      "grad_norm": 0.26741766929626465,
+      "learning_rate": 8.796691215008492e-05,
+      "loss": 0.7124,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24913574556349388,
+      "grad_norm": 0.28756871819496155,
+      "learning_rate": 8.794261185978219e-05,
+      "loss": 0.7144,
+      "step": 1081
+    },
+    {
+      "epoch": 0.24936621341322884,
+      "grad_norm": 0.29482561349868774,
+      "learning_rate": 8.79182904211783e-05,
+      "loss": 0.712,
+      "step": 1082
+    },
+    {
+      "epoch": 0.2495966812629638,
+      "grad_norm": 0.2730870842933655,
+      "learning_rate": 8.789394784782945e-05,
+      "loss": 0.7076,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24982714911269877,
+      "grad_norm": 0.2692764103412628,
+      "learning_rate": 8.786958415330359e-05,
+      "loss": 0.7024,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2500576169624337,
+      "grad_norm": 0.2866170108318329,
+      "learning_rate": 8.78451993511805e-05,
+      "loss": 0.7082,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2502880848121687,
+      "grad_norm": 0.300509512424469,
+      "learning_rate": 8.782079345505163e-05,
+      "loss": 0.7082,
+      "step": 1086
+    },
+    {
+      "epoch": 0.25051855266190365,
+      "grad_norm": 0.2729628086090088,
+      "learning_rate": 8.779636647852027e-05,
+      "loss": 0.7143,
+      "step": 1087
+    },
+    {
+      "epoch": 0.2507490205116386,
+      "grad_norm": 0.26800036430358887,
+      "learning_rate": 8.777191843520142e-05,
+      "loss": 0.7149,
+      "step": 1088
+    },
+    {
+      "epoch": 0.25097948836137357,
+      "grad_norm": 0.2820512056350708,
+      "learning_rate": 8.774744933872186e-05,
+      "loss": 0.7154,
+      "step": 1089
+    },
+    {
+      "epoch": 0.25120995621110853,
+      "grad_norm": 0.2764631509780884,
+      "learning_rate": 8.772295920272003e-05,
+      "loss": 0.7179,
+      "step": 1090
+    },
+    {
+      "epoch": 0.2514404240608435,
+      "grad_norm": 0.23512467741966248,
+      "learning_rate": 8.769844804084619e-05,
+      "loss": 0.7126,
+      "step": 1091
+    },
+    {
+      "epoch": 0.25167089191057845,
+      "grad_norm": 0.25535455346107483,
+      "learning_rate": 8.767391586676223e-05,
+      "loss": 0.7041,
+      "step": 1092
+    },
+    {
+      "epoch": 0.2519013597603134,
+      "grad_norm": 0.250704288482666,
+      "learning_rate": 8.764936269414184e-05,
+      "loss": 0.7144,
+      "step": 1093
+    },
+    {
+      "epoch": 0.2521318276100484,
+      "grad_norm": 0.353876531124115,
+      "learning_rate": 8.762478853667033e-05,
+      "loss": 0.7028,
+      "step": 1094
+    },
+    {
+      "epoch": 0.25236229545978334,
+      "grad_norm": 0.2544647455215454,
+      "learning_rate": 8.760019340804478e-05,
+      "loss": 0.7035,
+      "step": 1095
+    },
+    {
+      "epoch": 0.2525927633095183,
+      "grad_norm": 0.2700643241405487,
+      "learning_rate": 8.75755773219739e-05,
+      "loss": 0.7063,
+      "step": 1096
+    },
+    {
+      "epoch": 0.25282323115925326,
+      "grad_norm": 0.3172542154788971,
+      "learning_rate": 8.755094029217809e-05,
+      "loss": 0.7113,
+      "step": 1097
+    },
+    {
+      "epoch": 0.2530536990089882,
+      "grad_norm": 0.2928423583507538,
+      "learning_rate": 8.75262823323895e-05,
+      "loss": 0.7074,
+      "step": 1098
+    },
+    {
+      "epoch": 0.2532841668587232,
+      "grad_norm": 0.27802032232284546,
+      "learning_rate": 8.750160345635183e-05,
+      "loss": 0.7005,
+      "step": 1099
+    },
+    {
+      "epoch": 0.2535146347084582,
+      "grad_norm": 0.313966304063797,
+      "learning_rate": 8.74769036778205e-05,
+      "loss": 0.7019,
+      "step": 1100
+    },
+    {
+      "epoch": 0.25374510255819316,
+      "grad_norm": 0.3200061619281769,
+      "learning_rate": 8.74521830105626e-05,
+      "loss": 0.7082,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2539755704079281,
+      "grad_norm": 0.30982211232185364,
+      "learning_rate": 8.74274414683568e-05,
+      "loss": 0.7161,
+      "step": 1102
+    },
+    {
+      "epoch": 0.2542060382576631,
+      "grad_norm": 0.26937606930732727,
+      "learning_rate": 8.740267906499346e-05,
+      "loss": 0.7037,
+      "step": 1103
+    },
+    {
+      "epoch": 0.25443650610739804,
+      "grad_norm": 0.33216673135757446,
+      "learning_rate": 8.737789581427455e-05,
+      "loss": 0.7087,
+      "step": 1104
+    },
+    {
+      "epoch": 0.254666973957133,
+      "grad_norm": 0.31464841961860657,
+      "learning_rate": 8.73530917300137e-05,
+      "loss": 0.7027,
+      "step": 1105
+    },
+    {
+      "epoch": 0.25489744180686796,
+      "grad_norm": 0.2963610887527466,
+      "learning_rate": 8.732826682603603e-05,
+      "loss": 0.7244,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2551279096566029,
+      "grad_norm": 0.28852593898773193,
+      "learning_rate": 8.73034211161784e-05,
+      "loss": 0.703,
+      "step": 1107
+    },
+    {
+      "epoch": 0.2553583775063379,
+      "grad_norm": 0.2891587018966675,
+      "learning_rate": 8.72785546142892e-05,
+      "loss": 0.6958,
+      "step": 1108
+    },
+    {
+      "epoch": 0.25558884535607285,
+      "grad_norm": 0.3704643249511719,
+      "learning_rate": 8.725366733422842e-05,
+      "loss": 0.7071,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2558193132058078,
+      "grad_norm": 0.3873230516910553,
+      "learning_rate": 8.722875928986762e-05,
+      "loss": 0.708,
+      "step": 1110
+    },
+    {
+      "epoch": 0.25604978105554277,
+      "grad_norm": 0.29551389813423157,
+      "learning_rate": 8.720383049508997e-05,
+      "loss": 0.6972,
+      "step": 1111
+    },
+    {
+      "epoch": 0.25628024890527773,
+      "grad_norm": 0.2649083137512207,
+      "learning_rate": 8.717888096379018e-05,
+      "loss": 0.7027,
+      "step": 1112
+    },
+    {
+      "epoch": 0.2565107167550127,
+      "grad_norm": 0.33623412251472473,
+      "learning_rate": 8.71539107098745e-05,
+      "loss": 0.7038,
+      "step": 1113
+    },
+    {
+      "epoch": 0.25674118460474765,
+      "grad_norm": 0.450360506772995,
+      "learning_rate": 8.712891974726076e-05,
+      "loss": 0.7067,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2569716524544826,
+      "grad_norm": 1.8468785285949707,
+      "learning_rate": 8.710390808987833e-05,
+      "loss": 0.7232,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2572021203042176,
+      "grad_norm": 0.29168421030044556,
+      "learning_rate": 8.70788757516681e-05,
+      "loss": 0.6999,
+      "step": 1116
+    },
+    {
+      "epoch": 0.25743258815395254,
+      "grad_norm": 0.3151211738586426,
+      "learning_rate": 8.705382274658249e-05,
+      "loss": 0.7001,
+      "step": 1117
+    },
+    {
+      "epoch": 0.2576630560036875,
+      "grad_norm": 0.3864755630493164,
+      "learning_rate": 8.702874908858545e-05,
+      "loss": 0.7102,
+      "step": 1118
+    },
+    {
+      "epoch": 0.25789352385342246,
+      "grad_norm": 0.3055296242237091,
+      "learning_rate": 8.700365479165244e-05,
+      "loss": 0.6979,
+      "step": 1119
+    },
+    {
+      "epoch": 0.2581239917031574,
+      "grad_norm": 0.3239305317401886,
+      "learning_rate": 8.697853986977041e-05,
+      "loss": 0.7155,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2583544595528924,
+      "grad_norm": 0.29787853360176086,
+      "learning_rate": 8.695340433693781e-05,
+      "loss": 0.7129,
+      "step": 1121
+    },
+    {
+      "epoch": 0.25858492740262734,
+      "grad_norm": 0.2536768615245819,
+      "learning_rate": 8.692824820716461e-05,
+      "loss": 0.7032,
+      "step": 1122
+    },
+    {
+      "epoch": 0.2588153952523623,
+      "grad_norm": 0.31833314895629883,
+      "learning_rate": 8.690307149447221e-05,
+      "loss": 0.709,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25904586310209726,
+      "grad_norm": 0.30274853110313416,
+      "learning_rate": 8.68778742128935e-05,
+      "loss": 0.7057,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2592763309518322,
+      "grad_norm": 0.2601548135280609,
+      "learning_rate": 8.685265637647284e-05,
+      "loss": 0.7002,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2595067988015672,
+      "grad_norm": 0.26416724920272827,
+      "learning_rate": 8.682741799926609e-05,
+      "loss": 0.7109,
+      "step": 1126
+    },
+    {
+      "epoch": 0.25973726665130215,
+      "grad_norm": 0.3520905375480652,
+      "learning_rate": 8.680215909534044e-05,
+      "loss": 0.7183,
+      "step": 1127
+    },
+    {
+      "epoch": 0.2599677345010371,
+      "grad_norm": 0.2665805220603943,
+      "learning_rate": 8.677687967877466e-05,
+      "loss": 0.7108,
+      "step": 1128
+    },
+    {
+      "epoch": 0.26019820235077207,
+      "grad_norm": 0.3360787034034729,
+      "learning_rate": 8.675157976365886e-05,
+      "loss": 0.7033,
+      "step": 1129
+    },
+    {
+      "epoch": 0.26042867020050703,
+      "grad_norm": 0.28345534205436707,
+      "learning_rate": 8.672625936409462e-05,
+      "loss": 0.7068,
+      "step": 1130
+    },
+    {
+      "epoch": 0.260659138050242,
+      "grad_norm": 0.2835051119327545,
+      "learning_rate": 8.67009184941949e-05,
+      "loss": 0.7021,
+      "step": 1131
+    },
+    {
+      "epoch": 0.26088960589997695,
+      "grad_norm": 0.2675539553165436,
+      "learning_rate": 8.667555716808414e-05,
+      "loss": 0.7028,
+      "step": 1132
+    },
+    {
+      "epoch": 0.2611200737497119,
+      "grad_norm": 0.29232192039489746,
+      "learning_rate": 8.665017539989808e-05,
+      "loss": 0.7108,
+      "step": 1133
+    },
+    {
+      "epoch": 0.2613505415994469,
+      "grad_norm": 0.27097558975219727,
+      "learning_rate": 8.662477320378395e-05,
+      "loss": 0.7142,
+      "step": 1134
+    },
+    {
+      "epoch": 0.26158100944918183,
+      "grad_norm": 0.2470446676015854,
+      "learning_rate": 8.659935059390028e-05,
+      "loss": 0.7061,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2618114772989168,
+      "grad_norm": 0.2592596113681793,
+      "learning_rate": 8.657390758441708e-05,
+      "loss": 0.705,
+      "step": 1136
+    },
+    {
+      "epoch": 0.26204194514865176,
+      "grad_norm": 0.27617162466049194,
+      "learning_rate": 8.654844418951563e-05,
+      "loss": 0.7039,
+      "step": 1137
+    },
+    {
+      "epoch": 0.2622724129983867,
+      "grad_norm": 0.2819935977458954,
+      "learning_rate": 8.652296042338861e-05,
+      "loss": 0.7089,
+      "step": 1138
+    },
+    {
+      "epoch": 0.2625028808481217,
+      "grad_norm": 0.24662001430988312,
+      "learning_rate": 8.64974563002401e-05,
+      "loss": 0.7052,
+      "step": 1139
+    },
+    {
+      "epoch": 0.26273334869785664,
+      "grad_norm": 0.28374093770980835,
+      "learning_rate": 8.647193183428545e-05,
+      "loss": 0.7071,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2629638165475916,
+      "grad_norm": 0.2856753468513489,
+      "learning_rate": 8.64463870397514e-05,
+      "loss": 0.7074,
+      "step": 1141
+    },
+    {
+      "epoch": 0.26319428439732656,
+      "grad_norm": 0.2997211813926697,
+      "learning_rate": 8.6420821930876e-05,
+      "loss": 0.7084,
+      "step": 1142
+    },
+    {
+      "epoch": 0.2634247522470615,
+      "grad_norm": 0.26690179109573364,
+      "learning_rate": 8.639523652190864e-05,
+      "loss": 0.7025,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2636552200967965,
+      "grad_norm": 0.27999091148376465,
+      "learning_rate": 8.636963082710999e-05,
+      "loss": 0.7008,
+      "step": 1144
+    },
+    {
+      "epoch": 0.26388568794653144,
+      "grad_norm": 0.3172776997089386,
+      "learning_rate": 8.634400486075207e-05,
+      "loss": 0.7137,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2641161557962664,
+      "grad_norm": 0.30888575315475464,
+      "learning_rate": 8.631835863711817e-05,
+      "loss": 0.7105,
+      "step": 1146
+    },
+    {
+      "epoch": 0.26434662364600137,
+      "grad_norm": 0.28774750232696533,
+      "learning_rate": 8.629269217050289e-05,
+      "loss": 0.7057,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2645770914957363,
+      "grad_norm": 0.24954736232757568,
+      "learning_rate": 8.626700547521209e-05,
+      "loss": 0.7165,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2648075593454713,
+      "grad_norm": 0.2624035179615021,
+      "learning_rate": 8.624129856556291e-05,
+      "loss": 0.7006,
+      "step": 1149
+    },
+    {
+      "epoch": 0.26503802719520625,
+      "grad_norm": 0.2522975504398346,
+      "learning_rate": 8.62155714558838e-05,
+      "loss": 0.7096,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2652684950449412,
+      "grad_norm": 0.2908228039741516,
+      "learning_rate": 8.618982416051438e-05,
+      "loss": 0.7077,
+      "step": 1151
+    },
+    {
+      "epoch": 0.26549896289467617,
+      "grad_norm": 0.23220433294773102,
+      "learning_rate": 8.616405669380561e-05,
+      "loss": 0.7094,
+      "step": 1152
+    },
+    {
+      "epoch": 0.26572943074441113,
+      "grad_norm": 0.31848591566085815,
+      "learning_rate": 8.613826907011965e-05,
+      "loss": 0.7091,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2659598985941461,
+      "grad_norm": 0.2557988464832306,
+      "learning_rate": 8.611246130382992e-05,
+      "loss": 0.7028,
+      "step": 1154
+    },
+    {
+      "epoch": 0.26619036644388105,
+      "grad_norm": 0.24541644752025604,
+      "learning_rate": 8.608663340932104e-05,
+      "loss": 0.6998,
+      "step": 1155
+    },
+    {
+      "epoch": 0.266420834293616,
+      "grad_norm": 0.25283804535865784,
+      "learning_rate": 8.606078540098883e-05,
+      "loss": 0.7044,
+      "step": 1156
+    },
+    {
+      "epoch": 0.266651302143351,
+      "grad_norm": 0.24628406763076782,
+      "learning_rate": 8.603491729324039e-05,
+      "loss": 0.7136,
+      "step": 1157
+    },
+    {
+      "epoch": 0.26688176999308594,
+      "grad_norm": 0.24645006656646729,
+      "learning_rate": 8.600902910049393e-05,
+      "loss": 0.7039,
+      "step": 1158
+    },
+    {
+      "epoch": 0.2671122378428209,
+      "grad_norm": 0.23496174812316895,
+      "learning_rate": 8.598312083717896e-05,
+      "loss": 0.7146,
+      "step": 1159
+    },
+    {
+      "epoch": 0.2673427056925559,
+      "grad_norm": 0.26855039596557617,
+      "learning_rate": 8.59571925177361e-05,
+      "loss": 0.7036,
+      "step": 1160
+    },
+    {
+      "epoch": 0.2675731735422909,
+      "grad_norm": 0.21970762312412262,
+      "learning_rate": 8.59312441566172e-05,
+      "loss": 0.7109,
+      "step": 1161
+    },
+    {
+      "epoch": 0.26780364139202584,
+      "grad_norm": 0.2488812357187271,
+      "learning_rate": 8.59052757682852e-05,
+      "loss": 0.703,
+      "step": 1162
+    },
+    {
+      "epoch": 0.2680341092417608,
+      "grad_norm": 0.2618003487586975,
+      "learning_rate": 8.587928736721432e-05,
+      "loss": 0.695,
+      "step": 1163
+    },
+    {
+      "epoch": 0.26826457709149576,
+      "grad_norm": 0.25669237971305847,
+      "learning_rate": 8.585327896788981e-05,
+      "loss": 0.6989,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2684950449412307,
+      "grad_norm": 0.26360106468200684,
+      "learning_rate": 8.582725058480816e-05,
+      "loss": 0.7075,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2687255127909657,
+      "grad_norm": 0.24905282258987427,
+      "learning_rate": 8.580120223247697e-05,
+      "loss": 0.7065,
+      "step": 1166
+    },
+    {
+      "epoch": 0.26895598064070064,
+      "grad_norm": 0.343464195728302,
+      "learning_rate": 8.577513392541496e-05,
+      "loss": 0.6978,
+      "step": 1167
+    },
+    {
+      "epoch": 0.2691864484904356,
+      "grad_norm": 0.2586437463760376,
+      "learning_rate": 8.574904567815196e-05,
+      "loss": 0.7056,
+      "step": 1168
+    },
+    {
+      "epoch": 0.26941691634017056,
+      "grad_norm": 0.23977968096733093,
+      "learning_rate": 8.572293750522897e-05,
+      "loss": 0.7056,
+      "step": 1169
+    },
+    {
+      "epoch": 0.2696473841899055,
+      "grad_norm": 0.2433982640504837,
+      "learning_rate": 8.569680942119804e-05,
+      "loss": 0.7021,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2698778520396405,
+      "grad_norm": 0.26769572496414185,
+      "learning_rate": 8.567066144062232e-05,
+      "loss": 0.6999,
+      "step": 1171
+    },
+    {
+      "epoch": 0.27010831988937545,
+      "grad_norm": 0.24118681252002716,
+      "learning_rate": 8.564449357807608e-05,
+      "loss": 0.707,
+      "step": 1172
+    },
+    {
+      "epoch": 0.2703387877391104,
+      "grad_norm": 0.2884605824947357,
+      "learning_rate": 8.561830584814468e-05,
+      "loss": 0.7099,
+      "step": 1173
+    },
+    {
+      "epoch": 0.27056925558884537,
+      "grad_norm": 0.28720924258232117,
+      "learning_rate": 8.559209826542451e-05,
+      "loss": 0.6992,
+      "step": 1174
+    },
+    {
+      "epoch": 0.27079972343858033,
+      "grad_norm": 0.24960888922214508,
+      "learning_rate": 8.556587084452305e-05,
+      "loss": 0.7054,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2710301912883153,
+      "grad_norm": 0.2377004772424698,
+      "learning_rate": 8.553962360005882e-05,
+      "loss": 0.7014,
+      "step": 1176
+    },
+    {
+      "epoch": 0.27126065913805025,
+      "grad_norm": 0.2665834426879883,
+      "learning_rate": 8.551335654666144e-05,
+      "loss": 0.7019,
+      "step": 1177
+    },
+    {
+      "epoch": 0.2714911269877852,
+      "grad_norm": 0.25975316762924194,
+      "learning_rate": 8.54870696989715e-05,
+      "loss": 0.6971,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2717215948375202,
+      "grad_norm": 0.2380094677209854,
+      "learning_rate": 8.546076307164068e-05,
+      "loss": 0.6961,
+      "step": 1179
+    },
+    {
+      "epoch": 0.27195206268725514,
+      "grad_norm": 0.24472801387310028,
+      "learning_rate": 8.543443667933166e-05,
+      "loss": 0.7049,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2721825305369901,
+      "grad_norm": 0.2646014094352722,
+      "learning_rate": 8.540809053671812e-05,
+      "loss": 0.7148,
+      "step": 1181
+    },
+    {
+      "epoch": 0.27241299838672506,
+      "grad_norm": 0.23404592275619507,
+      "learning_rate": 8.538172465848479e-05,
+      "loss": 0.7094,
+      "step": 1182
+    },
+    {
+      "epoch": 0.27264346623646,
+      "grad_norm": 0.2830381691455841,
+      "learning_rate": 8.535533905932738e-05,
+      "loss": 0.7061,
+      "step": 1183
+    },
+    {
+      "epoch": 0.272873934086195,
+      "grad_norm": 0.2919880747795105,
+      "learning_rate": 8.532893375395257e-05,
+      "loss": 0.7036,
+      "step": 1184
+    },
+    {
+      "epoch": 0.27310440193592994,
+      "grad_norm": 0.24687440693378448,
+      "learning_rate": 8.530250875707806e-05,
+      "loss": 0.7004,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2733348697856649,
+      "grad_norm": 0.2443513572216034,
+      "learning_rate": 8.527606408343249e-05,
+      "loss": 0.6899,
+      "step": 1186
+    },
+    {
+      "epoch": 0.27356533763539986,
+      "grad_norm": 0.2695756256580353,
+      "learning_rate": 8.524959974775551e-05,
+      "loss": 0.7065,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2737958054851348,
+      "grad_norm": 0.24225404858589172,
+      "learning_rate": 8.522311576479768e-05,
+      "loss": 0.7019,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2740262733348698,
+      "grad_norm": 0.2715757191181183,
+      "learning_rate": 8.519661214932055e-05,
+      "loss": 0.7099,
+      "step": 1189
+    },
+    {
+      "epoch": 0.27425674118460475,
+      "grad_norm": 0.31206077337265015,
+      "learning_rate": 8.517008891609661e-05,
+      "loss": 0.7088,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2744872090343397,
+      "grad_norm": 0.2775088846683502,
+      "learning_rate": 8.514354607990926e-05,
+      "loss": 0.7031,
+      "step": 1191
+    },
+    {
+      "epoch": 0.27471767688407467,
+      "grad_norm": 0.2421555519104004,
+      "learning_rate": 8.511698365555285e-05,
+      "loss": 0.7112,
+      "step": 1192
+    },
+    {
+      "epoch": 0.27494814473380963,
+      "grad_norm": 0.31836438179016113,
+      "learning_rate": 8.509040165783263e-05,
+      "loss": 0.7034,
+      "step": 1193
+    },
+    {
+      "epoch": 0.2751786125835446,
+      "grad_norm": 0.3080047369003296,
+      "learning_rate": 8.506380010156476e-05,
+      "loss": 0.7053,
+      "step": 1194
+    },
+    {
+      "epoch": 0.27540908043327955,
+      "grad_norm": 0.2516467869281769,
+      "learning_rate": 8.503717900157632e-05,
+      "loss": 0.7077,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2756395482830145,
+      "grad_norm": 0.26254069805145264,
+      "learning_rate": 8.501053837270528e-05,
+      "loss": 0.7066,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2758700161327495,
+      "grad_norm": 0.3127571940422058,
+      "learning_rate": 8.498387822980046e-05,
+      "loss": 0.6957,
+      "step": 1197
+    },
+    {
+      "epoch": 0.27610048398248443,
+      "grad_norm": 0.2957381308078766,
+      "learning_rate": 8.49571985877216e-05,
+      "loss": 0.6919,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2763309518322194,
+      "grad_norm": 0.29202744364738464,
+      "learning_rate": 8.49304994613393e-05,
+      "loss": 0.7101,
+      "step": 1199
+    },
+    {
+      "epoch": 0.27656141968195436,
+      "grad_norm": 0.272582083940506,
+      "learning_rate": 8.490378086553499e-05,
+      "loss": 0.6988,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2767918875316893,
+      "grad_norm": 0.24930858612060547,
+      "learning_rate": 8.4877042815201e-05,
+      "loss": 0.699,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2770223553814243,
+      "grad_norm": 0.2611529231071472,
+      "learning_rate": 8.485028532524046e-05,
+      "loss": 0.7102,
+      "step": 1202
+    },
+    {
+      "epoch": 0.27725282323115924,
+      "grad_norm": 0.3048509359359741,
+      "learning_rate": 8.482350841056737e-05,
+      "loss": 0.7006,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2774832910808942,
+      "grad_norm": 0.26629912853240967,
+      "learning_rate": 8.479671208610653e-05,
+      "loss": 0.6953,
+      "step": 1204
+    },
+    {
+      "epoch": 0.27771375893062916,
+      "grad_norm": 0.27922797203063965,
+      "learning_rate": 8.476989636679355e-05,
+      "loss": 0.7101,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2779442267803641,
+      "grad_norm": 0.26804566383361816,
+      "learning_rate": 8.47430612675749e-05,
+      "loss": 0.7082,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2781746946300991,
+      "grad_norm": 0.3235054910182953,
+      "learning_rate": 8.471620680340779e-05,
+      "loss": 0.6958,
+      "step": 1207
+    },
+    {
+      "epoch": 0.27840516247983405,
+      "grad_norm": 0.260946124792099,
+      "learning_rate": 8.468933298926029e-05,
+      "loss": 0.6972,
+      "step": 1208
+    },
+    {
+      "epoch": 0.278635630329569,
+      "grad_norm": 0.2844315469264984,
+      "learning_rate": 8.466243984011117e-05,
+      "loss": 0.6987,
+      "step": 1209
+    },
+    {
+      "epoch": 0.27886609817930397,
+      "grad_norm": 0.2835547626018524,
+      "learning_rate": 8.463552737095007e-05,
+      "loss": 0.6999,
+      "step": 1210
+    },
+    {
+      "epoch": 0.27909656602903893,
+      "grad_norm": 0.2883650064468384,
+      "learning_rate": 8.460859559677734e-05,
+      "loss": 0.7028,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2793270338787739,
+      "grad_norm": 0.2849369943141937,
+      "learning_rate": 8.458164453260408e-05,
+      "loss": 0.7007,
+      "step": 1212
+    },
+    {
+      "epoch": 0.27955750172850885,
+      "grad_norm": 0.24277077615261078,
+      "learning_rate": 8.455467419345222e-05,
+      "loss": 0.698,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2797879695782438,
+      "grad_norm": 0.2855810821056366,
+      "learning_rate": 8.452768459435434e-05,
+      "loss": 0.6947,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2800184374279788,
+      "grad_norm": 0.2892487645149231,
+      "learning_rate": 8.450067575035378e-05,
+      "loss": 0.7041,
+      "step": 1215
+    },
+    {
+      "epoch": 0.28024890527771373,
+      "grad_norm": 0.2550940215587616,
+      "learning_rate": 8.447364767650468e-05,
+      "loss": 0.6982,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2804793731274487,
+      "grad_norm": 0.28303179144859314,
+      "learning_rate": 8.444660038787178e-05,
+      "loss": 0.6891,
+      "step": 1217
+    },
+    {
+      "epoch": 0.28070984097718366,
+      "grad_norm": 0.29353299736976624,
+      "learning_rate": 8.441953389953062e-05,
+      "loss": 0.699,
+      "step": 1218
+    },
+    {
+      "epoch": 0.28094030882691867,
+      "grad_norm": 0.2536478340625763,
+      "learning_rate": 8.43924482265674e-05,
+      "loss": 0.7046,
+      "step": 1219
+    },
+    {
+      "epoch": 0.28117077667665363,
+      "grad_norm": 0.32717081904411316,
+      "learning_rate": 8.436534338407902e-05,
+      "loss": 0.7079,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2814012445263886,
+      "grad_norm": 0.2975577116012573,
+      "learning_rate": 8.43382193871731e-05,
+      "loss": 0.7002,
+      "step": 1221
+    },
+    {
+      "epoch": 0.28163171237612356,
+      "grad_norm": 0.25119921565055847,
+      "learning_rate": 8.431107625096787e-05,
+      "loss": 0.702,
+      "step": 1222
+    },
+    {
+      "epoch": 0.2818621802258585,
+      "grad_norm": 0.30405279994010925,
+      "learning_rate": 8.428391399059228e-05,
+      "loss": 0.6982,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2820926480755935,
+      "grad_norm": 0.2723596692085266,
+      "learning_rate": 8.42567326211859e-05,
+      "loss": 0.7001,
+      "step": 1224
+    },
+    {
+      "epoch": 0.28232311592532844,
+      "grad_norm": 0.226847305893898,
+      "learning_rate": 8.4229532157899e-05,
+      "loss": 0.6981,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2825535837750634,
+      "grad_norm": 0.2864554226398468,
+      "learning_rate": 8.420231261589246e-05,
+      "loss": 0.7076,
+      "step": 1226
+    },
+    {
+      "epoch": 0.28278405162479836,
+      "grad_norm": 0.24330449104309082,
+      "learning_rate": 8.417507401033779e-05,
+      "loss": 0.6981,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2830145194745333,
+      "grad_norm": 0.3172352612018585,
+      "learning_rate": 8.414781635641714e-05,
+      "loss": 0.6957,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2832449873242683,
+      "grad_norm": 0.2853315770626068,
+      "learning_rate": 8.412053966932326e-05,
+      "loss": 0.7,
+      "step": 1229
+    },
+    {
+      "epoch": 0.28347545517400324,
+      "grad_norm": 0.3434127867221832,
+      "learning_rate": 8.409324396425954e-05,
+      "loss": 0.7085,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2837059230237382,
+      "grad_norm": 0.2804059088230133,
+      "learning_rate": 8.406592925643995e-05,
+      "loss": 0.7037,
+      "step": 1231
+    },
+    {
+      "epoch": 0.28393639087347317,
+      "grad_norm": 0.26557254791259766,
+      "learning_rate": 8.403859556108904e-05,
+      "loss": 0.699,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2841668587232081,
+      "grad_norm": 0.28296953439712524,
+      "learning_rate": 8.401124289344195e-05,
+      "loss": 0.6907,
+      "step": 1233
+    },
+    {
+      "epoch": 0.2843973265729431,
+      "grad_norm": 0.2905629873275757,
+      "learning_rate": 8.398387126874443e-05,
+      "loss": 0.6962,
+      "step": 1234
+    },
+    {
+      "epoch": 0.28462779442267805,
+      "grad_norm": 0.2673647403717041,
+      "learning_rate": 8.395648070225272e-05,
+      "loss": 0.6941,
+      "step": 1235
+    },
+    {
+      "epoch": 0.284858262272413,
+      "grad_norm": 0.30280497670173645,
+      "learning_rate": 8.392907120923373e-05,
+      "loss": 0.6993,
+      "step": 1236
+    },
+    {
+      "epoch": 0.28508873012214797,
+      "grad_norm": 0.28587183356285095,
+      "learning_rate": 8.39016428049648e-05,
+      "loss": 0.696,
+      "step": 1237
+    },
+    {
+      "epoch": 0.28531919797188293,
+      "grad_norm": 0.24387072026729584,
+      "learning_rate": 8.387419550473387e-05,
+      "loss": 0.7027,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2855496658216179,
+      "grad_norm": 0.2583426237106323,
+      "learning_rate": 8.384672932383942e-05,
+      "loss": 0.701,
+      "step": 1239
+    },
+    {
+      "epoch": 0.28578013367135285,
+      "grad_norm": 0.2734989821910858,
+      "learning_rate": 8.381924427759044e-05,
+      "loss": 0.6959,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2860106015210878,
+      "grad_norm": 0.2484637349843979,
+      "learning_rate": 8.379174038130643e-05,
+      "loss": 0.6997,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2862410693708228,
+      "grad_norm": 0.2433558702468872,
+      "learning_rate": 8.376421765031741e-05,
+      "loss": 0.6998,
+      "step": 1242
+    },
+    {
+      "epoch": 0.28647153722055774,
+      "grad_norm": 0.2775626480579376,
+      "learning_rate": 8.373667609996387e-05,
+      "loss": 0.6962,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2867020050702927,
+      "grad_norm": 0.26966193318367004,
+      "learning_rate": 8.370911574559683e-05,
+      "loss": 0.7024,
+      "step": 1244
+    },
+    {
+      "epoch": 0.28693247292002766,
+      "grad_norm": 0.2636489272117615,
+      "learning_rate": 8.368153660257776e-05,
+      "loss": 0.7041,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2871629407697626,
+      "grad_norm": 0.2798607051372528,
+      "learning_rate": 8.365393868627862e-05,
+      "loss": 0.6912,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2873934086194976,
+      "grad_norm": 0.26497724652290344,
+      "learning_rate": 8.362632201208181e-05,
+      "loss": 0.6995,
+      "step": 1247
+    },
+    {
+      "epoch": 0.28762387646923254,
+      "grad_norm": 0.26089614629745483,
+      "learning_rate": 8.359868659538022e-05,
+      "loss": 0.6994,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2878543443189675,
+      "grad_norm": 0.2735275328159332,
+      "learning_rate": 8.357103245157715e-05,
+      "loss": 0.6995,
+      "step": 1249
+    },
+    {
+      "epoch": 0.28808481216870246,
+      "grad_norm": 0.27281323075294495,
+      "learning_rate": 8.354335959608638e-05,
+      "loss": 0.6947,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2883152800184374,
+      "grad_norm": 0.2844576835632324,
+      "learning_rate": 8.351566804433207e-05,
+      "loss": 0.7073,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2885457478681724,
+      "grad_norm": 0.2630547285079956,
+      "learning_rate": 8.348795781174885e-05,
+      "loss": 0.7021,
+      "step": 1252
+    },
+    {
+      "epoch": 0.28877621571790735,
+      "grad_norm": 0.23190078139305115,
+      "learning_rate": 8.346022891378172e-05,
+      "loss": 0.7013,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2890066835676423,
+      "grad_norm": 0.28771939873695374,
+      "learning_rate": 8.34324813658861e-05,
+      "loss": 0.7027,
+      "step": 1254
+    },
+    {
+      "epoch": 0.28923715141737727,
+      "grad_norm": 0.2860211133956909,
+      "learning_rate": 8.340471518352781e-05,
+      "loss": 0.7008,
+      "step": 1255
+    },
+    {
+      "epoch": 0.28946761926711223,
+      "grad_norm": 0.26749560236930847,
+      "learning_rate": 8.337693038218308e-05,
+      "loss": 0.7003,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2896980871168472,
+      "grad_norm": 0.22241248190402985,
+      "learning_rate": 8.334912697733845e-05,
+      "loss": 0.7045,
+      "step": 1257
+    },
+    {
+      "epoch": 0.28992855496658215,
+      "grad_norm": 0.2874812185764313,
+      "learning_rate": 8.33213049844909e-05,
+      "loss": 0.696,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2901590228163171,
+      "grad_norm": 0.28802555799484253,
+      "learning_rate": 8.329346441914774e-05,
+      "loss": 0.6997,
+      "step": 1259
+    },
+    {
+      "epoch": 0.2903894906660521,
+      "grad_norm": 0.24786509573459625,
+      "learning_rate": 8.326560529682661e-05,
+      "loss": 0.6983,
+      "step": 1260
+    },
+    {
+      "epoch": 0.29061995851578704,
+      "grad_norm": 0.2500215768814087,
+      "learning_rate": 8.323772763305554e-05,
+      "loss": 0.7038,
+      "step": 1261
+    },
+    {
+      "epoch": 0.290850426365522,
+      "grad_norm": 0.22261877357959747,
+      "learning_rate": 8.320983144337286e-05,
+      "loss": 0.6928,
+      "step": 1262
+    },
+    {
+      "epoch": 0.29108089421525696,
+      "grad_norm": 0.26974332332611084,
+      "learning_rate": 8.318191674332724e-05,
+      "loss": 0.7024,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2913113620649919,
+      "grad_norm": 0.2661108374595642,
+      "learning_rate": 8.315398354847766e-05,
+      "loss": 0.6996,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2915418299147269,
+      "grad_norm": 0.2478044480085373,
+      "learning_rate": 8.31260318743934e-05,
+      "loss": 0.6913,
+      "step": 1265
+    },
+    {
+      "epoch": 0.29177229776446184,
+      "grad_norm": 0.28388431668281555,
+      "learning_rate": 8.309806173665409e-05,
+      "loss": 0.6914,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2920027656141968,
+      "grad_norm": 0.26973721385002136,
+      "learning_rate": 8.307007315084958e-05,
+      "loss": 0.706,
+      "step": 1267
+    },
+    {
+      "epoch": 0.29223323346393176,
+      "grad_norm": 0.2543336749076843,
+      "learning_rate": 8.304206613258003e-05,
+      "loss": 0.6988,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2924637013136667,
+      "grad_norm": 0.2953673303127289,
+      "learning_rate": 8.301404069745592e-05,
+      "loss": 0.6876,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2926941691634017,
+      "grad_norm": 0.24503032863140106,
+      "learning_rate": 8.298599686109792e-05,
+      "loss": 0.6943,
+      "step": 1270
+    },
+    {
+      "epoch": 0.29292463701313665,
+      "grad_norm": 0.2405981868505478,
+      "learning_rate": 8.2957934639137e-05,
+      "loss": 0.7012,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2931551048628716,
+      "grad_norm": 0.23449143767356873,
+      "learning_rate": 8.292985404721438e-05,
+      "loss": 0.6919,
+      "step": 1272
+    },
+    {
+      "epoch": 0.29338557271260657,
+      "grad_norm": 0.2791789472103119,
+      "learning_rate": 8.29017551009815e-05,
+      "loss": 0.6987,
+      "step": 1273
+    },
+    {
+      "epoch": 0.29361604056234153,
+      "grad_norm": 0.26588696241378784,
+      "learning_rate": 8.287363781610003e-05,
+      "loss": 0.7009,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2938465084120765,
+      "grad_norm": 0.2526492178440094,
+      "learning_rate": 8.284550220824187e-05,
+      "loss": 0.7001,
+      "step": 1275
+    },
+    {
+      "epoch": 0.29407697626181145,
+      "grad_norm": 0.23610974848270416,
+      "learning_rate": 8.281734829308914e-05,
+      "loss": 0.6982,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2943074441115464,
+      "grad_norm": 0.24589475989341736,
+      "learning_rate": 8.278917608633416e-05,
+      "loss": 0.7077,
+      "step": 1277
+    },
+    {
+      "epoch": 0.29453791196128143,
+      "grad_norm": 0.25068679451942444,
+      "learning_rate": 8.276098560367944e-05,
+      "loss": 0.6999,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2947683798110164,
+      "grad_norm": 0.24742679297924042,
+      "learning_rate": 8.273277686083767e-05,
+      "loss": 0.7018,
+      "step": 1279
+    },
+    {
+      "epoch": 0.29499884766075135,
+      "grad_norm": 0.25572851300239563,
+      "learning_rate": 8.270454987353173e-05,
+      "loss": 0.7108,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2952293155104863,
+      "grad_norm": 0.2460428923368454,
+      "learning_rate": 8.267630465749467e-05,
+      "loss": 0.6992,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2954597833602213,
+      "grad_norm": 0.2824234068393707,
+      "learning_rate": 8.264804122846969e-05,
+      "loss": 0.6957,
+      "step": 1282
+    },
+    {
+      "epoch": 0.29569025120995623,
+      "grad_norm": 0.23757845163345337,
+      "learning_rate": 8.261975960221017e-05,
+      "loss": 0.6955,
+      "step": 1283
+    },
+    {
+      "epoch": 0.2959207190596912,
+      "grad_norm": 0.2634974718093872,
+      "learning_rate": 8.259145979447957e-05,
+      "loss": 0.6982,
+      "step": 1284
+    },
+    {
+      "epoch": 0.29615118690942616,
+      "grad_norm": 0.26144111156463623,
+      "learning_rate": 8.256314182105155e-05,
+      "loss": 0.7052,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2963816547591611,
+      "grad_norm": 0.2734619081020355,
+      "learning_rate": 8.253480569770986e-05,
+      "loss": 0.705,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2966121226088961,
+      "grad_norm": 0.28824108839035034,
+      "learning_rate": 8.250645144024838e-05,
+      "loss": 0.7058,
+      "step": 1287
+    },
+    {
+      "epoch": 0.29684259045863104,
+      "grad_norm": 0.2391139417886734,
+      "learning_rate": 8.247807906447108e-05,
+      "loss": 0.6906,
+      "step": 1288
+    },
+    {
+      "epoch": 0.297073058308366,
+      "grad_norm": 0.2693355977535248,
+      "learning_rate": 8.244968858619206e-05,
+      "loss": 0.7039,
+      "step": 1289
+    },
+    {
+      "epoch": 0.29730352615810096,
+      "grad_norm": 0.2826348841190338,
+      "learning_rate": 8.242128002123548e-05,
+      "loss": 0.696,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2975339940078359,
+      "grad_norm": 0.22388167679309845,
+      "learning_rate": 8.239285338543558e-05,
+      "loss": 0.7022,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2977644618575709,
+      "grad_norm": 0.29318729043006897,
+      "learning_rate": 8.236440869463671e-05,
+      "loss": 0.6948,
+      "step": 1292
+    },
+    {
+      "epoch": 0.29799492970730584,
+      "grad_norm": 0.28733378648757935,
+      "learning_rate": 8.233594596469325e-05,
+      "loss": 0.7068,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2982253975570408,
+      "grad_norm": 0.2567828893661499,
+      "learning_rate": 8.230746521146963e-05,
+      "loss": 0.6986,
+      "step": 1294
+    },
+    {
+      "epoch": 0.29845586540677577,
+      "grad_norm": 0.2597467601299286,
+      "learning_rate": 8.227896645084036e-05,
+      "loss": 0.685,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2986863332565107,
+      "grad_norm": 0.26486629247665405,
+      "learning_rate": 8.225044969868994e-05,
+      "loss": 0.694,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2989168011062457,
+      "grad_norm": 0.2641950249671936,
+      "learning_rate": 8.222191497091294e-05,
+      "loss": 0.7005,
+      "step": 1297
+    },
+    {
+      "epoch": 0.29914726895598065,
+      "grad_norm": 0.3094748556613922,
+      "learning_rate": 8.219336228341392e-05,
+      "loss": 0.7035,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2993777368057156,
+      "grad_norm": 0.2441185563802719,
+      "learning_rate": 8.216479165210748e-05,
+      "loss": 0.6908,
+      "step": 1299
+    },
+    {
+      "epoch": 0.29960820465545057,
+      "grad_norm": 0.23661471903324127,
+      "learning_rate": 8.213620309291819e-05,
+      "loss": 0.7005,
+      "step": 1300
+    },
+    {
+      "epoch": 0.29983867250518553,
+      "grad_norm": 0.3035421371459961,
+      "learning_rate": 8.210759662178064e-05,
+      "loss": 0.6909,
+      "step": 1301
+    },
+    {
+      "epoch": 0.3000691403549205,
+      "grad_norm": 0.2798277735710144,
+      "learning_rate": 8.207897225463938e-05,
+      "loss": 0.6936,
+      "step": 1302
+    },
+    {
+      "epoch": 0.30029960820465545,
+      "grad_norm": 0.232849583029747,
+      "learning_rate": 8.205033000744894e-05,
+      "loss": 0.6943,
+      "step": 1303
+    },
+    {
+      "epoch": 0.3005300760543904,
+      "grad_norm": 0.26353907585144043,
+      "learning_rate": 8.202166989617383e-05,
+      "loss": 0.6917,
+      "step": 1304
+    },
+    {
+      "epoch": 0.3007605439041254,
+      "grad_norm": 0.27309128642082214,
+      "learning_rate": 8.199299193678851e-05,
+      "loss": 0.702,
+      "step": 1305
+    },
+    {
+      "epoch": 0.30099101175386034,
+      "grad_norm": 0.2615521252155304,
+      "learning_rate": 8.196429614527737e-05,
+      "loss": 0.6962,
+      "step": 1306
+    },
+    {
+      "epoch": 0.3012214796035953,
+      "grad_norm": 0.25091928243637085,
+      "learning_rate": 8.193558253763478e-05,
+      "loss": 0.6988,
+      "step": 1307
+    },
+    {
+      "epoch": 0.30145194745333026,
+      "grad_norm": 0.25908082723617554,
+      "learning_rate": 8.1906851129865e-05,
+      "loss": 0.6996,
+      "step": 1308
+    },
+    {
+      "epoch": 0.3016824153030652,
+      "grad_norm": 0.2851544916629791,
+      "learning_rate": 8.187810193798223e-05,
+      "loss": 0.6932,
+      "step": 1309
+    },
+    {
+      "epoch": 0.3019128831528002,
+      "grad_norm": 0.2424258440732956,
+      "learning_rate": 8.184933497801059e-05,
+      "loss": 0.6926,
+      "step": 1310
+    },
+    {
+      "epoch": 0.30214335100253514,
+      "grad_norm": 0.24669015407562256,
+      "learning_rate": 8.182055026598408e-05,
+      "loss": 0.701,
+      "step": 1311
+    },
+    {
+      "epoch": 0.3023738188522701,
+      "grad_norm": 0.27151423692703247,
+      "learning_rate": 8.179174781794658e-05,
+      "loss": 0.6909,
+      "step": 1312
+    },
+    {
+      "epoch": 0.30260428670200507,
+      "grad_norm": 0.2658521831035614,
+      "learning_rate": 8.176292764995195e-05,
+      "loss": 0.6906,
+      "step": 1313
+    },
+    {
+      "epoch": 0.30283475455174,
+      "grad_norm": 0.2563072443008423,
+      "learning_rate": 8.173408977806382e-05,
+      "loss": 0.6948,
+      "step": 1314
+    },
+    {
+      "epoch": 0.303065222401475,
+      "grad_norm": 0.27026668190956116,
+      "learning_rate": 8.170523421835572e-05,
+      "loss": 0.6986,
+      "step": 1315
+    },
+    {
+      "epoch": 0.30329569025120995,
+      "grad_norm": 0.2493722289800644,
+      "learning_rate": 8.167636098691105e-05,
+      "loss": 0.695,
+      "step": 1316
+    },
+    {
+      "epoch": 0.3035261581009449,
+      "grad_norm": 0.46334898471832275,
+      "learning_rate": 8.164747009982308e-05,
+      "loss": 0.6948,
+      "step": 1317
+    },
+    {
+      "epoch": 0.30375662595067987,
+      "grad_norm": 0.3177790343761444,
+      "learning_rate": 8.161856157319488e-05,
+      "loss": 0.692,
+      "step": 1318
+    },
+    {
+      "epoch": 0.30398709380041483,
+      "grad_norm": 0.25527554750442505,
+      "learning_rate": 8.158963542313937e-05,
+      "loss": 0.7024,
+      "step": 1319
+    },
+    {
+      "epoch": 0.3042175616501498,
+      "grad_norm": 0.2921774983406067,
+      "learning_rate": 8.156069166577928e-05,
+      "loss": 0.7102,
+      "step": 1320
+    },
+    {
+      "epoch": 0.30444802949988475,
+      "grad_norm": 0.3124329149723053,
+      "learning_rate": 8.153173031724719e-05,
+      "loss": 0.6898,
+      "step": 1321
+    },
+    {
+      "epoch": 0.3046784973496197,
+      "grad_norm": 0.2984555959701538,
+      "learning_rate": 8.150275139368544e-05,
+      "loss": 0.6944,
+      "step": 1322
+    },
+    {
+      "epoch": 0.3049089651993547,
+      "grad_norm": 0.2839076817035675,
+      "learning_rate": 8.14737549112462e-05,
+      "loss": 0.6978,
+      "step": 1323
+    },
+    {
+      "epoch": 0.30513943304908964,
+      "grad_norm": 0.25131693482398987,
+      "learning_rate": 8.144474088609141e-05,
+      "loss": 0.6931,
+      "step": 1324
+    },
+    {
+      "epoch": 0.3053699008988246,
+      "grad_norm": 0.2686595916748047,
+      "learning_rate": 8.141570933439277e-05,
+      "loss": 0.6873,
+      "step": 1325
+    },
+    {
+      "epoch": 0.30560036874855956,
+      "grad_norm": 0.2873900830745697,
+      "learning_rate": 8.138666027233181e-05,
+      "loss": 0.6959,
+      "step": 1326
+    },
+    {
+      "epoch": 0.3058308365982945,
+      "grad_norm": 0.26454657316207886,
+      "learning_rate": 8.135759371609976e-05,
+      "loss": 0.6977,
+      "step": 1327
+    },
+    {
+      "epoch": 0.3060613044480295,
+      "grad_norm": 0.27393782138824463,
+      "learning_rate": 8.132850968189762e-05,
+      "loss": 0.6977,
+      "step": 1328
+    },
+    {
+      "epoch": 0.30629177229776444,
+      "grad_norm": 0.29595717787742615,
+      "learning_rate": 8.129940818593612e-05,
+      "loss": 0.6941,
+      "step": 1329
+    },
+    {
+      "epoch": 0.3065222401474994,
+      "grad_norm": 0.2389097660779953,
+      "learning_rate": 8.127028924443573e-05,
+      "loss": 0.7001,
+      "step": 1330
+    },
+    {
+      "epoch": 0.30675270799723436,
+      "grad_norm": 0.7923142910003662,
+      "learning_rate": 8.12411528736267e-05,
+      "loss": 0.6915,
+      "step": 1331
+    },
+    {
+      "epoch": 0.3069831758469693,
+      "grad_norm": 0.40648335218429565,
+      "learning_rate": 8.12119990897489e-05,
+      "loss": 0.6905,
+      "step": 1332
+    },
+    {
+      "epoch": 0.3072136436967043,
+      "grad_norm": 0.2911185622215271,
+      "learning_rate": 8.118282790905197e-05,
+      "loss": 0.6953,
+      "step": 1333
+    },
+    {
+      "epoch": 0.30744411154643925,
+      "grad_norm": 0.2990292012691498,
+      "learning_rate": 8.11536393477952e-05,
+      "loss": 0.6953,
+      "step": 1334
+    },
+    {
+      "epoch": 0.3076745793961742,
+      "grad_norm": 0.28816699981689453,
+      "learning_rate": 8.112443342224761e-05,
+      "loss": 0.6954,
+      "step": 1335
+    },
+    {
+      "epoch": 0.30790504724590917,
+      "grad_norm": 0.32540613412857056,
+      "learning_rate": 8.109521014868789e-05,
+      "loss": 0.6935,
+      "step": 1336
+    },
+    {
+      "epoch": 0.30813551509564413,
+      "grad_norm": 0.29871416091918945,
+      "learning_rate": 8.106596954340438e-05,
+      "loss": 0.7029,
+      "step": 1337
+    },
+    {
+      "epoch": 0.30836598294537915,
+      "grad_norm": 0.29098618030548096,
+      "learning_rate": 8.10367116226951e-05,
+      "loss": 0.7083,
+      "step": 1338
+    },
+    {
+      "epoch": 0.3085964507951141,
+      "grad_norm": 0.27469146251678467,
+      "learning_rate": 8.100743640286768e-05,
+      "loss": 0.7025,
+      "step": 1339
+    },
+    {
+      "epoch": 0.30882691864484907,
+      "grad_norm": 0.34525662660598755,
+      "learning_rate": 8.097814390023947e-05,
+      "loss": 0.6983,
+      "step": 1340
+    },
+    {
+      "epoch": 0.30905738649458403,
+      "grad_norm": 0.3090488314628601,
+      "learning_rate": 8.09488341311374e-05,
+      "loss": 0.7003,
+      "step": 1341
+    },
+    {
+      "epoch": 0.309287854344319,
+      "grad_norm": 0.2643933892250061,
+      "learning_rate": 8.091950711189801e-05,
+      "loss": 0.689,
+      "step": 1342
+    },
+    {
+      "epoch": 0.30951832219405395,
+      "grad_norm": 0.3033878803253174,
+      "learning_rate": 8.089016285886748e-05,
+      "loss": 0.6958,
+      "step": 1343
+    },
+    {
+      "epoch": 0.3097487900437889,
+      "grad_norm": 0.29167822003364563,
+      "learning_rate": 8.086080138840162e-05,
+      "loss": 0.6976,
+      "step": 1344
+    },
+    {
+      "epoch": 0.3099792578935239,
+      "grad_norm": 0.29147741198539734,
+      "learning_rate": 8.083142271686577e-05,
+      "loss": 0.6985,
+      "step": 1345
+    },
+    {
+      "epoch": 0.31020972574325884,
+      "grad_norm": 0.2536722719669342,
+      "learning_rate": 8.080202686063492e-05,
+      "loss": 0.7038,
+      "step": 1346
+    },
+    {
+      "epoch": 0.3104401935929938,
+      "grad_norm": 0.2939794361591339,
+      "learning_rate": 8.077261383609363e-05,
+      "loss": 0.6915,
+      "step": 1347
+    },
+    {
+      "epoch": 0.31067066144272876,
+      "grad_norm": 0.260345458984375,
+      "learning_rate": 8.074318365963597e-05,
+      "loss": 0.6942,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3109011292924637,
+      "grad_norm": 0.25490236282348633,
+      "learning_rate": 8.071373634766563e-05,
+      "loss": 0.6947,
+      "step": 1349
+    },
+    {
+      "epoch": 0.3111315971421987,
+      "grad_norm": 0.2755090892314911,
+      "learning_rate": 8.068427191659586e-05,
+      "loss": 0.7035,
+      "step": 1350
+    },
+    {
+      "epoch": 0.31136206499193364,
+      "grad_norm": 0.2627999484539032,
+      "learning_rate": 8.065479038284942e-05,
+      "loss": 0.6992,
+      "step": 1351
+    },
+    {
+      "epoch": 0.3115925328416686,
+      "grad_norm": 0.25387728214263916,
+      "learning_rate": 8.062529176285858e-05,
+      "loss": 0.7024,
+      "step": 1352
+    },
+    {
+      "epoch": 0.31182300069140356,
+      "grad_norm": 0.27479302883148193,
+      "learning_rate": 8.059577607306518e-05,
+      "loss": 0.6976,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3120534685411385,
+      "grad_norm": 0.23171396553516388,
+      "learning_rate": 8.056624332992057e-05,
+      "loss": 0.6977,
+      "step": 1354
+    },
+    {
+      "epoch": 0.3122839363908735,
+      "grad_norm": 0.26171013712882996,
+      "learning_rate": 8.05366935498856e-05,
+      "loss": 0.6927,
+      "step": 1355
+    },
+    {
+      "epoch": 0.31251440424060845,
+      "grad_norm": 0.27590593695640564,
+      "learning_rate": 8.050712674943055e-05,
+      "loss": 0.6868,
+      "step": 1356
+    },
+    {
+      "epoch": 0.3127448720903434,
+      "grad_norm": 0.2615238428115845,
+      "learning_rate": 8.047754294503531e-05,
+      "loss": 0.6913,
+      "step": 1357
+    },
+    {
+      "epoch": 0.31297533994007837,
+      "grad_norm": 0.2482849508523941,
+      "learning_rate": 8.044794215318916e-05,
+      "loss": 0.6949,
+      "step": 1358
+    },
+    {
+      "epoch": 0.31320580778981333,
+      "grad_norm": 0.26599544286727905,
+      "learning_rate": 8.041832439039084e-05,
+      "loss": 0.6972,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3134362756395483,
+      "grad_norm": 0.2676306962966919,
+      "learning_rate": 8.038868967314863e-05,
+      "loss": 0.7023,
+      "step": 1360
+    },
+    {
+      "epoch": 0.31366674348928325,
+      "grad_norm": 0.2429536134004593,
+      "learning_rate": 8.035903801798018e-05,
+      "loss": 0.691,
+      "step": 1361
+    },
+    {
+      "epoch": 0.3138972113390182,
+      "grad_norm": 0.264345645904541,
+      "learning_rate": 8.032936944141261e-05,
+      "loss": 0.6948,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3141276791887532,
+      "grad_norm": 0.2679714262485504,
+      "learning_rate": 8.02996839599825e-05,
+      "loss": 0.6977,
+      "step": 1363
+    },
+    {
+      "epoch": 0.31435814703848813,
+      "grad_norm": 0.28790920972824097,
+      "learning_rate": 8.02699815902358e-05,
+      "loss": 0.6931,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3145886148882231,
+      "grad_norm": 0.23752626776695251,
+      "learning_rate": 8.02402623487279e-05,
+      "loss": 0.692,
+      "step": 1365
+    },
+    {
+      "epoch": 0.31481908273795806,
+      "grad_norm": 0.25893813371658325,
+      "learning_rate": 8.021052625202359e-05,
+      "loss": 0.7112,
+      "step": 1366
+    },
+    {
+      "epoch": 0.315049550587693,
+      "grad_norm": 0.29539117217063904,
+      "learning_rate": 8.018077331669706e-05,
+      "loss": 0.6844,
+      "step": 1367
+    },
+    {
+      "epoch": 0.315280018437428,
+      "grad_norm": 0.2858338952064514,
+      "learning_rate": 8.015100355933189e-05,
+      "loss": 0.6944,
+      "step": 1368
+    },
+    {
+      "epoch": 0.31551048628716294,
+      "grad_norm": 0.24025440216064453,
+      "learning_rate": 8.012121699652103e-05,
+      "loss": 0.6995,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3157409541368979,
+      "grad_norm": 0.2733021676540375,
+      "learning_rate": 8.009141364486679e-05,
+      "loss": 0.6892,
+      "step": 1370
+    },
+    {
+      "epoch": 0.31597142198663286,
+      "grad_norm": 0.286783903837204,
+      "learning_rate": 8.006159352098082e-05,
+      "loss": 0.6881,
+      "step": 1371
+    },
+    {
+      "epoch": 0.3162018898363678,
+      "grad_norm": 0.2588256001472473,
+      "learning_rate": 8.003175664148416e-05,
+      "loss": 0.6927,
+      "step": 1372
+    },
+    {
+      "epoch": 0.3164323576861028,
+      "grad_norm": 0.26540815830230713,
+      "learning_rate": 8.000190302300721e-05,
+      "loss": 0.6959,
+      "step": 1373
+    },
+    {
+      "epoch": 0.31666282553583774,
+      "grad_norm": 0.2769984006881714,
+      "learning_rate": 7.99720326821896e-05,
+      "loss": 0.6986,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3168932933855727,
+      "grad_norm": 0.23313313722610474,
+      "learning_rate": 7.994214563568036e-05,
+      "loss": 0.6915,
+      "step": 1375
+    },
+    {
+      "epoch": 0.31712376123530767,
+      "grad_norm": 0.24230745434761047,
+      "learning_rate": 7.991224190013782e-05,
+      "loss": 0.7006,
+      "step": 1376
+    },
+    {
+      "epoch": 0.3173542290850426,
+      "grad_norm": 0.2684518098831177,
+      "learning_rate": 7.988232149222959e-05,
+      "loss": 0.6968,
+      "step": 1377
+    },
+    {
+      "epoch": 0.3175846969347776,
+      "grad_norm": 0.24516130983829498,
+      "learning_rate": 7.985238442863261e-05,
+      "loss": 0.6957,
+      "step": 1378
+    },
+    {
+      "epoch": 0.31781516478451255,
+      "grad_norm": 0.22360306978225708,
+      "learning_rate": 7.982243072603306e-05,
+      "loss": 0.7003,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3180456326342475,
+      "grad_norm": 0.2547612190246582,
+      "learning_rate": 7.979246040112643e-05,
+      "loss": 0.6904,
+      "step": 1380
+    },
+    {
+      "epoch": 0.31827610048398247,
+      "grad_norm": 0.3107689619064331,
+      "learning_rate": 7.976247347061745e-05,
+      "loss": 0.6902,
+      "step": 1381
+    },
+    {
+      "epoch": 0.31850656833371743,
+      "grad_norm": 0.2589125335216522,
+      "learning_rate": 7.973246995122013e-05,
+      "loss": 0.6908,
+      "step": 1382
+    },
+    {
+      "epoch": 0.3187370361834524,
+      "grad_norm": 0.24334730207920074,
+      "learning_rate": 7.970244985965767e-05,
+      "loss": 0.6942,
+      "step": 1383
+    },
+    {
+      "epoch": 0.31896750403318735,
+      "grad_norm": 0.249411478638649,
+      "learning_rate": 7.96724132126626e-05,
+      "loss": 0.6968,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3191979718829223,
+      "grad_norm": 0.29693132638931274,
+      "learning_rate": 7.964236002697661e-05,
+      "loss": 0.6869,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3194284397326573,
+      "grad_norm": 0.2691076993942261,
+      "learning_rate": 7.96122903193506e-05,
+      "loss": 0.6914,
+      "step": 1386
+    },
+    {
+      "epoch": 0.31965890758239224,
+      "grad_norm": 0.2546314597129822,
+      "learning_rate": 7.958220410654475e-05,
+      "loss": 0.6944,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3198893754321272,
+      "grad_norm": 0.29681459069252014,
+      "learning_rate": 7.955210140532833e-05,
+      "loss": 0.6801,
+      "step": 1388
+    },
+    {
+      "epoch": 0.32011984328186216,
+      "grad_norm": 0.24881964921951294,
+      "learning_rate": 7.952198223247993e-05,
+      "loss": 0.6966,
+      "step": 1389
+    },
+    {
+      "epoch": 0.3203503111315971,
+      "grad_norm": 0.2598433196544647,
+      "learning_rate": 7.949184660478721e-05,
+      "loss": 0.6955,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3205807789813321,
+      "grad_norm": 0.2911115884780884,
+      "learning_rate": 7.946169453904706e-05,
+      "loss": 0.696,
+      "step": 1391
+    },
+    {
+      "epoch": 0.32081124683106704,
+      "grad_norm": 0.2831457853317261,
+      "learning_rate": 7.943152605206551e-05,
+      "loss": 0.6951,
+      "step": 1392
+    },
+    {
+      "epoch": 0.321041714680802,
+      "grad_norm": 0.25077253580093384,
+      "learning_rate": 7.940134116065776e-05,
+      "loss": 0.691,
+      "step": 1393
+    },
+    {
+      "epoch": 0.32127218253053696,
+      "grad_norm": 0.2983539402484894,
+      "learning_rate": 7.937113988164814e-05,
+      "loss": 0.69,
+      "step": 1394
+    },
+    {
+      "epoch": 0.3215026503802719,
+      "grad_norm": 0.28595560789108276,
+      "learning_rate": 7.93409222318701e-05,
+      "loss": 0.6948,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3217331182300069,
+      "grad_norm": 0.2612264156341553,
+      "learning_rate": 7.931068822816627e-05,
+      "loss": 0.6897,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3219635860797419,
+      "grad_norm": 0.2846459150314331,
+      "learning_rate": 7.928043788738835e-05,
+      "loss": 0.7021,
+      "step": 1397
+    },
+    {
+      "epoch": 0.32219405392947686,
+      "grad_norm": 0.26386579871177673,
+      "learning_rate": 7.925017122639712e-05,
+      "loss": 0.6866,
+      "step": 1398
+    },
+    {
+      "epoch": 0.3224245217792118,
+      "grad_norm": 0.24749140441417694,
+      "learning_rate": 7.921988826206252e-05,
+      "loss": 0.6953,
+      "step": 1399
+    },
+    {
+      "epoch": 0.3226549896289468,
+      "grad_norm": 0.24181775748729706,
+      "learning_rate": 7.918958901126354e-05,
+      "loss": 0.6928,
+      "step": 1400
+    },
+    {
+      "epoch": 0.32288545747868175,
+      "grad_norm": 0.26094570755958557,
+      "learning_rate": 7.915927349088825e-05,
+      "loss": 0.6997,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3231159253284167,
+      "grad_norm": 0.25756776332855225,
+      "learning_rate": 7.912894171783383e-05,
+      "loss": 0.6834,
+      "step": 1402
+    },
+    {
+      "epoch": 0.32334639317815167,
+      "grad_norm": 0.2835976183414459,
+      "learning_rate": 7.909859370900642e-05,
+      "loss": 0.6971,
+      "step": 1403
+    },
+    {
+      "epoch": 0.32357686102788663,
+      "grad_norm": 0.27739542722702026,
+      "learning_rate": 7.906822948132131e-05,
+      "loss": 0.6931,
+      "step": 1404
+    },
+    {
+      "epoch": 0.3238073288776216,
+      "grad_norm": 0.2250133603811264,
+      "learning_rate": 7.903784905170277e-05,
+      "loss": 0.6969,
+      "step": 1405
+    },
+    {
+      "epoch": 0.32403779672735655,
+      "grad_norm": 0.2541544735431671,
+      "learning_rate": 7.900745243708416e-05,
+      "loss": 0.6874,
+      "step": 1406
+    },
+    {
+      "epoch": 0.3242682645770915,
+      "grad_norm": 0.24582208693027496,
+      "learning_rate": 7.89770396544078e-05,
+      "loss": 0.688,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3244987324268265,
+      "grad_norm": 0.2596662938594818,
+      "learning_rate": 7.894661072062503e-05,
+      "loss": 0.7033,
+      "step": 1408
+    },
+    {
+      "epoch": 0.32472920027656144,
+      "grad_norm": 0.24980351328849792,
+      "learning_rate": 7.891616565269623e-05,
+      "loss": 0.6844,
+      "step": 1409
+    },
+    {
+      "epoch": 0.3249596681262964,
+      "grad_norm": 0.2832036316394806,
+      "learning_rate": 7.888570446759074e-05,
+      "loss": 0.6904,
+      "step": 1410
+    },
+    {
+      "epoch": 0.32519013597603136,
+      "grad_norm": 0.27952510118484497,
+      "learning_rate": 7.88552271822869e-05,
+      "loss": 0.6913,
+      "step": 1411
+    },
+    {
+      "epoch": 0.3254206038257663,
+      "grad_norm": 0.2454599142074585,
+      "learning_rate": 7.882473381377202e-05,
+      "loss": 0.6917,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3256510716755013,
+      "grad_norm": 0.219132199883461,
+      "learning_rate": 7.879422437904236e-05,
+      "loss": 0.6959,
+      "step": 1413
+    },
+    {
+      "epoch": 0.32588153952523624,
+      "grad_norm": 0.25890830159187317,
+      "learning_rate": 7.876369889510316e-05,
+      "loss": 0.6987,
+      "step": 1414
+    },
+    {
+      "epoch": 0.3261120073749712,
+      "grad_norm": 0.22995416820049286,
+      "learning_rate": 7.873315737896856e-05,
+      "loss": 0.6895,
+      "step": 1415
+    },
+    {
+      "epoch": 0.32634247522470616,
+      "grad_norm": 0.26824015378952026,
+      "learning_rate": 7.87025998476617e-05,
+      "loss": 0.6919,
+      "step": 1416
+    },
+    {
+      "epoch": 0.3265729430744411,
+      "grad_norm": 0.26897183060646057,
+      "learning_rate": 7.867202631821462e-05,
+      "loss": 0.6913,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3268034109241761,
+      "grad_norm": 0.24150390923023224,
+      "learning_rate": 7.864143680766826e-05,
+      "loss": 0.6967,
+      "step": 1418
+    },
+    {
+      "epoch": 0.32703387877391105,
+      "grad_norm": 0.26501011848449707,
+      "learning_rate": 7.861083133307247e-05,
+      "loss": 0.6967,
+      "step": 1419
+    },
+    {
+      "epoch": 0.327264346623646,
+      "grad_norm": 0.23892751336097717,
+      "learning_rate": 7.858020991148602e-05,
+      "loss": 0.6999,
+      "step": 1420
+    },
+    {
+      "epoch": 0.32749481447338097,
+      "grad_norm": 0.2643808126449585,
+      "learning_rate": 7.854957255997653e-05,
+      "loss": 0.6919,
+      "step": 1421
+    },
+    {
+      "epoch": 0.32772528232311593,
+      "grad_norm": 0.26628732681274414,
+      "learning_rate": 7.851891929562057e-05,
+      "loss": 0.6963,
+      "step": 1422
+    },
+    {
+      "epoch": 0.3279557501728509,
+      "grad_norm": 0.27681294083595276,
+      "learning_rate": 7.84882501355035e-05,
+      "loss": 0.6879,
+      "step": 1423
+    },
+    {
+      "epoch": 0.32818621802258585,
+      "grad_norm": 0.25981584191322327,
+      "learning_rate": 7.845756509671955e-05,
+      "loss": 0.6974,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3284166858723208,
+      "grad_norm": 0.2389248162508011,
+      "learning_rate": 7.842686419637189e-05,
+      "loss": 0.6961,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3286471537220558,
+      "grad_norm": 0.27327385544776917,
+      "learning_rate": 7.839614745157242e-05,
+      "loss": 0.6834,
+      "step": 1426
+    },
+    {
+      "epoch": 0.32887762157179073,
+      "grad_norm": 0.2433241307735443,
+      "learning_rate": 7.836541487944193e-05,
+      "loss": 0.6887,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3291080894215257,
+      "grad_norm": 0.23156027495861053,
+      "learning_rate": 7.833466649711002e-05,
+      "loss": 0.6872,
+      "step": 1428
+    },
+    {
+      "epoch": 0.32933855727126066,
+      "grad_norm": 0.27620989084243774,
+      "learning_rate": 7.83039023217151e-05,
+      "loss": 0.7045,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3295690251209956,
+      "grad_norm": 0.2832176983356476,
+      "learning_rate": 7.827312237040435e-05,
+      "loss": 0.6952,
+      "step": 1430
+    },
+    {
+      "epoch": 0.3297994929707306,
+      "grad_norm": 0.2531898319721222,
+      "learning_rate": 7.824232666033382e-05,
+      "loss": 0.6888,
+      "step": 1431
+    },
+    {
+      "epoch": 0.33002996082046554,
+      "grad_norm": 0.21987776458263397,
+      "learning_rate": 7.821151520866829e-05,
+      "loss": 0.687,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3302604286702005,
+      "grad_norm": 0.26392921805381775,
+      "learning_rate": 7.818068803258131e-05,
+      "loss": 0.6815,
+      "step": 1433
+    },
+    {
+      "epoch": 0.33049089651993546,
+      "grad_norm": 0.23336456716060638,
+      "learning_rate": 7.814984514925521e-05,
+      "loss": 0.6834,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3307213643696704,
+      "grad_norm": 0.25581541657447815,
+      "learning_rate": 7.811898657588109e-05,
+      "loss": 0.6811,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3309518322194054,
+      "grad_norm": 0.23013727366924286,
+      "learning_rate": 7.808811232965876e-05,
+      "loss": 0.676,
+      "step": 1436
+    },
+    {
+      "epoch": 0.33118230006914035,
+      "grad_norm": 0.222098246216774,
+      "learning_rate": 7.805722242779678e-05,
+      "loss": 0.6909,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3314127679188753,
+      "grad_norm": 0.24615336954593658,
+      "learning_rate": 7.802631688751247e-05,
+      "loss": 0.6924,
+      "step": 1438
+    },
+    {
+      "epoch": 0.33164323576861027,
+      "grad_norm": 0.23416768014431,
+      "learning_rate": 7.79953957260318e-05,
+      "loss": 0.7015,
+      "step": 1439
+    },
+    {
+      "epoch": 0.33187370361834523,
+      "grad_norm": 0.2338051199913025,
+      "learning_rate": 7.796445896058953e-05,
+      "loss": 0.6929,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3321041714680802,
+      "grad_norm": 0.24458903074264526,
+      "learning_rate": 7.793350660842902e-05,
+      "loss": 0.6852,
+      "step": 1441
+    },
+    {
+      "epoch": 0.33233463931781515,
+      "grad_norm": 0.2856997847557068,
+      "learning_rate": 7.790253868680237e-05,
+      "loss": 0.6926,
+      "step": 1442
+    },
+    {
+      "epoch": 0.3325651071675501,
+      "grad_norm": 0.24041031301021576,
+      "learning_rate": 7.78715552129704e-05,
+      "loss": 0.689,
+      "step": 1443
+    },
+    {
+      "epoch": 0.3327955750172851,
+      "grad_norm": 0.286312997341156,
+      "learning_rate": 7.784055620420251e-05,
+      "loss": 0.6954,
+      "step": 1444
+    },
+    {
+      "epoch": 0.33302604286702003,
+      "grad_norm": 0.25002339482307434,
+      "learning_rate": 7.780954167777684e-05,
+      "loss": 0.6929,
+      "step": 1445
+    },
+    {
+      "epoch": 0.333256510716755,
+      "grad_norm": 0.2585028409957886,
+      "learning_rate": 7.777851165098012e-05,
+      "loss": 0.6919,
+      "step": 1446
+    },
+    {
+      "epoch": 0.33348697856648996,
+      "grad_norm": 0.25653818249702454,
+      "learning_rate": 7.774746614110775e-05,
+      "loss": 0.6909,
+      "step": 1447
+    },
+    {
+      "epoch": 0.3337174464162249,
+      "grad_norm": 0.2610546052455902,
+      "learning_rate": 7.771640516546373e-05,
+      "loss": 0.6921,
+      "step": 1448
+    },
+    {
+      "epoch": 0.3339479142659599,
+      "grad_norm": 0.23188693821430206,
+      "learning_rate": 7.768532874136074e-05,
+      "loss": 0.6898,
+      "step": 1449
+    },
+    {
+      "epoch": 0.33417838211569484,
+      "grad_norm": 0.2562350332736969,
+      "learning_rate": 7.765423688612001e-05,
+      "loss": 0.6815,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3344088499654298,
+      "grad_norm": 0.2621154487133026,
+      "learning_rate": 7.762312961707141e-05,
+      "loss": 0.6978,
+      "step": 1451
+    },
+    {
+      "epoch": 0.33463931781516476,
+      "grad_norm": 0.23152048885822296,
+      "learning_rate": 7.759200695155336e-05,
+      "loss": 0.6961,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3348697856648997,
+      "grad_norm": 0.2416309267282486,
+      "learning_rate": 7.75608689069129e-05,
+      "loss": 0.6855,
+      "step": 1453
+    },
+    {
+      "epoch": 0.3351002535146347,
+      "grad_norm": 0.2472970187664032,
+      "learning_rate": 7.752971550050563e-05,
+      "loss": 0.6993,
+      "step": 1454
+    },
+    {
+      "epoch": 0.33533072136436964,
+      "grad_norm": 0.24622397124767303,
+      "learning_rate": 7.749854674969573e-05,
+      "loss": 0.7027,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3355611892141046,
+      "grad_norm": 0.21482546627521515,
+      "learning_rate": 7.746736267185587e-05,
+      "loss": 0.6861,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3357916570638396,
+      "grad_norm": 0.2635815143585205,
+      "learning_rate": 7.743616328436733e-05,
+      "loss": 0.6936,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3360221249135746,
+      "grad_norm": 0.22743616998195648,
+      "learning_rate": 7.740494860461991e-05,
+      "loss": 0.6925,
+      "step": 1458
+    },
+    {
+      "epoch": 0.33625259276330954,
+      "grad_norm": 0.24108994007110596,
+      "learning_rate": 7.73737186500119e-05,
+      "loss": 0.6888,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3364830606130445,
+      "grad_norm": 0.2540665566921234,
+      "learning_rate": 7.734247343795016e-05,
+      "loss": 0.6934,
+      "step": 1460
+    },
+    {
+      "epoch": 0.33671352846277947,
+      "grad_norm": 0.22767134010791779,
+      "learning_rate": 7.731121298585e-05,
+      "loss": 0.6871,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3369439963125144,
+      "grad_norm": 0.2467404305934906,
+      "learning_rate": 7.727993731113523e-05,
+      "loss": 0.6895,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3371744641622494,
+      "grad_norm": 0.23155854642391205,
+      "learning_rate": 7.724864643123819e-05,
+      "loss": 0.683,
+      "step": 1463
+    },
+    {
+      "epoch": 0.33740493201198435,
+      "grad_norm": 0.2407347857952118,
+      "learning_rate": 7.721734036359964e-05,
+      "loss": 0.6873,
+      "step": 1464
+    },
+    {
+      "epoch": 0.3376353998617193,
+      "grad_norm": 0.254625141620636,
+      "learning_rate": 7.718601912566887e-05,
+      "loss": 0.6934,
+      "step": 1465
+    },
+    {
+      "epoch": 0.33786586771145427,
+      "grad_norm": 0.2507195770740509,
+      "learning_rate": 7.715468273490354e-05,
+      "loss": 0.6922,
+      "step": 1466
+    },
+    {
+      "epoch": 0.33809633556118923,
+      "grad_norm": 0.2584257423877716,
+      "learning_rate": 7.712333120876983e-05,
+      "loss": 0.6855,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3383268034109242,
+      "grad_norm": 0.24727049469947815,
+      "learning_rate": 7.709196456474231e-05,
+      "loss": 0.6748,
+      "step": 1468
+    },
+    {
+      "epoch": 0.33855727126065915,
+      "grad_norm": 0.2488059401512146,
+      "learning_rate": 7.7060582820304e-05,
+      "loss": 0.6869,
+      "step": 1469
+    },
+    {
+      "epoch": 0.3387877391103941,
+      "grad_norm": 0.24992689490318298,
+      "learning_rate": 7.702918599294636e-05,
+      "loss": 0.6857,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3390182069601291,
+      "grad_norm": 0.2532345652580261,
+      "learning_rate": 7.69977741001692e-05,
+      "loss": 0.6923,
+      "step": 1471
+    },
+    {
+      "epoch": 0.33924867480986404,
+      "grad_norm": 0.2411700189113617,
+      "learning_rate": 7.696634715948072e-05,
+      "loss": 0.6909,
+      "step": 1472
+    },
+    {
+      "epoch": 0.339479142659599,
+      "grad_norm": 0.2590501606464386,
+      "learning_rate": 7.693490518839763e-05,
+      "loss": 0.6771,
+      "step": 1473
+    },
+    {
+      "epoch": 0.33970961050933396,
+      "grad_norm": 0.2995712161064148,
+      "learning_rate": 7.690344820444486e-05,
+      "loss": 0.6867,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3399400783590689,
+      "grad_norm": 0.26080286502838135,
+      "learning_rate": 7.68719762251558e-05,
+      "loss": 0.696,
+      "step": 1475
+    },
+    {
+      "epoch": 0.3401705462088039,
+      "grad_norm": 0.2599622309207916,
+      "learning_rate": 7.684048926807219e-05,
+      "loss": 0.6828,
+      "step": 1476
+    },
+    {
+      "epoch": 0.34040101405853884,
+      "grad_norm": 0.2792719006538391,
+      "learning_rate": 7.680898735074407e-05,
+      "loss": 0.6876,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3406314819082738,
+      "grad_norm": 0.25866684317588806,
+      "learning_rate": 7.677747049072987e-05,
+      "loss": 0.6922,
+      "step": 1478
+    },
+    {
+      "epoch": 0.34086194975800876,
+      "grad_norm": 0.23737114667892456,
+      "learning_rate": 7.674593870559634e-05,
+      "loss": 0.692,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3410924176077437,
+      "grad_norm": 0.26033899188041687,
+      "learning_rate": 7.671439201291853e-05,
+      "loss": 0.69,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3413228854574787,
+      "grad_norm": 0.24097611010074615,
+      "learning_rate": 7.668283043027982e-05,
+      "loss": 0.6896,
+      "step": 1481
+    },
+    {
+      "epoch": 0.34155335330721365,
+      "grad_norm": 0.26150503754615784,
+      "learning_rate": 7.665125397527187e-05,
+      "loss": 0.6919,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3417838211569486,
+      "grad_norm": 0.21474456787109375,
+      "learning_rate": 7.661966266549463e-05,
+      "loss": 0.7017,
+      "step": 1483
+    },
+    {
+      "epoch": 0.34201428900668357,
+      "grad_norm": 0.2468489408493042,
+      "learning_rate": 7.658805651855636e-05,
+      "loss": 0.6858,
+      "step": 1484
+    },
+    {
+      "epoch": 0.34224475685641853,
+      "grad_norm": 0.2597835659980774,
+      "learning_rate": 7.655643555207355e-05,
+      "loss": 0.6919,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3424752247061535,
+      "grad_norm": 0.24256932735443115,
+      "learning_rate": 7.652479978367097e-05,
+      "loss": 0.6915,
+      "step": 1486
+    },
+    {
+      "epoch": 0.34270569255588845,
+      "grad_norm": 0.2565065026283264,
+      "learning_rate": 7.649314923098164e-05,
+      "loss": 0.6863,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3429361604056234,
+      "grad_norm": 0.27215030789375305,
+      "learning_rate": 7.646148391164682e-05,
+      "loss": 0.6871,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3431666282553584,
+      "grad_norm": 0.27064988017082214,
+      "learning_rate": 7.6429803843316e-05,
+      "loss": 0.6914,
+      "step": 1489
+    },
+    {
+      "epoch": 0.34339709610509334,
+      "grad_norm": 0.2974797487258911,
+      "learning_rate": 7.63981090436469e-05,
+      "loss": 0.6805,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3436275639548283,
+      "grad_norm": 0.21655718982219696,
+      "learning_rate": 7.636639953030541e-05,
+      "loss": 0.688,
+      "step": 1491
+    },
+    {
+      "epoch": 0.34385803180456326,
+      "grad_norm": 0.26122844219207764,
+      "learning_rate": 7.633467532096567e-05,
+      "loss": 0.675,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3440884996542982,
+      "grad_norm": 0.22643567621707916,
+      "learning_rate": 7.630293643331001e-05,
+      "loss": 0.6917,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3443189675040332,
+      "grad_norm": 0.26686912775039673,
+      "learning_rate": 7.627118288502889e-05,
+      "loss": 0.6975,
+      "step": 1494
+    },
+    {
+      "epoch": 0.34454943535376814,
+      "grad_norm": 0.25299954414367676,
+      "learning_rate": 7.623941469382099e-05,
+      "loss": 0.6944,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3447799032035031,
+      "grad_norm": 0.2686970829963684,
+      "learning_rate": 7.620763187739315e-05,
+      "loss": 0.6879,
+      "step": 1496
+    },
+    {
+      "epoch": 0.34501037105323806,
+      "grad_norm": 0.2587553858757019,
+      "learning_rate": 7.617583445346033e-05,
+      "loss": 0.6885,
+      "step": 1497
+    },
+    {
+      "epoch": 0.345240838902973,
+      "grad_norm": 0.25488153100013733,
+      "learning_rate": 7.614402243974568e-05,
+      "loss": 0.7005,
+      "step": 1498
+    },
+    {
+      "epoch": 0.345471306752708,
+      "grad_norm": 0.2613796889781952,
+      "learning_rate": 7.61121958539804e-05,
+      "loss": 0.6872,
+      "step": 1499
+    },
+    {
+      "epoch": 0.34570177460244295,
+      "grad_norm": 0.21812820434570312,
+      "learning_rate": 7.608035471390394e-05,
+      "loss": 0.6909,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3459322424521779,
+      "grad_norm": 0.24839772284030914,
+      "learning_rate": 7.604849903726372e-05,
+      "loss": 0.6839,
+      "step": 1501
+    },
+    {
+      "epoch": 0.34616271030191287,
+      "grad_norm": 0.25548359751701355,
+      "learning_rate": 7.601662884181534e-05,
+      "loss": 0.6927,
+      "step": 1502
+    },
+    {
+      "epoch": 0.34639317815164783,
+      "grad_norm": 0.25990694761276245,
+      "learning_rate": 7.598474414532252e-05,
+      "loss": 0.6875,
+      "step": 1503
+    },
+    {
+      "epoch": 0.3466236460013828,
+      "grad_norm": 0.23840296268463135,
+      "learning_rate": 7.595284496555698e-05,
+      "loss": 0.6963,
+      "step": 1504
+    },
+    {
+      "epoch": 0.34685411385111775,
+      "grad_norm": 0.24960501492023468,
+      "learning_rate": 7.592093132029861e-05,
+      "loss": 0.6853,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3470845817008527,
+      "grad_norm": 0.22266826033592224,
+      "learning_rate": 7.588900322733526e-05,
+      "loss": 0.6832,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3473150495505877,
+      "grad_norm": 0.23632311820983887,
+      "learning_rate": 7.585706070446288e-05,
+      "loss": 0.6813,
+      "step": 1507
+    },
+    {
+      "epoch": 0.34754551740032263,
+      "grad_norm": 0.22808539867401123,
+      "learning_rate": 7.582510376948552e-05,
+      "loss": 0.6788,
+      "step": 1508
+    },
+    {
+      "epoch": 0.3477759852500576,
+      "grad_norm": 0.2298205941915512,
+      "learning_rate": 7.579313244021515e-05,
+      "loss": 0.6926,
+      "step": 1509
+    },
+    {
+      "epoch": 0.34800645309979256,
+      "grad_norm": 0.25376659631729126,
+      "learning_rate": 7.576114673447186e-05,
+      "loss": 0.6953,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3482369209495275,
+      "grad_norm": 0.24706676602363586,
+      "learning_rate": 7.57291466700837e-05,
+      "loss": 0.685,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3484673887992625,
+      "grad_norm": 0.25202682614326477,
+      "learning_rate": 7.569713226488674e-05,
+      "loss": 0.6867,
+      "step": 1512
+    },
+    {
+      "epoch": 0.34869785664899744,
+      "grad_norm": 0.2461564689874649,
+      "learning_rate": 7.566510353672504e-05,
+      "loss": 0.6946,
+      "step": 1513
+    },
+    {
+      "epoch": 0.3489283244987324,
+      "grad_norm": 0.26574838161468506,
+      "learning_rate": 7.563306050345062e-05,
+      "loss": 0.699,
+      "step": 1514
+    },
+    {
+      "epoch": 0.34915879234846736,
+      "grad_norm": 0.256676584482193,
+      "learning_rate": 7.560100318292355e-05,
+      "loss": 0.6896,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3493892601982024,
+      "grad_norm": 0.23021619021892548,
+      "learning_rate": 7.556893159301178e-05,
+      "loss": 0.693,
+      "step": 1516
+    },
+    {
+      "epoch": 0.34961972804793734,
+      "grad_norm": 0.23799525201320648,
+      "learning_rate": 7.553684575159124e-05,
+      "loss": 0.6855,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3498501958976723,
+      "grad_norm": 0.23646856844425201,
+      "learning_rate": 7.550474567654583e-05,
+      "loss": 0.6894,
+      "step": 1518
+    },
+    {
+      "epoch": 0.35008066374740726,
+      "grad_norm": 0.2123948484659195,
+      "learning_rate": 7.547263138576732e-05,
+      "loss": 0.6744,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3503111315971422,
+      "grad_norm": 0.22038936614990234,
+      "learning_rate": 7.54405028971555e-05,
+      "loss": 0.6812,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3505415994468772,
+      "grad_norm": 0.22953416407108307,
+      "learning_rate": 7.540836022861797e-05,
+      "loss": 0.679,
+      "step": 1521
+    },
+    {
+      "epoch": 0.35077206729661214,
+      "grad_norm": 0.23723843693733215,
+      "learning_rate": 7.53762033980703e-05,
+      "loss": 0.6828,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3510025351463471,
+      "grad_norm": 0.2359115332365036,
+      "learning_rate": 7.534403242343595e-05,
+      "loss": 0.6821,
+      "step": 1523
+    },
+    {
+      "epoch": 0.35123300299608207,
+      "grad_norm": 0.23382125794887543,
+      "learning_rate": 7.531184732264624e-05,
+      "loss": 0.6885,
+      "step": 1524
+    },
+    {
+      "epoch": 0.351463470845817,
+      "grad_norm": 0.2295057624578476,
+      "learning_rate": 7.527964811364035e-05,
+      "loss": 0.686,
+      "step": 1525
+    },
+    {
+      "epoch": 0.351693938695552,
+      "grad_norm": 0.21867454051971436,
+      "learning_rate": 7.524743481436537e-05,
+      "loss": 0.6864,
+      "step": 1526
+    },
+    {
+      "epoch": 0.35192440654528695,
+      "grad_norm": 0.21911703050136566,
+      "learning_rate": 7.521520744277623e-05,
+      "loss": 0.6882,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3521548743950219,
+      "grad_norm": 0.227437362074852,
+      "learning_rate": 7.518296601683567e-05,
+      "loss": 0.6894,
+      "step": 1528
+    },
+    {
+      "epoch": 0.35238534224475687,
+      "grad_norm": 0.23779958486557007,
+      "learning_rate": 7.515071055451429e-05,
+      "loss": 0.6749,
+      "step": 1529
+    },
+    {
+      "epoch": 0.35261581009449183,
+      "grad_norm": 0.22434940934181213,
+      "learning_rate": 7.511844107379052e-05,
+      "loss": 0.6897,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3528462779442268,
+      "grad_norm": 0.22103999555110931,
+      "learning_rate": 7.508615759265059e-05,
+      "loss": 0.68,
+      "step": 1531
+    },
+    {
+      "epoch": 0.35307674579396175,
+      "grad_norm": 0.22952067852020264,
+      "learning_rate": 7.505386012908853e-05,
+      "loss": 0.6892,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3533072136436967,
+      "grad_norm": 0.24926556646823883,
+      "learning_rate": 7.50215487011062e-05,
+      "loss": 0.6829,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3535376814934317,
+      "grad_norm": 0.2560986280441284,
+      "learning_rate": 7.498922332671317e-05,
+      "loss": 0.6971,
+      "step": 1534
+    },
+    {
+      "epoch": 0.35376814934316664,
+      "grad_norm": 0.256628155708313,
+      "learning_rate": 7.495688402392686e-05,
+      "loss": 0.6908,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3539986171929016,
+      "grad_norm": 0.23531514406204224,
+      "learning_rate": 7.492453081077241e-05,
+      "loss": 0.6811,
+      "step": 1536
+    },
+    {
+      "epoch": 0.35422908504263656,
+      "grad_norm": 0.24269689619541168,
+      "learning_rate": 7.489216370528273e-05,
+      "loss": 0.6824,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3544595528923715,
+      "grad_norm": 0.2383260726928711,
+      "learning_rate": 7.485978272549847e-05,
+      "loss": 0.6876,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3546900207421065,
+      "grad_norm": 0.25450727343559265,
+      "learning_rate": 7.482738788946799e-05,
+      "loss": 0.6852,
+      "step": 1539
+    },
+    {
+      "epoch": 0.35492048859184144,
+      "grad_norm": 0.21417759358882904,
+      "learning_rate": 7.479497921524741e-05,
+      "loss": 0.6848,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3551509564415764,
+      "grad_norm": 0.23762138187885284,
+      "learning_rate": 7.476255672090055e-05,
+      "loss": 0.6943,
+      "step": 1541
+    },
+    {
+      "epoch": 0.35538142429131137,
+      "grad_norm": 0.2219182252883911,
+      "learning_rate": 7.473012042449894e-05,
+      "loss": 0.687,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3556118921410463,
+      "grad_norm": 0.23807905614376068,
+      "learning_rate": 7.469767034412176e-05,
+      "loss": 0.6859,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3558423599907813,
+      "grad_norm": 0.24784959852695465,
+      "learning_rate": 7.466520649785593e-05,
+      "loss": 0.6787,
+      "step": 1544
+    },
+    {
+      "epoch": 0.35607282784051625,
+      "grad_norm": 0.2389153093099594,
+      "learning_rate": 7.463272890379602e-05,
+      "loss": 0.6922,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3563032956902512,
+      "grad_norm": 0.24661684036254883,
+      "learning_rate": 7.460023758004426e-05,
+      "loss": 0.6782,
+      "step": 1546
+    },
+    {
+      "epoch": 0.35653376353998617,
+      "grad_norm": 0.22186362743377686,
+      "learning_rate": 7.456773254471053e-05,
+      "loss": 0.684,
+      "step": 1547
+    },
+    {
+      "epoch": 0.35676423138972113,
+      "grad_norm": 0.2341771274805069,
+      "learning_rate": 7.453521381591233e-05,
+      "loss": 0.688,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3569946992394561,
+      "grad_norm": 0.2423945963382721,
+      "learning_rate": 7.450268141177486e-05,
+      "loss": 0.6738,
+      "step": 1549
+    },
+    {
+      "epoch": 0.35722516708919105,
+      "grad_norm": 0.2303714007139206,
+      "learning_rate": 7.44701353504309e-05,
+      "loss": 0.6826,
+      "step": 1550
+    },
+    {
+      "epoch": 0.357455634938926,
+      "grad_norm": 0.26467540860176086,
+      "learning_rate": 7.443757565002081e-05,
+      "loss": 0.6836,
+      "step": 1551
+    },
+    {
+      "epoch": 0.357686102788661,
+      "grad_norm": 0.25847116112709045,
+      "learning_rate": 7.440500232869262e-05,
+      "loss": 0.6886,
+      "step": 1552
+    },
+    {
+      "epoch": 0.35791657063839594,
+      "grad_norm": 0.22903743386268616,
+      "learning_rate": 7.43724154046019e-05,
+      "loss": 0.6929,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3581470384881309,
+      "grad_norm": 0.2548694312572479,
+      "learning_rate": 7.433981489591181e-05,
+      "loss": 0.6879,
+      "step": 1554
+    },
+    {
+      "epoch": 0.35837750633786586,
+      "grad_norm": 0.2554045617580414,
+      "learning_rate": 7.43072008207931e-05,
+      "loss": 0.6834,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3586079741876008,
+      "grad_norm": 0.23032429814338684,
+      "learning_rate": 7.42745731974241e-05,
+      "loss": 0.6771,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3588384420373358,
+      "grad_norm": 0.23225395381450653,
+      "learning_rate": 7.424193204399061e-05,
+      "loss": 0.6906,
+      "step": 1557
+    },
+    {
+      "epoch": 0.35906890988707074,
+      "grad_norm": 0.24884352087974548,
+      "learning_rate": 7.420927737868608e-05,
+      "loss": 0.6937,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3592993777368057,
+      "grad_norm": 0.22079841792583466,
+      "learning_rate": 7.417660921971141e-05,
+      "loss": 0.6965,
+      "step": 1559
+    },
+    {
+      "epoch": 0.35952984558654066,
+      "grad_norm": 0.2275458425283432,
+      "learning_rate": 7.414392758527504e-05,
+      "loss": 0.6788,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3597603134362756,
+      "grad_norm": 0.25046491622924805,
+      "learning_rate": 7.411123249359294e-05,
+      "loss": 0.6874,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3599907812860106,
+      "grad_norm": 0.24402521550655365,
+      "learning_rate": 7.407852396288857e-05,
+      "loss": 0.6775,
+      "step": 1562
+    },
+    {
+      "epoch": 0.36022124913574555,
+      "grad_norm": 0.2387145459651947,
+      "learning_rate": 7.404580201139286e-05,
+      "loss": 0.6817,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3604517169854805,
+      "grad_norm": 0.2291865348815918,
+      "learning_rate": 7.401306665734429e-05,
+      "loss": 0.691,
+      "step": 1564
+    },
+    {
+      "epoch": 0.36068218483521547,
+      "grad_norm": 0.24180877208709717,
+      "learning_rate": 7.398031791898872e-05,
+      "loss": 0.6863,
+      "step": 1565
+    },
+    {
+      "epoch": 0.36091265268495043,
+      "grad_norm": 0.2156798541545868,
+      "learning_rate": 7.394755581457949e-05,
+      "loss": 0.683,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3611431205346854,
+      "grad_norm": 0.2224697768688202,
+      "learning_rate": 7.391478036237747e-05,
+      "loss": 0.6863,
+      "step": 1567
+    },
+    {
+      "epoch": 0.36137358838442035,
+      "grad_norm": 0.23175424337387085,
+      "learning_rate": 7.388199158065086e-05,
+      "loss": 0.6896,
+      "step": 1568
+    },
+    {
+      "epoch": 0.3616040562341553,
+      "grad_norm": 0.20944839715957642,
+      "learning_rate": 7.384918948767538e-05,
+      "loss": 0.6949,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3618345240838903,
+      "grad_norm": 0.24800141155719757,
+      "learning_rate": 7.38163741017341e-05,
+      "loss": 0.6827,
+      "step": 1570
+    },
+    {
+      "epoch": 0.36206499193362524,
+      "grad_norm": 0.2055944949388504,
+      "learning_rate": 7.378354544111755e-05,
+      "loss": 0.6859,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3622954597833602,
+      "grad_norm": 0.23914718627929688,
+      "learning_rate": 7.37507035241236e-05,
+      "loss": 0.691,
+      "step": 1572
+    },
+    {
+      "epoch": 0.36252592763309516,
+      "grad_norm": 0.24815700948238373,
+      "learning_rate": 7.371784836905758e-05,
+      "loss": 0.6964,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3627563954828301,
+      "grad_norm": 0.22123336791992188,
+      "learning_rate": 7.368497999423216e-05,
+      "loss": 0.6821,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3629868633325651,
+      "grad_norm": 0.22671015560626984,
+      "learning_rate": 7.365209841796738e-05,
+      "loss": 0.6824,
+      "step": 1575
+    },
+    {
+      "epoch": 0.3632173311823001,
+      "grad_norm": 0.2351810485124588,
+      "learning_rate": 7.361920365859066e-05,
+      "loss": 0.6829,
+      "step": 1576
+    },
+    {
+      "epoch": 0.36344779903203506,
+      "grad_norm": 0.27589890360832214,
+      "learning_rate": 7.35862957344367e-05,
+      "loss": 0.6909,
+      "step": 1577
+    },
+    {
+      "epoch": 0.36367826688177,
+      "grad_norm": 0.24565422534942627,
+      "learning_rate": 7.355337466384761e-05,
+      "loss": 0.6882,
+      "step": 1578
+    },
+    {
+      "epoch": 0.363908734731505,
+      "grad_norm": 0.24734973907470703,
+      "learning_rate": 7.352044046517285e-05,
+      "loss": 0.694,
+      "step": 1579
+    },
+    {
+      "epoch": 0.36413920258123994,
+      "grad_norm": 0.23439617455005646,
+      "learning_rate": 7.34874931567691e-05,
+      "loss": 0.6883,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3643696704309749,
+      "grad_norm": 0.2454935908317566,
+      "learning_rate": 7.34545327570004e-05,
+      "loss": 0.692,
+      "step": 1581
+    },
+    {
+      "epoch": 0.36460013828070986,
+      "grad_norm": 0.21465803682804108,
+      "learning_rate": 7.342155928423812e-05,
+      "loss": 0.68,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3648306061304448,
+      "grad_norm": 0.24188017845153809,
+      "learning_rate": 7.338857275686084e-05,
+      "loss": 0.683,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3650610739801798,
+      "grad_norm": 0.2033400982618332,
+      "learning_rate": 7.335557319325449e-05,
+      "loss": 0.6855,
+      "step": 1584
+    },
+    {
+      "epoch": 0.36529154182991475,
+      "grad_norm": 0.2551295757293701,
+      "learning_rate": 7.332256061181222e-05,
+      "loss": 0.6839,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3655220096796497,
+      "grad_norm": 0.218764066696167,
+      "learning_rate": 7.328953503093446e-05,
+      "loss": 0.6823,
+      "step": 1586
+    },
+    {
+      "epoch": 0.36575247752938467,
+      "grad_norm": 0.23315462470054626,
+      "learning_rate": 7.325649646902887e-05,
+      "loss": 0.6884,
+      "step": 1587
+    },
+    {
+      "epoch": 0.36598294537911963,
+      "grad_norm": 0.2543986439704895,
+      "learning_rate": 7.322344494451034e-05,
+      "loss": 0.681,
+      "step": 1588
+    },
+    {
+      "epoch": 0.3662134132288546,
+      "grad_norm": 0.3002406060695648,
+      "learning_rate": 7.319038047580102e-05,
+      "loss": 0.6907,
+      "step": 1589
+    },
+    {
+      "epoch": 0.36644388107858955,
+      "grad_norm": 0.24670617282390594,
+      "learning_rate": 7.315730308133023e-05,
+      "loss": 0.6922,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3666743489283245,
+      "grad_norm": 0.23123899102210999,
+      "learning_rate": 7.312421277953454e-05,
+      "loss": 0.682,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3669048167780595,
+      "grad_norm": 0.3154587745666504,
+      "learning_rate": 7.309110958885768e-05,
+      "loss": 0.6841,
+      "step": 1592
+    },
+    {
+      "epoch": 0.36713528462779443,
+      "grad_norm": 0.24491167068481445,
+      "learning_rate": 7.305799352775055e-05,
+      "loss": 0.6869,
+      "step": 1593
+    },
+    {
+      "epoch": 0.3673657524775294,
+      "grad_norm": 0.24614278972148895,
+      "learning_rate": 7.302486461467128e-05,
+      "loss": 0.678,
+      "step": 1594
+    },
+    {
+      "epoch": 0.36759622032726436,
+      "grad_norm": 0.2878095507621765,
+      "learning_rate": 7.299172286808511e-05,
+      "loss": 0.6899,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3678266881769993,
+      "grad_norm": 0.26783236861228943,
+      "learning_rate": 7.295856830646446e-05,
+      "loss": 0.693,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3680571560267343,
+      "grad_norm": 0.2804826498031616,
+      "learning_rate": 7.29254009482889e-05,
+      "loss": 0.687,
+      "step": 1597
+    },
+    {
+      "epoch": 0.36828762387646924,
+      "grad_norm": 0.29956531524658203,
+      "learning_rate": 7.28922208120451e-05,
+      "loss": 0.6868,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3685180917262042,
+      "grad_norm": 0.22456161677837372,
+      "learning_rate": 7.285902791622688e-05,
+      "loss": 0.6829,
+      "step": 1599
+    },
+    {
+      "epoch": 0.36874855957593916,
+      "grad_norm": 0.2787242829799652,
+      "learning_rate": 7.282582227933517e-05,
+      "loss": 0.6845,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3689790274256741,
+      "grad_norm": 0.28973785042762756,
+      "learning_rate": 7.279260391987799e-05,
+      "loss": 0.6921,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3692094952754091,
+      "grad_norm": 0.25046855211257935,
+      "learning_rate": 7.275937285637044e-05,
+      "loss": 0.6825,
+      "step": 1602
+    },
+    {
+      "epoch": 0.36943996312514404,
+      "grad_norm": 0.25190284848213196,
+      "learning_rate": 7.272612910733475e-05,
+      "loss": 0.6882,
+      "step": 1603
+    },
+    {
+      "epoch": 0.369670430974879,
+      "grad_norm": 0.2506854832172394,
+      "learning_rate": 7.269287269130017e-05,
+      "loss": 0.6851,
+      "step": 1604
+    },
+    {
+      "epoch": 0.36990089882461397,
+      "grad_norm": 0.2547498643398285,
+      "learning_rate": 7.2659603626803e-05,
+      "loss": 0.6797,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3701313666743489,
+      "grad_norm": 0.23294281959533691,
+      "learning_rate": 7.262632193238668e-05,
+      "loss": 0.6831,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3703618345240839,
+      "grad_norm": 0.24481774866580963,
+      "learning_rate": 7.259302762660157e-05,
+      "loss": 0.6835,
+      "step": 1607
+    },
+    {
+      "epoch": 0.37059230237381885,
+      "grad_norm": 0.24376718699932098,
+      "learning_rate": 7.255972072800514e-05,
+      "loss": 0.6762,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3708227702235538,
+      "grad_norm": 0.22260530292987823,
+      "learning_rate": 7.252640125516189e-05,
+      "loss": 0.6831,
+      "step": 1609
+    },
+    {
+      "epoch": 0.37105323807328877,
+      "grad_norm": 0.23927612602710724,
+      "learning_rate": 7.249306922664322e-05,
+      "loss": 0.6926,
+      "step": 1610
+    },
+    {
+      "epoch": 0.37128370592302373,
+      "grad_norm": 0.23161613941192627,
+      "learning_rate": 7.245972466102766e-05,
+      "loss": 0.6795,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3715141737727587,
+      "grad_norm": 0.2783213257789612,
+      "learning_rate": 7.242636757690064e-05,
+      "loss": 0.6919,
+      "step": 1612
+    },
+    {
+      "epoch": 0.37174464162249365,
+      "grad_norm": 0.3050868809223175,
+      "learning_rate": 7.239299799285462e-05,
+      "loss": 0.6864,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3719751094722286,
+      "grad_norm": 0.2619262635707855,
+      "learning_rate": 7.235961592748901e-05,
+      "loss": 0.6822,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3722055773219636,
+      "grad_norm": 0.24144136905670166,
+      "learning_rate": 7.232622139941016e-05,
+      "loss": 0.6942,
+      "step": 1615
+    },
+    {
+      "epoch": 0.37243604517169854,
+      "grad_norm": 0.2681979238986969,
+      "learning_rate": 7.229281442723136e-05,
+      "loss": 0.686,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3726665130214335,
+      "grad_norm": 0.23774147033691406,
+      "learning_rate": 7.225939502957287e-05,
+      "loss": 0.6849,
+      "step": 1617
+    },
+    {
+      "epoch": 0.37289698087116846,
+      "grad_norm": 0.24318799376487732,
+      "learning_rate": 7.222596322506188e-05,
+      "loss": 0.6866,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3731274487209034,
+      "grad_norm": 0.2579040825366974,
+      "learning_rate": 7.219251903233246e-05,
+      "loss": 0.6873,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3733579165706384,
+      "grad_norm": 0.26902881264686584,
+      "learning_rate": 7.215906247002557e-05,
+      "loss": 0.684,
+      "step": 1620
+    },
+    {
+      "epoch": 0.37358838442037334,
+      "grad_norm": 0.22406496107578278,
+      "learning_rate": 7.212559355678915e-05,
+      "loss": 0.6909,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3738188522701083,
+      "grad_norm": 0.2592219114303589,
+      "learning_rate": 7.209211231127791e-05,
+      "loss": 0.6849,
+      "step": 1622
+    },
+    {
+      "epoch": 0.37404932011984326,
+      "grad_norm": 0.22760163247585297,
+      "learning_rate": 7.205861875215357e-05,
+      "loss": 0.6806,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3742797879695782,
+      "grad_norm": 0.24689623713493347,
+      "learning_rate": 7.202511289808456e-05,
+      "loss": 0.6842,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3745102558193132,
+      "grad_norm": 0.23074333369731903,
+      "learning_rate": 7.199159476774627e-05,
+      "loss": 0.6808,
+      "step": 1625
+    },
+    {
+      "epoch": 0.37474072366904815,
+      "grad_norm": 0.23396410048007965,
+      "learning_rate": 7.19580643798209e-05,
+      "loss": 0.6801,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3749711915187831,
+      "grad_norm": 0.21473470330238342,
+      "learning_rate": 7.192452175299748e-05,
+      "loss": 0.6891,
+      "step": 1627
+    },
+    {
+      "epoch": 0.37520165936851807,
+      "grad_norm": 0.24224236607551575,
+      "learning_rate": 7.189096690597188e-05,
+      "loss": 0.6844,
+      "step": 1628
+    },
+    {
+      "epoch": 0.37543212721825303,
+      "grad_norm": 0.21217916905879974,
+      "learning_rate": 7.185739985744675e-05,
+      "loss": 0.681,
+      "step": 1629
+    },
+    {
+      "epoch": 0.375662595067988,
+      "grad_norm": 0.24271690845489502,
+      "learning_rate": 7.182382062613156e-05,
+      "loss": 0.6834,
+      "step": 1630
+    },
+    {
+      "epoch": 0.37589306291772295,
+      "grad_norm": 0.2046801894903183,
+      "learning_rate": 7.179022923074258e-05,
+      "loss": 0.685,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3761235307674579,
+      "grad_norm": 0.23926329612731934,
+      "learning_rate": 7.175662569000282e-05,
+      "loss": 0.6768,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3763539986171929,
+      "grad_norm": 0.2330237329006195,
+      "learning_rate": 7.172301002264212e-05,
+      "loss": 0.6862,
+      "step": 1633
+    },
+    {
+      "epoch": 0.37658446646692784,
+      "grad_norm": 0.2538076937198639,
+      "learning_rate": 7.168938224739704e-05,
+      "loss": 0.6871,
+      "step": 1634
+    },
+    {
+      "epoch": 0.37681493431666285,
+      "grad_norm": 0.2078879177570343,
+      "learning_rate": 7.165574238301085e-05,
+      "loss": 0.6845,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3770454021663978,
+      "grad_norm": 0.23577922582626343,
+      "learning_rate": 7.162209044823367e-05,
+      "loss": 0.6771,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3772758700161328,
+      "grad_norm": 0.22798825800418854,
+      "learning_rate": 7.158842646182222e-05,
+      "loss": 0.6861,
+      "step": 1637
+    },
+    {
+      "epoch": 0.37750633786586774,
+      "grad_norm": 0.2441009134054184,
+      "learning_rate": 7.155475044254006e-05,
+      "loss": 0.6834,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3777368057156027,
+      "grad_norm": 0.2254721075296402,
+      "learning_rate": 7.152106240915735e-05,
+      "loss": 0.6814,
+      "step": 1639
+    },
+    {
+      "epoch": 0.37796727356533766,
+      "grad_norm": 0.23852422833442688,
+      "learning_rate": 7.148736238045098e-05,
+      "loss": 0.6843,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3781977414150726,
+      "grad_norm": 0.23249536752700806,
+      "learning_rate": 7.14536503752046e-05,
+      "loss": 0.69,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3784282092648076,
+      "grad_norm": 0.2591071128845215,
+      "learning_rate": 7.141992641220841e-05,
+      "loss": 0.6809,
+      "step": 1642
+    },
+    {
+      "epoch": 0.37865867711454254,
+      "grad_norm": 0.24217630922794342,
+      "learning_rate": 7.138619051025935e-05,
+      "loss": 0.6746,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3788891449642775,
+      "grad_norm": 0.24814555048942566,
+      "learning_rate": 7.135244268816102e-05,
+      "loss": 0.6856,
+      "step": 1644
+    },
+    {
+      "epoch": 0.37911961281401246,
+      "grad_norm": 0.24192722141742706,
+      "learning_rate": 7.131868296472366e-05,
+      "loss": 0.6856,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3793500806637474,
+      "grad_norm": 0.27118441462516785,
+      "learning_rate": 7.12849113587641e-05,
+      "loss": 0.6829,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3795805485134824,
+      "grad_norm": 0.2808471918106079,
+      "learning_rate": 7.125112788910581e-05,
+      "loss": 0.6844,
+      "step": 1647
+    },
+    {
+      "epoch": 0.37981101636321735,
+      "grad_norm": 0.23738627135753632,
+      "learning_rate": 7.121733257457893e-05,
+      "loss": 0.6808,
+      "step": 1648
+    },
+    {
+      "epoch": 0.3800414842129523,
+      "grad_norm": 0.241289883852005,
+      "learning_rate": 7.11835254340201e-05,
+      "loss": 0.6853,
+      "step": 1649
+    },
+    {
+      "epoch": 0.38027195206268727,
+      "grad_norm": 0.240788996219635,
+      "learning_rate": 7.114970648627267e-05,
+      "loss": 0.6805,
+      "step": 1650
+    },
+    {
+      "epoch": 0.38050241991242223,
+      "grad_norm": 0.2523535192012787,
+      "learning_rate": 7.111587575018648e-05,
+      "loss": 0.6764,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3807328877621572,
+      "grad_norm": 0.2584069073200226,
+      "learning_rate": 7.108203324461795e-05,
+      "loss": 0.6913,
+      "step": 1652
+    },
+    {
+      "epoch": 0.38096335561189215,
+      "grad_norm": 0.24159248173236847,
+      "learning_rate": 7.104817898843013e-05,
+      "loss": 0.6815,
+      "step": 1653
+    },
+    {
+      "epoch": 0.3811938234616271,
+      "grad_norm": 0.22381487488746643,
+      "learning_rate": 7.101431300049253e-05,
+      "loss": 0.6824,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3814242913113621,
+      "grad_norm": 0.25524625182151794,
+      "learning_rate": 7.098043529968123e-05,
+      "loss": 0.6801,
+      "step": 1655
+    },
+    {
+      "epoch": 0.38165475916109703,
+      "grad_norm": 0.24695929884910583,
+      "learning_rate": 7.09465459048789e-05,
+      "loss": 0.6799,
+      "step": 1656
+    },
+    {
+      "epoch": 0.381885227010832,
+      "grad_norm": 0.21066734194755554,
+      "learning_rate": 7.091264483497463e-05,
+      "loss": 0.6713,
+      "step": 1657
+    },
+    {
+      "epoch": 0.38211569486056696,
+      "grad_norm": 0.25811514258384705,
+      "learning_rate": 7.087873210886406e-05,
+      "loss": 0.6852,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3823461627103019,
+      "grad_norm": 0.21766284108161926,
+      "learning_rate": 7.084480774544937e-05,
+      "loss": 0.6869,
+      "step": 1659
+    },
+    {
+      "epoch": 0.3825766305600369,
+      "grad_norm": 0.2117396891117096,
+      "learning_rate": 7.081087176363916e-05,
+      "loss": 0.6815,
+      "step": 1660
+    },
+    {
+      "epoch": 0.38280709840977184,
+      "grad_norm": 0.23894590139389038,
+      "learning_rate": 7.077692418234852e-05,
+      "loss": 0.6919,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3830375662595068,
+      "grad_norm": 0.2442779392004013,
+      "learning_rate": 7.074296502049903e-05,
+      "loss": 0.6762,
+      "step": 1662
+    },
+    {
+      "epoch": 0.38326803410924176,
+      "grad_norm": 0.20315802097320557,
+      "learning_rate": 7.070899429701873e-05,
+      "loss": 0.6747,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3834985019589767,
+      "grad_norm": 0.22201156616210938,
+      "learning_rate": 7.067501203084203e-05,
+      "loss": 0.6802,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3837289698087117,
+      "grad_norm": 0.21427859365940094,
+      "learning_rate": 7.064101824090991e-05,
+      "loss": 0.6758,
+      "step": 1665
+    },
+    {
+      "epoch": 0.38395943765844665,
+      "grad_norm": 0.20603862404823303,
+      "learning_rate": 7.060701294616963e-05,
+      "loss": 0.6759,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3841899055081816,
+      "grad_norm": 0.2058810144662857,
+      "learning_rate": 7.057299616557493e-05,
+      "loss": 0.6883,
+      "step": 1667
+    },
+    {
+      "epoch": 0.38442037335791657,
+      "grad_norm": 0.20582230389118195,
+      "learning_rate": 7.053896791808598e-05,
+      "loss": 0.6883,
+      "step": 1668
+    },
+    {
+      "epoch": 0.38465084120765153,
+      "grad_norm": 0.20562750101089478,
+      "learning_rate": 7.050492822266929e-05,
+      "loss": 0.6827,
+      "step": 1669
+    },
+    {
+      "epoch": 0.3848813090573865,
+      "grad_norm": 0.1945854127407074,
+      "learning_rate": 7.047087709829777e-05,
+      "loss": 0.6869,
+      "step": 1670
+    },
+    {
+      "epoch": 0.38511177690712145,
+      "grad_norm": 0.22347545623779297,
+      "learning_rate": 7.043681456395068e-05,
+      "loss": 0.6801,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3853422447568564,
+      "grad_norm": 0.21079128980636597,
+      "learning_rate": 7.04027406386137e-05,
+      "loss": 0.6774,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3855727126065914,
+      "grad_norm": 0.2209564447402954,
+      "learning_rate": 7.036865534127879e-05,
+      "loss": 0.6855,
+      "step": 1673
+    },
+    {
+      "epoch": 0.38580318045632633,
+      "grad_norm": 0.226369246840477,
+      "learning_rate": 7.033455869094428e-05,
+      "loss": 0.6808,
+      "step": 1674
+    },
+    {
+      "epoch": 0.3860336483060613,
+      "grad_norm": 0.2145013064146042,
+      "learning_rate": 7.030045070661484e-05,
+      "loss": 0.6904,
+      "step": 1675
+    },
+    {
+      "epoch": 0.38626411615579626,
+      "grad_norm": 0.20242154598236084,
+      "learning_rate": 7.02663314073014e-05,
+      "loss": 0.6812,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3864945840055312,
+      "grad_norm": 0.2313269078731537,
+      "learning_rate": 7.023220081202128e-05,
+      "loss": 0.6791,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3867250518552662,
+      "grad_norm": 0.2067273110151291,
+      "learning_rate": 7.019805893979805e-05,
+      "loss": 0.6795,
+      "step": 1678
+    },
+    {
+      "epoch": 0.38695551970500114,
+      "grad_norm": 0.2214699238538742,
+      "learning_rate": 7.016390580966157e-05,
+      "loss": 0.6751,
+      "step": 1679
+    },
+    {
+      "epoch": 0.3871859875547361,
+      "grad_norm": 0.23702481389045715,
+      "learning_rate": 7.012974144064795e-05,
+      "loss": 0.6868,
+      "step": 1680
+    },
+    {
+      "epoch": 0.38741645540447106,
+      "grad_norm": 0.23808321356773376,
+      "learning_rate": 7.00955658517996e-05,
+      "loss": 0.6805,
+      "step": 1681
+    },
+    {
+      "epoch": 0.387646923254206,
+      "grad_norm": 0.22761982679367065,
+      "learning_rate": 7.00613790621652e-05,
+      "loss": 0.6794,
+      "step": 1682
+    },
+    {
+      "epoch": 0.387877391103941,
+      "grad_norm": 0.22034461796283722,
+      "learning_rate": 7.002718109079964e-05,
+      "loss": 0.6863,
+      "step": 1683
+    },
+    {
+      "epoch": 0.38810785895367594,
+      "grad_norm": 0.24931013584136963,
+      "learning_rate": 6.999297195676399e-05,
+      "loss": 0.6806,
+      "step": 1684
+    },
+    {
+      "epoch": 0.3883383268034109,
+      "grad_norm": 0.24123869836330414,
+      "learning_rate": 6.995875167912566e-05,
+      "loss": 0.6765,
+      "step": 1685
+    },
+    {
+      "epoch": 0.38856879465314587,
+      "grad_norm": 0.20375779271125793,
+      "learning_rate": 6.992452027695821e-05,
+      "loss": 0.677,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3887992625028808,
+      "grad_norm": 0.2386665791273117,
+      "learning_rate": 6.989027776934138e-05,
+      "loss": 0.6855,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3890297303526158,
+      "grad_norm": 0.21312054991722107,
+      "learning_rate": 6.985602417536112e-05,
+      "loss": 0.6694,
+      "step": 1688
+    },
+    {
+      "epoch": 0.38926019820235075,
+      "grad_norm": 0.21280424296855927,
+      "learning_rate": 6.982175951410957e-05,
+      "loss": 0.674,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3894906660520857,
+      "grad_norm": 0.24086953699588776,
+      "learning_rate": 6.978748380468504e-05,
+      "loss": 0.6634,
+      "step": 1690
+    },
+    {
+      "epoch": 0.38972113390182067,
+      "grad_norm": 0.21306774020195007,
+      "learning_rate": 6.975319706619197e-05,
+      "loss": 0.6797,
+      "step": 1691
+    },
+    {
+      "epoch": 0.38995160175155563,
+      "grad_norm": 0.22666017711162567,
+      "learning_rate": 6.971889931774098e-05,
+      "loss": 0.6839,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3901820696012906,
+      "grad_norm": 0.22966359555721283,
+      "learning_rate": 6.968459057844881e-05,
+      "loss": 0.6755,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3904125374510256,
+      "grad_norm": 0.2565663158893585,
+      "learning_rate": 6.96502708674383e-05,
+      "loss": 0.6831,
+      "step": 1694
+    },
+    {
+      "epoch": 0.39064300530076057,
+      "grad_norm": 0.21080222725868225,
+      "learning_rate": 6.961594020383848e-05,
+      "loss": 0.6786,
+      "step": 1695
+    },
+    {
+      "epoch": 0.39087347315049553,
+      "grad_norm": 0.22605536878108978,
+      "learning_rate": 6.958159860678439e-05,
+      "loss": 0.6843,
+      "step": 1696
+    },
+    {
+      "epoch": 0.3911039410002305,
+      "grad_norm": 0.23553648591041565,
+      "learning_rate": 6.954724609541727e-05,
+      "loss": 0.6775,
+      "step": 1697
+    },
+    {
+      "epoch": 0.39133440884996545,
+      "grad_norm": 0.2261909395456314,
+      "learning_rate": 6.951288268888431e-05,
+      "loss": 0.6903,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3915648766997004,
+      "grad_norm": 0.201494961977005,
+      "learning_rate": 6.947850840633892e-05,
+      "loss": 0.6812,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3917953445494354,
+      "grad_norm": 0.24182942509651184,
+      "learning_rate": 6.944412326694046e-05,
+      "loss": 0.6807,
+      "step": 1700
+    },
+    {
+      "epoch": 0.39202581239917034,
+      "grad_norm": 0.2183101326227188,
+      "learning_rate": 6.940972728985438e-05,
+      "loss": 0.6847,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3922562802489053,
+      "grad_norm": 0.21912945806980133,
+      "learning_rate": 6.93753204942522e-05,
+      "loss": 0.6768,
+      "step": 1702
+    },
+    {
+      "epoch": 0.39248674809864026,
+      "grad_norm": 0.24168923497200012,
+      "learning_rate": 6.93409028993114e-05,
+      "loss": 0.6825,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3927172159483752,
+      "grad_norm": 0.21193869411945343,
+      "learning_rate": 6.930647452421557e-05,
+      "loss": 0.6891,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3929476837981102,
+      "grad_norm": 0.23223961889743805,
+      "learning_rate": 6.927203538815422e-05,
+      "loss": 0.6724,
+      "step": 1705
+    },
+    {
+      "epoch": 0.39317815164784514,
+      "grad_norm": 0.2386191338300705,
+      "learning_rate": 6.923758551032291e-05,
+      "loss": 0.6857,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3934086194975801,
+      "grad_norm": 0.2278871089220047,
+      "learning_rate": 6.92031249099232e-05,
+      "loss": 0.6871,
+      "step": 1707
+    },
+    {
+      "epoch": 0.39363908734731506,
+      "grad_norm": 0.25711220502853394,
+      "learning_rate": 6.916865360616256e-05,
+      "loss": 0.6841,
+      "step": 1708
+    },
+    {
+      "epoch": 0.39386955519705,
+      "grad_norm": 0.21602317690849304,
+      "learning_rate": 6.91341716182545e-05,
+      "loss": 0.6761,
+      "step": 1709
+    },
+    {
+      "epoch": 0.394100023046785,
+      "grad_norm": 0.24047702550888062,
+      "learning_rate": 6.909967896541843e-05,
+      "loss": 0.6855,
+      "step": 1710
+    },
+    {
+      "epoch": 0.39433049089651995,
+      "grad_norm": 0.2949325740337372,
+      "learning_rate": 6.906517566687973e-05,
+      "loss": 0.6794,
+      "step": 1711
+    },
+    {
+      "epoch": 0.3945609587462549,
+      "grad_norm": 0.2666512727737427,
+      "learning_rate": 6.903066174186974e-05,
+      "loss": 0.685,
+      "step": 1712
+    },
+    {
+      "epoch": 0.39479142659598987,
+      "grad_norm": 0.22977623343467712,
+      "learning_rate": 6.899613720962566e-05,
+      "loss": 0.6855,
+      "step": 1713
+    },
+    {
+      "epoch": 0.39502189444572483,
+      "grad_norm": 0.2291439175605774,
+      "learning_rate": 6.896160208939063e-05,
+      "loss": 0.6755,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3952523622954598,
+      "grad_norm": 0.26552703976631165,
+      "learning_rate": 6.892705640041373e-05,
+      "loss": 0.6886,
+      "step": 1715
+    },
+    {
+      "epoch": 0.39548283014519475,
+      "grad_norm": 0.20833250880241394,
+      "learning_rate": 6.889250016194988e-05,
+      "loss": 0.6832,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3957132979949297,
+      "grad_norm": 0.23018650710582733,
+      "learning_rate": 6.88579333932599e-05,
+      "loss": 0.6717,
+      "step": 1717
+    },
+    {
+      "epoch": 0.3959437658446647,
+      "grad_norm": 0.23923005163669586,
+      "learning_rate": 6.882335611361046e-05,
+      "loss": 0.6858,
+      "step": 1718
+    },
+    {
+      "epoch": 0.39617423369439964,
+      "grad_norm": 0.22628365457057953,
+      "learning_rate": 6.878876834227413e-05,
+      "loss": 0.6803,
+      "step": 1719
+    },
+    {
+      "epoch": 0.3964047015441346,
+      "grad_norm": 0.232679545879364,
+      "learning_rate": 6.87541700985293e-05,
+      "loss": 0.6903,
+      "step": 1720
+    },
+    {
+      "epoch": 0.39663516939386956,
+      "grad_norm": 0.24142929911613464,
+      "learning_rate": 6.871956140166019e-05,
+      "loss": 0.676,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3968656372436045,
+      "grad_norm": 0.21797588467597961,
+      "learning_rate": 6.868494227095688e-05,
+      "loss": 0.6824,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3970961050933395,
+      "grad_norm": 0.24194461107254028,
+      "learning_rate": 6.86503127257152e-05,
+      "loss": 0.6743,
+      "step": 1723
+    },
+    {
+      "epoch": 0.39732657294307444,
+      "grad_norm": 0.21700245141983032,
+      "learning_rate": 6.861567278523687e-05,
+      "loss": 0.6746,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3975570407928094,
+      "grad_norm": 0.2519386410713196,
+      "learning_rate": 6.858102246882934e-05,
+      "loss": 0.6816,
+      "step": 1725
+    },
+    {
+      "epoch": 0.39778750864254436,
+      "grad_norm": 0.2088238000869751,
+      "learning_rate": 6.854636179580587e-05,
+      "loss": 0.6775,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3980179764922793,
+      "grad_norm": 0.23535437881946564,
+      "learning_rate": 6.851169078548549e-05,
+      "loss": 0.6774,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3982484443420143,
+      "grad_norm": 0.22251394391059875,
+      "learning_rate": 6.847700945719298e-05,
+      "loss": 0.68,
+      "step": 1728
+    },
+    {
+      "epoch": 0.39847891219174925,
+      "grad_norm": 0.21713124215602875,
+      "learning_rate": 6.844231783025888e-05,
+      "loss": 0.6772,
+      "step": 1729
+    },
+    {
+      "epoch": 0.3987093800414842,
+      "grad_norm": 0.21638789772987366,
+      "learning_rate": 6.840761592401948e-05,
+      "loss": 0.6854,
+      "step": 1730
+    },
+    {
+      "epoch": 0.39893984789121917,
+      "grad_norm": 0.22768516838550568,
+      "learning_rate": 6.837290375781678e-05,
+      "loss": 0.6828,
+      "step": 1731
+    },
+    {
+      "epoch": 0.39917031574095413,
+      "grad_norm": 0.22035862505435944,
+      "learning_rate": 6.833818135099852e-05,
+      "loss": 0.678,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3994007835906891,
+      "grad_norm": 0.2309066206216812,
+      "learning_rate": 6.830344872291812e-05,
+      "loss": 0.6797,
+      "step": 1733
+    },
+    {
+      "epoch": 0.39963125144042405,
+      "grad_norm": 0.2268085926771164,
+      "learning_rate": 6.826870589293474e-05,
+      "loss": 0.6789,
+      "step": 1734
+    },
+    {
+      "epoch": 0.399861719290159,
+      "grad_norm": 0.22533169388771057,
+      "learning_rate": 6.823395288041316e-05,
+      "loss": 0.6791,
+      "step": 1735
+    },
+    {
+      "epoch": 0.400092187139894,
+      "grad_norm": 0.22329865396022797,
+      "learning_rate": 6.819918970472391e-05,
+      "loss": 0.6779,
+      "step": 1736
+    },
+    {
+      "epoch": 0.40032265498962893,
+      "grad_norm": 0.22880035638809204,
+      "learning_rate": 6.816441638524315e-05,
+      "loss": 0.6845,
+      "step": 1737
+    },
+    {
+      "epoch": 0.4005531228393639,
+      "grad_norm": 0.20969046652317047,
+      "learning_rate": 6.812963294135265e-05,
+      "loss": 0.6739,
+      "step": 1738
+    },
+    {
+      "epoch": 0.40078359068909886,
+      "grad_norm": 0.20743465423583984,
+      "learning_rate": 6.809483939243992e-05,
+      "loss": 0.6761,
+      "step": 1739
+    },
+    {
+      "epoch": 0.4010140585388338,
+      "grad_norm": 0.22215551137924194,
+      "learning_rate": 6.806003575789801e-05,
+      "loss": 0.6696,
+      "step": 1740
+    },
+    {
+      "epoch": 0.4012445263885688,
+      "grad_norm": 0.20283448696136475,
+      "learning_rate": 6.802522205712564e-05,
+      "loss": 0.6786,
+      "step": 1741
+    },
+    {
+      "epoch": 0.40147499423830374,
+      "grad_norm": 0.21726839244365692,
+      "learning_rate": 6.799039830952718e-05,
+      "loss": 0.6883,
+      "step": 1742
+    },
+    {
+      "epoch": 0.4017054620880387,
+      "grad_norm": 0.20720744132995605,
+      "learning_rate": 6.795556453451246e-05,
+      "loss": 0.6844,
+      "step": 1743
+    },
+    {
+      "epoch": 0.40193592993777366,
+      "grad_norm": 0.21686407923698425,
+      "learning_rate": 6.792072075149707e-05,
+      "loss": 0.6776,
+      "step": 1744
+    },
+    {
+      "epoch": 0.4021663977875086,
+      "grad_norm": 0.21741721034049988,
+      "learning_rate": 6.788586697990206e-05,
+      "loss": 0.6675,
+      "step": 1745
+    },
+    {
+      "epoch": 0.4023968656372436,
+      "grad_norm": 0.22603777050971985,
+      "learning_rate": 6.785100323915409e-05,
+      "loss": 0.6708,
+      "step": 1746
+    },
+    {
+      "epoch": 0.40262733348697854,
+      "grad_norm": 0.2569337487220764,
+      "learning_rate": 6.781612954868538e-05,
+      "loss": 0.6787,
+      "step": 1747
+    },
+    {
+      "epoch": 0.4028578013367135,
+      "grad_norm": 0.21277722716331482,
+      "learning_rate": 6.778124592793368e-05,
+      "loss": 0.6772,
+      "step": 1748
+    },
+    {
+      "epoch": 0.40308826918644847,
+      "grad_norm": 0.23380626738071442,
+      "learning_rate": 6.774635239634229e-05,
+      "loss": 0.6775,
+      "step": 1749
+    },
+    {
+      "epoch": 0.40331873703618343,
+      "grad_norm": 0.22136418521404266,
+      "learning_rate": 6.771144897336001e-05,
+      "loss": 0.679,
+      "step": 1750
+    },
+    {
+      "epoch": 0.4035492048859184,
+      "grad_norm": 0.22446192800998688,
+      "learning_rate": 6.767653567844121e-05,
+      "loss": 0.6803,
+      "step": 1751
+    },
+    {
+      "epoch": 0.40377967273565335,
+      "grad_norm": 0.21622978150844574,
+      "learning_rate": 6.764161253104567e-05,
+      "loss": 0.6771,
+      "step": 1752
+    },
+    {
+      "epoch": 0.4040101405853883,
+      "grad_norm": 0.2152906060218811,
+      "learning_rate": 6.760667955063876e-05,
+      "loss": 0.6848,
+      "step": 1753
+    },
+    {
+      "epoch": 0.4042406084351233,
+      "grad_norm": 0.19828496873378754,
+      "learning_rate": 6.757173675669127e-05,
+      "loss": 0.6788,
+      "step": 1754
+    },
+    {
+      "epoch": 0.4044710762848583,
+      "grad_norm": 0.20052795112133026,
+      "learning_rate": 6.753678416867944e-05,
+      "loss": 0.6759,
+      "step": 1755
+    },
+    {
+      "epoch": 0.40470154413459325,
+      "grad_norm": 0.21270537376403809,
+      "learning_rate": 6.750182180608505e-05,
+      "loss": 0.6801,
+      "step": 1756
+    },
+    {
+      "epoch": 0.4049320119843282,
+      "grad_norm": 0.20280341804027557,
+      "learning_rate": 6.746684968839525e-05,
+      "loss": 0.6724,
+      "step": 1757
+    },
+    {
+      "epoch": 0.40516247983406317,
+      "grad_norm": 0.22918108105659485,
+      "learning_rate": 6.743186783510269e-05,
+      "loss": 0.6797,
+      "step": 1758
+    },
+    {
+      "epoch": 0.40539294768379813,
+      "grad_norm": 0.23558905720710754,
+      "learning_rate": 6.739687626570541e-05,
+      "loss": 0.6875,
+      "step": 1759
+    },
+    {
+      "epoch": 0.4056234155335331,
+      "grad_norm": 0.19638538360595703,
+      "learning_rate": 6.736187499970684e-05,
+      "loss": 0.6751,
+      "step": 1760
+    },
+    {
+      "epoch": 0.40585388338326805,
+      "grad_norm": 0.19849823415279388,
+      "learning_rate": 6.732686405661587e-05,
+      "loss": 0.6771,
+      "step": 1761
+    },
+    {
+      "epoch": 0.406084351233003,
+      "grad_norm": 0.2098354548215866,
+      "learning_rate": 6.729184345594677e-05,
+      "loss": 0.6775,
+      "step": 1762
+    },
+    {
+      "epoch": 0.406314819082738,
+      "grad_norm": 0.20642666518688202,
+      "learning_rate": 6.725681321721916e-05,
+      "loss": 0.6746,
+      "step": 1763
+    },
+    {
+      "epoch": 0.40654528693247294,
+      "grad_norm": 0.23004627227783203,
+      "learning_rate": 6.722177335995805e-05,
+      "loss": 0.6811,
+      "step": 1764
+    },
+    {
+      "epoch": 0.4067757547822079,
+      "grad_norm": 0.21482717990875244,
+      "learning_rate": 6.718672390369384e-05,
+      "loss": 0.6762,
+      "step": 1765
+    },
+    {
+      "epoch": 0.40700622263194286,
+      "grad_norm": 0.19076307117938995,
+      "learning_rate": 6.715166486796225e-05,
+      "loss": 0.6849,
+      "step": 1766
+    },
+    {
+      "epoch": 0.4072366904816778,
+      "grad_norm": 0.23472625017166138,
+      "learning_rate": 6.711659627230432e-05,
+      "loss": 0.6638,
+      "step": 1767
+    },
+    {
+      "epoch": 0.4074671583314128,
+      "grad_norm": 0.20769445598125458,
+      "learning_rate": 6.708151813626648e-05,
+      "loss": 0.6732,
+      "step": 1768
+    },
+    {
+      "epoch": 0.40769762618114774,
+      "grad_norm": 0.2301913946866989,
+      "learning_rate": 6.704643047940041e-05,
+      "loss": 0.6707,
+      "step": 1769
+    },
+    {
+      "epoch": 0.4079280940308827,
+      "grad_norm": 0.2207859307527542,
+      "learning_rate": 6.701133332126313e-05,
+      "loss": 0.6799,
+      "step": 1770
+    },
+    {
+      "epoch": 0.40815856188061767,
+      "grad_norm": 0.20174138247966766,
+      "learning_rate": 6.697622668141698e-05,
+      "loss": 0.6815,
+      "step": 1771
+    },
+    {
+      "epoch": 0.4083890297303526,
+      "grad_norm": 0.24854540824890137,
+      "learning_rate": 6.694111057942953e-05,
+      "loss": 0.6759,
+      "step": 1772
+    },
+    {
+      "epoch": 0.4086194975800876,
+      "grad_norm": 0.2082047462463379,
+      "learning_rate": 6.690598503487368e-05,
+      "loss": 0.6792,
+      "step": 1773
+    },
+    {
+      "epoch": 0.40884996542982255,
+      "grad_norm": 0.2406819462776184,
+      "learning_rate": 6.687085006732755e-05,
+      "loss": 0.6852,
+      "step": 1774
+    },
+    {
+      "epoch": 0.4090804332795575,
+      "grad_norm": 0.22995240986347198,
+      "learning_rate": 6.683570569637451e-05,
+      "loss": 0.675,
+      "step": 1775
+    },
+    {
+      "epoch": 0.40931090112929247,
+      "grad_norm": 0.21281692385673523,
+      "learning_rate": 6.680055194160323e-05,
+      "loss": 0.6801,
+      "step": 1776
+    },
+    {
+      "epoch": 0.40954136897902743,
+      "grad_norm": 0.2113114297389984,
+      "learning_rate": 6.676538882260752e-05,
+      "loss": 0.6841,
+      "step": 1777
+    },
+    {
+      "epoch": 0.4097718368287624,
+      "grad_norm": 0.22743813693523407,
+      "learning_rate": 6.67302163589865e-05,
+      "loss": 0.6817,
+      "step": 1778
+    },
+    {
+      "epoch": 0.41000230467849735,
+      "grad_norm": 0.23003269731998444,
+      "learning_rate": 6.669503457034446e-05,
+      "loss": 0.6772,
+      "step": 1779
+    },
+    {
+      "epoch": 0.4102327725282323,
+      "grad_norm": 0.22872349619865417,
+      "learning_rate": 6.665984347629083e-05,
+      "loss": 0.6782,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4104632403779673,
+      "grad_norm": 0.2347979098558426,
+      "learning_rate": 6.662464309644034e-05,
+      "loss": 0.678,
+      "step": 1781
+    },
+    {
+      "epoch": 0.41069370822770224,
+      "grad_norm": 0.21414168179035187,
+      "learning_rate": 6.658943345041279e-05,
+      "loss": 0.6821,
+      "step": 1782
+    },
+    {
+      "epoch": 0.4109241760774372,
+      "grad_norm": 0.21668468415737152,
+      "learning_rate": 6.655421455783324e-05,
+      "loss": 0.6705,
+      "step": 1783
+    },
+    {
+      "epoch": 0.41115464392717216,
+      "grad_norm": 0.22316017746925354,
+      "learning_rate": 6.651898643833182e-05,
+      "loss": 0.6775,
+      "step": 1784
+    },
+    {
+      "epoch": 0.4113851117769071,
+      "grad_norm": 0.21024148166179657,
+      "learning_rate": 6.648374911154385e-05,
+      "loss": 0.678,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4116155796266421,
+      "grad_norm": 0.2320425808429718,
+      "learning_rate": 6.644850259710978e-05,
+      "loss": 0.6739,
+      "step": 1786
+    },
+    {
+      "epoch": 0.41184604747637704,
+      "grad_norm": 0.21307185292243958,
+      "learning_rate": 6.641324691467514e-05,
+      "loss": 0.6781,
+      "step": 1787
+    },
+    {
+      "epoch": 0.412076515326112,
+      "grad_norm": 0.21239805221557617,
+      "learning_rate": 6.637798208389063e-05,
+      "loss": 0.671,
+      "step": 1788
+    },
+    {
+      "epoch": 0.41230698317584696,
+      "grad_norm": 0.20224244892597198,
+      "learning_rate": 6.634270812441201e-05,
+      "loss": 0.6801,
+      "step": 1789
+    },
+    {
+      "epoch": 0.4125374510255819,
+      "grad_norm": 0.228766068816185,
+      "learning_rate": 6.630742505590015e-05,
+      "loss": 0.6785,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4127679188753169,
+      "grad_norm": 0.24188366532325745,
+      "learning_rate": 6.627213289802098e-05,
+      "loss": 0.6889,
+      "step": 1791
+    },
+    {
+      "epoch": 0.41299838672505185,
+      "grad_norm": 0.2053944319486618,
+      "learning_rate": 6.623683167044548e-05,
+      "loss": 0.6838,
+      "step": 1792
+    },
+    {
+      "epoch": 0.4132288545747868,
+      "grad_norm": 0.22618556022644043,
+      "learning_rate": 6.620152139284974e-05,
+      "loss": 0.6778,
+      "step": 1793
+    },
+    {
+      "epoch": 0.41345932242452177,
+      "grad_norm": 0.20091716945171356,
+      "learning_rate": 6.616620208491482e-05,
+      "loss": 0.6835,
+      "step": 1794
+    },
+    {
+      "epoch": 0.41368979027425673,
+      "grad_norm": 0.2244507372379303,
+      "learning_rate": 6.61308737663269e-05,
+      "loss": 0.6756,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4139202581239917,
+      "grad_norm": 0.24260321259498596,
+      "learning_rate": 6.609553645677714e-05,
+      "loss": 0.6853,
+      "step": 1796
+    },
+    {
+      "epoch": 0.41415072597372665,
+      "grad_norm": 0.21369081735610962,
+      "learning_rate": 6.606019017596164e-05,
+      "loss": 0.6953,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4143811938234616,
+      "grad_norm": 0.20895330607891083,
+      "learning_rate": 6.602483494358164e-05,
+      "loss": 0.6743,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4146116616731966,
+      "grad_norm": 0.231523796916008,
+      "learning_rate": 6.598947077934326e-05,
+      "loss": 0.6732,
+      "step": 1799
+    },
+    {
+      "epoch": 0.41484212952293154,
+      "grad_norm": 0.23038837313652039,
+      "learning_rate": 6.595409770295765e-05,
+      "loss": 0.6803,
+      "step": 1800
+    },
+    {
+      "epoch": 0.4150725973726665,
+      "grad_norm": 0.2283259779214859,
+      "learning_rate": 6.591871573414093e-05,
+      "loss": 0.6748,
+      "step": 1801
+    },
+    {
+      "epoch": 0.41530306522240146,
+      "grad_norm": 0.23936063051223755,
+      "learning_rate": 6.588332489261412e-05,
+      "loss": 0.676,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4155335330721364,
+      "grad_norm": 0.20093607902526855,
+      "learning_rate": 6.584792519810325e-05,
+      "loss": 0.6767,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4157640009218714,
+      "grad_norm": 0.24578706920146942,
+      "learning_rate": 6.581251667033927e-05,
+      "loss": 0.678,
+      "step": 1804
+    },
+    {
+      "epoch": 0.41599446877160634,
+      "grad_norm": 0.21614333987236023,
+      "learning_rate": 6.577709932905802e-05,
+      "loss": 0.6824,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4162249366213413,
+      "grad_norm": 0.22684995830059052,
+      "learning_rate": 6.574167319400028e-05,
+      "loss": 0.6815,
+      "step": 1806
+    },
+    {
+      "epoch": 0.41645540447107626,
+      "grad_norm": 0.2361087203025818,
+      "learning_rate": 6.570623828491175e-05,
+      "loss": 0.681,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4166858723208112,
+      "grad_norm": 0.2213921695947647,
+      "learning_rate": 6.567079462154298e-05,
+      "loss": 0.6683,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4169163401705462,
+      "grad_norm": 0.22478877007961273,
+      "learning_rate": 6.563534222364941e-05,
+      "loss": 0.6801,
+      "step": 1809
+    },
+    {
+      "epoch": 0.41714680802028115,
+      "grad_norm": 0.22775450348854065,
+      "learning_rate": 6.559988111099139e-05,
+      "loss": 0.6733,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4173772758700161,
+      "grad_norm": 0.21958664059638977,
+      "learning_rate": 6.556441130333403e-05,
+      "loss": 0.6759,
+      "step": 1811
+    },
+    {
+      "epoch": 0.41760774371975107,
+      "grad_norm": 0.21274985373020172,
+      "learning_rate": 6.55289328204474e-05,
+      "loss": 0.6758,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4178382115694861,
+      "grad_norm": 0.2310141921043396,
+      "learning_rate": 6.549344568210636e-05,
+      "loss": 0.6769,
+      "step": 1813
+    },
+    {
+      "epoch": 0.41806867941922105,
+      "grad_norm": 0.20216527581214905,
+      "learning_rate": 6.545794990809056e-05,
+      "loss": 0.6665,
+      "step": 1814
+    },
+    {
+      "epoch": 0.418299147268956,
+      "grad_norm": 0.22055986523628235,
+      "learning_rate": 6.54224455181845e-05,
+      "loss": 0.6646,
+      "step": 1815
+    },
+    {
+      "epoch": 0.41852961511869097,
+      "grad_norm": 0.22518932819366455,
+      "learning_rate": 6.538693253217748e-05,
+      "loss": 0.6865,
+      "step": 1816
+    },
+    {
+      "epoch": 0.41876008296842593,
+      "grad_norm": 0.23729877173900604,
+      "learning_rate": 6.535141096986359e-05,
+      "loss": 0.6711,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4189905508181609,
+      "grad_norm": 0.2299748808145523,
+      "learning_rate": 6.531588085104169e-05,
+      "loss": 0.6771,
+      "step": 1818
+    },
+    {
+      "epoch": 0.41922101866789585,
+      "grad_norm": 0.248263880610466,
+      "learning_rate": 6.528034219551543e-05,
+      "loss": 0.6709,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4194514865176308,
+      "grad_norm": 0.22252872586250305,
+      "learning_rate": 6.524479502309315e-05,
+      "loss": 0.6814,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4196819543673658,
+      "grad_norm": 0.24352537095546722,
+      "learning_rate": 6.520923935358806e-05,
+      "loss": 0.6792,
+      "step": 1821
+    },
+    {
+      "epoch": 0.41991242221710073,
+      "grad_norm": 0.24360835552215576,
+      "learning_rate": 6.517367520681801e-05,
+      "loss": 0.6874,
+      "step": 1822
+    },
+    {
+      "epoch": 0.4201428900668357,
+      "grad_norm": 0.24170473217964172,
+      "learning_rate": 6.513810260260558e-05,
+      "loss": 0.6815,
+      "step": 1823
+    },
+    {
+      "epoch": 0.42037335791657066,
+      "grad_norm": 0.2242230921983719,
+      "learning_rate": 6.510252156077813e-05,
+      "loss": 0.676,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4206038257663056,
+      "grad_norm": 0.22861500084400177,
+      "learning_rate": 6.506693210116764e-05,
+      "loss": 0.671,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4208342936160406,
+      "grad_norm": 0.23668934404850006,
+      "learning_rate": 6.503133424361082e-05,
+      "loss": 0.6788,
+      "step": 1826
+    },
+    {
+      "epoch": 0.42106476146577554,
+      "grad_norm": 0.2199348360300064,
+      "learning_rate": 6.499572800794911e-05,
+      "loss": 0.6786,
+      "step": 1827
+    },
+    {
+      "epoch": 0.4212952293155105,
+      "grad_norm": 0.23262521624565125,
+      "learning_rate": 6.496011341402852e-05,
+      "loss": 0.6797,
+      "step": 1828
+    },
+    {
+      "epoch": 0.42152569716524546,
+      "grad_norm": 0.2463759034872055,
+      "learning_rate": 6.492449048169977e-05,
+      "loss": 0.6763,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4217561650149804,
+      "grad_norm": 0.2045678198337555,
+      "learning_rate": 6.488885923081827e-05,
+      "loss": 0.6856,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4219866328647154,
+      "grad_norm": 0.23405107855796814,
+      "learning_rate": 6.485321968124398e-05,
+      "loss": 0.671,
+      "step": 1831
+    },
+    {
+      "epoch": 0.42221710071445034,
+      "grad_norm": 0.21921029686927795,
+      "learning_rate": 6.481757185284157e-05,
+      "loss": 0.6799,
+      "step": 1832
+    },
+    {
+      "epoch": 0.4224475685641853,
+      "grad_norm": 0.218912273645401,
+      "learning_rate": 6.478191576548024e-05,
+      "loss": 0.672,
+      "step": 1833
+    },
+    {
+      "epoch": 0.42267803641392027,
+      "grad_norm": 0.20524494349956512,
+      "learning_rate": 6.474625143903387e-05,
+      "loss": 0.686,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4229085042636552,
+      "grad_norm": 0.2237442284822464,
+      "learning_rate": 6.471057889338089e-05,
+      "loss": 0.6793,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4231389721133902,
+      "grad_norm": 0.21341606974601746,
+      "learning_rate": 6.46748981484043e-05,
+      "loss": 0.678,
+      "step": 1836
+    },
+    {
+      "epoch": 0.42336943996312515,
+      "grad_norm": 0.258621484041214,
+      "learning_rate": 6.463920922399173e-05,
+      "loss": 0.6849,
+      "step": 1837
+    },
+    {
+      "epoch": 0.4235999078128601,
+      "grad_norm": 0.23613744974136353,
+      "learning_rate": 6.46035121400353e-05,
+      "loss": 0.6809,
+      "step": 1838
+    },
+    {
+      "epoch": 0.42383037566259507,
+      "grad_norm": 0.21923868358135223,
+      "learning_rate": 6.456780691643171e-05,
+      "loss": 0.6752,
+      "step": 1839
+    },
+    {
+      "epoch": 0.42406084351233003,
+      "grad_norm": 0.26361706852912903,
+      "learning_rate": 6.453209357308224e-05,
+      "loss": 0.6801,
+      "step": 1840
+    },
+    {
+      "epoch": 0.424291311362065,
+      "grad_norm": 0.22276267409324646,
+      "learning_rate": 6.449637212989256e-05,
+      "loss": 0.6755,
+      "step": 1841
+    },
+    {
+      "epoch": 0.42452177921179995,
+      "grad_norm": 0.2298184633255005,
+      "learning_rate": 6.446064260677303e-05,
+      "loss": 0.6751,
+      "step": 1842
+    },
+    {
+      "epoch": 0.4247522470615349,
+      "grad_norm": 0.2290985882282257,
+      "learning_rate": 6.442490502363838e-05,
+      "loss": 0.6794,
+      "step": 1843
+    },
+    {
+      "epoch": 0.4249827149112699,
+      "grad_norm": 0.2414330095052719,
+      "learning_rate": 6.438915940040791e-05,
+      "loss": 0.6783,
+      "step": 1844
+    },
+    {
+      "epoch": 0.42521318276100484,
+      "grad_norm": 0.21323642134666443,
+      "learning_rate": 6.435340575700536e-05,
+      "loss": 0.6757,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4254436506107398,
+      "grad_norm": 0.24010320007801056,
+      "learning_rate": 6.431764411335894e-05,
+      "loss": 0.6741,
+      "step": 1846
+    },
+    {
+      "epoch": 0.42567411846047476,
+      "grad_norm": 0.22818125784397125,
+      "learning_rate": 6.428187448940136e-05,
+      "loss": 0.676,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4259045863102097,
+      "grad_norm": 0.26238662004470825,
+      "learning_rate": 6.424609690506972e-05,
+      "loss": 0.6724,
+      "step": 1848
+    },
+    {
+      "epoch": 0.4261350541599447,
+      "grad_norm": 0.21544672548770905,
+      "learning_rate": 6.421031138030562e-05,
+      "loss": 0.6768,
+      "step": 1849
+    },
+    {
+      "epoch": 0.42636552200967964,
+      "grad_norm": 0.25987011194229126,
+      "learning_rate": 6.417451793505502e-05,
+      "loss": 0.6761,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4265959898594146,
+      "grad_norm": 0.23844222724437714,
+      "learning_rate": 6.413871658926833e-05,
+      "loss": 0.6709,
+      "step": 1851
+    },
+    {
+      "epoch": 0.42682645770914956,
+      "grad_norm": 0.25137385725975037,
+      "learning_rate": 6.41029073629004e-05,
+      "loss": 0.6798,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4270569255588845,
+      "grad_norm": 0.2629540264606476,
+      "learning_rate": 6.406709027591039e-05,
+      "loss": 0.6829,
+      "step": 1853
+    },
+    {
+      "epoch": 0.4272873934086195,
+      "grad_norm": 0.27725306153297424,
+      "learning_rate": 6.403126534826189e-05,
+      "loss": 0.6825,
+      "step": 1854
+    },
+    {
+      "epoch": 0.42751786125835445,
+      "grad_norm": 0.2338194102048874,
+      "learning_rate": 6.399543259992288e-05,
+      "loss": 0.6717,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4277483291080894,
+      "grad_norm": 0.26456090807914734,
+      "learning_rate": 6.395959205086564e-05,
+      "loss": 0.6735,
+      "step": 1856
+    },
+    {
+      "epoch": 0.42797879695782437,
+      "grad_norm": 0.22425277531147003,
+      "learning_rate": 6.392374372106686e-05,
+      "loss": 0.6744,
+      "step": 1857
+    },
+    {
+      "epoch": 0.42820926480755933,
+      "grad_norm": 0.26653608679771423,
+      "learning_rate": 6.388788763050753e-05,
+      "loss": 0.6773,
+      "step": 1858
+    },
+    {
+      "epoch": 0.4284397326572943,
+      "grad_norm": 0.29735860228538513,
+      "learning_rate": 6.385202379917297e-05,
+      "loss": 0.6769,
+      "step": 1859
+    },
+    {
+      "epoch": 0.42867020050702925,
+      "grad_norm": 0.25533658266067505,
+      "learning_rate": 6.381615224705283e-05,
+      "loss": 0.6652,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4289006683567642,
+      "grad_norm": 0.27232304215431213,
+      "learning_rate": 6.378027299414104e-05,
+      "loss": 0.6763,
+      "step": 1861
+    },
+    {
+      "epoch": 0.4291311362064992,
+      "grad_norm": 0.25832486152648926,
+      "learning_rate": 6.374438606043582e-05,
+      "loss": 0.6839,
+      "step": 1862
+    },
+    {
+      "epoch": 0.42936160405623414,
+      "grad_norm": 0.22207960486412048,
+      "learning_rate": 6.370849146593973e-05,
+      "loss": 0.6797,
+      "step": 1863
+    },
+    {
+      "epoch": 0.4295920719059691,
+      "grad_norm": 0.2290259748697281,
+      "learning_rate": 6.367258923065951e-05,
+      "loss": 0.6747,
+      "step": 1864
+    },
+    {
+      "epoch": 0.42982253975570406,
+      "grad_norm": 0.24262063205242157,
+      "learning_rate": 6.363667937460624e-05,
+      "loss": 0.6784,
+      "step": 1865
+    },
+    {
+      "epoch": 0.430053007605439,
+      "grad_norm": 0.21408423781394958,
+      "learning_rate": 6.360076191779519e-05,
+      "loss": 0.6623,
+      "step": 1866
+    },
+    {
+      "epoch": 0.430283475455174,
+      "grad_norm": 0.2212885320186615,
+      "learning_rate": 6.356483688024588e-05,
+      "loss": 0.6676,
+      "step": 1867
+    },
+    {
+      "epoch": 0.43051394330490894,
+      "grad_norm": 0.23267394304275513,
+      "learning_rate": 6.352890428198208e-05,
+      "loss": 0.6764,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4307444111546439,
+      "grad_norm": 0.21774008870124817,
+      "learning_rate": 6.349296414303176e-05,
+      "loss": 0.6752,
+      "step": 1869
+    },
+    {
+      "epoch": 0.43097487900437886,
+      "grad_norm": 0.22179049253463745,
+      "learning_rate": 6.345701648342709e-05,
+      "loss": 0.6886,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4312053468541138,
+      "grad_norm": 0.2196541428565979,
+      "learning_rate": 6.342106132320442e-05,
+      "loss": 0.6793,
+      "step": 1871
+    },
+    {
+      "epoch": 0.4314358147038488,
+      "grad_norm": 0.22896122932434082,
+      "learning_rate": 6.338509868240432e-05,
+      "loss": 0.6643,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4316662825535838,
+      "grad_norm": 0.22660693526268005,
+      "learning_rate": 6.334912858107147e-05,
+      "loss": 0.6817,
+      "step": 1873
+    },
+    {
+      "epoch": 0.43189675040331876,
+      "grad_norm": 0.1979314535856247,
+      "learning_rate": 6.331315103925475e-05,
+      "loss": 0.6786,
+      "step": 1874
+    },
+    {
+      "epoch": 0.4321272182530537,
+      "grad_norm": 0.22958844900131226,
+      "learning_rate": 6.327716607700719e-05,
+      "loss": 0.6719,
+      "step": 1875
+    },
+    {
+      "epoch": 0.4323576861027887,
+      "grad_norm": 0.20848886668682098,
+      "learning_rate": 6.324117371438593e-05,
+      "loss": 0.6718,
+      "step": 1876
+    },
+    {
+      "epoch": 0.43258815395252365,
+      "grad_norm": 0.21527381241321564,
+      "learning_rate": 6.320517397145228e-05,
+      "loss": 0.6845,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4328186218022586,
+      "grad_norm": 0.20677097141742706,
+      "learning_rate": 6.316916686827159e-05,
+      "loss": 0.6769,
+      "step": 1878
+    },
+    {
+      "epoch": 0.43304908965199357,
+      "grad_norm": 0.20856349170207977,
+      "learning_rate": 6.313315242491338e-05,
+      "loss": 0.6758,
+      "step": 1879
+    },
+    {
+      "epoch": 0.43327955750172853,
+      "grad_norm": 0.21471652388572693,
+      "learning_rate": 6.309713066145123e-05,
+      "loss": 0.6686,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4335100253514635,
+      "grad_norm": 0.22820867598056793,
+      "learning_rate": 6.306110159796282e-05,
+      "loss": 0.6764,
+      "step": 1881
+    },
+    {
+      "epoch": 0.43374049320119845,
+      "grad_norm": 0.21438585221767426,
+      "learning_rate": 6.302506525452986e-05,
+      "loss": 0.675,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4339709610509334,
+      "grad_norm": 0.21153029799461365,
+      "learning_rate": 6.298902165123815e-05,
+      "loss": 0.6701,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4342014289006684,
+      "grad_norm": 0.21586838364601135,
+      "learning_rate": 6.295297080817754e-05,
+      "loss": 0.6741,
+      "step": 1884
+    },
+    {
+      "epoch": 0.43443189675040333,
+      "grad_norm": 0.20780424773693085,
+      "learning_rate": 6.29169127454419e-05,
+      "loss": 0.6765,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4346623646001383,
+      "grad_norm": 0.2166220098733902,
+      "learning_rate": 6.288084748312915e-05,
+      "loss": 0.6734,
+      "step": 1886
+    },
+    {
+      "epoch": 0.43489283244987326,
+      "grad_norm": 0.209901362657547,
+      "learning_rate": 6.284477504134116e-05,
+      "loss": 0.6738,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4351233002996082,
+      "grad_norm": 0.21214714646339417,
+      "learning_rate": 6.280869544018385e-05,
+      "loss": 0.6764,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4353537681493432,
+      "grad_norm": 0.22302688658237457,
+      "learning_rate": 6.277260869976716e-05,
+      "loss": 0.6727,
+      "step": 1889
+    },
+    {
+      "epoch": 0.43558423599907814,
+      "grad_norm": 0.21137331426143646,
+      "learning_rate": 6.273651484020492e-05,
+      "loss": 0.6813,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4358147038488131,
+      "grad_norm": 0.23412688076496124,
+      "learning_rate": 6.270041388161503e-05,
+      "loss": 0.6748,
+      "step": 1891
+    },
+    {
+      "epoch": 0.43604517169854806,
+      "grad_norm": 0.21604032814502716,
+      "learning_rate": 6.26643058441193e-05,
+      "loss": 0.676,
+      "step": 1892
+    },
+    {
+      "epoch": 0.436275639548283,
+      "grad_norm": 0.2466844618320465,
+      "learning_rate": 6.262819074784343e-05,
+      "loss": 0.6804,
+      "step": 1893
+    },
+    {
+      "epoch": 0.436506107398018,
+      "grad_norm": 0.2448147088289261,
+      "learning_rate": 6.259206861291716e-05,
+      "loss": 0.6694,
+      "step": 1894
+    },
+    {
+      "epoch": 0.43673657524775295,
+      "grad_norm": 0.23058217763900757,
+      "learning_rate": 6.255593945947407e-05,
+      "loss": 0.6764,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4369670430974879,
+      "grad_norm": 0.2556455731391907,
+      "learning_rate": 6.251980330765171e-05,
+      "loss": 0.6752,
+      "step": 1896
+    },
+    {
+      "epoch": 0.43719751094722287,
+      "grad_norm": 0.251406729221344,
+      "learning_rate": 6.248366017759146e-05,
+      "loss": 0.6751,
+      "step": 1897
+    },
+    {
+      "epoch": 0.43742797879695783,
+      "grad_norm": 0.2086581438779831,
+      "learning_rate": 6.244751008943867e-05,
+      "loss": 0.6774,
+      "step": 1898
+    },
+    {
+      "epoch": 0.4376584466466928,
+      "grad_norm": 0.2590232789516449,
+      "learning_rate": 6.241135306334254e-05,
+      "loss": 0.6764,
+      "step": 1899
+    },
+    {
+      "epoch": 0.43788891449642775,
+      "grad_norm": 0.23755843937397003,
+      "learning_rate": 6.237518911945608e-05,
+      "loss": 0.675,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4381193823461627,
+      "grad_norm": 0.20899838209152222,
+      "learning_rate": 6.233901827793625e-05,
+      "loss": 0.6699,
+      "step": 1901
+    },
+    {
+      "epoch": 0.4383498501958977,
+      "grad_norm": 0.2480127364397049,
+      "learning_rate": 6.230284055894379e-05,
+      "loss": 0.6698,
+      "step": 1902
+    },
+    {
+      "epoch": 0.43858031804563263,
+      "grad_norm": 0.23556362092494965,
+      "learning_rate": 6.22666559826433e-05,
+      "loss": 0.6734,
+      "step": 1903
+    },
+    {
+      "epoch": 0.4388107858953676,
+      "grad_norm": 0.2578951418399811,
+      "learning_rate": 6.223046456920321e-05,
+      "loss": 0.6689,
+      "step": 1904
+    },
+    {
+      "epoch": 0.43904125374510256,
+      "grad_norm": 0.275614470243454,
+      "learning_rate": 6.21942663387957e-05,
+      "loss": 0.6684,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4392717215948375,
+      "grad_norm": 0.21438564360141754,
+      "learning_rate": 6.215806131159683e-05,
+      "loss": 0.6766,
+      "step": 1906
+    },
+    {
+      "epoch": 0.4395021894445725,
+      "grad_norm": 0.26986971497535706,
+      "learning_rate": 6.21218495077864e-05,
+      "loss": 0.6716,
+      "step": 1907
+    },
+    {
+      "epoch": 0.43973265729430744,
+      "grad_norm": 0.25460711121559143,
+      "learning_rate": 6.208563094754802e-05,
+      "loss": 0.6763,
+      "step": 1908
+    },
+    {
+      "epoch": 0.4399631251440424,
+      "grad_norm": 0.2263481318950653,
+      "learning_rate": 6.2049405651069e-05,
+      "loss": 0.6738,
+      "step": 1909
+    },
+    {
+      "epoch": 0.44019359299377736,
+      "grad_norm": 0.26499027013778687,
+      "learning_rate": 6.20131736385405e-05,
+      "loss": 0.6708,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4404240608435123,
+      "grad_norm": 0.2109207808971405,
+      "learning_rate": 6.197693493015734e-05,
+      "loss": 0.6737,
+      "step": 1911
+    },
+    {
+      "epoch": 0.4406545286932473,
+      "grad_norm": 0.22647859156131744,
+      "learning_rate": 6.194068954611814e-05,
+      "loss": 0.6719,
+      "step": 1912
+    },
+    {
+      "epoch": 0.44088499654298224,
+      "grad_norm": 0.24029402434825897,
+      "learning_rate": 6.190443750662518e-05,
+      "loss": 0.6752,
+      "step": 1913
+    },
+    {
+      "epoch": 0.4411154643927172,
+      "grad_norm": 0.20866861939430237,
+      "learning_rate": 6.186817883188449e-05,
+      "loss": 0.6746,
+      "step": 1914
+    },
+    {
+      "epoch": 0.44134593224245217,
+      "grad_norm": 0.20965643227100372,
+      "learning_rate": 6.183191354210577e-05,
+      "loss": 0.675,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4415764000921871,
+      "grad_norm": 0.23772816359996796,
+      "learning_rate": 6.179564165750244e-05,
+      "loss": 0.6806,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4418068679419221,
+      "grad_norm": 0.2268814593553543,
+      "learning_rate": 6.175936319829157e-05,
+      "loss": 0.6682,
+      "step": 1917
+    },
+    {
+      "epoch": 0.44203733579165705,
+      "grad_norm": 0.24102425575256348,
+      "learning_rate": 6.17230781846939e-05,
+      "loss": 0.679,
+      "step": 1918
+    },
+    {
+      "epoch": 0.442267803641392,
+      "grad_norm": 0.22678816318511963,
+      "learning_rate": 6.168678663693382e-05,
+      "loss": 0.6703,
+      "step": 1919
+    },
+    {
+      "epoch": 0.44249827149112697,
+      "grad_norm": 0.22168827056884766,
+      "learning_rate": 6.165048857523938e-05,
+      "loss": 0.6867,
+      "step": 1920
+    },
+    {
+      "epoch": 0.44272873934086193,
+      "grad_norm": 0.23149913549423218,
+      "learning_rate": 6.161418401984225e-05,
+      "loss": 0.6705,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4429592071905969,
+      "grad_norm": 0.22349926829338074,
+      "learning_rate": 6.157787299097771e-05,
+      "loss": 0.6817,
+      "step": 1922
+    },
+    {
+      "epoch": 0.44318967504033185,
+      "grad_norm": 0.2404746562242508,
+      "learning_rate": 6.154155550888466e-05,
+      "loss": 0.6758,
+      "step": 1923
+    },
+    {
+      "epoch": 0.4434201428900668,
+      "grad_norm": 0.2592242956161499,
+      "learning_rate": 6.150523159380558e-05,
+      "loss": 0.6769,
+      "step": 1924
+    },
+    {
+      "epoch": 0.4436506107398018,
+      "grad_norm": 0.20442263782024384,
+      "learning_rate": 6.146890126598657e-05,
+      "loss": 0.6711,
+      "step": 1925
+    },
+    {
+      "epoch": 0.44388107858953674,
+      "grad_norm": 0.22541704773902893,
+      "learning_rate": 6.143256454567727e-05,
+      "loss": 0.674,
+      "step": 1926
+    },
+    {
+      "epoch": 0.4441115464392717,
+      "grad_norm": 0.20725850760936737,
+      "learning_rate": 6.139622145313089e-05,
+      "loss": 0.6663,
+      "step": 1927
+    },
+    {
+      "epoch": 0.44434201428900666,
+      "grad_norm": 0.21723538637161255,
+      "learning_rate": 6.13598720086042e-05,
+      "loss": 0.675,
+      "step": 1928
+    },
+    {
+      "epoch": 0.4445724821387416,
+      "grad_norm": 0.21730880439281464,
+      "learning_rate": 6.132351623235753e-05,
+      "loss": 0.6746,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4448029499884766,
+      "grad_norm": 0.23787382245063782,
+      "learning_rate": 6.12871541446547e-05,
+      "loss": 0.6767,
+      "step": 1930
+    },
+    {
+      "epoch": 0.44503341783821154,
+      "grad_norm": 0.23728126287460327,
+      "learning_rate": 6.125078576576306e-05,
+      "loss": 0.6811,
+      "step": 1931
+    },
+    {
+      "epoch": 0.44526388568794656,
+      "grad_norm": 0.2103312760591507,
+      "learning_rate": 6.121441111595347e-05,
+      "loss": 0.6702,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4454943535376815,
+      "grad_norm": 0.23246845602989197,
+      "learning_rate": 6.117803021550028e-05,
+      "loss": 0.6714,
+      "step": 1933
+    },
+    {
+      "epoch": 0.4457248213874165,
+      "grad_norm": 0.2163802981376648,
+      "learning_rate": 6.114164308468136e-05,
+      "loss": 0.6749,
+      "step": 1934
+    },
+    {
+      "epoch": 0.44595528923715144,
+      "grad_norm": 0.22772881388664246,
+      "learning_rate": 6.110524974377802e-05,
+      "loss": 0.6758,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4461857570868864,
+      "grad_norm": 0.21084089577198029,
+      "learning_rate": 6.1068850213075e-05,
+      "loss": 0.6662,
+      "step": 1936
+    },
+    {
+      "epoch": 0.44641622493662136,
+      "grad_norm": 0.21491113305091858,
+      "learning_rate": 6.1032444512860556e-05,
+      "loss": 0.6724,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4466466927863563,
+      "grad_norm": 0.2205878645181656,
+      "learning_rate": 6.099603266342635e-05,
+      "loss": 0.6817,
+      "step": 1938
+    },
+    {
+      "epoch": 0.4468771606360913,
+      "grad_norm": 0.22424796223640442,
+      "learning_rate": 6.0959614685067444e-05,
+      "loss": 0.6858,
+      "step": 1939
+    },
+    {
+      "epoch": 0.44710762848582625,
+      "grad_norm": 0.2058129459619522,
+      "learning_rate": 6.092319059808238e-05,
+      "loss": 0.6659,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4473380963355612,
+      "grad_norm": 0.22552870213985443,
+      "learning_rate": 6.088676042277306e-05,
+      "loss": 0.6786,
+      "step": 1941
+    },
+    {
+      "epoch": 0.44756856418529617,
+      "grad_norm": 0.24123747646808624,
+      "learning_rate": 6.085032417944477e-05,
+      "loss": 0.6724,
+      "step": 1942
+    },
+    {
+      "epoch": 0.44779903203503113,
+      "grad_norm": 0.2038024514913559,
+      "learning_rate": 6.081388188840623e-05,
+      "loss": 0.6653,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4480294998847661,
+      "grad_norm": 0.21841315925121307,
+      "learning_rate": 6.077743356996947e-05,
+      "loss": 0.6684,
+      "step": 1944
+    },
+    {
+      "epoch": 0.44825996773450105,
+      "grad_norm": 0.2158724069595337,
+      "learning_rate": 6.074097924444992e-05,
+      "loss": 0.6716,
+      "step": 1945
+    },
+    {
+      "epoch": 0.448490435584236,
+      "grad_norm": 0.2177586555480957,
+      "learning_rate": 6.0704518932166356e-05,
+      "loss": 0.6714,
+      "step": 1946
+    },
+    {
+      "epoch": 0.448720903433971,
+      "grad_norm": 0.21237412095069885,
+      "learning_rate": 6.066805265344084e-05,
+      "loss": 0.6753,
+      "step": 1947
+    },
+    {
+      "epoch": 0.44895137128370594,
+      "grad_norm": 0.22231976687908173,
+      "learning_rate": 6.0631580428598864e-05,
+      "loss": 0.6767,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4491818391334409,
+      "grad_norm": 0.200894296169281,
+      "learning_rate": 6.059510227796912e-05,
+      "loss": 0.6752,
+      "step": 1949
+    },
+    {
+      "epoch": 0.44941230698317586,
+      "grad_norm": 0.22096696496009827,
+      "learning_rate": 6.0558618221883664e-05,
+      "loss": 0.6732,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4496427748329108,
+      "grad_norm": 0.1898728609085083,
+      "learning_rate": 6.052212828067787e-05,
+      "loss": 0.6762,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4498732426826458,
+      "grad_norm": 0.21051624417304993,
+      "learning_rate": 6.0485632474690304e-05,
+      "loss": 0.675,
+      "step": 1952
+    },
+    {
+      "epoch": 0.45010371053238074,
+      "grad_norm": 0.20307575166225433,
+      "learning_rate": 6.0449130824262864e-05,
+      "loss": 0.6638,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4503341783821157,
+      "grad_norm": 0.2037128061056137,
+      "learning_rate": 6.041262334974072e-05,
+      "loss": 0.6732,
+      "step": 1954
+    },
+    {
+      "epoch": 0.45056464623185066,
+      "grad_norm": 0.20359142124652863,
+      "learning_rate": 6.0376110071472234e-05,
+      "loss": 0.6662,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4507951140815856,
+      "grad_norm": 0.19084343314170837,
+      "learning_rate": 6.033959100980905e-05,
+      "loss": 0.6774,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4510255819313206,
+      "grad_norm": 0.2061081826686859,
+      "learning_rate": 6.0303066185106e-05,
+      "loss": 0.674,
+      "step": 1957
+    },
+    {
+      "epoch": 0.45125604978105555,
+      "grad_norm": 0.19711264967918396,
+      "learning_rate": 6.026653561772118e-05,
+      "loss": 0.6689,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4514865176307905,
+      "grad_norm": 0.19926705956459045,
+      "learning_rate": 6.022999932801579e-05,
+      "loss": 0.6764,
+      "step": 1959
+    },
+    {
+      "epoch": 0.45171698548052547,
+      "grad_norm": 0.20394420623779297,
+      "learning_rate": 6.019345733635433e-05,
+      "loss": 0.6714,
+      "step": 1960
+    },
+    {
+      "epoch": 0.45194745333026043,
+      "grad_norm": 0.19170762598514557,
+      "learning_rate": 6.015690966310441e-05,
+      "loss": 0.6676,
+      "step": 1961
+    },
+    {
+      "epoch": 0.4521779211799954,
+      "grad_norm": 0.20363013446331024,
+      "learning_rate": 6.012035632863683e-05,
+      "loss": 0.6661,
+      "step": 1962
+    },
+    {
+      "epoch": 0.45240838902973035,
+      "grad_norm": 0.1984439343214035,
+      "learning_rate": 6.008379735332556e-05,
+      "loss": 0.67,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4526388568794653,
+      "grad_norm": 0.20422999560832977,
+      "learning_rate": 6.0047232757547654e-05,
+      "loss": 0.6804,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4528693247292003,
+      "grad_norm": 0.2078217715024948,
+      "learning_rate": 6.001066256168337e-05,
+      "loss": 0.6704,
+      "step": 1965
+    },
+    {
+      "epoch": 0.45309979257893523,
+      "grad_norm": 0.21460744738578796,
+      "learning_rate": 5.997408678611606e-05,
+      "loss": 0.674,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4533302604286702,
+      "grad_norm": 0.2205062061548233,
+      "learning_rate": 5.9937505451232165e-05,
+      "loss": 0.6722,
+      "step": 1967
+    },
+    {
+      "epoch": 0.45356072827840516,
+      "grad_norm": 0.18799827992916107,
+      "learning_rate": 5.990091857742126e-05,
+      "loss": 0.6728,
+      "step": 1968
+    },
+    {
+      "epoch": 0.4537911961281401,
+      "grad_norm": 0.2258245050907135,
+      "learning_rate": 5.986432618507598e-05,
+      "loss": 0.6753,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4540216639778751,
+      "grad_norm": 0.20360851287841797,
+      "learning_rate": 5.982772829459204e-05,
+      "loss": 0.6741,
+      "step": 1970
+    },
+    {
+      "epoch": 0.45425213182761004,
+      "grad_norm": 0.21974268555641174,
+      "learning_rate": 5.979112492636824e-05,
+      "loss": 0.6732,
+      "step": 1971
+    },
+    {
+      "epoch": 0.454482599677345,
+      "grad_norm": 0.18492695689201355,
+      "learning_rate": 5.9754516100806423e-05,
+      "loss": 0.667,
+      "step": 1972
+    },
+    {
+      "epoch": 0.45471306752707996,
+      "grad_norm": 0.2193833887577057,
+      "learning_rate": 5.971790183831145e-05,
+      "loss": 0.6748,
+      "step": 1973
+    },
+    {
+      "epoch": 0.4549435353768149,
+      "grad_norm": 0.19800156354904175,
+      "learning_rate": 5.968128215929123e-05,
+      "loss": 0.6672,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4551740032265499,
+      "grad_norm": 0.2352752387523651,
+      "learning_rate": 5.964465708415673e-05,
+      "loss": 0.6753,
+      "step": 1975
+    },
+    {
+      "epoch": 0.45540447107628484,
+      "grad_norm": 0.22403676807880402,
+      "learning_rate": 5.9608026633321846e-05,
+      "loss": 0.6752,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4556349389260198,
+      "grad_norm": 0.22172044217586517,
+      "learning_rate": 5.957139082720353e-05,
+      "loss": 0.6721,
+      "step": 1977
+    },
+    {
+      "epoch": 0.45586540677575477,
+      "grad_norm": 0.2300235778093338,
+      "learning_rate": 5.9534749686221715e-05,
+      "loss": 0.6779,
+      "step": 1978
+    },
+    {
+      "epoch": 0.45609587462548973,
+      "grad_norm": 0.2155136615037918,
+      "learning_rate": 5.9498103230799274e-05,
+      "loss": 0.6717,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4563263424752247,
+      "grad_norm": 0.2152310460805893,
+      "learning_rate": 5.9461451481362054e-05,
+      "loss": 0.6599,
+      "step": 1980
+    },
+    {
+      "epoch": 0.45655681032495965,
+      "grad_norm": 0.23655883967876434,
+      "learning_rate": 5.942479445833887e-05,
+      "loss": 0.6705,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4567872781746946,
+      "grad_norm": 0.20639467239379883,
+      "learning_rate": 5.938813218216149e-05,
+      "loss": 0.674,
+      "step": 1982
+    },
+    {
+      "epoch": 0.45701774602442957,
+      "grad_norm": 0.24684162437915802,
+      "learning_rate": 5.9351464673264556e-05,
+      "loss": 0.6707,
+      "step": 1983
+    },
+    {
+      "epoch": 0.45724821387416453,
+      "grad_norm": 0.2441384196281433,
+      "learning_rate": 5.931479195208566e-05,
+      "loss": 0.6608,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4574786817238995,
+      "grad_norm": 0.20324283838272095,
+      "learning_rate": 5.927811403906531e-05,
+      "loss": 0.6749,
+      "step": 1985
+    },
+    {
+      "epoch": 0.45770914957363446,
+      "grad_norm": 0.24871313571929932,
+      "learning_rate": 5.9241430954646884e-05,
+      "loss": 0.6715,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4579396174233694,
+      "grad_norm": 0.20936191082000732,
+      "learning_rate": 5.9204742719276676e-05,
+      "loss": 0.6773,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4581700852731044,
+      "grad_norm": 0.2653404474258423,
+      "learning_rate": 5.916804935340379e-05,
+      "loss": 0.6766,
+      "step": 1988
+    },
+    {
+      "epoch": 0.45840055312283934,
+      "grad_norm": 0.24194635450839996,
+      "learning_rate": 5.913135087748025e-05,
+      "loss": 0.6688,
+      "step": 1989
+    },
+    {
+      "epoch": 0.4586310209725743,
+      "grad_norm": 0.20647594332695007,
+      "learning_rate": 5.90946473119609e-05,
+      "loss": 0.675,
+      "step": 1990
+    },
+    {
+      "epoch": 0.45886148882230926,
+      "grad_norm": 0.26715630292892456,
+      "learning_rate": 5.9057938677303435e-05,
+      "loss": 0.6729,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4590919566720443,
+      "grad_norm": 0.21820828318595886,
+      "learning_rate": 5.902122499396836e-05,
+      "loss": 0.6683,
+      "step": 1992
+    },
+    {
+      "epoch": 0.45932242452177924,
+      "grad_norm": 0.2330540269613266,
+      "learning_rate": 5.898450628241899e-05,
+      "loss": 0.6566,
+      "step": 1993
+    },
+    {
+      "epoch": 0.4595528923715142,
+      "grad_norm": 0.2669338583946228,
+      "learning_rate": 5.894778256312149e-05,
+      "loss": 0.6655,
+      "step": 1994
+    },
+    {
+      "epoch": 0.45978336022124916,
+      "grad_norm": 0.21196076273918152,
+      "learning_rate": 5.891105385654474e-05,
+      "loss": 0.6765,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4600138280709841,
+      "grad_norm": 0.23386694490909576,
+      "learning_rate": 5.887432018316045e-05,
+      "loss": 0.6729,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4602442959207191,
+      "grad_norm": 0.21577976644039154,
+      "learning_rate": 5.88375815634431e-05,
+      "loss": 0.6745,
+      "step": 1997
+    },
+    {
+      "epoch": 0.46047476377045404,
+      "grad_norm": 0.2088993936777115,
+      "learning_rate": 5.8800838017869886e-05,
+      "loss": 0.671,
+      "step": 1998
+    },
+    {
+      "epoch": 0.460705231620189,
+      "grad_norm": 0.2429114431142807,
+      "learning_rate": 5.876408956692083e-05,
+      "loss": 0.6655,
+      "step": 1999
+    },
+    {
+      "epoch": 0.46093569946992397,
+      "grad_norm": 0.19586068391799927,
+      "learning_rate": 5.87273362310786e-05,
+      "loss": 0.6832,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4611661673196589,
+      "grad_norm": 0.22119756042957306,
+      "learning_rate": 5.86905780308286e-05,
+      "loss": 0.674,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4613966351693939,
+      "grad_norm": 0.2294689267873764,
+      "learning_rate": 5.8653814986659026e-05,
+      "loss": 0.6654,
+      "step": 2002
+    },
+    {
+      "epoch": 0.46162710301912885,
+      "grad_norm": 0.20585887134075165,
+      "learning_rate": 5.861704711906067e-05,
+      "loss": 0.6715,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4618575708688638,
+      "grad_norm": 0.22693046927452087,
+      "learning_rate": 5.8580274448527094e-05,
+      "loss": 0.6639,
+      "step": 2004
+    },
+    {
+      "epoch": 0.46208803871859877,
+      "grad_norm": 0.23352816700935364,
+      "learning_rate": 5.854349699555448e-05,
+      "loss": 0.674,
+      "step": 2005
+    },
+    {
+      "epoch": 0.46231850656833373,
+      "grad_norm": 0.21101726591587067,
+      "learning_rate": 5.850671478064169e-05,
+      "loss": 0.6763,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4625489744180687,
+      "grad_norm": 0.21366450190544128,
+      "learning_rate": 5.846992782429027e-05,
+      "loss": 0.6712,
+      "step": 2007
+    },
+    {
+      "epoch": 0.46277944226780365,
+      "grad_norm": 0.20312468707561493,
+      "learning_rate": 5.843313614700438e-05,
+      "loss": 0.6683,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4630099101175386,
+      "grad_norm": 0.21990333497524261,
+      "learning_rate": 5.8396339769290795e-05,
+      "loss": 0.6773,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4632403779672736,
+      "grad_norm": 0.1999209225177765,
+      "learning_rate": 5.8359538711658976e-05,
+      "loss": 0.6837,
+      "step": 2010
+    },
+    {
+      "epoch": 0.46347084581700854,
+      "grad_norm": 0.22440659999847412,
+      "learning_rate": 5.832273299462092e-05,
+      "loss": 0.6775,
+      "step": 2011
+    },
+    {
+      "epoch": 0.4637013136667435,
+      "grad_norm": 0.23072776198387146,
+      "learning_rate": 5.8285922638691246e-05,
+      "loss": 0.6676,
+      "step": 2012
+    },
+    {
+      "epoch": 0.46393178151647846,
+      "grad_norm": 0.2057875543832779,
+      "learning_rate": 5.824910766438718e-05,
+      "loss": 0.6673,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4641622493662134,
+      "grad_norm": 0.23501604795455933,
+      "learning_rate": 5.8212288092228504e-05,
+      "loss": 0.6752,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4643927172159484,
+      "grad_norm": 0.22704358398914337,
+      "learning_rate": 5.817546394273754e-05,
+      "loss": 0.6716,
+      "step": 2015
+    },
+    {
+      "epoch": 0.46462318506568334,
+      "grad_norm": 0.20482905209064484,
+      "learning_rate": 5.8138635236439207e-05,
+      "loss": 0.6605,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4648536529154183,
+      "grad_norm": 0.22710943222045898,
+      "learning_rate": 5.810180199386096e-05,
+      "loss": 0.6605,
+      "step": 2017
+    },
+    {
+      "epoch": 0.46508412076515326,
+      "grad_norm": 0.20111124217510223,
+      "learning_rate": 5.8064964235532705e-05,
+      "loss": 0.668,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4653145886148882,
+      "grad_norm": 0.21100562810897827,
+      "learning_rate": 5.802812198198699e-05,
+      "loss": 0.6779,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4655450564646232,
+      "grad_norm": 0.22643785178661346,
+      "learning_rate": 5.799127525375876e-05,
+      "loss": 0.6764,
+      "step": 2020
+    },
+    {
+      "epoch": 0.46577552431435815,
+      "grad_norm": 0.20172421634197235,
+      "learning_rate": 5.7954424071385505e-05,
+      "loss": 0.6709,
+      "step": 2021
+    },
+    {
+      "epoch": 0.4660059921640931,
+      "grad_norm": 0.20389583706855774,
+      "learning_rate": 5.791756845540721e-05,
+      "loss": 0.6713,
+      "step": 2022
+    },
+    {
+      "epoch": 0.46623646001382807,
+      "grad_norm": 0.1967121958732605,
+      "learning_rate": 5.788070842636629e-05,
+      "loss": 0.6654,
+      "step": 2023
+    },
+    {
+      "epoch": 0.46646692786356303,
+      "grad_norm": 0.20393262803554535,
+      "learning_rate": 5.784384400480765e-05,
+      "loss": 0.6693,
+      "step": 2024
+    },
+    {
+      "epoch": 0.466697395713298,
+      "grad_norm": 0.22986234724521637,
+      "learning_rate": 5.780697521127862e-05,
+      "loss": 0.6671,
+      "step": 2025
+    },
+    {
+      "epoch": 0.46692786356303295,
+      "grad_norm": 0.2337898313999176,
+      "learning_rate": 5.7770102066329e-05,
+      "loss": 0.6716,
+      "step": 2026
+    },
+    {
+      "epoch": 0.4671583314127679,
+      "grad_norm": 0.21430113911628723,
+      "learning_rate": 5.773322459051098e-05,
+      "loss": 0.668,
+      "step": 2027
+    },
+    {
+      "epoch": 0.4673887992625029,
+      "grad_norm": 0.2504282295703888,
+      "learning_rate": 5.769634280437919e-05,
+      "loss": 0.6752,
+      "step": 2028
+    },
+    {
+      "epoch": 0.46761926711223784,
+      "grad_norm": 0.2045115828514099,
+      "learning_rate": 5.765945672849066e-05,
+      "loss": 0.6687,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4678497349619728,
+      "grad_norm": 0.26005640625953674,
+      "learning_rate": 5.7622566383404774e-05,
+      "loss": 0.6812,
+      "step": 2030
+    },
+    {
+      "epoch": 0.46808020281170776,
+      "grad_norm": 0.21131455898284912,
+      "learning_rate": 5.758567178968336e-05,
+      "loss": 0.6679,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4683106706614427,
+      "grad_norm": 0.24453459680080414,
+      "learning_rate": 5.754877296789056e-05,
+      "loss": 0.6715,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4685411385111777,
+      "grad_norm": 0.25293347239494324,
+      "learning_rate": 5.751186993859287e-05,
+      "loss": 0.6685,
+      "step": 2033
+    },
+    {
+      "epoch": 0.46877160636091264,
+      "grad_norm": 0.20162692666053772,
+      "learning_rate": 5.74749627223592e-05,
+      "loss": 0.6718,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4690020742106476,
+      "grad_norm": 0.2305351048707962,
+      "learning_rate": 5.743805133976071e-05,
+      "loss": 0.6754,
+      "step": 2035
+    },
+    {
+      "epoch": 0.46923254206038256,
+      "grad_norm": 0.22010259330272675,
+      "learning_rate": 5.740113581137094e-05,
+      "loss": 0.6668,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4694630099101175,
+      "grad_norm": 0.23104584217071533,
+      "learning_rate": 5.736421615776573e-05,
+      "loss": 0.6791,
+      "step": 2037
+    },
+    {
+      "epoch": 0.4696934777598525,
+      "grad_norm": 0.2572833299636841,
+      "learning_rate": 5.732729239952316e-05,
+      "loss": 0.6708,
+      "step": 2038
+    },
+    {
+      "epoch": 0.46992394560958745,
+      "grad_norm": 0.230802983045578,
+      "learning_rate": 5.7290364557223685e-05,
+      "loss": 0.6669,
+      "step": 2039
+    },
+    {
+      "epoch": 0.4701544134593224,
+      "grad_norm": 0.20792588591575623,
+      "learning_rate": 5.725343265144999e-05,
+      "loss": 0.6723,
+      "step": 2040
+    },
+    {
+      "epoch": 0.47038488130905737,
+      "grad_norm": 0.22476236522197723,
+      "learning_rate": 5.721649670278704e-05,
+      "loss": 0.6714,
+      "step": 2041
+    },
+    {
+      "epoch": 0.47061534915879233,
+      "grad_norm": 0.1980828195810318,
+      "learning_rate": 5.717955673182202e-05,
+      "loss": 0.6707,
+      "step": 2042
+    },
+    {
+      "epoch": 0.4708458170085273,
+      "grad_norm": 0.2145676165819168,
+      "learning_rate": 5.714261275914442e-05,
+      "loss": 0.6716,
+      "step": 2043
+    },
+    {
+      "epoch": 0.47107628485826225,
+      "grad_norm": 0.19914093613624573,
+      "learning_rate": 5.71056648053459e-05,
+      "loss": 0.6762,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4713067527079972,
+      "grad_norm": 0.21874842047691345,
+      "learning_rate": 5.706871289102036e-05,
+      "loss": 0.667,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4715372205577322,
+      "grad_norm": 0.20119917392730713,
+      "learning_rate": 5.7031757036763934e-05,
+      "loss": 0.6658,
+      "step": 2046
+    },
+    {
+      "epoch": 0.47176768840746713,
+      "grad_norm": 0.21794843673706055,
+      "learning_rate": 5.69947972631749e-05,
+      "loss": 0.6702,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4719981562572021,
+      "grad_norm": 0.20612038671970367,
+      "learning_rate": 5.695783359085377e-05,
+      "loss": 0.663,
+      "step": 2048
+    },
+    {
+      "epoch": 0.47222862410693706,
+      "grad_norm": 0.20984764397144318,
+      "learning_rate": 5.69208660404032e-05,
+      "loss": 0.6743,
+      "step": 2049
+    },
+    {
+      "epoch": 0.472459091956672,
+      "grad_norm": 0.22582849860191345,
+      "learning_rate": 5.6883894632428005e-05,
+      "loss": 0.6827,
+      "step": 2050
+    },
+    {
+      "epoch": 0.47268955980640703,
+      "grad_norm": 0.1955862045288086,
+      "learning_rate": 5.684691938753517e-05,
+      "loss": 0.6679,
+      "step": 2051
+    },
+    {
+      "epoch": 0.472920027656142,
+      "grad_norm": 0.22272345423698425,
+      "learning_rate": 5.680994032633381e-05,
+      "loss": 0.6657,
+      "step": 2052
+    },
+    {
+      "epoch": 0.47315049550587696,
+      "grad_norm": 0.208807110786438,
+      "learning_rate": 5.6772957469435176e-05,
+      "loss": 0.6673,
+      "step": 2053
+    },
+    {
+      "epoch": 0.4733809633556119,
+      "grad_norm": 0.21152140200138092,
+      "learning_rate": 5.67359708374526e-05,
+      "loss": 0.6739,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4736114312053469,
+      "grad_norm": 0.22693230211734772,
+      "learning_rate": 5.669898045100156e-05,
+      "loss": 0.6648,
+      "step": 2055
+    },
+    {
+      "epoch": 0.47384189905508184,
+      "grad_norm": 0.22745169699192047,
+      "learning_rate": 5.6661986330699615e-05,
+      "loss": 0.6723,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4740723669048168,
+      "grad_norm": 0.217105895280838,
+      "learning_rate": 5.662498849716636e-05,
+      "loss": 0.6751,
+      "step": 2057
+    },
+    {
+      "epoch": 0.47430283475455176,
+      "grad_norm": 0.2365586906671524,
+      "learning_rate": 5.6587986971023564e-05,
+      "loss": 0.6833,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4745333026042867,
+      "grad_norm": 0.22789350152015686,
+      "learning_rate": 5.655098177289496e-05,
+      "loss": 0.673,
+      "step": 2059
+    },
+    {
+      "epoch": 0.4747637704540217,
+      "grad_norm": 0.20388175547122955,
+      "learning_rate": 5.651397292340632e-05,
+      "loss": 0.6667,
+      "step": 2060
+    },
+    {
+      "epoch": 0.47499423830375664,
+      "grad_norm": 0.22663913667201996,
+      "learning_rate": 5.6476960443185546e-05,
+      "loss": 0.676,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4752247061534916,
+      "grad_norm": 0.19801859557628632,
+      "learning_rate": 5.6439944352862476e-05,
+      "loss": 0.6707,
+      "step": 2062
+    },
+    {
+      "epoch": 0.47545517400322657,
+      "grad_norm": 0.21870480477809906,
+      "learning_rate": 5.640292467306899e-05,
+      "loss": 0.6705,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4756856418529615,
+      "grad_norm": 0.21825294196605682,
+      "learning_rate": 5.6365901424438985e-05,
+      "loss": 0.6729,
+      "step": 2064
+    },
+    {
+      "epoch": 0.4759161097026965,
+      "grad_norm": 0.19710072875022888,
+      "learning_rate": 5.632887462760831e-05,
+      "loss": 0.6783,
+      "step": 2065
+    },
+    {
+      "epoch": 0.47614657755243145,
+      "grad_norm": 0.21672062575817108,
+      "learning_rate": 5.6291844303214826e-05,
+      "loss": 0.6733,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4763770454021664,
+      "grad_norm": 0.1974102258682251,
+      "learning_rate": 5.625481047189835e-05,
+      "loss": 0.6797,
+      "step": 2067
+    },
+    {
+      "epoch": 0.47660751325190137,
+      "grad_norm": 0.220162034034729,
+      "learning_rate": 5.6217773154300646e-05,
+      "loss": 0.6668,
+      "step": 2068
+    },
+    {
+      "epoch": 0.47683798110163633,
+      "grad_norm": 0.20030367374420166,
+      "learning_rate": 5.618073237106541e-05,
+      "loss": 0.6704,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4770684489513713,
+      "grad_norm": 0.1981177181005478,
+      "learning_rate": 5.614368814283831e-05,
+      "loss": 0.6609,
+      "step": 2070
+    },
+    {
+      "epoch": 0.47729891680110625,
+      "grad_norm": 0.19721218943595886,
+      "learning_rate": 5.6106640490266904e-05,
+      "loss": 0.665,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4775293846508412,
+      "grad_norm": 0.20160210132598877,
+      "learning_rate": 5.606958943400066e-05,
+      "loss": 0.6687,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4777598525005762,
+      "grad_norm": 0.20998936891555786,
+      "learning_rate": 5.6032534994690945e-05,
+      "loss": 0.6719,
+      "step": 2073
+    },
+    {
+      "epoch": 0.47799032035031114,
+      "grad_norm": 0.19076962769031525,
+      "learning_rate": 5.599547719299102e-05,
+      "loss": 0.6719,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4782207882000461,
+      "grad_norm": 0.2014273852109909,
+      "learning_rate": 5.595841604955601e-05,
+      "loss": 0.6757,
+      "step": 2075
+    },
+    {
+      "epoch": 0.47845125604978106,
+      "grad_norm": 0.20173531770706177,
+      "learning_rate": 5.5921351585042915e-05,
+      "loss": 0.6707,
+      "step": 2076
+    },
+    {
+      "epoch": 0.478681723899516,
+      "grad_norm": 0.2177407145500183,
+      "learning_rate": 5.588428382011055e-05,
+      "loss": 0.6611,
+      "step": 2077
+    },
+    {
+      "epoch": 0.478912191749251,
+      "grad_norm": 0.22235806286334991,
+      "learning_rate": 5.584721277541964e-05,
+      "loss": 0.6692,
+      "step": 2078
+    },
+    {
+      "epoch": 0.47914265959898594,
+      "grad_norm": 0.19388116896152496,
+      "learning_rate": 5.581013847163267e-05,
+      "loss": 0.6679,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4793731274487209,
+      "grad_norm": 0.23979228734970093,
+      "learning_rate": 5.577306092941397e-05,
+      "loss": 0.6729,
+      "step": 2080
+    },
+    {
+      "epoch": 0.47960359529845586,
+      "grad_norm": 0.22609175741672516,
+      "learning_rate": 5.573598016942968e-05,
+      "loss": 0.6638,
+      "step": 2081
+    },
+    {
+      "epoch": 0.4798340631481908,
+      "grad_norm": 0.20589977502822876,
+      "learning_rate": 5.569889621234771e-05,
+      "loss": 0.667,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4800645309979258,
+      "grad_norm": 0.23106610774993896,
+      "learning_rate": 5.566180907883777e-05,
+      "loss": 0.6704,
+      "step": 2083
+    },
+    {
+      "epoch": 0.48029499884766075,
+      "grad_norm": 0.19989155232906342,
+      "learning_rate": 5.562471878957135e-05,
+      "loss": 0.674,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4805254666973957,
+      "grad_norm": 0.22377397119998932,
+      "learning_rate": 5.55876253652217e-05,
+      "loss": 0.6624,
+      "step": 2085
+    },
+    {
+      "epoch": 0.48075593454713067,
+      "grad_norm": 0.21617180109024048,
+      "learning_rate": 5.5550528826463754e-05,
+      "loss": 0.6764,
+      "step": 2086
+    },
+    {
+      "epoch": 0.48098640239686563,
+      "grad_norm": 0.20691335201263428,
+      "learning_rate": 5.551342919397429e-05,
+      "loss": 0.6638,
+      "step": 2087
+    },
+    {
+      "epoch": 0.4812168702466006,
+      "grad_norm": 0.20599327981472015,
+      "learning_rate": 5.547632648843172e-05,
+      "loss": 0.665,
+      "step": 2088
+    },
+    {
+      "epoch": 0.48144733809633555,
+      "grad_norm": 0.21349526941776276,
+      "learning_rate": 5.54392207305162e-05,
+      "loss": 0.6734,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4816778059460705,
+      "grad_norm": 0.20899753272533417,
+      "learning_rate": 5.5402111940909595e-05,
+      "loss": 0.6702,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4819082737958055,
+      "grad_norm": 0.22185955941677094,
+      "learning_rate": 5.536500014029547e-05,
+      "loss": 0.6638,
+      "step": 2091
+    },
+    {
+      "epoch": 0.48213874164554044,
+      "grad_norm": 0.19926568865776062,
+      "learning_rate": 5.532788534935902e-05,
+      "loss": 0.6649,
+      "step": 2092
+    },
+    {
+      "epoch": 0.4823692094952754,
+      "grad_norm": 0.22930283844470978,
+      "learning_rate": 5.529076758878718e-05,
+      "loss": 0.6691,
+      "step": 2093
+    },
+    {
+      "epoch": 0.48259967734501036,
+      "grad_norm": 0.2192930430173874,
+      "learning_rate": 5.525364687926846e-05,
+      "loss": 0.6681,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4828301451947453,
+      "grad_norm": 0.23378318548202515,
+      "learning_rate": 5.521652324149307e-05,
+      "loss": 0.6736,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4830606130444803,
+      "grad_norm": 0.20761679112911224,
+      "learning_rate": 5.517939669615284e-05,
+      "loss": 0.6678,
+      "step": 2096
+    },
+    {
+      "epoch": 0.48329108089421524,
+      "grad_norm": 0.22278013825416565,
+      "learning_rate": 5.5142267263941204e-05,
+      "loss": 0.6649,
+      "step": 2097
+    },
+    {
+      "epoch": 0.4835215487439502,
+      "grad_norm": 0.2117822915315628,
+      "learning_rate": 5.510513496555322e-05,
+      "loss": 0.6664,
+      "step": 2098
+    },
+    {
+      "epoch": 0.48375201659368516,
+      "grad_norm": 0.20482023060321808,
+      "learning_rate": 5.506799982168553e-05,
+      "loss": 0.6709,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4839824844434201,
+      "grad_norm": 0.2443208247423172,
+      "learning_rate": 5.50308618530364e-05,
+      "loss": 0.6694,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4842129522931551,
+      "grad_norm": 0.1926216185092926,
+      "learning_rate": 5.4993721080305614e-05,
+      "loss": 0.6608,
+      "step": 2101
+    },
+    {
+      "epoch": 0.48444342014289005,
+      "grad_norm": 0.22089843451976776,
+      "learning_rate": 5.495657752419455e-05,
+      "loss": 0.6718,
+      "step": 2102
+    },
+    {
+      "epoch": 0.484673887992625,
+      "grad_norm": 0.22440844774246216,
+      "learning_rate": 5.491943120540616e-05,
+      "loss": 0.6648,
+      "step": 2103
+    },
+    {
+      "epoch": 0.48490435584235997,
+      "grad_norm": 0.19866810739040375,
+      "learning_rate": 5.488228214464487e-05,
+      "loss": 0.6596,
+      "step": 2104
+    },
+    {
+      "epoch": 0.48513482369209493,
+      "grad_norm": 0.22278371453285217,
+      "learning_rate": 5.484513036261671e-05,
+      "loss": 0.6685,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4853652915418299,
+      "grad_norm": 0.21866042912006378,
+      "learning_rate": 5.480797588002918e-05,
+      "loss": 0.6683,
+      "step": 2106
+    },
+    {
+      "epoch": 0.48559575939156485,
+      "grad_norm": 0.20206943154335022,
+      "learning_rate": 5.47708187175913e-05,
+      "loss": 0.6705,
+      "step": 2107
+    },
+    {
+      "epoch": 0.4858262272412998,
+      "grad_norm": 0.2211609184741974,
+      "learning_rate": 5.4733658896013575e-05,
+      "loss": 0.6669,
+      "step": 2108
+    },
+    {
+      "epoch": 0.4860566950910348,
+      "grad_norm": 0.20941609144210815,
+      "learning_rate": 5.4696496436008e-05,
+      "loss": 0.6641,
+      "step": 2109
+    },
+    {
+      "epoch": 0.48628716294076973,
+      "grad_norm": 0.21404802799224854,
+      "learning_rate": 5.465933135828802e-05,
+      "loss": 0.6754,
+      "step": 2110
+    },
+    {
+      "epoch": 0.48651763079050475,
+      "grad_norm": 0.2135351300239563,
+      "learning_rate": 5.4622163683568584e-05,
+      "loss": 0.6755,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4867480986402397,
+      "grad_norm": 0.22787488996982574,
+      "learning_rate": 5.4584993432566066e-05,
+      "loss": 0.6608,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4869785664899747,
+      "grad_norm": 0.22770029306411743,
+      "learning_rate": 5.4547820625998244e-05,
+      "loss": 0.6669,
+      "step": 2113
+    },
+    {
+      "epoch": 0.48720903433970963,
+      "grad_norm": 0.2112642526626587,
+      "learning_rate": 5.4510645284584364e-05,
+      "loss": 0.6636,
+      "step": 2114
+    },
+    {
+      "epoch": 0.4874395021894446,
+      "grad_norm": 0.22324852645397186,
+      "learning_rate": 5.447346742904508e-05,
+      "loss": 0.679,
+      "step": 2115
+    },
+    {
+      "epoch": 0.48766997003917956,
+      "grad_norm": 0.2138347178697586,
+      "learning_rate": 5.443628708010239e-05,
+      "loss": 0.6636,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4879004378889145,
+      "grad_norm": 0.22339309751987457,
+      "learning_rate": 5.439910425847979e-05,
+      "loss": 0.6648,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4881309057386495,
+      "grad_norm": 0.22265255451202393,
+      "learning_rate": 5.436191898490207e-05,
+      "loss": 0.6686,
+      "step": 2118
+    },
+    {
+      "epoch": 0.48836137358838444,
+      "grad_norm": 0.22078566253185272,
+      "learning_rate": 5.4324731280095374e-05,
+      "loss": 0.6704,
+      "step": 2119
+    },
+    {
+      "epoch": 0.4885918414381194,
+      "grad_norm": 0.19747048616409302,
+      "learning_rate": 5.428754116478729e-05,
+      "loss": 0.6723,
+      "step": 2120
+    },
+    {
+      "epoch": 0.48882230928785436,
+      "grad_norm": 0.22434230148792267,
+      "learning_rate": 5.425034865970666e-05,
+      "loss": 0.6769,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4890527771375893,
+      "grad_norm": 0.2001955658197403,
+      "learning_rate": 5.4213153785583705e-05,
+      "loss": 0.6748,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4892832449873243,
+      "grad_norm": 0.2093842476606369,
+      "learning_rate": 5.417595656314997e-05,
+      "loss": 0.666,
+      "step": 2123
+    },
+    {
+      "epoch": 0.48951371283705924,
+      "grad_norm": 0.18802720308303833,
+      "learning_rate": 5.413875701313825e-05,
+      "loss": 0.6667,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4897441806867942,
+      "grad_norm": 0.2099105566740036,
+      "learning_rate": 5.410155515628272e-05,
+      "loss": 0.6793,
+      "step": 2125
+    },
+    {
+      "epoch": 0.48997464853652917,
+      "grad_norm": 0.1877232789993286,
+      "learning_rate": 5.406435101331879e-05,
+      "loss": 0.6583,
+      "step": 2126
+    },
+    {
+      "epoch": 0.49020511638626413,
+      "grad_norm": 0.21019746363162994,
+      "learning_rate": 5.402714460498318e-05,
+      "loss": 0.6643,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4904355842359991,
+      "grad_norm": 0.19312387704849243,
+      "learning_rate": 5.39899359520138e-05,
+      "loss": 0.666,
+      "step": 2128
+    },
+    {
+      "epoch": 0.49066605208573405,
+      "grad_norm": 0.19418299198150635,
+      "learning_rate": 5.39527250751499e-05,
+      "loss": 0.6663,
+      "step": 2129
+    },
+    {
+      "epoch": 0.490896519935469,
+      "grad_norm": 0.19869892299175262,
+      "learning_rate": 5.391551199513192e-05,
+      "loss": 0.67,
+      "step": 2130
+    },
+    {
+      "epoch": 0.491126987785204,
+      "grad_norm": 0.19379819929599762,
+      "learning_rate": 5.3878296732701515e-05,
+      "loss": 0.6661,
+      "step": 2131
+    },
+    {
+      "epoch": 0.49135745563493893,
+      "grad_norm": 0.20687328279018402,
+      "learning_rate": 5.384107930860162e-05,
+      "loss": 0.6815,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4915879234846739,
+      "grad_norm": 0.20180322229862213,
+      "learning_rate": 5.38038597435763e-05,
+      "loss": 0.6631,
+      "step": 2133
+    },
+    {
+      "epoch": 0.49181839133440886,
+      "grad_norm": 0.19170688092708588,
+      "learning_rate": 5.3766638058370855e-05,
+      "loss": 0.6794,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4920488591841438,
+      "grad_norm": 0.1976936012506485,
+      "learning_rate": 5.372941427373178e-05,
+      "loss": 0.6709,
+      "step": 2135
+    },
+    {
+      "epoch": 0.4922793270338788,
+      "grad_norm": 0.1848510205745697,
+      "learning_rate": 5.3692188410406695e-05,
+      "loss": 0.6664,
+      "step": 2136
+    },
+    {
+      "epoch": 0.49250979488361374,
+      "grad_norm": 0.19432882964611053,
+      "learning_rate": 5.36549604891444e-05,
+      "loss": 0.6613,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4927402627333487,
+      "grad_norm": 0.19802847504615784,
+      "learning_rate": 5.361773053069487e-05,
+      "loss": 0.663,
+      "step": 2138
+    },
+    {
+      "epoch": 0.49297073058308366,
+      "grad_norm": 0.19331657886505127,
+      "learning_rate": 5.3580498555809163e-05,
+      "loss": 0.6743,
+      "step": 2139
+    },
+    {
+      "epoch": 0.4932011984328186,
+      "grad_norm": 0.20699891448020935,
+      "learning_rate": 5.354326458523952e-05,
+      "loss": 0.6685,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4934316662825536,
+      "grad_norm": 0.17387264966964722,
+      "learning_rate": 5.350602863973923e-05,
+      "loss": 0.6707,
+      "step": 2141
+    },
+    {
+      "epoch": 0.49366213413228854,
+      "grad_norm": 0.205754354596138,
+      "learning_rate": 5.346879074006271e-05,
+      "loss": 0.6634,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4938926019820235,
+      "grad_norm": 0.19446855783462524,
+      "learning_rate": 5.343155090696551e-05,
+      "loss": 0.6784,
+      "step": 2143
+    },
+    {
+      "epoch": 0.49412306983175847,
+      "grad_norm": 0.20317377150058746,
+      "learning_rate": 5.33943091612042e-05,
+      "loss": 0.6599,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4943535376814934,
+      "grad_norm": 0.1977040022611618,
+      "learning_rate": 5.335706552353643e-05,
+      "loss": 0.6631,
+      "step": 2145
+    },
+    {
+      "epoch": 0.4945840055312284,
+      "grad_norm": 0.19229522347450256,
+      "learning_rate": 5.331982001472091e-05,
+      "loss": 0.6602,
+      "step": 2146
+    },
+    {
+      "epoch": 0.49481447338096335,
+      "grad_norm": 0.20331966876983643,
+      "learning_rate": 5.3282572655517416e-05,
+      "loss": 0.6738,
+      "step": 2147
+    },
+    {
+      "epoch": 0.4950449412306983,
+      "grad_norm": 0.20568066835403442,
+      "learning_rate": 5.324532346668668e-05,
+      "loss": 0.6712,
+      "step": 2148
+    },
+    {
+      "epoch": 0.49527540908043327,
+      "grad_norm": 0.20340462028980255,
+      "learning_rate": 5.3208072468990555e-05,
+      "loss": 0.6651,
+      "step": 2149
+    },
+    {
+      "epoch": 0.49550587693016823,
+      "grad_norm": 0.2173525094985962,
+      "learning_rate": 5.317081968319185e-05,
+      "loss": 0.6651,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4957363447799032,
+      "grad_norm": 0.19157260656356812,
+      "learning_rate": 5.313356513005433e-05,
+      "loss": 0.6687,
+      "step": 2151
+    },
+    {
+      "epoch": 0.49596681262963815,
+      "grad_norm": 0.20568935573101044,
+      "learning_rate": 5.3096308830342844e-05,
+      "loss": 0.6671,
+      "step": 2152
+    },
+    {
+      "epoch": 0.4961972804793731,
+      "grad_norm": 0.19956238567829132,
+      "learning_rate": 5.305905080482312e-05,
+      "loss": 0.6649,
+      "step": 2153
+    },
+    {
+      "epoch": 0.4964277483291081,
+      "grad_norm": 0.19893619418144226,
+      "learning_rate": 5.302179107426191e-05,
+      "loss": 0.6692,
+      "step": 2154
+    },
+    {
+      "epoch": 0.49665821617884304,
+      "grad_norm": 0.20772667229175568,
+      "learning_rate": 5.298452965942687e-05,
+      "loss": 0.6686,
+      "step": 2155
+    },
+    {
+      "epoch": 0.496888684028578,
+      "grad_norm": 0.1938534379005432,
+      "learning_rate": 5.294726658108665e-05,
+      "loss": 0.6659,
+      "step": 2156
+    },
+    {
+      "epoch": 0.49711915187831296,
+      "grad_norm": 0.19728915393352509,
+      "learning_rate": 5.291000186001076e-05,
+      "loss": 0.6745,
+      "step": 2157
+    },
+    {
+      "epoch": 0.4973496197280479,
+      "grad_norm": 0.19719786942005157,
+      "learning_rate": 5.2872735516969695e-05,
+      "loss": 0.6713,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4975800875777829,
+      "grad_norm": 0.18405082821846008,
+      "learning_rate": 5.28354675727348e-05,
+      "loss": 0.6542,
+      "step": 2159
+    },
+    {
+      "epoch": 0.49781055542751784,
+      "grad_norm": 0.2271628975868225,
+      "learning_rate": 5.279819804807834e-05,
+      "loss": 0.6674,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4980410232772528,
+      "grad_norm": 0.18239864706993103,
+      "learning_rate": 5.2760926963773436e-05,
+      "loss": 0.6755,
+      "step": 2161
+    },
+    {
+      "epoch": 0.49827149112698776,
+      "grad_norm": 0.19707393646240234,
+      "learning_rate": 5.272365434059413e-05,
+      "loss": 0.6676,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4985019589767227,
+      "grad_norm": 0.21636438369750977,
+      "learning_rate": 5.2686380199315244e-05,
+      "loss": 0.6743,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4987324268264577,
+      "grad_norm": 0.2144385576248169,
+      "learning_rate": 5.2649104560712536e-05,
+      "loss": 0.6648,
+      "step": 2164
+    },
+    {
+      "epoch": 0.49896289467619265,
+      "grad_norm": 0.21302272379398346,
+      "learning_rate": 5.261182744556252e-05,
+      "loss": 0.6673,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4991933625259276,
+      "grad_norm": 0.22367598116397858,
+      "learning_rate": 5.257454887464258e-05,
+      "loss": 0.6754,
+      "step": 2166
+    },
+    {
+      "epoch": 0.49942383037566257,
+      "grad_norm": 0.1969953179359436,
+      "learning_rate": 5.253726886873089e-05,
+      "loss": 0.6759,
+      "step": 2167
+    },
+    {
+      "epoch": 0.49965429822539753,
+      "grad_norm": 0.22727370262145996,
+      "learning_rate": 5.2499987448606436e-05,
+      "loss": 0.6622,
+      "step": 2168
+    },
+    {
+      "epoch": 0.4998847660751325,
+      "grad_norm": 0.19945083558559418,
+      "learning_rate": 5.246270463504898e-05,
+      "loss": 0.6689,
+      "step": 2169
+    },
+    {
+      "epoch": 0.5001152339248675,
+      "grad_norm": 0.19829121232032776,
+      "learning_rate": 5.2425420448839055e-05,
+      "loss": 0.6707,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5003457017746025,
+      "grad_norm": 0.19157467782497406,
+      "learning_rate": 5.2388134910758015e-05,
+      "loss": 0.6674,
+      "step": 2171
+    },
+    {
+      "epoch": 0.5005761696243374,
+      "grad_norm": 0.19981160759925842,
+      "learning_rate": 5.235084804158787e-05,
+      "loss": 0.6703,
+      "step": 2172
+    },
+    {
+      "epoch": 0.5008066374740724,
+      "grad_norm": 0.2080090194940567,
+      "learning_rate": 5.231355986211146e-05,
+      "loss": 0.6659,
+      "step": 2173
+    },
+    {
+      "epoch": 0.5010371053238073,
+      "grad_norm": 0.19367657601833344,
+      "learning_rate": 5.2276270393112325e-05,
+      "loss": 0.6576,
+      "step": 2174
+    },
+    {
+      "epoch": 0.5012675731735423,
+      "grad_norm": 0.21728312969207764,
+      "learning_rate": 5.223897965537469e-05,
+      "loss": 0.6723,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5014980410232772,
+      "grad_norm": 0.19311760365962982,
+      "learning_rate": 5.220168766968355e-05,
+      "loss": 0.6622,
+      "step": 2176
+    },
+    {
+      "epoch": 0.5017285088730122,
+      "grad_norm": 0.20615801215171814,
+      "learning_rate": 5.216439445682455e-05,
+      "loss": 0.6652,
+      "step": 2177
+    },
+    {
+      "epoch": 0.5019589767227471,
+      "grad_norm": 0.1940842866897583,
+      "learning_rate": 5.212710003758401e-05,
+      "loss": 0.667,
+      "step": 2178
+    },
+    {
+      "epoch": 0.5021894445724822,
+      "grad_norm": 0.19929523766040802,
+      "learning_rate": 5.208980443274899e-05,
+      "loss": 0.6649,
+      "step": 2179
+    },
+    {
+      "epoch": 0.5024199124222171,
+      "grad_norm": 0.181834876537323,
+      "learning_rate": 5.205250766310712e-05,
+      "loss": 0.6654,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5026503802719521,
+      "grad_norm": 0.20930717885494232,
+      "learning_rate": 5.201520974944675e-05,
+      "loss": 0.6627,
+      "step": 2181
+    },
+    {
+      "epoch": 0.502880848121687,
+      "grad_norm": 0.18552522361278534,
+      "learning_rate": 5.1977910712556834e-05,
+      "loss": 0.6738,
+      "step": 2182
+    },
+    {
+      "epoch": 0.503111315971422,
+      "grad_norm": 0.21850310266017914,
+      "learning_rate": 5.1940610573226955e-05,
+      "loss": 0.6776,
+      "step": 2183
+    },
+    {
+      "epoch": 0.5033417838211569,
+      "grad_norm": 0.18333658576011658,
+      "learning_rate": 5.190330935224732e-05,
+      "loss": 0.6657,
+      "step": 2184
+    },
+    {
+      "epoch": 0.5035722516708919,
+      "grad_norm": 0.21493695676326752,
+      "learning_rate": 5.186600707040874e-05,
+      "loss": 0.6724,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5038027195206268,
+      "grad_norm": 0.19420945644378662,
+      "learning_rate": 5.1828703748502614e-05,
+      "loss": 0.6571,
+      "step": 2186
+    },
+    {
+      "epoch": 0.5040331873703618,
+      "grad_norm": 0.19441378116607666,
+      "learning_rate": 5.179139940732091e-05,
+      "loss": 0.6548,
+      "step": 2187
+    },
+    {
+      "epoch": 0.5042636552200968,
+      "grad_norm": 0.201436385512352,
+      "learning_rate": 5.1754094067656164e-05,
+      "loss": 0.6734,
+      "step": 2188
+    },
+    {
+      "epoch": 0.5044941230698318,
+      "grad_norm": 0.222244992852211,
+      "learning_rate": 5.17167877503015e-05,
+      "loss": 0.67,
+      "step": 2189
+    },
+    {
+      "epoch": 0.5047245909195667,
+      "grad_norm": 0.20612792670726776,
+      "learning_rate": 5.1679480476050525e-05,
+      "loss": 0.6591,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5049550587693017,
+      "grad_norm": 0.20908065140247345,
+      "learning_rate": 5.164217226569747e-05,
+      "loss": 0.6708,
+      "step": 2191
+    },
+    {
+      "epoch": 0.5051855266190366,
+      "grad_norm": 0.1922689974308014,
+      "learning_rate": 5.1604863140037e-05,
+      "loss": 0.6586,
+      "step": 2192
+    },
+    {
+      "epoch": 0.5054159944687716,
+      "grad_norm": 0.22337596118450165,
+      "learning_rate": 5.156755311986433e-05,
+      "loss": 0.6675,
+      "step": 2193
+    },
+    {
+      "epoch": 0.5056464623185065,
+      "grad_norm": 0.20729419589042664,
+      "learning_rate": 5.153024222597519e-05,
+      "loss": 0.6646,
+      "step": 2194
+    },
+    {
+      "epoch": 0.5058769301682415,
+      "grad_norm": 0.20641185343265533,
+      "learning_rate": 5.149293047916576e-05,
+      "loss": 0.6679,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5061073980179764,
+      "grad_norm": 0.20790040493011475,
+      "learning_rate": 5.14556179002327e-05,
+      "loss": 0.6635,
+      "step": 2196
+    },
+    {
+      "epoch": 0.5063378658677115,
+      "grad_norm": 0.19893454015254974,
+      "learning_rate": 5.141830450997316e-05,
+      "loss": 0.6687,
+      "step": 2197
+    },
+    {
+      "epoch": 0.5065683337174464,
+      "grad_norm": 0.2278234213590622,
+      "learning_rate": 5.138099032918475e-05,
+      "loss": 0.6625,
+      "step": 2198
+    },
+    {
+      "epoch": 0.5067988015671814,
+      "grad_norm": 0.20414131879806519,
+      "learning_rate": 5.1343675378665455e-05,
+      "loss": 0.6677,
+      "step": 2199
+    },
+    {
+      "epoch": 0.5070292694169164,
+      "grad_norm": 0.20839910209178925,
+      "learning_rate": 5.130635967921377e-05,
+      "loss": 0.6679,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5072597372666513,
+      "grad_norm": 0.1961502730846405,
+      "learning_rate": 5.1269043251628556e-05,
+      "loss": 0.6703,
+      "step": 2201
+    },
+    {
+      "epoch": 0.5074902051163863,
+      "grad_norm": 0.19088955223560333,
+      "learning_rate": 5.123172611670907e-05,
+      "loss": 0.6757,
+      "step": 2202
+    },
+    {
+      "epoch": 0.5077206729661212,
+      "grad_norm": 0.20697516202926636,
+      "learning_rate": 5.119440829525504e-05,
+      "loss": 0.6694,
+      "step": 2203
+    },
+    {
+      "epoch": 0.5079511408158562,
+      "grad_norm": 0.20802126824855804,
+      "learning_rate": 5.115708980806647e-05,
+      "loss": 0.6742,
+      "step": 2204
+    },
+    {
+      "epoch": 0.5081816086655911,
+      "grad_norm": 0.19614768028259277,
+      "learning_rate": 5.111977067594382e-05,
+      "loss": 0.6664,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5084120765153262,
+      "grad_norm": 0.1986846625804901,
+      "learning_rate": 5.1082450919687884e-05,
+      "loss": 0.6523,
+      "step": 2206
+    },
+    {
+      "epoch": 0.5086425443650611,
+      "grad_norm": 0.20315955579280853,
+      "learning_rate": 5.1045130560099776e-05,
+      "loss": 0.6673,
+      "step": 2207
+    },
+    {
+      "epoch": 0.5088730122147961,
+      "grad_norm": 0.20175747573375702,
+      "learning_rate": 5.100780961798098e-05,
+      "loss": 0.6709,
+      "step": 2208
+    },
+    {
+      "epoch": 0.509103480064531,
+      "grad_norm": 0.19836615025997162,
+      "learning_rate": 5.097048811413331e-05,
+      "loss": 0.6712,
+      "step": 2209
+    },
+    {
+      "epoch": 0.509333947914266,
+      "grad_norm": 0.2123645395040512,
+      "learning_rate": 5.093316606935883e-05,
+      "loss": 0.662,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5095644157640009,
+      "grad_norm": 0.19712963700294495,
+      "learning_rate": 5.0895843504460005e-05,
+      "loss": 0.6771,
+      "step": 2211
+    },
+    {
+      "epoch": 0.5097948836137359,
+      "grad_norm": 0.19324639439582825,
+      "learning_rate": 5.08585204402395e-05,
+      "loss": 0.6606,
+      "step": 2212
+    },
+    {
+      "epoch": 0.5100253514634708,
+      "grad_norm": 0.19227439165115356,
+      "learning_rate": 5.08211968975003e-05,
+      "loss": 0.6669,
+      "step": 2213
+    },
+    {
+      "epoch": 0.5102558193132058,
+      "grad_norm": 0.21038465201854706,
+      "learning_rate": 5.078387289704568e-05,
+      "loss": 0.6682,
+      "step": 2214
+    },
+    {
+      "epoch": 0.5104862871629408,
+      "grad_norm": 0.1920362412929535,
+      "learning_rate": 5.074654845967911e-05,
+      "loss": 0.6558,
+      "step": 2215
+    },
+    {
+      "epoch": 0.5107167550126758,
+      "grad_norm": 0.20892934501171112,
+      "learning_rate": 5.0709223606204345e-05,
+      "loss": 0.6653,
+      "step": 2216
+    },
+    {
+      "epoch": 0.5109472228624107,
+      "grad_norm": 0.1770077347755432,
+      "learning_rate": 5.0671898357425366e-05,
+      "loss": 0.6642,
+      "step": 2217
+    },
+    {
+      "epoch": 0.5111776907121457,
+      "grad_norm": 0.1956147402524948,
+      "learning_rate": 5.063457273414638e-05,
+      "loss": 0.6676,
+      "step": 2218
+    },
+    {
+      "epoch": 0.5114081585618806,
+      "grad_norm": 0.1727982610464096,
+      "learning_rate": 5.059724675717177e-05,
+      "loss": 0.6607,
+      "step": 2219
+    },
+    {
+      "epoch": 0.5116386264116156,
+      "grad_norm": 0.18610809743404388,
+      "learning_rate": 5.055992044730615e-05,
+      "loss": 0.6625,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5118690942613505,
+      "grad_norm": 0.1783786416053772,
+      "learning_rate": 5.0522593825354336e-05,
+      "loss": 0.6589,
+      "step": 2221
+    },
+    {
+      "epoch": 0.5120995621110855,
+      "grad_norm": 0.18354859948158264,
+      "learning_rate": 5.048526691212123e-05,
+      "loss": 0.6544,
+      "step": 2222
+    },
+    {
+      "epoch": 0.5123300299608204,
+      "grad_norm": 0.1823883056640625,
+      "learning_rate": 5.044793972841203e-05,
+      "loss": 0.6713,
+      "step": 2223
+    },
+    {
+      "epoch": 0.5125604978105555,
+      "grad_norm": 0.19022877514362335,
+      "learning_rate": 5.041061229503196e-05,
+      "loss": 0.6669,
+      "step": 2224
+    },
+    {
+      "epoch": 0.5127909656602904,
+      "grad_norm": 0.19528958201408386,
+      "learning_rate": 5.037328463278646e-05,
+      "loss": 0.6589,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5130214335100254,
+      "grad_norm": 0.1989632546901703,
+      "learning_rate": 5.033595676248106e-05,
+      "loss": 0.6609,
+      "step": 2226
+    },
+    {
+      "epoch": 0.5132519013597603,
+      "grad_norm": 0.17961591482162476,
+      "learning_rate": 5.029862870492142e-05,
+      "loss": 0.6652,
+      "step": 2227
+    },
+    {
+      "epoch": 0.5134823692094953,
+      "grad_norm": 0.18870680034160614,
+      "learning_rate": 5.026130048091331e-05,
+      "loss": 0.6658,
+      "step": 2228
+    },
+    {
+      "epoch": 0.5137128370592302,
+      "grad_norm": 0.2052677571773529,
+      "learning_rate": 5.0223972111262584e-05,
+      "loss": 0.6663,
+      "step": 2229
+    },
+    {
+      "epoch": 0.5139433049089652,
+      "grad_norm": 0.23451951146125793,
+      "learning_rate": 5.018664361677519e-05,
+      "loss": 0.6642,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5141737727587001,
+      "grad_norm": 0.18537165224552155,
+      "learning_rate": 5.01493150182571e-05,
+      "loss": 0.6698,
+      "step": 2231
+    },
+    {
+      "epoch": 0.5144042406084351,
+      "grad_norm": 0.21171410381793976,
+      "learning_rate": 5.011198633651442e-05,
+      "loss": 0.6687,
+      "step": 2232
+    },
+    {
+      "epoch": 0.51463470845817,
+      "grad_norm": 0.22823838889598846,
+      "learning_rate": 5.0074657592353246e-05,
+      "loss": 0.6657,
+      "step": 2233
+    },
+    {
+      "epoch": 0.5148651763079051,
+      "grad_norm": 0.1847493201494217,
+      "learning_rate": 5.003732880657971e-05,
+      "loss": 0.6638,
+      "step": 2234
+    },
+    {
+      "epoch": 0.51509564415764,
+      "grad_norm": 0.2114555537700653,
+      "learning_rate": 5e-05,
+      "loss": 0.668,
+      "step": 2235
+    },
+    {
+      "epoch": 0.515326112007375,
+      "grad_norm": 0.17518627643585205,
+      "learning_rate": 4.996267119342029e-05,
+      "loss": 0.6551,
+      "step": 2236
+    },
+    {
+      "epoch": 0.5155565798571099,
+      "grad_norm": 0.23371440172195435,
+      "learning_rate": 4.992534240764677e-05,
+      "loss": 0.6697,
+      "step": 2237
+    },
+    {
+      "epoch": 0.5157870477068449,
+      "grad_norm": 0.1804964542388916,
+      "learning_rate": 4.9888013663485586e-05,
+      "loss": 0.6622,
+      "step": 2238
+    },
+    {
+      "epoch": 0.5160175155565798,
+      "grad_norm": 0.22412839531898499,
+      "learning_rate": 4.98506849817429e-05,
+      "loss": 0.6576,
+      "step": 2239
+    },
+    {
+      "epoch": 0.5162479834063148,
+      "grad_norm": 0.2174907922744751,
+      "learning_rate": 4.981335638322484e-05,
+      "loss": 0.6597,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5164784512560497,
+      "grad_norm": 0.20137697458267212,
+      "learning_rate": 4.9776027888737434e-05,
+      "loss": 0.6747,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5167089191057848,
+      "grad_norm": 0.2253003716468811,
+      "learning_rate": 4.973869951908669e-05,
+      "loss": 0.6686,
+      "step": 2242
+    },
+    {
+      "epoch": 0.5169393869555197,
+      "grad_norm": 0.2619769871234894,
+      "learning_rate": 4.9701371295078603e-05,
+      "loss": 0.6616,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5171698548052547,
+      "grad_norm": 0.26056036353111267,
+      "learning_rate": 4.966404323751896e-05,
+      "loss": 0.6653,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5174003226549896,
+      "grad_norm": 0.21332813799381256,
+      "learning_rate": 4.962671536721355e-05,
+      "loss": 0.6609,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5176307905047246,
+      "grad_norm": 0.25862419605255127,
+      "learning_rate": 4.9589387704968054e-05,
+      "loss": 0.6685,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5178612583544595,
+      "grad_norm": 0.24932904541492462,
+      "learning_rate": 4.955206027158798e-05,
+      "loss": 0.6642,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5180917262041945,
+      "grad_norm": 0.21036414802074432,
+      "learning_rate": 4.951473308787876e-05,
+      "loss": 0.6636,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5183221940539294,
+      "grad_norm": 0.27057451009750366,
+      "learning_rate": 4.947740617464568e-05,
+      "loss": 0.6603,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5185526619036644,
+      "grad_norm": 0.2021026611328125,
+      "learning_rate": 4.9440079552693854e-05,
+      "loss": 0.6779,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5187831297533994,
+      "grad_norm": 0.24280990660190582,
+      "learning_rate": 4.940275324282824e-05,
+      "loss": 0.6528,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5190135976031344,
+      "grad_norm": 0.21159270405769348,
+      "learning_rate": 4.9365427265853644e-05,
+      "loss": 0.6706,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5192440654528693,
+      "grad_norm": 0.2115175724029541,
+      "learning_rate": 4.9328101642574646e-05,
+      "loss": 0.6645,
+      "step": 2253
+    },
+    {
+      "epoch": 0.5194745333026043,
+      "grad_norm": 0.21375073492527008,
+      "learning_rate": 4.929077639379566e-05,
+      "loss": 0.6633,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5197050011523392,
+      "grad_norm": 0.2012326568365097,
+      "learning_rate": 4.925345154032092e-05,
+      "loss": 0.658,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5199354690020742,
+      "grad_norm": 0.2035793513059616,
+      "learning_rate": 4.921612710295433e-05,
+      "loss": 0.6658,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5201659368518091,
+      "grad_norm": 0.2022104114294052,
+      "learning_rate": 4.91788031024997e-05,
+      "loss": 0.664,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5203964047015441,
+      "grad_norm": 0.2398633062839508,
+      "learning_rate": 4.9141479559760517e-05,
+      "loss": 0.6608,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5206268725512792,
+      "grad_norm": 0.1719241738319397,
+      "learning_rate": 4.910415649554001e-05,
+      "loss": 0.663,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5208573404010141,
+      "grad_norm": 0.22672203183174133,
+      "learning_rate": 4.906683393064117e-05,
+      "loss": 0.6632,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5210878082507491,
+      "grad_norm": 0.19581197202205658,
+      "learning_rate": 4.9029511885866717e-05,
+      "loss": 0.6602,
+      "step": 2261
+    },
+    {
+      "epoch": 0.521318276100484,
+      "grad_norm": 0.22291046380996704,
+      "learning_rate": 4.899219038201903e-05,
+      "loss": 0.6525,
+      "step": 2262
+    },
+    {
+      "epoch": 0.521548743950219,
+      "grad_norm": 0.19381098449230194,
+      "learning_rate": 4.895486943990023e-05,
+      "loss": 0.6722,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5217792117999539,
+      "grad_norm": 0.234474778175354,
+      "learning_rate": 4.891754908031213e-05,
+      "loss": 0.6639,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5220096796496889,
+      "grad_norm": 0.21784402430057526,
+      "learning_rate": 4.8880229324056184e-05,
+      "loss": 0.6639,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5222401474994238,
+      "grad_norm": 0.20804181694984436,
+      "learning_rate": 4.8842910191933526e-05,
+      "loss": 0.6595,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5224706153491588,
+      "grad_norm": 0.2230507731437683,
+      "learning_rate": 4.880559170474499e-05,
+      "loss": 0.6716,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5227010831988937,
+      "grad_norm": 0.21720674633979797,
+      "learning_rate": 4.876827388329094e-05,
+      "loss": 0.6584,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5229315510486288,
+      "grad_norm": 0.23036789894104004,
+      "learning_rate": 4.873095674837146e-05,
+      "loss": 0.6654,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5231620188983637,
+      "grad_norm": 0.21202722191810608,
+      "learning_rate": 4.869364032078625e-05,
+      "loss": 0.6704,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5233924867480987,
+      "grad_norm": 0.22500722110271454,
+      "learning_rate": 4.8656324621334557e-05,
+      "loss": 0.6703,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5236229545978336,
+      "grad_norm": 0.20429456233978271,
+      "learning_rate": 4.8619009670815265e-05,
+      "loss": 0.6656,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5238534224475686,
+      "grad_norm": 0.23129715025424957,
+      "learning_rate": 4.8581695490026845e-05,
+      "loss": 0.6689,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5240838902973035,
+      "grad_norm": 0.21389861404895782,
+      "learning_rate": 4.854438209976731e-05,
+      "loss": 0.6718,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5243143581470385,
+      "grad_norm": 0.20860935747623444,
+      "learning_rate": 4.850706952083426e-05,
+      "loss": 0.6725,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5245448259967734,
+      "grad_norm": 0.20308835804462433,
+      "learning_rate": 4.846975777402483e-05,
+      "loss": 0.6665,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5247752938465085,
+      "grad_norm": 0.2039281278848648,
+      "learning_rate": 4.843244688013568e-05,
+      "loss": 0.671,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5250057616962434,
+      "grad_norm": 0.21113687753677368,
+      "learning_rate": 4.839513685996301e-05,
+      "loss": 0.6605,
+      "step": 2278
+    },
+    {
+      "epoch": 0.5252362295459784,
+      "grad_norm": 0.1845845878124237,
+      "learning_rate": 4.835782773430255e-05,
+      "loss": 0.656,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5254666973957133,
+      "grad_norm": 0.22030138969421387,
+      "learning_rate": 4.832051952394948e-05,
+      "loss": 0.672,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5256971652454483,
+      "grad_norm": 0.19281761348247528,
+      "learning_rate": 4.8283212249698515e-05,
+      "loss": 0.6572,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5259276330951832,
+      "grad_norm": 0.21558502316474915,
+      "learning_rate": 4.824590593234386e-05,
+      "loss": 0.6674,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5261581009449182,
+      "grad_norm": 0.1933777928352356,
+      "learning_rate": 4.82086005926791e-05,
+      "loss": 0.6634,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5263885687946531,
+      "grad_norm": 0.19798806309700012,
+      "learning_rate": 4.81712962514974e-05,
+      "loss": 0.6658,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5266190366443881,
+      "grad_norm": 0.21283666789531708,
+      "learning_rate": 4.8133992929591265e-05,
+      "loss": 0.6731,
+      "step": 2285
+    },
+    {
+      "epoch": 0.526849504494123,
+      "grad_norm": 0.2087125927209854,
+      "learning_rate": 4.809669064775269e-05,
+      "loss": 0.6658,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5270799723438581,
+      "grad_norm": 0.21951022744178772,
+      "learning_rate": 4.805938942677306e-05,
+      "loss": 0.6581,
+      "step": 2287
+    },
+    {
+      "epoch": 0.527310440193593,
+      "grad_norm": 0.19603340327739716,
+      "learning_rate": 4.802208928744319e-05,
+      "loss": 0.6689,
+      "step": 2288
+    },
+    {
+      "epoch": 0.527540908043328,
+      "grad_norm": 0.20134252309799194,
+      "learning_rate": 4.798479025055327e-05,
+      "loss": 0.6639,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5277713758930629,
+      "grad_norm": 0.20186404883861542,
+      "learning_rate": 4.7947492336892894e-05,
+      "loss": 0.668,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5280018437427979,
+      "grad_norm": 0.1830950379371643,
+      "learning_rate": 4.791019556725104e-05,
+      "loss": 0.6719,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5282323115925328,
+      "grad_norm": 0.20880278944969177,
+      "learning_rate": 4.7872899962416e-05,
+      "loss": 0.6679,
+      "step": 2292
+    },
+    {
+      "epoch": 0.5284627794422678,
+      "grad_norm": 0.19241443276405334,
+      "learning_rate": 4.783560554317546e-05,
+      "loss": 0.6723,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5286932472920027,
+      "grad_norm": 0.2183217704296112,
+      "learning_rate": 4.779831233031647e-05,
+      "loss": 0.6537,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5289237151417377,
+      "grad_norm": 0.20928408205509186,
+      "learning_rate": 4.776102034462532e-05,
+      "loss": 0.6647,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5291541829914727,
+      "grad_norm": 0.18923519551753998,
+      "learning_rate": 4.772372960688768e-05,
+      "loss": 0.6654,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5293846508412077,
+      "grad_norm": 0.20825320482254028,
+      "learning_rate": 4.7686440137888555e-05,
+      "loss": 0.664,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5296151186909426,
+      "grad_norm": 0.21106471121311188,
+      "learning_rate": 4.764915195841214e-05,
+      "loss": 0.6672,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5298455865406776,
+      "grad_norm": 0.2134147584438324,
+      "learning_rate": 4.7611865089242004e-05,
+      "loss": 0.6697,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5300760543904125,
+      "grad_norm": 0.19978366792201996,
+      "learning_rate": 4.757457955116095e-05,
+      "loss": 0.66,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5303065222401475,
+      "grad_norm": 0.21647867560386658,
+      "learning_rate": 4.753729536495104e-05,
+      "loss": 0.6638,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5305369900898824,
+      "grad_norm": 0.1859664022922516,
+      "learning_rate": 4.750001255139358e-05,
+      "loss": 0.658,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5307674579396174,
+      "grad_norm": 0.1935039460659027,
+      "learning_rate": 4.7462731131269114e-05,
+      "loss": 0.6568,
+      "step": 2303
+    },
+    {
+      "epoch": 0.5309979257893523,
+      "grad_norm": 0.21178743243217468,
+      "learning_rate": 4.7425451125357435e-05,
+      "loss": 0.6627,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5312283936390874,
+      "grad_norm": 0.20362286269664764,
+      "learning_rate": 4.738817255443749e-05,
+      "loss": 0.6597,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5314588614888223,
+      "grad_norm": 0.20342570543289185,
+      "learning_rate": 4.735089543928746e-05,
+      "loss": 0.6692,
+      "step": 2306
+    },
+    {
+      "epoch": 0.5316893293385573,
+      "grad_norm": 0.1900341808795929,
+      "learning_rate": 4.731361980068476e-05,
+      "loss": 0.6613,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5319197971882922,
+      "grad_norm": 0.19465115666389465,
+      "learning_rate": 4.727634565940588e-05,
+      "loss": 0.6727,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5321502650380272,
+      "grad_norm": 0.19808262586593628,
+      "learning_rate": 4.723907303622656e-05,
+      "loss": 0.6605,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5323807328877621,
+      "grad_norm": 0.1885158270597458,
+      "learning_rate": 4.7201801951921676e-05,
+      "loss": 0.6574,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5326112007374971,
+      "grad_norm": 0.2195241004228592,
+      "learning_rate": 4.7164532427265204e-05,
+      "loss": 0.6655,
+      "step": 2311
+    },
+    {
+      "epoch": 0.532841668587232,
+      "grad_norm": 0.182617649435997,
+      "learning_rate": 4.712726448303031e-05,
+      "loss": 0.6608,
+      "step": 2312
+    },
+    {
+      "epoch": 0.533072136436967,
+      "grad_norm": 0.19868247210979462,
+      "learning_rate": 4.708999813998924e-05,
+      "loss": 0.656,
+      "step": 2313
+    },
+    {
+      "epoch": 0.533302604286702,
+      "grad_norm": 0.17593586444854736,
+      "learning_rate": 4.7052733418913366e-05,
+      "loss": 0.6577,
+      "step": 2314
+    },
+    {
+      "epoch": 0.533533072136437,
+      "grad_norm": 0.20548640191555023,
+      "learning_rate": 4.701547034057313e-05,
+      "loss": 0.6563,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5337635399861719,
+      "grad_norm": 0.20077823102474213,
+      "learning_rate": 4.697820892573811e-05,
+      "loss": 0.6607,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5339940078359069,
+      "grad_norm": 0.2142491191625595,
+      "learning_rate": 4.694094919517689e-05,
+      "loss": 0.6625,
+      "step": 2317
+    },
+    {
+      "epoch": 0.5342244756856418,
+      "grad_norm": 0.1935112625360489,
+      "learning_rate": 4.6903691169657154e-05,
+      "loss": 0.6687,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5344549435353768,
+      "grad_norm": 0.1958986222743988,
+      "learning_rate": 4.686643486994568e-05,
+      "loss": 0.6676,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5346854113851118,
+      "grad_norm": 0.1859496831893921,
+      "learning_rate": 4.6829180316808165e-05,
+      "loss": 0.6581,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5349158792348467,
+      "grad_norm": 0.21781980991363525,
+      "learning_rate": 4.6791927531009436e-05,
+      "loss": 0.6678,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5351463470845818,
+      "grad_norm": 0.19916626811027527,
+      "learning_rate": 4.675467653331333e-05,
+      "loss": 0.6597,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5353768149343167,
+      "grad_norm": 0.21010006964206696,
+      "learning_rate": 4.67174273444826e-05,
+      "loss": 0.6651,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5356072827840517,
+      "grad_norm": 0.19810812175273895,
+      "learning_rate": 4.668017998527909e-05,
+      "loss": 0.6601,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5358377506337866,
+      "grad_norm": 0.21458233892917633,
+      "learning_rate": 4.664293447646358e-05,
+      "loss": 0.6638,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5360682184835216,
+      "grad_norm": 0.1828889101743698,
+      "learning_rate": 4.660569083879581e-05,
+      "loss": 0.6631,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5362986863332565,
+      "grad_norm": 0.21997065842151642,
+      "learning_rate": 4.656844909303449e-05,
+      "loss": 0.6602,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5365291541829915,
+      "grad_norm": 0.18035876750946045,
+      "learning_rate": 4.653120925993729e-05,
+      "loss": 0.6686,
+      "step": 2328
+    },
+    {
+      "epoch": 0.5367596220327264,
+      "grad_norm": 0.1994839310646057,
+      "learning_rate": 4.649397136026079e-05,
+      "loss": 0.66,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5369900898824614,
+      "grad_norm": 0.1981378048658371,
+      "learning_rate": 4.645673541476049e-05,
+      "loss": 0.6705,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5372205577321963,
+      "grad_norm": 0.1892865151166916,
+      "learning_rate": 4.641950144419085e-05,
+      "loss": 0.6569,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5374510255819314,
+      "grad_norm": 0.19472159445285797,
+      "learning_rate": 4.6382269469305143e-05,
+      "loss": 0.6662,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5376814934316663,
+      "grad_norm": 0.20402705669403076,
+      "learning_rate": 4.634503951085559e-05,
+      "loss": 0.6651,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5379119612814013,
+      "grad_norm": 0.19563409686088562,
+      "learning_rate": 4.630781158959332e-05,
+      "loss": 0.6596,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5381424291311362,
+      "grad_norm": 0.20649971067905426,
+      "learning_rate": 4.627058572626823e-05,
+      "loss": 0.6567,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5383728969808712,
+      "grad_norm": 0.178875133395195,
+      "learning_rate": 4.6233361941629136e-05,
+      "loss": 0.672,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5386033648306061,
+      "grad_norm": 0.20355398952960968,
+      "learning_rate": 4.619614025642371e-05,
+      "loss": 0.6639,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5388338326803411,
+      "grad_norm": 0.19655583798885345,
+      "learning_rate": 4.615892069139839e-05,
+      "loss": 0.6685,
+      "step": 2338
+    },
+    {
+      "epoch": 0.539064300530076,
+      "grad_norm": 0.1810363531112671,
+      "learning_rate": 4.612170326729849e-05,
+      "loss": 0.6592,
+      "step": 2339
+    },
+    {
+      "epoch": 0.539294768379811,
+      "grad_norm": 0.19939152896404266,
+      "learning_rate": 4.6084488004868105e-05,
+      "loss": 0.6565,
+      "step": 2340
+    },
+    {
+      "epoch": 0.539525236229546,
+      "grad_norm": 0.18242943286895752,
+      "learning_rate": 4.604727492485011e-05,
+      "loss": 0.6669,
+      "step": 2341
+    },
+    {
+      "epoch": 0.539755704079281,
+      "grad_norm": 0.20084112882614136,
+      "learning_rate": 4.601006404798621e-05,
+      "loss": 0.6603,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5399861719290159,
+      "grad_norm": 0.18553341925144196,
+      "learning_rate": 4.597285539501684e-05,
+      "loss": 0.655,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5402166397787509,
+      "grad_norm": 0.18034282326698303,
+      "learning_rate": 4.5935648986681215e-05,
+      "loss": 0.6612,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5404471076284858,
+      "grad_norm": 0.20821136236190796,
+      "learning_rate": 4.5898444843717275e-05,
+      "loss": 0.6679,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5406775754782208,
+      "grad_norm": 0.18303748965263367,
+      "learning_rate": 4.586124298686177e-05,
+      "loss": 0.6627,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5409080433279557,
+      "grad_norm": 0.2008381336927414,
+      "learning_rate": 4.582404343685005e-05,
+      "loss": 0.6603,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5411385111776907,
+      "grad_norm": 0.18936984241008759,
+      "learning_rate": 4.578684621441629e-05,
+      "loss": 0.6602,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5413689790274256,
+      "grad_norm": 0.22308258712291718,
+      "learning_rate": 4.574965134029335e-05,
+      "loss": 0.6597,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5415994468771607,
+      "grad_norm": 0.1877029538154602,
+      "learning_rate": 4.5712458835212716e-05,
+      "loss": 0.6599,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5418299147268956,
+      "grad_norm": 0.21585100889205933,
+      "learning_rate": 4.5675268719904624e-05,
+      "loss": 0.6659,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5420603825766306,
+      "grad_norm": 0.20248469710350037,
+      "learning_rate": 4.5638081015097956e-05,
+      "loss": 0.6575,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5422908504263655,
+      "grad_norm": 0.19426268339157104,
+      "learning_rate": 4.560089574152021e-05,
+      "loss": 0.6667,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5425213182761005,
+      "grad_norm": 0.22340944409370422,
+      "learning_rate": 4.5563712919897606e-05,
+      "loss": 0.6619,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5427517861258354,
+      "grad_norm": 0.1823105663061142,
+      "learning_rate": 4.552653257095495e-05,
+      "loss": 0.6642,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5429822539755704,
+      "grad_norm": 0.23891690373420715,
+      "learning_rate": 4.548935471541565e-05,
+      "loss": 0.6654,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5432127218253053,
+      "grad_norm": 0.21343962848186493,
+      "learning_rate": 4.545217937400177e-05,
+      "loss": 0.6641,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5434431896750404,
+      "grad_norm": 0.2255440205335617,
+      "learning_rate": 4.541500656743396e-05,
+      "loss": 0.6682,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5436736575247753,
+      "grad_norm": 0.24012601375579834,
+      "learning_rate": 4.537783631643143e-05,
+      "loss": 0.6729,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5439041253745103,
+      "grad_norm": 0.18320710957050323,
+      "learning_rate": 4.534066864171198e-05,
+      "loss": 0.6559,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5441345932242452,
+      "grad_norm": 0.22548629343509674,
+      "learning_rate": 4.530350356399203e-05,
+      "loss": 0.6601,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5443650610739802,
+      "grad_norm": 0.20092882215976715,
+      "learning_rate": 4.5266341103986436e-05,
+      "loss": 0.661,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5445955289237151,
+      "grad_norm": 0.18173551559448242,
+      "learning_rate": 4.5229181282408705e-05,
+      "loss": 0.6608,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5448259967734501,
+      "grad_norm": 0.19635426998138428,
+      "learning_rate": 4.519202411997083e-05,
+      "loss": 0.6637,
+      "step": 2364
+    },
+    {
+      "epoch": 0.545056464623185,
+      "grad_norm": 0.1842816174030304,
+      "learning_rate": 4.515486963738329e-05,
+      "loss": 0.664,
+      "step": 2365
+    },
+    {
+      "epoch": 0.54528693247292,
+      "grad_norm": 0.19681525230407715,
+      "learning_rate": 4.511771785535513e-05,
+      "loss": 0.6692,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5455174003226549,
+      "grad_norm": 0.21367399394512177,
+      "learning_rate": 4.5080568794593865e-05,
+      "loss": 0.6541,
+      "step": 2367
+    },
+    {
+      "epoch": 0.54574786817239,
+      "grad_norm": 0.19998280704021454,
+      "learning_rate": 4.504342247580546e-05,
+      "loss": 0.6609,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5459783360221249,
+      "grad_norm": 0.20067118108272552,
+      "learning_rate": 4.50062789196944e-05,
+      "loss": 0.6614,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5462088038718599,
+      "grad_norm": 0.19367091357707977,
+      "learning_rate": 4.4969138146963625e-05,
+      "loss": 0.663,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5464392717215948,
+      "grad_norm": 0.19135761260986328,
+      "learning_rate": 4.493200017831448e-05,
+      "loss": 0.6648,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5466697395713298,
+      "grad_norm": 0.21066224575042725,
+      "learning_rate": 4.4894865034446784e-05,
+      "loss": 0.6591,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5469002074210647,
+      "grad_norm": 0.19303461909294128,
+      "learning_rate": 4.4857732736058814e-05,
+      "loss": 0.6621,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5471306752707997,
+      "grad_norm": 0.2217341661453247,
+      "learning_rate": 4.482060330384716e-05,
+      "loss": 0.6596,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5473611431205346,
+      "grad_norm": 0.20083634555339813,
+      "learning_rate": 4.478347675850693e-05,
+      "loss": 0.6644,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5475916109702696,
+      "grad_norm": 0.22229771316051483,
+      "learning_rate": 4.474635312073155e-05,
+      "loss": 0.6623,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5478220788200046,
+      "grad_norm": 0.20578941702842712,
+      "learning_rate": 4.470923241121283e-05,
+      "loss": 0.654,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5480525466697396,
+      "grad_norm": 0.21209031343460083,
+      "learning_rate": 4.467211465064097e-05,
+      "loss": 0.6662,
+      "step": 2378
+    },
+    {
+      "epoch": 0.5482830145194746,
+      "grad_norm": 0.2017475962638855,
+      "learning_rate": 4.4634999859704546e-05,
+      "loss": 0.6629,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5485134823692095,
+      "grad_norm": 0.19928203523159027,
+      "learning_rate": 4.459788805909041e-05,
+      "loss": 0.6593,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5487439502189445,
+      "grad_norm": 0.19713225960731506,
+      "learning_rate": 4.456077926948381e-05,
+      "loss": 0.6611,
+      "step": 2381
+    },
+    {
+      "epoch": 0.5489744180686794,
+      "grad_norm": 0.198060542345047,
+      "learning_rate": 4.45236735115683e-05,
+      "loss": 0.6569,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5492048859184144,
+      "grad_norm": 0.19292356073856354,
+      "learning_rate": 4.448657080602573e-05,
+      "loss": 0.6666,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5494353537681493,
+      "grad_norm": 0.18970555067062378,
+      "learning_rate": 4.444947117353625e-05,
+      "loss": 0.6612,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5496658216178844,
+      "grad_norm": 0.19588854908943176,
+      "learning_rate": 4.441237463477833e-05,
+      "loss": 0.6697,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5498962894676193,
+      "grad_norm": 0.20147165656089783,
+      "learning_rate": 4.437528121042866e-05,
+      "loss": 0.6583,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5501267573173543,
+      "grad_norm": 0.21111683547496796,
+      "learning_rate": 4.433819092116223e-05,
+      "loss": 0.6589,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5503572251670892,
+      "grad_norm": 0.17622540891170502,
+      "learning_rate": 4.430110378765232e-05,
+      "loss": 0.6652,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5505876930168242,
+      "grad_norm": 0.22303026914596558,
+      "learning_rate": 4.4264019830570334e-05,
+      "loss": 0.6743,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5508181608665591,
+      "grad_norm": 0.19800515472888947,
+      "learning_rate": 4.4226939070586035e-05,
+      "loss": 0.6707,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5510486287162941,
+      "grad_norm": 0.19506274163722992,
+      "learning_rate": 4.418986152836734e-05,
+      "loss": 0.6626,
+      "step": 2391
+    },
+    {
+      "epoch": 0.551279096566029,
+      "grad_norm": 0.197477325797081,
+      "learning_rate": 4.4152787224580364e-05,
+      "loss": 0.6636,
+      "step": 2392
+    },
+    {
+      "epoch": 0.551509564415764,
+      "grad_norm": 0.2039872705936432,
+      "learning_rate": 4.411571617988945e-05,
+      "loss": 0.6642,
+      "step": 2393
+    },
+    {
+      "epoch": 0.551740032265499,
+      "grad_norm": 0.19217687845230103,
+      "learning_rate": 4.407864841495711e-05,
+      "loss": 0.6634,
+      "step": 2394
+    },
+    {
+      "epoch": 0.551970500115234,
+      "grad_norm": 0.1925874948501587,
+      "learning_rate": 4.4041583950444004e-05,
+      "loss": 0.6565,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5522009679649689,
+      "grad_norm": 0.20073387026786804,
+      "learning_rate": 4.400452280700899e-05,
+      "loss": 0.6565,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5524314358147039,
+      "grad_norm": 0.19481593370437622,
+      "learning_rate": 4.3967465005309073e-05,
+      "loss": 0.6638,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5526619036644388,
+      "grad_norm": 0.18334874510765076,
+      "learning_rate": 4.393041056599936e-05,
+      "loss": 0.6625,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5528923715141738,
+      "grad_norm": 0.187678724527359,
+      "learning_rate": 4.38933595097331e-05,
+      "loss": 0.6575,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5531228393639087,
+      "grad_norm": 0.20493364334106445,
+      "learning_rate": 4.3856311857161705e-05,
+      "loss": 0.6672,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5533533072136437,
+      "grad_norm": 0.1791137009859085,
+      "learning_rate": 4.381926762893459e-05,
+      "loss": 0.6603,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5535837750633786,
+      "grad_norm": 0.19737230241298676,
+      "learning_rate": 4.378222684569937e-05,
+      "loss": 0.6646,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5538142429131137,
+      "grad_norm": 0.22899094223976135,
+      "learning_rate": 4.374518952810167e-05,
+      "loss": 0.66,
+      "step": 2403
+    },
+    {
+      "epoch": 0.5540447107628486,
+      "grad_norm": 0.18935391306877136,
+      "learning_rate": 4.3708155696785186e-05,
+      "loss": 0.6459,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5542751786125836,
+      "grad_norm": 0.17492151260375977,
+      "learning_rate": 4.3671125372391704e-05,
+      "loss": 0.6542,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5545056464623185,
+      "grad_norm": 0.21164800226688385,
+      "learning_rate": 4.363409857556104e-05,
+      "loss": 0.6638,
+      "step": 2406
+    },
+    {
+      "epoch": 0.5547361143120535,
+      "grad_norm": 0.2046276330947876,
+      "learning_rate": 4.359707532693102e-05,
+      "loss": 0.6591,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5549665821617884,
+      "grad_norm": 0.20502637326717377,
+      "learning_rate": 4.3560055647137536e-05,
+      "loss": 0.6636,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5551970500115234,
+      "grad_norm": 0.19672958552837372,
+      "learning_rate": 4.352303955681447e-05,
+      "loss": 0.6641,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5554275178612583,
+      "grad_norm": 0.2050708532333374,
+      "learning_rate": 4.348602707659369e-05,
+      "loss": 0.6539,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5556579857109933,
+      "grad_norm": 0.20396442711353302,
+      "learning_rate": 4.344901822710506e-05,
+      "loss": 0.6589,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5558884535607282,
+      "grad_norm": 0.20161718130111694,
+      "learning_rate": 4.341201302897645e-05,
+      "loss": 0.6672,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5561189214104633,
+      "grad_norm": 0.19954702258110046,
+      "learning_rate": 4.337501150283364e-05,
+      "loss": 0.6578,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5563493892601982,
+      "grad_norm": 0.18954674899578094,
+      "learning_rate": 4.3338013669300396e-05,
+      "loss": 0.6642,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5565798571099332,
+      "grad_norm": 0.2116153985261917,
+      "learning_rate": 4.3301019548998465e-05,
+      "loss": 0.6569,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5568103249596681,
+      "grad_norm": 0.2053348273038864,
+      "learning_rate": 4.326402916254741e-05,
+      "loss": 0.6547,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5570407928094031,
+      "grad_norm": 0.21485279500484467,
+      "learning_rate": 4.322704253056483e-05,
+      "loss": 0.6657,
+      "step": 2417
+    },
+    {
+      "epoch": 0.557271260659138,
+      "grad_norm": 0.21614287793636322,
+      "learning_rate": 4.31900596736662e-05,
+      "loss": 0.664,
+      "step": 2418
+    },
+    {
+      "epoch": 0.557501728508873,
+      "grad_norm": 0.18734347820281982,
+      "learning_rate": 4.3153080612464835e-05,
+      "loss": 0.6608,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5577321963586079,
+      "grad_norm": 0.2166302502155304,
+      "learning_rate": 4.3116105367572e-05,
+      "loss": 0.666,
+      "step": 2420
+    },
+    {
+      "epoch": 0.557962664208343,
+      "grad_norm": 0.18908052146434784,
+      "learning_rate": 4.3079133959596825e-05,
+      "loss": 0.6561,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5581931320580779,
+      "grad_norm": 0.21493315696716309,
+      "learning_rate": 4.304216640914625e-05,
+      "loss": 0.658,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5584235999078129,
+      "grad_norm": 0.17654471099376678,
+      "learning_rate": 4.300520273682511e-05,
+      "loss": 0.6602,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5586540677575478,
+      "grad_norm": 0.22438247501850128,
+      "learning_rate": 4.2968242963236084e-05,
+      "loss": 0.6561,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5588845356072828,
+      "grad_norm": 0.19954872131347656,
+      "learning_rate": 4.293128710897965e-05,
+      "loss": 0.6625,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5591150034570177,
+      "grad_norm": 0.2289677858352661,
+      "learning_rate": 4.28943351946541e-05,
+      "loss": 0.6567,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5593454713067527,
+      "grad_norm": 0.19166941940784454,
+      "learning_rate": 4.28573872408556e-05,
+      "loss": 0.6689,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5595759391564876,
+      "grad_norm": 0.21529777348041534,
+      "learning_rate": 4.282044326817798e-05,
+      "loss": 0.6652,
+      "step": 2428
+    },
+    {
+      "epoch": 0.5598064070062226,
+      "grad_norm": 0.2114769071340561,
+      "learning_rate": 4.2783503297212975e-05,
+      "loss": 0.6591,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5600368748559575,
+      "grad_norm": 0.19013917446136475,
+      "learning_rate": 4.2746567348550024e-05,
+      "loss": 0.6567,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5602673427056926,
+      "grad_norm": 0.2237829715013504,
+      "learning_rate": 4.2709635442776333e-05,
+      "loss": 0.6673,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5604978105554275,
+      "grad_norm": 0.18001073598861694,
+      "learning_rate": 4.267270760047685e-05,
+      "loss": 0.663,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5607282784051625,
+      "grad_norm": 0.21743640303611755,
+      "learning_rate": 4.263578384223429e-05,
+      "loss": 0.6747,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5609587462548974,
+      "grad_norm": 0.1813591569662094,
+      "learning_rate": 4.2598864188629065e-05,
+      "loss": 0.6593,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5611892141046324,
+      "grad_norm": 0.19372020661830902,
+      "learning_rate": 4.256194866023929e-05,
+      "loss": 0.6569,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5614196819543673,
+      "grad_norm": 0.17591120302677155,
+      "learning_rate": 4.25250372776408e-05,
+      "loss": 0.6594,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5616501498041023,
+      "grad_norm": 0.19922824203968048,
+      "learning_rate": 4.248813006140714e-05,
+      "loss": 0.6646,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5618806176538373,
+      "grad_norm": 0.1966255158185959,
+      "learning_rate": 4.245122703210945e-05,
+      "loss": 0.6507,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5621110855035722,
+      "grad_norm": 0.18422779440879822,
+      "learning_rate": 4.241432821031664e-05,
+      "loss": 0.6578,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5623415533533073,
+      "grad_norm": 0.1995001882314682,
+      "learning_rate": 4.2377433616595244e-05,
+      "loss": 0.6682,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5625720212030422,
+      "grad_norm": 0.1927601844072342,
+      "learning_rate": 4.234054327150935e-05,
+      "loss": 0.6568,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5628024890527772,
+      "grad_norm": 0.20397785305976868,
+      "learning_rate": 4.2303657195620804e-05,
+      "loss": 0.6562,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5630329569025121,
+      "grad_norm": 0.18410134315490723,
+      "learning_rate": 4.2266775409489023e-05,
+      "loss": 0.6654,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5632634247522471,
+      "grad_norm": 0.18105067312717438,
+      "learning_rate": 4.2229897933671006e-05,
+      "loss": 0.6571,
+      "step": 2444
+    },
+    {
+      "epoch": 0.563493892601982,
+      "grad_norm": 0.16961467266082764,
+      "learning_rate": 4.219302478872138e-05,
+      "loss": 0.6645,
+      "step": 2445
+    },
+    {
+      "epoch": 0.563724360451717,
+      "grad_norm": 0.19589442014694214,
+      "learning_rate": 4.2156155995192364e-05,
+      "loss": 0.6596,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5639548283014519,
+      "grad_norm": 0.17968355119228363,
+      "learning_rate": 4.211929157363372e-05,
+      "loss": 0.6652,
+      "step": 2447
+    },
+    {
+      "epoch": 0.564185296151187,
+      "grad_norm": 0.19263651967048645,
+      "learning_rate": 4.208243154459279e-05,
+      "loss": 0.6628,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5644157640009219,
+      "grad_norm": 0.22454223036766052,
+      "learning_rate": 4.20455759286145e-05,
+      "loss": 0.6667,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5646462318506569,
+      "grad_norm": 0.18612834811210632,
+      "learning_rate": 4.200872474624125e-05,
+      "loss": 0.6652,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5648766997003918,
+      "grad_norm": 0.1997823268175125,
+      "learning_rate": 4.197187801801301e-05,
+      "loss": 0.6521,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5651071675501268,
+      "grad_norm": 0.1961279809474945,
+      "learning_rate": 4.1935035764467306e-05,
+      "loss": 0.6687,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5653376353998617,
+      "grad_norm": 0.20370718836784363,
+      "learning_rate": 4.189819800613906e-05,
+      "loss": 0.6631,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5655681032495967,
+      "grad_norm": 0.2019963413476944,
+      "learning_rate": 4.1861364763560785e-05,
+      "loss": 0.6546,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5657985710993316,
+      "grad_norm": 0.2010907530784607,
+      "learning_rate": 4.182453605726247e-05,
+      "loss": 0.6675,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5660290389490666,
+      "grad_norm": 0.19456058740615845,
+      "learning_rate": 4.178771190777151e-05,
+      "loss": 0.6623,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5662595067988015,
+      "grad_norm": 0.19006440043449402,
+      "learning_rate": 4.175089233561282e-05,
+      "loss": 0.6699,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5664899746485366,
+      "grad_norm": 0.19521141052246094,
+      "learning_rate": 4.171407736130876e-05,
+      "loss": 0.6509,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5667204424982715,
+      "grad_norm": 0.1771191656589508,
+      "learning_rate": 4.167726700537909e-05,
+      "loss": 0.6619,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5669509103480065,
+      "grad_norm": 0.21703976392745972,
+      "learning_rate": 4.164046128834103e-05,
+      "loss": 0.6601,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5671813781977414,
+      "grad_norm": 0.18353256583213806,
+      "learning_rate": 4.160366023070921e-05,
+      "loss": 0.6564,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5674118460474764,
+      "grad_norm": 0.20448799431324005,
+      "learning_rate": 4.1566863852995635e-05,
+      "loss": 0.6552,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5676423138972113,
+      "grad_norm": 0.18381868302822113,
+      "learning_rate": 4.153007217570973e-05,
+      "loss": 0.6629,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5678727817469463,
+      "grad_norm": 0.20256194472312927,
+      "learning_rate": 4.149328521935832e-05,
+      "loss": 0.662,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5681032495966812,
+      "grad_norm": 0.19661514461040497,
+      "learning_rate": 4.145650300444553e-05,
+      "loss": 0.6552,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5683337174464163,
+      "grad_norm": 0.19295766949653625,
+      "learning_rate": 4.141972555147291e-05,
+      "loss": 0.6603,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5685641852961512,
+      "grad_norm": 0.19877560436725616,
+      "learning_rate": 4.1382952880939346e-05,
+      "loss": 0.6532,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5687946531458862,
+      "grad_norm": 0.20013585686683655,
+      "learning_rate": 4.1346185013340985e-05,
+      "loss": 0.6658,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5690251209956211,
+      "grad_norm": 0.1775607168674469,
+      "learning_rate": 4.13094219691714e-05,
+      "loss": 0.6679,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5692555888453561,
+      "grad_norm": 0.1954258233308792,
+      "learning_rate": 4.1272663768921426e-05,
+      "loss": 0.6593,
+      "step": 2470
+    },
+    {
+      "epoch": 0.569486056695091,
+      "grad_norm": 0.18076026439666748,
+      "learning_rate": 4.123591043307918e-05,
+      "loss": 0.66,
+      "step": 2471
+    },
+    {
+      "epoch": 0.569716524544826,
+      "grad_norm": 0.18170256912708282,
+      "learning_rate": 4.1199161982130105e-05,
+      "loss": 0.6651,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5699469923945609,
+      "grad_norm": 0.19181379675865173,
+      "learning_rate": 4.116241843655692e-05,
+      "loss": 0.6556,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5701774602442959,
+      "grad_norm": 0.1833178848028183,
+      "learning_rate": 4.1125679816839564e-05,
+      "loss": 0.6556,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5704079280940308,
+      "grad_norm": 0.19886410236358643,
+      "learning_rate": 4.108894614345527e-05,
+      "loss": 0.6608,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5706383959437659,
+      "grad_norm": 0.18102464079856873,
+      "learning_rate": 4.105221743687853e-05,
+      "loss": 0.6604,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5708688637935008,
+      "grad_norm": 0.18585921823978424,
+      "learning_rate": 4.101549371758101e-05,
+      "loss": 0.6546,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5710993316432358,
+      "grad_norm": 0.18972377479076385,
+      "learning_rate": 4.097877500603164e-05,
+      "loss": 0.6664,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5713297994929707,
+      "grad_norm": 0.18075346946716309,
+      "learning_rate": 4.0942061322696584e-05,
+      "loss": 0.6573,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5715602673427057,
+      "grad_norm": 0.185418963432312,
+      "learning_rate": 4.09053526880391e-05,
+      "loss": 0.6631,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5717907351924406,
+      "grad_norm": 0.1810043305158615,
+      "learning_rate": 4.0868649122519756e-05,
+      "loss": 0.6557,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5720212030421756,
+      "grad_norm": 0.1840018481016159,
+      "learning_rate": 4.083195064659622e-05,
+      "loss": 0.6653,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5722516708919105,
+      "grad_norm": 0.19294288754463196,
+      "learning_rate": 4.079525728072334e-05,
+      "loss": 0.6607,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5724821387416456,
+      "grad_norm": 0.1916273981332779,
+      "learning_rate": 4.0758569045353114e-05,
+      "loss": 0.6646,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5727126065913805,
+      "grad_norm": 0.18346838653087616,
+      "learning_rate": 4.07218859609347e-05,
+      "loss": 0.6529,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5729430744411155,
+      "grad_norm": 0.19156713783740997,
+      "learning_rate": 4.0685208047914346e-05,
+      "loss": 0.6647,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5731735422908504,
+      "grad_norm": 0.18695339560508728,
+      "learning_rate": 4.064853532673546e-05,
+      "loss": 0.6557,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5734040101405854,
+      "grad_norm": 0.20183449983596802,
+      "learning_rate": 4.061186781783853e-05,
+      "loss": 0.6653,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5736344779903203,
+      "grad_norm": 0.19282028079032898,
+      "learning_rate": 4.0575205541661135e-05,
+      "loss": 0.6603,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5738649458400553,
+      "grad_norm": 0.20781488716602325,
+      "learning_rate": 4.0538548518637944e-05,
+      "loss": 0.6528,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5740954136897902,
+      "grad_norm": 0.21310502290725708,
+      "learning_rate": 4.050189676920075e-05,
+      "loss": 0.6605,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5743258815395252,
+      "grad_norm": 0.2098270207643509,
+      "learning_rate": 4.046525031377829e-05,
+      "loss": 0.6566,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5745563493892601,
+      "grad_norm": 0.19469839334487915,
+      "learning_rate": 4.042860917279647e-05,
+      "loss": 0.6613,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5747868172389952,
+      "grad_norm": 0.20969228446483612,
+      "learning_rate": 4.039197336667816e-05,
+      "loss": 0.661,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5750172850887301,
+      "grad_norm": 0.20014849305152893,
+      "learning_rate": 4.035534291584328e-05,
+      "loss": 0.6572,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5752477529384651,
+      "grad_norm": 0.20189215242862701,
+      "learning_rate": 4.0318717840708766e-05,
+      "loss": 0.6551,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5754782207882001,
+      "grad_norm": 0.20174747705459595,
+      "learning_rate": 4.028209816168857e-05,
+      "loss": 0.663,
+      "step": 2497
+    },
+    {
+      "epoch": 0.575708688637935,
+      "grad_norm": 0.18536518514156342,
+      "learning_rate": 4.0245483899193595e-05,
+      "loss": 0.6579,
+      "step": 2498
+    },
+    {
+      "epoch": 0.57593915648767,
+      "grad_norm": 0.2104630321264267,
+      "learning_rate": 4.0208875073631767e-05,
+      "loss": 0.6583,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5761696243374049,
+      "grad_norm": 0.20068304240703583,
+      "learning_rate": 4.017227170540797e-05,
+      "loss": 0.6642,
+      "step": 2500
+    },
+    {
+      "epoch": 0.57640009218714,
+      "grad_norm": 0.2080548256635666,
+      "learning_rate": 4.013567381492404e-05,
+      "loss": 0.6574,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5766305600368749,
+      "grad_norm": 0.1887027621269226,
+      "learning_rate": 4.009908142257875e-05,
+      "loss": 0.6603,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5768610278866099,
+      "grad_norm": 0.2056475281715393,
+      "learning_rate": 4.006249454876785e-05,
+      "loss": 0.6586,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5770914957363448,
+      "grad_norm": 0.19129523634910583,
+      "learning_rate": 4.002591321388395e-05,
+      "loss": 0.6669,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5773219635860798,
+      "grad_norm": 0.20545504987239838,
+      "learning_rate": 3.998933743831663e-05,
+      "loss": 0.6603,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5775524314358147,
+      "grad_norm": 0.1966099590063095,
+      "learning_rate": 3.9952767242452364e-05,
+      "loss": 0.6575,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5777828992855497,
+      "grad_norm": 0.20709416270256042,
+      "learning_rate": 3.9916202646674454e-05,
+      "loss": 0.6619,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5780133671352846,
+      "grad_norm": 0.20298060774803162,
+      "learning_rate": 3.987964367136317e-05,
+      "loss": 0.6602,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5782438349850196,
+      "grad_norm": 0.19368208944797516,
+      "learning_rate": 3.98430903368956e-05,
+      "loss": 0.6637,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5784743028347545,
+      "grad_norm": 0.21980130672454834,
+      "learning_rate": 3.980654266364569e-05,
+      "loss": 0.6596,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5787047706844896,
+      "grad_norm": 0.20827548205852509,
+      "learning_rate": 3.977000067198422e-05,
+      "loss": 0.6553,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5789352385342245,
+      "grad_norm": 0.2053709775209427,
+      "learning_rate": 3.9733464382278854e-05,
+      "loss": 0.6532,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5791657063839595,
+      "grad_norm": 0.2115817666053772,
+      "learning_rate": 3.969693381489401e-05,
+      "loss": 0.6573,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5793961742336944,
+      "grad_norm": 0.20240476727485657,
+      "learning_rate": 3.966040899019096e-05,
+      "loss": 0.6572,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5796266420834294,
+      "grad_norm": 0.20553265511989594,
+      "learning_rate": 3.962388992852778e-05,
+      "loss": 0.6593,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5798571099331643,
+      "grad_norm": 0.20434992015361786,
+      "learning_rate": 3.95873766502593e-05,
+      "loss": 0.657,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5800875777828993,
+      "grad_norm": 0.19303986430168152,
+      "learning_rate": 3.955086917573714e-05,
+      "loss": 0.6632,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5803180456326342,
+      "grad_norm": 0.18680398166179657,
+      "learning_rate": 3.951436752530973e-05,
+      "loss": 0.6624,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5805485134823692,
+      "grad_norm": 0.19972355663776398,
+      "learning_rate": 3.947787171932215e-05,
+      "loss": 0.656,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5807789813321041,
+      "grad_norm": 0.18271125853061676,
+      "learning_rate": 3.944138177811633e-05,
+      "loss": 0.6557,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5810094491818392,
+      "grad_norm": 0.21198797225952148,
+      "learning_rate": 3.9404897722030886e-05,
+      "loss": 0.6563,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5812399170315741,
+      "grad_norm": 0.18243679404258728,
+      "learning_rate": 3.936841957140115e-05,
+      "loss": 0.6591,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5814703848813091,
+      "grad_norm": 0.21174080669879913,
+      "learning_rate": 3.933194734655916e-05,
+      "loss": 0.6634,
+      "step": 2523
+    },
+    {
+      "epoch": 0.581700852731044,
+      "grad_norm": 0.19727849960327148,
+      "learning_rate": 3.929548106783367e-05,
+      "loss": 0.6504,
+      "step": 2524
+    },
+    {
+      "epoch": 0.581931320580779,
+      "grad_norm": 0.19438689947128296,
+      "learning_rate": 3.925902075555009e-05,
+      "loss": 0.6678,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5821617884305139,
+      "grad_norm": 0.20954324305057526,
+      "learning_rate": 3.9222566430030536e-05,
+      "loss": 0.6592,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5823922562802489,
+      "grad_norm": 0.1915033459663391,
+      "learning_rate": 3.918611811159379e-05,
+      "loss": 0.6503,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5826227241299838,
+      "grad_norm": 0.21163514256477356,
+      "learning_rate": 3.914967582055524e-05,
+      "loss": 0.6506,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5828531919797189,
+      "grad_norm": 0.19895008206367493,
+      "learning_rate": 3.911323957722694e-05,
+      "loss": 0.659,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5830836598294538,
+      "grad_norm": 0.1981336772441864,
+      "learning_rate": 3.907680940191764e-05,
+      "loss": 0.6543,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5833141276791888,
+      "grad_norm": 0.19202502071857452,
+      "learning_rate": 3.904038531493257e-05,
+      "loss": 0.6523,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5835445955289237,
+      "grad_norm": 0.2164509892463684,
+      "learning_rate": 3.900396733657366e-05,
+      "loss": 0.6538,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5837750633786587,
+      "grad_norm": 0.18908166885375977,
+      "learning_rate": 3.896755548713946e-05,
+      "loss": 0.6575,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5840055312283936,
+      "grad_norm": 0.21118958294391632,
+      "learning_rate": 3.8931149786925e-05,
+      "loss": 0.6605,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5842359990781286,
+      "grad_norm": 0.1941760927438736,
+      "learning_rate": 3.889475025622199e-05,
+      "loss": 0.6668,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5844664669278635,
+      "grad_norm": 0.20846354961395264,
+      "learning_rate": 3.885835691531865e-05,
+      "loss": 0.6565,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5846969347775985,
+      "grad_norm": 0.19090569019317627,
+      "learning_rate": 3.882196978449972e-05,
+      "loss": 0.6618,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5849274026273334,
+      "grad_norm": 0.2103571891784668,
+      "learning_rate": 3.878558888404655e-05,
+      "loss": 0.6521,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5851578704770685,
+      "grad_norm": 0.19896583259105682,
+      "learning_rate": 3.874921423423697e-05,
+      "loss": 0.6568,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5853883383268034,
+      "grad_norm": 0.18624718487262726,
+      "learning_rate": 3.8712845855345325e-05,
+      "loss": 0.6519,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5856188061765384,
+      "grad_norm": 0.19909995794296265,
+      "learning_rate": 3.867648376764248e-05,
+      "loss": 0.6624,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5858492740262733,
+      "grad_norm": 0.18448121845722198,
+      "learning_rate": 3.864012799139581e-05,
+      "loss": 0.6622,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5860797418760083,
+      "grad_norm": 0.20322075486183167,
+      "learning_rate": 3.860377854686913e-05,
+      "loss": 0.6648,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5863102097257432,
+      "grad_norm": 0.19707633554935455,
+      "learning_rate": 3.8567435454322745e-05,
+      "loss": 0.6581,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5865406775754782,
+      "grad_norm": 0.18212544918060303,
+      "learning_rate": 3.853109873401346e-05,
+      "loss": 0.6537,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5867711454252131,
+      "grad_norm": 0.19013650715351105,
+      "learning_rate": 3.849476840619443e-05,
+      "loss": 0.6569,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5870016132749482,
+      "grad_norm": 0.20143313705921173,
+      "learning_rate": 3.845844449111535e-05,
+      "loss": 0.6615,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5872320811246831,
+      "grad_norm": 0.20364569127559662,
+      "learning_rate": 3.842212700902231e-05,
+      "loss": 0.6565,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5874625489744181,
+      "grad_norm": 0.1817772090435028,
+      "learning_rate": 3.838581598015776e-05,
+      "loss": 0.6533,
+      "step": 2549
+    },
+    {
+      "epoch": 0.587693016824153,
+      "grad_norm": 0.1907220184803009,
+      "learning_rate": 3.834951142476063e-05,
+      "loss": 0.6604,
+      "step": 2550
+    },
+    {
+      "epoch": 0.587923484673888,
+      "grad_norm": 0.19262264668941498,
+      "learning_rate": 3.83132133630662e-05,
+      "loss": 0.656,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5881539525236229,
+      "grad_norm": 0.18963393568992615,
+      "learning_rate": 3.827692181530612e-05,
+      "loss": 0.6608,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5883844203733579,
+      "grad_norm": 0.20289820432662964,
+      "learning_rate": 3.8240636801708444e-05,
+      "loss": 0.6643,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5886148882230928,
+      "grad_norm": 0.17837589979171753,
+      "learning_rate": 3.8204358342497584e-05,
+      "loss": 0.6626,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5888453560728278,
+      "grad_norm": 0.2154434323310852,
+      "learning_rate": 3.816808645789425e-05,
+      "loss": 0.6643,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5890758239225629,
+      "grad_norm": 0.19185711443424225,
+      "learning_rate": 3.813182116811552e-05,
+      "loss": 0.6568,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5893062917722978,
+      "grad_norm": 0.18890288472175598,
+      "learning_rate": 3.809556249337484e-05,
+      "loss": 0.652,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5895367596220328,
+      "grad_norm": 0.20426598191261292,
+      "learning_rate": 3.805931045388188e-05,
+      "loss": 0.6576,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5897672274717677,
+      "grad_norm": 0.18901459872722626,
+      "learning_rate": 3.8023065069842653e-05,
+      "loss": 0.6577,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5899976953215027,
+      "grad_norm": 0.19619275629520416,
+      "learning_rate": 3.7986826361459524e-05,
+      "loss": 0.6664,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5902281631712376,
+      "grad_norm": 0.1934560090303421,
+      "learning_rate": 3.795059434893101e-05,
+      "loss": 0.6597,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5904586310209726,
+      "grad_norm": 0.1980142444372177,
+      "learning_rate": 3.7914369052452006e-05,
+      "loss": 0.6619,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5906890988707075,
+      "grad_norm": 0.1985742449760437,
+      "learning_rate": 3.787815049221361e-05,
+      "loss": 0.6524,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5909195667204425,
+      "grad_norm": 0.1910460889339447,
+      "learning_rate": 3.784193868840318e-05,
+      "loss": 0.6597,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5911500345701775,
+      "grad_norm": 0.1858750432729721,
+      "learning_rate": 3.7805733661204306e-05,
+      "loss": 0.6636,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5913805024199125,
+      "grad_norm": 0.20883874595165253,
+      "learning_rate": 3.77695354307968e-05,
+      "loss": 0.6565,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5916109702696474,
+      "grad_norm": 0.18166543543338776,
+      "learning_rate": 3.77333440173567e-05,
+      "loss": 0.6525,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5918414381193824,
+      "grad_norm": 0.1856246143579483,
+      "learning_rate": 3.7697159441056205e-05,
+      "loss": 0.6615,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5920719059691173,
+      "grad_norm": 0.19537490606307983,
+      "learning_rate": 3.7660981722063745e-05,
+      "loss": 0.6563,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5923023738188523,
+      "grad_norm": 0.19945433735847473,
+      "learning_rate": 3.762481088054393e-05,
+      "loss": 0.6573,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5925328416685872,
+      "grad_norm": 0.20831739902496338,
+      "learning_rate": 3.758864693665748e-05,
+      "loss": 0.6633,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5927633095183222,
+      "grad_norm": 0.21567675471305847,
+      "learning_rate": 3.7552489910561326e-05,
+      "loss": 0.6572,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5929937773680571,
+      "grad_norm": 0.215605229139328,
+      "learning_rate": 3.751633982240855e-05,
+      "loss": 0.6625,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5932242452177922,
+      "grad_norm": 0.19509264826774597,
+      "learning_rate": 3.7480196692348315e-05,
+      "loss": 0.6619,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5934547130675271,
+      "grad_norm": 0.20573195815086365,
+      "learning_rate": 3.744406054052594e-05,
+      "loss": 0.6536,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5936851809172621,
+      "grad_norm": 0.18587082624435425,
+      "learning_rate": 3.740793138708285e-05,
+      "loss": 0.6453,
+      "step": 2576
+    },
+    {
+      "epoch": 0.593915648766997,
+      "grad_norm": 0.20044218003749847,
+      "learning_rate": 3.737180925215658e-05,
+      "loss": 0.6622,
+      "step": 2577
+    },
+    {
+      "epoch": 0.594146116616732,
+      "grad_norm": 0.20294342935085297,
+      "learning_rate": 3.733569415588071e-05,
+      "loss": 0.6649,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5943765844664669,
+      "grad_norm": 0.1900031715631485,
+      "learning_rate": 3.729958611838496e-05,
+      "loss": 0.6607,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5946070523162019,
+      "grad_norm": 0.21229608356952667,
+      "learning_rate": 3.7263485159795075e-05,
+      "loss": 0.6624,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5948375201659368,
+      "grad_norm": 0.18590863049030304,
+      "learning_rate": 3.7227391300232845e-05,
+      "loss": 0.6529,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5950679880156718,
+      "grad_norm": 0.20582017302513123,
+      "learning_rate": 3.7191304559816165e-05,
+      "loss": 0.6516,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5952984558654068,
+      "grad_norm": 0.18912175297737122,
+      "learning_rate": 3.715522495865885e-05,
+      "loss": 0.6669,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5955289237151418,
+      "grad_norm": 0.2016802579164505,
+      "learning_rate": 3.711915251687086e-05,
+      "loss": 0.6562,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5957593915648767,
+      "grad_norm": 0.18063144385814667,
+      "learning_rate": 3.7083087254558116e-05,
+      "loss": 0.6661,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5959898594146117,
+      "grad_norm": 0.20485232770442963,
+      "learning_rate": 3.7047029191822455e-05,
+      "loss": 0.6597,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5962203272643466,
+      "grad_norm": 0.17791825532913208,
+      "learning_rate": 3.701097834876185e-05,
+      "loss": 0.6555,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5964507951140816,
+      "grad_norm": 0.2008713036775589,
+      "learning_rate": 3.697493474547016e-05,
+      "loss": 0.6649,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5966812629638165,
+      "grad_norm": 0.19855642318725586,
+      "learning_rate": 3.693889840203719e-05,
+      "loss": 0.6608,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5969117308135515,
+      "grad_norm": 0.1890726089477539,
+      "learning_rate": 3.690286933854877e-05,
+      "loss": 0.6555,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5971421986632864,
+      "grad_norm": 0.19872720539569855,
+      "learning_rate": 3.686684757508663e-05,
+      "loss": 0.6596,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5973726665130215,
+      "grad_norm": 0.1832602471113205,
+      "learning_rate": 3.683083313172841e-05,
+      "loss": 0.6586,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5976031343627564,
+      "grad_norm": 0.2064426690340042,
+      "learning_rate": 3.679482602854773e-05,
+      "loss": 0.6593,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5978336022124914,
+      "grad_norm": 0.17292998731136322,
+      "learning_rate": 3.675882628561408e-05,
+      "loss": 0.6601,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5980640700622263,
+      "grad_norm": 0.2106057107448578,
+      "learning_rate": 3.672283392299282e-05,
+      "loss": 0.6503,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5982945379119613,
+      "grad_norm": 0.1862911581993103,
+      "learning_rate": 3.6686848960745244e-05,
+      "loss": 0.6562,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5985250057616962,
+      "grad_norm": 0.1884952336549759,
+      "learning_rate": 3.6650871418928556e-05,
+      "loss": 0.6618,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5987554736114312,
+      "grad_norm": 0.17884479463100433,
+      "learning_rate": 3.66149013175957e-05,
+      "loss": 0.6558,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5989859414611661,
+      "grad_norm": 0.1988133192062378,
+      "learning_rate": 3.6578938676795575e-05,
+      "loss": 0.6535,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5992164093109011,
+      "grad_norm": 0.18073412775993347,
+      "learning_rate": 3.6542983516572915e-05,
+      "loss": 0.6506,
+      "step": 2600
+    },
+    {
+      "epoch": 0.599446877160636,
+      "grad_norm": 0.19410258531570435,
+      "learning_rate": 3.650703585696824e-05,
+      "loss": 0.6578,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5996773450103711,
+      "grad_norm": 0.1835651844739914,
+      "learning_rate": 3.647109571801792e-05,
+      "loss": 0.6606,
+      "step": 2602
+    },
+    {
+      "epoch": 0.599907812860106,
+      "grad_norm": 0.21423934400081635,
+      "learning_rate": 3.643516311975413e-05,
+      "loss": 0.6515,
+      "step": 2603
+    },
+    {
+      "epoch": 0.600138280709841,
+      "grad_norm": 0.19270770251750946,
+      "learning_rate": 3.639923808220483e-05,
+      "loss": 0.6521,
+      "step": 2604
+    },
+    {
+      "epoch": 0.6003687485595759,
+      "grad_norm": 0.18155768513679504,
+      "learning_rate": 3.636332062539378e-05,
+      "loss": 0.6557,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6005992164093109,
+      "grad_norm": 0.2070186883211136,
+      "learning_rate": 3.6327410769340505e-05,
+      "loss": 0.6535,
+      "step": 2606
+    },
+    {
+      "epoch": 0.6008296842590458,
+      "grad_norm": 0.17192067205905914,
+      "learning_rate": 3.6291508534060296e-05,
+      "loss": 0.6545,
+      "step": 2607
+    },
+    {
+      "epoch": 0.6010601521087808,
+      "grad_norm": 0.20923547446727753,
+      "learning_rate": 3.625561393956417e-05,
+      "loss": 0.6645,
+      "step": 2608
+    },
+    {
+      "epoch": 0.6012906199585157,
+      "grad_norm": 0.18093320727348328,
+      "learning_rate": 3.621972700585898e-05,
+      "loss": 0.653,
+      "step": 2609
+    },
+    {
+      "epoch": 0.6015210878082508,
+      "grad_norm": 0.19666415452957153,
+      "learning_rate": 3.618384775294718e-05,
+      "loss": 0.6512,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6017515556579857,
+      "grad_norm": 0.1838379055261612,
+      "learning_rate": 3.614797620082703e-05,
+      "loss": 0.658,
+      "step": 2611
+    },
+    {
+      "epoch": 0.6019820235077207,
+      "grad_norm": 0.19848011434078217,
+      "learning_rate": 3.611211236949248e-05,
+      "loss": 0.6495,
+      "step": 2612
+    },
+    {
+      "epoch": 0.6022124913574556,
+      "grad_norm": 0.19375987350940704,
+      "learning_rate": 3.6076256278933145e-05,
+      "loss": 0.6578,
+      "step": 2613
+    },
+    {
+      "epoch": 0.6024429592071906,
+      "grad_norm": 0.18482361733913422,
+      "learning_rate": 3.604040794913437e-05,
+      "loss": 0.6623,
+      "step": 2614
+    },
+    {
+      "epoch": 0.6026734270569255,
+      "grad_norm": 0.18653304874897003,
+      "learning_rate": 3.600456740007714e-05,
+      "loss": 0.6625,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6029038949066605,
+      "grad_norm": 0.17585410177707672,
+      "learning_rate": 3.5968734651738124e-05,
+      "loss": 0.6578,
+      "step": 2616
+    },
+    {
+      "epoch": 0.6031343627563955,
+      "grad_norm": 0.18889284133911133,
+      "learning_rate": 3.5932909724089626e-05,
+      "loss": 0.6487,
+      "step": 2617
+    },
+    {
+      "epoch": 0.6033648306061304,
+      "grad_norm": 0.1867912858724594,
+      "learning_rate": 3.589709263709963e-05,
+      "loss": 0.65,
+      "step": 2618
+    },
+    {
+      "epoch": 0.6035952984558655,
+      "grad_norm": 0.18602915108203888,
+      "learning_rate": 3.586128341073167e-05,
+      "loss": 0.6587,
+      "step": 2619
+    },
+    {
+      "epoch": 0.6038257663056004,
+      "grad_norm": 0.16733336448669434,
+      "learning_rate": 3.582548206494499e-05,
+      "loss": 0.6489,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6040562341553354,
+      "grad_norm": 0.1732863187789917,
+      "learning_rate": 3.578968861969441e-05,
+      "loss": 0.6551,
+      "step": 2621
+    },
+    {
+      "epoch": 0.6042867020050703,
+      "grad_norm": 0.18000958859920502,
+      "learning_rate": 3.575390309493029e-05,
+      "loss": 0.6646,
+      "step": 2622
+    },
+    {
+      "epoch": 0.6045171698548053,
+      "grad_norm": 0.17645825445652008,
+      "learning_rate": 3.5718125510598645e-05,
+      "loss": 0.6579,
+      "step": 2623
+    },
+    {
+      "epoch": 0.6047476377045402,
+      "grad_norm": 0.18449077010154724,
+      "learning_rate": 3.5682355886641075e-05,
+      "loss": 0.6467,
+      "step": 2624
+    },
+    {
+      "epoch": 0.6049781055542752,
+      "grad_norm": 0.1781143993139267,
+      "learning_rate": 3.5646594242994646e-05,
+      "loss": 0.6607,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6052085734040101,
+      "grad_norm": 0.1856512576341629,
+      "learning_rate": 3.5610840599592096e-05,
+      "loss": 0.6588,
+      "step": 2626
+    },
+    {
+      "epoch": 0.6054390412537451,
+      "grad_norm": 0.18362030386924744,
+      "learning_rate": 3.5575094976361625e-05,
+      "loss": 0.6636,
+      "step": 2627
+    },
+    {
+      "epoch": 0.60566950910348,
+      "grad_norm": 0.172404944896698,
+      "learning_rate": 3.553935739322698e-05,
+      "loss": 0.6543,
+      "step": 2628
+    },
+    {
+      "epoch": 0.6058999769532151,
+      "grad_norm": 0.18513377010822296,
+      "learning_rate": 3.550362787010744e-05,
+      "loss": 0.6549,
+      "step": 2629
+    },
+    {
+      "epoch": 0.60613044480295,
+      "grad_norm": 0.1736796498298645,
+      "learning_rate": 3.5467906426917795e-05,
+      "loss": 0.6516,
+      "step": 2630
+    },
+    {
+      "epoch": 0.606360912652685,
+      "grad_norm": 0.18415535986423492,
+      "learning_rate": 3.54321930835683e-05,
+      "loss": 0.6537,
+      "step": 2631
+    },
+    {
+      "epoch": 0.6065913805024199,
+      "grad_norm": 0.1662503480911255,
+      "learning_rate": 3.539648785996471e-05,
+      "loss": 0.6554,
+      "step": 2632
+    },
+    {
+      "epoch": 0.6068218483521549,
+      "grad_norm": 0.17874038219451904,
+      "learning_rate": 3.536079077600829e-05,
+      "loss": 0.6543,
+      "step": 2633
+    },
+    {
+      "epoch": 0.6070523162018898,
+      "grad_norm": 0.17104828357696533,
+      "learning_rate": 3.532510185159572e-05,
+      "loss": 0.6532,
+      "step": 2634
+    },
+    {
+      "epoch": 0.6072827840516248,
+      "grad_norm": 0.1689852625131607,
+      "learning_rate": 3.5289421106619126e-05,
+      "loss": 0.6549,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6075132519013597,
+      "grad_norm": 0.17767558991909027,
+      "learning_rate": 3.525374856096616e-05,
+      "loss": 0.6609,
+      "step": 2636
+    },
+    {
+      "epoch": 0.6077437197510948,
+      "grad_norm": 0.1855451464653015,
+      "learning_rate": 3.5218084234519764e-05,
+      "loss": 0.6619,
+      "step": 2637
+    },
+    {
+      "epoch": 0.6079741876008297,
+      "grad_norm": 0.17707620561122894,
+      "learning_rate": 3.518242814715844e-05,
+      "loss": 0.6615,
+      "step": 2638
+    },
+    {
+      "epoch": 0.6082046554505647,
+      "grad_norm": 0.18121664226055145,
+      "learning_rate": 3.514678031875602e-05,
+      "loss": 0.663,
+      "step": 2639
+    },
+    {
+      "epoch": 0.6084351233002996,
+      "grad_norm": 0.1841084361076355,
+      "learning_rate": 3.5111140769181735e-05,
+      "loss": 0.6526,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6086655911500346,
+      "grad_norm": 0.18913474678993225,
+      "learning_rate": 3.5075509518300224e-05,
+      "loss": 0.6576,
+      "step": 2641
+    },
+    {
+      "epoch": 0.6088960589997695,
+      "grad_norm": 0.1743757128715515,
+      "learning_rate": 3.5039886585971496e-05,
+      "loss": 0.6534,
+      "step": 2642
+    },
+    {
+      "epoch": 0.6091265268495045,
+      "grad_norm": 0.1718021035194397,
+      "learning_rate": 3.500427199205091e-05,
+      "loss": 0.667,
+      "step": 2643
+    },
+    {
+      "epoch": 0.6093569946992394,
+      "grad_norm": 0.18834391236305237,
+      "learning_rate": 3.4968665756389174e-05,
+      "loss": 0.6608,
+      "step": 2644
+    },
+    {
+      "epoch": 0.6095874625489744,
+      "grad_norm": 0.17941658198833466,
+      "learning_rate": 3.4933067898832375e-05,
+      "loss": 0.6586,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6098179303987094,
+      "grad_norm": 0.1901034712791443,
+      "learning_rate": 3.489747843922189e-05,
+      "loss": 0.6558,
+      "step": 2646
+    },
+    {
+      "epoch": 0.6100483982484444,
+      "grad_norm": 0.1816999912261963,
+      "learning_rate": 3.4861897397394415e-05,
+      "loss": 0.6595,
+      "step": 2647
+    },
+    {
+      "epoch": 0.6102788660981793,
+      "grad_norm": 0.187819242477417,
+      "learning_rate": 3.482632479318201e-05,
+      "loss": 0.6464,
+      "step": 2648
+    },
+    {
+      "epoch": 0.6105093339479143,
+      "grad_norm": 0.20110177993774414,
+      "learning_rate": 3.479076064641195e-05,
+      "loss": 0.6544,
+      "step": 2649
+    },
+    {
+      "epoch": 0.6107398017976492,
+      "grad_norm": 0.19999714195728302,
+      "learning_rate": 3.475520497690684e-05,
+      "loss": 0.6536,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6109702696473842,
+      "grad_norm": 0.20862232148647308,
+      "learning_rate": 3.471965780448461e-05,
+      "loss": 0.6516,
+      "step": 2651
+    },
+    {
+      "epoch": 0.6112007374971191,
+      "grad_norm": 0.1882224977016449,
+      "learning_rate": 3.4684119148958314e-05,
+      "loss": 0.6561,
+      "step": 2652
+    },
+    {
+      "epoch": 0.6114312053468541,
+      "grad_norm": 0.22021904587745667,
+      "learning_rate": 3.464858903013641e-05,
+      "loss": 0.6543,
+      "step": 2653
+    },
+    {
+      "epoch": 0.611661673196589,
+      "grad_norm": 0.1685914695262909,
+      "learning_rate": 3.461306746782253e-05,
+      "loss": 0.6563,
+      "step": 2654
+    },
+    {
+      "epoch": 0.6118921410463241,
+      "grad_norm": 0.19287706911563873,
+      "learning_rate": 3.457755448181551e-05,
+      "loss": 0.647,
+      "step": 2655
+    },
+    {
+      "epoch": 0.612122608896059,
+      "grad_norm": 0.18482106924057007,
+      "learning_rate": 3.454205009190945e-05,
+      "loss": 0.6534,
+      "step": 2656
+    },
+    {
+      "epoch": 0.612353076745794,
+      "grad_norm": 0.18831658363342285,
+      "learning_rate": 3.450655431789366e-05,
+      "loss": 0.6577,
+      "step": 2657
+    },
+    {
+      "epoch": 0.6125835445955289,
+      "grad_norm": 0.1809016317129135,
+      "learning_rate": 3.447106717955261e-05,
+      "loss": 0.6562,
+      "step": 2658
+    },
+    {
+      "epoch": 0.6128140124452639,
+      "grad_norm": 0.19347211718559265,
+      "learning_rate": 3.443558869666598e-05,
+      "loss": 0.649,
+      "step": 2659
+    },
+    {
+      "epoch": 0.6130444802949988,
+      "grad_norm": 0.19009865820407867,
+      "learning_rate": 3.4400118889008646e-05,
+      "loss": 0.6468,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6132749481447338,
+      "grad_norm": 0.19617056846618652,
+      "learning_rate": 3.4364657776350605e-05,
+      "loss": 0.658,
+      "step": 2661
+    },
+    {
+      "epoch": 0.6135054159944687,
+      "grad_norm": 0.19712841510772705,
+      "learning_rate": 3.432920537845703e-05,
+      "loss": 0.6584,
+      "step": 2662
+    },
+    {
+      "epoch": 0.6137358838442037,
+      "grad_norm": 0.18756158649921417,
+      "learning_rate": 3.429376171508827e-05,
+      "loss": 0.6505,
+      "step": 2663
+    },
+    {
+      "epoch": 0.6139663516939387,
+      "grad_norm": 0.19545352458953857,
+      "learning_rate": 3.425832680599972e-05,
+      "loss": 0.6567,
+      "step": 2664
+    },
+    {
+      "epoch": 0.6141968195436737,
+      "grad_norm": 0.19188739359378815,
+      "learning_rate": 3.4222900670941995e-05,
+      "loss": 0.6664,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6144272873934086,
+      "grad_norm": 0.19670219719409943,
+      "learning_rate": 3.4187483329660755e-05,
+      "loss": 0.6572,
+      "step": 2666
+    },
+    {
+      "epoch": 0.6146577552431436,
+      "grad_norm": 0.18227505683898926,
+      "learning_rate": 3.415207480189676e-05,
+      "loss": 0.6524,
+      "step": 2667
+    },
+    {
+      "epoch": 0.6148882230928785,
+      "grad_norm": 0.19035394489765167,
+      "learning_rate": 3.4116675107385885e-05,
+      "loss": 0.6538,
+      "step": 2668
+    },
+    {
+      "epoch": 0.6151186909426135,
+      "grad_norm": 0.18735790252685547,
+      "learning_rate": 3.408128426585909e-05,
+      "loss": 0.6554,
+      "step": 2669
+    },
+    {
+      "epoch": 0.6153491587923484,
+      "grad_norm": 0.20038571953773499,
+      "learning_rate": 3.4045902297042356e-05,
+      "loss": 0.6627,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6155796266420834,
+      "grad_norm": 0.17390772700309753,
+      "learning_rate": 3.401052922065675e-05,
+      "loss": 0.6551,
+      "step": 2671
+    },
+    {
+      "epoch": 0.6158100944918183,
+      "grad_norm": 0.19239278137683868,
+      "learning_rate": 3.397516505641838e-05,
+      "loss": 0.654,
+      "step": 2672
+    },
+    {
+      "epoch": 0.6160405623415534,
+      "grad_norm": 0.1672012060880661,
+      "learning_rate": 3.393980982403837e-05,
+      "loss": 0.6537,
+      "step": 2673
+    },
+    {
+      "epoch": 0.6162710301912883,
+      "grad_norm": 0.19565598666667938,
+      "learning_rate": 3.3904463543222876e-05,
+      "loss": 0.6523,
+      "step": 2674
+    },
+    {
+      "epoch": 0.6165014980410233,
+      "grad_norm": 0.17734399437904358,
+      "learning_rate": 3.386912623367311e-05,
+      "loss": 0.6533,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6167319658907583,
+      "grad_norm": 0.18294768035411835,
+      "learning_rate": 3.383379791508519e-05,
+      "loss": 0.6492,
+      "step": 2676
+    },
+    {
+      "epoch": 0.6169624337404932,
+      "grad_norm": 0.187822625041008,
+      "learning_rate": 3.3798478607150274e-05,
+      "loss": 0.6568,
+      "step": 2677
+    },
+    {
+      "epoch": 0.6171929015902282,
+      "grad_norm": 0.17213211953639984,
+      "learning_rate": 3.376316832955454e-05,
+      "loss": 0.6634,
+      "step": 2678
+    },
+    {
+      "epoch": 0.6174233694399631,
+      "grad_norm": 0.18802215158939362,
+      "learning_rate": 3.3727867101979036e-05,
+      "loss": 0.6512,
+      "step": 2679
+    },
+    {
+      "epoch": 0.6176538372896981,
+      "grad_norm": 0.16777820885181427,
+      "learning_rate": 3.369257494409985e-05,
+      "loss": 0.6475,
+      "step": 2680
+    },
+    {
+      "epoch": 0.617884305139433,
+      "grad_norm": 0.19203679263591766,
+      "learning_rate": 3.3657291875587996e-05,
+      "loss": 0.6552,
+      "step": 2681
+    },
+    {
+      "epoch": 0.6181147729891681,
+      "grad_norm": 0.19129416346549988,
+      "learning_rate": 3.3622017916109376e-05,
+      "loss": 0.6645,
+      "step": 2682
+    },
+    {
+      "epoch": 0.618345240838903,
+      "grad_norm": 0.17216765880584717,
+      "learning_rate": 3.358675308532486e-05,
+      "loss": 0.6611,
+      "step": 2683
+    },
+    {
+      "epoch": 0.618575708688638,
+      "grad_norm": 0.18110480904579163,
+      "learning_rate": 3.355149740289024e-05,
+      "loss": 0.6505,
+      "step": 2684
+    },
+    {
+      "epoch": 0.6188061765383729,
+      "grad_norm": 0.18171606957912445,
+      "learning_rate": 3.3516250888456165e-05,
+      "loss": 0.6439,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6190366443881079,
+      "grad_norm": 0.17510759830474854,
+      "learning_rate": 3.348101356166819e-05,
+      "loss": 0.6511,
+      "step": 2686
+    },
+    {
+      "epoch": 0.6192671122378428,
+      "grad_norm": 0.18047381937503815,
+      "learning_rate": 3.344578544216678e-05,
+      "loss": 0.6561,
+      "step": 2687
+    },
+    {
+      "epoch": 0.6194975800875778,
+      "grad_norm": 0.17559714615345,
+      "learning_rate": 3.341056654958722e-05,
+      "loss": 0.6538,
+      "step": 2688
+    },
+    {
+      "epoch": 0.6197280479373127,
+      "grad_norm": 0.18052367866039276,
+      "learning_rate": 3.3375356903559676e-05,
+      "loss": 0.6521,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6199585157870477,
+      "grad_norm": 0.17809216678142548,
+      "learning_rate": 3.3340156523709185e-05,
+      "loss": 0.657,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6201889836367827,
+      "grad_norm": 0.18209248781204224,
+      "learning_rate": 3.330496542965556e-05,
+      "loss": 0.651,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6204194514865177,
+      "grad_norm": 0.1755290925502777,
+      "learning_rate": 3.32697836410135e-05,
+      "loss": 0.6581,
+      "step": 2692
+    },
+    {
+      "epoch": 0.6206499193362526,
+      "grad_norm": 0.16924412548542023,
+      "learning_rate": 3.323461117739248e-05,
+      "loss": 0.6497,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6208803871859876,
+      "grad_norm": 0.18715137243270874,
+      "learning_rate": 3.3199448058396786e-05,
+      "loss": 0.6562,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6211108550357225,
+      "grad_norm": 0.17708511650562286,
+      "learning_rate": 3.3164294303625486e-05,
+      "loss": 0.6584,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6213413228854575,
+      "grad_norm": 0.1799190789461136,
+      "learning_rate": 3.312914993267246e-05,
+      "loss": 0.6591,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6215717907351924,
+      "grad_norm": 0.19582661986351013,
+      "learning_rate": 3.309401496512633e-05,
+      "loss": 0.6479,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6218022585849274,
+      "grad_norm": 0.17881685495376587,
+      "learning_rate": 3.305888942057047e-05,
+      "loss": 0.653,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6220327264346623,
+      "grad_norm": 0.1858767569065094,
+      "learning_rate": 3.302377331858302e-05,
+      "loss": 0.652,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6222631942843974,
+      "grad_norm": 0.16495144367218018,
+      "learning_rate": 3.298866667873688e-05,
+      "loss": 0.6534,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6224936621341323,
+      "grad_norm": 0.1828274130821228,
+      "learning_rate": 3.2953569520599606e-05,
+      "loss": 0.6452,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6227241299838673,
+      "grad_norm": 0.1840105652809143,
+      "learning_rate": 3.291848186373353e-05,
+      "loss": 0.6509,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6229545978336022,
+      "grad_norm": 0.1832989603281021,
+      "learning_rate": 3.2883403727695694e-05,
+      "loss": 0.6469,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6231850656833372,
+      "grad_norm": 0.18650995194911957,
+      "learning_rate": 3.2848335132037763e-05,
+      "loss": 0.654,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6234155335330721,
+      "grad_norm": 0.17295216023921967,
+      "learning_rate": 3.2813276096306156e-05,
+      "loss": 0.654,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6236460013828071,
+      "grad_norm": 0.17538554966449738,
+      "learning_rate": 3.2778226640041956e-05,
+      "loss": 0.6558,
+      "step": 2706
+    },
+    {
+      "epoch": 0.623876469232542,
+      "grad_norm": 0.15956903994083405,
+      "learning_rate": 3.274318678278085e-05,
+      "loss": 0.6538,
+      "step": 2707
+    },
+    {
+      "epoch": 0.624106937082277,
+      "grad_norm": 0.18525734543800354,
+      "learning_rate": 3.270815654405324e-05,
+      "loss": 0.6538,
+      "step": 2708
+    },
+    {
+      "epoch": 0.624337404932012,
+      "grad_norm": 0.16766776144504547,
+      "learning_rate": 3.267313594338414e-05,
+      "loss": 0.6503,
+      "step": 2709
+    },
+    {
+      "epoch": 0.624567872781747,
+      "grad_norm": 0.19396458566188812,
+      "learning_rate": 3.2638125000293165e-05,
+      "loss": 0.6553,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6247983406314819,
+      "grad_norm": 0.1728057563304901,
+      "learning_rate": 3.2603123734294606e-05,
+      "loss": 0.6621,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6250288084812169,
+      "grad_norm": 0.179381862282753,
+      "learning_rate": 3.2568132164897316e-05,
+      "loss": 0.6579,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6252592763309518,
+      "grad_norm": 0.1791330873966217,
+      "learning_rate": 3.253315031160475e-05,
+      "loss": 0.66,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6254897441806868,
+      "grad_norm": 0.18653972446918488,
+      "learning_rate": 3.249817819391495e-05,
+      "loss": 0.6593,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6257202120304217,
+      "grad_norm": 0.17987924814224243,
+      "learning_rate": 3.246321583132058e-05,
+      "loss": 0.6539,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6259506798801567,
+      "grad_norm": 0.19162432849407196,
+      "learning_rate": 3.242826324330875e-05,
+      "loss": 0.6572,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6261811477298916,
+      "grad_norm": 0.19111159443855286,
+      "learning_rate": 3.2393320449361246e-05,
+      "loss": 0.6546,
+      "step": 2717
+    },
+    {
+      "epoch": 0.6264116155796267,
+      "grad_norm": 0.17735035717487335,
+      "learning_rate": 3.235838746895434e-05,
+      "loss": 0.6561,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6266420834293616,
+      "grad_norm": 0.1921262890100479,
+      "learning_rate": 3.2323464321558804e-05,
+      "loss": 0.6506,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6268725512790966,
+      "grad_norm": 0.1894858479499817,
+      "learning_rate": 3.2288551026639986e-05,
+      "loss": 0.6502,
+      "step": 2720
+    },
+    {
+      "epoch": 0.6271030191288315,
+      "grad_norm": 0.19639429450035095,
+      "learning_rate": 3.225364760365772e-05,
+      "loss": 0.6546,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6273334869785665,
+      "grad_norm": 0.1766640543937683,
+      "learning_rate": 3.221875407206633e-05,
+      "loss": 0.66,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6275639548283014,
+      "grad_norm": 0.18973618745803833,
+      "learning_rate": 3.2183870451314624e-05,
+      "loss": 0.6616,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6277944226780364,
+      "grad_norm": 0.19568420946598053,
+      "learning_rate": 3.2148996760845924e-05,
+      "loss": 0.653,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6280248905277713,
+      "grad_norm": 0.18159765005111694,
+      "learning_rate": 3.2114133020097945e-05,
+      "loss": 0.6494,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6282553583775063,
+      "grad_norm": 0.2033732682466507,
+      "learning_rate": 3.2079279248502925e-05,
+      "loss": 0.6529,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6284858262272413,
+      "grad_norm": 0.17295779287815094,
+      "learning_rate": 3.204443546548754e-05,
+      "loss": 0.65,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6287162940769763,
+      "grad_norm": 0.1940968781709671,
+      "learning_rate": 3.2009601690472834e-05,
+      "loss": 0.6572,
+      "step": 2728
+    },
+    {
+      "epoch": 0.6289467619267112,
+      "grad_norm": 0.17500554025173187,
+      "learning_rate": 3.197477794287435e-05,
+      "loss": 0.6498,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6291772297764462,
+      "grad_norm": 0.18749865889549255,
+      "learning_rate": 3.1939964242102004e-05,
+      "loss": 0.65,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6294076976261811,
+      "grad_norm": 0.19215604662895203,
+      "learning_rate": 3.190516060756009e-05,
+      "loss": 0.6679,
+      "step": 2731
+    },
+    {
+      "epoch": 0.6296381654759161,
+      "grad_norm": 0.18513312935829163,
+      "learning_rate": 3.187036705864736e-05,
+      "loss": 0.6617,
+      "step": 2732
+    },
+    {
+      "epoch": 0.629868633325651,
+      "grad_norm": 0.20599323511123657,
+      "learning_rate": 3.183558361475687e-05,
+      "loss": 0.652,
+      "step": 2733
+    },
+    {
+      "epoch": 0.630099101175386,
+      "grad_norm": 0.18766285479068756,
+      "learning_rate": 3.1800810295276095e-05,
+      "loss": 0.6536,
+      "step": 2734
+    },
+    {
+      "epoch": 0.630329569025121,
+      "grad_norm": 0.19500939548015594,
+      "learning_rate": 3.1766047119586845e-05,
+      "loss": 0.6591,
+      "step": 2735
+    },
+    {
+      "epoch": 0.630560036874856,
+      "grad_norm": 0.19361793994903564,
+      "learning_rate": 3.173129410706528e-05,
+      "loss": 0.6434,
+      "step": 2736
+    },
+    {
+      "epoch": 0.630790504724591,
+      "grad_norm": 0.17595571279525757,
+      "learning_rate": 3.169655127708189e-05,
+      "loss": 0.6405,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6310209725743259,
+      "grad_norm": 0.19558657705783844,
+      "learning_rate": 3.1661818649001494e-05,
+      "loss": 0.6524,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6312514404240609,
+      "grad_norm": 0.18385924398899078,
+      "learning_rate": 3.1627096242183243e-05,
+      "loss": 0.6614,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6314819082737958,
+      "grad_norm": 0.19142907857894897,
+      "learning_rate": 3.159238407598054e-05,
+      "loss": 0.657,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6317123761235308,
+      "grad_norm": 0.18829232454299927,
+      "learning_rate": 3.155768216974112e-05,
+      "loss": 0.6483,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6319428439732657,
+      "grad_norm": 0.1925613284111023,
+      "learning_rate": 3.1522990542807046e-05,
+      "loss": 0.6478,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6321733118230007,
+      "grad_norm": 0.17510046064853668,
+      "learning_rate": 3.148830921451452e-05,
+      "loss": 0.6551,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6324037796727356,
+      "grad_norm": 0.17866365611553192,
+      "learning_rate": 3.1453638204194135e-05,
+      "loss": 0.6585,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6326342475224707,
+      "grad_norm": 0.18686874210834503,
+      "learning_rate": 3.141897753117067e-05,
+      "loss": 0.6517,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6328647153722056,
+      "grad_norm": 0.16817151010036469,
+      "learning_rate": 3.1384327214763135e-05,
+      "loss": 0.6591,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6330951832219406,
+      "grad_norm": 0.1925475150346756,
+      "learning_rate": 3.13496872742848e-05,
+      "loss": 0.6576,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6333256510716755,
+      "grad_norm": 0.19002029299736023,
+      "learning_rate": 3.131505772904314e-05,
+      "loss": 0.6528,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6335561189214105,
+      "grad_norm": 0.1708815097808838,
+      "learning_rate": 3.128043859833981e-05,
+      "loss": 0.6493,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6337865867711454,
+      "grad_norm": 0.19003167748451233,
+      "learning_rate": 3.12458299014707e-05,
+      "loss": 0.6568,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6340170546208804,
+      "grad_norm": 0.1745605319738388,
+      "learning_rate": 3.121123165772588e-05,
+      "loss": 0.6547,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6342475224706153,
+      "grad_norm": 0.18086248636245728,
+      "learning_rate": 3.1176643886389554e-05,
+      "loss": 0.6454,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6344779903203503,
+      "grad_norm": 0.17058676481246948,
+      "learning_rate": 3.114206660674011e-05,
+      "loss": 0.6611,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6347084581700853,
+      "grad_norm": 0.1747319996356964,
+      "learning_rate": 3.1107499838050146e-05,
+      "loss": 0.6428,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6349389260198203,
+      "grad_norm": 0.17745546996593475,
+      "learning_rate": 3.107294359958628e-05,
+      "loss": 0.6602,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6351693938695552,
+      "grad_norm": 0.16670075058937073,
+      "learning_rate": 3.103839791060937e-05,
+      "loss": 0.6608,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6353998617192902,
+      "grad_norm": 0.1686505526304245,
+      "learning_rate": 3.1003862790374364e-05,
+      "loss": 0.6535,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6356303295690251,
+      "grad_norm": 0.1769195795059204,
+      "learning_rate": 3.0969338258130274e-05,
+      "loss": 0.6537,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6358607974187601,
+      "grad_norm": 0.16672547161579132,
+      "learning_rate": 3.093482433312027e-05,
+      "loss": 0.644,
+      "step": 2759
+    },
+    {
+      "epoch": 0.636091265268495,
+      "grad_norm": 0.17630359530448914,
+      "learning_rate": 3.090032103458159e-05,
+      "loss": 0.6616,
+      "step": 2760
+    },
+    {
+      "epoch": 0.63632173311823,
+      "grad_norm": 0.1749248057603836,
+      "learning_rate": 3.086582838174551e-05,
+      "loss": 0.6506,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6365522009679649,
+      "grad_norm": 0.16957643628120422,
+      "learning_rate": 3.0831346393837445e-05,
+      "loss": 0.6552,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6367826688177,
+      "grad_norm": 0.16888821125030518,
+      "learning_rate": 3.079687509007682e-05,
+      "loss": 0.6457,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6370131366674349,
+      "grad_norm": 0.17261339724063873,
+      "learning_rate": 3.076241448967709e-05,
+      "loss": 0.6568,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6372436045171699,
+      "grad_norm": 0.18425433337688446,
+      "learning_rate": 3.072796461184579e-05,
+      "loss": 0.6558,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6374740723669048,
+      "grad_norm": 0.1994054913520813,
+      "learning_rate": 3.069352547578445e-05,
+      "loss": 0.6559,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6377045402166398,
+      "grad_norm": 0.17605717480182648,
+      "learning_rate": 3.0659097100688607e-05,
+      "loss": 0.6496,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6379350080663747,
+      "grad_norm": 0.18694092333316803,
+      "learning_rate": 3.062467950574781e-05,
+      "loss": 0.6579,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6381654759161097,
+      "grad_norm": 0.18670520186424255,
+      "learning_rate": 3.059027271014564e-05,
+      "loss": 0.6445,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6383959437658446,
+      "grad_norm": 0.18028950691223145,
+      "learning_rate": 3.055587673305955e-05,
+      "loss": 0.6488,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6386264116155796,
+      "grad_norm": 0.18050305545330048,
+      "learning_rate": 3.052149159366109e-05,
+      "loss": 0.6566,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6388568794653146,
+      "grad_norm": 0.18738004565238953,
+      "learning_rate": 3.0487117311115687e-05,
+      "loss": 0.6465,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6390873473150496,
+      "grad_norm": 0.1994061917066574,
+      "learning_rate": 3.0452753904582747e-05,
+      "loss": 0.65,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6393178151647845,
+      "grad_norm": 0.1728476881980896,
+      "learning_rate": 3.04184013932156e-05,
+      "loss": 0.6544,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6395482830145195,
+      "grad_norm": 0.20216725766658783,
+      "learning_rate": 3.0384059796161535e-05,
+      "loss": 0.6532,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6397787508642544,
+      "grad_norm": 0.17996247112751007,
+      "learning_rate": 3.03497291325617e-05,
+      "loss": 0.6548,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6400092187139894,
+      "grad_norm": 0.20604853332042694,
+      "learning_rate": 3.03154094215512e-05,
+      "loss": 0.6472,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6402396865637243,
+      "grad_norm": 0.1871616393327713,
+      "learning_rate": 3.0281100682259032e-05,
+      "loss": 0.6537,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6404701544134593,
+      "grad_norm": 0.18597093224525452,
+      "learning_rate": 3.024680293380804e-05,
+      "loss": 0.6539,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6407006222631942,
+      "grad_norm": 0.2045731246471405,
+      "learning_rate": 3.0212516195314955e-05,
+      "loss": 0.6574,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6409310901129293,
+      "grad_norm": 0.19084343314170837,
+      "learning_rate": 3.017824048589044e-05,
+      "loss": 0.6615,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6411615579626642,
+      "grad_norm": 0.17811843752861023,
+      "learning_rate": 3.0143975824638887e-05,
+      "loss": 0.6502,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6413920258123992,
+      "grad_norm": 0.19356852769851685,
+      "learning_rate": 3.0109722230658633e-05,
+      "loss": 0.6544,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6416224936621341,
+      "grad_norm": 0.18267957866191864,
+      "learning_rate": 3.0075479723041804e-05,
+      "loss": 0.6531,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6418529615118691,
+      "grad_norm": 0.18512317538261414,
+      "learning_rate": 3.0041248320874343e-05,
+      "loss": 0.6616,
+      "step": 2785
+    },
+    {
+      "epoch": 0.642083429361604,
+      "grad_norm": 0.17581506073474884,
+      "learning_rate": 3.0007028043236013e-05,
+      "loss": 0.6548,
+      "step": 2786
+    },
+    {
+      "epoch": 0.642313897211339,
+      "grad_norm": 0.17305217683315277,
+      "learning_rate": 2.9972818909200396e-05,
+      "loss": 0.6558,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6425443650610739,
+      "grad_norm": 0.1679084450006485,
+      "learning_rate": 2.9938620937834804e-05,
+      "loss": 0.6515,
+      "step": 2788
+    },
+    {
+      "epoch": 0.642774832910809,
+      "grad_norm": 0.16354484856128693,
+      "learning_rate": 2.9904434148200395e-05,
+      "loss": 0.6591,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6430053007605439,
+      "grad_norm": 0.171888068318367,
+      "learning_rate": 2.9870258559352067e-05,
+      "loss": 0.6477,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6432357686102789,
+      "grad_norm": 0.1731024980545044,
+      "learning_rate": 2.9836094190338448e-05,
+      "loss": 0.6491,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6434662364600138,
+      "grad_norm": 0.1673286110162735,
+      "learning_rate": 2.9801941060201944e-05,
+      "loss": 0.6474,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6436967043097488,
+      "grad_norm": 0.18038569390773773,
+      "learning_rate": 2.976779918797873e-05,
+      "loss": 0.6492,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6439271721594838,
+      "grad_norm": 0.1763104796409607,
+      "learning_rate": 2.973366859269861e-05,
+      "loss": 0.6451,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6441576400092187,
+      "grad_norm": 0.17095611989498138,
+      "learning_rate": 2.9699549293385176e-05,
+      "loss": 0.6561,
+      "step": 2795
+    },
+    {
+      "epoch": 0.6443881078589537,
+      "grad_norm": 0.16650767624378204,
+      "learning_rate": 2.966544130905574e-05,
+      "loss": 0.665,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6446185757086886,
+      "grad_norm": 0.1730603128671646,
+      "learning_rate": 2.9631344658721215e-05,
+      "loss": 0.6566,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6448490435584237,
+      "grad_norm": 0.17055529356002808,
+      "learning_rate": 2.9597259361386305e-05,
+      "loss": 0.6559,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6450795114081586,
+      "grad_norm": 0.246446430683136,
+      "learning_rate": 2.9563185436049323e-05,
+      "loss": 0.659,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6453099792578936,
+      "grad_norm": 0.16206693649291992,
+      "learning_rate": 2.9529122901702244e-05,
+      "loss": 0.6473,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6455404471076285,
+      "grad_norm": 0.17527292668819427,
+      "learning_rate": 2.9495071777330717e-05,
+      "loss": 0.6558,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6457709149573635,
+      "grad_norm": 0.16923247277736664,
+      "learning_rate": 2.946103208191403e-05,
+      "loss": 0.6567,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6460013828070984,
+      "grad_norm": 0.18780246376991272,
+      "learning_rate": 2.9427003834425075e-05,
+      "loss": 0.6511,
+      "step": 2803
+    },
+    {
+      "epoch": 0.6462318506568334,
+      "grad_norm": 0.18644435703754425,
+      "learning_rate": 2.939298705383039e-05,
+      "loss": 0.6495,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6464623185065683,
+      "grad_norm": 0.1641910821199417,
+      "learning_rate": 2.935898175909012e-05,
+      "loss": 0.65,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6466927863563033,
+      "grad_norm": 0.18704113364219666,
+      "learning_rate": 2.9324987969157974e-05,
+      "loss": 0.6479,
+      "step": 2806
+    },
+    {
+      "epoch": 0.6469232542060382,
+      "grad_norm": 0.18336251378059387,
+      "learning_rate": 2.9291005702981288e-05,
+      "loss": 0.6506,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6471537220557733,
+      "grad_norm": 0.18807761371135712,
+      "learning_rate": 2.9257034979500986e-05,
+      "loss": 0.6477,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6473841899055082,
+      "grad_norm": 0.18197950720787048,
+      "learning_rate": 2.922307581765149e-05,
+      "loss": 0.6592,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6476146577552432,
+      "grad_norm": 0.20037148892879486,
+      "learning_rate": 2.9189128236360852e-05,
+      "loss": 0.6513,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6478451256049781,
+      "grad_norm": 0.17539845407009125,
+      "learning_rate": 2.915519225455065e-05,
+      "loss": 0.646,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6480755934547131,
+      "grad_norm": 0.19264990091323853,
+      "learning_rate": 2.9121267891135952e-05,
+      "loss": 0.6494,
+      "step": 2812
+    },
+    {
+      "epoch": 0.648306061304448,
+      "grad_norm": 0.1686212420463562,
+      "learning_rate": 2.908735516502537e-05,
+      "loss": 0.6511,
+      "step": 2813
+    },
+    {
+      "epoch": 0.648536529154183,
+      "grad_norm": 0.1705242097377777,
+      "learning_rate": 2.905345409512112e-05,
+      "loss": 0.6491,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6487669970039179,
+      "grad_norm": 0.1819016933441162,
+      "learning_rate": 2.9019564700318768e-05,
+      "loss": 0.6421,
+      "step": 2815
+    },
+    {
+      "epoch": 0.648997464853653,
+      "grad_norm": 0.17861327528953552,
+      "learning_rate": 2.8985686999507488e-05,
+      "loss": 0.6532,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6492279327033879,
+      "grad_norm": 0.2011132538318634,
+      "learning_rate": 2.89518210115699e-05,
+      "loss": 0.6426,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6494584005531229,
+      "grad_norm": 0.16986913979053497,
+      "learning_rate": 2.8917966755382048e-05,
+      "loss": 0.6544,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6496888684028578,
+      "grad_norm": 0.19294098019599915,
+      "learning_rate": 2.8884124249813526e-05,
+      "loss": 0.655,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6499193362525928,
+      "grad_norm": 0.18572697043418884,
+      "learning_rate": 2.885029351372735e-05,
+      "loss": 0.6563,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6501498041023277,
+      "grad_norm": 0.19467751681804657,
+      "learning_rate": 2.881647456597991e-05,
+      "loss": 0.6606,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6503802719520627,
+      "grad_norm": 0.18023499846458435,
+      "learning_rate": 2.8782667425421096e-05,
+      "loss": 0.6527,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6506107398017976,
+      "grad_norm": 0.19630929827690125,
+      "learning_rate": 2.87488721108942e-05,
+      "loss": 0.6475,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6508412076515326,
+      "grad_norm": 0.17777912318706512,
+      "learning_rate": 2.871508864123592e-05,
+      "loss": 0.6543,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6510716755012675,
+      "grad_norm": 0.20080679655075073,
+      "learning_rate": 2.868131703527635e-05,
+      "loss": 0.6556,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6513021433510026,
+      "grad_norm": 0.1675284504890442,
+      "learning_rate": 2.864755731183899e-05,
+      "loss": 0.6482,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6515326112007375,
+      "grad_norm": 0.1921423077583313,
+      "learning_rate": 2.8613809489740662e-05,
+      "loss": 0.6511,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6517630790504725,
+      "grad_norm": 0.17145875096321106,
+      "learning_rate": 2.8580073587791596e-05,
+      "loss": 0.6484,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6519935469002074,
+      "grad_norm": 0.19045400619506836,
+      "learning_rate": 2.8546349624795404e-05,
+      "loss": 0.6555,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6522240147499424,
+      "grad_norm": 0.18223097920417786,
+      "learning_rate": 2.8512637619549022e-05,
+      "loss": 0.643,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6524544825996773,
+      "grad_norm": 0.17837011814117432,
+      "learning_rate": 2.847893759084267e-05,
+      "loss": 0.653,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6526849504494123,
+      "grad_norm": 0.17700053751468658,
+      "learning_rate": 2.8445249557459953e-05,
+      "loss": 0.6537,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6529154182991472,
+      "grad_norm": 0.17131763696670532,
+      "learning_rate": 2.8411573538177772e-05,
+      "loss": 0.6515,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6531458861488822,
+      "grad_norm": 0.1803225427865982,
+      "learning_rate": 2.837790955176634e-05,
+      "loss": 0.6526,
+      "step": 2834
+    },
+    {
+      "epoch": 0.6533763539986172,
+      "grad_norm": 0.17444370687007904,
+      "learning_rate": 2.8344257616989144e-05,
+      "loss": 0.6564,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6536068218483522,
+      "grad_norm": 0.18952690064907074,
+      "learning_rate": 2.8310617752602996e-05,
+      "loss": 0.6535,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6538372896980871,
+      "grad_norm": 0.16605597734451294,
+      "learning_rate": 2.8276989977357894e-05,
+      "loss": 0.6439,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6540677575478221,
+      "grad_norm": 0.19188542664051056,
+      "learning_rate": 2.8243374309997166e-05,
+      "loss": 0.654,
+      "step": 2838
+    },
+    {
+      "epoch": 0.654298225397557,
+      "grad_norm": 0.17647671699523926,
+      "learning_rate": 2.8209770769257437e-05,
+      "loss": 0.643,
+      "step": 2839
+    },
+    {
+      "epoch": 0.654528693247292,
+      "grad_norm": 0.19883349537849426,
+      "learning_rate": 2.8176179373868443e-05,
+      "loss": 0.662,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6547591610970269,
+      "grad_norm": 0.17992442846298218,
+      "learning_rate": 2.8142600142553254e-05,
+      "loss": 0.6503,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6549896289467619,
+      "grad_norm": 0.19443680346012115,
+      "learning_rate": 2.8109033094028126e-05,
+      "loss": 0.6492,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6552200967964968,
+      "grad_norm": 0.18997105956077576,
+      "learning_rate": 2.8075478247002518e-05,
+      "loss": 0.6544,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6554505646462319,
+      "grad_norm": 0.2033330202102661,
+      "learning_rate": 2.8041935620179105e-05,
+      "loss": 0.6607,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6556810324959668,
+      "grad_norm": 0.18512104451656342,
+      "learning_rate": 2.8008405232253752e-05,
+      "loss": 0.6491,
+      "step": 2845
+    },
+    {
+      "epoch": 0.6559115003457018,
+      "grad_norm": 0.1780814528465271,
+      "learning_rate": 2.7974887101915458e-05,
+      "loss": 0.6497,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6561419681954367,
+      "grad_norm": 0.20027370750904083,
+      "learning_rate": 2.7941381247846453e-05,
+      "loss": 0.6509,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6563724360451717,
+      "grad_norm": 0.1773580014705658,
+      "learning_rate": 2.7907887688722085e-05,
+      "loss": 0.6508,
+      "step": 2848
+    },
+    {
+      "epoch": 0.6566029038949066,
+      "grad_norm": 0.20906241238117218,
+      "learning_rate": 2.7874406443210866e-05,
+      "loss": 0.6531,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6568333717446416,
+      "grad_norm": 0.17152054607868195,
+      "learning_rate": 2.784093752997443e-05,
+      "loss": 0.6519,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6570638395943765,
+      "grad_norm": 0.20871195197105408,
+      "learning_rate": 2.7807480967667576e-05,
+      "loss": 0.6517,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6572943074441115,
+      "grad_norm": 0.18561682105064392,
+      "learning_rate": 2.7774036774938138e-05,
+      "loss": 0.6447,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6575247752938465,
+      "grad_norm": 0.19581186771392822,
+      "learning_rate": 2.774060497042712e-05,
+      "loss": 0.6593,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6577552431435815,
+      "grad_norm": 0.1915295422077179,
+      "learning_rate": 2.7707185572768656e-05,
+      "loss": 0.6545,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6579857109933165,
+      "grad_norm": 0.19746430218219757,
+      "learning_rate": 2.7673778600589862e-05,
+      "loss": 0.6521,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6582161788430514,
+      "grad_norm": 0.17735208570957184,
+      "learning_rate": 2.7640384072510994e-05,
+      "loss": 0.6554,
+      "step": 2856
+    },
+    {
+      "epoch": 0.6584466466927864,
+      "grad_norm": 0.19774432480335236,
+      "learning_rate": 2.7607002007145377e-05,
+      "loss": 0.6499,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6586771145425213,
+      "grad_norm": 0.17123498022556305,
+      "learning_rate": 2.7573632423099355e-05,
+      "loss": 0.6539,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6589075823922563,
+      "grad_norm": 0.19462165236473083,
+      "learning_rate": 2.7540275338972343e-05,
+      "loss": 0.6533,
+      "step": 2859
+    },
+    {
+      "epoch": 0.6591380502419912,
+      "grad_norm": 0.1761290431022644,
+      "learning_rate": 2.75069307733568e-05,
+      "loss": 0.6443,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6593685180917263,
+      "grad_norm": 0.17915204167366028,
+      "learning_rate": 2.7473598744838146e-05,
+      "loss": 0.6526,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6595989859414612,
+      "grad_norm": 0.170929417014122,
+      "learning_rate": 2.744027927199486e-05,
+      "loss": 0.6531,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6598294537911962,
+      "grad_norm": 0.18236766755580902,
+      "learning_rate": 2.7406972373398443e-05,
+      "loss": 0.6508,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6600599216409311,
+      "grad_norm": 0.16858653724193573,
+      "learning_rate": 2.737367806761334e-05,
+      "loss": 0.6535,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6602903894906661,
+      "grad_norm": 0.1613864302635193,
+      "learning_rate": 2.7340396373196996e-05,
+      "loss": 0.6503,
+      "step": 2865
+    },
+    {
+      "epoch": 0.660520857340401,
+      "grad_norm": 0.17449095845222473,
+      "learning_rate": 2.7307127308699865e-05,
+      "loss": 0.6549,
+      "step": 2866
+    },
+    {
+      "epoch": 0.660751325190136,
+      "grad_norm": 0.16756795346736908,
+      "learning_rate": 2.7273870892665253e-05,
+      "loss": 0.6488,
+      "step": 2867
+    },
+    {
+      "epoch": 0.6609817930398709,
+      "grad_norm": 0.16483251750469208,
+      "learning_rate": 2.724062714362955e-05,
+      "loss": 0.6564,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6612122608896059,
+      "grad_norm": 0.16283339262008667,
+      "learning_rate": 2.7207396080122028e-05,
+      "loss": 0.655,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6614427287393408,
+      "grad_norm": 0.16669294238090515,
+      "learning_rate": 2.717417772066484e-05,
+      "loss": 0.6508,
+      "step": 2870
+    },
+    {
+      "epoch": 0.6616731965890759,
+      "grad_norm": 0.18657532334327698,
+      "learning_rate": 2.7140972083773124e-05,
+      "loss": 0.6463,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6619036644388108,
+      "grad_norm": 0.17012335360050201,
+      "learning_rate": 2.7107779187954908e-05,
+      "loss": 0.6552,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6621341322885458,
+      "grad_norm": 0.17444707453250885,
+      "learning_rate": 2.7074599051711108e-05,
+      "loss": 0.6454,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6623646001382807,
+      "grad_norm": 0.17587541043758392,
+      "learning_rate": 2.704143169353554e-05,
+      "loss": 0.6565,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6625950679880157,
+      "grad_norm": 0.16666634380817413,
+      "learning_rate": 2.7008277131914916e-05,
+      "loss": 0.6483,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6628255358377506,
+      "grad_norm": 0.16995498538017273,
+      "learning_rate": 2.6975135385328743e-05,
+      "loss": 0.6459,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6630560036874856,
+      "grad_norm": 0.17786766588687897,
+      "learning_rate": 2.6942006472249447e-05,
+      "loss": 0.6482,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6632864715372205,
+      "grad_norm": 0.15976811945438385,
+      "learning_rate": 2.690889041114234e-05,
+      "loss": 0.6426,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6635169393869556,
+      "grad_norm": 0.17242498695850372,
+      "learning_rate": 2.6875787220465463e-05,
+      "loss": 0.6514,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6637474072366905,
+      "grad_norm": 0.18254664540290833,
+      "learning_rate": 2.6842696918669768e-05,
+      "loss": 0.6541,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6639778750864255,
+      "grad_norm": 0.18394513428211212,
+      "learning_rate": 2.6809619524199004e-05,
+      "loss": 0.6491,
+      "step": 2881
+    },
+    {
+      "epoch": 0.6642083429361604,
+      "grad_norm": 0.17171119153499603,
+      "learning_rate": 2.677655505548966e-05,
+      "loss": 0.6435,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6644388107858954,
+      "grad_norm": 0.17666159570217133,
+      "learning_rate": 2.6743503530971138e-05,
+      "loss": 0.6427,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6646692786356303,
+      "grad_norm": 0.18384835124015808,
+      "learning_rate": 2.6710464969065563e-05,
+      "loss": 0.6508,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6648997464853653,
+      "grad_norm": 0.16794157028198242,
+      "learning_rate": 2.6677439388187796e-05,
+      "loss": 0.6547,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6651302143351002,
+      "grad_norm": 0.17705510556697845,
+      "learning_rate": 2.6644426806745526e-05,
+      "loss": 0.6458,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6653606821848352,
+      "grad_norm": 0.17769016325473785,
+      "learning_rate": 2.6611427243139166e-05,
+      "loss": 0.6518,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6655911500345701,
+      "grad_norm": 0.16915744543075562,
+      "learning_rate": 2.6578440715761894e-05,
+      "loss": 0.65,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6658216178843052,
+      "grad_norm": 0.17128948867321014,
+      "learning_rate": 2.65454672429996e-05,
+      "loss": 0.6476,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6660520857340401,
+      "grad_norm": 0.16499844193458557,
+      "learning_rate": 2.6512506843230922e-05,
+      "loss": 0.661,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6662825535837751,
+      "grad_norm": 0.1791979968547821,
+      "learning_rate": 2.6479559534827168e-05,
+      "loss": 0.6513,
+      "step": 2891
+    },
+    {
+      "epoch": 0.66651302143351,
+      "grad_norm": 0.16432271897792816,
+      "learning_rate": 2.6446625336152364e-05,
+      "loss": 0.6536,
+      "step": 2892
+    },
+    {
+      "epoch": 0.666743489283245,
+      "grad_norm": 0.17049454152584076,
+      "learning_rate": 2.641370426556331e-05,
+      "loss": 0.6504,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6669739571329799,
+      "grad_norm": 0.17584246397018433,
+      "learning_rate": 2.6380796341409364e-05,
+      "loss": 0.6444,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6672044249827149,
+      "grad_norm": 0.1704198718070984,
+      "learning_rate": 2.6347901582032623e-05,
+      "loss": 0.6566,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6674348928324498,
+      "grad_norm": 0.1673116832971573,
+      "learning_rate": 2.6315020005767843e-05,
+      "loss": 0.6508,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6676653606821848,
+      "grad_norm": 0.16023887693881989,
+      "learning_rate": 2.628215163094242e-05,
+      "loss": 0.6522,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6678958285319198,
+      "grad_norm": 0.16509434580802917,
+      "learning_rate": 2.6249296475876407e-05,
+      "loss": 0.6573,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6681262963816548,
+      "grad_norm": 0.17283135652542114,
+      "learning_rate": 2.6216454558882486e-05,
+      "loss": 0.6567,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6683567642313897,
+      "grad_norm": 0.15373359620571136,
+      "learning_rate": 2.6183625898265918e-05,
+      "loss": 0.6509,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6685872320811247,
+      "grad_norm": 0.17244744300842285,
+      "learning_rate": 2.6150810512324637e-05,
+      "loss": 0.6472,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6688176999308596,
+      "grad_norm": 0.17626379430294037,
+      "learning_rate": 2.6118008419349148e-05,
+      "loss": 0.6575,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6690481677805946,
+      "grad_norm": 0.16538438200950623,
+      "learning_rate": 2.6085219637622544e-05,
+      "loss": 0.6507,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6692786356303295,
+      "grad_norm": 0.18583250045776367,
+      "learning_rate": 2.605244418542051e-05,
+      "loss": 0.6585,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6695091034800645,
+      "grad_norm": 0.17038129270076752,
+      "learning_rate": 2.601968208101132e-05,
+      "loss": 0.6591,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6697395713297994,
+      "grad_norm": 0.17771609127521515,
+      "learning_rate": 2.5986933342655717e-05,
+      "loss": 0.6568,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6699700391795345,
+      "grad_norm": 0.17149491608142853,
+      "learning_rate": 2.5954197988607133e-05,
+      "loss": 0.6535,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6702005070292694,
+      "grad_norm": 0.18247468769550323,
+      "learning_rate": 2.5921476037111448e-05,
+      "loss": 0.6533,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6704309748790044,
+      "grad_norm": 0.16240952908992767,
+      "learning_rate": 2.5888767506407075e-05,
+      "loss": 0.6556,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6706614427287393,
+      "grad_norm": 0.18257564306259155,
+      "learning_rate": 2.5856072414724974e-05,
+      "loss": 0.6539,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6708919105784743,
+      "grad_norm": 0.16340118646621704,
+      "learning_rate": 2.5823390780288604e-05,
+      "loss": 0.6605,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6711223784282092,
+      "grad_norm": 0.18919028341770172,
+      "learning_rate": 2.5790722621313924e-05,
+      "loss": 0.6558,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6713528462779442,
+      "grad_norm": 0.1639140248298645,
+      "learning_rate": 2.575806795600938e-05,
+      "loss": 0.6472,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6715833141276792,
+      "grad_norm": 0.18610194325447083,
+      "learning_rate": 2.5725426802575925e-05,
+      "loss": 0.6476,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6718137819774141,
+      "grad_norm": 0.18196745216846466,
+      "learning_rate": 2.5692799179206906e-05,
+      "loss": 0.6472,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6720442498271492,
+      "grad_norm": 0.1704276204109192,
+      "learning_rate": 2.5660185104088207e-05,
+      "loss": 0.6442,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6722747176768841,
+      "grad_norm": 0.1694510579109192,
+      "learning_rate": 2.5627584595398125e-05,
+      "loss": 0.6528,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6725051855266191,
+      "grad_norm": 0.16790097951889038,
+      "learning_rate": 2.5594997671307397e-05,
+      "loss": 0.6555,
+      "step": 2918
+    },
+    {
+      "epoch": 0.672735653376354,
+      "grad_norm": 0.17335295677185059,
+      "learning_rate": 2.5562424349979198e-05,
+      "loss": 0.6553,
+      "step": 2919
+    },
+    {
+      "epoch": 0.672966121226089,
+      "grad_norm": 0.1660207062959671,
+      "learning_rate": 2.552986464956913e-05,
+      "loss": 0.648,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6731965890758239,
+      "grad_norm": 0.1704881191253662,
+      "learning_rate": 2.5497318588225137e-05,
+      "loss": 0.6468,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6734270569255589,
+      "grad_norm": 0.17709209024906158,
+      "learning_rate": 2.5464786184087665e-05,
+      "loss": 0.6567,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6736575247752938,
+      "grad_norm": 0.16337838768959045,
+      "learning_rate": 2.5432267455289503e-05,
+      "loss": 0.6459,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6738879926250289,
+      "grad_norm": 0.1824062317609787,
+      "learning_rate": 2.5399762419955764e-05,
+      "loss": 0.6545,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6741184604747638,
+      "grad_norm": 0.16532284021377563,
+      "learning_rate": 2.5367271096203994e-05,
+      "loss": 0.6471,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6743489283244988,
+      "grad_norm": 0.18639203906059265,
+      "learning_rate": 2.5334793502144077e-05,
+      "loss": 0.6585,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6745793961742337,
+      "grad_norm": 0.1666010320186615,
+      "learning_rate": 2.5302329655878244e-05,
+      "loss": 0.656,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6748098640239687,
+      "grad_norm": 0.17926305532455444,
+      "learning_rate": 2.5269879575501072e-05,
+      "loss": 0.6496,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6750403318737036,
+      "grad_norm": 0.1780381202697754,
+      "learning_rate": 2.523744327909947e-05,
+      "loss": 0.6528,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6752707997234386,
+      "grad_norm": 0.17732974886894226,
+      "learning_rate": 2.520502078475261e-05,
+      "loss": 0.6516,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6755012675731735,
+      "grad_norm": 0.17732085287570953,
+      "learning_rate": 2.5172612110532012e-05,
+      "loss": 0.6591,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6757317354229085,
+      "grad_norm": 0.18524208664894104,
+      "learning_rate": 2.5140217274501555e-05,
+      "loss": 0.648,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6759622032726434,
+      "grad_norm": 0.18076317012310028,
+      "learning_rate": 2.510783629471728e-05,
+      "loss": 0.6565,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6761926711223785,
+      "grad_norm": 0.1893717348575592,
+      "learning_rate": 2.5075469189227597e-05,
+      "loss": 0.6421,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6764231389721134,
+      "grad_norm": 0.17758077383041382,
+      "learning_rate": 2.5043115976073167e-05,
+      "loss": 0.646,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6766536068218484,
+      "grad_norm": 0.17987452447414398,
+      "learning_rate": 2.5010776673286834e-05,
+      "loss": 0.6401,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6768840746715833,
+      "grad_norm": 0.1678553968667984,
+      "learning_rate": 2.497845129889381e-05,
+      "loss": 0.647,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6771145425213183,
+      "grad_norm": 0.1824767142534256,
+      "learning_rate": 2.4946139870911482e-05,
+      "loss": 0.6409,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6773450103710532,
+      "grad_norm": 0.1831919103860855,
+      "learning_rate": 2.491384240734943e-05,
+      "loss": 0.6454,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6775754782207882,
+      "grad_norm": 0.17052872478961945,
+      "learning_rate": 2.48815589262095e-05,
+      "loss": 0.6512,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6778059460705231,
+      "grad_norm": 0.17829042673110962,
+      "learning_rate": 2.4849289445485723e-05,
+      "loss": 0.6466,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6780364139202582,
+      "grad_norm": 0.170830637216568,
+      "learning_rate": 2.4817033983164345e-05,
+      "loss": 0.6609,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6782668817699931,
+      "grad_norm": 0.16118223965168,
+      "learning_rate": 2.478479255722378e-05,
+      "loss": 0.6512,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6784973496197281,
+      "grad_norm": 0.1625395268201828,
+      "learning_rate": 2.4752565185634645e-05,
+      "loss": 0.6502,
+      "step": 2944
+    },
+    {
+      "epoch": 0.678727817469463,
+      "grad_norm": 0.17522037029266357,
+      "learning_rate": 2.472035188635967e-05,
+      "loss": 0.6458,
+      "step": 2945
+    },
+    {
+      "epoch": 0.678958285319198,
+      "grad_norm": 0.16705302894115448,
+      "learning_rate": 2.468815267735377e-05,
+      "loss": 0.6545,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6791887531689329,
+      "grad_norm": 0.1653483510017395,
+      "learning_rate": 2.4655967576564064e-05,
+      "loss": 0.6473,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6794192210186679,
+      "grad_norm": 0.16879500448703766,
+      "learning_rate": 2.4623796601929712e-05,
+      "loss": 0.652,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6796496888684028,
+      "grad_norm": 0.16743235290050507,
+      "learning_rate": 2.4591639771382043e-05,
+      "loss": 0.6524,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6798801567181378,
+      "grad_norm": 0.17148716747760773,
+      "learning_rate": 2.4559497102844514e-05,
+      "loss": 0.6468,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6801106245678727,
+      "grad_norm": 0.15919731557369232,
+      "learning_rate": 2.452736861423268e-05,
+      "loss": 0.6566,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6803410924176078,
+      "grad_norm": 0.17846500873565674,
+      "learning_rate": 2.449525432345418e-05,
+      "loss": 0.6603,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6805715602673427,
+      "grad_norm": 0.16970974206924438,
+      "learning_rate": 2.4463154248408777e-05,
+      "loss": 0.6434,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6808020281170777,
+      "grad_norm": 0.1619907021522522,
+      "learning_rate": 2.4431068406988238e-05,
+      "loss": 0.6526,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6810324959668126,
+      "grad_norm": 0.1664748191833496,
+      "learning_rate": 2.439899681707646e-05,
+      "loss": 0.648,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6812629638165476,
+      "grad_norm": 0.16397912800312042,
+      "learning_rate": 2.4366939496549378e-05,
+      "loss": 0.6482,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6814934316662825,
+      "grad_norm": 0.16301095485687256,
+      "learning_rate": 2.4334896463274976e-05,
+      "loss": 0.6452,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6817238995160175,
+      "grad_norm": 0.16809646785259247,
+      "learning_rate": 2.430286773511327e-05,
+      "loss": 0.6518,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6819543673657524,
+      "grad_norm": 0.15831275284290314,
+      "learning_rate": 2.4270853329916304e-05,
+      "loss": 0.6463,
+      "step": 2959
+    },
+    {
+      "epoch": 0.6821848352154875,
+      "grad_norm": 0.1629604548215866,
+      "learning_rate": 2.4238853265528143e-05,
+      "loss": 0.6586,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6824153030652224,
+      "grad_norm": 0.1589616984128952,
+      "learning_rate": 2.4206867559784847e-05,
+      "loss": 0.6457,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6826457709149574,
+      "grad_norm": 0.17117534577846527,
+      "learning_rate": 2.417489623051448e-05,
+      "loss": 0.6551,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6828762387646923,
+      "grad_norm": 0.1638975292444229,
+      "learning_rate": 2.4142939295537126e-05,
+      "loss": 0.649,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6831067066144273,
+      "grad_norm": 0.1592676341533661,
+      "learning_rate": 2.411099677266476e-05,
+      "loss": 0.6549,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6833371744641622,
+      "grad_norm": 0.1702510565519333,
+      "learning_rate": 2.407906867970141e-05,
+      "loss": 0.644,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6835676423138972,
+      "grad_norm": 0.17177124321460724,
+      "learning_rate": 2.404715503444302e-05,
+      "loss": 0.6515,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6837981101636321,
+      "grad_norm": 0.17113865911960602,
+      "learning_rate": 2.4015255854677488e-05,
+      "loss": 0.6572,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6840285780133671,
+      "grad_norm": 0.17294318974018097,
+      "learning_rate": 2.398337115818466e-05,
+      "loss": 0.6504,
+      "step": 2968
+    },
+    {
+      "epoch": 0.684259045863102,
+      "grad_norm": 0.16647367179393768,
+      "learning_rate": 2.3951500962736312e-05,
+      "loss": 0.6593,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6844895137128371,
+      "grad_norm": 0.16758325695991516,
+      "learning_rate": 2.391964528609609e-05,
+      "loss": 0.6422,
+      "step": 2970
+    },
+    {
+      "epoch": 0.684719981562572,
+      "grad_norm": 0.18237681686878204,
+      "learning_rate": 2.388780414601959e-05,
+      "loss": 0.6479,
+      "step": 2971
+    },
+    {
+      "epoch": 0.684950449412307,
+      "grad_norm": 0.16424010694026947,
+      "learning_rate": 2.3855977560254338e-05,
+      "loss": 0.6441,
+      "step": 2972
+    },
+    {
+      "epoch": 0.685180917262042,
+      "grad_norm": 0.1714371293783188,
+      "learning_rate": 2.3824165546539673e-05,
+      "loss": 0.6534,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6854113851117769,
+      "grad_norm": 0.17393812537193298,
+      "learning_rate": 2.3792368122606856e-05,
+      "loss": 0.65,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6856418529615119,
+      "grad_norm": 0.18081031739711761,
+      "learning_rate": 2.3760585306179012e-05,
+      "loss": 0.6493,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6858723208112468,
+      "grad_norm": 0.16579000651836395,
+      "learning_rate": 2.3728817114971118e-05,
+      "loss": 0.6448,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6861027886609818,
+      "grad_norm": 0.17078308761119843,
+      "learning_rate": 2.3697063566689998e-05,
+      "loss": 0.6526,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6863332565107167,
+      "grad_norm": 0.16518385708332062,
+      "learning_rate": 2.3665324679034344e-05,
+      "loss": 0.6374,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6865637243604518,
+      "grad_norm": 0.16557030379772186,
+      "learning_rate": 2.3633600469694606e-05,
+      "loss": 0.6436,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6867941922101867,
+      "grad_norm": 0.17224299907684326,
+      "learning_rate": 2.3601890956353118e-05,
+      "loss": 0.6588,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6870246600599217,
+      "grad_norm": 0.17414404451847076,
+      "learning_rate": 2.3570196156684005e-05,
+      "loss": 0.6473,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6872551279096566,
+      "grad_norm": 0.16957977414131165,
+      "learning_rate": 2.353851608835318e-05,
+      "loss": 0.6469,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6874855957593916,
+      "grad_norm": 0.16680394113063812,
+      "learning_rate": 2.3506850769018363e-05,
+      "loss": 0.648,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6877160636091265,
+      "grad_norm": 0.1786261647939682,
+      "learning_rate": 2.3475200216329052e-05,
+      "loss": 0.6516,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6879465314588615,
+      "grad_norm": 0.17838843166828156,
+      "learning_rate": 2.3443564447926447e-05,
+      "loss": 0.6538,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6881769993085964,
+      "grad_norm": 0.16465331614017487,
+      "learning_rate": 2.3411943481443637e-05,
+      "loss": 0.6512,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6884074671583315,
+      "grad_norm": 0.16617000102996826,
+      "learning_rate": 2.3380337334505374e-05,
+      "loss": 0.6529,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6886379350080664,
+      "grad_norm": 0.16464616358280182,
+      "learning_rate": 2.3348746024728142e-05,
+      "loss": 0.6582,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6888684028578014,
+      "grad_norm": 0.1769178956747055,
+      "learning_rate": 2.3317169569720187e-05,
+      "loss": 0.6525,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6890988707075363,
+      "grad_norm": 0.16962438821792603,
+      "learning_rate": 2.328560798708147e-05,
+      "loss": 0.65,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6893293385572713,
+      "grad_norm": 0.16247595846652985,
+      "learning_rate": 2.3254061294403663e-05,
+      "loss": 0.6484,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6895598064070062,
+      "grad_norm": 0.17956919968128204,
+      "learning_rate": 2.322252950927013e-05,
+      "loss": 0.6464,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6897902742567412,
+      "grad_norm": 0.17907127737998962,
+      "learning_rate": 2.3191012649255956e-05,
+      "loss": 0.653,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6900207421064761,
+      "grad_norm": 0.16521626710891724,
+      "learning_rate": 2.3159510731927842e-05,
+      "loss": 0.6548,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6902512099562111,
+      "grad_norm": 0.18841508030891418,
+      "learning_rate": 2.3128023774844194e-05,
+      "loss": 0.6502,
+      "step": 2995
+    },
+    {
+      "epoch": 0.690481677805946,
+      "grad_norm": 0.17101146280765533,
+      "learning_rate": 2.3096551795555155e-05,
+      "loss": 0.6417,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6907121456556811,
+      "grad_norm": 0.19493061304092407,
+      "learning_rate": 2.3065094811602383e-05,
+      "loss": 0.6547,
+      "step": 2997
+    },
+    {
+      "epoch": 0.690942613505416,
+      "grad_norm": 0.17434021830558777,
+      "learning_rate": 2.3033652840519272e-05,
+      "loss": 0.6511,
+      "step": 2998
+    },
+    {
+      "epoch": 0.691173081355151,
+      "grad_norm": 0.17021694779396057,
+      "learning_rate": 2.300222589983084e-05,
+      "loss": 0.6542,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6914035492048859,
+      "grad_norm": 0.1962118148803711,
+      "learning_rate": 2.297081400705365e-05,
+      "loss": 0.6532,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6916340170546209,
+      "grad_norm": 0.17823426425457,
+      "learning_rate": 2.293941717969599e-05,
+      "loss": 0.6437,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6918644849043558,
+      "grad_norm": 0.1798575520515442,
+      "learning_rate": 2.2908035435257706e-05,
+      "loss": 0.6528,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6920949527540908,
+      "grad_norm": 0.1800197809934616,
+      "learning_rate": 2.2876668791230193e-05,
+      "loss": 0.6527,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6923254206038257,
+      "grad_norm": 0.17531868815422058,
+      "learning_rate": 2.2845317265096474e-05,
+      "loss": 0.651,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6925558884535608,
+      "grad_norm": 0.19876468181610107,
+      "learning_rate": 2.2813980874331146e-05,
+      "loss": 0.6485,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6927863563032957,
+      "grad_norm": 0.16582679748535156,
+      "learning_rate": 2.2782659636400355e-05,
+      "loss": 0.6532,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6930168241530307,
+      "grad_norm": 0.18158891797065735,
+      "learning_rate": 2.2751353568761813e-05,
+      "loss": 0.6509,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6932472920027656,
+      "grad_norm": 0.17845842242240906,
+      "learning_rate": 2.272006268886479e-05,
+      "loss": 0.6502,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6934777598525006,
+      "grad_norm": 0.17071473598480225,
+      "learning_rate": 2.2688787014150027e-05,
+      "loss": 0.6491,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6937082277022355,
+      "grad_norm": 0.17339655756950378,
+      "learning_rate": 2.2657526562049836e-05,
+      "loss": 0.6563,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6939386955519705,
+      "grad_norm": 0.16644999384880066,
+      "learning_rate": 2.2626281349988103e-05,
+      "loss": 0.6527,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6941691634017054,
+      "grad_norm": 0.16771528124809265,
+      "learning_rate": 2.25950513953801e-05,
+      "loss": 0.6455,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6943996312514404,
+      "grad_norm": 0.17318162322044373,
+      "learning_rate": 2.2563836715632676e-05,
+      "loss": 0.6634,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6946300991011753,
+      "grad_norm": 0.15834879875183105,
+      "learning_rate": 2.253263732814414e-05,
+      "loss": 0.6532,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6948605669509104,
+      "grad_norm": 0.1758836805820465,
+      "learning_rate": 2.2501453250304283e-05,
+      "loss": 0.6399,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6950910348006453,
+      "grad_norm": 0.1595585197210312,
+      "learning_rate": 2.2470284499494364e-05,
+      "loss": 0.6525,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6953215026503803,
+      "grad_norm": 0.1615588665008545,
+      "learning_rate": 2.2439131093087113e-05,
+      "loss": 0.643,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6955519705001152,
+      "grad_norm": 0.16813978552818298,
+      "learning_rate": 2.2407993048446656e-05,
+      "loss": 0.6448,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6957824383498502,
+      "grad_norm": 0.16986919939517975,
+      "learning_rate": 2.2376870382928607e-05,
+      "loss": 0.6473,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6960129061995851,
+      "grad_norm": 0.17389391362667084,
+      "learning_rate": 2.2345763113879996e-05,
+      "loss": 0.6468,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6962433740493201,
+      "grad_norm": 0.17148233950138092,
+      "learning_rate": 2.2314671258639263e-05,
+      "loss": 0.6585,
+      "step": 3021
+    },
+    {
+      "epoch": 0.696473841899055,
+      "grad_norm": 0.18123234808444977,
+      "learning_rate": 2.228359483453627e-05,
+      "loss": 0.6487,
+      "step": 3022
+    },
+    {
+      "epoch": 0.69670430974879,
+      "grad_norm": 0.16465289890766144,
+      "learning_rate": 2.2252533858892277e-05,
+      "loss": 0.6497,
+      "step": 3023
+    },
+    {
+      "epoch": 0.696934777598525,
+      "grad_norm": 0.1766558438539505,
+      "learning_rate": 2.2221488349019903e-05,
+      "loss": 0.6505,
+      "step": 3024
+    },
+    {
+      "epoch": 0.69716524544826,
+      "grad_norm": 0.16725115478038788,
+      "learning_rate": 2.2190458322223163e-05,
+      "loss": 0.6475,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6973957132979949,
+      "grad_norm": 0.17288514971733093,
+      "learning_rate": 2.2159443795797497e-05,
+      "loss": 0.6519,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6976261811477299,
+      "grad_norm": 0.165741428732872,
+      "learning_rate": 2.2128444787029618e-05,
+      "loss": 0.6421,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6978566489974648,
+      "grad_norm": 0.17508132755756378,
+      "learning_rate": 2.209746131319764e-05,
+      "loss": 0.6495,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6980871168471998,
+      "grad_norm": 0.1722734570503235,
+      "learning_rate": 2.2066493391570996e-05,
+      "loss": 0.6473,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6983175846969347,
+      "grad_norm": 0.16845974326133728,
+      "learning_rate": 2.2035541039410483e-05,
+      "loss": 0.6449,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6985480525466697,
+      "grad_norm": 0.18813900649547577,
+      "learning_rate": 2.200460427396819e-05,
+      "loss": 0.647,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6987785203964048,
+      "grad_norm": 0.18029405176639557,
+      "learning_rate": 2.1973683112487544e-05,
+      "loss": 0.6461,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6990089882461397,
+      "grad_norm": 0.18963518738746643,
+      "learning_rate": 2.194277757220322e-05,
+      "loss": 0.6387,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6992394560958747,
+      "grad_norm": 0.17914853990077972,
+      "learning_rate": 2.191188767034125e-05,
+      "loss": 0.642,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6994699239456096,
+      "grad_norm": 0.20282970368862152,
+      "learning_rate": 2.1881013424118922e-05,
+      "loss": 0.6574,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6997003917953446,
+      "grad_norm": 0.17879828810691833,
+      "learning_rate": 2.18501548507448e-05,
+      "loss": 0.6485,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6999308596450795,
+      "grad_norm": 0.18236006796360016,
+      "learning_rate": 2.1819311967418697e-05,
+      "loss": 0.6525,
+      "step": 3037
+    },
+    {
+      "epoch": 0.7001613274948145,
+      "grad_norm": 0.18954585492610931,
+      "learning_rate": 2.1788484791331738e-05,
+      "loss": 0.6411,
+      "step": 3038
+    },
+    {
+      "epoch": 0.7003917953445494,
+      "grad_norm": 0.19611108303070068,
+      "learning_rate": 2.1757673339666178e-05,
+      "loss": 0.6474,
+      "step": 3039
+    },
+    {
+      "epoch": 0.7006222631942844,
+      "grad_norm": 0.1711956113576889,
+      "learning_rate": 2.172687762959565e-05,
+      "loss": 0.6458,
+      "step": 3040
+    },
+    {
+      "epoch": 0.7008527310440194,
+      "grad_norm": 0.1677607148885727,
+      "learning_rate": 2.169609767828493e-05,
+      "loss": 0.6419,
+      "step": 3041
+    },
+    {
+      "epoch": 0.7010831988937544,
+      "grad_norm": 0.19762775301933289,
+      "learning_rate": 2.1665333502889994e-05,
+      "loss": 0.6508,
+      "step": 3042
+    },
+    {
+      "epoch": 0.7013136667434893,
+      "grad_norm": 0.1680401712656021,
+      "learning_rate": 2.1634585120558078e-05,
+      "loss": 0.6456,
+      "step": 3043
+    },
+    {
+      "epoch": 0.7015441345932243,
+      "grad_norm": 0.1744004786014557,
+      "learning_rate": 2.1603852548427582e-05,
+      "loss": 0.6496,
+      "step": 3044
+    },
+    {
+      "epoch": 0.7017746024429592,
+      "grad_norm": 0.179367795586586,
+      "learning_rate": 2.1573135803628114e-05,
+      "loss": 0.6438,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7020050702926942,
+      "grad_norm": 0.16695451736450195,
+      "learning_rate": 2.154243490328044e-05,
+      "loss": 0.6508,
+      "step": 3046
+    },
+    {
+      "epoch": 0.7022355381424291,
+      "grad_norm": 0.18172797560691833,
+      "learning_rate": 2.1511749864496534e-05,
+      "loss": 0.6414,
+      "step": 3047
+    },
+    {
+      "epoch": 0.7024660059921641,
+      "grad_norm": 0.17104046046733856,
+      "learning_rate": 2.148108070437945e-05,
+      "loss": 0.649,
+      "step": 3048
+    },
+    {
+      "epoch": 0.702696473841899,
+      "grad_norm": 0.17437295615673065,
+      "learning_rate": 2.1450427440023456e-05,
+      "loss": 0.6541,
+      "step": 3049
+    },
+    {
+      "epoch": 0.702926941691634,
+      "grad_norm": 0.1672581136226654,
+      "learning_rate": 2.1419790088513998e-05,
+      "loss": 0.6469,
+      "step": 3050
+    },
+    {
+      "epoch": 0.703157409541369,
+      "grad_norm": 0.17412829399108887,
+      "learning_rate": 2.138916866692754e-05,
+      "loss": 0.6396,
+      "step": 3051
+    },
+    {
+      "epoch": 0.703387877391104,
+      "grad_norm": 0.1715768724679947,
+      "learning_rate": 2.1358563192331747e-05,
+      "loss": 0.6532,
+      "step": 3052
+    },
+    {
+      "epoch": 0.7036183452408389,
+      "grad_norm": 0.16486719250679016,
+      "learning_rate": 2.1327973681785397e-05,
+      "loss": 0.653,
+      "step": 3053
+    },
+    {
+      "epoch": 0.7038488130905739,
+      "grad_norm": 0.1680760681629181,
+      "learning_rate": 2.1297400152338286e-05,
+      "loss": 0.6564,
+      "step": 3054
+    },
+    {
+      "epoch": 0.7040792809403088,
+      "grad_norm": 0.16449017822742462,
+      "learning_rate": 2.1266842621031434e-05,
+      "loss": 0.6408,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7043097487900438,
+      "grad_norm": 0.1720770299434662,
+      "learning_rate": 2.1236301104896866e-05,
+      "loss": 0.6612,
+      "step": 3056
+    },
+    {
+      "epoch": 0.7045402166397787,
+      "grad_norm": 0.15701450407505035,
+      "learning_rate": 2.1205775620957652e-05,
+      "loss": 0.6421,
+      "step": 3057
+    },
+    {
+      "epoch": 0.7047706844895137,
+      "grad_norm": 0.17185088992118835,
+      "learning_rate": 2.1175266186227987e-05,
+      "loss": 0.6487,
+      "step": 3058
+    },
+    {
+      "epoch": 0.7050011523392486,
+      "grad_norm": 0.18234078586101532,
+      "learning_rate": 2.1144772817713103e-05,
+      "loss": 0.6472,
+      "step": 3059
+    },
+    {
+      "epoch": 0.7052316201889837,
+      "grad_norm": 0.16597682237625122,
+      "learning_rate": 2.1114295532409263e-05,
+      "loss": 0.6511,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7054620880387186,
+      "grad_norm": 0.17775195837020874,
+      "learning_rate": 2.1083834347303772e-05,
+      "loss": 0.6451,
+      "step": 3061
+    },
+    {
+      "epoch": 0.7056925558884536,
+      "grad_norm": 0.1636880785226822,
+      "learning_rate": 2.1053389279374987e-05,
+      "loss": 0.6505,
+      "step": 3062
+    },
+    {
+      "epoch": 0.7059230237381885,
+      "grad_norm": 0.16783006489276886,
+      "learning_rate": 2.1022960345592223e-05,
+      "loss": 0.6394,
+      "step": 3063
+    },
+    {
+      "epoch": 0.7061534915879235,
+      "grad_norm": 0.17598304152488708,
+      "learning_rate": 2.0992547562915838e-05,
+      "loss": 0.6499,
+      "step": 3064
+    },
+    {
+      "epoch": 0.7063839594376584,
+      "grad_norm": 0.16829919815063477,
+      "learning_rate": 2.096215094829723e-05,
+      "loss": 0.6499,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7066144272873934,
+      "grad_norm": 0.1717023104429245,
+      "learning_rate": 2.0931770518678707e-05,
+      "loss": 0.6479,
+      "step": 3066
+    },
+    {
+      "epoch": 0.7068448951371283,
+      "grad_norm": 0.17776085436344147,
+      "learning_rate": 2.0901406290993598e-05,
+      "loss": 0.6604,
+      "step": 3067
+    },
+    {
+      "epoch": 0.7070753629868634,
+      "grad_norm": 0.17955102026462555,
+      "learning_rate": 2.087105828216619e-05,
+      "loss": 0.6479,
+      "step": 3068
+    },
+    {
+      "epoch": 0.7073058308365983,
+      "grad_norm": 0.17219555377960205,
+      "learning_rate": 2.0840726509111748e-05,
+      "loss": 0.6486,
+      "step": 3069
+    },
+    {
+      "epoch": 0.7075362986863333,
+      "grad_norm": 0.16756534576416016,
+      "learning_rate": 2.081041098873646e-05,
+      "loss": 0.636,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7077667665360682,
+      "grad_norm": 0.17784403264522552,
+      "learning_rate": 2.0780111737937497e-05,
+      "loss": 0.6599,
+      "step": 3071
+    },
+    {
+      "epoch": 0.7079972343858032,
+      "grad_norm": 0.16979734599590302,
+      "learning_rate": 2.0749828773602898e-05,
+      "loss": 0.6566,
+      "step": 3072
+    },
+    {
+      "epoch": 0.7082277022355381,
+      "grad_norm": 0.18243227899074554,
+      "learning_rate": 2.0719562112611675e-05,
+      "loss": 0.6452,
+      "step": 3073
+    },
+    {
+      "epoch": 0.7084581700852731,
+      "grad_norm": 0.16543544828891754,
+      "learning_rate": 2.0689311771833737e-05,
+      "loss": 0.6489,
+      "step": 3074
+    },
+    {
+      "epoch": 0.708688637935008,
+      "grad_norm": 0.17076393961906433,
+      "learning_rate": 2.0659077768129898e-05,
+      "loss": 0.6472,
+      "step": 3075
+    },
+    {
+      "epoch": 0.708919105784743,
+      "grad_norm": 0.17454828321933746,
+      "learning_rate": 2.0628860118351874e-05,
+      "loss": 0.6464,
+      "step": 3076
+    },
+    {
+      "epoch": 0.709149573634478,
+      "grad_norm": 0.16081586480140686,
+      "learning_rate": 2.0598658839342266e-05,
+      "loss": 0.6488,
+      "step": 3077
+    },
+    {
+      "epoch": 0.709380041484213,
+      "grad_norm": 0.17376914620399475,
+      "learning_rate": 2.0568473947934498e-05,
+      "loss": 0.6416,
+      "step": 3078
+    },
+    {
+      "epoch": 0.7096105093339479,
+      "grad_norm": 0.17064812779426575,
+      "learning_rate": 2.0538305460952945e-05,
+      "loss": 0.6464,
+      "step": 3079
+    },
+    {
+      "epoch": 0.7098409771836829,
+      "grad_norm": 0.18025130033493042,
+      "learning_rate": 2.050815339521281e-05,
+      "loss": 0.6501,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7100714450334178,
+      "grad_norm": 0.16442714631557465,
+      "learning_rate": 2.0478017767520087e-05,
+      "loss": 0.647,
+      "step": 3081
+    },
+    {
+      "epoch": 0.7103019128831528,
+      "grad_norm": 0.1766822636127472,
+      "learning_rate": 2.0447898594671667e-05,
+      "loss": 0.6456,
+      "step": 3082
+    },
+    {
+      "epoch": 0.7105323807328877,
+      "grad_norm": 0.17419351637363434,
+      "learning_rate": 2.0417795893455265e-05,
+      "loss": 0.6503,
+      "step": 3083
+    },
+    {
+      "epoch": 0.7107628485826227,
+      "grad_norm": 0.1673542559146881,
+      "learning_rate": 2.0387709680649397e-05,
+      "loss": 0.6526,
+      "step": 3084
+    },
+    {
+      "epoch": 0.7109933164323576,
+      "grad_norm": 0.1788138598203659,
+      "learning_rate": 2.0357639973023396e-05,
+      "loss": 0.6441,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7112237842820927,
+      "grad_norm": 0.16873040795326233,
+      "learning_rate": 2.032758678733741e-05,
+      "loss": 0.6467,
+      "step": 3086
+    },
+    {
+      "epoch": 0.7114542521318276,
+      "grad_norm": 0.17903602123260498,
+      "learning_rate": 2.0297550140342338e-05,
+      "loss": 0.647,
+      "step": 3087
+    },
+    {
+      "epoch": 0.7116847199815626,
+      "grad_norm": 0.15698370337486267,
+      "learning_rate": 2.0267530048779896e-05,
+      "loss": 0.6465,
+      "step": 3088
+    },
+    {
+      "epoch": 0.7119151878312975,
+      "grad_norm": 0.15955950319766998,
+      "learning_rate": 2.023752652938256e-05,
+      "loss": 0.6401,
+      "step": 3089
+    },
+    {
+      "epoch": 0.7121456556810325,
+      "grad_norm": 0.17143939435482025,
+      "learning_rate": 2.020753959887358e-05,
+      "loss": 0.6558,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7123761235307675,
+      "grad_norm": 0.16940918564796448,
+      "learning_rate": 2.0177569273966945e-05,
+      "loss": 0.6542,
+      "step": 3091
+    },
+    {
+      "epoch": 0.7126065913805024,
+      "grad_norm": 0.1974901705980301,
+      "learning_rate": 2.01476155713674e-05,
+      "loss": 0.6397,
+      "step": 3092
+    },
+    {
+      "epoch": 0.7128370592302374,
+      "grad_norm": 0.16214625537395477,
+      "learning_rate": 2.0117678507770416e-05,
+      "loss": 0.6492,
+      "step": 3093
+    },
+    {
+      "epoch": 0.7130675270799723,
+      "grad_norm": 0.19105158746242523,
+      "learning_rate": 2.0087758099862192e-05,
+      "loss": 0.6497,
+      "step": 3094
+    },
+    {
+      "epoch": 0.7132979949297074,
+      "grad_norm": 0.17240196466445923,
+      "learning_rate": 2.0057854364319646e-05,
+      "loss": 0.6336,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7135284627794423,
+      "grad_norm": 0.17806990444660187,
+      "learning_rate": 2.0027967317810426e-05,
+      "loss": 0.6529,
+      "step": 3096
+    },
+    {
+      "epoch": 0.7137589306291773,
+      "grad_norm": 0.1702287495136261,
+      "learning_rate": 1.9998096976992812e-05,
+      "loss": 0.6559,
+      "step": 3097
+    },
+    {
+      "epoch": 0.7139893984789122,
+      "grad_norm": 0.17224186658859253,
+      "learning_rate": 1.9968243358515837e-05,
+      "loss": 0.639,
+      "step": 3098
+    },
+    {
+      "epoch": 0.7142198663286472,
+      "grad_norm": 0.18311919271945953,
+      "learning_rate": 1.9938406479019183e-05,
+      "loss": 0.6348,
+      "step": 3099
+    },
+    {
+      "epoch": 0.7144503341783821,
+      "grad_norm": 0.1758183240890503,
+      "learning_rate": 1.9908586355133223e-05,
+      "loss": 0.6486,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7146808020281171,
+      "grad_norm": 0.1789383888244629,
+      "learning_rate": 1.9878783003478975e-05,
+      "loss": 0.649,
+      "step": 3101
+    },
+    {
+      "epoch": 0.714911269877852,
+      "grad_norm": 0.18306902050971985,
+      "learning_rate": 1.9848996440668123e-05,
+      "loss": 0.6387,
+      "step": 3102
+    },
+    {
+      "epoch": 0.715141737727587,
+      "grad_norm": 0.1843525767326355,
+      "learning_rate": 1.981922668330293e-05,
+      "loss": 0.6538,
+      "step": 3103
+    },
+    {
+      "epoch": 0.715372205577322,
+      "grad_norm": 0.16771547496318817,
+      "learning_rate": 1.9789473747976412e-05,
+      "loss": 0.6386,
+      "step": 3104
+    },
+    {
+      "epoch": 0.715602673427057,
+      "grad_norm": 0.19470912218093872,
+      "learning_rate": 1.975973765127212e-05,
+      "loss": 0.6467,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7158331412767919,
+      "grad_norm": 0.16712939739227295,
+      "learning_rate": 1.9730018409764218e-05,
+      "loss": 0.646,
+      "step": 3106
+    },
+    {
+      "epoch": 0.7160636091265269,
+      "grad_norm": 0.18347413837909698,
+      "learning_rate": 1.9700316040017515e-05,
+      "loss": 0.6378,
+      "step": 3107
+    },
+    {
+      "epoch": 0.7162940769762618,
+      "grad_norm": 0.18481610715389252,
+      "learning_rate": 1.967063055858739e-05,
+      "loss": 0.6492,
+      "step": 3108
+    },
+    {
+      "epoch": 0.7165245448259968,
+      "grad_norm": 0.15738435089588165,
+      "learning_rate": 1.9640961982019825e-05,
+      "loss": 0.6417,
+      "step": 3109
+    },
+    {
+      "epoch": 0.7167550126757317,
+      "grad_norm": 0.17788545787334442,
+      "learning_rate": 1.9611310326851373e-05,
+      "loss": 0.6416,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7169854805254667,
+      "grad_norm": 0.1677306592464447,
+      "learning_rate": 1.9581675609609173e-05,
+      "loss": 0.6449,
+      "step": 3111
+    },
+    {
+      "epoch": 0.7172159483752016,
+      "grad_norm": 0.17086228728294373,
+      "learning_rate": 1.9552057846810866e-05,
+      "loss": 0.6473,
+      "step": 3112
+    },
+    {
+      "epoch": 0.7174464162249367,
+      "grad_norm": 0.1743098348379135,
+      "learning_rate": 1.9522457054964683e-05,
+      "loss": 0.6488,
+      "step": 3113
+    },
+    {
+      "epoch": 0.7176768840746716,
+      "grad_norm": 0.1679782271385193,
+      "learning_rate": 1.949287325056945e-05,
+      "loss": 0.6525,
+      "step": 3114
+    },
+    {
+      "epoch": 0.7179073519244066,
+      "grad_norm": 0.15719559788703918,
+      "learning_rate": 1.9463306450114416e-05,
+      "loss": 0.6499,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7181378197741415,
+      "grad_norm": 0.16834014654159546,
+      "learning_rate": 1.9433756670079423e-05,
+      "loss": 0.6547,
+      "step": 3116
+    },
+    {
+      "epoch": 0.7183682876238765,
+      "grad_norm": 0.17471790313720703,
+      "learning_rate": 1.9404223926934828e-05,
+      "loss": 0.6499,
+      "step": 3117
+    },
+    {
+      "epoch": 0.7185987554736114,
+      "grad_norm": 0.16939441859722137,
+      "learning_rate": 1.9374708237141413e-05,
+      "loss": 0.6544,
+      "step": 3118
+    },
+    {
+      "epoch": 0.7188292233233464,
+      "grad_norm": 0.17127317190170288,
+      "learning_rate": 1.9345209617150577e-05,
+      "loss": 0.6403,
+      "step": 3119
+    },
+    {
+      "epoch": 0.7190596911730813,
+      "grad_norm": 0.16861754655838013,
+      "learning_rate": 1.9315728083404145e-05,
+      "loss": 0.6465,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7192901590228163,
+      "grad_norm": 0.17622019350528717,
+      "learning_rate": 1.9286263652334368e-05,
+      "loss": 0.6353,
+      "step": 3121
+    },
+    {
+      "epoch": 0.7195206268725512,
+      "grad_norm": 0.16617770493030548,
+      "learning_rate": 1.925681634036404e-05,
+      "loss": 0.6502,
+      "step": 3122
+    },
+    {
+      "epoch": 0.7197510947222863,
+      "grad_norm": 0.16550083458423615,
+      "learning_rate": 1.922738616390639e-05,
+      "loss": 0.6483,
+      "step": 3123
+    },
+    {
+      "epoch": 0.7199815625720212,
+      "grad_norm": 0.15915584564208984,
+      "learning_rate": 1.9197973139365083e-05,
+      "loss": 0.6441,
+      "step": 3124
+    },
+    {
+      "epoch": 0.7202120304217562,
+      "grad_norm": 0.16623058915138245,
+      "learning_rate": 1.9168577283134232e-05,
+      "loss": 0.6493,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7204424982714911,
+      "grad_norm": 0.16962306201457977,
+      "learning_rate": 1.9139198611598404e-05,
+      "loss": 0.6456,
+      "step": 3126
+    },
+    {
+      "epoch": 0.7206729661212261,
+      "grad_norm": 0.1750573068857193,
+      "learning_rate": 1.910983714113253e-05,
+      "loss": 0.644,
+      "step": 3127
+    },
+    {
+      "epoch": 0.720903433970961,
+      "grad_norm": 0.17163607478141785,
+      "learning_rate": 1.9080492888101993e-05,
+      "loss": 0.641,
+      "step": 3128
+    },
+    {
+      "epoch": 0.721133901820696,
+      "grad_norm": 0.6170444488525391,
+      "learning_rate": 1.9051165868862615e-05,
+      "loss": 0.6669,
+      "step": 3129
+    },
+    {
+      "epoch": 0.7213643696704309,
+      "grad_norm": 0.17929589748382568,
+      "learning_rate": 1.9021856099760533e-05,
+      "loss": 0.6479,
+      "step": 3130
+    },
+    {
+      "epoch": 0.721594837520166,
+      "grad_norm": 0.16149266064167023,
+      "learning_rate": 1.8992563597132323e-05,
+      "loss": 0.6498,
+      "step": 3131
+    },
+    {
+      "epoch": 0.7218253053699009,
+      "grad_norm": 0.16591522097587585,
+      "learning_rate": 1.8963288377304916e-05,
+      "loss": 0.6493,
+      "step": 3132
+    },
+    {
+      "epoch": 0.7220557732196359,
+      "grad_norm": 0.1631591022014618,
+      "learning_rate": 1.8934030456595625e-05,
+      "loss": 0.6517,
+      "step": 3133
+    },
+    {
+      "epoch": 0.7222862410693708,
+      "grad_norm": 0.1634044647216797,
+      "learning_rate": 1.890478985131211e-05,
+      "loss": 0.6413,
+      "step": 3134
+    },
+    {
+      "epoch": 0.7225167089191058,
+      "grad_norm": 0.1616399884223938,
+      "learning_rate": 1.88755665777524e-05,
+      "loss": 0.6432,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7227471767688407,
+      "grad_norm": 0.16510508954524994,
+      "learning_rate": 1.8846360652204816e-05,
+      "loss": 0.6428,
+      "step": 3136
+    },
+    {
+      "epoch": 0.7229776446185757,
+      "grad_norm": 0.15748094022274017,
+      "learning_rate": 1.881717209094805e-05,
+      "loss": 0.6463,
+      "step": 3137
+    },
+    {
+      "epoch": 0.7232081124683106,
+      "grad_norm": 0.17742668092250824,
+      "learning_rate": 1.8788000910251103e-05,
+      "loss": 0.6451,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7234385803180456,
+      "grad_norm": 0.16781218349933624,
+      "learning_rate": 1.8758847126373303e-05,
+      "loss": 0.6499,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7236690481677805,
+      "grad_norm": 0.1616116166114807,
+      "learning_rate": 1.8729710755564257e-05,
+      "loss": 0.6435,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7238995160175156,
+      "grad_norm": 0.1795063465833664,
+      "learning_rate": 1.8700591814063905e-05,
+      "loss": 0.6404,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7241299838672505,
+      "grad_norm": 0.16432838141918182,
+      "learning_rate": 1.867149031810241e-05,
+      "loss": 0.641,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7243604517169855,
+      "grad_norm": 0.1743660271167755,
+      "learning_rate": 1.864240628390024e-05,
+      "loss": 0.6405,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7245909195667204,
+      "grad_norm": 0.17047765851020813,
+      "learning_rate": 1.8613339727668194e-05,
+      "loss": 0.6471,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7248213874164554,
+      "grad_norm": 0.15893389284610748,
+      "learning_rate": 1.8584290665607228e-05,
+      "loss": 0.6404,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7250518552661903,
+      "grad_norm": 0.20478779077529907,
+      "learning_rate": 1.8555259113908597e-05,
+      "loss": 0.6507,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7252823231159253,
+      "grad_norm": 0.17662855982780457,
+      "learning_rate": 1.85262450887538e-05,
+      "loss": 0.6537,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7255127909656602,
+      "grad_norm": 0.17941661179065704,
+      "learning_rate": 1.849724860631456e-05,
+      "loss": 0.6383,
+      "step": 3148
+    },
+    {
+      "epoch": 0.7257432588153953,
+      "grad_norm": 0.15811479091644287,
+      "learning_rate": 1.846826968275281e-05,
+      "loss": 0.6488,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7259737266651302,
+      "grad_norm": 0.1792788952589035,
+      "learning_rate": 1.843930833422073e-05,
+      "loss": 0.6522,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7262041945148652,
+      "grad_norm": 0.16518691182136536,
+      "learning_rate": 1.8410364576860646e-05,
+      "loss": 0.6368,
+      "step": 3151
+    },
+    {
+      "epoch": 0.7264346623646002,
+      "grad_norm": 0.16625241935253143,
+      "learning_rate": 1.838143842680513e-05,
+      "loss": 0.6339,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7266651302143351,
+      "grad_norm": 0.17410995066165924,
+      "learning_rate": 1.8352529900176923e-05,
+      "loss": 0.6421,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7268955980640701,
+      "grad_norm": 0.16226623952388763,
+      "learning_rate": 1.832363901308895e-05,
+      "loss": 0.6515,
+      "step": 3154
+    },
+    {
+      "epoch": 0.727126065913805,
+      "grad_norm": 0.16737419366836548,
+      "learning_rate": 1.8294765781644285e-05,
+      "loss": 0.639,
+      "step": 3155
+    },
+    {
+      "epoch": 0.72735653376354,
+      "grad_norm": 0.1635529100894928,
+      "learning_rate": 1.8265910221936206e-05,
+      "loss": 0.6342,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7275870016132749,
+      "grad_norm": 0.17363202571868896,
+      "learning_rate": 1.823707235004805e-05,
+      "loss": 0.6494,
+      "step": 3157
+    },
+    {
+      "epoch": 0.72781746946301,
+      "grad_norm": 0.1653600037097931,
+      "learning_rate": 1.8208252182053403e-05,
+      "loss": 0.6471,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7280479373127449,
+      "grad_norm": 0.1543000489473343,
+      "learning_rate": 1.8179449734015948e-05,
+      "loss": 0.6399,
+      "step": 3159
+    },
+    {
+      "epoch": 0.7282784051624799,
+      "grad_norm": 0.1642274111509323,
+      "learning_rate": 1.8150665021989426e-05,
+      "loss": 0.6496,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7285088730122148,
+      "grad_norm": 0.16651754081249237,
+      "learning_rate": 1.812189806201778e-05,
+      "loss": 0.6451,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7287393408619498,
+      "grad_norm": 0.1647575944662094,
+      "learning_rate": 1.8093148870135e-05,
+      "loss": 0.6464,
+      "step": 3162
+    },
+    {
+      "epoch": 0.7289698087116847,
+      "grad_norm": 0.16285358369350433,
+      "learning_rate": 1.8064417462365226e-05,
+      "loss": 0.6495,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7292002765614197,
+      "grad_norm": 0.18453556299209595,
+      "learning_rate": 1.8035703854722623e-05,
+      "loss": 0.6457,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7294307444111546,
+      "grad_norm": 0.16734904050827026,
+      "learning_rate": 1.800700806321151e-05,
+      "loss": 0.6464,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7296612122608896,
+      "grad_norm": 0.167644202709198,
+      "learning_rate": 1.7978330103826184e-05,
+      "loss": 0.6494,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7298916801106246,
+      "grad_norm": 0.17859847843647003,
+      "learning_rate": 1.7949669992551053e-05,
+      "loss": 0.6481,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7301221479603596,
+      "grad_norm": 0.16591614484786987,
+      "learning_rate": 1.792102774536063e-05,
+      "loss": 0.6398,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7303526158100945,
+      "grad_norm": 0.15973828732967377,
+      "learning_rate": 1.7892403378219364e-05,
+      "loss": 0.6475,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7305830836598295,
+      "grad_norm": 0.16776901483535767,
+      "learning_rate": 1.786379690708181e-05,
+      "loss": 0.6442,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7308135515095644,
+      "grad_norm": 0.15858832001686096,
+      "learning_rate": 1.7835208347892535e-05,
+      "loss": 0.6528,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7310440193592994,
+      "grad_norm": 0.16048070788383484,
+      "learning_rate": 1.7806637716586073e-05,
+      "loss": 0.6393,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7312744872090343,
+      "grad_norm": 0.1663476973772049,
+      "learning_rate": 1.777808502908706e-05,
+      "loss": 0.6438,
+      "step": 3173
+    },
+    {
+      "epoch": 0.7315049550587693,
+      "grad_norm": 0.1595594882965088,
+      "learning_rate": 1.7749550301310074e-05,
+      "loss": 0.6347,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7317354229085042,
+      "grad_norm": 0.16028767824172974,
+      "learning_rate": 1.7721033549159655e-05,
+      "loss": 0.6495,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7319658907582393,
+      "grad_norm": 0.16454143822193146,
+      "learning_rate": 1.7692534788530374e-05,
+      "loss": 0.653,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7321963586079742,
+      "grad_norm": 0.16430748999118805,
+      "learning_rate": 1.7664054035306756e-05,
+      "loss": 0.6383,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7324268264577092,
+      "grad_norm": 0.16365505754947662,
+      "learning_rate": 1.7635591305363292e-05,
+      "loss": 0.6416,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7326572943074441,
+      "grad_norm": 0.1731487661600113,
+      "learning_rate": 1.7607146614564418e-05,
+      "loss": 0.6551,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7328877621571791,
+      "grad_norm": 0.1614767163991928,
+      "learning_rate": 1.7578719978764545e-05,
+      "loss": 0.6446,
+      "step": 3180
+    },
+    {
+      "epoch": 0.733118230006914,
+      "grad_norm": 0.16245704889297485,
+      "learning_rate": 1.755031141380796e-05,
+      "loss": 0.6451,
+      "step": 3181
+    },
+    {
+      "epoch": 0.733348697856649,
+      "grad_norm": 0.15642160177230835,
+      "learning_rate": 1.7521920935528917e-05,
+      "loss": 0.6422,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7335791657063839,
+      "grad_norm": 0.162491574883461,
+      "learning_rate": 1.749354855975164e-05,
+      "loss": 0.6495,
+      "step": 3183
+    },
+    {
+      "epoch": 0.733809633556119,
+      "grad_norm": 0.15852685272693634,
+      "learning_rate": 1.746519430229015e-05,
+      "loss": 0.6452,
+      "step": 3184
+    },
+    {
+      "epoch": 0.7340401014058539,
+      "grad_norm": 0.16387516260147095,
+      "learning_rate": 1.7436858178948457e-05,
+      "loss": 0.6464,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7342705692555889,
+      "grad_norm": 0.15998558700084686,
+      "learning_rate": 1.7408540205520436e-05,
+      "loss": 0.648,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7345010371053238,
+      "grad_norm": 0.17251843214035034,
+      "learning_rate": 1.7380240397789836e-05,
+      "loss": 0.6492,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7347315049550588,
+      "grad_norm": 0.1601380854845047,
+      "learning_rate": 1.7351958771530298e-05,
+      "loss": 0.6511,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7349619728047937,
+      "grad_norm": 0.15078677237033844,
+      "learning_rate": 1.7323695342505342e-05,
+      "loss": 0.6492,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7351924406545287,
+      "grad_norm": 0.16088485717773438,
+      "learning_rate": 1.729545012646828e-05,
+      "loss": 0.6434,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7354229085042636,
+      "grad_norm": 0.17706431448459625,
+      "learning_rate": 1.7267223139162342e-05,
+      "loss": 0.6491,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7356533763539986,
+      "grad_norm": 0.15498369932174683,
+      "learning_rate": 1.7239014396320574e-05,
+      "loss": 0.646,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7358838442037335,
+      "grad_norm": 0.1685081273317337,
+      "learning_rate": 1.7210823913665852e-05,
+      "loss": 0.6454,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7361143120534686,
+      "grad_norm": 0.16466312110424042,
+      "learning_rate": 1.718265170691087e-05,
+      "loss": 0.6512,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7363447799032035,
+      "grad_norm": 0.1716887503862381,
+      "learning_rate": 1.7154497791758157e-05,
+      "loss": 0.6552,
+      "step": 3195
+    },
+    {
+      "epoch": 0.7365752477529385,
+      "grad_norm": 0.17006585001945496,
+      "learning_rate": 1.7126362183899986e-05,
+      "loss": 0.6402,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7368057156026734,
+      "grad_norm": 0.15960989892482758,
+      "learning_rate": 1.7098244899018512e-05,
+      "loss": 0.6478,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7370361834524084,
+      "grad_norm": 0.16518181562423706,
+      "learning_rate": 1.707014595278564e-05,
+      "loss": 0.6585,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7372666513021433,
+      "grad_norm": 0.15796171128749847,
+      "learning_rate": 1.7042065360863007e-05,
+      "loss": 0.6493,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7374971191518783,
+      "grad_norm": 0.1603865772485733,
+      "learning_rate": 1.7014003138902092e-05,
+      "loss": 0.6361,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7377275870016132,
+      "grad_norm": 0.1591259092092514,
+      "learning_rate": 1.698595930254409e-05,
+      "loss": 0.6409,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7379580548513482,
+      "grad_norm": 0.16630783677101135,
+      "learning_rate": 1.6957933867419966e-05,
+      "loss": 0.6435,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7381885227010831,
+      "grad_norm": 0.17076456546783447,
+      "learning_rate": 1.6929926849150428e-05,
+      "loss": 0.6506,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7384189905508182,
+      "grad_norm": 0.16689413785934448,
+      "learning_rate": 1.6901938263345934e-05,
+      "loss": 0.6521,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7386494584005531,
+      "grad_norm": 0.16302435100078583,
+      "learning_rate": 1.687396812560661e-05,
+      "loss": 0.6446,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7388799262502881,
+      "grad_norm": 0.1644594967365265,
+      "learning_rate": 1.6846016451522362e-05,
+      "loss": 0.6389,
+      "step": 3206
+    },
+    {
+      "epoch": 0.739110394100023,
+      "grad_norm": 0.16946260631084442,
+      "learning_rate": 1.681808325667278e-05,
+      "loss": 0.6493,
+      "step": 3207
+    },
+    {
+      "epoch": 0.739340861949758,
+      "grad_norm": 0.16244158148765564,
+      "learning_rate": 1.6790168556627156e-05,
+      "loss": 0.6507,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7395713297994929,
+      "grad_norm": 0.18265312910079956,
+      "learning_rate": 1.6762272366944472e-05,
+      "loss": 0.6404,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7398017976492279,
+      "grad_norm": 0.16505154967308044,
+      "learning_rate": 1.673439470317341e-05,
+      "loss": 0.6464,
+      "step": 3210
+    },
+    {
+      "epoch": 0.740032265498963,
+      "grad_norm": 0.1631191074848175,
+      "learning_rate": 1.6706535580852267e-05,
+      "loss": 0.6385,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7402627333486979,
+      "grad_norm": 0.17808805406093597,
+      "learning_rate": 1.66786950155091e-05,
+      "loss": 0.6458,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7404932011984329,
+      "grad_norm": 0.16241349279880524,
+      "learning_rate": 1.6650873022661563e-05,
+      "loss": 0.6487,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7407236690481678,
+      "grad_norm": 0.1839081048965454,
+      "learning_rate": 1.662306961781694e-05,
+      "loss": 0.6478,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7409541368979028,
+      "grad_norm": 0.16764585673809052,
+      "learning_rate": 1.6595284816472195e-05,
+      "loss": 0.6424,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7411846047476377,
+      "grad_norm": 0.15650008618831635,
+      "learning_rate": 1.6567518634113916e-05,
+      "loss": 0.6455,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7414150725973727,
+      "grad_norm": 0.17013835906982422,
+      "learning_rate": 1.65397710862183e-05,
+      "loss": 0.6422,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7416455404471076,
+      "grad_norm": 0.15841804444789886,
+      "learning_rate": 1.6512042188251164e-05,
+      "loss": 0.6402,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7418760082968426,
+      "grad_norm": 0.1625175029039383,
+      "learning_rate": 1.6484331955667947e-05,
+      "loss": 0.6491,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7421064761465775,
+      "grad_norm": 0.17238567769527435,
+      "learning_rate": 1.6456640403913638e-05,
+      "loss": 0.6331,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7423369439963126,
+      "grad_norm": 0.15374347567558289,
+      "learning_rate": 1.642896754842284e-05,
+      "loss": 0.656,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7425674118460475,
+      "grad_norm": 0.17570993304252625,
+      "learning_rate": 1.640131340461978e-05,
+      "loss": 0.641,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7427978796957825,
+      "grad_norm": 0.15289098024368286,
+      "learning_rate": 1.6373677987918195e-05,
+      "loss": 0.6542,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7430283475455174,
+      "grad_norm": 0.16534923017024994,
+      "learning_rate": 1.634606131372139e-05,
+      "loss": 0.6461,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7432588153952524,
+      "grad_norm": 0.1682467758655548,
+      "learning_rate": 1.631846339742224e-05,
+      "loss": 0.65,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7434892832449873,
+      "grad_norm": 0.15141364932060242,
+      "learning_rate": 1.629088425440317e-05,
+      "loss": 0.6427,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7437197510947223,
+      "grad_norm": 0.16520178318023682,
+      "learning_rate": 1.6263323900036126e-05,
+      "loss": 0.6427,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7439502189444572,
+      "grad_norm": 0.1559877246618271,
+      "learning_rate": 1.6235782349682592e-05,
+      "loss": 0.6452,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7441806867941922,
+      "grad_norm": 0.15489576756954193,
+      "learning_rate": 1.6208259618693583e-05,
+      "loss": 0.6445,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7444111546439272,
+      "grad_norm": 0.1593122035264969,
+      "learning_rate": 1.618075572240957e-05,
+      "loss": 0.6506,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7446416224936622,
+      "grad_norm": 0.15990042686462402,
+      "learning_rate": 1.615327067616057e-05,
+      "loss": 0.6488,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7448720903433971,
+      "grad_norm": 0.15468232333660126,
+      "learning_rate": 1.612580449526614e-05,
+      "loss": 0.6389,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7451025581931321,
+      "grad_norm": 0.15509067475795746,
+      "learning_rate": 1.6098357195035212e-05,
+      "loss": 0.6439,
+      "step": 3233
+    },
+    {
+      "epoch": 0.745333026042867,
+      "grad_norm": 0.17395319044589996,
+      "learning_rate": 1.6070928790766275e-05,
+      "loss": 0.638,
+      "step": 3234
+    },
+    {
+      "epoch": 0.745563493892602,
+      "grad_norm": 0.151578888297081,
+      "learning_rate": 1.6043519297747285e-05,
+      "loss": 0.6367,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7457939617423369,
+      "grad_norm": 0.15744104981422424,
+      "learning_rate": 1.6016128731255575e-05,
+      "loss": 0.6429,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7460244295920719,
+      "grad_norm": 0.17480209469795227,
+      "learning_rate": 1.5988757106558043e-05,
+      "loss": 0.6509,
+      "step": 3237
+    },
+    {
+      "epoch": 0.7462548974418068,
+      "grad_norm": 0.16368581354618073,
+      "learning_rate": 1.5961404438910976e-05,
+      "loss": 0.6411,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7464853652915419,
+      "grad_norm": 0.15094834566116333,
+      "learning_rate": 1.5934070743560065e-05,
+      "loss": 0.64,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7467158331412768,
+      "grad_norm": 0.16196765005588531,
+      "learning_rate": 1.590675603574046e-05,
+      "loss": 0.6455,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7469463009910118,
+      "grad_norm": 0.17574729025363922,
+      "learning_rate": 1.5879460330676743e-05,
+      "loss": 0.6494,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7471767688407467,
+      "grad_norm": 0.1642741709947586,
+      "learning_rate": 1.5852183643582868e-05,
+      "loss": 0.6417,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7474072366904817,
+      "grad_norm": 0.1686857044696808,
+      "learning_rate": 1.5824925989662216e-05,
+      "loss": 0.6415,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7476377045402166,
+      "grad_norm": 0.17271287739276886,
+      "learning_rate": 1.5797687384107558e-05,
+      "loss": 0.6514,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7478681723899516,
+      "grad_norm": 0.16826404631137848,
+      "learning_rate": 1.577046784210101e-05,
+      "loss": 0.6418,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7480986402396865,
+      "grad_norm": 0.5926986932754517,
+      "learning_rate": 1.574326737881409e-05,
+      "loss": 0.6596,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7483291080894215,
+      "grad_norm": 0.16710159182548523,
+      "learning_rate": 1.571608600940774e-05,
+      "loss": 0.6324,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7485595759391565,
+      "grad_norm": 0.16913191974163055,
+      "learning_rate": 1.568892374903214e-05,
+      "loss": 0.6437,
+      "step": 3248
+    },
+    {
+      "epoch": 0.7487900437888915,
+      "grad_norm": 0.21093370020389557,
+      "learning_rate": 1.566178061282691e-05,
+      "loss": 0.6346,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7490205116386264,
+      "grad_norm": 0.15978707373142242,
+      "learning_rate": 1.5634656615920974e-05,
+      "loss": 0.6546,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7492509794883614,
+      "grad_norm": 0.17602089047431946,
+      "learning_rate": 1.56075517734326e-05,
+      "loss": 0.6429,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7494814473380963,
+      "grad_norm": 0.16324278712272644,
+      "learning_rate": 1.558046610046938e-05,
+      "loss": 0.6439,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7497119151878313,
+      "grad_norm": 0.1663910150527954,
+      "learning_rate": 1.5553399612128234e-05,
+      "loss": 0.6506,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7499423830375662,
+      "grad_norm": 0.17245832085609436,
+      "learning_rate": 1.5526352323495336e-05,
+      "loss": 0.6397,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7501728508873012,
+      "grad_norm": 0.1658082902431488,
+      "learning_rate": 1.549932424964622e-05,
+      "loss": 0.6503,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7504033187370361,
+      "grad_norm": 0.16782449185848236,
+      "learning_rate": 1.547231540564567e-05,
+      "loss": 0.6483,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7506337865867712,
+      "grad_norm": 0.16004249453544617,
+      "learning_rate": 1.5445325806547782e-05,
+      "loss": 0.6485,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7508642544365061,
+      "grad_norm": 0.1629590541124344,
+      "learning_rate": 1.5418355467395906e-05,
+      "loss": 0.6478,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7510947222862411,
+      "grad_norm": 0.1687597781419754,
+      "learning_rate": 1.5391404403222676e-05,
+      "loss": 0.6482,
+      "step": 3259
+    },
+    {
+      "epoch": 0.751325190135976,
+      "grad_norm": 0.16578345000743866,
+      "learning_rate": 1.536447262904994e-05,
+      "loss": 0.648,
+      "step": 3260
+    },
+    {
+      "epoch": 0.751555657985711,
+      "grad_norm": 0.15404678881168365,
+      "learning_rate": 1.533756015988882e-05,
+      "loss": 0.6436,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7517861258354459,
+      "grad_norm": 0.18324708938598633,
+      "learning_rate": 1.5310667010739726e-05,
+      "loss": 0.6466,
+      "step": 3262
+    },
+    {
+      "epoch": 0.7520165936851809,
+      "grad_norm": 0.16745442152023315,
+      "learning_rate": 1.5283793196592212e-05,
+      "loss": 0.6459,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7522470615349158,
+      "grad_norm": 0.16149534285068512,
+      "learning_rate": 1.5256938732425107e-05,
+      "loss": 0.6426,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7524775293846508,
+      "grad_norm": 0.14993827044963837,
+      "learning_rate": 1.5230103633206449e-05,
+      "loss": 0.6334,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7527079972343858,
+      "grad_norm": 0.15830254554748535,
+      "learning_rate": 1.5203287913893478e-05,
+      "loss": 0.6382,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7529384650841208,
+      "grad_norm": 0.15330778062343597,
+      "learning_rate": 1.5176491589432628e-05,
+      "loss": 0.6551,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7531689329338557,
+      "grad_norm": 0.15092666447162628,
+      "learning_rate": 1.5149714674759546e-05,
+      "loss": 0.6435,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7533994007835907,
+      "grad_norm": 0.16142503917217255,
+      "learning_rate": 1.5122957184799007e-05,
+      "loss": 0.6423,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7536298686333257,
+      "grad_norm": 0.15477819740772247,
+      "learning_rate": 1.5096219134465017e-05,
+      "loss": 0.6452,
+      "step": 3270
+    },
+    {
+      "epoch": 0.7538603364830606,
+      "grad_norm": 0.16160400211811066,
+      "learning_rate": 1.5069500538660713e-05,
+      "loss": 0.6451,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7540908043327956,
+      "grad_norm": 0.17091768980026245,
+      "learning_rate": 1.5042801412278412e-05,
+      "loss": 0.6416,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7543212721825305,
+      "grad_norm": 0.1987750232219696,
+      "learning_rate": 1.5016121770199553e-05,
+      "loss": 0.64,
+      "step": 3273
+    },
+    {
+      "epoch": 0.7545517400322655,
+      "grad_norm": 0.16774295270442963,
+      "learning_rate": 1.4989461627294755e-05,
+      "loss": 0.639,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7547822078820005,
+      "grad_norm": 0.1618596464395523,
+      "learning_rate": 1.4962820998423683e-05,
+      "loss": 0.6448,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7550126757317355,
+      "grad_norm": 0.1566198170185089,
+      "learning_rate": 1.4936199898435238e-05,
+      "loss": 0.642,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7552431435814704,
+      "grad_norm": 0.17685432732105255,
+      "learning_rate": 1.4909598342167385e-05,
+      "loss": 0.6434,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7554736114312054,
+      "grad_norm": 0.15080119669437408,
+      "learning_rate": 1.488301634444716e-05,
+      "loss": 0.6513,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7557040792809403,
+      "grad_norm": 0.15888544917106628,
+      "learning_rate": 1.485645392009074e-05,
+      "loss": 0.6462,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7559345471306753,
+      "grad_norm": 0.16779987514019012,
+      "learning_rate": 1.4829911083903386e-05,
+      "loss": 0.6443,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7561650149804102,
+      "grad_norm": 0.15703189373016357,
+      "learning_rate": 1.4803387850679445e-05,
+      "loss": 0.6497,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7563954828301452,
+      "grad_norm": 0.1584496796131134,
+      "learning_rate": 1.477688423520232e-05,
+      "loss": 0.6466,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7566259506798801,
+      "grad_norm": 0.16846555471420288,
+      "learning_rate": 1.4750400252244511e-05,
+      "loss": 0.6452,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7568564185296152,
+      "grad_norm": 0.1615644246339798,
+      "learning_rate": 1.4723935916567522e-05,
+      "loss": 0.6491,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7570868863793501,
+      "grad_norm": 0.15736454725265503,
+      "learning_rate": 1.469749124292194e-05,
+      "loss": 0.6436,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7573173542290851,
+      "grad_norm": 0.15067139267921448,
+      "learning_rate": 1.4671066246047438e-05,
+      "loss": 0.6353,
+      "step": 3286
+    },
+    {
+      "epoch": 0.75754782207882,
+      "grad_norm": 0.16569077968597412,
+      "learning_rate": 1.4644660940672627e-05,
+      "loss": 0.6491,
+      "step": 3287
+    },
+    {
+      "epoch": 0.757778289928555,
+      "grad_norm": 0.15051403641700745,
+      "learning_rate": 1.461827534151521e-05,
+      "loss": 0.6394,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7580087577782899,
+      "grad_norm": 0.15739889442920685,
+      "learning_rate": 1.4591909463281894e-05,
+      "loss": 0.6364,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7582392256280249,
+      "grad_norm": 0.16223600506782532,
+      "learning_rate": 1.4565563320668346e-05,
+      "loss": 0.6472,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7584696934777598,
+      "grad_norm": 0.159046471118927,
+      "learning_rate": 1.4539236928359318e-05,
+      "loss": 0.654,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7587001613274948,
+      "grad_norm": 0.16086021065711975,
+      "learning_rate": 1.451293030102851e-05,
+      "loss": 0.6456,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7589306291772298,
+      "grad_norm": 0.15306276082992554,
+      "learning_rate": 1.4486643453338571e-05,
+      "loss": 0.656,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7591610970269648,
+      "grad_norm": 0.16488851606845856,
+      "learning_rate": 1.4460376399941184e-05,
+      "loss": 0.6519,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7593915648766997,
+      "grad_norm": 0.16087326407432556,
+      "learning_rate": 1.443412915547696e-05,
+      "loss": 0.6482,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7596220327264347,
+      "grad_norm": 0.15743768215179443,
+      "learning_rate": 1.4407901734575496e-05,
+      "loss": 0.6472,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7598525005761696,
+      "grad_norm": 0.16286884248256683,
+      "learning_rate": 1.4381694151855318e-05,
+      "loss": 0.6464,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7600829684259046,
+      "grad_norm": 0.15299223363399506,
+      "learning_rate": 1.4355506421923926e-05,
+      "loss": 0.6535,
+      "step": 3298
+    },
+    {
+      "epoch": 0.7603134362756395,
+      "grad_norm": 0.16617640852928162,
+      "learning_rate": 1.4329338559377691e-05,
+      "loss": 0.6402,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7605439041253745,
+      "grad_norm": 0.17199209332466125,
+      "learning_rate": 1.4303190578801967e-05,
+      "loss": 0.6406,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7607743719751094,
+      "grad_norm": 0.1543896347284317,
+      "learning_rate": 1.4277062494771044e-05,
+      "loss": 0.646,
+      "step": 3301
+    },
+    {
+      "epoch": 0.7610048398248445,
+      "grad_norm": 0.16141347587108612,
+      "learning_rate": 1.4250954321848043e-05,
+      "loss": 0.6484,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7612353076745794,
+      "grad_norm": 0.17056810855865479,
+      "learning_rate": 1.4224866074585053e-05,
+      "loss": 0.641,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7614657755243144,
+      "grad_norm": 0.17288394272327423,
+      "learning_rate": 1.4198797767523037e-05,
+      "loss": 0.6421,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7616962433740493,
+      "grad_norm": 0.16343657672405243,
+      "learning_rate": 1.4172749415191844e-05,
+      "loss": 0.6313,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7619267112237843,
+      "grad_norm": 0.16413342952728271,
+      "learning_rate": 1.41467210321102e-05,
+      "loss": 0.6415,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7621571790735192,
+      "grad_norm": 0.17422489821910858,
+      "learning_rate": 1.412071263278571e-05,
+      "loss": 0.6417,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7623876469232542,
+      "grad_norm": 0.16758742928504944,
+      "learning_rate": 1.4094724231714812e-05,
+      "loss": 0.6429,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7626181147729891,
+      "grad_norm": 0.16555525362491608,
+      "learning_rate": 1.406875584338282e-05,
+      "loss": 0.641,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7628485826227241,
+      "grad_norm": 0.168875589966774,
+      "learning_rate": 1.4042807482263904e-05,
+      "loss": 0.6468,
+      "step": 3310
+    },
+    {
+      "epoch": 0.763079050472459,
+      "grad_norm": 0.16183854639530182,
+      "learning_rate": 1.4016879162821044e-05,
+      "loss": 0.644,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7633095183221941,
+      "grad_norm": 0.15983159840106964,
+      "learning_rate": 1.3990970899506072e-05,
+      "loss": 0.6481,
+      "step": 3312
+    },
+    {
+      "epoch": 0.763539986171929,
+      "grad_norm": 0.1633015125989914,
+      "learning_rate": 1.3965082706759646e-05,
+      "loss": 0.642,
+      "step": 3313
+    },
+    {
+      "epoch": 0.763770454021664,
+      "grad_norm": 0.15642528235912323,
+      "learning_rate": 1.3939214599011174e-05,
+      "loss": 0.6431,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7640009218713989,
+      "grad_norm": 0.15615011751651764,
+      "learning_rate": 1.3913366590678966e-05,
+      "loss": 0.6502,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7642313897211339,
+      "grad_norm": 0.15808658301830292,
+      "learning_rate": 1.3887538696170089e-05,
+      "loss": 0.6522,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7644618575708688,
+      "grad_norm": 0.15133537352085114,
+      "learning_rate": 1.3861730929880346e-05,
+      "loss": 0.6448,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7646923254206038,
+      "grad_norm": 0.152401402592659,
+      "learning_rate": 1.3835943306194393e-05,
+      "loss": 0.6524,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7649227932703387,
+      "grad_norm": 0.15733560919761658,
+      "learning_rate": 1.3810175839485628e-05,
+      "loss": 0.6479,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7651532611200738,
+      "grad_norm": 0.16682898998260498,
+      "learning_rate": 1.3784428544116218e-05,
+      "loss": 0.6391,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7653837289698087,
+      "grad_norm": 0.14624935388565063,
+      "learning_rate": 1.3758701434437088e-05,
+      "loss": 0.6499,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7656141968195437,
+      "grad_norm": 0.1629732847213745,
+      "learning_rate": 1.3732994524787935e-05,
+      "loss": 0.6372,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7658446646692786,
+      "grad_norm": 0.1721165031194687,
+      "learning_rate": 1.370730782949713e-05,
+      "loss": 0.6376,
+      "step": 3323
+    },
+    {
+      "epoch": 0.7660751325190136,
+      "grad_norm": 0.1540745496749878,
+      "learning_rate": 1.3681641362881842e-05,
+      "loss": 0.6442,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7663056003687485,
+      "grad_norm": 0.1577741801738739,
+      "learning_rate": 1.365599513924794e-05,
+      "loss": 0.6475,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7665360682184835,
+      "grad_norm": 0.15906648337841034,
+      "learning_rate": 1.3630369172890017e-05,
+      "loss": 0.6426,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7667665360682184,
+      "grad_norm": 0.14891204237937927,
+      "learning_rate": 1.3604763478091375e-05,
+      "loss": 0.6394,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7669970039179534,
+      "grad_norm": 0.1566314697265625,
+      "learning_rate": 1.3579178069124021e-05,
+      "loss": 0.6483,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7672274717676885,
+      "grad_norm": 0.16238120198249817,
+      "learning_rate": 1.3553612960248607e-05,
+      "loss": 0.6396,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7674579396174234,
+      "grad_norm": 0.16790124773979187,
+      "learning_rate": 1.3528068165714552e-05,
+      "loss": 0.6512,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7676884074671584,
+      "grad_norm": 0.15930506587028503,
+      "learning_rate": 1.3502543699759917e-05,
+      "loss": 0.6483,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7679188753168933,
+      "grad_norm": 0.15827548503875732,
+      "learning_rate": 1.34770395766114e-05,
+      "loss": 0.6445,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7681493431666283,
+      "grad_norm": 0.16675326228141785,
+      "learning_rate": 1.3451555810484389e-05,
+      "loss": 0.6528,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7683798110163632,
+      "grad_norm": 0.16960427165031433,
+      "learning_rate": 1.3426092415582936e-05,
+      "loss": 0.6565,
+      "step": 3334
+    },
+    {
+      "epoch": 0.7686102788660982,
+      "grad_norm": 0.1545518934726715,
+      "learning_rate": 1.3400649406099719e-05,
+      "loss": 0.6461,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7688407467158331,
+      "grad_norm": 0.1682869791984558,
+      "learning_rate": 1.337522679621606e-05,
+      "loss": 0.6409,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7690712145655682,
+      "grad_norm": 0.1694023311138153,
+      "learning_rate": 1.3349824600101934e-05,
+      "loss": 0.6443,
+      "step": 3337
+    },
+    {
+      "epoch": 0.7693016824153031,
+      "grad_norm": 0.15705664455890656,
+      "learning_rate": 1.3324442831915878e-05,
+      "loss": 0.6416,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7695321502650381,
+      "grad_norm": 0.16872623562812805,
+      "learning_rate": 1.3299081505805088e-05,
+      "loss": 0.6481,
+      "step": 3339
+    },
+    {
+      "epoch": 0.769762618114773,
+      "grad_norm": 0.15904702246189117,
+      "learning_rate": 1.3273740635905397e-05,
+      "loss": 0.6493,
+      "step": 3340
+    },
+    {
+      "epoch": 0.769993085964508,
+      "grad_norm": 0.165695458650589,
+      "learning_rate": 1.3248420236341147e-05,
+      "loss": 0.6417,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7702235538142429,
+      "grad_norm": 0.16277405619621277,
+      "learning_rate": 1.3223120321225351e-05,
+      "loss": 0.644,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7704540216639779,
+      "grad_norm": 0.15501618385314941,
+      "learning_rate": 1.319784090465958e-05,
+      "loss": 0.6462,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7706844895137128,
+      "grad_norm": 0.16321119666099548,
+      "learning_rate": 1.317258200073393e-05,
+      "loss": 0.6355,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7709149573634478,
+      "grad_norm": 0.1589282751083374,
+      "learning_rate": 1.314734362352716e-05,
+      "loss": 0.6329,
+      "step": 3345
+    },
+    {
+      "epoch": 0.7711454252131827,
+      "grad_norm": 0.15726275742053986,
+      "learning_rate": 1.3122125787106521e-05,
+      "loss": 0.632,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7713758930629178,
+      "grad_norm": 0.17059746384620667,
+      "learning_rate": 1.3096928505527811e-05,
+      "loss": 0.6396,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7716063609126527,
+      "grad_norm": 0.15518136322498322,
+      "learning_rate": 1.3071751792835402e-05,
+      "loss": 0.651,
+      "step": 3348
+    },
+    {
+      "epoch": 0.7718368287623877,
+      "grad_norm": 0.16409626603126526,
+      "learning_rate": 1.3046595663062188e-05,
+      "loss": 0.6485,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7720672966121226,
+      "grad_norm": 0.15897217392921448,
+      "learning_rate": 1.3021460130229596e-05,
+      "loss": 0.6549,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7722977644618576,
+      "grad_norm": 0.15244118869304657,
+      "learning_rate": 1.2996345208347565e-05,
+      "loss": 0.6445,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7725282323115925,
+      "grad_norm": 0.15054090321063995,
+      "learning_rate": 1.2971250911414567e-05,
+      "loss": 0.6433,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7727587001613275,
+      "grad_norm": 0.16574056446552277,
+      "learning_rate": 1.2946177253417525e-05,
+      "loss": 0.6394,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7729891680110624,
+      "grad_norm": 0.16119171679019928,
+      "learning_rate": 1.2921124248331901e-05,
+      "loss": 0.6408,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7732196358607974,
+      "grad_norm": 0.15423963963985443,
+      "learning_rate": 1.2896091910121666e-05,
+      "loss": 0.6454,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7734501037105324,
+      "grad_norm": 0.15891356766223907,
+      "learning_rate": 1.2871080252739247e-05,
+      "loss": 0.6464,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7736805715602674,
+      "grad_norm": 0.15453371405601501,
+      "learning_rate": 1.2846089290125507e-05,
+      "loss": 0.6365,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7739110394100023,
+      "grad_norm": 0.15373654663562775,
+      "learning_rate": 1.2821119036209827e-05,
+      "loss": 0.6484,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7741415072597373,
+      "grad_norm": 0.15394984185695648,
+      "learning_rate": 1.2796169504910028e-05,
+      "loss": 0.6376,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7743719751094722,
+      "grad_norm": 0.1592058390378952,
+      "learning_rate": 1.2771240710132375e-05,
+      "loss": 0.642,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7746024429592072,
+      "grad_norm": 0.1550869345664978,
+      "learning_rate": 1.2746332665771587e-05,
+      "loss": 0.6512,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7748329108089421,
+      "grad_norm": 0.16978594660758972,
+      "learning_rate": 1.2721445385710818e-05,
+      "loss": 0.6474,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7750633786586771,
+      "grad_norm": 0.16240744292736053,
+      "learning_rate": 1.2696578883821614e-05,
+      "loss": 0.6442,
+      "step": 3363
+    },
+    {
+      "epoch": 0.775293846508412,
+      "grad_norm": 0.14979274570941925,
+      "learning_rate": 1.2671733173963968e-05,
+      "loss": 0.6478,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7755243143581471,
+      "grad_norm": 0.16255247592926025,
+      "learning_rate": 1.2646908269986318e-05,
+      "loss": 0.6445,
+      "step": 3365
+    },
+    {
+      "epoch": 0.775754782207882,
+      "grad_norm": 0.15680548548698425,
+      "learning_rate": 1.2622104185725441e-05,
+      "loss": 0.6426,
+      "step": 3366
+    },
+    {
+      "epoch": 0.775985250057617,
+      "grad_norm": 0.15975458920001984,
+      "learning_rate": 1.259732093500654e-05,
+      "loss": 0.6461,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7762157179073519,
+      "grad_norm": 0.15226735174655914,
+      "learning_rate": 1.2572558531643208e-05,
+      "loss": 0.6403,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7764461857570869,
+      "grad_norm": 0.1590985804796219,
+      "learning_rate": 1.2547816989437416e-05,
+      "loss": 0.645,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7766766536068218,
+      "grad_norm": 0.1611194908618927,
+      "learning_rate": 1.2523096322179501e-05,
+      "loss": 0.6515,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7769071214565568,
+      "grad_norm": 0.16382546722888947,
+      "learning_rate": 1.2498396543648195e-05,
+      "loss": 0.6497,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7771375893062917,
+      "grad_norm": 0.15387794375419617,
+      "learning_rate": 1.2473717667610519e-05,
+      "loss": 0.6458,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7773680571560267,
+      "grad_norm": 0.15444159507751465,
+      "learning_rate": 1.2449059707821904e-05,
+      "loss": 0.6327,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7775985250057617,
+      "grad_norm": 0.16839702427387238,
+      "learning_rate": 1.2424422678026116e-05,
+      "loss": 0.6397,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7778289928554967,
+      "grad_norm": 0.16093666851520538,
+      "learning_rate": 1.2399806591955227e-05,
+      "loss": 0.6428,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7780594607052316,
+      "grad_norm": 0.1614423543214798,
+      "learning_rate": 1.2375211463329666e-05,
+      "loss": 0.6453,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7782899285549666,
+      "grad_norm": 0.1571039855480194,
+      "learning_rate": 1.2350637305858176e-05,
+      "loss": 0.6286,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7785203964047015,
+      "grad_norm": 0.1751880794763565,
+      "learning_rate": 1.2326084133237774e-05,
+      "loss": 0.6371,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7787508642544365,
+      "grad_norm": 0.16066774725914001,
+      "learning_rate": 1.2301551959153813e-05,
+      "loss": 0.6501,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7789813321041714,
+      "grad_norm": 0.16211430728435516,
+      "learning_rate": 1.2277040797279976e-05,
+      "loss": 0.6437,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7792117999539064,
+      "grad_norm": 0.16450825333595276,
+      "learning_rate": 1.2252550661278156e-05,
+      "loss": 0.6409,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7794422678036413,
+      "grad_norm": 0.16443844139575958,
+      "learning_rate": 1.2228081564798583e-05,
+      "loss": 0.6472,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7796727356533764,
+      "grad_norm": 0.17358103394508362,
+      "learning_rate": 1.2203633521479735e-05,
+      "loss": 0.6389,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7799032035031113,
+      "grad_norm": 0.16539937257766724,
+      "learning_rate": 1.2179206544948374e-05,
+      "loss": 0.6404,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7801336713528463,
+      "grad_norm": 0.15908871591091156,
+      "learning_rate": 1.2154800648819508e-05,
+      "loss": 0.6396,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7803641392025812,
+      "grad_norm": 0.17083421349525452,
+      "learning_rate": 1.2130415846696414e-05,
+      "loss": 0.6408,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7805946070523162,
+      "grad_norm": 0.17554426193237305,
+      "learning_rate": 1.2106052152170561e-05,
+      "loss": 0.6402,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7808250749020512,
+      "grad_norm": 0.16431263089179993,
+      "learning_rate": 1.2081709578821709e-05,
+      "loss": 0.6394,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7810555427517861,
+      "grad_norm": 0.16662830114364624,
+      "learning_rate": 1.2057388140217818e-05,
+      "loss": 0.6456,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7812860106015211,
+      "grad_norm": 0.16625608503818512,
+      "learning_rate": 1.2033087849915076e-05,
+      "loss": 0.6474,
+      "step": 3390
+    },
+    {
+      "epoch": 0.781516478451256,
+      "grad_norm": 0.1618298441171646,
+      "learning_rate": 1.2008808721457881e-05,
+      "loss": 0.642,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7817469463009911,
+      "grad_norm": 0.1568860411643982,
+      "learning_rate": 1.1984550768378856e-05,
+      "loss": 0.6519,
+      "step": 3392
+    },
+    {
+      "epoch": 0.781977414150726,
+      "grad_norm": 0.15564152598381042,
+      "learning_rate": 1.1960314004198752e-05,
+      "loss": 0.6391,
+      "step": 3393
+    },
+    {
+      "epoch": 0.782207882000461,
+      "grad_norm": 0.16593000292778015,
+      "learning_rate": 1.1936098442426608e-05,
+      "loss": 0.6472,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7824383498501959,
+      "grad_norm": 0.1633983701467514,
+      "learning_rate": 1.1911904096559589e-05,
+      "loss": 0.6497,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7826688176999309,
+      "grad_norm": 0.15584073960781097,
+      "learning_rate": 1.1887730980083023e-05,
+      "loss": 0.6437,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7828992855496658,
+      "grad_norm": 0.15618832409381866,
+      "learning_rate": 1.1863579106470434e-05,
+      "loss": 0.6372,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7831297533994008,
+      "grad_norm": 0.16305984556674957,
+      "learning_rate": 1.1839448489183503e-05,
+      "loss": 0.6409,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7833602212491357,
+      "grad_norm": 0.1538400650024414,
+      "learning_rate": 1.181533914167205e-05,
+      "loss": 0.6457,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7835906890988708,
+      "grad_norm": 0.15274596214294434,
+      "learning_rate": 1.1791251077374043e-05,
+      "loss": 0.6412,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7838211569486057,
+      "grad_norm": 0.1625516712665558,
+      "learning_rate": 1.1767184309715618e-05,
+      "loss": 0.6418,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7840516247983407,
+      "grad_norm": 0.14723870158195496,
+      "learning_rate": 1.1743138852110969e-05,
+      "loss": 0.6439,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7842820926480756,
+      "grad_norm": 0.15598557889461517,
+      "learning_rate": 1.1719114717962476e-05,
+      "loss": 0.638,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7845125604978106,
+      "grad_norm": 0.16605372726917267,
+      "learning_rate": 1.169511192066064e-05,
+      "loss": 0.6421,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7847430283475455,
+      "grad_norm": 0.1595214158296585,
+      "learning_rate": 1.1671130473584013e-05,
+      "loss": 0.6449,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7849734961972805,
+      "grad_norm": 0.15899643301963806,
+      "learning_rate": 1.1647170390099283e-05,
+      "loss": 0.6373,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7852039640470154,
+      "grad_norm": 0.16188432276248932,
+      "learning_rate": 1.1623231683561247e-05,
+      "loss": 0.6418,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7854344318967504,
+      "grad_norm": 0.16148293018341064,
+      "learning_rate": 1.1599314367312725e-05,
+      "loss": 0.6438,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7856648997464853,
+      "grad_norm": 0.14987516403198242,
+      "learning_rate": 1.157541845468469e-05,
+      "loss": 0.6383,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7858953675962204,
+      "grad_norm": 0.160349503159523,
+      "learning_rate": 1.1551543958996148e-05,
+      "loss": 0.6452,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7861258354459553,
+      "grad_norm": 0.1663265824317932,
+      "learning_rate": 1.1527690893554156e-05,
+      "loss": 0.6423,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7863563032956903,
+      "grad_norm": 0.15916943550109863,
+      "learning_rate": 1.1503859271653839e-05,
+      "loss": 0.6423,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7865867711454252,
+      "grad_norm": 0.16647377610206604,
+      "learning_rate": 1.1480049106578377e-05,
+      "loss": 0.652,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7868172389951602,
+      "grad_norm": 0.15861909091472626,
+      "learning_rate": 1.1456260411598984e-05,
+      "loss": 0.6446,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7870477068448951,
+      "grad_norm": 0.15791212022304535,
+      "learning_rate": 1.143249319997491e-05,
+      "loss": 0.6436,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7872781746946301,
+      "grad_norm": 0.17296838760375977,
+      "learning_rate": 1.1408747484953442e-05,
+      "loss": 0.6361,
+      "step": 3416
+    },
+    {
+      "epoch": 0.787508642544365,
+      "grad_norm": 0.16724546253681183,
+      "learning_rate": 1.138502327976984e-05,
+      "loss": 0.6409,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7877391103941,
+      "grad_norm": 0.1602526307106018,
+      "learning_rate": 1.1361320597647407e-05,
+      "loss": 0.636,
+      "step": 3418
+    },
+    {
+      "epoch": 0.787969578243835,
+      "grad_norm": 0.15437106788158417,
+      "learning_rate": 1.1337639451797494e-05,
+      "loss": 0.6457,
+      "step": 3419
+    },
+    {
+      "epoch": 0.78820004609357,
+      "grad_norm": 0.15734903514385223,
+      "learning_rate": 1.1313979855419359e-05,
+      "loss": 0.6389,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7884305139433049,
+      "grad_norm": 0.16602694988250732,
+      "learning_rate": 1.1290341821700313e-05,
+      "loss": 0.6401,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7886609817930399,
+      "grad_norm": 0.1661614030599594,
+      "learning_rate": 1.1266725363815623e-05,
+      "loss": 0.655,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7888914496427748,
+      "grad_norm": 0.1512596160173416,
+      "learning_rate": 1.1243130494928533e-05,
+      "loss": 0.6383,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7891219174925098,
+      "grad_norm": 0.14931842684745789,
+      "learning_rate": 1.1219557228190258e-05,
+      "loss": 0.6418,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7893523853422447,
+      "grad_norm": 0.1699947863817215,
+      "learning_rate": 1.1196005576739993e-05,
+      "loss": 0.6467,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7895828531919797,
+      "grad_norm": 0.1627425104379654,
+      "learning_rate": 1.1172475553704826e-05,
+      "loss": 0.6307,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7898133210417146,
+      "grad_norm": 0.17060711979866028,
+      "learning_rate": 1.1148967172199848e-05,
+      "loss": 0.652,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7900437888914497,
+      "grad_norm": 0.16025568544864655,
+      "learning_rate": 1.1125480445328057e-05,
+      "loss": 0.6434,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7902742567411846,
+      "grad_norm": 0.1679077297449112,
+      "learning_rate": 1.1102015386180404e-05,
+      "loss": 0.6428,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7905047245909196,
+      "grad_norm": 0.1602405458688736,
+      "learning_rate": 1.1078572007835735e-05,
+      "loss": 0.6387,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7907351924406545,
+      "grad_norm": 0.14988547563552856,
+      "learning_rate": 1.1055150323360852e-05,
+      "loss": 0.6398,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7909656602903895,
+      "grad_norm": 0.16209550201892853,
+      "learning_rate": 1.10317503458104e-05,
+      "loss": 0.6433,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7911961281401244,
+      "grad_norm": 0.16234451532363892,
+      "learning_rate": 1.1008372088226992e-05,
+      "loss": 0.6415,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7914265959898594,
+      "grad_norm": 0.14663785696029663,
+      "learning_rate": 1.098501556364112e-05,
+      "loss": 0.6472,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7916570638395943,
+      "grad_norm": 0.15545038878917694,
+      "learning_rate": 1.0961680785071116e-05,
+      "loss": 0.6441,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7918875316893293,
+      "grad_norm": 0.17046032845973969,
+      "learning_rate": 1.0938367765523244e-05,
+      "loss": 0.6392,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7921179995390643,
+      "grad_norm": 0.15083369612693787,
+      "learning_rate": 1.0915076517991617e-05,
+      "loss": 0.6356,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7923484673887993,
+      "grad_norm": 0.169847771525383,
+      "learning_rate": 1.0891807055458226e-05,
+      "loss": 0.6369,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7925789352385342,
+      "grad_norm": 0.16202600300312042,
+      "learning_rate": 1.0868559390892902e-05,
+      "loss": 0.6494,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7928094030882692,
+      "grad_norm": 0.1544782817363739,
+      "learning_rate": 1.0845333537253349e-05,
+      "loss": 0.6499,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7930398709380041,
+      "grad_norm": 0.15906357765197754,
+      "learning_rate": 1.082212950748508e-05,
+      "loss": 0.6448,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7932703387877391,
+      "grad_norm": 0.1550484299659729,
+      "learning_rate": 1.0798947314521468e-05,
+      "loss": 0.6422,
+      "step": 3442
+    },
+    {
+      "epoch": 0.793500806637474,
+      "grad_norm": 0.15442262589931488,
+      "learning_rate": 1.0775786971283725e-05,
+      "loss": 0.64,
+      "step": 3443
+    },
+    {
+      "epoch": 0.793731274487209,
+      "grad_norm": 0.1611274778842926,
+      "learning_rate": 1.0752648490680856e-05,
+      "loss": 0.642,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7939617423369439,
+      "grad_norm": 0.15969206392765045,
+      "learning_rate": 1.07295318856097e-05,
+      "loss": 0.641,
+      "step": 3445
+    },
+    {
+      "epoch": 0.794192210186679,
+      "grad_norm": 0.15087677538394928,
+      "learning_rate": 1.0706437168954913e-05,
+      "loss": 0.6345,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7944226780364139,
+      "grad_norm": 0.1505325436592102,
+      "learning_rate": 1.0683364353588898e-05,
+      "loss": 0.639,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7946531458861489,
+      "grad_norm": 0.16190530359745026,
+      "learning_rate": 1.0660313452371922e-05,
+      "loss": 0.6442,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7948836137358839,
+      "grad_norm": 0.15547344088554382,
+      "learning_rate": 1.063728447815201e-05,
+      "loss": 0.6475,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7951140815856188,
+      "grad_norm": 0.1529662311077118,
+      "learning_rate": 1.0614277443764925e-05,
+      "loss": 0.642,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7953445494353538,
+      "grad_norm": 0.15717142820358276,
+      "learning_rate": 1.0591292362034255e-05,
+      "loss": 0.6335,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7955750172850887,
+      "grad_norm": 0.15520958602428436,
+      "learning_rate": 1.0568329245771336e-05,
+      "loss": 0.6368,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7958054851348237,
+      "grad_norm": 0.16604743897914886,
+      "learning_rate": 1.054538810777525e-05,
+      "loss": 0.6456,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7960359529845586,
+      "grad_norm": 0.15997354686260223,
+      "learning_rate": 1.0522468960832842e-05,
+      "loss": 0.6438,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7962664208342937,
+      "grad_norm": 0.1553625613451004,
+      "learning_rate": 1.0499571817718707e-05,
+      "loss": 0.6391,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7964968886840286,
+      "grad_norm": 0.2511269748210907,
+      "learning_rate": 1.0476696691195138e-05,
+      "loss": 0.6387,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7967273565337636,
+      "grad_norm": 0.14997459948062897,
+      "learning_rate": 1.0453843594012175e-05,
+      "loss": 0.6303,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7969578243834985,
+      "grad_norm": 0.14934763312339783,
+      "learning_rate": 1.0431012538907631e-05,
+      "loss": 0.6377,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7971882922332335,
+      "grad_norm": 0.15451250970363617,
+      "learning_rate": 1.0408203538606948e-05,
+      "loss": 0.6484,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7974187600829684,
+      "grad_norm": 0.15496329963207245,
+      "learning_rate": 1.0385416605823323e-05,
+      "loss": 0.6356,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7976492279327034,
+      "grad_norm": 0.15835978090763092,
+      "learning_rate": 1.0362651753257668e-05,
+      "loss": 0.6446,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7978796957824383,
+      "grad_norm": 0.16143862903118134,
+      "learning_rate": 1.0339908993598518e-05,
+      "loss": 0.6367,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7981101636321734,
+      "grad_norm": 0.1462894082069397,
+      "learning_rate": 1.0317188339522188e-05,
+      "loss": 0.647,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7983406314819083,
+      "grad_norm": 0.14987273514270782,
+      "learning_rate": 1.029448980369262e-05,
+      "loss": 0.6468,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7985710993316433,
+      "grad_norm": 0.1489580124616623,
+      "learning_rate": 1.0271813398761405e-05,
+      "loss": 0.6474,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7988015671813782,
+      "grad_norm": 0.1617879867553711,
+      "learning_rate": 1.0249159137367842e-05,
+      "loss": 0.6429,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7990320350311132,
+      "grad_norm": 0.1757936030626297,
+      "learning_rate": 1.0226527032138878e-05,
+      "loss": 0.6353,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7992625028808481,
+      "grad_norm": 0.15790215134620667,
+      "learning_rate": 1.0203917095689097e-05,
+      "loss": 0.6442,
+      "step": 3468
+    },
+    {
+      "epoch": 0.7994929707305831,
+      "grad_norm": 0.155640110373497,
+      "learning_rate": 1.0181329340620743e-05,
+      "loss": 0.643,
+      "step": 3469
+    },
+    {
+      "epoch": 0.799723438580318,
+      "grad_norm": 0.16087128221988678,
+      "learning_rate": 1.0158763779523695e-05,
+      "loss": 0.6401,
+      "step": 3470
+    },
+    {
+      "epoch": 0.799953906430053,
+      "grad_norm": 0.16374313831329346,
+      "learning_rate": 1.0136220424975435e-05,
+      "loss": 0.6386,
+      "step": 3471
+    },
+    {
+      "epoch": 0.800184374279788,
+      "grad_norm": 0.16088031232357025,
+      "learning_rate": 1.011369928954108e-05,
+      "loss": 0.6392,
+      "step": 3472
+    },
+    {
+      "epoch": 0.800414842129523,
+      "grad_norm": 0.1503087878227234,
+      "learning_rate": 1.0091200385773408e-05,
+      "loss": 0.6402,
+      "step": 3473
+    },
+    {
+      "epoch": 0.8006453099792579,
+      "grad_norm": 0.16075649857521057,
+      "learning_rate": 1.0068723726212742e-05,
+      "loss": 0.6463,
+      "step": 3474
+    },
+    {
+      "epoch": 0.8008757778289929,
+      "grad_norm": 0.16603581607341766,
+      "learning_rate": 1.0046269323387036e-05,
+      "loss": 0.643,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8011062456787278,
+      "grad_norm": 0.15744850039482117,
+      "learning_rate": 1.0023837189811835e-05,
+      "loss": 0.6437,
+      "step": 3476
+    },
+    {
+      "epoch": 0.8013367135284628,
+      "grad_norm": 0.16175509989261627,
+      "learning_rate": 1.0001427337990277e-05,
+      "loss": 0.6434,
+      "step": 3477
+    },
+    {
+      "epoch": 0.8015671813781977,
+      "grad_norm": 0.17071296274662018,
+      "learning_rate": 9.979039780413068e-06,
+      "loss": 0.6449,
+      "step": 3478
+    },
+    {
+      "epoch": 0.8017976492279327,
+      "grad_norm": 0.15305480360984802,
+      "learning_rate": 9.956674529558518e-06,
+      "loss": 0.6427,
+      "step": 3479
+    },
+    {
+      "epoch": 0.8020281170776676,
+      "grad_norm": 0.16541430354118347,
+      "learning_rate": 9.934331597892448e-06,
+      "loss": 0.6406,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8022585849274027,
+      "grad_norm": 0.15421070158481598,
+      "learning_rate": 9.912010997868287e-06,
+      "loss": 0.6473,
+      "step": 3481
+    },
+    {
+      "epoch": 0.8024890527771376,
+      "grad_norm": 0.15333889424800873,
+      "learning_rate": 9.889712741926998e-06,
+      "loss": 0.6405,
+      "step": 3482
+    },
+    {
+      "epoch": 0.8027195206268726,
+      "grad_norm": 0.15218707919120789,
+      "learning_rate": 9.867436842497103e-06,
+      "loss": 0.632,
+      "step": 3483
+    },
+    {
+      "epoch": 0.8029499884766075,
+      "grad_norm": 0.15634161233901978,
+      "learning_rate": 9.845183311994637e-06,
+      "loss": 0.6378,
+      "step": 3484
+    },
+    {
+      "epoch": 0.8031804563263425,
+      "grad_norm": 0.157097727060318,
+      "learning_rate": 9.822952162823201e-06,
+      "loss": 0.6448,
+      "step": 3485
+    },
+    {
+      "epoch": 0.8034109241760774,
+      "grad_norm": 0.15480422973632812,
+      "learning_rate": 9.800743407373896e-06,
+      "loss": 0.6354,
+      "step": 3486
+    },
+    {
+      "epoch": 0.8036413920258124,
+      "grad_norm": 0.15486778318881989,
+      "learning_rate": 9.778557058025356e-06,
+      "loss": 0.6309,
+      "step": 3487
+    },
+    {
+      "epoch": 0.8038718598755473,
+      "grad_norm": 0.15796728432178497,
+      "learning_rate": 9.756393127143709e-06,
+      "loss": 0.6334,
+      "step": 3488
+    },
+    {
+      "epoch": 0.8041023277252823,
+      "grad_norm": 0.15152916312217712,
+      "learning_rate": 9.734251627082613e-06,
+      "loss": 0.6429,
+      "step": 3489
+    },
+    {
+      "epoch": 0.8043327955750172,
+      "grad_norm": 0.1551668345928192,
+      "learning_rate": 9.71213257018319e-06,
+      "loss": 0.6454,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8045632634247523,
+      "grad_norm": 0.16736368834972382,
+      "learning_rate": 9.69003596877408e-06,
+      "loss": 0.6362,
+      "step": 3491
+    },
+    {
+      "epoch": 0.8047937312744872,
+      "grad_norm": 0.14687344431877136,
+      "learning_rate": 9.667961835171402e-06,
+      "loss": 0.6366,
+      "step": 3492
+    },
+    {
+      "epoch": 0.8050241991242222,
+      "grad_norm": 0.1573791354894638,
+      "learning_rate": 9.645910181678741e-06,
+      "loss": 0.6338,
+      "step": 3493
+    },
+    {
+      "epoch": 0.8052546669739571,
+      "grad_norm": 0.1550455242395401,
+      "learning_rate": 9.62388102058716e-06,
+      "loss": 0.6381,
+      "step": 3494
+    },
+    {
+      "epoch": 0.8054851348236921,
+      "grad_norm": 0.15506303310394287,
+      "learning_rate": 9.601874364175206e-06,
+      "loss": 0.6391,
+      "step": 3495
+    },
+    {
+      "epoch": 0.805715602673427,
+      "grad_norm": 0.15035480260849,
+      "learning_rate": 9.579890224708826e-06,
+      "loss": 0.6455,
+      "step": 3496
+    },
+    {
+      "epoch": 0.805946070523162,
+      "grad_norm": 0.1552959680557251,
+      "learning_rate": 9.557928614441458e-06,
+      "loss": 0.641,
+      "step": 3497
+    },
+    {
+      "epoch": 0.8061765383728969,
+      "grad_norm": 0.17842282354831696,
+      "learning_rate": 9.535989545614016e-06,
+      "loss": 0.6346,
+      "step": 3498
+    },
+    {
+      "epoch": 0.806407006222632,
+      "grad_norm": 0.15210920572280884,
+      "learning_rate": 9.514073030454762e-06,
+      "loss": 0.6414,
+      "step": 3499
+    },
+    {
+      "epoch": 0.8066374740723669,
+      "grad_norm": 0.16027803719043732,
+      "learning_rate": 9.49217908117946e-06,
+      "loss": 0.6539,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8068679419221019,
+      "grad_norm": 0.1576579064130783,
+      "learning_rate": 9.470307709991267e-06,
+      "loss": 0.6397,
+      "step": 3501
+    },
+    {
+      "epoch": 0.8070984097718368,
+      "grad_norm": 0.15845666825771332,
+      "learning_rate": 9.448458929080756e-06,
+      "loss": 0.6395,
+      "step": 3502
+    },
+    {
+      "epoch": 0.8073288776215718,
+      "grad_norm": 0.15758542716503143,
+      "learning_rate": 9.426632750625918e-06,
+      "loss": 0.6437,
+      "step": 3503
+    },
+    {
+      "epoch": 0.8075593454713067,
+      "grad_norm": 0.15715070068836212,
+      "learning_rate": 9.404829186792152e-06,
+      "loss": 0.6347,
+      "step": 3504
+    },
+    {
+      "epoch": 0.8077898133210417,
+      "grad_norm": 0.15392762422561646,
+      "learning_rate": 9.383048249732217e-06,
+      "loss": 0.6407,
+      "step": 3505
+    },
+    {
+      "epoch": 0.8080202811707766,
+      "grad_norm": 0.15316665172576904,
+      "learning_rate": 9.3612899515863e-06,
+      "loss": 0.6394,
+      "step": 3506
+    },
+    {
+      "epoch": 0.8082507490205116,
+      "grad_norm": 0.15614305436611176,
+      "learning_rate": 9.339554304481951e-06,
+      "loss": 0.6348,
+      "step": 3507
+    },
+    {
+      "epoch": 0.8084812168702467,
+      "grad_norm": 0.16622847318649292,
+      "learning_rate": 9.317841320534092e-06,
+      "loss": 0.6378,
+      "step": 3508
+    },
+    {
+      "epoch": 0.8087116847199816,
+      "grad_norm": 0.1508837789297104,
+      "learning_rate": 9.296151011845034e-06,
+      "loss": 0.6428,
+      "step": 3509
+    },
+    {
+      "epoch": 0.8089421525697166,
+      "grad_norm": 0.15079739689826965,
+      "learning_rate": 9.27448339050443e-06,
+      "loss": 0.6462,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8091726204194515,
+      "grad_norm": 0.15400461852550507,
+      "learning_rate": 9.252838468589265e-06,
+      "loss": 0.6382,
+      "step": 3511
+    },
+    {
+      "epoch": 0.8094030882691865,
+      "grad_norm": 0.16339372098445892,
+      "learning_rate": 9.231216258163939e-06,
+      "loss": 0.6392,
+      "step": 3512
+    },
+    {
+      "epoch": 0.8096335561189214,
+      "grad_norm": 0.15596559643745422,
+      "learning_rate": 9.209616771280139e-06,
+      "loss": 0.6341,
+      "step": 3513
+    },
+    {
+      "epoch": 0.8098640239686564,
+      "grad_norm": 0.1570686399936676,
+      "learning_rate": 9.18804001997689e-06,
+      "loss": 0.6424,
+      "step": 3514
+    },
+    {
+      "epoch": 0.8100944918183913,
+      "grad_norm": 0.1516939401626587,
+      "learning_rate": 9.166486016280562e-06,
+      "loss": 0.6523,
+      "step": 3515
+    },
+    {
+      "epoch": 0.8103249596681263,
+      "grad_norm": 0.15133339166641235,
+      "learning_rate": 9.14495477220484e-06,
+      "loss": 0.6376,
+      "step": 3516
+    },
+    {
+      "epoch": 0.8105554275178612,
+      "grad_norm": 0.1549098789691925,
+      "learning_rate": 9.12344629975072e-06,
+      "loss": 0.6424,
+      "step": 3517
+    },
+    {
+      "epoch": 0.8107858953675963,
+      "grad_norm": 0.1573808342218399,
+      "learning_rate": 9.101960610906519e-06,
+      "loss": 0.6441,
+      "step": 3518
+    },
+    {
+      "epoch": 0.8110163632173312,
+      "grad_norm": 0.16286353766918182,
+      "learning_rate": 9.08049771764784e-06,
+      "loss": 0.639,
+      "step": 3519
+    },
+    {
+      "epoch": 0.8112468310670662,
+      "grad_norm": 0.15851277112960815,
+      "learning_rate": 9.059057631937567e-06,
+      "loss": 0.642,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8114772989168011,
+      "grad_norm": 0.15502890944480896,
+      "learning_rate": 9.037640365725897e-06,
+      "loss": 0.6326,
+      "step": 3521
+    },
+    {
+      "epoch": 0.8117077667665361,
+      "grad_norm": 0.15994207561016083,
+      "learning_rate": 9.01624593095033e-06,
+      "loss": 0.6448,
+      "step": 3522
+    },
+    {
+      "epoch": 0.811938234616271,
+      "grad_norm": 0.1541290283203125,
+      "learning_rate": 8.994874339535569e-06,
+      "loss": 0.6481,
+      "step": 3523
+    },
+    {
+      "epoch": 0.812168702466006,
+      "grad_norm": 0.14977748692035675,
+      "learning_rate": 8.973525603393645e-06,
+      "loss": 0.6415,
+      "step": 3524
+    },
+    {
+      "epoch": 0.8123991703157409,
+      "grad_norm": 0.15313945710659027,
+      "learning_rate": 8.952199734423843e-06,
+      "loss": 0.6383,
+      "step": 3525
+    },
+    {
+      "epoch": 0.812629638165476,
+      "grad_norm": 0.15545286238193512,
+      "learning_rate": 8.930896744512652e-06,
+      "loss": 0.6438,
+      "step": 3526
+    },
+    {
+      "epoch": 0.8128601060152109,
+      "grad_norm": 0.15130577981472015,
+      "learning_rate": 8.909616645533886e-06,
+      "loss": 0.6466,
+      "step": 3527
+    },
+    {
+      "epoch": 0.8130905738649459,
+      "grad_norm": 0.15956194698810577,
+      "learning_rate": 8.888359449348555e-06,
+      "loss": 0.6353,
+      "step": 3528
+    },
+    {
+      "epoch": 0.8133210417146808,
+      "grad_norm": 0.15058216452598572,
+      "learning_rate": 8.867125167804896e-06,
+      "loss": 0.6374,
+      "step": 3529
+    },
+    {
+      "epoch": 0.8135515095644158,
+      "grad_norm": 0.14475852251052856,
+      "learning_rate": 8.845913812738394e-06,
+      "loss": 0.6442,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8137819774141507,
+      "grad_norm": 0.14834396541118622,
+      "learning_rate": 8.824725395971745e-06,
+      "loss": 0.6469,
+      "step": 3531
+    },
+    {
+      "epoch": 0.8140124452638857,
+      "grad_norm": 0.15428373217582703,
+      "learning_rate": 8.803559929314869e-06,
+      "loss": 0.6378,
+      "step": 3532
+    },
+    {
+      "epoch": 0.8142429131136206,
+      "grad_norm": 0.15115933120250702,
+      "learning_rate": 8.782417424564893e-06,
+      "loss": 0.6415,
+      "step": 3533
+    },
+    {
+      "epoch": 0.8144733809633556,
+      "grad_norm": 0.14465634524822235,
+      "learning_rate": 8.761297893506149e-06,
+      "loss": 0.6432,
+      "step": 3534
+    },
+    {
+      "epoch": 0.8147038488130905,
+      "grad_norm": 0.15278743207454681,
+      "learning_rate": 8.740201347910132e-06,
+      "loss": 0.6305,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8149343166628256,
+      "grad_norm": 0.15596990287303925,
+      "learning_rate": 8.719127799535547e-06,
+      "loss": 0.6451,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8151647845125605,
+      "grad_norm": 0.15721940994262695,
+      "learning_rate": 8.698077260128329e-06,
+      "loss": 0.636,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8153952523622955,
+      "grad_norm": 0.15862002968788147,
+      "learning_rate": 8.677049741421506e-06,
+      "loss": 0.643,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8156257202120304,
+      "grad_norm": 0.15502506494522095,
+      "learning_rate": 8.656045255135314e-06,
+      "loss": 0.641,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8158561880617654,
+      "grad_norm": 0.1488857865333557,
+      "learning_rate": 8.635063812977156e-06,
+      "loss": 0.6389,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8160866559115003,
+      "grad_norm": 0.1511356681585312,
+      "learning_rate": 8.61410542664159e-06,
+      "loss": 0.6383,
+      "step": 3541
+    },
+    {
+      "epoch": 0.8163171237612353,
+      "grad_norm": 0.15279202163219452,
+      "learning_rate": 8.593170107810312e-06,
+      "loss": 0.633,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8165475916109702,
+      "grad_norm": 0.1624889373779297,
+      "learning_rate": 8.572257868152172e-06,
+      "loss": 0.6456,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8167780594607053,
+      "grad_norm": 0.15917298197746277,
+      "learning_rate": 8.551368719323139e-06,
+      "loss": 0.6379,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8170085273104402,
+      "grad_norm": 0.16097372770309448,
+      "learning_rate": 8.530502672966328e-06,
+      "loss": 0.6475,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8172389951601752,
+      "grad_norm": 0.14943112432956696,
+      "learning_rate": 8.509659740711973e-06,
+      "loss": 0.646,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8174694630099101,
+      "grad_norm": 0.15080755949020386,
+      "learning_rate": 8.488839934177422e-06,
+      "loss": 0.6347,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8176999308596451,
+      "grad_norm": 0.16167816519737244,
+      "learning_rate": 8.46804326496714e-06,
+      "loss": 0.6356,
+      "step": 3548
+    },
+    {
+      "epoch": 0.81793039870938,
+      "grad_norm": 0.15981729328632355,
+      "learning_rate": 8.447269744672703e-06,
+      "loss": 0.6338,
+      "step": 3549
+    },
+    {
+      "epoch": 0.818160866559115,
+      "grad_norm": 0.1530756950378418,
+      "learning_rate": 8.426519384872733e-06,
+      "loss": 0.6392,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8183913344088499,
+      "grad_norm": 0.1670352965593338,
+      "learning_rate": 8.405792197133022e-06,
+      "loss": 0.6457,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8186218022585849,
+      "grad_norm": 0.1554369479417801,
+      "learning_rate": 8.385088193006407e-06,
+      "loss": 0.6376,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8188522701083198,
+      "grad_norm": 0.15348638594150543,
+      "learning_rate": 8.364407384032775e-06,
+      "loss": 0.6355,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8190827379580549,
+      "grad_norm": 0.15805724263191223,
+      "learning_rate": 8.343749781739125e-06,
+      "loss": 0.6443,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8193132058077898,
+      "grad_norm": 0.15739752352237701,
+      "learning_rate": 8.323115397639513e-06,
+      "loss": 0.6445,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8195436736575248,
+      "grad_norm": 0.16260041296482086,
+      "learning_rate": 8.302504243235043e-06,
+      "loss": 0.6431,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8197741415072597,
+      "grad_norm": 0.14841555058956146,
+      "learning_rate": 8.281916330013889e-06,
+      "loss": 0.6307,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8200046093569947,
+      "grad_norm": 0.15895935893058777,
+      "learning_rate": 8.261351669451256e-06,
+      "loss": 0.6315,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8202350772067296,
+      "grad_norm": 0.14655686914920807,
+      "learning_rate": 8.240810273009381e-06,
+      "loss": 0.6334,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8204655450564646,
+      "grad_norm": 0.1525675356388092,
+      "learning_rate": 8.220292152137554e-06,
+      "loss": 0.6474,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8206960129061995,
+      "grad_norm": 0.15452326834201813,
+      "learning_rate": 8.199797318272085e-06,
+      "loss": 0.6402,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8209264807559346,
+      "grad_norm": 0.15377068519592285,
+      "learning_rate": 8.179325782836295e-06,
+      "loss": 0.6416,
+      "step": 3562
+    },
+    {
+      "epoch": 0.8211569486056695,
+      "grad_norm": 0.15333302319049835,
+      "learning_rate": 8.158877557240529e-06,
+      "loss": 0.6398,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8213874164554045,
+      "grad_norm": 0.14731299877166748,
+      "learning_rate": 8.138452652882156e-06,
+      "loss": 0.6459,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8216178843051394,
+      "grad_norm": 0.15332041680812836,
+      "learning_rate": 8.118051081145484e-06,
+      "loss": 0.6408,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8218483521548744,
+      "grad_norm": 0.15887399017810822,
+      "learning_rate": 8.097672853401894e-06,
+      "loss": 0.643,
+      "step": 3566
+    },
+    {
+      "epoch": 0.8220788200046094,
+      "grad_norm": 0.14689992368221283,
+      "learning_rate": 8.07731798100973e-06,
+      "loss": 0.6382,
+      "step": 3567
+    },
+    {
+      "epoch": 0.8223092878543443,
+      "grad_norm": 0.15063437819480896,
+      "learning_rate": 8.056986475314283e-06,
+      "loss": 0.6366,
+      "step": 3568
+    },
+    {
+      "epoch": 0.8225397557040793,
+      "grad_norm": 0.15097613632678986,
+      "learning_rate": 8.036678347647853e-06,
+      "loss": 0.6489,
+      "step": 3569
+    },
+    {
+      "epoch": 0.8227702235538142,
+      "grad_norm": 0.14924323558807373,
+      "learning_rate": 8.016393609329703e-06,
+      "loss": 0.6397,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8230006914035493,
+      "grad_norm": 0.14785458147525787,
+      "learning_rate": 7.996132271666062e-06,
+      "loss": 0.6405,
+      "step": 3571
+    },
+    {
+      "epoch": 0.8232311592532842,
+      "grad_norm": 0.1523989886045456,
+      "learning_rate": 7.975894345950114e-06,
+      "loss": 0.6478,
+      "step": 3572
+    },
+    {
+      "epoch": 0.8234616271030192,
+      "grad_norm": 0.152211531996727,
+      "learning_rate": 7.955679843462005e-06,
+      "loss": 0.6379,
+      "step": 3573
+    },
+    {
+      "epoch": 0.8236920949527541,
+      "grad_norm": 0.14990997314453125,
+      "learning_rate": 7.935488775468791e-06,
+      "loss": 0.6435,
+      "step": 3574
+    },
+    {
+      "epoch": 0.8239225628024891,
+      "grad_norm": 0.1513289213180542,
+      "learning_rate": 7.915321153224487e-06,
+      "loss": 0.6429,
+      "step": 3575
+    },
+    {
+      "epoch": 0.824153030652224,
+      "grad_norm": 0.150692880153656,
+      "learning_rate": 7.895176987970071e-06,
+      "loss": 0.6436,
+      "step": 3576
+    },
+    {
+      "epoch": 0.824383498501959,
+      "grad_norm": 0.14454329013824463,
+      "learning_rate": 7.875056290933391e-06,
+      "loss": 0.6294,
+      "step": 3577
+    },
+    {
+      "epoch": 0.8246139663516939,
+      "grad_norm": 0.14884406328201294,
+      "learning_rate": 7.85495907332925e-06,
+      "loss": 0.639,
+      "step": 3578
+    },
+    {
+      "epoch": 0.8248444342014289,
+      "grad_norm": 0.15509894490242004,
+      "learning_rate": 7.83488534635936e-06,
+      "loss": 0.6402,
+      "step": 3579
+    },
+    {
+      "epoch": 0.8250749020511638,
+      "grad_norm": 0.15603148937225342,
+      "learning_rate": 7.814835121212305e-06,
+      "loss": 0.6336,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8253053699008989,
+      "grad_norm": 0.1463710367679596,
+      "learning_rate": 7.794808409063636e-06,
+      "loss": 0.6454,
+      "step": 3581
+    },
+    {
+      "epoch": 0.8255358377506338,
+      "grad_norm": 0.15044938027858734,
+      "learning_rate": 7.774805221075764e-06,
+      "loss": 0.6402,
+      "step": 3582
+    },
+    {
+      "epoch": 0.8257663056003688,
+      "grad_norm": 0.1514042317867279,
+      "learning_rate": 7.754825568397955e-06,
+      "loss": 0.6413,
+      "step": 3583
+    },
+    {
+      "epoch": 0.8259967734501037,
+      "grad_norm": 0.15234024822711945,
+      "learning_rate": 7.7348694621664e-06,
+      "loss": 0.6444,
+      "step": 3584
+    },
+    {
+      "epoch": 0.8262272412998387,
+      "grad_norm": 0.14573800563812256,
+      "learning_rate": 7.71493691350416e-06,
+      "loss": 0.6449,
+      "step": 3585
+    },
+    {
+      "epoch": 0.8264577091495736,
+      "grad_norm": 0.14887578785419464,
+      "learning_rate": 7.695027933521154e-06,
+      "loss": 0.6321,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8266881769993086,
+      "grad_norm": 0.15681131184101105,
+      "learning_rate": 7.675142533314172e-06,
+      "loss": 0.6428,
+      "step": 3587
+    },
+    {
+      "epoch": 0.8269186448490435,
+      "grad_norm": 0.15508340299129486,
+      "learning_rate": 7.655280723966868e-06,
+      "loss": 0.6383,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8271491126987786,
+      "grad_norm": 0.15364862978458405,
+      "learning_rate": 7.635442516549719e-06,
+      "loss": 0.636,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8273795805485135,
+      "grad_norm": 0.1485067903995514,
+      "learning_rate": 7.615627922120056e-06,
+      "loss": 0.6457,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8276100483982485,
+      "grad_norm": 0.14897732436656952,
+      "learning_rate": 7.5958369517221075e-06,
+      "loss": 0.6437,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8278405162479834,
+      "grad_norm": 0.1528944969177246,
+      "learning_rate": 7.576069616386838e-06,
+      "loss": 0.6447,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8280709840977184,
+      "grad_norm": 0.15931028127670288,
+      "learning_rate": 7.556325927132102e-06,
+      "loss": 0.6421,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8283014519474533,
+      "grad_norm": 0.15149252116680145,
+      "learning_rate": 7.536605894962562e-06,
+      "loss": 0.6398,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8285319197971883,
+      "grad_norm": 0.14745914936065674,
+      "learning_rate": 7.5169095308696865e-06,
+      "loss": 0.65,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8287623876469232,
+      "grad_norm": 0.1486995816230774,
+      "learning_rate": 7.497236845831751e-06,
+      "loss": 0.6366,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8289928554966582,
+      "grad_norm": 0.14624127745628357,
+      "learning_rate": 7.4775878508138606e-06,
+      "loss": 0.6405,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8292233233463931,
+      "grad_norm": 0.1482323855161667,
+      "learning_rate": 7.457962556767866e-06,
+      "loss": 0.6405,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8294537911961282,
+      "grad_norm": 0.14901240170001984,
+      "learning_rate": 7.438360974632441e-06,
+      "loss": 0.6402,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8296842590458631,
+      "grad_norm": 0.15405641496181488,
+      "learning_rate": 7.418783115333045e-06,
+      "loss": 0.6439,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8299147268955981,
+      "grad_norm": 0.14017413556575775,
+      "learning_rate": 7.399228989781898e-06,
+      "loss": 0.6411,
+      "step": 3601
+    },
+    {
+      "epoch": 0.830145194745333,
+      "grad_norm": 0.14487187564373016,
+      "learning_rate": 7.3796986088780105e-06,
+      "loss": 0.6315,
+      "step": 3602
+    },
+    {
+      "epoch": 0.830375662595068,
+      "grad_norm": 0.15042652189731598,
+      "learning_rate": 7.360191983507153e-06,
+      "loss": 0.6384,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8306061304448029,
+      "grad_norm": 0.15280528366565704,
+      "learning_rate": 7.340709124541817e-06,
+      "loss": 0.6446,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8308365982945379,
+      "grad_norm": 0.1426979899406433,
+      "learning_rate": 7.321250042841316e-06,
+      "loss": 0.6448,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8310670661442728,
+      "grad_norm": 0.14999282360076904,
+      "learning_rate": 7.301814749251679e-06,
+      "loss": 0.6387,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8312975339940079,
+      "grad_norm": 0.151020810008049,
+      "learning_rate": 7.2824032546056354e-06,
+      "loss": 0.6489,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8315280018437428,
+      "grad_norm": 0.15103960037231445,
+      "learning_rate": 7.2630155697227146e-06,
+      "loss": 0.6448,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8317584696934778,
+      "grad_norm": 0.14460238814353943,
+      "learning_rate": 7.243651705409132e-06,
+      "loss": 0.6409,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8319889375432127,
+      "grad_norm": 0.14994823932647705,
+      "learning_rate": 7.2243116724578565e-06,
+      "loss": 0.6336,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8322194053929477,
+      "grad_norm": 0.15003249049186707,
+      "learning_rate": 7.2049954816485465e-06,
+      "loss": 0.6397,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8324498732426826,
+      "grad_norm": 0.15217861533164978,
+      "learning_rate": 7.185703143747596e-06,
+      "loss": 0.6425,
+      "step": 3612
+    },
+    {
+      "epoch": 0.8326803410924176,
+      "grad_norm": 0.1463039368391037,
+      "learning_rate": 7.166434669508071e-06,
+      "loss": 0.6341,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8329108089421525,
+      "grad_norm": 0.1474071443080902,
+      "learning_rate": 7.147190069669768e-06,
+      "loss": 0.6397,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8331412767918875,
+      "grad_norm": 0.14856012165546417,
+      "learning_rate": 7.127969354959163e-06,
+      "loss": 0.6444,
+      "step": 3615
+    },
+    {
+      "epoch": 0.8333717446416224,
+      "grad_norm": 0.147612527012825,
+      "learning_rate": 7.1087725360894195e-06,
+      "loss": 0.6347,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8336022124913575,
+      "grad_norm": 0.15374769270420074,
+      "learning_rate": 7.08959962376039e-06,
+      "loss": 0.6435,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8338326803410924,
+      "grad_norm": 0.14860646426677704,
+      "learning_rate": 7.070450628658592e-06,
+      "loss": 0.646,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8340631481908274,
+      "grad_norm": 0.1480385959148407,
+      "learning_rate": 7.051325561457217e-06,
+      "loss": 0.6384,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8342936160405623,
+      "grad_norm": 0.1520072966814041,
+      "learning_rate": 7.0322244328161144e-06,
+      "loss": 0.6458,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8345240838902973,
+      "grad_norm": 0.14834314584732056,
+      "learning_rate": 7.013147253381797e-06,
+      "loss": 0.6341,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8347545517400322,
+      "grad_norm": 0.14721745252609253,
+      "learning_rate": 6.99409403378744e-06,
+      "loss": 0.6417,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8349850195897672,
+      "grad_norm": 0.1415952891111374,
+      "learning_rate": 6.975064784652829e-06,
+      "loss": 0.6443,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8352154874395021,
+      "grad_norm": 0.14462348818778992,
+      "learning_rate": 6.9560595165844175e-06,
+      "loss": 0.633,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8354459552892372,
+      "grad_norm": 0.1447225958108902,
+      "learning_rate": 6.937078240175287e-06,
+      "loss": 0.6442,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8356764231389722,
+      "grad_norm": 0.1561073660850525,
+      "learning_rate": 6.918120966005148e-06,
+      "loss": 0.6468,
+      "step": 3626
+    },
+    {
+      "epoch": 0.8359068909887071,
+      "grad_norm": 0.1541508585214615,
+      "learning_rate": 6.899187704640325e-06,
+      "loss": 0.6415,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8361373588384421,
+      "grad_norm": 0.15150463581085205,
+      "learning_rate": 6.880278466633783e-06,
+      "loss": 0.6413,
+      "step": 3628
+    },
+    {
+      "epoch": 0.836367826688177,
+      "grad_norm": 0.14670462906360626,
+      "learning_rate": 6.861393262525035e-06,
+      "loss": 0.6488,
+      "step": 3629
+    },
+    {
+      "epoch": 0.836598294537912,
+      "grad_norm": 0.14568495750427246,
+      "learning_rate": 6.842532102840277e-06,
+      "loss": 0.6333,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8368287623876469,
+      "grad_norm": 0.15337829291820526,
+      "learning_rate": 6.823694998092273e-06,
+      "loss": 0.6408,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8370592302373819,
+      "grad_norm": 0.14713677763938904,
+      "learning_rate": 6.80488195878034e-06,
+      "loss": 0.6367,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8372896980871168,
+      "grad_norm": 0.14200221002101898,
+      "learning_rate": 6.786092995390436e-06,
+      "loss": 0.6303,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8375201659368519,
+      "grad_norm": 0.15284350514411926,
+      "learning_rate": 6.7673281183950665e-06,
+      "loss": 0.6438,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8377506337865868,
+      "grad_norm": 0.147255539894104,
+      "learning_rate": 6.748587338253337e-06,
+      "loss": 0.643,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8379811016363218,
+      "grad_norm": 0.14654171466827393,
+      "learning_rate": 6.729870665410898e-06,
+      "loss": 0.6404,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8382115694860567,
+      "grad_norm": 0.15386711061000824,
+      "learning_rate": 6.711178110299993e-06,
+      "loss": 0.6394,
+      "step": 3637
+    },
+    {
+      "epoch": 0.8384420373357917,
+      "grad_norm": 0.15187443792819977,
+      "learning_rate": 6.692509683339371e-06,
+      "loss": 0.6411,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8386725051855266,
+      "grad_norm": 0.14688163995742798,
+      "learning_rate": 6.673865394934376e-06,
+      "loss": 0.6341,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8389029730352616,
+      "grad_norm": 0.15309026837348938,
+      "learning_rate": 6.655245255476911e-06,
+      "loss": 0.6373,
+      "step": 3640
+    },
+    {
+      "epoch": 0.8391334408849965,
+      "grad_norm": 0.15117040276527405,
+      "learning_rate": 6.6366492753453695e-06,
+      "loss": 0.6554,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8393639087347315,
+      "grad_norm": 0.15723955631256104,
+      "learning_rate": 6.61807746490471e-06,
+      "loss": 0.6349,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8395943765844665,
+      "grad_norm": 0.15363402664661407,
+      "learning_rate": 6.59952983450643e-06,
+      "loss": 0.641,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8398248444342015,
+      "grad_norm": 0.14676257967948914,
+      "learning_rate": 6.581006394488493e-06,
+      "loss": 0.6338,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8400553122839364,
+      "grad_norm": 0.14586247503757477,
+      "learning_rate": 6.562507155175457e-06,
+      "loss": 0.6316,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8402857801336714,
+      "grad_norm": 0.1537342369556427,
+      "learning_rate": 6.544032126878358e-06,
+      "loss": 0.6407,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8405162479834063,
+      "grad_norm": 0.14989496767520905,
+      "learning_rate": 6.525581319894703e-06,
+      "loss": 0.6309,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8407467158331413,
+      "grad_norm": 0.15396440029144287,
+      "learning_rate": 6.507154744508548e-06,
+      "loss": 0.6486,
+      "step": 3648
+    },
+    {
+      "epoch": 0.8409771836828762,
+      "grad_norm": 0.14274229109287262,
+      "learning_rate": 6.488752410990417e-06,
+      "loss": 0.6309,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8412076515326112,
+      "grad_norm": 0.14844496548175812,
+      "learning_rate": 6.470374329597334e-06,
+      "loss": 0.6343,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8414381193823461,
+      "grad_norm": 0.15259870886802673,
+      "learning_rate": 6.452020510572798e-06,
+      "loss": 0.6426,
+      "step": 3651
+    },
+    {
+      "epoch": 0.8416685872320812,
+      "grad_norm": 0.1534135341644287,
+      "learning_rate": 6.433690964146799e-06,
+      "loss": 0.6334,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8418990550818161,
+      "grad_norm": 0.14689849317073822,
+      "learning_rate": 6.415385700535764e-06,
+      "loss": 0.6429,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8421295229315511,
+      "grad_norm": 0.14878277480602264,
+      "learning_rate": 6.39710472994261e-06,
+      "loss": 0.6417,
+      "step": 3654
+    },
+    {
+      "epoch": 0.842359990781286,
+      "grad_norm": 0.1506912261247635,
+      "learning_rate": 6.378848062556741e-06,
+      "loss": 0.6453,
+      "step": 3655
+    },
+    {
+      "epoch": 0.842590458631021,
+      "grad_norm": 0.15941332280635834,
+      "learning_rate": 6.360615708553952e-06,
+      "loss": 0.6397,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8428209264807559,
+      "grad_norm": 0.15402281284332275,
+      "learning_rate": 6.342407678096534e-06,
+      "loss": 0.6413,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8430513943304909,
+      "grad_norm": 0.14595894515514374,
+      "learning_rate": 6.324223981333199e-06,
+      "loss": 0.6448,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8432818621802258,
+      "grad_norm": 0.16061711311340332,
+      "learning_rate": 6.3060646283991106e-06,
+      "loss": 0.646,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8435123300299608,
+      "grad_norm": 0.16146208345890045,
+      "learning_rate": 6.287929629415856e-06,
+      "loss": 0.6377,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8437427978796957,
+      "grad_norm": 0.15115495026111603,
+      "learning_rate": 6.269818994491455e-06,
+      "loss": 0.6467,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8439732657294308,
+      "grad_norm": 0.13822171092033386,
+      "learning_rate": 6.251732733720323e-06,
+      "loss": 0.639,
+      "step": 3662
+    },
+    {
+      "epoch": 0.8442037335791657,
+      "grad_norm": 0.14701151847839355,
+      "learning_rate": 6.23367085718331e-06,
+      "loss": 0.6439,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8444342014289007,
+      "grad_norm": 0.1508350670337677,
+      "learning_rate": 6.215633374947683e-06,
+      "loss": 0.6387,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8446646692786356,
+      "grad_norm": 0.15383285284042358,
+      "learning_rate": 6.197620297067097e-06,
+      "loss": 0.6405,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8448951371283706,
+      "grad_norm": 0.1498384326696396,
+      "learning_rate": 6.179631633581612e-06,
+      "loss": 0.6461,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8451256049781055,
+      "grad_norm": 0.15178535878658295,
+      "learning_rate": 6.1616673945176836e-06,
+      "loss": 0.6454,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8453560728278405,
+      "grad_norm": 0.15506048500537872,
+      "learning_rate": 6.143727589888126e-06,
+      "loss": 0.6449,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8455865406775754,
+      "grad_norm": 0.1453298181295395,
+      "learning_rate": 6.125812229692162e-06,
+      "loss": 0.6417,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8458170085273105,
+      "grad_norm": 0.1514149308204651,
+      "learning_rate": 6.107921323915411e-06,
+      "loss": 0.6433,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8460474763770454,
+      "grad_norm": 0.14425182342529297,
+      "learning_rate": 6.0900548825298e-06,
+      "loss": 0.6461,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8462779442267804,
+      "grad_norm": 0.1488070785999298,
+      "learning_rate": 6.072212915493669e-06,
+      "loss": 0.6473,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8465084120765153,
+      "grad_norm": 0.14930947124958038,
+      "learning_rate": 6.054395432751703e-06,
+      "loss": 0.6418,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8467388799262503,
+      "grad_norm": 0.15142978727817535,
+      "learning_rate": 6.036602444234935e-06,
+      "loss": 0.638,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8469693477759852,
+      "grad_norm": 0.13954748213291168,
+      "learning_rate": 6.018833959860753e-06,
+      "loss": 0.6415,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8471998156257202,
+      "grad_norm": 0.14654095470905304,
+      "learning_rate": 6.001089989532893e-06,
+      "loss": 0.6398,
+      "step": 3676
+    },
+    {
+      "epoch": 0.8474302834754551,
+      "grad_norm": 0.14327792823314667,
+      "learning_rate": 5.9833705431413975e-06,
+      "loss": 0.6393,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8476607513251901,
+      "grad_norm": 0.15117095410823822,
+      "learning_rate": 5.965675630562672e-06,
+      "loss": 0.6375,
+      "step": 3678
+    },
+    {
+      "epoch": 0.847891219174925,
+      "grad_norm": 0.14697414636611938,
+      "learning_rate": 5.948005261659434e-06,
+      "loss": 0.6385,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8481216870246601,
+      "grad_norm": 0.14706385135650635,
+      "learning_rate": 5.930359446280726e-06,
+      "loss": 0.6391,
+      "step": 3680
+    },
+    {
+      "epoch": 0.848352154874395,
+      "grad_norm": 0.1475333720445633,
+      "learning_rate": 5.912738194261902e-06,
+      "loss": 0.636,
+      "step": 3681
+    },
+    {
+      "epoch": 0.84858262272413,
+      "grad_norm": 0.15489518642425537,
+      "learning_rate": 5.895141515424629e-06,
+      "loss": 0.6409,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8488130905738649,
+      "grad_norm": 0.1442023515701294,
+      "learning_rate": 5.87756941957685e-06,
+      "loss": 0.6374,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8490435584235999,
+      "grad_norm": 0.14666606485843658,
+      "learning_rate": 5.860021916512859e-06,
+      "loss": 0.6388,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8492740262733348,
+      "grad_norm": 0.1419772058725357,
+      "learning_rate": 5.842499016013209e-06,
+      "loss": 0.6369,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8495044941230698,
+      "grad_norm": 0.14429770410060883,
+      "learning_rate": 5.8250007278447205e-06,
+      "loss": 0.637,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8497349619728048,
+      "grad_norm": 0.15203547477722168,
+      "learning_rate": 5.807527061760543e-06,
+      "loss": 0.6346,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8499654298225398,
+      "grad_norm": 0.14554864168167114,
+      "learning_rate": 5.790078027500068e-06,
+      "loss": 0.6423,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8501958976722748,
+      "grad_norm": 0.15379323065280914,
+      "learning_rate": 5.772653634788971e-06,
+      "loss": 0.645,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8504263655220097,
+      "grad_norm": 0.15001268684864044,
+      "learning_rate": 5.755253893339185e-06,
+      "loss": 0.6349,
+      "step": 3690
+    },
+    {
+      "epoch": 0.8506568333717447,
+      "grad_norm": 0.15877242386341095,
+      "learning_rate": 5.737878812848929e-06,
+      "loss": 0.6364,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8508873012214796,
+      "grad_norm": 0.15002503991127014,
+      "learning_rate": 5.720528403002634e-06,
+      "loss": 0.6319,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8511177690712146,
+      "grad_norm": 0.1416318565607071,
+      "learning_rate": 5.703202673470992e-06,
+      "loss": 0.6432,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8513482369209495,
+      "grad_norm": 0.15145903825759888,
+      "learning_rate": 5.685901633910989e-06,
+      "loss": 0.6373,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8515787047706845,
+      "grad_norm": 0.1428060680627823,
+      "learning_rate": 5.668625293965774e-06,
+      "loss": 0.6411,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8518091726204194,
+      "grad_norm": 0.1449238806962967,
+      "learning_rate": 5.6513736632647695e-06,
+      "loss": 0.6353,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8520396404701545,
+      "grad_norm": 0.15147364139556885,
+      "learning_rate": 5.634146751423647e-06,
+      "loss": 0.6433,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8522701083198894,
+      "grad_norm": 0.14954763650894165,
+      "learning_rate": 5.616944568044225e-06,
+      "loss": 0.6279,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8525005761696244,
+      "grad_norm": 0.14778868854045868,
+      "learning_rate": 5.599767122714627e-06,
+      "loss": 0.6368,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8527310440193593,
+      "grad_norm": 0.14133252203464508,
+      "learning_rate": 5.582614425009153e-06,
+      "loss": 0.6416,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8529615118690943,
+      "grad_norm": 0.15429161489009857,
+      "learning_rate": 5.565486484488275e-06,
+      "loss": 0.6418,
+      "step": 3701
+    },
+    {
+      "epoch": 0.8531919797188292,
+      "grad_norm": 0.1481894999742508,
+      "learning_rate": 5.548383310698707e-06,
+      "loss": 0.635,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8534224475685642,
+      "grad_norm": 0.14938104152679443,
+      "learning_rate": 5.531304913173357e-06,
+      "loss": 0.6503,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8536529154182991,
+      "grad_norm": 0.14781972765922546,
+      "learning_rate": 5.514251301431306e-06,
+      "loss": 0.6352,
+      "step": 3704
+    },
+    {
+      "epoch": 0.8538833832680341,
+      "grad_norm": 0.1436912715435028,
+      "learning_rate": 5.497222484977826e-06,
+      "loss": 0.6331,
+      "step": 3705
+    },
+    {
+      "epoch": 0.854113851117769,
+      "grad_norm": 0.14232562482357025,
+      "learning_rate": 5.480218473304388e-06,
+      "loss": 0.6355,
+      "step": 3706
+    },
+    {
+      "epoch": 0.8543443189675041,
+      "grad_norm": 0.1478080451488495,
+      "learning_rate": 5.4632392758885985e-06,
+      "loss": 0.6371,
+      "step": 3707
+    },
+    {
+      "epoch": 0.854574786817239,
+      "grad_norm": 0.14770181477069855,
+      "learning_rate": 5.446284902194249e-06,
+      "loss": 0.6365,
+      "step": 3708
+    },
+    {
+      "epoch": 0.854805254666974,
+      "grad_norm": 0.14597497880458832,
+      "learning_rate": 5.429355361671335e-06,
+      "loss": 0.641,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8550357225167089,
+      "grad_norm": 0.1463644802570343,
+      "learning_rate": 5.412450663755941e-06,
+      "loss": 0.6343,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8552661903664439,
+      "grad_norm": 0.14388582110404968,
+      "learning_rate": 5.39557081787036e-06,
+      "loss": 0.6475,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8554966582161788,
+      "grad_norm": 0.1477576494216919,
+      "learning_rate": 5.378715833423004e-06,
+      "loss": 0.6357,
+      "step": 3712
+    },
+    {
+      "epoch": 0.8557271260659138,
+      "grad_norm": 0.15977248549461365,
+      "learning_rate": 5.36188571980844e-06,
+      "loss": 0.6389,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8559575939156487,
+      "grad_norm": 0.14352811872959137,
+      "learning_rate": 5.3450804864073665e-06,
+      "loss": 0.6389,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8561880617653838,
+      "grad_norm": 0.13531458377838135,
+      "learning_rate": 5.328300142586629e-06,
+      "loss": 0.6385,
+      "step": 3715
+    },
+    {
+      "epoch": 0.8564185296151187,
+      "grad_norm": 0.15272001922130585,
+      "learning_rate": 5.311544697699172e-06,
+      "loss": 0.635,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8566489974648537,
+      "grad_norm": 0.15149231255054474,
+      "learning_rate": 5.294814161084083e-06,
+      "loss": 0.634,
+      "step": 3717
+    },
+    {
+      "epoch": 0.8568794653145886,
+      "grad_norm": 0.13878074288368225,
+      "learning_rate": 5.278108542066562e-06,
+      "loss": 0.6344,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8571099331643236,
+      "grad_norm": 0.143055722117424,
+      "learning_rate": 5.261427849957928e-06,
+      "loss": 0.6459,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8573404010140585,
+      "grad_norm": 0.14342109858989716,
+      "learning_rate": 5.244772094055589e-06,
+      "loss": 0.6344,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8575708688637935,
+      "grad_norm": 0.1506274938583374,
+      "learning_rate": 5.228141283643073e-06,
+      "loss": 0.6441,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8578013367135284,
+      "grad_norm": 0.13922744989395142,
+      "learning_rate": 5.211535427989972e-06,
+      "loss": 0.6469,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8580318045632634,
+      "grad_norm": 0.13600236177444458,
+      "learning_rate": 5.194954536352021e-06,
+      "loss": 0.6365,
+      "step": 3723
+    },
+    {
+      "epoch": 0.8582622724129984,
+      "grad_norm": 0.140496164560318,
+      "learning_rate": 5.178398617971003e-06,
+      "loss": 0.6401,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8584927402627334,
+      "grad_norm": 0.14220167696475983,
+      "learning_rate": 5.161867682074773e-06,
+      "loss": 0.6382,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8587232081124683,
+      "grad_norm": 0.14413121342658997,
+      "learning_rate": 5.145361737877291e-06,
+      "loss": 0.6251,
+      "step": 3726
+    },
+    {
+      "epoch": 0.8589536759622033,
+      "grad_norm": 0.14119820296764374,
+      "learning_rate": 5.128880794578572e-06,
+      "loss": 0.6368,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8591841438119382,
+      "grad_norm": 0.14928089082241058,
+      "learning_rate": 5.112424861364701e-06,
+      "loss": 0.6405,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8594146116616732,
+      "grad_norm": 0.1459737867116928,
+      "learning_rate": 5.095993947407818e-06,
+      "loss": 0.6333,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8596450795114081,
+      "grad_norm": 0.1493435502052307,
+      "learning_rate": 5.079588061866125e-06,
+      "loss": 0.635,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8598755473611431,
+      "grad_norm": 0.14926058053970337,
+      "learning_rate": 5.0632072138838584e-06,
+      "loss": 0.6535,
+      "step": 3731
+    },
+    {
+      "epoch": 0.860106015210878,
+      "grad_norm": 0.14926740527153015,
+      "learning_rate": 5.046851412591314e-06,
+      "loss": 0.6416,
+      "step": 3732
+    },
+    {
+      "epoch": 0.860336483060613,
+      "grad_norm": 0.148090198636055,
+      "learning_rate": 5.030520667104821e-06,
+      "loss": 0.637,
+      "step": 3733
+    },
+    {
+      "epoch": 0.860566950910348,
+      "grad_norm": 0.14074482023715973,
+      "learning_rate": 5.01421498652675e-06,
+      "loss": 0.6391,
+      "step": 3734
+    },
+    {
+      "epoch": 0.860797418760083,
+      "grad_norm": 0.14252926409244537,
+      "learning_rate": 4.997934379945491e-06,
+      "loss": 0.638,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8610278866098179,
+      "grad_norm": 0.14792828261852264,
+      "learning_rate": 4.981678856435479e-06,
+      "loss": 0.6389,
+      "step": 3736
+    },
+    {
+      "epoch": 0.8612583544595529,
+      "grad_norm": 0.15630586445331573,
+      "learning_rate": 4.965448425057118e-06,
+      "loss": 0.6465,
+      "step": 3737
+    },
+    {
+      "epoch": 0.8614888223092878,
+      "grad_norm": 0.14542563259601593,
+      "learning_rate": 4.949243094856892e-06,
+      "loss": 0.64,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8617192901590228,
+      "grad_norm": 0.1466866284608841,
+      "learning_rate": 4.933062874867267e-06,
+      "loss": 0.6441,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8619497580087577,
+      "grad_norm": 0.14459478855133057,
+      "learning_rate": 4.916907774106683e-06,
+      "loss": 0.6341,
+      "step": 3740
+    },
+    {
+      "epoch": 0.8621802258584927,
+      "grad_norm": 0.152002215385437,
+      "learning_rate": 4.90077780157962e-06,
+      "loss": 0.6486,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8624106937082276,
+      "grad_norm": 0.14948220551013947,
+      "learning_rate": 4.884672966276538e-06,
+      "loss": 0.6398,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8626411615579627,
+      "grad_norm": 0.14576594531536102,
+      "learning_rate": 4.868593277173878e-06,
+      "loss": 0.6509,
+      "step": 3743
+    },
+    {
+      "epoch": 0.8628716294076976,
+      "grad_norm": 0.1429005265235901,
+      "learning_rate": 4.852538743234081e-06,
+      "loss": 0.6391,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8631020972574326,
+      "grad_norm": 0.145622119307518,
+      "learning_rate": 4.836509373405568e-06,
+      "loss": 0.6404,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8633325651071676,
+      "grad_norm": 0.1396528035402298,
+      "learning_rate": 4.820505176622697e-06,
+      "loss": 0.6299,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8635630329569025,
+      "grad_norm": 0.14236707985401154,
+      "learning_rate": 4.804526161805833e-06,
+      "loss": 0.6399,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8637935008066375,
+      "grad_norm": 0.14210587739944458,
+      "learning_rate": 4.788572337861313e-06,
+      "loss": 0.6344,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8640239686563724,
+      "grad_norm": 0.1455959975719452,
+      "learning_rate": 4.772643713681413e-06,
+      "loss": 0.6388,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8642544365061074,
+      "grad_norm": 0.14806394279003143,
+      "learning_rate": 4.756740298144346e-06,
+      "loss": 0.6385,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8644849043558424,
+      "grad_norm": 0.13723453879356384,
+      "learning_rate": 4.740862100114307e-06,
+      "loss": 0.6385,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8647153722055774,
+      "grad_norm": 0.15280765295028687,
+      "learning_rate": 4.725009128441421e-06,
+      "loss": 0.6454,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8649458400553123,
+      "grad_norm": 0.14481478929519653,
+      "learning_rate": 4.709181391961753e-06,
+      "loss": 0.6413,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8651763079050473,
+      "grad_norm": 0.14188295602798462,
+      "learning_rate": 4.693378899497303e-06,
+      "loss": 0.6528,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8654067757547822,
+      "grad_norm": 0.1486392468214035,
+      "learning_rate": 4.6776016598560124e-06,
+      "loss": 0.6364,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8656372436045172,
+      "grad_norm": 0.1420852243900299,
+      "learning_rate": 4.6618496818317145e-06,
+      "loss": 0.6371,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8658677114542521,
+      "grad_norm": 0.1463935822248459,
+      "learning_rate": 4.646122974204187e-06,
+      "loss": 0.6368,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8660981793039871,
+      "grad_norm": 0.14199309051036835,
+      "learning_rate": 4.630421545739144e-06,
+      "loss": 0.6407,
+      "step": 3758
+    },
+    {
+      "epoch": 0.866328647153722,
+      "grad_norm": 0.13984660804271698,
+      "learning_rate": 4.6147454051881585e-06,
+      "loss": 0.6389,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8665591150034571,
+      "grad_norm": 0.14572177827358246,
+      "learning_rate": 4.5990945612887415e-06,
+      "loss": 0.6353,
+      "step": 3760
+    },
+    {
+      "epoch": 0.866789582853192,
+      "grad_norm": 0.14711903035640717,
+      "learning_rate": 4.583469022764314e-06,
+      "loss": 0.6408,
+      "step": 3761
+    },
+    {
+      "epoch": 0.867020050702927,
+      "grad_norm": 0.14538122713565826,
+      "learning_rate": 4.567868798324143e-06,
+      "loss": 0.6373,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8672505185526619,
+      "grad_norm": 0.14264172315597534,
+      "learning_rate": 4.552293896663451e-06,
+      "loss": 0.6357,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8674809864023969,
+      "grad_norm": 0.14745338261127472,
+      "learning_rate": 4.536744326463304e-06,
+      "loss": 0.6415,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8677114542521318,
+      "grad_norm": 0.1417376846075058,
+      "learning_rate": 4.521220096390655e-06,
+      "loss": 0.6299,
+      "step": 3765
+    },
+    {
+      "epoch": 0.8679419221018668,
+      "grad_norm": 0.13753369450569153,
+      "learning_rate": 4.505721215098335e-06,
+      "loss": 0.6364,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8681723899516017,
+      "grad_norm": 0.14812487363815308,
+      "learning_rate": 4.490247691225058e-06,
+      "loss": 0.6494,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8684028578013367,
+      "grad_norm": 0.14765094220638275,
+      "learning_rate": 4.4747995333953855e-06,
+      "loss": 0.6363,
+      "step": 3768
+    },
+    {
+      "epoch": 0.8686333256510717,
+      "grad_norm": 0.14122313261032104,
+      "learning_rate": 4.459376750219757e-06,
+      "loss": 0.6319,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8688637935008067,
+      "grad_norm": 0.13697822391986847,
+      "learning_rate": 4.443979350294463e-06,
+      "loss": 0.6328,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8690942613505416,
+      "grad_norm": 0.1433285027742386,
+      "learning_rate": 4.428607342201635e-06,
+      "loss": 0.643,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8693247292002766,
+      "grad_norm": 0.13931584358215332,
+      "learning_rate": 4.4132607345092555e-06,
+      "loss": 0.6451,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8695551970500115,
+      "grad_norm": 0.1486472487449646,
+      "learning_rate": 4.397939535771189e-06,
+      "loss": 0.6352,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8697856648997465,
+      "grad_norm": 0.14353391528129578,
+      "learning_rate": 4.382643754527072e-06,
+      "loss": 0.6599,
+      "step": 3774
+    },
+    {
+      "epoch": 0.8700161327494814,
+      "grad_norm": 0.14892083406448364,
+      "learning_rate": 4.36737339930241e-06,
+      "loss": 0.6315,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8702466005992164,
+      "grad_norm": 0.1448989063501358,
+      "learning_rate": 4.352128478608541e-06,
+      "loss": 0.6396,
+      "step": 3776
+    },
+    {
+      "epoch": 0.8704770684489513,
+      "grad_norm": 0.1473805457353592,
+      "learning_rate": 4.3369090009426185e-06,
+      "loss": 0.6484,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8707075362986864,
+      "grad_norm": 0.14433009922504425,
+      "learning_rate": 4.321714974787605e-06,
+      "loss": 0.6427,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8709380041484213,
+      "grad_norm": 0.14538750052452087,
+      "learning_rate": 4.306546408612306e-06,
+      "loss": 0.6443,
+      "step": 3779
+    },
+    {
+      "epoch": 0.8711684719981563,
+      "grad_norm": 0.13906364142894745,
+      "learning_rate": 4.291403310871284e-06,
+      "loss": 0.6368,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8713989398478912,
+      "grad_norm": 0.15096677839756012,
+      "learning_rate": 4.276285690004961e-06,
+      "loss": 0.6428,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8716294076976262,
+      "grad_norm": 0.14877794682979584,
+      "learning_rate": 4.26119355443953e-06,
+      "loss": 0.6374,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8718598755473611,
+      "grad_norm": 0.14585071802139282,
+      "learning_rate": 4.24612691258699e-06,
+      "loss": 0.6383,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8720903433970961,
+      "grad_norm": 0.14376723766326904,
+      "learning_rate": 4.231085772845117e-06,
+      "loss": 0.6423,
+      "step": 3784
+    },
+    {
+      "epoch": 0.872320811246831,
+      "grad_norm": 0.1447305977344513,
+      "learning_rate": 4.2160701435974945e-06,
+      "loss": 0.6369,
+      "step": 3785
+    },
+    {
+      "epoch": 0.872551279096566,
+      "grad_norm": 0.14197836816310883,
+      "learning_rate": 4.201080033213456e-06,
+      "loss": 0.6373,
+      "step": 3786
+    },
+    {
+      "epoch": 0.872781746946301,
+      "grad_norm": 0.14709553122520447,
+      "learning_rate": 4.186115450048128e-06,
+      "loss": 0.6385,
+      "step": 3787
+    },
+    {
+      "epoch": 0.873012214796036,
+      "grad_norm": 0.1427144557237625,
+      "learning_rate": 4.171176402442445e-06,
+      "loss": 0.6398,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8732426826457709,
+      "grad_norm": 0.14940115809440613,
+      "learning_rate": 4.156262898723034e-06,
+      "loss": 0.6333,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8734731504955059,
+      "grad_norm": 0.1392568051815033,
+      "learning_rate": 4.141374947202336e-06,
+      "loss": 0.6419,
+      "step": 3790
+    },
+    {
+      "epoch": 0.8737036183452408,
+      "grad_norm": 0.1471542865037918,
+      "learning_rate": 4.1265125561785465e-06,
+      "loss": 0.6299,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8739340861949758,
+      "grad_norm": 0.14388377964496613,
+      "learning_rate": 4.1116757339355995e-06,
+      "loss": 0.6431,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8741645540447107,
+      "grad_norm": 0.1433270126581192,
+      "learning_rate": 4.0968644887431795e-06,
+      "loss": 0.6342,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8743950218944457,
+      "grad_norm": 0.13722054660320282,
+      "learning_rate": 4.082078828856733e-06,
+      "loss": 0.6373,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8746254897441806,
+      "grad_norm": 0.14640505611896515,
+      "learning_rate": 4.0673187625174195e-06,
+      "loss": 0.6395,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8748559575939157,
+      "grad_norm": 0.1450256109237671,
+      "learning_rate": 4.052584297952145e-06,
+      "loss": 0.6384,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8750864254436506,
+      "grad_norm": 0.14310050010681152,
+      "learning_rate": 4.037875443373546e-06,
+      "loss": 0.6361,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8753168932933856,
+      "grad_norm": 0.142232283949852,
+      "learning_rate": 4.023192206979992e-06,
+      "loss": 0.6378,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8755473611431205,
+      "grad_norm": 0.14268440008163452,
+      "learning_rate": 4.008534596955565e-06,
+      "loss": 0.6358,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8757778289928555,
+      "grad_norm": 0.14434240758419037,
+      "learning_rate": 3.9939026214700695e-06,
+      "loss": 0.6394,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8760082968425904,
+      "grad_norm": 0.14181116223335266,
+      "learning_rate": 3.979296288678996e-06,
+      "loss": 0.6414,
+      "step": 3801
+    },
+    {
+      "epoch": 0.8762387646923254,
+      "grad_norm": 0.14815394580364227,
+      "learning_rate": 3.964715606723585e-06,
+      "loss": 0.6415,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8764692325420603,
+      "grad_norm": 0.14605048298835754,
+      "learning_rate": 3.950160583730761e-06,
+      "loss": 0.6353,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8766997003917953,
+      "grad_norm": 0.14189860224723816,
+      "learning_rate": 3.93563122781313e-06,
+      "loss": 0.6441,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8769301682415304,
+      "grad_norm": 0.14701896905899048,
+      "learning_rate": 3.921127547069014e-06,
+      "loss": 0.6457,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8771606360912653,
+      "grad_norm": 0.141061931848526,
+      "learning_rate": 3.906649549582414e-06,
+      "loss": 0.6324,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8773911039410003,
+      "grad_norm": 0.14271359145641327,
+      "learning_rate": 3.8921972434230185e-06,
+      "loss": 0.6447,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8776215717907352,
+      "grad_norm": 0.1448899656534195,
+      "learning_rate": 3.8777706366462e-06,
+      "loss": 0.6401,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8778520396404702,
+      "grad_norm": 0.13925962150096893,
+      "learning_rate": 3.863369737293005e-06,
+      "loss": 0.6411,
+      "step": 3809
+    },
+    {
+      "epoch": 0.8780825074902051,
+      "grad_norm": 0.1407134234905243,
+      "learning_rate": 3.848994553390134e-06,
+      "loss": 0.6445,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8783129753399401,
+      "grad_norm": 0.1439376175403595,
+      "learning_rate": 3.834645092949973e-06,
+      "loss": 0.6367,
+      "step": 3811
+    },
+    {
+      "epoch": 0.878543443189675,
+      "grad_norm": 0.14084570109844208,
+      "learning_rate": 3.8203213639705915e-06,
+      "loss": 0.6304,
+      "step": 3812
+    },
+    {
+      "epoch": 0.87877391103941,
+      "grad_norm": 0.14097703993320465,
+      "learning_rate": 3.8060233744356633e-06,
+      "loss": 0.638,
+      "step": 3813
+    },
+    {
+      "epoch": 0.879004378889145,
+      "grad_norm": 0.1431102752685547,
+      "learning_rate": 3.7917511323145584e-06,
+      "loss": 0.6446,
+      "step": 3814
+    },
+    {
+      "epoch": 0.87923484673888,
+      "grad_norm": 0.14385107159614563,
+      "learning_rate": 3.7775046455622855e-06,
+      "loss": 0.6343,
+      "step": 3815
+    },
+    {
+      "epoch": 0.8794653145886149,
+      "grad_norm": 0.14835472404956818,
+      "learning_rate": 3.7632839221194706e-06,
+      "loss": 0.6442,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8796957824383499,
+      "grad_norm": 0.1427791565656662,
+      "learning_rate": 3.749088969912429e-06,
+      "loss": 0.6434,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8799262502880848,
+      "grad_norm": 0.13777343928813934,
+      "learning_rate": 3.734919796853087e-06,
+      "loss": 0.6434,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8801567181378198,
+      "grad_norm": 0.15572573244571686,
+      "learning_rate": 3.720776410838983e-06,
+      "loss": 0.6361,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8803871859875547,
+      "grad_norm": 0.14312800765037537,
+      "learning_rate": 3.7066588197533115e-06,
+      "loss": 0.6331,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8806176538372897,
+      "grad_norm": 0.14728614687919617,
+      "learning_rate": 3.6925670314648775e-06,
+      "loss": 0.6391,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8808481216870246,
+      "grad_norm": 0.14271116256713867,
+      "learning_rate": 3.6785010538281093e-06,
+      "loss": 0.6359,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8810785895367597,
+      "grad_norm": 0.14498567581176758,
+      "learning_rate": 3.664460894683036e-06,
+      "loss": 0.6383,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8813090573864946,
+      "grad_norm": 0.14503896236419678,
+      "learning_rate": 3.650446561855325e-06,
+      "loss": 0.6293,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8815395252362296,
+      "grad_norm": 0.14249297976493835,
+      "learning_rate": 3.6364580631562063e-06,
+      "loss": 0.6329,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8817699930859645,
+      "grad_norm": 0.14754772186279297,
+      "learning_rate": 3.622495406382531e-06,
+      "loss": 0.6412,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8820004609356995,
+      "grad_norm": 0.17180697619915009,
+      "learning_rate": 3.6085585993167805e-06,
+      "loss": 0.6378,
+      "step": 3827
+    },
+    {
+      "epoch": 0.8822309287854344,
+      "grad_norm": 0.14113938808441162,
+      "learning_rate": 3.594647649726962e-06,
+      "loss": 0.6425,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8824613966351694,
+      "grad_norm": 0.1466577649116516,
+      "learning_rate": 3.5807625653667243e-06,
+      "loss": 0.6484,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8826918644849043,
+      "grad_norm": 0.14053227007389069,
+      "learning_rate": 3.566903353975276e-06,
+      "loss": 0.6409,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8829223323346393,
+      "grad_norm": 0.14258506894111633,
+      "learning_rate": 3.553070023277405e-06,
+      "loss": 0.6373,
+      "step": 3831
+    },
+    {
+      "epoch": 0.8831528001843743,
+      "grad_norm": 0.1399742215871811,
+      "learning_rate": 3.5392625809834823e-06,
+      "loss": 0.6461,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8833832680341093,
+      "grad_norm": 0.13795405626296997,
+      "learning_rate": 3.525481034789446e-06,
+      "loss": 0.6329,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8836137358838442,
+      "grad_norm": 0.14222726225852966,
+      "learning_rate": 3.5117253923767967e-06,
+      "loss": 0.6454,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8838442037335792,
+      "grad_norm": 0.14100506901741028,
+      "learning_rate": 3.4979956614125953e-06,
+      "loss": 0.6318,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8840746715833141,
+      "grad_norm": 0.13950549066066742,
+      "learning_rate": 3.4842918495494646e-06,
+      "loss": 0.6348,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8843051394330491,
+      "grad_norm": 0.17627927660942078,
+      "learning_rate": 3.4706139644255897e-06,
+      "loss": 0.639,
+      "step": 3837
+    },
+    {
+      "epoch": 0.884535607282784,
+      "grad_norm": 0.1371656060218811,
+      "learning_rate": 3.4569620136646886e-06,
+      "loss": 0.64,
+      "step": 3838
+    },
+    {
+      "epoch": 0.884766075132519,
+      "grad_norm": 0.13862690329551697,
+      "learning_rate": 3.4433360048760357e-06,
+      "loss": 0.6335,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8849965429822539,
+      "grad_norm": 0.14692457020282745,
+      "learning_rate": 3.4297359456544276e-06,
+      "loss": 0.629,
+      "step": 3840
+    },
+    {
+      "epoch": 0.885227010831989,
+      "grad_norm": 0.13511526584625244,
+      "learning_rate": 3.4161618435802233e-06,
+      "loss": 0.6377,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8854574786817239,
+      "grad_norm": 0.14511561393737793,
+      "learning_rate": 3.4026137062193097e-06,
+      "loss": 0.645,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8856879465314589,
+      "grad_norm": 0.1404416412115097,
+      "learning_rate": 3.389091541123074e-06,
+      "loss": 0.6393,
+      "step": 3843
+    },
+    {
+      "epoch": 0.8859184143811938,
+      "grad_norm": 0.14536434412002563,
+      "learning_rate": 3.375595355828454e-06,
+      "loss": 0.6376,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8861488822309288,
+      "grad_norm": 0.14059731364250183,
+      "learning_rate": 3.362125157857904e-06,
+      "loss": 0.6335,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8863793500806637,
+      "grad_norm": 0.14244748651981354,
+      "learning_rate": 3.34868095471938e-06,
+      "loss": 0.6391,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8866098179303987,
+      "grad_norm": 0.13621114194393158,
+      "learning_rate": 3.335262753906371e-06,
+      "loss": 0.6436,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8868402857801336,
+      "grad_norm": 0.14168091118335724,
+      "learning_rate": 3.3218705628978554e-06,
+      "loss": 0.6382,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8870707536298686,
+      "grad_norm": 0.14067873358726501,
+      "learning_rate": 3.3085043891583125e-06,
+      "loss": 0.6332,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8873012214796036,
+      "grad_norm": 0.14555051922798157,
+      "learning_rate": 3.295164240137727e-06,
+      "loss": 0.6424,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8875316893293386,
+      "grad_norm": 0.14118191599845886,
+      "learning_rate": 3.2818501232715794e-06,
+      "loss": 0.6465,
+      "step": 3851
+    },
+    {
+      "epoch": 0.8877621571790735,
+      "grad_norm": 0.13933710753917694,
+      "learning_rate": 3.268562045980844e-06,
+      "loss": 0.6429,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8879926250288085,
+      "grad_norm": 0.1409381926059723,
+      "learning_rate": 3.2553000156719747e-06,
+      "loss": 0.6372,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8882230928785434,
+      "grad_norm": 0.13979056477546692,
+      "learning_rate": 3.242064039736914e-06,
+      "loss": 0.6342,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8884535607282784,
+      "grad_norm": 0.14258621633052826,
+      "learning_rate": 3.2288541255530545e-06,
+      "loss": 0.6412,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8886840285780133,
+      "grad_norm": 0.1503862887620926,
+      "learning_rate": 3.215670280483307e-06,
+      "loss": 0.6461,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8889144964277483,
+      "grad_norm": 0.14357925951480865,
+      "learning_rate": 3.202512511876038e-06,
+      "loss": 0.6381,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8891449642774832,
+      "grad_norm": 0.1408717930316925,
+      "learning_rate": 3.189380827065047e-06,
+      "loss": 0.6363,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8893754321272183,
+      "grad_norm": 0.14140430092811584,
+      "learning_rate": 3.1762752333696297e-06,
+      "loss": 0.6376,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8896058999769532,
+      "grad_norm": 0.14138497412204742,
+      "learning_rate": 3.163195738094532e-06,
+      "loss": 0.6396,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8898363678266882,
+      "grad_norm": 0.14000347256660461,
+      "learning_rate": 3.150142348529955e-06,
+      "loss": 0.6386,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8900668356764231,
+      "grad_norm": 0.14265434443950653,
+      "learning_rate": 3.1371150719515354e-06,
+      "loss": 0.6347,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8902973035261581,
+      "grad_norm": 0.14289766550064087,
+      "learning_rate": 3.1241139156203746e-06,
+      "loss": 0.6392,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8905277713758931,
+      "grad_norm": 0.14235663414001465,
+      "learning_rate": 3.111138886782994e-06,
+      "loss": 0.6365,
+      "step": 3864
+    },
+    {
+      "epoch": 0.890758239225628,
+      "grad_norm": 0.1410188376903534,
+      "learning_rate": 3.0981899926713574e-06,
+      "loss": 0.6476,
+      "step": 3865
+    },
+    {
+      "epoch": 0.890988707075363,
+      "grad_norm": 0.1419987976551056,
+      "learning_rate": 3.0852672405028984e-06,
+      "loss": 0.6372,
+      "step": 3866
+    },
+    {
+      "epoch": 0.891219174925098,
+      "grad_norm": 0.14122051000595093,
+      "learning_rate": 3.072370637480415e-06,
+      "loss": 0.6286,
+      "step": 3867
+    },
+    {
+      "epoch": 0.891449642774833,
+      "grad_norm": 0.13950307667255402,
+      "learning_rate": 3.059500190792186e-06,
+      "loss": 0.6365,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8916801106245679,
+      "grad_norm": 0.1390613466501236,
+      "learning_rate": 3.0466559076118837e-06,
+      "loss": 0.639,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8919105784743029,
+      "grad_norm": 0.14016972482204437,
+      "learning_rate": 3.0338377950985875e-06,
+      "loss": 0.6386,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8921410463240378,
+      "grad_norm": 0.14392688870429993,
+      "learning_rate": 3.0210458603968263e-06,
+      "loss": 0.6441,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8923715141737728,
+      "grad_norm": 0.13809806108474731,
+      "learning_rate": 3.0082801106365187e-06,
+      "loss": 0.6364,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8926019820235077,
+      "grad_norm": 0.14696134626865387,
+      "learning_rate": 2.995540552932974e-06,
+      "loss": 0.6413,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8928324498732427,
+      "grad_norm": 0.14118346571922302,
+      "learning_rate": 2.982827194386917e-06,
+      "loss": 0.6405,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8930629177229776,
+      "grad_norm": 0.14888674020767212,
+      "learning_rate": 2.9701400420844737e-06,
+      "loss": 0.6413,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8932933855727127,
+      "grad_norm": 0.14854340255260468,
+      "learning_rate": 2.9574791030971604e-06,
+      "loss": 0.6442,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8935238534224476,
+      "grad_norm": 0.1380283385515213,
+      "learning_rate": 2.944844384481871e-06,
+      "loss": 0.6342,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8937543212721826,
+      "grad_norm": 0.13916511833667755,
+      "learning_rate": 2.9322358932809157e-06,
+      "loss": 0.6426,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8939847891219175,
+      "grad_norm": 0.1366412490606308,
+      "learning_rate": 2.919653636521935e-06,
+      "loss": 0.6418,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8942152569716525,
+      "grad_norm": 0.13759376108646393,
+      "learning_rate": 2.907097621217986e-06,
+      "loss": 0.6409,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8944457248213874,
+      "grad_norm": 0.1421494036912918,
+      "learning_rate": 2.894567854367508e-06,
+      "loss": 0.6452,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8946761926711224,
+      "grad_norm": 0.13909095525741577,
+      "learning_rate": 2.8820643429542825e-06,
+      "loss": 0.6435,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8949066605208573,
+      "grad_norm": 0.1391996145248413,
+      "learning_rate": 2.8695870939474624e-06,
+      "loss": 0.6409,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8951371283705923,
+      "grad_norm": 0.13931603729724884,
+      "learning_rate": 2.857136114301562e-06,
+      "loss": 0.6327,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8953675962203272,
+      "grad_norm": 0.1370822936296463,
+      "learning_rate": 2.844711410956469e-06,
+      "loss": 0.642,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8955980640700623,
+      "grad_norm": 0.14227797091007233,
+      "learning_rate": 2.83231299083741e-06,
+      "loss": 0.6278,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8958285319197972,
+      "grad_norm": 0.1375594437122345,
+      "learning_rate": 2.8199408608549695e-06,
+      "loss": 0.6328,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8960589997695322,
+      "grad_norm": 0.14295582473278046,
+      "learning_rate": 2.8075950279050855e-06,
+      "loss": 0.646,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8962894676192671,
+      "grad_norm": 0.13916680216789246,
+      "learning_rate": 2.7952754988690046e-06,
+      "loss": 0.647,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8965199354690021,
+      "grad_norm": 0.1418418288230896,
+      "learning_rate": 2.782982280613344e-06,
+      "loss": 0.6377,
+      "step": 3890
+    },
+    {
+      "epoch": 0.896750403318737,
+      "grad_norm": 0.1433732807636261,
+      "learning_rate": 2.770715379990069e-06,
+      "loss": 0.6426,
+      "step": 3891
+    },
+    {
+      "epoch": 0.896980871168472,
+      "grad_norm": 0.1423547863960266,
+      "learning_rate": 2.7584748038364303e-06,
+      "loss": 0.635,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8972113390182069,
+      "grad_norm": 0.1434488594532013,
+      "learning_rate": 2.7462605589750443e-06,
+      "loss": 0.6382,
+      "step": 3893
+    },
+    {
+      "epoch": 0.897441806867942,
+      "grad_norm": 0.14350055158138275,
+      "learning_rate": 2.734072652213837e-06,
+      "loss": 0.6414,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8976722747176769,
+      "grad_norm": 0.13868124783039093,
+      "learning_rate": 2.7219110903460523e-06,
+      "loss": 0.6353,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8979027425674119,
+      "grad_norm": 0.13998155295848846,
+      "learning_rate": 2.7097758801502506e-06,
+      "loss": 0.6369,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8981332104171468,
+      "grad_norm": 0.1414719521999359,
+      "learning_rate": 2.6976670283903215e-06,
+      "loss": 0.6281,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8983636782668818,
+      "grad_norm": 0.13889610767364502,
+      "learning_rate": 2.685584541815428e-06,
+      "loss": 0.6447,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8985941461166167,
+      "grad_norm": 0.14231473207473755,
+      "learning_rate": 2.6735284271600657e-06,
+      "loss": 0.6355,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8988246139663517,
+      "grad_norm": 0.1469811350107193,
+      "learning_rate": 2.6614986911440264e-06,
+      "loss": 0.6374,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8990550818160866,
+      "grad_norm": 0.13894438743591309,
+      "learning_rate": 2.6494953404723965e-06,
+      "loss": 0.6362,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8992855496658216,
+      "grad_norm": 0.1419995129108429,
+      "learning_rate": 2.637518381835552e-06,
+      "loss": 0.6354,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8995160175155565,
+      "grad_norm": 0.14471600949764252,
+      "learning_rate": 2.625567821909175e-06,
+      "loss": 0.6406,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8997464853652916,
+      "grad_norm": 0.1410285234451294,
+      "learning_rate": 2.6136436673541986e-06,
+      "loss": 0.6399,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8999769532150265,
+      "grad_norm": 0.14543531835079193,
+      "learning_rate": 2.601745924816862e-06,
+      "loss": 0.6349,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9002074210647615,
+      "grad_norm": 0.14382287859916687,
+      "learning_rate": 2.58987460092871e-06,
+      "loss": 0.6331,
+      "step": 3906
+    },
+    {
+      "epoch": 0.9004378889144964,
+      "grad_norm": 0.14252907037734985,
+      "learning_rate": 2.5780297023065057e-06,
+      "loss": 0.6367,
+      "step": 3907
+    },
+    {
+      "epoch": 0.9006683567642314,
+      "grad_norm": 0.1392444670200348,
+      "learning_rate": 2.5662112355523183e-06,
+      "loss": 0.6349,
+      "step": 3908
+    },
+    {
+      "epoch": 0.9008988246139663,
+      "grad_norm": 0.13947899639606476,
+      "learning_rate": 2.5544192072534835e-06,
+      "loss": 0.6431,
+      "step": 3909
+    },
+    {
+      "epoch": 0.9011292924637013,
+      "grad_norm": 0.14244791865348816,
+      "learning_rate": 2.542653623982588e-06,
+      "loss": 0.6393,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9013597603134362,
+      "grad_norm": 0.1407129168510437,
+      "learning_rate": 2.530914492297487e-06,
+      "loss": 0.637,
+      "step": 3911
+    },
+    {
+      "epoch": 0.9015902281631712,
+      "grad_norm": 0.13949137926101685,
+      "learning_rate": 2.519201818741301e-06,
+      "loss": 0.6359,
+      "step": 3912
+    },
+    {
+      "epoch": 0.9018206960129062,
+      "grad_norm": 0.13872455060482025,
+      "learning_rate": 2.507515609842376e-06,
+      "loss": 0.6415,
+      "step": 3913
+    },
+    {
+      "epoch": 0.9020511638626412,
+      "grad_norm": 0.13934583961963654,
+      "learning_rate": 2.495855872114333e-06,
+      "loss": 0.6298,
+      "step": 3914
+    },
+    {
+      "epoch": 0.9022816317123761,
+      "grad_norm": 0.13912633061408997,
+      "learning_rate": 2.4842226120560255e-06,
+      "loss": 0.6321,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9025120995621111,
+      "grad_norm": 0.1382398009300232,
+      "learning_rate": 2.4726158361515593e-06,
+      "loss": 0.6343,
+      "step": 3916
+    },
+    {
+      "epoch": 0.902742567411846,
+      "grad_norm": 0.14063194394111633,
+      "learning_rate": 2.461035550870272e-06,
+      "loss": 0.6505,
+      "step": 3917
+    },
+    {
+      "epoch": 0.902973035261581,
+      "grad_norm": 0.14703617990016937,
+      "learning_rate": 2.4494817626667442e-06,
+      "loss": 0.64,
+      "step": 3918
+    },
+    {
+      "epoch": 0.9032035031113159,
+      "grad_norm": 0.14321446418762207,
+      "learning_rate": 2.437954477980753e-06,
+      "loss": 0.6332,
+      "step": 3919
+    },
+    {
+      "epoch": 0.9034339709610509,
+      "grad_norm": 0.14084553718566895,
+      "learning_rate": 2.426453703237358e-06,
+      "loss": 0.6363,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9036644388107858,
+      "grad_norm": 0.14175479114055634,
+      "learning_rate": 2.41497944484681e-06,
+      "loss": 0.6319,
+      "step": 3921
+    },
+    {
+      "epoch": 0.9038949066605209,
+      "grad_norm": 0.13902176916599274,
+      "learning_rate": 2.4035317092045763e-06,
+      "loss": 0.6345,
+      "step": 3922
+    },
+    {
+      "epoch": 0.9041253745102559,
+      "grad_norm": 0.1377190500497818,
+      "learning_rate": 2.3921105026913527e-06,
+      "loss": 0.6291,
+      "step": 3923
+    },
+    {
+      "epoch": 0.9043558423599908,
+      "grad_norm": 0.13996082544326782,
+      "learning_rate": 2.380715831673047e-06,
+      "loss": 0.6375,
+      "step": 3924
+    },
+    {
+      "epoch": 0.9045863102097258,
+      "grad_norm": 0.1362973004579544,
+      "learning_rate": 2.369347702500774e-06,
+      "loss": 0.6454,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9048167780594607,
+      "grad_norm": 0.13769546151161194,
+      "learning_rate": 2.3580061215108585e-06,
+      "loss": 0.635,
+      "step": 3926
+    },
+    {
+      "epoch": 0.9050472459091957,
+      "grad_norm": 0.13530834019184113,
+      "learning_rate": 2.3466910950248332e-06,
+      "loss": 0.634,
+      "step": 3927
+    },
+    {
+      "epoch": 0.9052777137589306,
+      "grad_norm": 0.14218156039714813,
+      "learning_rate": 2.3354026293494034e-06,
+      "loss": 0.6334,
+      "step": 3928
+    },
+    {
+      "epoch": 0.9055081816086656,
+      "grad_norm": 0.13567234575748444,
+      "learning_rate": 2.324140730776497e-06,
+      "loss": 0.6383,
+      "step": 3929
+    },
+    {
+      "epoch": 0.9057386494584005,
+      "grad_norm": 0.14312051236629486,
+      "learning_rate": 2.3129054055832376e-06,
+      "loss": 0.6454,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9059691173081356,
+      "grad_norm": 0.13520599901676178,
+      "learning_rate": 2.3016966600319154e-06,
+      "loss": 0.6408,
+      "step": 3931
+    },
+    {
+      "epoch": 0.9061995851578705,
+      "grad_norm": 0.13904651999473572,
+      "learning_rate": 2.290514500370011e-06,
+      "loss": 0.6371,
+      "step": 3932
+    },
+    {
+      "epoch": 0.9064300530076055,
+      "grad_norm": 0.14134728908538818,
+      "learning_rate": 2.2793589328302056e-06,
+      "loss": 0.641,
+      "step": 3933
+    },
+    {
+      "epoch": 0.9066605208573404,
+      "grad_norm": 0.13760726153850555,
+      "learning_rate": 2.268229963630325e-06,
+      "loss": 0.6408,
+      "step": 3934
+    },
+    {
+      "epoch": 0.9068909887070754,
+      "grad_norm": 0.1370038092136383,
+      "learning_rate": 2.2571275989734076e-06,
+      "loss": 0.6401,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9071214565568103,
+      "grad_norm": 0.13965949416160583,
+      "learning_rate": 2.2460518450476474e-06,
+      "loss": 0.6404,
+      "step": 3936
+    },
+    {
+      "epoch": 0.9073519244065453,
+      "grad_norm": 0.1396677941083908,
+      "learning_rate": 2.2350027080263845e-06,
+      "loss": 0.6266,
+      "step": 3937
+    },
+    {
+      "epoch": 0.9075823922562802,
+      "grad_norm": 0.13712947070598602,
+      "learning_rate": 2.223980194068159e-06,
+      "loss": 0.6334,
+      "step": 3938
+    },
+    {
+      "epoch": 0.9078128601060153,
+      "grad_norm": 0.1409660130739212,
+      "learning_rate": 2.212984309316646e-06,
+      "loss": 0.6407,
+      "step": 3939
+    },
+    {
+      "epoch": 0.9080433279557502,
+      "grad_norm": 0.14090897142887115,
+      "learning_rate": 2.2020150599006916e-06,
+      "loss": 0.6377,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9082737958054852,
+      "grad_norm": 0.1404808610677719,
+      "learning_rate": 2.191072451934295e-06,
+      "loss": 0.6421,
+      "step": 3941
+    },
+    {
+      "epoch": 0.9085042636552201,
+      "grad_norm": 0.1363547444343567,
+      "learning_rate": 2.180156491516605e-06,
+      "loss": 0.6403,
+      "step": 3942
+    },
+    {
+      "epoch": 0.9087347315049551,
+      "grad_norm": 0.13701926171779633,
+      "learning_rate": 2.1692671847319048e-06,
+      "loss": 0.6341,
+      "step": 3943
+    },
+    {
+      "epoch": 0.90896519935469,
+      "grad_norm": 0.19568116962909698,
+      "learning_rate": 2.1584045376496385e-06,
+      "loss": 0.6294,
+      "step": 3944
+    },
+    {
+      "epoch": 0.909195667204425,
+      "grad_norm": 0.13929463922977448,
+      "learning_rate": 2.147568556324392e-06,
+      "loss": 0.6354,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9094261350541599,
+      "grad_norm": 0.14168402552604675,
+      "learning_rate": 2.136759246795872e-06,
+      "loss": 0.6367,
+      "step": 3946
+    },
+    {
+      "epoch": 0.9096566029038949,
+      "grad_norm": 0.13594500720500946,
+      "learning_rate": 2.125976615088926e-06,
+      "loss": 0.6406,
+      "step": 3947
+    },
+    {
+      "epoch": 0.9098870707536298,
+      "grad_norm": 0.14707335829734802,
+      "learning_rate": 2.1152206672135465e-06,
+      "loss": 0.6409,
+      "step": 3948
+    },
+    {
+      "epoch": 0.9101175386033649,
+      "grad_norm": 0.13740447163581848,
+      "learning_rate": 2.104491409164827e-06,
+      "loss": 0.6325,
+      "step": 3949
+    },
+    {
+      "epoch": 0.9103480064530998,
+      "grad_norm": 0.1381852775812149,
+      "learning_rate": 2.0937888469230115e-06,
+      "loss": 0.6356,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9105784743028348,
+      "grad_norm": 0.13955402374267578,
+      "learning_rate": 2.083112986453445e-06,
+      "loss": 0.6406,
+      "step": 3951
+    },
+    {
+      "epoch": 0.9108089421525697,
+      "grad_norm": 0.14137601852416992,
+      "learning_rate": 2.072463833706595e-06,
+      "loss": 0.6395,
+      "step": 3952
+    },
+    {
+      "epoch": 0.9110394100023047,
+      "grad_norm": 0.14015664160251617,
+      "learning_rate": 2.061841394618036e-06,
+      "loss": 0.6381,
+      "step": 3953
+    },
+    {
+      "epoch": 0.9112698778520396,
+      "grad_norm": 0.1479710191488266,
+      "learning_rate": 2.0512456751084763e-06,
+      "loss": 0.6534,
+      "step": 3954
+    },
+    {
+      "epoch": 0.9115003457017746,
+      "grad_norm": 0.14452481269836426,
+      "learning_rate": 2.040676681083703e-06,
+      "loss": 0.6372,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9117308135515095,
+      "grad_norm": 0.1414709836244583,
+      "learning_rate": 2.030134418434626e-06,
+      "loss": 0.6353,
+      "step": 3956
+    },
+    {
+      "epoch": 0.9119612814012446,
+      "grad_norm": 0.1382758766412735,
+      "learning_rate": 2.0196188930372563e-06,
+      "loss": 0.6351,
+      "step": 3957
+    },
+    {
+      "epoch": 0.9121917492509795,
+      "grad_norm": 0.14365412294864655,
+      "learning_rate": 2.0091301107526774e-06,
+      "loss": 0.6343,
+      "step": 3958
+    },
+    {
+      "epoch": 0.9124222171007145,
+      "grad_norm": 0.13868463039398193,
+      "learning_rate": 1.998668077427096e-06,
+      "loss": 0.6444,
+      "step": 3959
+    },
+    {
+      "epoch": 0.9126526849504494,
+      "grad_norm": 0.14234106242656708,
+      "learning_rate": 1.9882327988918038e-06,
+      "loss": 0.6295,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9128831528001844,
+      "grad_norm": 0.16360850632190704,
+      "learning_rate": 1.977824280963164e-06,
+      "loss": 0.6482,
+      "step": 3961
+    },
+    {
+      "epoch": 0.9131136206499193,
+      "grad_norm": 0.13877460360527039,
+      "learning_rate": 1.967442529442637e-06,
+      "loss": 0.641,
+      "step": 3962
+    },
+    {
+      "epoch": 0.9133440884996543,
+      "grad_norm": 0.13784171640872955,
+      "learning_rate": 1.957087550116765e-06,
+      "loss": 0.6391,
+      "step": 3963
+    },
+    {
+      "epoch": 0.9135745563493892,
+      "grad_norm": 0.1386898308992386,
+      "learning_rate": 1.94675934875716e-06,
+      "loss": 0.6399,
+      "step": 3964
+    },
+    {
+      "epoch": 0.9138050241991242,
+      "grad_norm": 0.14157116413116455,
+      "learning_rate": 1.936457931120522e-06,
+      "loss": 0.6383,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9140354920488591,
+      "grad_norm": 0.13940860331058502,
+      "learning_rate": 1.9261833029486088e-06,
+      "loss": 0.6441,
+      "step": 3966
+    },
+    {
+      "epoch": 0.9142659598985942,
+      "grad_norm": 0.13591444492340088,
+      "learning_rate": 1.9159354699682497e-06,
+      "loss": 0.6382,
+      "step": 3967
+    },
+    {
+      "epoch": 0.9144964277483291,
+      "grad_norm": 0.14465990662574768,
+      "learning_rate": 1.9057144378913427e-06,
+      "loss": 0.6278,
+      "step": 3968
+    },
+    {
+      "epoch": 0.9147268955980641,
+      "grad_norm": 0.13699522614479065,
+      "learning_rate": 1.895520212414842e-06,
+      "loss": 0.6409,
+      "step": 3969
+    },
+    {
+      "epoch": 0.914957363447799,
+      "grad_norm": 0.14070484042167664,
+      "learning_rate": 1.8853527992207742e-06,
+      "loss": 0.6438,
+      "step": 3970
+    },
+    {
+      "epoch": 0.915187831297534,
+      "grad_norm": 0.1463005095720291,
+      "learning_rate": 1.875212203976201e-06,
+      "loss": 0.6356,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9154182991472689,
+      "grad_norm": 0.1436576545238495,
+      "learning_rate": 1.8650984323332566e-06,
+      "loss": 0.6466,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9156487669970039,
+      "grad_norm": 0.13998855650424957,
+      "learning_rate": 1.8550114899290983e-06,
+      "loss": 0.6379,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9158792348467388,
+      "grad_norm": 0.13579371571540833,
+      "learning_rate": 1.8449513823859622e-06,
+      "loss": 0.6415,
+      "step": 3974
+    },
+    {
+      "epoch": 0.9161097026964738,
+      "grad_norm": 0.14221090078353882,
+      "learning_rate": 1.8349181153111073e-06,
+      "loss": 0.6376,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9163401705462088,
+      "grad_norm": 0.136023610830307,
+      "learning_rate": 1.8249116942968325e-06,
+      "loss": 0.6349,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9165706383959438,
+      "grad_norm": 0.1378968209028244,
+      "learning_rate": 1.8149321249204765e-06,
+      "loss": 0.6463,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9168011062456787,
+      "grad_norm": 0.13700707256793976,
+      "learning_rate": 1.8049794127444119e-06,
+      "loss": 0.6424,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9170315740954137,
+      "grad_norm": 0.1433679163455963,
+      "learning_rate": 1.7950535633160403e-06,
+      "loss": 0.6281,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9172620419451486,
+      "grad_norm": 0.14099539816379547,
+      "learning_rate": 1.7851545821677973e-06,
+      "loss": 0.6381,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9174925097948836,
+      "grad_norm": 0.13756389915943146,
+      "learning_rate": 1.7752824748171415e-06,
+      "loss": 0.6286,
+      "step": 3981
+    },
+    {
+      "epoch": 0.9177229776446185,
+      "grad_norm": 0.13646264374256134,
+      "learning_rate": 1.7654372467665325e-06,
+      "loss": 0.626,
+      "step": 3982
+    },
+    {
+      "epoch": 0.9179534454943535,
+      "grad_norm": 0.1390201449394226,
+      "learning_rate": 1.7556189035034644e-06,
+      "loss": 0.6433,
+      "step": 3983
+    },
+    {
+      "epoch": 0.9181839133440886,
+      "grad_norm": 0.1405726969242096,
+      "learning_rate": 1.7458274505004702e-06,
+      "loss": 0.6378,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9184143811938235,
+      "grad_norm": 0.13295190036296844,
+      "learning_rate": 1.7360628932150512e-06,
+      "loss": 0.635,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9186448490435585,
+      "grad_norm": 0.13831880688667297,
+      "learning_rate": 1.7263252370897377e-06,
+      "loss": 0.6451,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9188753168932934,
+      "grad_norm": 0.1404305100440979,
+      "learning_rate": 1.7166144875520763e-06,
+      "loss": 0.6413,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9191057847430284,
+      "grad_norm": 0.13879506289958954,
+      "learning_rate": 1.7069306500145875e-06,
+      "loss": 0.6319,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9193362525927633,
+      "grad_norm": 0.13712599873542786,
+      "learning_rate": 1.6972737298748266e-06,
+      "loss": 0.6433,
+      "step": 3989
+    },
+    {
+      "epoch": 0.9195667204424983,
+      "grad_norm": 0.13845130801200867,
+      "learning_rate": 1.6876437325153261e-06,
+      "loss": 0.6356,
+      "step": 3990
+    },
+    {
+      "epoch": 0.9197971882922332,
+      "grad_norm": 0.13963979482650757,
+      "learning_rate": 1.6780406633036095e-06,
+      "loss": 0.6461,
+      "step": 3991
+    },
+    {
+      "epoch": 0.9200276561419682,
+      "grad_norm": 0.1397661715745926,
+      "learning_rate": 1.6684645275922007e-06,
+      "loss": 0.6437,
+      "step": 3992
+    },
+    {
+      "epoch": 0.9202581239917031,
+      "grad_norm": 0.13445524871349335,
+      "learning_rate": 1.6589153307186078e-06,
+      "loss": 0.6337,
+      "step": 3993
+    },
+    {
+      "epoch": 0.9204885918414382,
+      "grad_norm": 0.14014729857444763,
+      "learning_rate": 1.6493930780053235e-06,
+      "loss": 0.6396,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9207190596911731,
+      "grad_norm": 0.14015179872512817,
+      "learning_rate": 1.6398977747598243e-06,
+      "loss": 0.6377,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9209495275409081,
+      "grad_norm": 0.13548614084720612,
+      "learning_rate": 1.6304294262745656e-06,
+      "loss": 0.635,
+      "step": 3996
+    },
+    {
+      "epoch": 0.921179995390643,
+      "grad_norm": 0.1403760462999344,
+      "learning_rate": 1.6209880378269705e-06,
+      "loss": 0.6384,
+      "step": 3997
+    },
+    {
+      "epoch": 0.921410463240378,
+      "grad_norm": 0.1437310427427292,
+      "learning_rate": 1.6115736146794402e-06,
+      "loss": 0.634,
+      "step": 3998
+    },
+    {
+      "epoch": 0.9216409310901129,
+      "grad_norm": 0.1346195787191391,
+      "learning_rate": 1.6021861620793666e-06,
+      "loss": 0.6337,
+      "step": 3999
+    },
+    {
+      "epoch": 0.9218713989398479,
+      "grad_norm": 0.13712115585803986,
+      "learning_rate": 1.5928256852590751e-06,
+      "loss": 0.6279,
+      "step": 4000
+    },
+    {
+      "epoch": 0.9221018667895828,
+      "grad_norm": 0.13652116060256958,
+      "learning_rate": 1.5834921894358701e-06,
+      "loss": 0.644,
+      "step": 4001
+    },
+    {
+      "epoch": 0.9223323346393179,
+      "grad_norm": 0.13506993651390076,
+      "learning_rate": 1.574185679812029e-06,
+      "loss": 0.6371,
+      "step": 4002
+    },
+    {
+      "epoch": 0.9225628024890528,
+      "grad_norm": 0.13801835477352142,
+      "learning_rate": 1.564906161574764e-06,
+      "loss": 0.6401,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9227932703387878,
+      "grad_norm": 0.13854144513607025,
+      "learning_rate": 1.555653639896265e-06,
+      "loss": 0.6415,
+      "step": 4004
+    },
+    {
+      "epoch": 0.9230237381885227,
+      "grad_norm": 0.13430531322956085,
+      "learning_rate": 1.5464281199336683e-06,
+      "loss": 0.6296,
+      "step": 4005
+    },
+    {
+      "epoch": 0.9232542060382577,
+      "grad_norm": 0.13435374200344086,
+      "learning_rate": 1.5372296068290493e-06,
+      "loss": 0.6304,
+      "step": 4006
+    },
+    {
+      "epoch": 0.9234846738879926,
+      "grad_norm": 0.13501442968845367,
+      "learning_rate": 1.5280581057094346e-06,
+      "loss": 0.6355,
+      "step": 4007
+    },
+    {
+      "epoch": 0.9237151417377276,
+      "grad_norm": 0.13476891815662384,
+      "learning_rate": 1.518913621686807e-06,
+      "loss": 0.6398,
+      "step": 4008
+    },
+    {
+      "epoch": 0.9239456095874625,
+      "grad_norm": 0.1361880600452423,
+      "learning_rate": 1.5097961598580845e-06,
+      "loss": 0.6492,
+      "step": 4009
+    },
+    {
+      "epoch": 0.9241760774371975,
+      "grad_norm": 0.13403740525245667,
+      "learning_rate": 1.5007057253051127e-06,
+      "loss": 0.6257,
+      "step": 4010
+    },
+    {
+      "epoch": 0.9244065452869324,
+      "grad_norm": 0.13556121289730072,
+      "learning_rate": 1.4916423230946885e-06,
+      "loss": 0.6413,
+      "step": 4011
+    },
+    {
+      "epoch": 0.9246370131366675,
+      "grad_norm": 0.13702526688575745,
+      "learning_rate": 1.4826059582785324e-06,
+      "loss": 0.6427,
+      "step": 4012
+    },
+    {
+      "epoch": 0.9248674809864024,
+      "grad_norm": 0.13721799850463867,
+      "learning_rate": 1.473596635893293e-06,
+      "loss": 0.6494,
+      "step": 4013
+    },
+    {
+      "epoch": 0.9250979488361374,
+      "grad_norm": 0.1371307075023651,
+      "learning_rate": 1.464614360960559e-06,
+      "loss": 0.6482,
+      "step": 4014
+    },
+    {
+      "epoch": 0.9253284166858723,
+      "grad_norm": 0.1363903433084488,
+      "learning_rate": 1.4556591384868367e-06,
+      "loss": 0.6343,
+      "step": 4015
+    },
+    {
+      "epoch": 0.9255588845356073,
+      "grad_norm": 0.14017435908317566,
+      "learning_rate": 1.4467309734635393e-06,
+      "loss": 0.6399,
+      "step": 4016
+    },
+    {
+      "epoch": 0.9257893523853422,
+      "grad_norm": 0.13649635016918182,
+      "learning_rate": 1.437829870867019e-06,
+      "loss": 0.6372,
+      "step": 4017
+    },
+    {
+      "epoch": 0.9260198202350772,
+      "grad_norm": 0.13363580405712128,
+      "learning_rate": 1.4289558356585353e-06,
+      "loss": 0.6357,
+      "step": 4018
+    },
+    {
+      "epoch": 0.9262502880848121,
+      "grad_norm": 0.1385689377784729,
+      "learning_rate": 1.4201088727842648e-06,
+      "loss": 0.6344,
+      "step": 4019
+    },
+    {
+      "epoch": 0.9264807559345472,
+      "grad_norm": 0.13566775619983673,
+      "learning_rate": 1.411288987175291e-06,
+      "loss": 0.646,
+      "step": 4020
+    },
+    {
+      "epoch": 0.9267112237842821,
+      "grad_norm": 0.13955044746398926,
+      "learning_rate": 1.4024961837476092e-06,
+      "loss": 0.6411,
+      "step": 4021
+    },
+    {
+      "epoch": 0.9269416916340171,
+      "grad_norm": 0.13665254414081573,
+      "learning_rate": 1.39373046740211e-06,
+      "loss": 0.6417,
+      "step": 4022
+    },
+    {
+      "epoch": 0.927172159483752,
+      "grad_norm": 0.13750925660133362,
+      "learning_rate": 1.384991843024591e-06,
+      "loss": 0.6351,
+      "step": 4023
+    },
+    {
+      "epoch": 0.927402627333487,
+      "grad_norm": 0.1418118178844452,
+      "learning_rate": 1.3762803154857729e-06,
+      "loss": 0.6324,
+      "step": 4024
+    },
+    {
+      "epoch": 0.9276330951832219,
+      "grad_norm": 0.13481222093105316,
+      "learning_rate": 1.3675958896412267e-06,
+      "loss": 0.6325,
+      "step": 4025
+    },
+    {
+      "epoch": 0.9278635630329569,
+      "grad_norm": 0.1390499770641327,
+      "learning_rate": 1.3589385703314529e-06,
+      "loss": 0.6369,
+      "step": 4026
+    },
+    {
+      "epoch": 0.9280940308826918,
+      "grad_norm": 0.13821762800216675,
+      "learning_rate": 1.3503083623818412e-06,
+      "loss": 0.6386,
+      "step": 4027
+    },
+    {
+      "epoch": 0.9283244987324268,
+      "grad_norm": 0.141762837767601,
+      "learning_rate": 1.34170527060265e-06,
+      "loss": 0.6422,
+      "step": 4028
+    },
+    {
+      "epoch": 0.9285549665821617,
+      "grad_norm": 0.1358966827392578,
+      "learning_rate": 1.3331292997890377e-06,
+      "loss": 0.6365,
+      "step": 4029
+    },
+    {
+      "epoch": 0.9287854344318968,
+      "grad_norm": 0.13388477265834808,
+      "learning_rate": 1.3245804547210582e-06,
+      "loss": 0.6495,
+      "step": 4030
+    },
+    {
+      "epoch": 0.9290159022816317,
+      "grad_norm": 0.1355268657207489,
+      "learning_rate": 1.3160587401636171e-06,
+      "loss": 0.6362,
+      "step": 4031
+    },
+    {
+      "epoch": 0.9292463701313667,
+      "grad_norm": 0.1358853578567505,
+      "learning_rate": 1.3075641608665202e-06,
+      "loss": 0.6367,
+      "step": 4032
+    },
+    {
+      "epoch": 0.9294768379811016,
+      "grad_norm": 0.1385107785463333,
+      "learning_rate": 1.2990967215644412e-06,
+      "loss": 0.6408,
+      "step": 4033
+    },
+    {
+      "epoch": 0.9297073058308366,
+      "grad_norm": 0.1369771957397461,
+      "learning_rate": 1.2906564269769217e-06,
+      "loss": 0.6382,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9299377736805715,
+      "grad_norm": 0.14050287008285522,
+      "learning_rate": 1.282243281808393e-06,
+      "loss": 0.6412,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9301682415303065,
+      "grad_norm": 0.1366300880908966,
+      "learning_rate": 1.2738572907481315e-06,
+      "loss": 0.6419,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9303987093800414,
+      "grad_norm": 0.1353761851787567,
+      "learning_rate": 1.2654984584702766e-06,
+      "loss": 0.6393,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9306291772297764,
+      "grad_norm": 0.18008558452129364,
+      "learning_rate": 1.2571667896338624e-06,
+      "loss": 0.6354,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9308596450795114,
+      "grad_norm": 0.13674607872962952,
+      "learning_rate": 1.2488622888827517e-06,
+      "loss": 0.6426,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9310901129292464,
+      "grad_norm": 0.13963568210601807,
+      "learning_rate": 1.24058496084567e-06,
+      "loss": 0.6333,
+      "step": 4040
+    },
+    {
+      "epoch": 0.9313205807789813,
+      "grad_norm": 0.1362845003604889,
+      "learning_rate": 1.2323348101362043e-06,
+      "loss": 0.6435,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9315510486287163,
+      "grad_norm": 0.1407993733882904,
+      "learning_rate": 1.224111841352793e-06,
+      "loss": 0.6403,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9317815164784513,
+      "grad_norm": 0.13926169276237488,
+      "learning_rate": 1.2159160590787143e-06,
+      "loss": 0.6412,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9320119843281862,
+      "grad_norm": 0.13823188841342926,
+      "learning_rate": 1.2077474678821088e-06,
+      "loss": 0.6452,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9322424521779212,
+      "grad_norm": 0.1374104917049408,
+      "learning_rate": 1.1996060723159508e-06,
+      "loss": 0.6343,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9324729200276561,
+      "grad_norm": 0.13301199674606323,
+      "learning_rate": 1.1914918769180606e-06,
+      "loss": 0.6373,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9327033878773912,
+      "grad_norm": 0.136099711060524,
+      "learning_rate": 1.1834048862110814e-06,
+      "loss": 0.6435,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9329338557271261,
+      "grad_norm": 0.140198215842247,
+      "learning_rate": 1.17534510470253e-06,
+      "loss": 0.6392,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9331643235768611,
+      "grad_norm": 0.1340055614709854,
+      "learning_rate": 1.1673125368847238e-06,
+      "loss": 0.6425,
+      "step": 4049
+    },
+    {
+      "epoch": 0.933394791426596,
+      "grad_norm": 0.14197507500648499,
+      "learning_rate": 1.1593071872348204e-06,
+      "loss": 0.6369,
+      "step": 4050
+    },
+    {
+      "epoch": 0.933625259276331,
+      "grad_norm": 0.13812677562236786,
+      "learning_rate": 1.1513290602148174e-06,
+      "loss": 0.6403,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9338557271260659,
+      "grad_norm": 0.1445024460554123,
+      "learning_rate": 1.1433781602715189e-06,
+      "loss": 0.6346,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9340861949758009,
+      "grad_norm": 0.13985450565814972,
+      "learning_rate": 1.1354544918365795e-06,
+      "loss": 0.643,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9343166628255358,
+      "grad_norm": 0.13591651618480682,
+      "learning_rate": 1.1275580593264611e-06,
+      "loss": 0.6331,
+      "step": 4054
+    },
+    {
+      "epoch": 0.9345471306752708,
+      "grad_norm": 0.13689643144607544,
+      "learning_rate": 1.1196888671424377e-06,
+      "loss": 0.6402,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9347775985250057,
+      "grad_norm": 0.13568592071533203,
+      "learning_rate": 1.111846919670606e-06,
+      "loss": 0.6353,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9350080663747408,
+      "grad_norm": 0.13714367151260376,
+      "learning_rate": 1.1040322212818922e-06,
+      "loss": 0.636,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9352385342244757,
+      "grad_norm": 0.13835318386554718,
+      "learning_rate": 1.096244776332006e-06,
+      "loss": 0.6393,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9354690020742107,
+      "grad_norm": 0.14133679866790771,
+      "learning_rate": 1.0884845891614925e-06,
+      "loss": 0.6446,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9356994699239456,
+      "grad_norm": 0.13992494344711304,
+      "learning_rate": 1.0807516640956972e-06,
+      "loss": 0.6326,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9359299377736806,
+      "grad_norm": 0.13520614802837372,
+      "learning_rate": 1.0730460054447612e-06,
+      "loss": 0.6325,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9361604056234155,
+      "grad_norm": 0.13457529246807098,
+      "learning_rate": 1.065367617503621e-06,
+      "loss": 0.634,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9363908734731505,
+      "grad_norm": 0.1412680298089981,
+      "learning_rate": 1.057716504552053e-06,
+      "loss": 0.6391,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9366213413228854,
+      "grad_norm": 0.13816601037979126,
+      "learning_rate": 1.0500926708545855e-06,
+      "loss": 0.6334,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9368518091726205,
+      "grad_norm": 0.138847216963768,
+      "learning_rate": 1.0424961206605632e-06,
+      "loss": 0.6358,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9370822770223554,
+      "grad_norm": 0.13748426735401154,
+      "learning_rate": 1.0349268582041161e-06,
+      "loss": 0.6406,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9373127448720904,
+      "grad_norm": 0.13643766939640045,
+      "learning_rate": 1.0273848877041802e-06,
+      "loss": 0.6331,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9375432127218253,
+      "grad_norm": 0.1338902860879898,
+      "learning_rate": 1.0198702133644656e-06,
+      "loss": 0.639,
+      "step": 4068
+    },
+    {
+      "epoch": 0.9377736805715603,
+      "grad_norm": 0.13689804077148438,
+      "learning_rate": 1.0123828393734714e-06,
+      "loss": 0.6414,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9380041484212952,
+      "grad_norm": 0.14404194056987762,
+      "learning_rate": 1.0049227699044762e-06,
+      "loss": 0.6427,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9382346162710302,
+      "grad_norm": 0.1352960616350174,
+      "learning_rate": 9.974900091155425e-07,
+      "loss": 0.6478,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9384650841207651,
+      "grad_norm": 0.1446170061826706,
+      "learning_rate": 9.90084561149518e-07,
+      "loss": 0.6336,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9386955519705001,
+      "grad_norm": 0.14939038455486298,
+      "learning_rate": 9.827064301340228e-07,
+      "loss": 0.6342,
+      "step": 4073
+    },
+    {
+      "epoch": 0.938926019820235,
+      "grad_norm": 0.1382802277803421,
+      "learning_rate": 9.75355620181445e-07,
+      "loss": 0.6345,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9391564876699701,
+      "grad_norm": 0.1343778669834137,
+      "learning_rate": 9.680321353889576e-07,
+      "loss": 0.6438,
+      "step": 4075
+    },
+    {
+      "epoch": 0.939386955519705,
+      "grad_norm": 0.13062545657157898,
+      "learning_rate": 9.607359798384785e-07,
+      "loss": 0.6332,
+      "step": 4076
+    },
+    {
+      "epoch": 0.93961742336944,
+      "grad_norm": 0.1372269243001938,
+      "learning_rate": 9.534671575967213e-07,
+      "loss": 0.6382,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9398478912191749,
+      "grad_norm": 0.13971804082393646,
+      "learning_rate": 9.46225672715162e-07,
+      "loss": 0.6306,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9400783590689099,
+      "grad_norm": 0.13551059365272522,
+      "learning_rate": 9.390115292300162e-07,
+      "loss": 0.6438,
+      "step": 4079
+    },
+    {
+      "epoch": 0.9403088269186448,
+      "grad_norm": 0.13725043833255768,
+      "learning_rate": 9.318247311622785e-07,
+      "loss": 0.6437,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9405392947683798,
+      "grad_norm": 0.13926532864570618,
+      "learning_rate": 9.246652825176949e-07,
+      "loss": 0.6427,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9407697626181147,
+      "grad_norm": 0.14660166203975677,
+      "learning_rate": 9.175331872867732e-07,
+      "loss": 0.6462,
+      "step": 4082
+    },
+    {
+      "epoch": 0.9410002304678498,
+      "grad_norm": 0.13765586912631989,
+      "learning_rate": 9.104284494447779e-07,
+      "loss": 0.6419,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9412306983175847,
+      "grad_norm": 0.13550062477588654,
+      "learning_rate": 9.033510729517136e-07,
+      "loss": 0.6526,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9414611661673197,
+      "grad_norm": 0.20040088891983032,
+      "learning_rate": 8.96301061752336e-07,
+      "loss": 0.6438,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9416916340170546,
+      "grad_norm": 0.13654790818691254,
+      "learning_rate": 8.892784197761572e-07,
+      "loss": 0.629,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9419221018667896,
+      "grad_norm": 0.13696341216564178,
+      "learning_rate": 8.822831509374297e-07,
+      "loss": 0.6424,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9421525697165245,
+      "grad_norm": 0.13563202321529388,
+      "learning_rate": 8.753152591351455e-07,
+      "loss": 0.642,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9423830375662595,
+      "grad_norm": 0.13841907680034637,
+      "learning_rate": 8.683747482530424e-07,
+      "loss": 0.6395,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9426135054159944,
+      "grad_norm": 0.13629648089408875,
+      "learning_rate": 8.614616221595983e-07,
+      "loss": 0.6394,
+      "step": 4090
+    },
+    {
+      "epoch": 0.9428439732657294,
+      "grad_norm": 0.13527196645736694,
+      "learning_rate": 8.545758847080143e-07,
+      "loss": 0.6384,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9430744411154643,
+      "grad_norm": 0.13530802726745605,
+      "learning_rate": 8.47717539736237e-07,
+      "loss": 0.6322,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9433049089651994,
+      "grad_norm": 0.1388748586177826,
+      "learning_rate": 8.408865910669583e-07,
+      "loss": 0.6369,
+      "step": 4093
+    },
+    {
+      "epoch": 0.9435353768149343,
+      "grad_norm": 0.13674351572990417,
+      "learning_rate": 8.340830425075663e-07,
+      "loss": 0.6384,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9437658446646693,
+      "grad_norm": 0.13656572997570038,
+      "learning_rate": 8.273068978501996e-07,
+      "loss": 0.6455,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9439963125144042,
+      "grad_norm": 0.13279734551906586,
+      "learning_rate": 8.205581608717261e-07,
+      "loss": 0.633,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9442267803641392,
+      "grad_norm": 0.14017152786254883,
+      "learning_rate": 8.138368353337255e-07,
+      "loss": 0.6338,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9444572482138741,
+      "grad_norm": 0.13749238848686218,
+      "learning_rate": 8.071429249825013e-07,
+      "loss": 0.6366,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9446877160636091,
+      "grad_norm": 0.1360204666852951,
+      "learning_rate": 8.004764335490856e-07,
+      "loss": 0.6405,
+      "step": 4099
+    },
+    {
+      "epoch": 0.944918183913344,
+      "grad_norm": 0.14251859486103058,
+      "learning_rate": 7.938373647492115e-07,
+      "loss": 0.6426,
+      "step": 4100
+    },
+    {
+      "epoch": 0.945148651763079,
+      "grad_norm": 0.13560590147972107,
+      "learning_rate": 7.872257222833357e-07,
+      "loss": 0.6376,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9453791196128141,
+      "grad_norm": 0.1355055868625641,
+      "learning_rate": 7.806415098366438e-07,
+      "loss": 0.6379,
+      "step": 4102
+    },
+    {
+      "epoch": 0.945609587462549,
+      "grad_norm": 0.13701874017715454,
+      "learning_rate": 7.74084731079e-07,
+      "loss": 0.6298,
+      "step": 4103
+    },
+    {
+      "epoch": 0.945840055312284,
+      "grad_norm": 0.13840676844120026,
+      "learning_rate": 7.67555389665009e-07,
+      "loss": 0.629,
+      "step": 4104
+    },
+    {
+      "epoch": 0.9460705231620189,
+      "grad_norm": 0.13396546244621277,
+      "learning_rate": 7.61053489233965e-07,
+      "loss": 0.6323,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9463009910117539,
+      "grad_norm": 0.1397581547498703,
+      "learning_rate": 7.545790334098579e-07,
+      "loss": 0.6427,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9465314588614888,
+      "grad_norm": 0.13133475184440613,
+      "learning_rate": 7.481320258014124e-07,
+      "loss": 0.6424,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9467619267112238,
+      "grad_norm": 0.13664543628692627,
+      "learning_rate": 7.417124700020373e-07,
+      "loss": 0.6419,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9469923945609587,
+      "grad_norm": 0.13687404990196228,
+      "learning_rate": 7.353203695898203e-07,
+      "loss": 0.631,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9472228624106938,
+      "grad_norm": 0.13796484470367432,
+      "learning_rate": 7.289557281275782e-07,
+      "loss": 0.6416,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9474533302604287,
+      "grad_norm": 0.13329102098941803,
+      "learning_rate": 7.226185491628069e-07,
+      "loss": 0.6261,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9476837981101637,
+      "grad_norm": 0.13241469860076904,
+      "learning_rate": 7.163088362276971e-07,
+      "loss": 0.6364,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9479142659598986,
+      "grad_norm": 0.13166265189647675,
+      "learning_rate": 7.100265928391303e-07,
+      "loss": 0.6389,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9481447338096336,
+      "grad_norm": 0.13284790515899658,
+      "learning_rate": 7.037718224986833e-07,
+      "loss": 0.6334,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9483752016593685,
+      "grad_norm": 0.13939420878887177,
+      "learning_rate": 6.975445286926063e-07,
+      "loss": 0.6227,
+      "step": 4115
+    },
+    {
+      "epoch": 0.9486056695091035,
+      "grad_norm": 0.13803811371326447,
+      "learning_rate": 6.913447148918506e-07,
+      "loss": 0.6405,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9488361373588384,
+      "grad_norm": 0.13367731869220734,
+      "learning_rate": 6.851723845520408e-07,
+      "loss": 0.6311,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9490666052085734,
+      "grad_norm": 0.13658930361270905,
+      "learning_rate": 6.790275411134861e-07,
+      "loss": 0.6459,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9492970730583083,
+      "grad_norm": 0.13773563504219055,
+      "learning_rate": 6.729101880011746e-07,
+      "loss": 0.6382,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9495275409080434,
+      "grad_norm": 0.13418744504451752,
+      "learning_rate": 6.668203286247732e-07,
+      "loss": 0.6321,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9497580087577783,
+      "grad_norm": 0.13385219871997833,
+      "learning_rate": 6.607579663786223e-07,
+      "loss": 0.6343,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9499884766075133,
+      "grad_norm": 0.13633115589618683,
+      "learning_rate": 6.547231046417357e-07,
+      "loss": 0.6265,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9502189444572482,
+      "grad_norm": 0.13809369504451752,
+      "learning_rate": 6.48715746777806e-07,
+      "loss": 0.6444,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9504494123069832,
+      "grad_norm": 0.13296149671077728,
+      "learning_rate": 6.42735896135177e-07,
+      "loss": 0.6357,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9506798801567181,
+      "grad_norm": 0.13449513912200928,
+      "learning_rate": 6.367835560468938e-07,
+      "loss": 0.6401,
+      "step": 4125
+    },
+    {
+      "epoch": 0.9509103480064531,
+      "grad_norm": 0.13610585033893585,
+      "learning_rate": 6.308587298306301e-07,
+      "loss": 0.6376,
+      "step": 4126
+    },
+    {
+      "epoch": 0.951140815856188,
+      "grad_norm": 0.13235846161842346,
+      "learning_rate": 6.2496142078875e-07,
+      "loss": 0.6397,
+      "step": 4127
+    },
+    {
+      "epoch": 0.951371283705923,
+      "grad_norm": 0.13302072882652283,
+      "learning_rate": 6.19091632208274e-07,
+      "loss": 0.634,
+      "step": 4128
+    },
+    {
+      "epoch": 0.951601751555658,
+      "grad_norm": 0.13481658697128296,
+      "learning_rate": 6.132493673608797e-07,
+      "loss": 0.6375,
+      "step": 4129
+    },
+    {
+      "epoch": 0.951832219405393,
+      "grad_norm": 0.139765664935112,
+      "learning_rate": 6.074346295028955e-07,
+      "loss": 0.6346,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9520626872551279,
+      "grad_norm": 0.13490067422389984,
+      "learning_rate": 6.016474218753288e-07,
+      "loss": 0.6376,
+      "step": 4131
+    },
+    {
+      "epoch": 0.9522931551048629,
+      "grad_norm": 0.13335686922073364,
+      "learning_rate": 5.958877477038327e-07,
+      "loss": 0.6462,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9525236229545978,
+      "grad_norm": 0.1309853196144104,
+      "learning_rate": 5.901556101987005e-07,
+      "loss": 0.632,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9527540908043328,
+      "grad_norm": 0.13043320178985596,
+      "learning_rate": 5.84451012554893e-07,
+      "loss": 0.6474,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9529845586540677,
+      "grad_norm": 0.13605879247188568,
+      "learning_rate": 5.787739579520113e-07,
+      "loss": 0.6376,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9532150265038027,
+      "grad_norm": 0.13562782108783722,
+      "learning_rate": 5.731244495543186e-07,
+      "loss": 0.6292,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9534454943535376,
+      "grad_norm": 0.1376948058605194,
+      "learning_rate": 5.675024905107129e-07,
+      "loss": 0.641,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9536759622032727,
+      "grad_norm": 0.13430991768836975,
+      "learning_rate": 5.619080839547375e-07,
+      "loss": 0.6371,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9539064300530076,
+      "grad_norm": 0.1342536360025406,
+      "learning_rate": 5.563412330045758e-07,
+      "loss": 0.6447,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9541368979027426,
+      "grad_norm": 0.13564327359199524,
+      "learning_rate": 5.508019407630572e-07,
+      "loss": 0.6422,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9543673657524775,
+      "grad_norm": 0.1371186226606369,
+      "learning_rate": 5.452902103176616e-07,
+      "loss": 0.6401,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9545978336022125,
+      "grad_norm": 0.13882949948310852,
+      "learning_rate": 5.398060447404818e-07,
+      "loss": 0.6294,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9548283014519474,
+      "grad_norm": 0.14043927192687988,
+      "learning_rate": 5.343494470882671e-07,
+      "loss": 0.6451,
+      "step": 4143
+    },
+    {
+      "epoch": 0.9550587693016824,
+      "grad_norm": 0.13908624649047852,
+      "learning_rate": 5.289204204023957e-07,
+      "loss": 0.6357,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9552892371514173,
+      "grad_norm": 0.13900399208068848,
+      "learning_rate": 5.235189677088692e-07,
+      "loss": 0.6432,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9555197050011524,
+      "grad_norm": 0.13600599765777588,
+      "learning_rate": 5.18145092018335e-07,
+      "loss": 0.6311,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9557501728508873,
+      "grad_norm": 0.1335585117340088,
+      "learning_rate": 5.127987963260583e-07,
+      "loss": 0.6355,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9559806407006223,
+      "grad_norm": 0.1321658492088318,
+      "learning_rate": 5.074800836119442e-07,
+      "loss": 0.6348,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9562111085503572,
+      "grad_norm": 0.13960915803909302,
+      "learning_rate": 5.021889568404991e-07,
+      "loss": 0.6346,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9564415764000922,
+      "grad_norm": 0.13837899267673492,
+      "learning_rate": 4.969254189608863e-07,
+      "loss": 0.6393,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9566720442498271,
+      "grad_norm": 0.13521605730056763,
+      "learning_rate": 4.916894729068644e-07,
+      "loss": 0.6338,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9569025120995621,
+      "grad_norm": 0.13600218296051025,
+      "learning_rate": 4.864811215968324e-07,
+      "loss": 0.6392,
+      "step": 4152
+    },
+    {
+      "epoch": 0.957132979949297,
+      "grad_norm": 0.1411590576171875,
+      "learning_rate": 4.813003679337957e-07,
+      "loss": 0.6352,
+      "step": 4153
+    },
+    {
+      "epoch": 0.957363447799032,
+      "grad_norm": 0.13098092377185822,
+      "learning_rate": 4.761472148053836e-07,
+      "loss": 0.637,
+      "step": 4154
+    },
+    {
+      "epoch": 0.957593915648767,
+      "grad_norm": 0.13644741475582123,
+      "learning_rate": 4.710216650838317e-07,
+      "loss": 0.6348,
+      "step": 4155
+    },
+    {
+      "epoch": 0.957824383498502,
+      "grad_norm": 0.13410955667495728,
+      "learning_rate": 4.6592372162601037e-07,
+      "loss": 0.6394,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9580548513482369,
+      "grad_norm": 0.13613279163837433,
+      "learning_rate": 4.608533872733911e-07,
+      "loss": 0.642,
+      "step": 4157
+    },
+    {
+      "epoch": 0.9582853191979719,
+      "grad_norm": 0.13792695105075836,
+      "learning_rate": 4.558106648520466e-07,
+      "loss": 0.6296,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9585157870477068,
+      "grad_norm": 0.14079563319683075,
+      "learning_rate": 4.5079555717267854e-07,
+      "loss": 0.649,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9587462548974418,
+      "grad_norm": 0.13565976917743683,
+      "learning_rate": 4.4580806703057865e-07,
+      "loss": 0.6424,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9589767227471768,
+      "grad_norm": 0.13566036522388458,
+      "learning_rate": 4.408481972056622e-07,
+      "loss": 0.6388,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9592071905969117,
+      "grad_norm": 0.1372612714767456,
+      "learning_rate": 4.3591595046243994e-07,
+      "loss": 0.6375,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9594376584466467,
+      "grad_norm": 0.13461288809776306,
+      "learning_rate": 4.3101132955002396e-07,
+      "loss": 0.635,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9596681262963817,
+      "grad_norm": 0.1348971575498581,
+      "learning_rate": 4.2613433720213316e-07,
+      "loss": 0.6376,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9598985941461167,
+      "grad_norm": 0.13625575602054596,
+      "learning_rate": 4.212849761370874e-07,
+      "loss": 0.6393,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9601290619958516,
+      "grad_norm": 0.13304929435253143,
+      "learning_rate": 4.16463249057808e-07,
+      "loss": 0.6365,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9603595298455866,
+      "grad_norm": 0.13466964662075043,
+      "learning_rate": 4.116691586518062e-07,
+      "loss": 0.6374,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9605899976953215,
+      "grad_norm": 0.1362065076828003,
+      "learning_rate": 4.0690270759119464e-07,
+      "loss": 0.6364,
+      "step": 4168
+    },
+    {
+      "epoch": 0.9608204655450565,
+      "grad_norm": 0.13293537497520447,
+      "learning_rate": 4.021638985326759e-07,
+      "loss": 0.6401,
+      "step": 4169
+    },
+    {
+      "epoch": 0.9610509333947914,
+      "grad_norm": 0.13475392758846283,
+      "learning_rate": 3.974527341175427e-07,
+      "loss": 0.6275,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9612814012445264,
+      "grad_norm": 0.13077649474143982,
+      "learning_rate": 3.9276921697169455e-07,
+      "loss": 0.6325,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9615118690942613,
+      "grad_norm": 0.13671356439590454,
+      "learning_rate": 3.8811334970561553e-07,
+      "loss": 0.6435,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9617423369439964,
+      "grad_norm": 0.1355324536561966,
+      "learning_rate": 3.834851349143631e-07,
+      "loss": 0.6361,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9619728047937313,
+      "grad_norm": 0.13314592838287354,
+      "learning_rate": 3.78884575177596e-07,
+      "loss": 0.6362,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9622032726434663,
+      "grad_norm": 0.14188657701015472,
+      "learning_rate": 3.743116730595575e-07,
+      "loss": 0.6382,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9624337404932012,
+      "grad_norm": 0.133900985121727,
+      "learning_rate": 3.697664311090754e-07,
+      "loss": 0.6374,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9626642083429362,
+      "grad_norm": 0.1333308219909668,
+      "learning_rate": 3.6524885185955647e-07,
+      "loss": 0.6408,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9628946761926711,
+      "grad_norm": 0.13134267926216125,
+      "learning_rate": 3.6075893782899217e-07,
+      "loss": 0.632,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9631251440424061,
+      "grad_norm": 0.13069699704647064,
+      "learning_rate": 3.5629669151994725e-07,
+      "loss": 0.6364,
+      "step": 4179
+    },
+    {
+      "epoch": 0.963355611892141,
+      "grad_norm": 0.1325685828924179,
+      "learning_rate": 3.518621154195767e-07,
+      "loss": 0.6371,
+      "step": 4180
+    },
+    {
+      "epoch": 0.963586079741876,
+      "grad_norm": 0.13809578120708466,
+      "learning_rate": 3.4745521199960884e-07,
+      "loss": 0.6378,
+      "step": 4181
+    },
+    {
+      "epoch": 0.963816547591611,
+      "grad_norm": 0.13212169706821442,
+      "learning_rate": 3.4307598371633445e-07,
+      "loss": 0.6327,
+      "step": 4182
+    },
+    {
+      "epoch": 0.964047015441346,
+      "grad_norm": 0.1352756917476654,
+      "learning_rate": 3.387244330106454e-07,
+      "loss": 0.6443,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9642774832910809,
+      "grad_norm": 0.1354503035545349,
+      "learning_rate": 3.3440056230797933e-07,
+      "loss": 0.6386,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9645079511408159,
+      "grad_norm": 0.13389207422733307,
+      "learning_rate": 3.30104374018364e-07,
+      "loss": 0.6328,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9647384189905508,
+      "grad_norm": 0.13191154599189758,
+      "learning_rate": 3.2583587053638955e-07,
+      "loss": 0.6486,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9649688868402858,
+      "grad_norm": 0.1348099559545517,
+      "learning_rate": 3.2159505424122495e-07,
+      "loss": 0.6407,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9651993546900207,
+      "grad_norm": 0.135834202170372,
+      "learning_rate": 3.1738192749658503e-07,
+      "loss": 0.6347,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9654298225397557,
+      "grad_norm": 0.13616180419921875,
+      "learning_rate": 3.131964926507747e-07,
+      "loss": 0.6374,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9656602903894906,
+      "grad_norm": 0.13497985899448395,
+      "learning_rate": 3.0903875203665556e-07,
+      "loss": 0.6316,
+      "step": 4190
+    },
+    {
+      "epoch": 0.9658907582392257,
+      "grad_norm": 0.13490994274616241,
+      "learning_rate": 3.049087079716462e-07,
+      "loss": 0.639,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9661212260889606,
+      "grad_norm": 0.13283471763134003,
+      "learning_rate": 3.0080636275774397e-07,
+      "loss": 0.6342,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9663516939386956,
+      "grad_norm": 0.1357053518295288,
+      "learning_rate": 2.967317186814922e-07,
+      "loss": 0.6337,
+      "step": 4193
+    },
+    {
+      "epoch": 0.9665821617884305,
+      "grad_norm": 0.13123416900634766,
+      "learning_rate": 2.926847780139907e-07,
+      "loss": 0.6394,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9668126296381655,
+      "grad_norm": 0.13529904186725616,
+      "learning_rate": 2.8866554301091866e-07,
+      "loss": 0.6258,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9670430974879004,
+      "grad_norm": 0.13570614159107208,
+      "learning_rate": 2.846740159125061e-07,
+      "loss": 0.6341,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9672735653376354,
+      "grad_norm": 0.13483721017837524,
+      "learning_rate": 2.807101989435179e-07,
+      "loss": 0.6394,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9675040331873703,
+      "grad_norm": 0.13158808648586273,
+      "learning_rate": 2.767740943133035e-07,
+      "loss": 0.6317,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9677345010371053,
+      "grad_norm": 0.135604590177536,
+      "learning_rate": 2.7286570421574677e-07,
+      "loss": 0.6362,
+      "step": 4199
+    },
+    {
+      "epoch": 0.9679649688868402,
+      "grad_norm": 0.13495594263076782,
+      "learning_rate": 2.6898503082929406e-07,
+      "loss": 0.6278,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9681954367365753,
+      "grad_norm": 0.13914699852466583,
+      "learning_rate": 2.6513207631693184e-07,
+      "loss": 0.6368,
+      "step": 4201
+    },
+    {
+      "epoch": 0.9684259045863102,
+      "grad_norm": 0.1364966183900833,
+      "learning_rate": 2.6130684282621995e-07,
+      "loss": 0.6334,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9686563724360452,
+      "grad_norm": 0.1355530023574829,
+      "learning_rate": 2.575093324892364e-07,
+      "loss": 0.637,
+      "step": 4203
+    },
+    {
+      "epoch": 0.9688868402857801,
+      "grad_norm": 0.1372985690832138,
+      "learning_rate": 2.5373954742263227e-07,
+      "loss": 0.6267,
+      "step": 4204
+    },
+    {
+      "epoch": 0.9691173081355151,
+      "grad_norm": 0.1355789601802826,
+      "learning_rate": 2.4999748972758805e-07,
+      "loss": 0.6418,
+      "step": 4205
+    },
+    {
+      "epoch": 0.96934777598525,
+      "grad_norm": 0.13223963975906372,
+      "learning_rate": 2.462831614898409e-07,
+      "loss": 0.6339,
+      "step": 4206
+    },
+    {
+      "epoch": 0.969578243834985,
+      "grad_norm": 0.1328386813402176,
+      "learning_rate": 2.42596564779668e-07,
+      "loss": 0.6352,
+      "step": 4207
+    },
+    {
+      "epoch": 0.9698087116847199,
+      "grad_norm": 0.13580402731895447,
+      "learning_rate": 2.3893770165189235e-07,
+      "loss": 0.634,
+      "step": 4208
+    },
+    {
+      "epoch": 0.970039179534455,
+      "grad_norm": 0.13300515711307526,
+      "learning_rate": 2.3530657414586598e-07,
+      "loss": 0.6409,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9702696473841899,
+      "grad_norm": 0.13167451322078705,
+      "learning_rate": 2.317031842855033e-07,
+      "loss": 0.6353,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9705001152339249,
+      "grad_norm": 0.13538329303264618,
+      "learning_rate": 2.281275340792477e-07,
+      "loss": 0.6342,
+      "step": 4211
+    },
+    {
+      "epoch": 0.9707305830836598,
+      "grad_norm": 0.1796192079782486,
+      "learning_rate": 2.2457962552007162e-07,
+      "loss": 0.6347,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9709610509333948,
+      "grad_norm": 0.13724969327449799,
+      "learning_rate": 2.2105946058549876e-07,
+      "loss": 0.634,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9711915187831297,
+      "grad_norm": 0.13621461391448975,
+      "learning_rate": 2.1756704123758742e-07,
+      "loss": 0.641,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9714219866328647,
+      "grad_norm": 0.13638457655906677,
+      "learning_rate": 2.141023694229305e-07,
+      "loss": 0.6366,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9716524544825996,
+      "grad_norm": 0.1362111121416092,
+      "learning_rate": 2.1066544707264435e-07,
+      "loss": 0.6483,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9718829223323346,
+      "grad_norm": 0.13554121553897858,
+      "learning_rate": 2.0725627610239107e-07,
+      "loss": 0.6455,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9721133901820695,
+      "grad_norm": 0.13165734708309174,
+      "learning_rate": 2.038748584123562e-07,
+      "loss": 0.6362,
+      "step": 4218
+    },
+    {
+      "epoch": 0.9723438580318046,
+      "grad_norm": 0.13413430750370026,
+      "learning_rate": 2.0052119588727103e-07,
+      "loss": 0.6415,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9725743258815395,
+      "grad_norm": 0.13295698165893555,
+      "learning_rate": 1.9719529039637919e-07,
+      "loss": 0.6413,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9728047937312745,
+      "grad_norm": 0.1375885158777237,
+      "learning_rate": 1.9389714379346446e-07,
+      "loss": 0.6374,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9730352615810095,
+      "grad_norm": 0.1361759752035141,
+      "learning_rate": 1.906267579168286e-07,
+      "loss": 0.6349,
+      "step": 4222
+    },
+    {
+      "epoch": 0.9732657294307444,
+      "grad_norm": 0.13398383557796478,
+      "learning_rate": 1.8738413458931347e-07,
+      "loss": 0.6336,
+      "step": 4223
+    },
+    {
+      "epoch": 0.9734961972804794,
+      "grad_norm": 0.13173305988311768,
+      "learning_rate": 1.8416927561827336e-07,
+      "loss": 0.6304,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9737266651302143,
+      "grad_norm": 0.13570670783519745,
+      "learning_rate": 1.8098218279559708e-07,
+      "loss": 0.6381,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9739571329799493,
+      "grad_norm": 0.13517099618911743,
+      "learning_rate": 1.7782285789769147e-07,
+      "loss": 0.6361,
+      "step": 4226
+    },
+    {
+      "epoch": 0.9741876008296843,
+      "grad_norm": 0.1324930042028427,
+      "learning_rate": 1.7469130268549238e-07,
+      "loss": 0.6377,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9744180686794193,
+      "grad_norm": 0.1336393505334854,
+      "learning_rate": 1.7158751890444803e-07,
+      "loss": 0.6356,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9746485365291542,
+      "grad_norm": 0.13054290413856506,
+      "learning_rate": 1.6851150828453566e-07,
+      "loss": 0.64,
+      "step": 4229
+    },
+    {
+      "epoch": 0.9748790043788892,
+      "grad_norm": 0.130965456366539,
+      "learning_rate": 1.6546327254025052e-07,
+      "loss": 0.6411,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9751094722286241,
+      "grad_norm": 0.13424457609653473,
+      "learning_rate": 1.6244281337060574e-07,
+      "loss": 0.6407,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9753399400783591,
+      "grad_norm": 0.1327114850282669,
+      "learning_rate": 1.5945013245913799e-07,
+      "loss": 0.646,
+      "step": 4232
+    },
+    {
+      "epoch": 0.975570407928094,
+      "grad_norm": 0.13304896652698517,
+      "learning_rate": 1.5648523147388516e-07,
+      "loss": 0.6394,
+      "step": 4233
+    },
+    {
+      "epoch": 0.975800875777829,
+      "grad_norm": 0.1386221945285797,
+      "learning_rate": 1.5354811206741427e-07,
+      "loss": 0.645,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9760313436275639,
+      "grad_norm": 0.13558200001716614,
+      "learning_rate": 1.5063877587681019e-07,
+      "loss": 0.6303,
+      "step": 4235
+    },
+    {
+      "epoch": 0.976261811477299,
+      "grad_norm": 0.13395509123802185,
+      "learning_rate": 1.4775722452366468e-07,
+      "loss": 0.6372,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9764922793270339,
+      "grad_norm": 0.1313483864068985,
+      "learning_rate": 1.4490345961408746e-07,
+      "loss": 0.634,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9767227471767689,
+      "grad_norm": 0.13300201296806335,
+      "learning_rate": 1.4207748273868948e-07,
+      "loss": 0.6448,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9769532150265038,
+      "grad_norm": 0.1316831409931183,
+      "learning_rate": 1.3927929547261632e-07,
+      "loss": 0.6384,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9771836828762388,
+      "grad_norm": 0.13369691371917725,
+      "learning_rate": 1.365088993755037e-07,
+      "loss": 0.6355,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9774141507259737,
+      "grad_norm": 0.13882844150066376,
+      "learning_rate": 1.337662959914998e-07,
+      "loss": 0.6482,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9776446185757087,
+      "grad_norm": 0.13777481019496918,
+      "learning_rate": 1.3105148684927072e-07,
+      "loss": 0.6451,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9778750864254436,
+      "grad_norm": 0.1354648768901825,
+      "learning_rate": 1.283644734619893e-07,
+      "loss": 0.6485,
+      "step": 4243
+    },
+    {
+      "epoch": 0.9781055542751786,
+      "grad_norm": 0.13480645418167114,
+      "learning_rate": 1.257052573273243e-07,
+      "loss": 0.6384,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9783360221249136,
+      "grad_norm": 0.1347130984067917,
+      "learning_rate": 1.2307383992746225e-07,
+      "loss": 0.6335,
+      "step": 4245
+    },
+    {
+      "epoch": 0.9785664899746486,
+      "grad_norm": 0.13522972166538239,
+      "learning_rate": 1.2047022272909102e-07,
+      "loss": 0.6338,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9787969578243835,
+      "grad_norm": 0.137363463640213,
+      "learning_rate": 1.1789440718341093e-07,
+      "loss": 0.6306,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9790274256741185,
+      "grad_norm": 0.13565847277641296,
+      "learning_rate": 1.1534639472611242e-07,
+      "loss": 0.6334,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9792578935238534,
+      "grad_norm": 0.1334226429462433,
+      "learning_rate": 1.1282618677739831e-07,
+      "loss": 0.6332,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9794883613735884,
+      "grad_norm": 0.13705900311470032,
+      "learning_rate": 1.1033378474197276e-07,
+      "loss": 0.6332,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9797188292233233,
+      "grad_norm": 0.1340656876564026,
+      "learning_rate": 1.0786919000903562e-07,
+      "loss": 0.6424,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9799492970730583,
+      "grad_norm": 0.13730952143669128,
+      "learning_rate": 1.054324039523047e-07,
+      "loss": 0.6377,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9801797649227932,
+      "grad_norm": 0.1353830099105835,
+      "learning_rate": 1.0302342792997688e-07,
+      "loss": 0.6389,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9804102327725283,
+      "grad_norm": 0.13165293633937836,
+      "learning_rate": 1.0064226328476145e-07,
+      "loss": 0.6366,
+      "step": 4254
+    },
+    {
+      "epoch": 0.9806407006222632,
+      "grad_norm": 0.12924884259700775,
+      "learning_rate": 9.828891134385786e-08,
+      "loss": 0.6363,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9808711684719982,
+      "grad_norm": 0.13422457873821259,
+      "learning_rate": 9.596337341897243e-08,
+      "loss": 0.635,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9811016363217331,
+      "grad_norm": 0.1396060436964035,
+      "learning_rate": 9.366565080630163e-08,
+      "loss": 0.6328,
+      "step": 4257
+    },
+    {
+      "epoch": 0.9813321041714681,
+      "grad_norm": 0.1325884759426117,
+      "learning_rate": 9.139574478654322e-08,
+      "loss": 0.6278,
+      "step": 4258
+    },
+    {
+      "epoch": 0.981562572021203,
+      "grad_norm": 0.13009661436080933,
+      "learning_rate": 8.915365662488518e-08,
+      "loss": 0.6329,
+      "step": 4259
+    },
+    {
+      "epoch": 0.981793039870938,
+      "grad_norm": 0.13182280957698822,
+      "learning_rate": 8.693938757101672e-08,
+      "loss": 0.6293,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9820235077206729,
+      "grad_norm": 0.13452081382274628,
+      "learning_rate": 8.475293885911173e-08,
+      "loss": 0.6298,
+      "step": 4261
+    },
+    {
+      "epoch": 0.982253975570408,
+      "grad_norm": 0.13344185054302216,
+      "learning_rate": 8.259431170785647e-08,
+      "loss": 0.6335,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9824844434201429,
+      "grad_norm": 0.1369807869195938,
+      "learning_rate": 8.046350732041075e-08,
+      "loss": 0.632,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9827149112698779,
+      "grad_norm": 0.13430048525333405,
+      "learning_rate": 7.836052688443007e-08,
+      "loss": 0.6378,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9829453791196128,
+      "grad_norm": 0.13494747877120972,
+      "learning_rate": 7.628537157207128e-08,
+      "loss": 0.6394,
+      "step": 4265
+    },
+    {
+      "epoch": 0.9831758469693478,
+      "grad_norm": 0.13669085502624512,
+      "learning_rate": 7.423804253997579e-08,
+      "loss": 0.6392,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9834063148190827,
+      "grad_norm": 0.13560469448566437,
+      "learning_rate": 7.221854092926971e-08,
+      "loss": 0.6522,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9836367826688177,
+      "grad_norm": 0.13016438484191895,
+      "learning_rate": 7.022686786558042e-08,
+      "loss": 0.6298,
+      "step": 4268
+    },
+    {
+      "epoch": 0.9838672505185526,
+      "grad_norm": 0.13416080176830292,
+      "learning_rate": 6.826302445901989e-08,
+      "loss": 0.6433,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9840977183682876,
+      "grad_norm": 0.13181835412979126,
+      "learning_rate": 6.632701180418476e-08,
+      "loss": 0.6342,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9843281862180225,
+      "grad_norm": 0.13344010710716248,
+      "learning_rate": 6.441883098015633e-08,
+      "loss": 0.6337,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9845586540677576,
+      "grad_norm": 0.13779869675636292,
+      "learning_rate": 6.253848305052268e-08,
+      "loss": 0.6446,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9847891219174925,
+      "grad_norm": 0.13841082155704498,
+      "learning_rate": 6.06859690633288e-08,
+      "loss": 0.6393,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9850195897672275,
+      "grad_norm": 0.13432055711746216,
+      "learning_rate": 5.886129005113206e-08,
+      "loss": 0.6352,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9852500576169624,
+      "grad_norm": 0.1333775520324707,
+      "learning_rate": 5.706444703096336e-08,
+      "loss": 0.6288,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9854805254666974,
+      "grad_norm": 0.13299840688705444,
+      "learning_rate": 5.5295441004332704e-08,
+      "loss": 0.6288,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9857109933164323,
+      "grad_norm": 0.1346500962972641,
+      "learning_rate": 5.355427295725135e-08,
+      "loss": 0.6369,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9859414611661673,
+      "grad_norm": 0.13148584961891174,
+      "learning_rate": 5.184094386019855e-08,
+      "loss": 0.6326,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9861719290159022,
+      "grad_norm": 0.1373184323310852,
+      "learning_rate": 5.0155454668149304e-08,
+      "loss": 0.6345,
+      "step": 4279
+    },
+    {
+      "epoch": 0.9864023968656372,
+      "grad_norm": 0.1355026662349701,
+      "learning_rate": 4.849780632054657e-08,
+      "loss": 0.6454,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9866328647153723,
+      "grad_norm": 0.13367265462875366,
+      "learning_rate": 4.6867999741323496e-08,
+      "loss": 0.6447,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9868633325651072,
+      "grad_norm": 0.134071946144104,
+      "learning_rate": 4.5266035838903434e-08,
+      "loss": 0.6395,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9870938004148422,
+      "grad_norm": 0.1350136399269104,
+      "learning_rate": 4.3691915506177686e-08,
+      "loss": 0.6453,
+      "step": 4283
+    },
+    {
+      "epoch": 0.9873242682645771,
+      "grad_norm": 0.13533277809619904,
+      "learning_rate": 4.21456396205222e-08,
+      "loss": 0.6347,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9875547361143121,
+      "grad_norm": 0.13973018527030945,
+      "learning_rate": 4.062720904379757e-08,
+      "loss": 0.6393,
+      "step": 4285
+    },
+    {
+      "epoch": 0.987785203964047,
+      "grad_norm": 0.13376709818840027,
+      "learning_rate": 3.913662462233791e-08,
+      "loss": 0.6352,
+      "step": 4286
+    },
+    {
+      "epoch": 0.988015671813782,
+      "grad_norm": 0.13301095366477966,
+      "learning_rate": 3.767388718696197e-08,
+      "loss": 0.6446,
+      "step": 4287
+    },
+    {
+      "epoch": 0.9882461396635169,
+      "grad_norm": 0.13637404143810272,
+      "learning_rate": 3.6238997552956456e-08,
+      "loss": 0.6414,
+      "step": 4288
+    },
+    {
+      "epoch": 0.988476607513252,
+      "grad_norm": 0.13000015914440155,
+      "learning_rate": 3.483195652010385e-08,
+      "loss": 0.6326,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9887070753629869,
+      "grad_norm": 0.1287553757429123,
+      "learning_rate": 3.3452764872649036e-08,
+      "loss": 0.6359,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9889375432127219,
+      "grad_norm": 0.13659948110580444,
+      "learning_rate": 3.210142337932709e-08,
+      "loss": 0.6306,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9891680110624568,
+      "grad_norm": 0.13787119090557098,
+      "learning_rate": 3.0777932793335516e-08,
+      "loss": 0.6243,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9893984789121918,
+      "grad_norm": 0.13213881850242615,
+      "learning_rate": 2.948229385236201e-08,
+      "loss": 0.6381,
+      "step": 4293
+    },
+    {
+      "epoch": 0.9896289467619267,
+      "grad_norm": 0.13811156153678894,
+      "learning_rate": 2.8214507278556678e-08,
+      "loss": 0.6412,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9898594146116617,
+      "grad_norm": 0.1367308348417282,
+      "learning_rate": 2.6974573778565383e-08,
+      "loss": 0.6411,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9900898824613966,
+      "grad_norm": 0.1384420096874237,
+      "learning_rate": 2.5762494043485296e-08,
+      "loss": 0.6355,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9903203503111316,
+      "grad_norm": 0.13190501928329468,
+      "learning_rate": 2.4578268748909338e-08,
+      "loss": 0.6296,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9905508181608665,
+      "grad_norm": 0.13569803535938263,
+      "learning_rate": 2.3421898554892852e-08,
+      "loss": 0.6342,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9907812860106016,
+      "grad_norm": 0.13529643416404724,
+      "learning_rate": 2.229338410597026e-08,
+      "loss": 0.6311,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9910117538603365,
+      "grad_norm": 0.13517087697982788,
+      "learning_rate": 2.1192726031143974e-08,
+      "loss": 0.6434,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9912422217100715,
+      "grad_norm": 0.13566431403160095,
+      "learning_rate": 2.0119924943901025e-08,
+      "loss": 0.6371,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9914726895598064,
+      "grad_norm": 0.1320231556892395,
+      "learning_rate": 1.9074981442185336e-08,
+      "loss": 0.6346,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9917031574095414,
+      "grad_norm": 0.13202235102653503,
+      "learning_rate": 1.8057896108436558e-08,
+      "loss": 0.6322,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9919336252592763,
+      "grad_norm": 0.13193956017494202,
+      "learning_rate": 1.7068669509545665e-08,
+      "loss": 0.6393,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9921640931090113,
+      "grad_norm": 0.13464799523353577,
+      "learning_rate": 1.6107302196882724e-08,
+      "loss": 0.6363,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9923945609587462,
+      "grad_norm": 0.13650983572006226,
+      "learning_rate": 1.5173794706291324e-08,
+      "loss": 0.6332,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9926250288084812,
+      "grad_norm": 0.1360529065132141,
+      "learning_rate": 1.4268147558088585e-08,
+      "loss": 0.6251,
+      "step": 4307
+    },
+    {
+      "epoch": 0.9928554966582162,
+      "grad_norm": 0.1333739310503006,
+      "learning_rate": 1.3390361257059614e-08,
+      "loss": 0.6345,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9930859645079512,
+      "grad_norm": 0.13426019251346588,
+      "learning_rate": 1.2540436292463043e-08,
+      "loss": 0.6416,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9933164323576861,
+      "grad_norm": 0.13250836730003357,
+      "learning_rate": 1.1718373138019933e-08,
+      "loss": 0.6278,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9935469002074211,
+      "grad_norm": 0.13704052567481995,
+      "learning_rate": 1.0924172251941534e-08,
+      "loss": 0.6428,
+      "step": 4311
+    },
+    {
+      "epoch": 0.993777368057156,
+      "grad_norm": 0.13682594895362854,
+      "learning_rate": 1.0157834076879313e-08,
+      "loss": 0.6392,
+      "step": 4312
+    },
+    {
+      "epoch": 0.994007835906891,
+      "grad_norm": 0.13360513746738434,
+      "learning_rate": 9.419359039986032e-09,
+      "loss": 0.6351,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9942383037566259,
+      "grad_norm": 0.1286047101020813,
+      "learning_rate": 8.70874755286577e-09,
+      "loss": 0.6381,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9944687716063609,
+      "grad_norm": 0.13860009610652924,
+      "learning_rate": 8.026000011596146e-09,
+      "loss": 0.642,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9946992394560958,
+      "grad_norm": 0.1374213993549347,
+      "learning_rate": 7.371116796717203e-09,
+      "loss": 0.6356,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9949297073058309,
+      "grad_norm": 0.13426125049591064,
+      "learning_rate": 6.74409827325917e-09,
+      "loss": 0.6325,
+      "step": 4317
+    },
+    {
+      "epoch": 0.9951601751555658,
+      "grad_norm": 0.13406482338905334,
+      "learning_rate": 6.144944790692497e-09,
+      "loss": 0.6422,
+      "step": 4318
+    },
+    {
+      "epoch": 0.9953906430053008,
+      "grad_norm": 0.1396821290254593,
+      "learning_rate": 5.573656682977824e-09,
+      "loss": 0.6443,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9956211108550357,
+      "grad_norm": 0.12982842326164246,
+      "learning_rate": 5.030234268543765e-09,
+      "loss": 0.6245,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9958515787047707,
+      "grad_norm": 0.1334078311920166,
+      "learning_rate": 4.514677850270266e-09,
+      "loss": 0.6401,
+      "step": 4321
+    },
+    {
+      "epoch": 0.9960820465545056,
+      "grad_norm": 0.13134385645389557,
+      "learning_rate": 4.0269877155219016e-09,
+      "loss": 0.6331,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9963125144042406,
+      "grad_norm": 0.1397440880537033,
+      "learning_rate": 3.567164136120127e-09,
+      "loss": 0.6252,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9965429822539755,
+      "grad_norm": 0.13442939519882202,
+      "learning_rate": 3.1352073683654783e-09,
+      "loss": 0.635,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9967734501037105,
+      "grad_norm": 0.13672947883605957,
+      "learning_rate": 2.7311176530209203e-09,
+      "loss": 0.6365,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9970039179534455,
+      "grad_norm": 0.13293464481830597,
+      "learning_rate": 2.3548952153118476e-09,
+      "loss": 0.6337,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9972343858031805,
+      "grad_norm": 0.1351083517074585,
+      "learning_rate": 2.0065402649371845e-09,
+      "loss": 0.633,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9974648536529154,
+      "grad_norm": 0.13411957025527954,
+      "learning_rate": 1.6860529960638361e-09,
+      "loss": 0.6376,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9976953215026504,
+      "grad_norm": 0.1313239187002182,
+      "learning_rate": 1.3934335873155846e-09,
+      "loss": 0.6344,
+      "step": 4329
+    },
+    {
+      "epoch": 0.9979257893523853,
+      "grad_norm": 0.13476663827896118,
+      "learning_rate": 1.1286822018008458e-09,
+      "loss": 0.6432,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9981562572021203,
+      "grad_norm": 0.13341675698757172,
+      "learning_rate": 8.917989870849131e-10,
+      "loss": 0.6288,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9983867250518552,
+      "grad_norm": 0.13334329426288605,
+      "learning_rate": 6.827840751955083e-10,
+      "loss": 0.6403,
+      "step": 4332
+    },
+    {
+      "epoch": 0.9986171929015902,
+      "grad_norm": 0.135373055934906,
+      "learning_rate": 5.016375826394359e-10,
+      "loss": 0.6377,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9988476607513251,
+      "grad_norm": 0.13281653821468353,
+      "learning_rate": 3.4835961037482655e-10,
+      "loss": 0.6261,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9990781286010602,
+      "grad_norm": 0.1364908516407013,
+      "learning_rate": 2.2295024383889306e-10,
+      "loss": 0.6386,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9993085964507951,
+      "grad_norm": 0.1349654197692871,
+      "learning_rate": 1.2540955293682822e-10,
+      "loss": 0.6366,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9995390643005301,
+      "grad_norm": 0.14018984138965607,
+      "learning_rate": 5.573759202515127e-11,
+      "loss": 0.6342,
+      "step": 4337
+    },
+    {
+      "epoch": 0.999769532150265,
+      "grad_norm": 0.1340036541223526,
+      "learning_rate": 1.3934399950565891e-11,
+      "loss": 0.6367,
+      "step": 4338
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.13604800403118134,
+      "learning_rate": 0.0,
+      "loss": 0.6404,
+      "step": 4339
+    },
+    {
+      "epoch": 1.0,
+      "step": 4339,
+      "total_flos": 1.6526087271915035e+20,
+      "train_loss": 0.7063895864925178,
+      "train_runtime": 30118.2929,
+      "train_samples_per_second": 589.978,
+      "train_steps_per_second": 0.144
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 4339,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 3000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6526087271915035e+20,
+  "train_batch_size": 32,
+  "trial_name": null,
+  "trial_params": null
+}