diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20629 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2941,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00034002040122407346,
+      "grad_norm": 7.957215845157106,
+      "learning_rate": 1.1235955056179776e-07,
+      "loss": 1.265,
+      "step": 1
+    },
+    {
+      "epoch": 0.0006800408024481469,
+      "grad_norm": 13.156657121104576,
+      "learning_rate": 2.247191011235955e-07,
+      "loss": 1.1163,
+      "step": 2
+    },
+    {
+      "epoch": 0.0010200612036722204,
+      "grad_norm": 8.234411807150657,
+      "learning_rate": 3.3707865168539325e-07,
+      "loss": 1.1117,
+      "step": 3
+    },
+    {
+      "epoch": 0.0013600816048962938,
+      "grad_norm": 8.378852124626784,
+      "learning_rate": 4.49438202247191e-07,
+      "loss": 1.0045,
+      "step": 4
+    },
+    {
+      "epoch": 0.0017001020061203672,
+      "grad_norm": 5.0871456226741785,
+      "learning_rate": 5.617977528089888e-07,
+      "loss": 1.1061,
+      "step": 5
+    },
+    {
+      "epoch": 0.002040122407344441,
+      "grad_norm": 12.113028237990287,
+      "learning_rate": 6.741573033707865e-07,
+      "loss": 0.978,
+      "step": 6
+    },
+    {
+      "epoch": 0.0023801428085685142,
+      "grad_norm": 7.182446599540591,
+      "learning_rate": 7.865168539325843e-07,
+      "loss": 1.2117,
+      "step": 7
+    },
+    {
+      "epoch": 0.0027201632097925877,
+      "grad_norm": 8.794165320435372,
+      "learning_rate": 8.98876404494382e-07,
+      "loss": 1.0302,
+      "step": 8
+    },
+    {
+      "epoch": 0.003060183611016661,
+      "grad_norm": 6.19432478948948,
+      "learning_rate": 1.01123595505618e-06,
+      "loss": 1.1328,
+      "step": 9
+    },
+    {
+      "epoch": 0.0034002040122407345,
+      "grad_norm": 8.815917265181259,
+      "learning_rate": 1.1235955056179777e-06,
+      "loss": 1.0271,
+      "step": 10
+    },
+    {
+      "epoch": 0.003740224413464808,
+      "grad_norm": 3.4552395124739923,
+      "learning_rate": 1.2359550561797752e-06,
+      "loss": 1.1775,
+      "step": 11
+    },
+    {
+      "epoch": 0.004080244814688882,
+      "grad_norm": 3.920780003164372,
+      "learning_rate": 1.348314606741573e-06,
+      "loss": 1.1197,
+      "step": 12
+    },
+    {
+      "epoch": 0.004420265215912955,
+      "grad_norm": 9.503601873541784,
+      "learning_rate": 1.4606741573033708e-06,
+      "loss": 1.0117,
+      "step": 13
+    },
+    {
+      "epoch": 0.0047602856171370285,
+      "grad_norm": 3.8392791998777698,
+      "learning_rate": 1.5730337078651686e-06,
+      "loss": 1.1943,
+      "step": 14
+    },
+    {
+      "epoch": 0.0051003060183611015,
+      "grad_norm": 16.880294672259833,
+      "learning_rate": 1.6853932584269663e-06,
+      "loss": 0.9484,
+      "step": 15
+    },
+    {
+      "epoch": 0.005440326419585175,
+      "grad_norm": 3.7585745209663135,
+      "learning_rate": 1.797752808988764e-06,
+      "loss": 1.0903,
+      "step": 16
+    },
+    {
+      "epoch": 0.005780346820809248,
+      "grad_norm": 4.0811771911781705,
+      "learning_rate": 1.910112359550562e-06,
+      "loss": 0.9225,
+      "step": 17
+    },
+    {
+      "epoch": 0.006120367222033322,
+      "grad_norm": 5.5863711220456835,
+      "learning_rate": 2.02247191011236e-06,
+      "loss": 1.0673,
+      "step": 18
+    },
+    {
+      "epoch": 0.006460387623257395,
+      "grad_norm": 3.1676452672990454,
+      "learning_rate": 2.1348314606741574e-06,
+      "loss": 0.9291,
+      "step": 19
+    },
+    {
+      "epoch": 0.006800408024481469,
+      "grad_norm": 3.611139125930507,
+      "learning_rate": 2.2471910112359554e-06,
+      "loss": 0.8766,
+      "step": 20
+    },
+    {
+      "epoch": 0.007140428425705543,
+      "grad_norm": 2.4985925611214093,
+      "learning_rate": 2.359550561797753e-06,
+      "loss": 0.8919,
+      "step": 21
+    },
+    {
+      "epoch": 0.007480448826929616,
+      "grad_norm": 3.1107777199577735,
+      "learning_rate": 2.4719101123595505e-06,
+      "loss": 0.9591,
+      "step": 22
+    },
+    {
+      "epoch": 0.007820469228153689,
+      "grad_norm": 2.4206200369165054,
+      "learning_rate": 2.584269662921349e-06,
+      "loss": 0.9344,
+      "step": 23
+    },
+    {
+      "epoch": 0.008160489629377763,
+      "grad_norm": 3.006651205729068,
+      "learning_rate": 2.696629213483146e-06,
+      "loss": 0.929,
+      "step": 24
+    },
+    {
+      "epoch": 0.008500510030601836,
+      "grad_norm": 4.211242050455176,
+      "learning_rate": 2.8089887640449444e-06,
+      "loss": 0.922,
+      "step": 25
+    },
+    {
+      "epoch": 0.00884053043182591,
+      "grad_norm": 2.928334204696846,
+      "learning_rate": 2.9213483146067416e-06,
+      "loss": 0.8109,
+      "step": 26
+    },
+    {
+      "epoch": 0.009180550833049982,
+      "grad_norm": 2.920501950315103,
+      "learning_rate": 3.03370786516854e-06,
+      "loss": 0.9918,
+      "step": 27
+    },
+    {
+      "epoch": 0.009520571234274057,
+      "grad_norm": 2.6627931747349463,
+      "learning_rate": 3.146067415730337e-06,
+      "loss": 1.0249,
+      "step": 28
+    },
+    {
+      "epoch": 0.00986059163549813,
+      "grad_norm": 2.2391523692025945,
+      "learning_rate": 3.258426966292135e-06,
+      "loss": 0.8657,
+      "step": 29
+    },
+    {
+      "epoch": 0.010200612036722203,
+      "grad_norm": 2.3439476348012924,
+      "learning_rate": 3.3707865168539327e-06,
+      "loss": 0.8669,
+      "step": 30
+    },
+    {
+      "epoch": 0.010540632437946278,
+      "grad_norm": 2.6299632657140646,
+      "learning_rate": 3.4831460674157306e-06,
+      "loss": 0.9279,
+      "step": 31
+    },
+    {
+      "epoch": 0.01088065283917035,
+      "grad_norm": 2.5299601625034702,
+      "learning_rate": 3.595505617977528e-06,
+      "loss": 0.9859,
+      "step": 32
+    },
+    {
+      "epoch": 0.011220673240394424,
+      "grad_norm": 1.869054216542114,
+      "learning_rate": 3.707865168539326e-06,
+      "loss": 0.9683,
+      "step": 33
+    },
+    {
+      "epoch": 0.011560693641618497,
+      "grad_norm": 2.641701125076694,
+      "learning_rate": 3.820224719101124e-06,
+      "loss": 0.9049,
+      "step": 34
+    },
+    {
+      "epoch": 0.011900714042842571,
+      "grad_norm": 2.3075105518731167,
+      "learning_rate": 3.932584269662922e-06,
+      "loss": 0.95,
+      "step": 35
+    },
+    {
+      "epoch": 0.012240734444066644,
+      "grad_norm": 2.3217690220079943,
+      "learning_rate": 4.04494382022472e-06,
+      "loss": 0.9225,
+      "step": 36
+    },
+    {
+      "epoch": 0.012580754845290717,
+      "grad_norm": 2.480896383955115,
+      "learning_rate": 4.157303370786518e-06,
+      "loss": 0.8839,
+      "step": 37
+    },
+    {
+      "epoch": 0.01292077524651479,
+      "grad_norm": 2.151833792105812,
+      "learning_rate": 4.269662921348315e-06,
+      "loss": 0.7775,
+      "step": 38
+    },
+    {
+      "epoch": 0.013260795647738865,
+      "grad_norm": 4.211040389687315,
+      "learning_rate": 4.382022471910113e-06,
+      "loss": 0.9096,
+      "step": 39
+    },
+    {
+      "epoch": 0.013600816048962938,
+      "grad_norm": 2.064947551303069,
+      "learning_rate": 4.494382022471911e-06,
+      "loss": 0.9284,
+      "step": 40
+    },
+    {
+      "epoch": 0.01394083645018701,
+      "grad_norm": 2.0660909038422526,
+      "learning_rate": 4.606741573033709e-06,
+      "loss": 0.925,
+      "step": 41
+    },
+    {
+      "epoch": 0.014280856851411085,
+      "grad_norm": 4.901236022148853,
+      "learning_rate": 4.719101123595506e-06,
+      "loss": 0.8714,
+      "step": 42
+    },
+    {
+      "epoch": 0.014620877252635158,
+      "grad_norm": 4.502266868053889,
+      "learning_rate": 4.831460674157304e-06,
+      "loss": 0.8047,
+      "step": 43
+    },
+    {
+      "epoch": 0.014960897653859231,
+      "grad_norm": 2.4077409650266337,
+      "learning_rate": 4.943820224719101e-06,
+      "loss": 0.913,
+      "step": 44
+    },
+    {
+      "epoch": 0.015300918055083304,
+      "grad_norm": 4.793105624055668,
+      "learning_rate": 5.0561797752809e-06,
+      "loss": 0.8195,
+      "step": 45
+    },
+    {
+      "epoch": 0.015640938456307377,
+      "grad_norm": 4.106559075152421,
+      "learning_rate": 5.168539325842698e-06,
+      "loss": 0.771,
+      "step": 46
+    },
+    {
+      "epoch": 0.015980958857531452,
+      "grad_norm": 2.95021911633751,
+      "learning_rate": 5.280898876404494e-06,
+      "loss": 0.8879,
+      "step": 47
+    },
+    {
+      "epoch": 0.016320979258755527,
+      "grad_norm": 2.5286647211984956,
+      "learning_rate": 5.393258426966292e-06,
+      "loss": 0.8658,
+      "step": 48
+    },
+    {
+      "epoch": 0.016660999659979598,
+      "grad_norm": 2.8336405976774373,
+      "learning_rate": 5.50561797752809e-06,
+      "loss": 0.8408,
+      "step": 49
+    },
+    {
+      "epoch": 0.017001020061203673,
+      "grad_norm": 2.1850180954836187,
+      "learning_rate": 5.617977528089889e-06,
+      "loss": 0.9112,
+      "step": 50
+    },
+    {
+      "epoch": 0.017341040462427744,
+      "grad_norm": 2.06887498219481,
+      "learning_rate": 5.730337078651685e-06,
+      "loss": 0.7897,
+      "step": 51
+    },
+    {
+      "epoch": 0.01768106086365182,
+      "grad_norm": 4.3341123167435125,
+      "learning_rate": 5.842696629213483e-06,
+      "loss": 0.811,
+      "step": 52
+    },
+    {
+      "epoch": 0.018021081264875893,
+      "grad_norm": 2.14419920475312,
+      "learning_rate": 5.955056179775281e-06,
+      "loss": 0.9512,
+      "step": 53
+    },
+    {
+      "epoch": 0.018361101666099965,
+      "grad_norm": 2.5786480499173177,
+      "learning_rate": 6.06741573033708e-06,
+      "loss": 0.9029,
+      "step": 54
+    },
+    {
+      "epoch": 0.01870112206732404,
+      "grad_norm": 3.3621255490379833,
+      "learning_rate": 6.179775280898876e-06,
+      "loss": 0.8179,
+      "step": 55
+    },
+    {
+      "epoch": 0.019041142468548114,
+      "grad_norm": 2.0145825477143675,
+      "learning_rate": 6.292134831460674e-06,
+      "loss": 0.924,
+      "step": 56
+    },
+    {
+      "epoch": 0.019381162869772185,
+      "grad_norm": 2.1614514347624625,
+      "learning_rate": 6.404494382022472e-06,
+      "loss": 0.8009,
+      "step": 57
+    },
+    {
+      "epoch": 0.01972118327099626,
+      "grad_norm": 2.056785775033519,
+      "learning_rate": 6.51685393258427e-06,
+      "loss": 0.8702,
+      "step": 58
+    },
+    {
+      "epoch": 0.020061203672220335,
+      "grad_norm": 2.3474912782953172,
+      "learning_rate": 6.629213483146067e-06,
+      "loss": 0.8707,
+      "step": 59
+    },
+    {
+      "epoch": 0.020401224073444406,
+      "grad_norm": 2.735017646466919,
+      "learning_rate": 6.741573033707865e-06,
+      "loss": 0.8172,
+      "step": 60
+    },
+    {
+      "epoch": 0.02074124447466848,
+      "grad_norm": 2.422521293220476,
+      "learning_rate": 6.853932584269663e-06,
+      "loss": 0.9169,
+      "step": 61
+    },
+    {
+      "epoch": 0.021081264875892555,
+      "grad_norm": 3.808006273477038,
+      "learning_rate": 6.966292134831461e-06,
+      "loss": 0.9115,
+      "step": 62
+    },
+    {
+      "epoch": 0.021421285277116626,
+      "grad_norm": 6.385476467042873,
+      "learning_rate": 7.078651685393258e-06,
+      "loss": 0.9251,
+      "step": 63
+    },
+    {
+      "epoch": 0.0217613056783407,
+      "grad_norm": 2.056529939521199,
+      "learning_rate": 7.191011235955056e-06,
+      "loss": 0.871,
+      "step": 64
+    },
+    {
+      "epoch": 0.022101326079564772,
+      "grad_norm": 3.488639073828488,
+      "learning_rate": 7.303370786516854e-06,
+      "loss": 0.8088,
+      "step": 65
+    },
+    {
+      "epoch": 0.022441346480788847,
+      "grad_norm": 2.085309229849527,
+      "learning_rate": 7.415730337078652e-06,
+      "loss": 0.9016,
+      "step": 66
+    },
+    {
+      "epoch": 0.022781366882012922,
+      "grad_norm": 5.044997989735368,
+      "learning_rate": 7.5280898876404495e-06,
+      "loss": 1.0085,
+      "step": 67
+    },
+    {
+      "epoch": 0.023121387283236993,
+      "grad_norm": 4.002898882880522,
+      "learning_rate": 7.640449438202247e-06,
+      "loss": 0.7977,
+      "step": 68
+    },
+    {
+      "epoch": 0.023461407684461068,
+      "grad_norm": 2.189417532279016,
+      "learning_rate": 7.752808988764046e-06,
+      "loss": 0.749,
+      "step": 69
+    },
+    {
+      "epoch": 0.023801428085685142,
+      "grad_norm": 1.9250509612156395,
+      "learning_rate": 7.865168539325843e-06,
+      "loss": 0.9402,
+      "step": 70
+    },
+    {
+      "epoch": 0.024141448486909214,
+      "grad_norm": 8.213532430857596,
+      "learning_rate": 7.97752808988764e-06,
+      "loss": 0.8401,
+      "step": 71
+    },
+    {
+      "epoch": 0.02448146888813329,
+      "grad_norm": 2.0693669751843085,
+      "learning_rate": 8.08988764044944e-06,
+      "loss": 0.8317,
+      "step": 72
+    },
+    {
+      "epoch": 0.024821489289357363,
+      "grad_norm": 4.854843546953979,
+      "learning_rate": 8.202247191011237e-06,
+      "loss": 0.7864,
+      "step": 73
+    },
+    {
+      "epoch": 0.025161509690581434,
+      "grad_norm": 1.9256245489125812,
+      "learning_rate": 8.314606741573035e-06,
+      "loss": 0.8605,
+      "step": 74
+    },
+    {
+      "epoch": 0.02550153009180551,
+      "grad_norm": 3.3854334242945736,
+      "learning_rate": 8.426966292134832e-06,
+      "loss": 0.994,
+      "step": 75
+    },
+    {
+      "epoch": 0.02584155049302958,
+      "grad_norm": 3.4183811142773135,
+      "learning_rate": 8.53932584269663e-06,
+      "loss": 0.8632,
+      "step": 76
+    },
+    {
+      "epoch": 0.026181570894253655,
+      "grad_norm": 2.2762352809156416,
+      "learning_rate": 8.651685393258428e-06,
+      "loss": 0.8118,
+      "step": 77
+    },
+    {
+      "epoch": 0.02652159129547773,
+      "grad_norm": 2.297584618749765,
+      "learning_rate": 8.764044943820226e-06,
+      "loss": 0.8672,
+      "step": 78
+    },
+    {
+      "epoch": 0.0268616116967018,
+      "grad_norm": 3.23414736176019,
+      "learning_rate": 8.876404494382023e-06,
+      "loss": 0.6984,
+      "step": 79
+    },
+    {
+      "epoch": 0.027201632097925876,
+      "grad_norm": 2.970866674193053,
+      "learning_rate": 8.988764044943822e-06,
+      "loss": 0.9025,
+      "step": 80
+    },
+    {
+      "epoch": 0.02754165249914995,
+      "grad_norm": 2.5731909072039727,
+      "learning_rate": 9.101123595505619e-06,
+      "loss": 0.8281,
+      "step": 81
+    },
+    {
+      "epoch": 0.02788167290037402,
+      "grad_norm": 2.087322150347932,
+      "learning_rate": 9.213483146067417e-06,
+      "loss": 0.9687,
+      "step": 82
+    },
+    {
+      "epoch": 0.028221693301598096,
+      "grad_norm": 2.1556977588073867,
+      "learning_rate": 9.325842696629213e-06,
+      "loss": 0.801,
+      "step": 83
+    },
+    {
+      "epoch": 0.02856171370282217,
+      "grad_norm": 2.0239695578983237,
+      "learning_rate": 9.438202247191012e-06,
+      "loss": 0.8507,
+      "step": 84
+    },
+    {
+      "epoch": 0.028901734104046242,
+      "grad_norm": 2.8375743876638633,
+      "learning_rate": 9.55056179775281e-06,
+      "loss": 0.8769,
+      "step": 85
+    },
+    {
+      "epoch": 0.029241754505270317,
+      "grad_norm": 2.258009150645209,
+      "learning_rate": 9.662921348314608e-06,
+      "loss": 0.7804,
+      "step": 86
+    },
+    {
+      "epoch": 0.029581774906494388,
+      "grad_norm": 2.468189158455857,
+      "learning_rate": 9.775280898876405e-06,
+      "loss": 0.879,
+      "step": 87
+    },
+    {
+      "epoch": 0.029921795307718463,
+      "grad_norm": 2.5014523937082482,
+      "learning_rate": 9.887640449438202e-06,
+      "loss": 0.8627,
+      "step": 88
+    },
+    {
+      "epoch": 0.030261815708942538,
+      "grad_norm": 2.134005133865966,
+      "learning_rate": 1e-05,
+      "loss": 0.7324,
+      "step": 89
+    },
+    {
+      "epoch": 0.03060183611016661,
+      "grad_norm": 3.2988329246550396,
+      "learning_rate": 9.999996966523272e-06,
+      "loss": 0.8279,
+      "step": 90
+    },
+    {
+      "epoch": 0.030941856511390683,
+      "grad_norm": 3.35459533797809,
+      "learning_rate": 9.999987866096762e-06,
+      "loss": 0.8842,
+      "step": 91
+    },
+    {
+      "epoch": 0.031281876912614755,
+      "grad_norm": 2.346746709615949,
+      "learning_rate": 9.999972698731516e-06,
+      "loss": 0.8541,
+      "step": 92
+    },
+    {
+      "epoch": 0.03162189731383883,
+      "grad_norm": 2.5077318815610816,
+      "learning_rate": 9.999951464445938e-06,
+      "loss": 0.874,
+      "step": 93
+    },
+    {
+      "epoch": 0.031961917715062904,
+      "grad_norm": 6.007848319671871,
+      "learning_rate": 9.99992416326579e-06,
+      "loss": 0.7933,
+      "step": 94
+    },
+    {
+      "epoch": 0.032301938116286975,
+      "grad_norm": 3.1450126786992882,
+      "learning_rate": 9.999890795224206e-06,
+      "loss": 0.8713,
+      "step": 95
+    },
+    {
+      "epoch": 0.032641958517511054,
+      "grad_norm": 2.0402024490282633,
+      "learning_rate": 9.999851360361666e-06,
+      "loss": 0.801,
+      "step": 96
+    },
+    {
+      "epoch": 0.032981978918735125,
+      "grad_norm": 4.045944747267887,
+      "learning_rate": 9.999805858726026e-06,
+      "loss": 0.8282,
+      "step": 97
+    },
+    {
+      "epoch": 0.033321999319959196,
+      "grad_norm": 2.7667039685581947,
+      "learning_rate": 9.999754290372496e-06,
+      "loss": 0.9823,
+      "step": 98
+    },
+    {
+      "epoch": 0.033662019721183274,
+      "grad_norm": 1.9693833148302338,
+      "learning_rate": 9.999696655363646e-06,
+      "loss": 0.8958,
+      "step": 99
+    },
+    {
+      "epoch": 0.034002040122407345,
+      "grad_norm": 1.8144739667930376,
+      "learning_rate": 9.999632953769413e-06,
+      "loss": 0.865,
+      "step": 100
+    },
+    {
+      "epoch": 0.03434206052363142,
+      "grad_norm": 2.054769480683066,
+      "learning_rate": 9.99956318566709e-06,
+      "loss": 0.7375,
+      "step": 101
+    },
+    {
+      "epoch": 0.03468208092485549,
+      "grad_norm": 2.3634788615414535,
+      "learning_rate": 9.999487351141333e-06,
+      "loss": 0.7926,
+      "step": 102
+    },
+    {
+      "epoch": 0.035022101326079566,
+      "grad_norm": 2.432946102833199,
+      "learning_rate": 9.999405450284161e-06,
+      "loss": 0.9227,
+      "step": 103
+    },
+    {
+      "epoch": 0.03536212172730364,
+      "grad_norm": 2.01036703041841,
+      "learning_rate": 9.999317483194948e-06,
+      "loss": 0.836,
+      "step": 104
+    },
+    {
+      "epoch": 0.03570214212852771,
+      "grad_norm": 11.065759310951304,
+      "learning_rate": 9.999223449980434e-06,
+      "loss": 0.8741,
+      "step": 105
+    },
+    {
+      "epoch": 0.03604216252975179,
+      "grad_norm": 3.1475568662058984,
+      "learning_rate": 9.999123350754722e-06,
+      "loss": 0.8076,
+      "step": 106
+    },
+    {
+      "epoch": 0.03638218293097586,
+      "grad_norm": 2.087445939358465,
+      "learning_rate": 9.999017185639266e-06,
+      "loss": 0.9328,
+      "step": 107
+    },
+    {
+      "epoch": 0.03672220333219993,
+      "grad_norm": 5.085608104666051,
+      "learning_rate": 9.99890495476289e-06,
+      "loss": 0.8641,
+      "step": 108
+    },
+    {
+      "epoch": 0.03706222373342401,
+      "grad_norm": 3.1741923963644827,
+      "learning_rate": 9.99878665826177e-06,
+      "loss": 0.8019,
+      "step": 109
+    },
+    {
+      "epoch": 0.03740224413464808,
+      "grad_norm": 2.355452993498913,
+      "learning_rate": 9.998662296279447e-06,
+      "loss": 0.8653,
+      "step": 110
+    },
+    {
+      "epoch": 0.03774226453587215,
+      "grad_norm": 2.413762001760075,
+      "learning_rate": 9.998531868966822e-06,
+      "loss": 0.7411,
+      "step": 111
+    },
+    {
+      "epoch": 0.03808228493709623,
+      "grad_norm": 2.0339220046950524,
+      "learning_rate": 9.998395376482152e-06,
+      "loss": 0.8552,
+      "step": 112
+    },
+    {
+      "epoch": 0.0384223053383203,
+      "grad_norm": 2.079277005071041,
+      "learning_rate": 9.998252818991062e-06,
+      "loss": 0.8222,
+      "step": 113
+    },
+    {
+      "epoch": 0.03876232573954437,
+      "grad_norm": 5.632785104475984,
+      "learning_rate": 9.99810419666652e-06,
+      "loss": 0.8776,
+      "step": 114
+    },
+    {
+      "epoch": 0.03910234614076845,
+      "grad_norm": 1.8509337002126818,
+      "learning_rate": 9.997949509688871e-06,
+      "loss": 0.8431,
+      "step": 115
+    },
+    {
+      "epoch": 0.03944236654199252,
+      "grad_norm": 2.573221870698461,
+      "learning_rate": 9.997788758245808e-06,
+      "loss": 0.9841,
+      "step": 116
+    },
+    {
+      "epoch": 0.03978238694321659,
+      "grad_norm": 2.572525731553236,
+      "learning_rate": 9.997621942532383e-06,
+      "loss": 0.8367,
+      "step": 117
+    },
+    {
+      "epoch": 0.04012240734444067,
+      "grad_norm": 2.2742345681308938,
+      "learning_rate": 9.997449062751012e-06,
+      "loss": 0.7897,
+      "step": 118
+    },
+    {
+      "epoch": 0.04046242774566474,
+      "grad_norm": 4.327044372060793,
+      "learning_rate": 9.997270119111467e-06,
+      "loss": 0.8457,
+      "step": 119
+    },
+    {
+      "epoch": 0.04080244814688881,
+      "grad_norm": 2.2772066238104376,
+      "learning_rate": 9.99708511183087e-06,
+      "loss": 0.8247,
+      "step": 120
+    },
+    {
+      "epoch": 0.04114246854811289,
+      "grad_norm": 2.601789407750634,
+      "learning_rate": 9.996894041133715e-06,
+      "loss": 0.825,
+      "step": 121
+    },
+    {
+      "epoch": 0.04148248894933696,
+      "grad_norm": 2.0585977344964284,
+      "learning_rate": 9.99669690725184e-06,
+      "loss": 0.7796,
+      "step": 122
+    },
+    {
+      "epoch": 0.04182250935056103,
+      "grad_norm": 1.8855361410264628,
+      "learning_rate": 9.996493710424447e-06,
+      "loss": 0.8304,
+      "step": 123
+    },
+    {
+      "epoch": 0.04216252975178511,
+      "grad_norm": 1.837257670311997,
+      "learning_rate": 9.996284450898093e-06,
+      "loss": 0.9445,
+      "step": 124
+    },
+    {
+      "epoch": 0.04250255015300918,
+      "grad_norm": 3.0656012239893964,
+      "learning_rate": 9.996069128926691e-06,
+      "loss": 0.8702,
+      "step": 125
+    },
+    {
+      "epoch": 0.04284257055423325,
+      "grad_norm": 2.438953538165285,
+      "learning_rate": 9.995847744771514e-06,
+      "loss": 0.7872,
+      "step": 126
+    },
+    {
+      "epoch": 0.043182590955457324,
+      "grad_norm": 2.279269747200056,
+      "learning_rate": 9.995620298701183e-06,
+      "loss": 0.8613,
+      "step": 127
+    },
+    {
+      "epoch": 0.0435226113566814,
+      "grad_norm": 3.213489993275219,
+      "learning_rate": 9.99538679099168e-06,
+      "loss": 0.7127,
+      "step": 128
+    },
+    {
+      "epoch": 0.043862631757905474,
+      "grad_norm": 2.5232149812128637,
+      "learning_rate": 9.995147221926343e-06,
+      "loss": 0.8698,
+      "step": 129
+    },
+    {
+      "epoch": 0.044202652159129545,
+      "grad_norm": 2.839717095850297,
+      "learning_rate": 9.994901591795863e-06,
+      "loss": 0.85,
+      "step": 130
+    },
+    {
+      "epoch": 0.04454267256035362,
+      "grad_norm": 7.648211408560256,
+      "learning_rate": 9.994649900898283e-06,
+      "loss": 0.9204,
+      "step": 131
+    },
+    {
+      "epoch": 0.044882692961577694,
+      "grad_norm": 1.8745457296274581,
+      "learning_rate": 9.994392149539003e-06,
+      "loss": 0.8267,
+      "step": 132
+    },
+    {
+      "epoch": 0.045222713362801766,
+      "grad_norm": 1.9836289186862535,
+      "learning_rate": 9.994128338030778e-06,
+      "loss": 0.9781,
+      "step": 133
+    },
+    {
+      "epoch": 0.045562733764025844,
+      "grad_norm": 2.126401684741174,
+      "learning_rate": 9.993858466693712e-06,
+      "loss": 1.0093,
+      "step": 134
+    },
+    {
+      "epoch": 0.045902754165249915,
+      "grad_norm": 1.983933252004279,
+      "learning_rate": 9.993582535855265e-06,
+      "loss": 0.8227,
+      "step": 135
+    },
+    {
+      "epoch": 0.046242774566473986,
+      "grad_norm": 2.088706279304247,
+      "learning_rate": 9.99330054585025e-06,
+      "loss": 0.8529,
+      "step": 136
+    },
+    {
+      "epoch": 0.046582794967698064,
+      "grad_norm": 2.4156135313351363,
+      "learning_rate": 9.993012497020831e-06,
+      "loss": 0.8666,
+      "step": 137
+    },
+    {
+      "epoch": 0.046922815368922136,
+      "grad_norm": 2.4388488859277584,
+      "learning_rate": 9.992718389716521e-06,
+      "loss": 0.7969,
+      "step": 138
+    },
+    {
+      "epoch": 0.04726283577014621,
+      "grad_norm": 2.169555997324188,
+      "learning_rate": 9.992418224294191e-06,
+      "loss": 0.893,
+      "step": 139
+    },
+    {
+      "epoch": 0.047602856171370285,
+      "grad_norm": 1.9240543034696431,
+      "learning_rate": 9.992112001118058e-06,
+      "loss": 0.9161,
+      "step": 140
+    },
+    {
+      "epoch": 0.047942876572594356,
+      "grad_norm": 3.4020949955907636,
+      "learning_rate": 9.991799720559687e-06,
+      "loss": 0.7416,
+      "step": 141
+    },
+    {
+      "epoch": 0.04828289697381843,
+      "grad_norm": 2.4604945354615153,
+      "learning_rate": 9.991481382998001e-06,
+      "loss": 0.9075,
+      "step": 142
+    },
+    {
+      "epoch": 0.048622917375042506,
+      "grad_norm": 2.255946457618047,
+      "learning_rate": 9.991156988819264e-06,
+      "loss": 0.9905,
+      "step": 143
+    },
+    {
+      "epoch": 0.04896293777626658,
+      "grad_norm": 2.024969438240585,
+      "learning_rate": 9.990826538417095e-06,
+      "loss": 0.9332,
+      "step": 144
+    },
+    {
+      "epoch": 0.04930295817749065,
+      "grad_norm": 2.2370360110433025,
+      "learning_rate": 9.99049003219246e-06,
+      "loss": 0.8148,
+      "step": 145
+    },
+    {
+      "epoch": 0.049642978578714726,
+      "grad_norm": 2.2743555041984407,
+      "learning_rate": 9.99014747055367e-06,
+      "loss": 0.8756,
+      "step": 146
+    },
+    {
+      "epoch": 0.0499829989799388,
+      "grad_norm": 3.95316960278758,
+      "learning_rate": 9.989798853916388e-06,
+      "loss": 0.8292,
+      "step": 147
+    },
+    {
+      "epoch": 0.05032301938116287,
+      "grad_norm": 3.6392856446793833,
+      "learning_rate": 9.989444182703623e-06,
+      "loss": 0.7666,
+      "step": 148
+    },
+    {
+      "epoch": 0.05066303978238694,
+      "grad_norm": 2.230929282152931,
+      "learning_rate": 9.989083457345727e-06,
+      "loss": 0.8501,
+      "step": 149
+    },
+    {
+      "epoch": 0.05100306018361102,
+      "grad_norm": 3.0061135776554826,
+      "learning_rate": 9.988716678280403e-06,
+      "loss": 0.9064,
+      "step": 150
+    },
+    {
+      "epoch": 0.05134308058483509,
+      "grad_norm": 2.1445457506712824,
+      "learning_rate": 9.988343845952697e-06,
+      "loss": 0.8971,
+      "step": 151
+    },
+    {
+      "epoch": 0.05168310098605916,
+      "grad_norm": 2.437139941905991,
+      "learning_rate": 9.987964960815e-06,
+      "loss": 0.9558,
+      "step": 152
+    },
+    {
+      "epoch": 0.05202312138728324,
+      "grad_norm": 4.873429755031727,
+      "learning_rate": 9.987580023327046e-06,
+      "loss": 0.8671,
+      "step": 153
+    },
+    {
+      "epoch": 0.05236314178850731,
+      "grad_norm": 3.944162605474988,
+      "learning_rate": 9.987189033955918e-06,
+      "loss": 0.8272,
+      "step": 154
+    },
+    {
+      "epoch": 0.05270316218973138,
+      "grad_norm": 2.8250974202447985,
+      "learning_rate": 9.986791993176035e-06,
+      "loss": 0.8342,
+      "step": 155
+    },
+    {
+      "epoch": 0.05304318259095546,
+      "grad_norm": 2.3354440033626345,
+      "learning_rate": 9.986388901469167e-06,
+      "loss": 0.7322,
+      "step": 156
+    },
+    {
+      "epoch": 0.05338320299217953,
+      "grad_norm": 3.5022852363593198,
+      "learning_rate": 9.985979759324418e-06,
+      "loss": 0.9182,
+      "step": 157
+    },
+    {
+      "epoch": 0.0537232233934036,
+      "grad_norm": 2.022608573939002,
+      "learning_rate": 9.985564567238237e-06,
+      "loss": 0.8555,
+      "step": 158
+    },
+    {
+      "epoch": 0.05406324379462768,
+      "grad_norm": 2.4546285068255744,
+      "learning_rate": 9.985143325714419e-06,
+      "loss": 0.8571,
+      "step": 159
+    },
+    {
+      "epoch": 0.05440326419585175,
+      "grad_norm": 1.7140624575103907,
+      "learning_rate": 9.984716035264089e-06,
+      "loss": 0.7834,
+      "step": 160
+    },
+    {
+      "epoch": 0.05474328459707582,
+      "grad_norm": 2.007966583007289,
+      "learning_rate": 9.98428269640572e-06,
+      "loss": 0.8762,
+      "step": 161
+    },
+    {
+      "epoch": 0.0550833049982999,
+      "grad_norm": 2.543305674356758,
+      "learning_rate": 9.983843309665122e-06,
+      "loss": 0.8685,
+      "step": 162
+    },
+    {
+      "epoch": 0.05542332539952397,
+      "grad_norm": 2.080393891117657,
+      "learning_rate": 9.983397875575442e-06,
+      "loss": 0.8911,
+      "step": 163
+    },
+    {
+      "epoch": 0.05576334580074804,
+      "grad_norm": 2.294684609497304,
+      "learning_rate": 9.982946394677165e-06,
+      "loss": 0.9734,
+      "step": 164
+    },
+    {
+      "epoch": 0.05610336620197212,
+      "grad_norm": 1.804144176865855,
+      "learning_rate": 9.982488867518112e-06,
+      "loss": 0.819,
+      "step": 165
+    },
+    {
+      "epoch": 0.05644338660319619,
+      "grad_norm": 3.657476774208594,
+      "learning_rate": 9.982025294653445e-06,
+      "loss": 0.8307,
+      "step": 166
+    },
+    {
+      "epoch": 0.056783407004420264,
+      "grad_norm": 2.074991736382199,
+      "learning_rate": 9.98155567664566e-06,
+      "loss": 0.8015,
+      "step": 167
+    },
+    {
+      "epoch": 0.05712342740564434,
+      "grad_norm": 2.9051744319346255,
+      "learning_rate": 9.981080014064584e-06,
+      "loss": 0.7693,
+      "step": 168
+    },
+    {
+      "epoch": 0.05746344780686841,
+      "grad_norm": 1.9099484180793749,
+      "learning_rate": 9.980598307487383e-06,
+      "loss": 0.7955,
+      "step": 169
+    },
+    {
+      "epoch": 0.057803468208092484,
+      "grad_norm": 2.146831708028932,
+      "learning_rate": 9.980110557498556e-06,
+      "loss": 0.8993,
+      "step": 170
+    },
+    {
+      "epoch": 0.058143488609316556,
+      "grad_norm": 2.61617139387018,
+      "learning_rate": 9.979616764689932e-06,
+      "loss": 0.8702,
+      "step": 171
+    },
+    {
+      "epoch": 0.058483509010540634,
+      "grad_norm": 2.1196184076767057,
+      "learning_rate": 9.979116929660677e-06,
+      "loss": 0.9441,
+      "step": 172
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 3.7671698090532866,
+      "learning_rate": 9.978611053017286e-06,
+      "loss": 0.9364,
+      "step": 173
+    },
+    {
+      "epoch": 0.059163549812988776,
+      "grad_norm": 2.6799358731624268,
+      "learning_rate": 9.978099135373584e-06,
+      "loss": 0.8686,
+      "step": 174
+    },
+    {
+      "epoch": 0.059503570214212854,
+      "grad_norm": 1.8432414885158408,
+      "learning_rate": 9.977581177350726e-06,
+      "loss": 0.8512,
+      "step": 175
+    },
+    {
+      "epoch": 0.059843590615436926,
+      "grad_norm": 2.6747034963076963,
+      "learning_rate": 9.977057179577199e-06,
+      "loss": 0.7921,
+      "step": 176
+    },
+    {
+      "epoch": 0.060183611016661,
+      "grad_norm": 3.038911682651883,
+      "learning_rate": 9.976527142688818e-06,
+      "loss": 0.849,
+      "step": 177
+    },
+    {
+      "epoch": 0.060523631417885075,
+      "grad_norm": 3.497473396206265,
+      "learning_rate": 9.975991067328722e-06,
+      "loss": 0.8514,
+      "step": 178
+    },
+    {
+      "epoch": 0.060863651819109146,
+      "grad_norm": 2.872664876776745,
+      "learning_rate": 9.975448954147383e-06,
+      "loss": 0.76,
+      "step": 179
+    },
+    {
+      "epoch": 0.06120367222033322,
+      "grad_norm": 2.265420234730481,
+      "learning_rate": 9.974900803802595e-06,
+      "loss": 0.8893,
+      "step": 180
+    },
+    {
+      "epoch": 0.061543692621557296,
+      "grad_norm": 2.2490770407309273,
+      "learning_rate": 9.974346616959476e-06,
+      "loss": 0.9017,
+      "step": 181
+    },
+    {
+      "epoch": 0.06188371302278137,
+      "grad_norm": 2.279036329894871,
+      "learning_rate": 9.973786394290475e-06,
+      "loss": 0.7729,
+      "step": 182
+    },
+    {
+      "epoch": 0.06222373342400544,
+      "grad_norm": 2.2273111686177693,
+      "learning_rate": 9.973220136475359e-06,
+      "loss": 0.9491,
+      "step": 183
+    },
+    {
+      "epoch": 0.06256375382522951,
+      "grad_norm": 1.8442525974273993,
+      "learning_rate": 9.97264784420122e-06,
+      "loss": 0.8584,
+      "step": 184
+    },
+    {
+      "epoch": 0.06290377422645359,
+      "grad_norm": 2.0159433278766876,
+      "learning_rate": 9.972069518162472e-06,
+      "loss": 0.855,
+      "step": 185
+    },
+    {
+      "epoch": 0.06324379462767767,
+      "grad_norm": 2.2639684727572336,
+      "learning_rate": 9.971485159060851e-06,
+      "loss": 0.9352,
+      "step": 186
+    },
+    {
+      "epoch": 0.06358381502890173,
+      "grad_norm": 2.5527740776494574,
+      "learning_rate": 9.970894767605412e-06,
+      "loss": 0.7912,
+      "step": 187
+    },
+    {
+      "epoch": 0.06392383543012581,
+      "grad_norm": 1.9379101706278465,
+      "learning_rate": 9.970298344512533e-06,
+      "loss": 0.8189,
+      "step": 188
+    },
+    {
+      "epoch": 0.06426385583134989,
+      "grad_norm": 3.1376756923399824,
+      "learning_rate": 9.969695890505904e-06,
+      "loss": 0.9007,
+      "step": 189
+    },
+    {
+      "epoch": 0.06460387623257395,
+      "grad_norm": 2.270221919510738,
+      "learning_rate": 9.96908740631654e-06,
+      "loss": 0.8274,
+      "step": 190
+    },
+    {
+      "epoch": 0.06494389663379803,
+      "grad_norm": 10.152532884466371,
+      "learning_rate": 9.96847289268277e-06,
+      "loss": 0.9628,
+      "step": 191
+    },
+    {
+      "epoch": 0.06528391703502211,
+      "grad_norm": 2.2780092143082395,
+      "learning_rate": 9.967852350350239e-06,
+      "loss": 0.8497,
+      "step": 192
+    },
+    {
+      "epoch": 0.06562393743624617,
+      "grad_norm": 2.212416833604225,
+      "learning_rate": 9.967225780071908e-06,
+      "loss": 0.8529,
+      "step": 193
+    },
+    {
+      "epoch": 0.06596395783747025,
+      "grad_norm": 2.5787177735483833,
+      "learning_rate": 9.966593182608048e-06,
+      "loss": 0.8931,
+      "step": 194
+    },
+    {
+      "epoch": 0.06630397823869433,
+      "grad_norm": 2.222235355514418,
+      "learning_rate": 9.965954558726249e-06,
+      "loss": 0.8334,
+      "step": 195
+    },
+    {
+      "epoch": 0.06664399863991839,
+      "grad_norm": 3.0045508316166742,
+      "learning_rate": 9.965309909201414e-06,
+      "loss": 0.8262,
+      "step": 196
+    },
+    {
+      "epoch": 0.06698401904114247,
+      "grad_norm": 2.071849506932709,
+      "learning_rate": 9.964659234815752e-06,
+      "loss": 0.9124,
+      "step": 197
+    },
+    {
+      "epoch": 0.06732403944236655,
+      "grad_norm": 2.67102771314228,
+      "learning_rate": 9.964002536358784e-06,
+      "loss": 0.8469,
+      "step": 198
+    },
+    {
+      "epoch": 0.06766405984359061,
+      "grad_norm": 3.45508048201135,
+      "learning_rate": 9.963339814627344e-06,
+      "loss": 0.862,
+      "step": 199
+    },
+    {
+      "epoch": 0.06800408024481469,
+      "grad_norm": 1.95482979584485,
+      "learning_rate": 9.962671070425573e-06,
+      "loss": 0.832,
+      "step": 200
+    },
+    {
+      "epoch": 0.06834410064603877,
+      "grad_norm": 2.6045722790965,
+      "learning_rate": 9.961996304564916e-06,
+      "loss": 0.8735,
+      "step": 201
+    },
+    {
+      "epoch": 0.06868412104726283,
+      "grad_norm": 2.198649716123807,
+      "learning_rate": 9.961315517864131e-06,
+      "loss": 0.8463,
+      "step": 202
+    },
+    {
+      "epoch": 0.06902414144848691,
+      "grad_norm": 3.00819531461859,
+      "learning_rate": 9.960628711149276e-06,
+      "loss": 0.7847,
+      "step": 203
+    },
+    {
+      "epoch": 0.06936416184971098,
+      "grad_norm": 2.7123733460067068,
+      "learning_rate": 9.959935885253715e-06,
+      "loss": 0.8133,
+      "step": 204
+    },
+    {
+      "epoch": 0.06970418225093505,
+      "grad_norm": 1.9582131537045544,
+      "learning_rate": 9.95923704101812e-06,
+      "loss": 0.9498,
+      "step": 205
+    },
+    {
+      "epoch": 0.07004420265215913,
+      "grad_norm": 2.0898364433787577,
+      "learning_rate": 9.958532179290458e-06,
+      "loss": 0.7157,
+      "step": 206
+    },
+    {
+      "epoch": 0.0703842230533832,
+      "grad_norm": 2.0874129936819688,
+      "learning_rate": 9.957821300926007e-06,
+      "loss": 0.8845,
+      "step": 207
+    },
+    {
+      "epoch": 0.07072424345460727,
+      "grad_norm": 2.565402516349025,
+      "learning_rate": 9.957104406787335e-06,
+      "loss": 0.8621,
+      "step": 208
+    },
+    {
+      "epoch": 0.07106426385583135,
+      "grad_norm": 3.070517613442607,
+      "learning_rate": 9.956381497744317e-06,
+      "loss": 0.8068,
+      "step": 209
+    },
+    {
+      "epoch": 0.07140428425705542,
+      "grad_norm": 1.7745161730599397,
+      "learning_rate": 9.955652574674122e-06,
+      "loss": 0.774,
+      "step": 210
+    },
+    {
+      "epoch": 0.0717443046582795,
+      "grad_norm": 2.2664861311135835,
+      "learning_rate": 9.954917638461221e-06,
+      "loss": 0.7881,
+      "step": 211
+    },
+    {
+      "epoch": 0.07208432505950357,
+      "grad_norm": 2.8326430149609103,
+      "learning_rate": 9.954176689997379e-06,
+      "loss": 0.8248,
+      "step": 212
+    },
+    {
+      "epoch": 0.07242434546072764,
+      "grad_norm": 5.345259266036334,
+      "learning_rate": 9.953429730181653e-06,
+      "loss": 0.8863,
+      "step": 213
+    },
+    {
+      "epoch": 0.07276436586195172,
+      "grad_norm": 2.0087383584289404,
+      "learning_rate": 9.952676759920401e-06,
+      "loss": 0.9046,
+      "step": 214
+    },
+    {
+      "epoch": 0.0731043862631758,
+      "grad_norm": 2.7067250680131503,
+      "learning_rate": 9.951917780127268e-06,
+      "loss": 0.7835,
+      "step": 215
+    },
+    {
+      "epoch": 0.07344440666439986,
+      "grad_norm": 2.0141720241519963,
+      "learning_rate": 9.951152791723193e-06,
+      "loss": 0.8934,
+      "step": 216
+    },
+    {
+      "epoch": 0.07378442706562394,
+      "grad_norm": 1.8777488263439555,
+      "learning_rate": 9.950381795636406e-06,
+      "loss": 0.8121,
+      "step": 217
+    },
+    {
+      "epoch": 0.07412444746684801,
+      "grad_norm": 3.255807583902176,
+      "learning_rate": 9.949604792802425e-06,
+      "loss": 0.8676,
+      "step": 218
+    },
+    {
+      "epoch": 0.07446446786807208,
+      "grad_norm": 2.67736933247142,
+      "learning_rate": 9.94882178416406e-06,
+      "loss": 0.7393,
+      "step": 219
+    },
+    {
+      "epoch": 0.07480448826929616,
+      "grad_norm": 4.008487179200106,
+      "learning_rate": 9.948032770671405e-06,
+      "loss": 0.8465,
+      "step": 220
+    },
+    {
+      "epoch": 0.07514450867052024,
+      "grad_norm": 2.940928392932845,
+      "learning_rate": 9.947237753281845e-06,
+      "loss": 0.8339,
+      "step": 221
+    },
+    {
+      "epoch": 0.0754845290717443,
+      "grad_norm": 2.051172980166151,
+      "learning_rate": 9.946436732960042e-06,
+      "loss": 0.9295,
+      "step": 222
+    },
+    {
+      "epoch": 0.07582454947296838,
+      "grad_norm": 2.424428707199302,
+      "learning_rate": 9.945629710677949e-06,
+      "loss": 0.8197,
+      "step": 223
+    },
+    {
+      "epoch": 0.07616456987419246,
+      "grad_norm": 2.0727299530697763,
+      "learning_rate": 9.9448166874148e-06,
+      "loss": 0.8643,
+      "step": 224
+    },
+    {
+      "epoch": 0.07650459027541652,
+      "grad_norm": 2.770127723386738,
+      "learning_rate": 9.943997664157108e-06,
+      "loss": 0.8465,
+      "step": 225
+    },
+    {
+      "epoch": 0.0768446106766406,
+      "grad_norm": 2.555973204555607,
+      "learning_rate": 9.943172641898669e-06,
+      "loss": 0.8517,
+      "step": 226
+    },
+    {
+      "epoch": 0.07718463107786468,
+      "grad_norm": 1.842390623944307,
+      "learning_rate": 9.942341621640558e-06,
+      "loss": 0.85,
+      "step": 227
+    },
+    {
+      "epoch": 0.07752465147908874,
+      "grad_norm": 2.128091117333538,
+      "learning_rate": 9.941504604391126e-06,
+      "loss": 0.8292,
+      "step": 228
+    },
+    {
+      "epoch": 0.07786467188031282,
+      "grad_norm": 1.7281395666728367,
+      "learning_rate": 9.940661591166003e-06,
+      "loss": 0.8231,
+      "step": 229
+    },
+    {
+      "epoch": 0.0782046922815369,
+      "grad_norm": 2.9385281630654285,
+      "learning_rate": 9.939812582988094e-06,
+      "loss": 0.7502,
+      "step": 230
+    },
+    {
+      "epoch": 0.07854471268276096,
+      "grad_norm": 2.1296804812289603,
+      "learning_rate": 9.938957580887575e-06,
+      "loss": 0.8717,
+      "step": 231
+    },
+    {
+      "epoch": 0.07888473308398504,
+      "grad_norm": 2.14717718646388,
+      "learning_rate": 9.9380965859019e-06,
+      "loss": 0.8894,
+      "step": 232
+    },
+    {
+      "epoch": 0.07922475348520912,
+      "grad_norm": 1.9538584630326372,
+      "learning_rate": 9.937229599075791e-06,
+      "loss": 0.8824,
+      "step": 233
+    },
+    {
+      "epoch": 0.07956477388643318,
+      "grad_norm": 1.7865365933867121,
+      "learning_rate": 9.936356621461243e-06,
+      "loss": 0.8454,
+      "step": 234
+    },
+    {
+      "epoch": 0.07990479428765726,
+      "grad_norm": 1.7416149475546014,
+      "learning_rate": 9.935477654117518e-06,
+      "loss": 0.8576,
+      "step": 235
+    },
+    {
+      "epoch": 0.08024481468888134,
+      "grad_norm": 1.768841666763347,
+      "learning_rate": 9.934592698111148e-06,
+      "loss": 0.9265,
+      "step": 236
+    },
+    {
+      "epoch": 0.0805848350901054,
+      "grad_norm": 2.4152772171644656,
+      "learning_rate": 9.933701754515928e-06,
+      "loss": 0.8519,
+      "step": 237
+    },
+    {
+      "epoch": 0.08092485549132948,
+      "grad_norm": 2.2519504055783814,
+      "learning_rate": 9.932804824412922e-06,
+      "loss": 0.9161,
+      "step": 238
+    },
+    {
+      "epoch": 0.08126487589255356,
+      "grad_norm": 2.450596425363767,
+      "learning_rate": 9.931901908890457e-06,
+      "loss": 0.8091,
+      "step": 239
+    },
+    {
+      "epoch": 0.08160489629377762,
+      "grad_norm": 1.8041419356227175,
+      "learning_rate": 9.930993009044123e-06,
+      "loss": 0.8561,
+      "step": 240
+    },
+    {
+      "epoch": 0.0819449166950017,
+      "grad_norm": 1.7358105710203156,
+      "learning_rate": 9.930078125976767e-06,
+      "loss": 0.9872,
+      "step": 241
+    },
+    {
+      "epoch": 0.08228493709622578,
+      "grad_norm": 1.9725454367197321,
+      "learning_rate": 9.929157260798504e-06,
+      "loss": 0.796,
+      "step": 242
+    },
+    {
+      "epoch": 0.08262495749744984,
+      "grad_norm": 2.0911715350232987,
+      "learning_rate": 9.9282304146267e-06,
+      "loss": 0.7842,
+      "step": 243
+    },
+    {
+      "epoch": 0.08296497789867392,
+      "grad_norm": 1.7083869843606982,
+      "learning_rate": 9.927297588585984e-06,
+      "loss": 0.7561,
+      "step": 244
+    },
+    {
+      "epoch": 0.083304998299898,
+      "grad_norm": 1.8485993258822613,
+      "learning_rate": 9.926358783808238e-06,
+      "loss": 0.7767,
+      "step": 245
+    },
+    {
+      "epoch": 0.08364501870112206,
+      "grad_norm": 2.4843083297610655,
+      "learning_rate": 9.925414001432599e-06,
+      "loss": 0.8209,
+      "step": 246
+    },
+    {
+      "epoch": 0.08398503910234614,
+      "grad_norm": 2.514523920882766,
+      "learning_rate": 9.924463242605454e-06,
+      "loss": 0.75,
+      "step": 247
+    },
+    {
+      "epoch": 0.08432505950357022,
+      "grad_norm": 2.038961464493961,
+      "learning_rate": 9.92350650848045e-06,
+      "loss": 0.8024,
+      "step": 248
+    },
+    {
+      "epoch": 0.08466507990479429,
+      "grad_norm": 5.967821891887257,
+      "learning_rate": 9.922543800218474e-06,
+      "loss": 0.8777,
+      "step": 249
+    },
+    {
+      "epoch": 0.08500510030601836,
+      "grad_norm": 2.1398795095312946,
+      "learning_rate": 9.921575118987672e-06,
+      "loss": 0.8732,
+      "step": 250
+    },
+    {
+      "epoch": 0.08534512070724243,
+      "grad_norm": 2.9133926895356277,
+      "learning_rate": 9.92060046596343e-06,
+      "loss": 0.7944,
+      "step": 251
+    },
+    {
+      "epoch": 0.0856851411084665,
+      "grad_norm": 1.9889465507980215,
+      "learning_rate": 9.919619842328383e-06,
+      "loss": 0.8085,
+      "step": 252
+    },
+    {
+      "epoch": 0.08602516150969058,
+      "grad_norm": 3.150534939883254,
+      "learning_rate": 9.918633249272412e-06,
+      "loss": 0.8113,
+      "step": 253
+    },
+    {
+      "epoch": 0.08636518191091465,
+      "grad_norm": 2.2835881883036335,
+      "learning_rate": 9.917640687992638e-06,
+      "loss": 0.787,
+      "step": 254
+    },
+    {
+      "epoch": 0.08670520231213873,
+      "grad_norm": 2.4230018775230904,
+      "learning_rate": 9.916642159693428e-06,
+      "loss": 0.8945,
+      "step": 255
+    },
+    {
+      "epoch": 0.0870452227133628,
+      "grad_norm": 2.2893689337255108,
+      "learning_rate": 9.915637665586386e-06,
+      "loss": 0.8661,
+      "step": 256
+    },
+    {
+      "epoch": 0.08738524311458687,
+      "grad_norm": 2.4221470788760713,
+      "learning_rate": 9.914627206890352e-06,
+      "loss": 0.8282,
+      "step": 257
+    },
+    {
+      "epoch": 0.08772526351581095,
+      "grad_norm": 2.376986510975116,
+      "learning_rate": 9.913610784831415e-06,
+      "loss": 0.8371,
+      "step": 258
+    },
+    {
+      "epoch": 0.08806528391703503,
+      "grad_norm": 2.4413551231465624,
+      "learning_rate": 9.912588400642884e-06,
+      "loss": 0.8826,
+      "step": 259
+    },
+    {
+      "epoch": 0.08840530431825909,
+      "grad_norm": 2.644042881659926,
+      "learning_rate": 9.911560055565316e-06,
+      "loss": 0.7355,
+      "step": 260
+    },
+    {
+      "epoch": 0.08874532471948317,
+      "grad_norm": 1.9367674986615202,
+      "learning_rate": 9.910525750846494e-06,
+      "loss": 0.8337,
+      "step": 261
+    },
+    {
+      "epoch": 0.08908534512070725,
+      "grad_norm": 3.6307557122329657,
+      "learning_rate": 9.909485487741432e-06,
+      "loss": 0.8818,
+      "step": 262
+    },
+    {
+      "epoch": 0.08942536552193131,
+      "grad_norm": 2.1386654842761468,
+      "learning_rate": 9.908439267512378e-06,
+      "loss": 0.8079,
+      "step": 263
+    },
+    {
+      "epoch": 0.08976538592315539,
+      "grad_norm": 2.05052351091725,
+      "learning_rate": 9.907387091428803e-06,
+      "loss": 0.808,
+      "step": 264
+    },
+    {
+      "epoch": 0.09010540632437947,
+      "grad_norm": 2.1166857344624455,
+      "learning_rate": 9.906328960767409e-06,
+      "loss": 0.8604,
+      "step": 265
+    },
+    {
+      "epoch": 0.09044542672560353,
+      "grad_norm": 2.179398767884701,
+      "learning_rate": 9.905264876812123e-06,
+      "loss": 0.7211,
+      "step": 266
+    },
+    {
+      "epoch": 0.09078544712682761,
+      "grad_norm": 2.8274324506005213,
+      "learning_rate": 9.904194840854094e-06,
+      "loss": 0.9274,
+      "step": 267
+    },
+    {
+      "epoch": 0.09112546752805169,
+      "grad_norm": 2.2046170596471315,
+      "learning_rate": 9.903118854191693e-06,
+      "loss": 0.8147,
+      "step": 268
+    },
+    {
+      "epoch": 0.09146548792927575,
+      "grad_norm": 2.7523217330730505,
+      "learning_rate": 9.902036918130514e-06,
+      "loss": 0.8264,
+      "step": 269
+    },
+    {
+      "epoch": 0.09180550833049983,
+      "grad_norm": 2.2273055464770812,
+      "learning_rate": 9.900949033983366e-06,
+      "loss": 0.866,
+      "step": 270
+    },
+    {
+      "epoch": 0.09214552873172391,
+      "grad_norm": 2.2499475354374607,
+      "learning_rate": 9.899855203070278e-06,
+      "loss": 0.818,
+      "step": 271
+    },
+    {
+      "epoch": 0.09248554913294797,
+      "grad_norm": 1.9616287518680806,
+      "learning_rate": 9.898755426718493e-06,
+      "loss": 0.8311,
+      "step": 272
+    },
+    {
+      "epoch": 0.09282556953417205,
+      "grad_norm": 2.5970292191674655,
+      "learning_rate": 9.897649706262474e-06,
+      "loss": 0.8518,
+      "step": 273
+    },
+    {
+      "epoch": 0.09316558993539613,
+      "grad_norm": 1.7660827251520181,
+      "learning_rate": 9.896538043043887e-06,
+      "loss": 0.8273,
+      "step": 274
+    },
+    {
+      "epoch": 0.09350561033662019,
+      "grad_norm": 2.6210906521485824,
+      "learning_rate": 9.895420438411616e-06,
+      "loss": 0.8951,
+      "step": 275
+    },
+    {
+      "epoch": 0.09384563073784427,
+      "grad_norm": 1.8800676289752767,
+      "learning_rate": 9.89429689372175e-06,
+      "loss": 0.8584,
+      "step": 276
+    },
+    {
+      "epoch": 0.09418565113906835,
+      "grad_norm": 3.1773516067582186,
+      "learning_rate": 9.893167410337591e-06,
+      "loss": 0.8535,
+      "step": 277
+    },
+    {
+      "epoch": 0.09452567154029241,
+      "grad_norm": 3.404016837148496,
+      "learning_rate": 9.892031989629642e-06,
+      "loss": 0.8279,
+      "step": 278
+    },
+    {
+      "epoch": 0.09486569194151649,
+      "grad_norm": 2.173356538675513,
+      "learning_rate": 9.890890632975612e-06,
+      "loss": 0.8635,
+      "step": 279
+    },
+    {
+      "epoch": 0.09520571234274057,
+      "grad_norm": 2.261938469413284,
+      "learning_rate": 9.889743341760412e-06,
+      "loss": 0.7996,
+      "step": 280
+    },
+    {
+      "epoch": 0.09554573274396463,
+      "grad_norm": 2.1477511262078233,
+      "learning_rate": 9.888590117376154e-06,
+      "loss": 0.9334,
+      "step": 281
+    },
+    {
+      "epoch": 0.09588575314518871,
+      "grad_norm": 2.458576128018538,
+      "learning_rate": 9.887430961222153e-06,
+      "loss": 0.88,
+      "step": 282
+    },
+    {
+      "epoch": 0.09622577354641279,
+      "grad_norm": 4.04358503504034,
+      "learning_rate": 9.886265874704914e-06,
+      "loss": 0.8699,
+      "step": 283
+    },
+    {
+      "epoch": 0.09656579394763685,
+      "grad_norm": 2.1028398707450293,
+      "learning_rate": 9.885094859238145e-06,
+      "loss": 1.0212,
+      "step": 284
+    },
+    {
+      "epoch": 0.09690581434886093,
+      "grad_norm": 2.208984051301497,
+      "learning_rate": 9.883917916242744e-06,
+      "loss": 0.8778,
+      "step": 285
+    },
+    {
+      "epoch": 0.09724583475008501,
+      "grad_norm": 5.2221607237155245,
+      "learning_rate": 9.882735047146803e-06,
+      "loss": 0.9002,
+      "step": 286
+    },
+    {
+      "epoch": 0.09758585515130908,
+      "grad_norm": 1.976265426760934,
+      "learning_rate": 9.881546253385603e-06,
+      "loss": 0.8457,
+      "step": 287
+    },
+    {
+      "epoch": 0.09792587555253315,
+      "grad_norm": 2.298479530013667,
+      "learning_rate": 9.880351536401617e-06,
+      "loss": 0.8554,
+      "step": 288
+    },
+    {
+      "epoch": 0.09826589595375723,
+      "grad_norm": 2.055015991219169,
+      "learning_rate": 9.879150897644504e-06,
+      "loss": 0.833,
+      "step": 289
+    },
+    {
+      "epoch": 0.0986059163549813,
+      "grad_norm": 2.147764160856317,
+      "learning_rate": 9.877944338571108e-06,
+      "loss": 0.8516,
+      "step": 290
+    },
+    {
+      "epoch": 0.09894593675620537,
+      "grad_norm": 1.9705786404454357,
+      "learning_rate": 9.876731860645454e-06,
+      "loss": 0.8118,
+      "step": 291
+    },
+    {
+      "epoch": 0.09928595715742945,
+      "grad_norm": 2.502835283880914,
+      "learning_rate": 9.875513465338754e-06,
+      "loss": 0.8403,
+      "step": 292
+    },
+    {
+      "epoch": 0.09962597755865352,
+      "grad_norm": 1.9113225330806665,
+      "learning_rate": 9.874289154129396e-06,
+      "loss": 0.8076,
+      "step": 293
+    },
+    {
+      "epoch": 0.0999659979598776,
+      "grad_norm": 2.282725376690771,
+      "learning_rate": 9.873058928502948e-06,
+      "loss": 0.9446,
+      "step": 294
+    },
+    {
+      "epoch": 0.10030601836110166,
+      "grad_norm": 2.2144235263707217,
+      "learning_rate": 9.871822789952155e-06,
+      "loss": 0.8268,
+      "step": 295
+    },
+    {
+      "epoch": 0.10064603876232574,
+      "grad_norm": 2.0635404740469525,
+      "learning_rate": 9.870580739976936e-06,
+      "loss": 0.8726,
+      "step": 296
+    },
+    {
+      "epoch": 0.10098605916354982,
+      "grad_norm": 1.9237725683316835,
+      "learning_rate": 9.869332780084383e-06,
+      "loss": 0.8556,
+      "step": 297
+    },
+    {
+      "epoch": 0.10132607956477388,
+      "grad_norm": 2.2476614190445514,
+      "learning_rate": 9.868078911788756e-06,
+      "loss": 0.9219,
+      "step": 298
+    },
+    {
+      "epoch": 0.10166609996599796,
+      "grad_norm": 2.5057546873897882,
+      "learning_rate": 9.866819136611492e-06,
+      "loss": 0.767,
+      "step": 299
+    },
+    {
+      "epoch": 0.10200612036722204,
+      "grad_norm": 2.355080917010462,
+      "learning_rate": 9.865553456081188e-06,
+      "loss": 0.7392,
+      "step": 300
+    },
+    {
+      "epoch": 0.1023461407684461,
+      "grad_norm": 2.1056856473326273,
+      "learning_rate": 9.864281871733608e-06,
+      "loss": 0.9198,
+      "step": 301
+    },
+    {
+      "epoch": 0.10268616116967018,
+      "grad_norm": 3.9555249729605917,
+      "learning_rate": 9.863004385111683e-06,
+      "loss": 0.7841,
+      "step": 302
+    },
+    {
+      "epoch": 0.10302618157089426,
+      "grad_norm": 2.2790309051441144,
+      "learning_rate": 9.8617209977655e-06,
+      "loss": 0.785,
+      "step": 303
+    },
+    {
+      "epoch": 0.10336620197211832,
+      "grad_norm": 2.117633945883531,
+      "learning_rate": 9.860431711252312e-06,
+      "loss": 0.8726,
+      "step": 304
+    },
+    {
+      "epoch": 0.1037062223733424,
+      "grad_norm": 2.5389939700471347,
+      "learning_rate": 9.859136527136525e-06,
+      "loss": 0.8982,
+      "step": 305
+    },
+    {
+      "epoch": 0.10404624277456648,
+      "grad_norm": 2.022741013697183,
+      "learning_rate": 9.857835446989708e-06,
+      "loss": 0.7907,
+      "step": 306
+    },
+    {
+      "epoch": 0.10438626317579054,
+      "grad_norm": 3.2009092222933715,
+      "learning_rate": 9.856528472390576e-06,
+      "loss": 0.6933,
+      "step": 307
+    },
+    {
+      "epoch": 0.10472628357701462,
+      "grad_norm": 2.2334333165472984,
+      "learning_rate": 9.855215604925e-06,
+      "loss": 0.8256,
+      "step": 308
+    },
+    {
+      "epoch": 0.1050663039782387,
+      "grad_norm": 2.7821304676829834,
+      "learning_rate": 9.853896846186e-06,
+      "loss": 0.8078,
+      "step": 309
+    },
+    {
+      "epoch": 0.10540632437946276,
+      "grad_norm": 1.960403992508658,
+      "learning_rate": 9.852572197773746e-06,
+      "loss": 0.7848,
+      "step": 310
+    },
+    {
+      "epoch": 0.10574634478068684,
+      "grad_norm": 2.0477907520683476,
+      "learning_rate": 9.851241661295558e-06,
+      "loss": 0.7813,
+      "step": 311
+    },
+    {
+      "epoch": 0.10608636518191092,
+      "grad_norm": 2.0911122056100164,
+      "learning_rate": 9.84990523836589e-06,
+      "loss": 0.8461,
+      "step": 312
+    },
+    {
+      "epoch": 0.10642638558313498,
+      "grad_norm": 2.0728170947217492,
+      "learning_rate": 9.848562930606353e-06,
+      "loss": 0.8832,
+      "step": 313
+    },
+    {
+      "epoch": 0.10676640598435906,
+      "grad_norm": 2.3673123838424,
+      "learning_rate": 9.847214739645684e-06,
+      "loss": 0.8177,
+      "step": 314
+    },
+    {
+      "epoch": 0.10710642638558314,
+      "grad_norm": 6.272345076835982,
+      "learning_rate": 9.845860667119769e-06,
+      "loss": 0.8795,
+      "step": 315
+    },
+    {
+      "epoch": 0.1074464467868072,
+      "grad_norm": 2.1579047581155466,
+      "learning_rate": 9.844500714671625e-06,
+      "loss": 0.7415,
+      "step": 316
+    },
+    {
+      "epoch": 0.10778646718803128,
+      "grad_norm": 1.9340927610624175,
+      "learning_rate": 9.843134883951405e-06,
+      "loss": 0.7208,
+      "step": 317
+    },
+    {
+      "epoch": 0.10812648758925536,
+      "grad_norm": 2.582635490684408,
+      "learning_rate": 9.8417631766164e-06,
+      "loss": 0.788,
+      "step": 318
+    },
+    {
+      "epoch": 0.10846650799047942,
+      "grad_norm": 2.5034059769021617,
+      "learning_rate": 9.840385594331022e-06,
+      "loss": 0.8107,
+      "step": 319
+    },
+    {
+      "epoch": 0.1088065283917035,
+      "grad_norm": 1.8885099464740926,
+      "learning_rate": 9.839002138766818e-06,
+      "loss": 0.8251,
+      "step": 320
+    },
+    {
+      "epoch": 0.10914654879292758,
+      "grad_norm": 2.1105158457858915,
+      "learning_rate": 9.837612811602462e-06,
+      "loss": 0.8193,
+      "step": 321
+    },
+    {
+      "epoch": 0.10948656919415165,
+      "grad_norm": 2.35579939491151,
+      "learning_rate": 9.836217614523747e-06,
+      "loss": 0.8366,
+      "step": 322
+    },
+    {
+      "epoch": 0.10982658959537572,
+      "grad_norm": 2.0838710319308174,
+      "learning_rate": 9.834816549223595e-06,
+      "loss": 0.8519,
+      "step": 323
+    },
+    {
+      "epoch": 0.1101666099965998,
+      "grad_norm": 1.9779789707319297,
+      "learning_rate": 9.833409617402044e-06,
+      "loss": 0.8505,
+      "step": 324
+    },
+    {
+      "epoch": 0.11050663039782387,
+      "grad_norm": 2.3766207833663784,
+      "learning_rate": 9.831996820766255e-06,
+      "loss": 0.8322,
+      "step": 325
+    },
+    {
+      "epoch": 0.11084665079904794,
+      "grad_norm": 2.1149686177574343,
+      "learning_rate": 9.830578161030498e-06,
+      "loss": 0.7337,
+      "step": 326
+    },
+    {
+      "epoch": 0.11118667120027202,
+      "grad_norm": 1.805489375141993,
+      "learning_rate": 9.829153639916162e-06,
+      "loss": 0.8365,
+      "step": 327
+    },
+    {
+      "epoch": 0.11152669160149609,
+      "grad_norm": 1.8967483785377255,
+      "learning_rate": 9.827723259151752e-06,
+      "loss": 0.8414,
+      "step": 328
+    },
+    {
+      "epoch": 0.11186671200272016,
+      "grad_norm": 1.797366163019355,
+      "learning_rate": 9.826287020472873e-06,
+      "loss": 0.7744,
+      "step": 329
+    },
+    {
+      "epoch": 0.11220673240394424,
+      "grad_norm": 2.0302882503199373,
+      "learning_rate": 9.82484492562225e-06,
+      "loss": 0.7147,
+      "step": 330
+    },
+    {
+      "epoch": 0.1125467528051683,
+      "grad_norm": 2.5063229156854723,
+      "learning_rate": 9.823396976349702e-06,
+      "loss": 0.9023,
+      "step": 331
+    },
+    {
+      "epoch": 0.11288677320639239,
+      "grad_norm": 1.8249477324783674,
+      "learning_rate": 9.821943174412159e-06,
+      "loss": 0.8074,
+      "step": 332
+    },
+    {
+      "epoch": 0.11322679360761646,
+      "grad_norm": 1.577389057439917,
+      "learning_rate": 9.82048352157365e-06,
+      "loss": 0.8605,
+      "step": 333
+    },
+    {
+      "epoch": 0.11356681400884053,
+      "grad_norm": 2.760876889647242,
+      "learning_rate": 9.819018019605306e-06,
+      "loss": 0.8667,
+      "step": 334
+    },
+    {
+      "epoch": 0.1139068344100646,
+      "grad_norm": 1.8963064743986586,
+      "learning_rate": 9.817546670285353e-06,
+      "loss": 0.7706,
+      "step": 335
+    },
+    {
+      "epoch": 0.11424685481128868,
+      "grad_norm": 2.2030659264137773,
+      "learning_rate": 9.816069475399113e-06,
+      "loss": 0.8123,
+      "step": 336
+    },
+    {
+      "epoch": 0.11458687521251275,
+      "grad_norm": 2.0354951728501685,
+      "learning_rate": 9.814586436738998e-06,
+      "loss": 0.8086,
+      "step": 337
+    },
+    {
+      "epoch": 0.11492689561373683,
+      "grad_norm": 1.9773497047211561,
+      "learning_rate": 9.813097556104514e-06,
+      "loss": 0.7746,
+      "step": 338
+    },
+    {
+      "epoch": 0.1152669160149609,
+      "grad_norm": 2.0274305580007628,
+      "learning_rate": 9.811602835302257e-06,
+      "loss": 0.8596,
+      "step": 339
+    },
+    {
+      "epoch": 0.11560693641618497,
+      "grad_norm": 1.8336317454746485,
+      "learning_rate": 9.810102276145907e-06,
+      "loss": 0.9853,
+      "step": 340
+    },
+    {
+      "epoch": 0.11594695681740905,
+      "grad_norm": 2.39198257850906,
+      "learning_rate": 9.808595880456226e-06,
+      "loss": 0.856,
+      "step": 341
+    },
+    {
+      "epoch": 0.11628697721863311,
+      "grad_norm": 2.245187215344967,
+      "learning_rate": 9.807083650061063e-06,
+      "loss": 0.8427,
+      "step": 342
+    },
+    {
+      "epoch": 0.11662699761985719,
+      "grad_norm": 2.972553693877441,
+      "learning_rate": 9.805565586795343e-06,
+      "loss": 0.7884,
+      "step": 343
+    },
+    {
+      "epoch": 0.11696701802108127,
+      "grad_norm": 2.1953868509755776,
+      "learning_rate": 9.804041692501071e-06,
+      "loss": 0.8486,
+      "step": 344
+    },
+    {
+      "epoch": 0.11730703842230533,
+      "grad_norm": 1.8166601796451167,
+      "learning_rate": 9.802511969027325e-06,
+      "loss": 0.8615,
+      "step": 345
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 1.9173045797680728,
+      "learning_rate": 9.800976418230257e-06,
+      "loss": 0.8542,
+      "step": 346
+    },
+    {
+      "epoch": 0.11798707922475349,
+      "grad_norm": 1.8134758747889816,
+      "learning_rate": 9.799435041973092e-06,
+      "loss": 0.8473,
+      "step": 347
+    },
+    {
+      "epoch": 0.11832709962597755,
+      "grad_norm": 4.312106023493087,
+      "learning_rate": 9.797887842126119e-06,
+      "loss": 0.8589,
+      "step": 348
+    },
+    {
+      "epoch": 0.11866712002720163,
+      "grad_norm": 2.0162211971459514,
+      "learning_rate": 9.796334820566697e-06,
+      "loss": 0.8679,
+      "step": 349
+    },
+    {
+      "epoch": 0.11900714042842571,
+      "grad_norm": 2.0564343236552873,
+      "learning_rate": 9.79477597917925e-06,
+      "loss": 0.8577,
+      "step": 350
+    },
+    {
+      "epoch": 0.11934716082964977,
+      "grad_norm": 2.133819909568558,
+      "learning_rate": 9.793211319855258e-06,
+      "loss": 0.8628,
+      "step": 351
+    },
+    {
+      "epoch": 0.11968718123087385,
+      "grad_norm": 1.9311795244885308,
+      "learning_rate": 9.791640844493267e-06,
+      "loss": 0.8469,
+      "step": 352
+    },
+    {
+      "epoch": 0.12002720163209793,
+      "grad_norm": 2.967519343303041,
+      "learning_rate": 9.790064554998875e-06,
+      "loss": 0.8362,
+      "step": 353
+    },
+    {
+      "epoch": 0.120367222033322,
+      "grad_norm": 4.208154882236529,
+      "learning_rate": 9.788482453284737e-06,
+      "loss": 0.9199,
+      "step": 354
+    },
+    {
+      "epoch": 0.12070724243454607,
+      "grad_norm": 2.617909166707279,
+      "learning_rate": 9.786894541270563e-06,
+      "loss": 0.7771,
+      "step": 355
+    },
+    {
+      "epoch": 0.12104726283577015,
+      "grad_norm": 1.8137174275503192,
+      "learning_rate": 9.785300820883108e-06,
+      "loss": 0.7556,
+      "step": 356
+    },
+    {
+      "epoch": 0.12138728323699421,
+      "grad_norm": 1.9432369984112166,
+      "learning_rate": 9.78370129405618e-06,
+      "loss": 0.8698,
+      "step": 357
+    },
+    {
+      "epoch": 0.12172730363821829,
+      "grad_norm": 2.5064308483815316,
+      "learning_rate": 9.782095962730628e-06,
+      "loss": 0.8607,
+      "step": 358
+    },
+    {
+      "epoch": 0.12206732403944237,
+      "grad_norm": 5.205967310399993,
+      "learning_rate": 9.780484828854346e-06,
+      "loss": 0.8711,
+      "step": 359
+    },
+    {
+      "epoch": 0.12240734444066644,
+      "grad_norm": 3.2761307002799462,
+      "learning_rate": 9.77886789438227e-06,
+      "loss": 0.8533,
+      "step": 360
+    },
+    {
+      "epoch": 0.12274736484189051,
+      "grad_norm": 1.8164118637880238,
+      "learning_rate": 9.777245161276372e-06,
+      "loss": 0.8407,
+      "step": 361
+    },
+    {
+      "epoch": 0.12308738524311459,
+      "grad_norm": 1.933148024538711,
+      "learning_rate": 9.775616631505663e-06,
+      "loss": 0.8629,
+      "step": 362
+    },
+    {
+      "epoch": 0.12342740564433866,
+      "grad_norm": 2.0021553446281395,
+      "learning_rate": 9.773982307046185e-06,
+      "loss": 0.7911,
+      "step": 363
+    },
+    {
+      "epoch": 0.12376742604556273,
+      "grad_norm": 2.4770275582280155,
+      "learning_rate": 9.772342189881012e-06,
+      "loss": 0.8485,
+      "step": 364
+    },
+    {
+      "epoch": 0.12410744644678681,
+      "grad_norm": 2.0858568863354474,
+      "learning_rate": 9.770696282000245e-06,
+      "loss": 0.8311,
+      "step": 365
+    },
+    {
+      "epoch": 0.12444746684801088,
+      "grad_norm": 1.8817539415546063,
+      "learning_rate": 9.769044585401017e-06,
+      "loss": 0.8769,
+      "step": 366
+    },
+    {
+      "epoch": 0.12478748724923495,
+      "grad_norm": 1.7596502920409212,
+      "learning_rate": 9.767387102087477e-06,
+      "loss": 0.8521,
+      "step": 367
+    },
+    {
+      "epoch": 0.12512750765045902,
+      "grad_norm": 4.783494590295243,
+      "learning_rate": 9.765723834070805e-06,
+      "loss": 0.8325,
+      "step": 368
+    },
+    {
+      "epoch": 0.1254675280516831,
+      "grad_norm": 4.166385331832633,
+      "learning_rate": 9.764054783369191e-06,
+      "loss": 0.862,
+      "step": 369
+    },
+    {
+      "epoch": 0.12580754845290718,
+      "grad_norm": 2.460165725461386,
+      "learning_rate": 9.762379952007847e-06,
+      "loss": 0.8256,
+      "step": 370
+    },
+    {
+      "epoch": 0.12614756885413125,
+      "grad_norm": 2.066282248436767,
+      "learning_rate": 9.760699342018997e-06,
+      "loss": 0.8975,
+      "step": 371
+    },
+    {
+      "epoch": 0.12648758925535533,
+      "grad_norm": 2.4732231692866793,
+      "learning_rate": 9.759012955441877e-06,
+      "loss": 0.8474,
+      "step": 372
+    },
+    {
+      "epoch": 0.12682760965657938,
+      "grad_norm": 2.1370791962949034,
+      "learning_rate": 9.757320794322736e-06,
+      "loss": 0.8541,
+      "step": 373
+    },
+    {
+      "epoch": 0.12716763005780346,
+      "grad_norm": 2.545444534776287,
+      "learning_rate": 9.755622860714824e-06,
+      "loss": 0.7436,
+      "step": 374
+    },
+    {
+      "epoch": 0.12750765045902754,
+      "grad_norm": 4.486941686450132,
+      "learning_rate": 9.753919156678397e-06,
+      "loss": 0.9077,
+      "step": 375
+    },
+    {
+      "epoch": 0.12784767086025162,
+      "grad_norm": 2.3634783435481004,
+      "learning_rate": 9.752209684280717e-06,
+      "loss": 0.9126,
+      "step": 376
+    },
+    {
+      "epoch": 0.1281876912614757,
+      "grad_norm": 2.155370098640075,
+      "learning_rate": 9.750494445596039e-06,
+      "loss": 0.9266,
+      "step": 377
+    },
+    {
+      "epoch": 0.12852771166269977,
+      "grad_norm": 3.380923933190232,
+      "learning_rate": 9.748773442705617e-06,
+      "loss": 0.859,
+      "step": 378
+    },
+    {
+      "epoch": 0.12886773206392382,
+      "grad_norm": 2.0842951308423805,
+      "learning_rate": 9.747046677697703e-06,
+      "loss": 0.8092,
+      "step": 379
+    },
+    {
+      "epoch": 0.1292077524651479,
+      "grad_norm": 2.2756970171219026,
+      "learning_rate": 9.745314152667532e-06,
+      "loss": 0.8278,
+      "step": 380
+    },
+    {
+      "epoch": 0.12954777286637198,
+      "grad_norm": 2.224154665108581,
+      "learning_rate": 9.743575869717343e-06,
+      "loss": 0.9301,
+      "step": 381
+    },
+    {
+      "epoch": 0.12988779326759606,
+      "grad_norm": 1.8618826446938581,
+      "learning_rate": 9.741831830956344e-06,
+      "loss": 0.7506,
+      "step": 382
+    },
+    {
+      "epoch": 0.13022781366882014,
+      "grad_norm": 2.359471897466931,
+      "learning_rate": 9.740082038500738e-06,
+      "loss": 0.8335,
+      "step": 383
+    },
+    {
+      "epoch": 0.13056783407004421,
+      "grad_norm": 2.2530265023397003,
+      "learning_rate": 9.738326494473708e-06,
+      "loss": 0.6634,
+      "step": 384
+    },
+    {
+      "epoch": 0.13090785447126826,
+      "grad_norm": 3.4483730372310313,
+      "learning_rate": 9.736565201005415e-06,
+      "loss": 0.796,
+      "step": 385
+    },
+    {
+      "epoch": 0.13124787487249234,
+      "grad_norm": 2.0533540536094264,
+      "learning_rate": 9.734798160232994e-06,
+      "loss": 0.7612,
+      "step": 386
+    },
+    {
+      "epoch": 0.13158789527371642,
+      "grad_norm": 2.0190748769698907,
+      "learning_rate": 9.733025374300556e-06,
+      "loss": 0.9188,
+      "step": 387
+    },
+    {
+      "epoch": 0.1319279156749405,
+      "grad_norm": 3.434702747485103,
+      "learning_rate": 9.731246845359187e-06,
+      "loss": 0.7621,
+      "step": 388
+    },
+    {
+      "epoch": 0.13226793607616458,
+      "grad_norm": 2.418994252917101,
+      "learning_rate": 9.729462575566931e-06,
+      "loss": 0.8452,
+      "step": 389
+    },
+    {
+      "epoch": 0.13260795647738866,
+      "grad_norm": 3.0894287674893337,
+      "learning_rate": 9.727672567088809e-06,
+      "loss": 0.9177,
+      "step": 390
+    },
+    {
+      "epoch": 0.1329479768786127,
+      "grad_norm": 2.087903312379049,
+      "learning_rate": 9.725876822096798e-06,
+      "loss": 0.8296,
+      "step": 391
+    },
+    {
+      "epoch": 0.13328799727983678,
+      "grad_norm": 1.9043558027810883,
+      "learning_rate": 9.724075342769841e-06,
+      "loss": 0.889,
+      "step": 392
+    },
+    {
+      "epoch": 0.13362801768106086,
+      "grad_norm": 2.455862323906618,
+      "learning_rate": 9.722268131293835e-06,
+      "loss": 0.7897,
+      "step": 393
+    },
+    {
+      "epoch": 0.13396803808228494,
+      "grad_norm": 2.0215817915535346,
+      "learning_rate": 9.720455189861634e-06,
+      "loss": 0.8734,
+      "step": 394
+    },
+    {
+      "epoch": 0.13430805848350902,
+      "grad_norm": 2.6158996648329347,
+      "learning_rate": 9.718636520673042e-06,
+      "loss": 0.7503,
+      "step": 395
+    },
+    {
+      "epoch": 0.1346480788847331,
+      "grad_norm": 1.9769125457115064,
+      "learning_rate": 9.716812125934818e-06,
+      "loss": 0.8262,
+      "step": 396
+    },
+    {
+      "epoch": 0.13498809928595715,
+      "grad_norm": 2.1779733445105514,
+      "learning_rate": 9.714982007860666e-06,
+      "loss": 0.859,
+      "step": 397
+    },
+    {
+      "epoch": 0.13532811968718123,
+      "grad_norm": 1.831021062504083,
+      "learning_rate": 9.713146168671229e-06,
+      "loss": 0.7766,
+      "step": 398
+    },
+    {
+      "epoch": 0.1356681400884053,
+      "grad_norm": 2.8959475128601224,
+      "learning_rate": 9.711304610594104e-06,
+      "loss": 0.7802,
+      "step": 399
+    },
+    {
+      "epoch": 0.13600816048962938,
+      "grad_norm": 2.1321321783640004,
+      "learning_rate": 9.709457335863815e-06,
+      "loss": 0.7941,
+      "step": 400
+    },
+    {
+      "epoch": 0.13634818089085346,
+      "grad_norm": 1.9938357040475916,
+      "learning_rate": 9.707604346721833e-06,
+      "loss": 0.8127,
+      "step": 401
+    },
+    {
+      "epoch": 0.13668820129207754,
+      "grad_norm": 1.7330981671482506,
+      "learning_rate": 9.705745645416553e-06,
+      "loss": 0.7544,
+      "step": 402
+    },
+    {
+      "epoch": 0.1370282216933016,
+      "grad_norm": 2.584162403732432,
+      "learning_rate": 9.703881234203309e-06,
+      "loss": 0.843,
+      "step": 403
+    },
+    {
+      "epoch": 0.13736824209452567,
+      "grad_norm": 2.3450929065189596,
+      "learning_rate": 9.702011115344359e-06,
+      "loss": 0.8568,
+      "step": 404
+    },
+    {
+      "epoch": 0.13770826249574974,
+      "grad_norm": 1.8906128272198697,
+      "learning_rate": 9.70013529110889e-06,
+      "loss": 0.7954,
+      "step": 405
+    },
+    {
+      "epoch": 0.13804828289697382,
+      "grad_norm": 2.005544234328828,
+      "learning_rate": 9.698253763773005e-06,
+      "loss": 0.8552,
+      "step": 406
+    },
+    {
+      "epoch": 0.1383883032981979,
+      "grad_norm": 2.1791404886973083,
+      "learning_rate": 9.696366535619735e-06,
+      "loss": 0.7682,
+      "step": 407
+    },
+    {
+      "epoch": 0.13872832369942195,
+      "grad_norm": 2.1254255310117216,
+      "learning_rate": 9.694473608939024e-06,
+      "loss": 0.9305,
+      "step": 408
+    },
+    {
+      "epoch": 0.13906834410064603,
+      "grad_norm": 2.065032167374917,
+      "learning_rate": 9.692574986027733e-06,
+      "loss": 0.8186,
+      "step": 409
+    },
+    {
+      "epoch": 0.1394083645018701,
+      "grad_norm": 1.976786103840094,
+      "learning_rate": 9.690670669189632e-06,
+      "loss": 0.9253,
+      "step": 410
+    },
+    {
+      "epoch": 0.13974838490309419,
+      "grad_norm": 9.33711461350008,
+      "learning_rate": 9.688760660735403e-06,
+      "loss": 0.8761,
+      "step": 411
+    },
+    {
+      "epoch": 0.14008840530431826,
+      "grad_norm": 2.0857389613573027,
+      "learning_rate": 9.68684496298263e-06,
+      "loss": 0.8343,
+      "step": 412
+    },
+    {
+      "epoch": 0.14042842570554234,
+      "grad_norm": 2.496185713769945,
+      "learning_rate": 9.684923578255806e-06,
+      "loss": 0.8012,
+      "step": 413
+    },
+    {
+      "epoch": 0.1407684461067664,
+      "grad_norm": 3.227539517727669,
+      "learning_rate": 9.682996508886318e-06,
+      "loss": 0.8353,
+      "step": 414
+    },
+    {
+      "epoch": 0.14110846650799047,
+      "grad_norm": 2.713651534237373,
+      "learning_rate": 9.681063757212455e-06,
+      "loss": 0.7775,
+      "step": 415
+    },
+    {
+      "epoch": 0.14144848690921455,
+      "grad_norm": 1.6905920998598611,
+      "learning_rate": 9.679125325579402e-06,
+      "loss": 0.79,
+      "step": 416
+    },
+    {
+      "epoch": 0.14178850731043863,
+      "grad_norm": 3.2275016714057947,
+      "learning_rate": 9.67718121633923e-06,
+      "loss": 0.8604,
+      "step": 417
+    },
+    {
+      "epoch": 0.1421285277116627,
+      "grad_norm": 2.6920966680993503,
+      "learning_rate": 9.675231431850907e-06,
+      "loss": 0.793,
+      "step": 418
+    },
+    {
+      "epoch": 0.14246854811288678,
+      "grad_norm": 2.6498430598374583,
+      "learning_rate": 9.673275974480282e-06,
+      "loss": 0.9103,
+      "step": 419
+    },
+    {
+      "epoch": 0.14280856851411083,
+      "grad_norm": 2.7892979782736864,
+      "learning_rate": 9.671314846600088e-06,
+      "loss": 0.826,
+      "step": 420
+    },
+    {
+      "epoch": 0.1431485889153349,
+      "grad_norm": 1.7387600466875632,
+      "learning_rate": 9.66934805058994e-06,
+      "loss": 0.7522,
+      "step": 421
+    },
+    {
+      "epoch": 0.143488609316559,
+      "grad_norm": 1.9252648709916258,
+      "learning_rate": 9.667375588836329e-06,
+      "loss": 0.9249,
+      "step": 422
+    },
+    {
+      "epoch": 0.14382862971778307,
+      "grad_norm": 2.0337042623648784,
+      "learning_rate": 9.665397463732623e-06,
+      "loss": 0.832,
+      "step": 423
+    },
+    {
+      "epoch": 0.14416865011900715,
+      "grad_norm": 2.355326076923748,
+      "learning_rate": 9.66341367767906e-06,
+      "loss": 0.7972,
+      "step": 424
+    },
+    {
+      "epoch": 0.14450867052023122,
+      "grad_norm": 4.252381455801173,
+      "learning_rate": 9.661424233082748e-06,
+      "loss": 0.8571,
+      "step": 425
+    },
+    {
+      "epoch": 0.14484869092145528,
+      "grad_norm": 2.2203148724753503,
+      "learning_rate": 9.65942913235766e-06,
+      "loss": 0.8049,
+      "step": 426
+    },
+    {
+      "epoch": 0.14518871132267935,
+      "grad_norm": 2.250620096493986,
+      "learning_rate": 9.657428377924632e-06,
+      "loss": 0.8665,
+      "step": 427
+    },
+    {
+      "epoch": 0.14552873172390343,
+      "grad_norm": 1.7641375389756913,
+      "learning_rate": 9.655421972211362e-06,
+      "loss": 0.8509,
+      "step": 428
+    },
+    {
+      "epoch": 0.1458687521251275,
+      "grad_norm": 1.917741070851843,
+      "learning_rate": 9.653409917652406e-06,
+      "loss": 0.8852,
+      "step": 429
+    },
+    {
+      "epoch": 0.1462087725263516,
+      "grad_norm": 2.1235005797818425,
+      "learning_rate": 9.651392216689167e-06,
+      "loss": 0.939,
+      "step": 430
+    },
+    {
+      "epoch": 0.14654879292757567,
+      "grad_norm": 2.423393500928274,
+      "learning_rate": 9.649368871769908e-06,
+      "loss": 0.7891,
+      "step": 431
+    },
+    {
+      "epoch": 0.14688881332879972,
+      "grad_norm": 2.29836823405829,
+      "learning_rate": 9.647339885349736e-06,
+      "loss": 0.8961,
+      "step": 432
+    },
+    {
+      "epoch": 0.1472288337300238,
+      "grad_norm": 3.041323557001672,
+      "learning_rate": 9.645305259890606e-06,
+      "loss": 0.6884,
+      "step": 433
+    },
+    {
+      "epoch": 0.14756885413124787,
+      "grad_norm": 2.1635297868724486,
+      "learning_rate": 9.643264997861312e-06,
+      "loss": 0.8554,
+      "step": 434
+    },
+    {
+      "epoch": 0.14790887453247195,
+      "grad_norm": 2.0623308876529896,
+      "learning_rate": 9.641219101737489e-06,
+      "loss": 0.8993,
+      "step": 435
+    },
+    {
+      "epoch": 0.14824889493369603,
+      "grad_norm": 1.9701367858930552,
+      "learning_rate": 9.639167574001608e-06,
+      "loss": 0.7581,
+      "step": 436
+    },
+    {
+      "epoch": 0.1485889153349201,
+      "grad_norm": 2.2500245249695365,
+      "learning_rate": 9.637110417142975e-06,
+      "loss": 0.7519,
+      "step": 437
+    },
+    {
+      "epoch": 0.14892893573614416,
+      "grad_norm": 2.2284060545707187,
+      "learning_rate": 9.635047633657723e-06,
+      "loss": 0.9183,
+      "step": 438
+    },
+    {
+      "epoch": 0.14926895613736824,
+      "grad_norm": 2.286832136889049,
+      "learning_rate": 9.632979226048816e-06,
+      "loss": 0.8386,
+      "step": 439
+    },
+    {
+      "epoch": 0.14960897653859231,
+      "grad_norm": 2.2951771970261143,
+      "learning_rate": 9.630905196826039e-06,
+      "loss": 0.8065,
+      "step": 440
+    },
+    {
+      "epoch": 0.1499489969398164,
+      "grad_norm": 2.131373688797792,
+      "learning_rate": 9.628825548506002e-06,
+      "loss": 0.7767,
+      "step": 441
+    },
+    {
+      "epoch": 0.15028901734104047,
+      "grad_norm": 2.091224107238558,
+      "learning_rate": 9.62674028361213e-06,
+      "loss": 0.8655,
+      "step": 442
+    },
+    {
+      "epoch": 0.15062903774226455,
+      "grad_norm": 3.2172828773538993,
+      "learning_rate": 9.624649404674661e-06,
+      "loss": 0.9147,
+      "step": 443
+    },
+    {
+      "epoch": 0.1509690581434886,
+      "grad_norm": 2.025151440715302,
+      "learning_rate": 9.622552914230655e-06,
+      "loss": 0.9121,
+      "step": 444
+    },
+    {
+      "epoch": 0.15130907854471268,
+      "grad_norm": 1.9528957154547468,
+      "learning_rate": 9.620450814823966e-06,
+      "loss": 0.8995,
+      "step": 445
+    },
+    {
+      "epoch": 0.15164909894593676,
+      "grad_norm": 2.2336331464426245,
+      "learning_rate": 9.618343109005266e-06,
+      "loss": 0.7953,
+      "step": 446
+    },
+    {
+      "epoch": 0.15198911934716083,
+      "grad_norm": 2.247828787686121,
+      "learning_rate": 9.616229799332026e-06,
+      "loss": 0.9126,
+      "step": 447
+    },
+    {
+      "epoch": 0.1523291397483849,
+      "grad_norm": 3.0710622235399967,
+      "learning_rate": 9.614110888368515e-06,
+      "loss": 0.7671,
+      "step": 448
+    },
+    {
+      "epoch": 0.152669160149609,
+      "grad_norm": 2.240205069427427,
+      "learning_rate": 9.6119863786858e-06,
+      "loss": 0.8534,
+      "step": 449
+    },
+    {
+      "epoch": 0.15300918055083304,
+      "grad_norm": 2.3440964305088374,
+      "learning_rate": 9.609856272861742e-06,
+      "loss": 0.8859,
+      "step": 450
+    },
+    {
+      "epoch": 0.15334920095205712,
+      "grad_norm": 2.947306903901322,
+      "learning_rate": 9.607720573480991e-06,
+      "loss": 0.8971,
+      "step": 451
+    },
+    {
+      "epoch": 0.1536892213532812,
+      "grad_norm": 1.9814248823269929,
+      "learning_rate": 9.605579283134985e-06,
+      "loss": 0.8666,
+      "step": 452
+    },
+    {
+      "epoch": 0.15402924175450527,
+      "grad_norm": 1.9414054426785363,
+      "learning_rate": 9.603432404421947e-06,
+      "loss": 0.83,
+      "step": 453
+    },
+    {
+      "epoch": 0.15436926215572935,
+      "grad_norm": 1.9508100977088108,
+      "learning_rate": 9.601279939946874e-06,
+      "loss": 0.7941,
+      "step": 454
+    },
+    {
+      "epoch": 0.1547092825569534,
+      "grad_norm": 2.738798819972234,
+      "learning_rate": 9.599121892321554e-06,
+      "loss": 0.7554,
+      "step": 455
+    },
+    {
+      "epoch": 0.15504930295817748,
+      "grad_norm": 2.2388309465744274,
+      "learning_rate": 9.59695826416454e-06,
+      "loss": 0.8047,
+      "step": 456
+    },
+    {
+      "epoch": 0.15538932335940156,
+      "grad_norm": 1.9102221923547757,
+      "learning_rate": 9.594789058101154e-06,
+      "loss": 0.7742,
+      "step": 457
+    },
+    {
+      "epoch": 0.15572934376062564,
+      "grad_norm": 2.176425662506411,
+      "learning_rate": 9.592614276763494e-06,
+      "loss": 0.8392,
+      "step": 458
+    },
+    {
+      "epoch": 0.15606936416184972,
+      "grad_norm": 2.075696818549719,
+      "learning_rate": 9.590433922790418e-06,
+      "loss": 0.8328,
+      "step": 459
+    },
+    {
+      "epoch": 0.1564093845630738,
+      "grad_norm": 2.45612251063091,
+      "learning_rate": 9.58824799882755e-06,
+      "loss": 0.8015,
+      "step": 460
+    },
+    {
+      "epoch": 0.15674940496429784,
+      "grad_norm": 1.8178552528317342,
+      "learning_rate": 9.586056507527266e-06,
+      "loss": 0.9039,
+      "step": 461
+    },
+    {
+      "epoch": 0.15708942536552192,
+      "grad_norm": 2.124214107555732,
+      "learning_rate": 9.583859451548703e-06,
+      "loss": 0.8113,
+      "step": 462
+    },
+    {
+      "epoch": 0.157429445766746,
+      "grad_norm": 1.5843205381627385,
+      "learning_rate": 9.581656833557749e-06,
+      "loss": 0.8248,
+      "step": 463
+    },
+    {
+      "epoch": 0.15776946616797008,
+      "grad_norm": 1.9306313527246615,
+      "learning_rate": 9.57944865622704e-06,
+      "loss": 0.7318,
+      "step": 464
+    },
+    {
+      "epoch": 0.15810948656919416,
+      "grad_norm": 2.9413967318596943,
+      "learning_rate": 9.577234922235954e-06,
+      "loss": 0.8524,
+      "step": 465
+    },
+    {
+      "epoch": 0.15844950697041824,
+      "grad_norm": 1.571426184030293,
+      "learning_rate": 9.575015634270619e-06,
+      "loss": 0.9224,
+      "step": 466
+    },
+    {
+      "epoch": 0.1587895273716423,
+      "grad_norm": 1.8730771122977774,
+      "learning_rate": 9.5727907950239e-06,
+      "loss": 0.7957,
+      "step": 467
+    },
+    {
+      "epoch": 0.15912954777286636,
+      "grad_norm": 2.376107493345504,
+      "learning_rate": 9.570560407195392e-06,
+      "loss": 0.7542,
+      "step": 468
+    },
+    {
+      "epoch": 0.15946956817409044,
+      "grad_norm": 1.9384094700182535,
+      "learning_rate": 9.568324473491431e-06,
+      "loss": 0.7407,
+      "step": 469
+    },
+    {
+      "epoch": 0.15980958857531452,
+      "grad_norm": 1.7937843614169016,
+      "learning_rate": 9.566082996625072e-06,
+      "loss": 0.7993,
+      "step": 470
+    },
+    {
+      "epoch": 0.1601496089765386,
+      "grad_norm": 3.3373628850176127,
+      "learning_rate": 9.56383597931611e-06,
+      "loss": 0.7848,
+      "step": 471
+    },
+    {
+      "epoch": 0.16048962937776268,
+      "grad_norm": 2.1659690728359697,
+      "learning_rate": 9.561583424291048e-06,
+      "loss": 0.8287,
+      "step": 472
+    },
+    {
+      "epoch": 0.16082964977898673,
+      "grad_norm": 2.2389421306989234,
+      "learning_rate": 9.55932533428312e-06,
+      "loss": 0.7587,
+      "step": 473
+    },
+    {
+      "epoch": 0.1611696701802108,
+      "grad_norm": 2.818398736688453,
+      "learning_rate": 9.557061712032269e-06,
+      "loss": 0.8222,
+      "step": 474
+    },
+    {
+      "epoch": 0.16150969058143488,
+      "grad_norm": 2.441252664404201,
+      "learning_rate": 9.554792560285152e-06,
+      "loss": 0.734,
+      "step": 475
+    },
+    {
+      "epoch": 0.16184971098265896,
+      "grad_norm": 2.295687270495865,
+      "learning_rate": 9.552517881795142e-06,
+      "loss": 0.8626,
+      "step": 476
+    },
+    {
+      "epoch": 0.16218973138388304,
+      "grad_norm": 13.760896349921383,
+      "learning_rate": 9.550237679322308e-06,
+      "loss": 0.8463,
+      "step": 477
+    },
+    {
+      "epoch": 0.16252975178510712,
+      "grad_norm": 2.1715258052291047,
+      "learning_rate": 9.547951955633428e-06,
+      "loss": 0.7491,
+      "step": 478
+    },
+    {
+      "epoch": 0.16286977218633117,
+      "grad_norm": 2.2255593479400906,
+      "learning_rate": 9.545660713501975e-06,
+      "loss": 0.9064,
+      "step": 479
+    },
+    {
+      "epoch": 0.16320979258755525,
+      "grad_norm": 2.21806151643282,
+      "learning_rate": 9.543363955708124e-06,
+      "loss": 0.8289,
+      "step": 480
+    },
+    {
+      "epoch": 0.16354981298877933,
+      "grad_norm": 2.188397294600766,
+      "learning_rate": 9.541061685038742e-06,
+      "loss": 0.8429,
+      "step": 481
+    },
+    {
+      "epoch": 0.1638898333900034,
+      "grad_norm": 2.166972985318867,
+      "learning_rate": 9.538753904287376e-06,
+      "loss": 0.9443,
+      "step": 482
+    },
+    {
+      "epoch": 0.16422985379122748,
+      "grad_norm": 2.455788846295091,
+      "learning_rate": 9.53644061625427e-06,
+      "loss": 0.7398,
+      "step": 483
+    },
+    {
+      "epoch": 0.16456987419245156,
+      "grad_norm": 2.4033301722625886,
+      "learning_rate": 9.534121823746348e-06,
+      "loss": 0.8728,
+      "step": 484
+    },
+    {
+      "epoch": 0.1649098945936756,
+      "grad_norm": 2.2276672135131634,
+      "learning_rate": 9.531797529577205e-06,
+      "loss": 0.9371,
+      "step": 485
+    },
+    {
+      "epoch": 0.1652499149948997,
+      "grad_norm": 2.434555018080122,
+      "learning_rate": 9.529467736567124e-06,
+      "loss": 0.9057,
+      "step": 486
+    },
+    {
+      "epoch": 0.16558993539612377,
+      "grad_norm": 2.427937801753027,
+      "learning_rate": 9.527132447543051e-06,
+      "loss": 0.8455,
+      "step": 487
+    },
+    {
+      "epoch": 0.16592995579734784,
+      "grad_norm": 3.4006562033751817,
+      "learning_rate": 9.524791665338606e-06,
+      "loss": 0.8247,
+      "step": 488
+    },
+    {
+      "epoch": 0.16626997619857192,
+      "grad_norm": 2.7613015303421466,
+      "learning_rate": 9.522445392794069e-06,
+      "loss": 0.8169,
+      "step": 489
+    },
+    {
+      "epoch": 0.166609996599796,
+      "grad_norm": 2.6236255693220323,
+      "learning_rate": 9.520093632756388e-06,
+      "loss": 0.7666,
+      "step": 490
+    },
+    {
+      "epoch": 0.16695001700102005,
+      "grad_norm": 2.389219152903732,
+      "learning_rate": 9.517736388079169e-06,
+      "loss": 0.8067,
+      "step": 491
+    },
+    {
+      "epoch": 0.16729003740224413,
+      "grad_norm": 1.7505382569470098,
+      "learning_rate": 9.515373661622665e-06,
+      "loss": 0.8714,
+      "step": 492
+    },
+    {
+      "epoch": 0.1676300578034682,
+      "grad_norm": 2.0417205501733795,
+      "learning_rate": 9.51300545625379e-06,
+      "loss": 0.8,
+      "step": 493
+    },
+    {
+      "epoch": 0.16797007820469229,
+      "grad_norm": 2.245402590882533,
+      "learning_rate": 9.510631774846099e-06,
+      "loss": 0.762,
+      "step": 494
+    },
+    {
+      "epoch": 0.16831009860591636,
+      "grad_norm": 2.82928636204283,
+      "learning_rate": 9.5082526202798e-06,
+      "loss": 0.6621,
+      "step": 495
+    },
+    {
+      "epoch": 0.16865011900714044,
+      "grad_norm": 2.387297511941106,
+      "learning_rate": 9.505867995441734e-06,
+      "loss": 0.8231,
+      "step": 496
+    },
+    {
+      "epoch": 0.1689901394083645,
+      "grad_norm": 3.8013261588105927,
+      "learning_rate": 9.503477903225382e-06,
+      "loss": 0.8885,
+      "step": 497
+    },
+    {
+      "epoch": 0.16933015980958857,
+      "grad_norm": 1.9582651742957375,
+      "learning_rate": 9.501082346530864e-06,
+      "loss": 0.7235,
+      "step": 498
+    },
+    {
+      "epoch": 0.16967018021081265,
+      "grad_norm": 2.0197184013240497,
+      "learning_rate": 9.498681328264919e-06,
+      "loss": 0.8888,
+      "step": 499
+    },
+    {
+      "epoch": 0.17001020061203673,
+      "grad_norm": 1.9383228814810216,
+      "learning_rate": 9.496274851340926e-06,
+      "loss": 0.7643,
+      "step": 500
+    },
+    {
+      "epoch": 0.1703502210132608,
+      "grad_norm": 2.2775734934604728,
+      "learning_rate": 9.49386291867888e-06,
+      "loss": 0.7317,
+      "step": 501
+    },
+    {
+      "epoch": 0.17069024141448486,
+      "grad_norm": 1.6625027584786152,
+      "learning_rate": 9.491445533205397e-06,
+      "loss": 0.8367,
+      "step": 502
+    },
+    {
+      "epoch": 0.17103026181570893,
+      "grad_norm": 2.1475307975125113,
+      "learning_rate": 9.48902269785371e-06,
+      "loss": 0.8449,
+      "step": 503
+    },
+    {
+      "epoch": 0.171370282216933,
+      "grad_norm": 2.347052710314186,
+      "learning_rate": 9.486594415563665e-06,
+      "loss": 0.867,
+      "step": 504
+    },
+    {
+      "epoch": 0.1717103026181571,
+      "grad_norm": 1.6939167607337662,
+      "learning_rate": 9.484160689281718e-06,
+      "loss": 0.8089,
+      "step": 505
+    },
+    {
+      "epoch": 0.17205032301938117,
+      "grad_norm": 2.7074248206478786,
+      "learning_rate": 9.48172152196093e-06,
+      "loss": 0.9275,
+      "step": 506
+    },
+    {
+      "epoch": 0.17239034342060525,
+      "grad_norm": 1.9623542563803935,
+      "learning_rate": 9.47927691656096e-06,
+      "loss": 0.7276,
+      "step": 507
+    },
+    {
+      "epoch": 0.1727303638218293,
+      "grad_norm": 2.2577590094858433,
+      "learning_rate": 9.476826876048076e-06,
+      "loss": 0.8322,
+      "step": 508
+    },
+    {
+      "epoch": 0.17307038422305338,
+      "grad_norm": 2.610689837249583,
+      "learning_rate": 9.474371403395129e-06,
+      "loss": 0.7989,
+      "step": 509
+    },
+    {
+      "epoch": 0.17341040462427745,
+      "grad_norm": 2.981168203823366,
+      "learning_rate": 9.47191050158157e-06,
+      "loss": 0.8787,
+      "step": 510
+    },
+    {
+      "epoch": 0.17375042502550153,
+      "grad_norm": 2.0910514651851937,
+      "learning_rate": 9.469444173593433e-06,
+      "loss": 0.8342,
+      "step": 511
+    },
+    {
+      "epoch": 0.1740904454267256,
+      "grad_norm": 2.327443732050833,
+      "learning_rate": 9.466972422423338e-06,
+      "loss": 0.7471,
+      "step": 512
+    },
+    {
+      "epoch": 0.1744304658279497,
+      "grad_norm": 1.8968205078548026,
+      "learning_rate": 9.464495251070483e-06,
+      "loss": 0.8071,
+      "step": 513
+    },
+    {
+      "epoch": 0.17477048622917374,
+      "grad_norm": 5.35775514905284,
+      "learning_rate": 9.462012662540645e-06,
+      "loss": 0.7672,
+      "step": 514
+    },
+    {
+      "epoch": 0.17511050663039782,
+      "grad_norm": 1.8205374867928585,
+      "learning_rate": 9.459524659846176e-06,
+      "loss": 0.8094,
+      "step": 515
+    },
+    {
+      "epoch": 0.1754505270316219,
+      "grad_norm": 2.70503593128861,
+      "learning_rate": 9.457031246005994e-06,
+      "loss": 0.8121,
+      "step": 516
+    },
+    {
+      "epoch": 0.17579054743284597,
+      "grad_norm": 2.097674274994687,
+      "learning_rate": 9.454532424045585e-06,
+      "loss": 0.7831,
+      "step": 517
+    },
+    {
+      "epoch": 0.17613056783407005,
+      "grad_norm": 2.541108967039887,
+      "learning_rate": 9.452028196996994e-06,
+      "loss": 0.7744,
+      "step": 518
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 2.448989357418417,
+      "learning_rate": 9.449518567898827e-06,
+      "loss": 0.8201,
+      "step": 519
+    },
+    {
+      "epoch": 0.17681060863651818,
+      "grad_norm": 2.4183579990716972,
+      "learning_rate": 9.44700353979625e-06,
+      "loss": 0.9491,
+      "step": 520
+    },
+    {
+      "epoch": 0.17715062903774226,
+      "grad_norm": 2.0371443097616204,
+      "learning_rate": 9.444483115740968e-06,
+      "loss": 0.8665,
+      "step": 521
+    },
+    {
+      "epoch": 0.17749064943896634,
+      "grad_norm": 2.417951434607015,
+      "learning_rate": 9.441957298791243e-06,
+      "loss": 0.7236,
+      "step": 522
+    },
+    {
+      "epoch": 0.17783066984019041,
+      "grad_norm": 3.2234895274833817,
+      "learning_rate": 9.439426092011877e-06,
+      "loss": 0.8275,
+      "step": 523
+    },
+    {
+      "epoch": 0.1781706902414145,
+      "grad_norm": 2.6553913479919773,
+      "learning_rate": 9.436889498474213e-06,
+      "loss": 0.8412,
+      "step": 524
+    },
+    {
+      "epoch": 0.17851071064263857,
+      "grad_norm": 3.1030470159238392,
+      "learning_rate": 9.434347521256131e-06,
+      "loss": 0.832,
+      "step": 525
+    },
+    {
+      "epoch": 0.17885073104386262,
+      "grad_norm": 1.8409627332088008,
+      "learning_rate": 9.431800163442043e-06,
+      "loss": 0.843,
+      "step": 526
+    },
+    {
+      "epoch": 0.1791907514450867,
+      "grad_norm": 3.186355638430994,
+      "learning_rate": 9.429247428122886e-06,
+      "loss": 0.707,
+      "step": 527
+    },
+    {
+      "epoch": 0.17953077184631078,
+      "grad_norm": 2.0444710139532516,
+      "learning_rate": 9.426689318396128e-06,
+      "loss": 0.8321,
+      "step": 528
+    },
+    {
+      "epoch": 0.17987079224753486,
+      "grad_norm": 1.7029237292350985,
+      "learning_rate": 9.424125837365754e-06,
+      "loss": 0.8387,
+      "step": 529
+    },
+    {
+      "epoch": 0.18021081264875893,
+      "grad_norm": 2.2003972951001427,
+      "learning_rate": 9.42155698814227e-06,
+      "loss": 0.7516,
+      "step": 530
+    },
+    {
+      "epoch": 0.180550833049983,
+      "grad_norm": 2.557957310252339,
+      "learning_rate": 9.41898277384269e-06,
+      "loss": 0.8797,
+      "step": 531
+    },
+    {
+      "epoch": 0.18089085345120706,
+      "grad_norm": 2.895779678922211,
+      "learning_rate": 9.416403197590547e-06,
+      "loss": 0.823,
+      "step": 532
+    },
+    {
+      "epoch": 0.18123087385243114,
+      "grad_norm": 2.0167366993376947,
+      "learning_rate": 9.41381826251587e-06,
+      "loss": 0.9268,
+      "step": 533
+    },
+    {
+      "epoch": 0.18157089425365522,
+      "grad_norm": 2.2879605857864265,
+      "learning_rate": 9.411227971755197e-06,
+      "loss": 0.9309,
+      "step": 534
+    },
+    {
+      "epoch": 0.1819109146548793,
+      "grad_norm": 5.454574823683741,
+      "learning_rate": 9.408632328451565e-06,
+      "loss": 0.8586,
+      "step": 535
+    },
+    {
+      "epoch": 0.18225093505610337,
+      "grad_norm": 2.145392976793967,
+      "learning_rate": 9.4060313357545e-06,
+      "loss": 0.8267,
+      "step": 536
+    },
+    {
+      "epoch": 0.18259095545732745,
+      "grad_norm": 2.4134835759822377,
+      "learning_rate": 9.403424996820024e-06,
+      "loss": 0.8951,
+      "step": 537
+    },
+    {
+      "epoch": 0.1829309758585515,
+      "grad_norm": 2.490795625384553,
+      "learning_rate": 9.400813314810644e-06,
+      "loss": 0.8217,
+      "step": 538
+    },
+    {
+      "epoch": 0.18327099625977558,
+      "grad_norm": 4.105744879893649,
+      "learning_rate": 9.39819629289535e-06,
+      "loss": 0.7641,
+      "step": 539
+    },
+    {
+      "epoch": 0.18361101666099966,
+      "grad_norm": 2.663813984562291,
+      "learning_rate": 9.395573934249614e-06,
+      "loss": 0.8811,
+      "step": 540
+    },
+    {
+      "epoch": 0.18395103706222374,
+      "grad_norm": 2.3379526523751037,
+      "learning_rate": 9.392946242055379e-06,
+      "loss": 0.8157,
+      "step": 541
+    },
+    {
+      "epoch": 0.18429105746344782,
+      "grad_norm": 2.460601877280579,
+      "learning_rate": 9.390313219501061e-06,
+      "loss": 0.8666,
+      "step": 542
+    },
+    {
+      "epoch": 0.18463107786467187,
+      "grad_norm": 2.4088692423050144,
+      "learning_rate": 9.38767486978155e-06,
+      "loss": 0.876,
+      "step": 543
+    },
+    {
+      "epoch": 0.18497109826589594,
+      "grad_norm": 1.5738286525981031,
+      "learning_rate": 9.385031196098194e-06,
+      "loss": 0.7488,
+      "step": 544
+    },
+    {
+      "epoch": 0.18531111866712002,
+      "grad_norm": 4.532392662900621,
+      "learning_rate": 9.3823822016588e-06,
+      "loss": 0.8693,
+      "step": 545
+    },
+    {
+      "epoch": 0.1856511390683441,
+      "grad_norm": 2.237189674784233,
+      "learning_rate": 9.379727889677632e-06,
+      "loss": 0.8958,
+      "step": 546
+    },
+    {
+      "epoch": 0.18599115946956818,
+      "grad_norm": 1.554107484360944,
+      "learning_rate": 9.377068263375411e-06,
+      "loss": 0.7866,
+      "step": 547
+    },
+    {
+      "epoch": 0.18633117987079226,
+      "grad_norm": 2.4467881402130125,
+      "learning_rate": 9.374403325979301e-06,
+      "loss": 0.8856,
+      "step": 548
+    },
+    {
+      "epoch": 0.1866712002720163,
+      "grad_norm": 2.4507416677378213,
+      "learning_rate": 9.371733080722911e-06,
+      "loss": 0.7532,
+      "step": 549
+    },
+    {
+      "epoch": 0.18701122067324039,
+      "grad_norm": 1.427343424523804,
+      "learning_rate": 9.369057530846294e-06,
+      "loss": 0.8418,
+      "step": 550
+    },
+    {
+      "epoch": 0.18735124107446446,
+      "grad_norm": 3.24841097172593,
+      "learning_rate": 9.366376679595936e-06,
+      "loss": 0.8738,
+      "step": 551
+    },
+    {
+      "epoch": 0.18769126147568854,
+      "grad_norm": 1.8294498367391419,
+      "learning_rate": 9.363690530224757e-06,
+      "loss": 0.9536,
+      "step": 552
+    },
+    {
+      "epoch": 0.18803128187691262,
+      "grad_norm": 3.147835292974788,
+      "learning_rate": 9.360999085992106e-06,
+      "loss": 0.9387,
+      "step": 553
+    },
+    {
+      "epoch": 0.1883713022781367,
+      "grad_norm": 1.7595455512097826,
+      "learning_rate": 9.358302350163758e-06,
+      "loss": 0.893,
+      "step": 554
+    },
+    {
+      "epoch": 0.18871132267936075,
+      "grad_norm": 2.0297149767885587,
+      "learning_rate": 9.355600326011903e-06,
+      "loss": 0.8648,
+      "step": 555
+    },
+    {
+      "epoch": 0.18905134308058483,
+      "grad_norm": 2.669731699705077,
+      "learning_rate": 9.352893016815155e-06,
+      "loss": 0.8835,
+      "step": 556
+    },
+    {
+      "epoch": 0.1893913634818089,
+      "grad_norm": 5.446794892812443,
+      "learning_rate": 9.350180425858538e-06,
+      "loss": 0.7767,
+      "step": 557
+    },
+    {
+      "epoch": 0.18973138388303298,
+      "grad_norm": 1.8268464771652575,
+      "learning_rate": 9.347462556433483e-06,
+      "loss": 0.7565,
+      "step": 558
+    },
+    {
+      "epoch": 0.19007140428425706,
+      "grad_norm": 1.642166985619522,
+      "learning_rate": 9.34473941183783e-06,
+      "loss": 0.8424,
+      "step": 559
+    },
+    {
+      "epoch": 0.19041142468548114,
+      "grad_norm": 2.0901831217312625,
+      "learning_rate": 9.342010995375811e-06,
+      "loss": 0.7805,
+      "step": 560
+    },
+    {
+      "epoch": 0.1907514450867052,
+      "grad_norm": 1.90871026730968,
+      "learning_rate": 9.33927731035807e-06,
+      "loss": 0.7668,
+      "step": 561
+    },
+    {
+      "epoch": 0.19109146548792927,
+      "grad_norm": 3.4307001624182316,
+      "learning_rate": 9.336538360101631e-06,
+      "loss": 0.8382,
+      "step": 562
+    },
+    {
+      "epoch": 0.19143148588915335,
+      "grad_norm": 2.393265808564416,
+      "learning_rate": 9.333794147929907e-06,
+      "loss": 0.7788,
+      "step": 563
+    },
+    {
+      "epoch": 0.19177150629037742,
+      "grad_norm": 2.022566960473697,
+      "learning_rate": 9.331044677172705e-06,
+      "loss": 0.744,
+      "step": 564
+    },
+    {
+      "epoch": 0.1921115266916015,
+      "grad_norm": 3.3062922036782836,
+      "learning_rate": 9.328289951166205e-06,
+      "loss": 0.8229,
+      "step": 565
+    },
+    {
+      "epoch": 0.19245154709282558,
+      "grad_norm": 3.0841654373647516,
+      "learning_rate": 9.325529973252967e-06,
+      "loss": 0.6495,
+      "step": 566
+    },
+    {
+      "epoch": 0.19279156749404963,
+      "grad_norm": 5.4892675893239105,
+      "learning_rate": 9.32276474678192e-06,
+      "loss": 0.8487,
+      "step": 567
+    },
+    {
+      "epoch": 0.1931315878952737,
+      "grad_norm": 2.013856083284362,
+      "learning_rate": 9.319994275108365e-06,
+      "loss": 0.9441,
+      "step": 568
+    },
+    {
+      "epoch": 0.1934716082964978,
+      "grad_norm": 2.4066098547440897,
+      "learning_rate": 9.31721856159397e-06,
+      "loss": 0.8576,
+      "step": 569
+    },
+    {
+      "epoch": 0.19381162869772187,
+      "grad_norm": 2.0530822158137023,
+      "learning_rate": 9.314437609606754e-06,
+      "loss": 0.7699,
+      "step": 570
+    },
+    {
+      "epoch": 0.19415164909894594,
+      "grad_norm": 1.7520902212648368,
+      "learning_rate": 9.311651422521103e-06,
+      "loss": 0.8794,
+      "step": 571
+    },
+    {
+      "epoch": 0.19449166950017002,
+      "grad_norm": 2.034565210374195,
+      "learning_rate": 9.308860003717748e-06,
+      "loss": 0.8773,
+      "step": 572
+    },
+    {
+      "epoch": 0.19483168990139407,
+      "grad_norm": 1.8451050446755255,
+      "learning_rate": 9.306063356583772e-06,
+      "loss": 0.7947,
+      "step": 573
+    },
+    {
+      "epoch": 0.19517171030261815,
+      "grad_norm": 2.010334432898869,
+      "learning_rate": 9.3032614845126e-06,
+      "loss": 0.8639,
+      "step": 574
+    },
+    {
+      "epoch": 0.19551173070384223,
+      "grad_norm": 1.8970535978324625,
+      "learning_rate": 9.300454390903999e-06,
+      "loss": 0.74,
+      "step": 575
+    },
+    {
+      "epoch": 0.1958517511050663,
+      "grad_norm": 2.3148877609826544,
+      "learning_rate": 9.297642079164067e-06,
+      "loss": 0.8328,
+      "step": 576
+    },
+    {
+      "epoch": 0.19619177150629039,
+      "grad_norm": 1.8964922349773508,
+      "learning_rate": 9.294824552705238e-06,
+      "loss": 0.7799,
+      "step": 577
+    },
+    {
+      "epoch": 0.19653179190751446,
+      "grad_norm": 1.970465125409274,
+      "learning_rate": 9.292001814946275e-06,
+      "loss": 0.8337,
+      "step": 578
+    },
+    {
+      "epoch": 0.19687181230873851,
+      "grad_norm": 1.8923831770943373,
+      "learning_rate": 9.289173869312259e-06,
+      "loss": 0.9365,
+      "step": 579
+    },
+    {
+      "epoch": 0.1972118327099626,
+      "grad_norm": 5.110858953274893,
+      "learning_rate": 9.286340719234592e-06,
+      "loss": 0.8185,
+      "step": 580
+    },
+    {
+      "epoch": 0.19755185311118667,
+      "grad_norm": 2.090608423644068,
+      "learning_rate": 9.283502368150996e-06,
+      "loss": 0.8934,
+      "step": 581
+    },
+    {
+      "epoch": 0.19789187351241075,
+      "grad_norm": 2.016487940198634,
+      "learning_rate": 9.280658819505495e-06,
+      "loss": 0.7756,
+      "step": 582
+    },
+    {
+      "epoch": 0.19823189391363483,
+      "grad_norm": 2.221575121411304,
+      "learning_rate": 9.277810076748427e-06,
+      "loss": 0.821,
+      "step": 583
+    },
+    {
+      "epoch": 0.1985719143148589,
+      "grad_norm": 1.8543548108937444,
+      "learning_rate": 9.274956143336433e-06,
+      "loss": 0.873,
+      "step": 584
+    },
+    {
+      "epoch": 0.19891193471608296,
+      "grad_norm": 2.259730694848992,
+      "learning_rate": 9.272097022732444e-06,
+      "loss": 0.8786,
+      "step": 585
+    },
+    {
+      "epoch": 0.19925195511730703,
+      "grad_norm": 1.848711170198966,
+      "learning_rate": 9.269232718405692e-06,
+      "loss": 0.8858,
+      "step": 586
+    },
+    {
+      "epoch": 0.1995919755185311,
+      "grad_norm": 1.866200683323318,
+      "learning_rate": 9.266363233831697e-06,
+      "loss": 0.8016,
+      "step": 587
+    },
+    {
+      "epoch": 0.1999319959197552,
+      "grad_norm": 14.97505212921413,
+      "learning_rate": 9.263488572492267e-06,
+      "loss": 0.7263,
+      "step": 588
+    },
+    {
+      "epoch": 0.20027201632097927,
+      "grad_norm": 2.15282534325055,
+      "learning_rate": 9.260608737875487e-06,
+      "loss": 0.83,
+      "step": 589
+    },
+    {
+      "epoch": 0.20061203672220332,
+      "grad_norm": 3.0728740665252308,
+      "learning_rate": 9.257723733475723e-06,
+      "loss": 0.8643,
+      "step": 590
+    },
+    {
+      "epoch": 0.2009520571234274,
+      "grad_norm": 1.9579947280234309,
+      "learning_rate": 9.25483356279361e-06,
+      "loss": 0.7628,
+      "step": 591
+    },
+    {
+      "epoch": 0.20129207752465147,
+      "grad_norm": 2.0595756759778645,
+      "learning_rate": 9.251938229336057e-06,
+      "loss": 0.7825,
+      "step": 592
+    },
+    {
+      "epoch": 0.20163209792587555,
+      "grad_norm": 3.9099946836071258,
+      "learning_rate": 9.249037736616235e-06,
+      "loss": 0.9511,
+      "step": 593
+    },
+    {
+      "epoch": 0.20197211832709963,
+      "grad_norm": 1.9453043208275265,
+      "learning_rate": 9.24613208815357e-06,
+      "loss": 0.8037,
+      "step": 594
+    },
+    {
+      "epoch": 0.2023121387283237,
+      "grad_norm": 2.067317387394065,
+      "learning_rate": 9.243221287473755e-06,
+      "loss": 0.8915,
+      "step": 595
+    },
+    {
+      "epoch": 0.20265215912954776,
+      "grad_norm": 1.7871868965358906,
+      "learning_rate": 9.240305338108726e-06,
+      "loss": 0.9012,
+      "step": 596
+    },
+    {
+      "epoch": 0.20299217953077184,
+      "grad_norm": 1.742758087393586,
+      "learning_rate": 9.237384243596667e-06,
+      "loss": 0.7241,
+      "step": 597
+    },
+    {
+      "epoch": 0.20333219993199592,
+      "grad_norm": 3.9506555587545105,
+      "learning_rate": 9.23445800748201e-06,
+      "loss": 0.8664,
+      "step": 598
+    },
+    {
+      "epoch": 0.20367222033322,
+      "grad_norm": 2.1447546142847704,
+      "learning_rate": 9.231526633315419e-06,
+      "loss": 0.8176,
+      "step": 599
+    },
+    {
+      "epoch": 0.20401224073444407,
+      "grad_norm": 2.0350086212804848,
+      "learning_rate": 9.2285901246538e-06,
+      "loss": 0.858,
+      "step": 600
+    },
+    {
+      "epoch": 0.20435226113566815,
+      "grad_norm": 2.0664182960390494,
+      "learning_rate": 9.225648485060283e-06,
+      "loss": 0.7872,
+      "step": 601
+    },
+    {
+      "epoch": 0.2046922815368922,
+      "grad_norm": 2.5061940461297714,
+      "learning_rate": 9.222701718104226e-06,
+      "loss": 0.7595,
+      "step": 602
+    },
+    {
+      "epoch": 0.20503230193811628,
+      "grad_norm": 2.0379087741894084,
+      "learning_rate": 9.21974982736121e-06,
+      "loss": 0.8303,
+      "step": 603
+    },
+    {
+      "epoch": 0.20537232233934036,
+      "grad_norm": 2.4684024083448697,
+      "learning_rate": 9.21679281641303e-06,
+      "loss": 0.7877,
+      "step": 604
+    },
+    {
+      "epoch": 0.20571234274056444,
+      "grad_norm": 1.541601484308322,
+      "learning_rate": 9.2138306888477e-06,
+      "loss": 0.8159,
+      "step": 605
+    },
+    {
+      "epoch": 0.2060523631417885,
+      "grad_norm": 2.855409839214273,
+      "learning_rate": 9.21086344825943e-06,
+      "loss": 0.892,
+      "step": 606
+    },
+    {
+      "epoch": 0.2063923835430126,
+      "grad_norm": 2.0094424248725584,
+      "learning_rate": 9.207891098248648e-06,
+      "loss": 0.8376,
+      "step": 607
+    },
+    {
+      "epoch": 0.20673240394423664,
+      "grad_norm": 2.3628874201292103,
+      "learning_rate": 9.204913642421977e-06,
+      "loss": 0.8384,
+      "step": 608
+    },
+    {
+      "epoch": 0.20707242434546072,
+      "grad_norm": 2.7503706346744408,
+      "learning_rate": 9.20193108439223e-06,
+      "loss": 0.8241,
+      "step": 609
+    },
+    {
+      "epoch": 0.2074124447466848,
+      "grad_norm": 1.8912481720048837,
+      "learning_rate": 9.198943427778415e-06,
+      "loss": 0.7518,
+      "step": 610
+    },
+    {
+      "epoch": 0.20775246514790888,
+      "grad_norm": 1.8807659764591915,
+      "learning_rate": 9.19595067620573e-06,
+      "loss": 0.9048,
+      "step": 611
+    },
+    {
+      "epoch": 0.20809248554913296,
+      "grad_norm": 2.5399683576595744,
+      "learning_rate": 9.19295283330555e-06,
+      "loss": 0.787,
+      "step": 612
+    },
+    {
+      "epoch": 0.20843250595035703,
+      "grad_norm": 2.2042041124830756,
+      "learning_rate": 9.189949902715432e-06,
+      "loss": 0.7788,
+      "step": 613
+    },
+    {
+      "epoch": 0.20877252635158108,
+      "grad_norm": 3.1080504073340185,
+      "learning_rate": 9.1869418880791e-06,
+      "loss": 0.9674,
+      "step": 614
+    },
+    {
+      "epoch": 0.20911254675280516,
+      "grad_norm": 4.302778372345475,
+      "learning_rate": 9.183928793046456e-06,
+      "loss": 0.8179,
+      "step": 615
+    },
+    {
+      "epoch": 0.20945256715402924,
+      "grad_norm": 4.5320039265679375,
+      "learning_rate": 9.180910621273555e-06,
+      "loss": 0.8254,
+      "step": 616
+    },
+    {
+      "epoch": 0.20979258755525332,
+      "grad_norm": 1.8992179347695721,
+      "learning_rate": 9.177887376422624e-06,
+      "loss": 0.7908,
+      "step": 617
+    },
+    {
+      "epoch": 0.2101326079564774,
+      "grad_norm": 1.9899814372282567,
+      "learning_rate": 9.174859062162037e-06,
+      "loss": 0.7912,
+      "step": 618
+    },
+    {
+      "epoch": 0.21047262835770147,
+      "grad_norm": 2.476667419018384,
+      "learning_rate": 9.171825682166325e-06,
+      "loss": 0.8038,
+      "step": 619
+    },
+    {
+      "epoch": 0.21081264875892552,
+      "grad_norm": 2.295861238237628,
+      "learning_rate": 9.168787240116162e-06,
+      "loss": 0.8047,
+      "step": 620
+    },
+    {
+      "epoch": 0.2111526691601496,
+      "grad_norm": 2.0894123097550104,
+      "learning_rate": 9.165743739698364e-06,
+      "loss": 0.8888,
+      "step": 621
+    },
+    {
+      "epoch": 0.21149268956137368,
+      "grad_norm": 1.8237207223507654,
+      "learning_rate": 9.162695184605887e-06,
+      "loss": 0.8017,
+      "step": 622
+    },
+    {
+      "epoch": 0.21183270996259776,
+      "grad_norm": 6.442654526204107,
+      "learning_rate": 9.15964157853782e-06,
+      "loss": 0.858,
+      "step": 623
+    },
+    {
+      "epoch": 0.21217273036382184,
+      "grad_norm": 1.7241002709899875,
+      "learning_rate": 9.15658292519938e-06,
+      "loss": 0.8993,
+      "step": 624
+    },
+    {
+      "epoch": 0.21251275076504592,
+      "grad_norm": 2.4396884010337767,
+      "learning_rate": 9.153519228301907e-06,
+      "loss": 0.8945,
+      "step": 625
+    },
+    {
+      "epoch": 0.21285277116626997,
+      "grad_norm": 3.270858990699396,
+      "learning_rate": 9.150450491562864e-06,
+      "loss": 0.8649,
+      "step": 626
+    },
+    {
+      "epoch": 0.21319279156749404,
+      "grad_norm": 3.3296457820914904,
+      "learning_rate": 9.147376718705825e-06,
+      "loss": 0.8044,
+      "step": 627
+    },
+    {
+      "epoch": 0.21353281196871812,
+      "grad_norm": 1.8050916328292812,
+      "learning_rate": 9.144297913460481e-06,
+      "loss": 0.7789,
+      "step": 628
+    },
+    {
+      "epoch": 0.2138728323699422,
+      "grad_norm": 1.6022202427177714,
+      "learning_rate": 9.141214079562624e-06,
+      "loss": 0.7811,
+      "step": 629
+    },
+    {
+      "epoch": 0.21421285277116628,
+      "grad_norm": 3.8320152900329987,
+      "learning_rate": 9.13812522075415e-06,
+      "loss": 0.8481,
+      "step": 630
+    },
+    {
+      "epoch": 0.21455287317239036,
+      "grad_norm": 2.968430851574413,
+      "learning_rate": 9.13503134078305e-06,
+      "loss": 0.8705,
+      "step": 631
+    },
+    {
+      "epoch": 0.2148928935736144,
+      "grad_norm": 2.7320514154269016,
+      "learning_rate": 9.13193244340341e-06,
+      "loss": 0.7574,
+      "step": 632
+    },
+    {
+      "epoch": 0.21523291397483849,
+      "grad_norm": 2.5381990072952987,
+      "learning_rate": 9.128828532375404e-06,
+      "loss": 0.758,
+      "step": 633
+    },
+    {
+      "epoch": 0.21557293437606256,
+      "grad_norm": 4.664824544519933,
+      "learning_rate": 9.125719611465287e-06,
+      "loss": 0.9002,
+      "step": 634
+    },
+    {
+      "epoch": 0.21591295477728664,
+      "grad_norm": 1.8271666953396724,
+      "learning_rate": 9.122605684445397e-06,
+      "loss": 0.8619,
+      "step": 635
+    },
+    {
+      "epoch": 0.21625297517851072,
+      "grad_norm": 1.9433634039375405,
+      "learning_rate": 9.119486755094143e-06,
+      "loss": 0.8429,
+      "step": 636
+    },
+    {
+      "epoch": 0.21659299557973477,
+      "grad_norm": 1.8859622430454264,
+      "learning_rate": 9.116362827196002e-06,
+      "loss": 0.7708,
+      "step": 637
+    },
+    {
+      "epoch": 0.21693301598095885,
+      "grad_norm": 2.211237168392557,
+      "learning_rate": 9.113233904541524e-06,
+      "loss": 0.8633,
+      "step": 638
+    },
+    {
+      "epoch": 0.21727303638218293,
+      "grad_norm": 1.767226866716121,
+      "learning_rate": 9.110099990927311e-06,
+      "loss": 0.9302,
+      "step": 639
+    },
+    {
+      "epoch": 0.217613056783407,
+      "grad_norm": 1.989789546199134,
+      "learning_rate": 9.106961090156026e-06,
+      "loss": 0.8603,
+      "step": 640
+    },
+    {
+      "epoch": 0.21795307718463108,
+      "grad_norm": 4.520549862271871,
+      "learning_rate": 9.103817206036383e-06,
+      "loss": 0.8902,
+      "step": 641
+    },
+    {
+      "epoch": 0.21829309758585516,
+      "grad_norm": 1.938477575541222,
+      "learning_rate": 9.100668342383138e-06,
+      "loss": 0.8366,
+      "step": 642
+    },
+    {
+      "epoch": 0.2186331179870792,
+      "grad_norm": 2.327565546501374,
+      "learning_rate": 9.097514503017098e-06,
+      "loss": 0.7551,
+      "step": 643
+    },
+    {
+      "epoch": 0.2189731383883033,
+      "grad_norm": 2.1109166908567936,
+      "learning_rate": 9.0943556917651e-06,
+      "loss": 0.8257,
+      "step": 644
+    },
+    {
+      "epoch": 0.21931315878952737,
+      "grad_norm": 1.8110479549556548,
+      "learning_rate": 9.091191912460014e-06,
+      "loss": 0.9507,
+      "step": 645
+    },
+    {
+      "epoch": 0.21965317919075145,
+      "grad_norm": 2.251697261146959,
+      "learning_rate": 9.088023168940743e-06,
+      "loss": 0.6991,
+      "step": 646
+    },
+    {
+      "epoch": 0.21999319959197552,
+      "grad_norm": 1.7773244864207038,
+      "learning_rate": 9.08484946505221e-06,
+      "loss": 0.7976,
+      "step": 647
+    },
+    {
+      "epoch": 0.2203332199931996,
+      "grad_norm": 2.3077357797688176,
+      "learning_rate": 9.08167080464536e-06,
+      "loss": 0.8666,
+      "step": 648
+    },
+    {
+      "epoch": 0.22067324039442365,
+      "grad_norm": 2.28086178140718,
+      "learning_rate": 9.078487191577146e-06,
+      "loss": 0.94,
+      "step": 649
+    },
+    {
+      "epoch": 0.22101326079564773,
+      "grad_norm": 1.5485886049410107,
+      "learning_rate": 9.075298629710536e-06,
+      "loss": 0.8475,
+      "step": 650
+    },
+    {
+      "epoch": 0.2213532811968718,
+      "grad_norm": 1.9290472754703258,
+      "learning_rate": 9.072105122914502e-06,
+      "loss": 0.8813,
+      "step": 651
+    },
+    {
+      "epoch": 0.2216933015980959,
+      "grad_norm": 8.21841805972392,
+      "learning_rate": 9.068906675064016e-06,
+      "loss": 0.7745,
+      "step": 652
+    },
+    {
+      "epoch": 0.22203332199931997,
+      "grad_norm": 2.9628280561922713,
+      "learning_rate": 9.065703290040043e-06,
+      "loss": 0.6788,
+      "step": 653
+    },
+    {
+      "epoch": 0.22237334240054404,
+      "grad_norm": 3.0384149610010076,
+      "learning_rate": 9.062494971729542e-06,
+      "loss": 0.8977,
+      "step": 654
+    },
+    {
+      "epoch": 0.2227133628017681,
+      "grad_norm": 3.001689658067599,
+      "learning_rate": 9.059281724025455e-06,
+      "loss": 0.7856,
+      "step": 655
+    },
+    {
+      "epoch": 0.22305338320299217,
+      "grad_norm": 2.4216473597244557,
+      "learning_rate": 9.056063550826708e-06,
+      "loss": 0.8248,
+      "step": 656
+    },
+    {
+      "epoch": 0.22339340360421625,
+      "grad_norm": 2.4916553224709115,
+      "learning_rate": 9.052840456038204e-06,
+      "loss": 0.8426,
+      "step": 657
+    },
+    {
+      "epoch": 0.22373342400544033,
+      "grad_norm": 2.030963651676548,
+      "learning_rate": 9.049612443570814e-06,
+      "loss": 0.8562,
+      "step": 658
+    },
+    {
+      "epoch": 0.2240734444066644,
+      "grad_norm": 2.8064288123845516,
+      "learning_rate": 9.046379517341378e-06,
+      "loss": 0.8298,
+      "step": 659
+    },
+    {
+      "epoch": 0.22441346480788849,
+      "grad_norm": 2.279614877119188,
+      "learning_rate": 9.0431416812727e-06,
+      "loss": 0.8515,
+      "step": 660
+    },
+    {
+      "epoch": 0.22475348520911254,
+      "grad_norm": 2.1516052872817553,
+      "learning_rate": 9.039898939293539e-06,
+      "loss": 0.7463,
+      "step": 661
+    },
+    {
+      "epoch": 0.2250935056103366,
+      "grad_norm": 2.343567497267246,
+      "learning_rate": 9.036651295338608e-06,
+      "loss": 0.8554,
+      "step": 662
+    },
+    {
+      "epoch": 0.2254335260115607,
+      "grad_norm": 2.6761262055968795,
+      "learning_rate": 9.033398753348569e-06,
+      "loss": 0.8184,
+      "step": 663
+    },
+    {
+      "epoch": 0.22577354641278477,
+      "grad_norm": 2.0464322670596493,
+      "learning_rate": 9.030141317270026e-06,
+      "loss": 0.7478,
+      "step": 664
+    },
+    {
+      "epoch": 0.22611356681400885,
+      "grad_norm": 2.2236911397789414,
+      "learning_rate": 9.026878991055521e-06,
+      "loss": 0.9156,
+      "step": 665
+    },
+    {
+      "epoch": 0.22645358721523293,
+      "grad_norm": 2.1101918658429124,
+      "learning_rate": 9.02361177866353e-06,
+      "loss": 0.782,
+      "step": 666
+    },
+    {
+      "epoch": 0.22679360761645698,
+      "grad_norm": 3.224573101223046,
+      "learning_rate": 9.020339684058459e-06,
+      "loss": 0.8831,
+      "step": 667
+    },
+    {
+      "epoch": 0.22713362801768106,
+      "grad_norm": 2.155983593275046,
+      "learning_rate": 9.017062711210638e-06,
+      "loss": 0.8461,
+      "step": 668
+    },
+    {
+      "epoch": 0.22747364841890513,
+      "grad_norm": 3.742735571520871,
+      "learning_rate": 9.013780864096313e-06,
+      "loss": 0.9233,
+      "step": 669
+    },
+    {
+      "epoch": 0.2278136688201292,
+      "grad_norm": 2.11951387339584,
+      "learning_rate": 9.010494146697648e-06,
+      "loss": 0.8415,
+      "step": 670
+    },
+    {
+      "epoch": 0.2281536892213533,
+      "grad_norm": 4.046208706917889,
+      "learning_rate": 9.007202563002715e-06,
+      "loss": 0.8367,
+      "step": 671
+    },
+    {
+      "epoch": 0.22849370962257737,
+      "grad_norm": 2.5726534936217353,
+      "learning_rate": 9.003906117005489e-06,
+      "loss": 0.7983,
+      "step": 672
+    },
+    {
+      "epoch": 0.22883373002380142,
+      "grad_norm": 2.3143058688116334,
+      "learning_rate": 9.000604812705854e-06,
+      "loss": 0.7471,
+      "step": 673
+    },
+    {
+      "epoch": 0.2291737504250255,
+      "grad_norm": 7.689486779358199,
+      "learning_rate": 8.997298654109573e-06,
+      "loss": 0.7961,
+      "step": 674
+    },
+    {
+      "epoch": 0.22951377082624957,
+      "grad_norm": 1.9154873476566034,
+      "learning_rate": 8.993987645228313e-06,
+      "loss": 0.8463,
+      "step": 675
+    },
+    {
+      "epoch": 0.22985379122747365,
+      "grad_norm": 1.9308286258150182,
+      "learning_rate": 8.99067179007962e-06,
+      "loss": 0.7293,
+      "step": 676
+    },
+    {
+      "epoch": 0.23019381162869773,
+      "grad_norm": 3.112154181656327,
+      "learning_rate": 8.987351092686923e-06,
+      "loss": 0.8588,
+      "step": 677
+    },
+    {
+      "epoch": 0.2305338320299218,
+      "grad_norm": 1.7834759216247464,
+      "learning_rate": 8.984025557079523e-06,
+      "loss": 0.8339,
+      "step": 678
+    },
+    {
+      "epoch": 0.23087385243114586,
+      "grad_norm": 2.316905251892595,
+      "learning_rate": 8.980695187292598e-06,
+      "loss": 0.7621,
+      "step": 679
+    },
+    {
+      "epoch": 0.23121387283236994,
+      "grad_norm": 3.101167924048224,
+      "learning_rate": 8.977359987367182e-06,
+      "loss": 0.8604,
+      "step": 680
+    },
+    {
+      "epoch": 0.23155389323359402,
+      "grad_norm": 2.2874629757251634,
+      "learning_rate": 8.97401996135018e-06,
+      "loss": 0.8787,
+      "step": 681
+    },
+    {
+      "epoch": 0.2318939136348181,
+      "grad_norm": 2.4403671006704677,
+      "learning_rate": 8.970675113294348e-06,
+      "loss": 0.9373,
+      "step": 682
+    },
+    {
+      "epoch": 0.23223393403604217,
+      "grad_norm": 2.4115030369064274,
+      "learning_rate": 8.967325447258292e-06,
+      "loss": 0.7396,
+      "step": 683
+    },
+    {
+      "epoch": 0.23257395443726622,
+      "grad_norm": 2.1720889668199708,
+      "learning_rate": 8.963970967306466e-06,
+      "loss": 0.843,
+      "step": 684
+    },
+    {
+      "epoch": 0.2329139748384903,
+      "grad_norm": 1.9401953765991744,
+      "learning_rate": 8.960611677509166e-06,
+      "loss": 0.8625,
+      "step": 685
+    },
+    {
+      "epoch": 0.23325399523971438,
+      "grad_norm": 2.068432286888123,
+      "learning_rate": 8.95724758194252e-06,
+      "loss": 0.8402,
+      "step": 686
+    },
+    {
+      "epoch": 0.23359401564093846,
+      "grad_norm": 1.591243068030512,
+      "learning_rate": 8.953878684688492e-06,
+      "loss": 0.7842,
+      "step": 687
+    },
+    {
+      "epoch": 0.23393403604216254,
+      "grad_norm": 1.7715469138687294,
+      "learning_rate": 8.950504989834873e-06,
+      "loss": 0.8833,
+      "step": 688
+    },
+    {
+      "epoch": 0.2342740564433866,
+      "grad_norm": 2.0026773241901537,
+      "learning_rate": 8.94712650147527e-06,
+      "loss": 0.8189,
+      "step": 689
+    },
+    {
+      "epoch": 0.23461407684461066,
+      "grad_norm": 2.8052765922906917,
+      "learning_rate": 8.943743223709109e-06,
+      "loss": 0.7157,
+      "step": 690
+    },
+    {
+      "epoch": 0.23495409724583474,
+      "grad_norm": 1.7526459634636724,
+      "learning_rate": 8.94035516064163e-06,
+      "loss": 0.7976,
+      "step": 691
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 1.5061919498395846,
+      "learning_rate": 8.936962316383876e-06,
+      "loss": 0.7932,
+      "step": 692
+    },
+    {
+      "epoch": 0.2356341380482829,
+      "grad_norm": 2.5099728137526736,
+      "learning_rate": 8.933564695052692e-06,
+      "loss": 0.7652,
+      "step": 693
+    },
+    {
+      "epoch": 0.23597415844950698,
+      "grad_norm": 1.8950918439469646,
+      "learning_rate": 8.930162300770721e-06,
+      "loss": 0.7014,
+      "step": 694
+    },
+    {
+      "epoch": 0.23631417885073105,
+      "grad_norm": 2.1074334438205358,
+      "learning_rate": 8.926755137666396e-06,
+      "loss": 0.8158,
+      "step": 695
+    },
+    {
+      "epoch": 0.2366541992519551,
+      "grad_norm": 2.093263958473567,
+      "learning_rate": 8.923343209873937e-06,
+      "loss": 0.8099,
+      "step": 696
+    },
+    {
+      "epoch": 0.23699421965317918,
+      "grad_norm": 1.8542225188561288,
+      "learning_rate": 8.919926521533346e-06,
+      "loss": 0.8189,
+      "step": 697
+    },
+    {
+      "epoch": 0.23733424005440326,
+      "grad_norm": 2.087963516864021,
+      "learning_rate": 8.9165050767904e-06,
+      "loss": 0.8313,
+      "step": 698
+    },
+    {
+      "epoch": 0.23767426045562734,
+      "grad_norm": 2.047018307961187,
+      "learning_rate": 8.913078879796648e-06,
+      "loss": 0.8662,
+      "step": 699
+    },
+    {
+      "epoch": 0.23801428085685142,
+      "grad_norm": 2.7094333692876837,
+      "learning_rate": 8.90964793470941e-06,
+      "loss": 0.8143,
+      "step": 700
+    },
+    {
+      "epoch": 0.2383543012580755,
+      "grad_norm": 2.7906165665573837,
+      "learning_rate": 8.906212245691755e-06,
+      "loss": 0.8905,
+      "step": 701
+    },
+    {
+      "epoch": 0.23869432165929955,
+      "grad_norm": 2.797092594783242,
+      "learning_rate": 8.902771816912521e-06,
+      "loss": 0.879,
+      "step": 702
+    },
+    {
+      "epoch": 0.23903434206052362,
+      "grad_norm": 2.5180099645451066,
+      "learning_rate": 8.899326652546292e-06,
+      "loss": 0.7547,
+      "step": 703
+    },
+    {
+      "epoch": 0.2393743624617477,
+      "grad_norm": 1.7769222125627648,
+      "learning_rate": 8.895876756773398e-06,
+      "loss": 0.9269,
+      "step": 704
+    },
+    {
+      "epoch": 0.23971438286297178,
+      "grad_norm": 1.8153973558094076,
+      "learning_rate": 8.89242213377991e-06,
+      "loss": 0.8157,
+      "step": 705
+    },
+    {
+      "epoch": 0.24005440326419586,
+      "grad_norm": 1.7008522196194176,
+      "learning_rate": 8.888962787757636e-06,
+      "loss": 0.8323,
+      "step": 706
+    },
+    {
+      "epoch": 0.24039442366541994,
+      "grad_norm": 3.122832621811037,
+      "learning_rate": 8.885498722904114e-06,
+      "loss": 0.8148,
+      "step": 707
+    },
+    {
+      "epoch": 0.240734444066644,
+      "grad_norm": 2.6291893036900045,
+      "learning_rate": 8.882029943422605e-06,
+      "loss": 0.8432,
+      "step": 708
+    },
+    {
+      "epoch": 0.24107446446786807,
+      "grad_norm": 2.273799903298918,
+      "learning_rate": 8.8785564535221e-06,
+      "loss": 0.8374,
+      "step": 709
+    },
+    {
+      "epoch": 0.24141448486909214,
+      "grad_norm": 1.8887964125341279,
+      "learning_rate": 8.875078257417294e-06,
+      "loss": 0.8395,
+      "step": 710
+    },
+    {
+      "epoch": 0.24175450527031622,
+      "grad_norm": 1.9167482586163092,
+      "learning_rate": 8.871595359328603e-06,
+      "loss": 0.8333,
+      "step": 711
+    },
+    {
+      "epoch": 0.2420945256715403,
+      "grad_norm": 2.0043550475154777,
+      "learning_rate": 8.868107763482137e-06,
+      "loss": 0.8465,
+      "step": 712
+    },
+    {
+      "epoch": 0.24243454607276438,
+      "grad_norm": 1.8537150982530552,
+      "learning_rate": 8.864615474109715e-06,
+      "loss": 0.7761,
+      "step": 713
+    },
+    {
+      "epoch": 0.24277456647398843,
+      "grad_norm": 2.2270421393649777,
+      "learning_rate": 8.861118495448847e-06,
+      "loss": 0.7535,
+      "step": 714
+    },
+    {
+      "epoch": 0.2431145868752125,
+      "grad_norm": 1.9397964001880972,
+      "learning_rate": 8.857616831742739e-06,
+      "loss": 0.751,
+      "step": 715
+    },
+    {
+      "epoch": 0.24345460727643659,
+      "grad_norm": 2.797785675978316,
+      "learning_rate": 8.854110487240275e-06,
+      "loss": 0.7928,
+      "step": 716
+    },
+    {
+      "epoch": 0.24379462767766066,
+      "grad_norm": 3.0971738774972604,
+      "learning_rate": 8.850599466196018e-06,
+      "loss": 0.7754,
+      "step": 717
+    },
+    {
+      "epoch": 0.24413464807888474,
+      "grad_norm": 3.154393286576707,
+      "learning_rate": 8.847083772870209e-06,
+      "loss": 0.7009,
+      "step": 718
+    },
+    {
+      "epoch": 0.24447466848010882,
+      "grad_norm": 2.605914049831787,
+      "learning_rate": 8.84356341152876e-06,
+      "loss": 0.7458,
+      "step": 719
+    },
+    {
+      "epoch": 0.24481468888133287,
+      "grad_norm": 2.3256897705301522,
+      "learning_rate": 8.840038386443243e-06,
+      "loss": 0.7355,
+      "step": 720
+    },
+    {
+      "epoch": 0.24515470928255695,
+      "grad_norm": 2.092998344309539,
+      "learning_rate": 8.836508701890892e-06,
+      "loss": 0.859,
+      "step": 721
+    },
+    {
+      "epoch": 0.24549472968378103,
+      "grad_norm": 1.8695733446918772,
+      "learning_rate": 8.832974362154592e-06,
+      "loss": 0.8425,
+      "step": 722
+    },
+    {
+      "epoch": 0.2458347500850051,
+      "grad_norm": 1.9623481565876253,
+      "learning_rate": 8.829435371522879e-06,
+      "loss": 0.7531,
+      "step": 723
+    },
+    {
+      "epoch": 0.24617477048622918,
+      "grad_norm": 1.7281382412711035,
+      "learning_rate": 8.82589173428993e-06,
+      "loss": 0.8975,
+      "step": 724
+    },
+    {
+      "epoch": 0.24651479088745323,
+      "grad_norm": 1.6071614173391668,
+      "learning_rate": 8.822343454755562e-06,
+      "loss": 0.8718,
+      "step": 725
+    },
+    {
+      "epoch": 0.2468548112886773,
+      "grad_norm": 1.9394291672670085,
+      "learning_rate": 8.818790537225224e-06,
+      "loss": 0.7458,
+      "step": 726
+    },
+    {
+      "epoch": 0.2471948316899014,
+      "grad_norm": 2.0539393000556343,
+      "learning_rate": 8.815232986009994e-06,
+      "loss": 0.8104,
+      "step": 727
+    },
+    {
+      "epoch": 0.24753485209112547,
+      "grad_norm": 2.047963973266287,
+      "learning_rate": 8.81167080542657e-06,
+      "loss": 0.8877,
+      "step": 728
+    },
+    {
+      "epoch": 0.24787487249234955,
+      "grad_norm": 3.0026729958135134,
+      "learning_rate": 8.80810399979727e-06,
+      "loss": 0.8977,
+      "step": 729
+    },
+    {
+      "epoch": 0.24821489289357362,
+      "grad_norm": 1.7334258967558374,
+      "learning_rate": 8.804532573450024e-06,
+      "loss": 0.7311,
+      "step": 730
+    },
+    {
+      "epoch": 0.24855491329479767,
+      "grad_norm": 2.2237717981886056,
+      "learning_rate": 8.800956530718365e-06,
+      "loss": 0.8934,
+      "step": 731
+    },
+    {
+      "epoch": 0.24889493369602175,
+      "grad_norm": 1.4722862638834173,
+      "learning_rate": 8.797375875941431e-06,
+      "loss": 0.7578,
+      "step": 732
+    },
+    {
+      "epoch": 0.24923495409724583,
+      "grad_norm": 2.076977010854489,
+      "learning_rate": 8.793790613463956e-06,
+      "loss": 0.8266,
+      "step": 733
+    },
+    {
+      "epoch": 0.2495749744984699,
+      "grad_norm": 1.9019007451038732,
+      "learning_rate": 8.790200747636261e-06,
+      "loss": 0.817,
+      "step": 734
+    },
+    {
+      "epoch": 0.249914994899694,
+      "grad_norm": 3.2523001391650603,
+      "learning_rate": 8.78660628281426e-06,
+      "loss": 0.823,
+      "step": 735
+    },
+    {
+      "epoch": 0.25025501530091804,
+      "grad_norm": 2.2606861111968706,
+      "learning_rate": 8.78300722335944e-06,
+      "loss": 0.7769,
+      "step": 736
+    },
+    {
+      "epoch": 0.25059503570214214,
+      "grad_norm": 2.4433629042102467,
+      "learning_rate": 8.77940357363887e-06,
+      "loss": 0.7904,
+      "step": 737
+    },
+    {
+      "epoch": 0.2509350561033662,
+      "grad_norm": 2.108341620407174,
+      "learning_rate": 8.77579533802518e-06,
+      "loss": 0.8316,
+      "step": 738
+    },
+    {
+      "epoch": 0.2512750765045903,
+      "grad_norm": 4.98477571143613,
+      "learning_rate": 8.772182520896573e-06,
+      "loss": 0.8266,
+      "step": 739
+    },
+    {
+      "epoch": 0.25161509690581435,
+      "grad_norm": 1.9399774055291894,
+      "learning_rate": 8.768565126636806e-06,
+      "loss": 0.8225,
+      "step": 740
+    },
+    {
+      "epoch": 0.2519551173070384,
+      "grad_norm": 2.594549449894867,
+      "learning_rate": 8.764943159635193e-06,
+      "loss": 0.7238,
+      "step": 741
+    },
+    {
+      "epoch": 0.2522951377082625,
+      "grad_norm": 2.6610598581449247,
+      "learning_rate": 8.761316624286593e-06,
+      "loss": 0.7797,
+      "step": 742
+    },
+    {
+      "epoch": 0.25263515810948656,
+      "grad_norm": 1.7706557977888584,
+      "learning_rate": 8.757685524991414e-06,
+      "loss": 0.8875,
+      "step": 743
+    },
+    {
+      "epoch": 0.25297517851071066,
+      "grad_norm": 2.367385917663463,
+      "learning_rate": 8.754049866155594e-06,
+      "loss": 0.8251,
+      "step": 744
+    },
+    {
+      "epoch": 0.2533151989119347,
+      "grad_norm": 2.0014441027718557,
+      "learning_rate": 8.750409652190609e-06,
+      "loss": 0.8519,
+      "step": 745
+    },
+    {
+      "epoch": 0.25365521931315876,
+      "grad_norm": 1.8473603569334116,
+      "learning_rate": 8.74676488751346e-06,
+      "loss": 0.8601,
+      "step": 746
+    },
+    {
+      "epoch": 0.25399523971438287,
+      "grad_norm": 2.477807792703976,
+      "learning_rate": 8.743115576546672e-06,
+      "loss": 0.9798,
+      "step": 747
+    },
+    {
+      "epoch": 0.2543352601156069,
+      "grad_norm": 2.1596648623116694,
+      "learning_rate": 8.739461723718286e-06,
+      "loss": 0.9241,
+      "step": 748
+    },
+    {
+      "epoch": 0.254675280516831,
+      "grad_norm": 2.271967660622451,
+      "learning_rate": 8.73580333346185e-06,
+      "loss": 0.9333,
+      "step": 749
+    },
+    {
+      "epoch": 0.2550153009180551,
+      "grad_norm": 1.89409525964846,
+      "learning_rate": 8.732140410216422e-06,
+      "loss": 0.9235,
+      "step": 750
+    },
+    {
+      "epoch": 0.2553553213192792,
+      "grad_norm": 1.9160319862426827,
+      "learning_rate": 8.72847295842656e-06,
+      "loss": 0.8362,
+      "step": 751
+    },
+    {
+      "epoch": 0.25569534172050323,
+      "grad_norm": 2.75041804529313,
+      "learning_rate": 8.724800982542313e-06,
+      "loss": 0.8281,
+      "step": 752
+    },
+    {
+      "epoch": 0.2560353621217273,
+      "grad_norm": 1.9660343049850402,
+      "learning_rate": 8.721124487019226e-06,
+      "loss": 0.8134,
+      "step": 753
+    },
+    {
+      "epoch": 0.2563753825229514,
+      "grad_norm": 1.7476864872494857,
+      "learning_rate": 8.717443476318322e-06,
+      "loss": 0.7963,
+      "step": 754
+    },
+    {
+      "epoch": 0.25671540292417544,
+      "grad_norm": 1.7181600952027278,
+      "learning_rate": 8.713757954906105e-06,
+      "loss": 0.6619,
+      "step": 755
+    },
+    {
+      "epoch": 0.25705542332539955,
+      "grad_norm": 4.689603340381868,
+      "learning_rate": 8.710067927254555e-06,
+      "loss": 0.8325,
+      "step": 756
+    },
+    {
+      "epoch": 0.2573954437266236,
+      "grad_norm": 2.0670743417962014,
+      "learning_rate": 8.706373397841114e-06,
+      "loss": 0.7841,
+      "step": 757
+    },
+    {
+      "epoch": 0.25773546412784765,
+      "grad_norm": 1.9345516631091482,
+      "learning_rate": 8.702674371148692e-06,
+      "loss": 0.7412,
+      "step": 758
+    },
+    {
+      "epoch": 0.25807548452907175,
+      "grad_norm": 2.3058554102539865,
+      "learning_rate": 8.698970851665652e-06,
+      "loss": 0.8672,
+      "step": 759
+    },
+    {
+      "epoch": 0.2584155049302958,
+      "grad_norm": 1.906875691115053,
+      "learning_rate": 8.695262843885812e-06,
+      "loss": 0.7907,
+      "step": 760
+    },
+    {
+      "epoch": 0.2587555253315199,
+      "grad_norm": 1.8081498930839859,
+      "learning_rate": 8.691550352308431e-06,
+      "loss": 0.7257,
+      "step": 761
+    },
+    {
+      "epoch": 0.25909554573274396,
+      "grad_norm": 2.0456832516321377,
+      "learning_rate": 8.687833381438215e-06,
+      "loss": 0.8767,
+      "step": 762
+    },
+    {
+      "epoch": 0.25943556613396807,
+      "grad_norm": 4.818955286864829,
+      "learning_rate": 8.684111935785299e-06,
+      "loss": 0.809,
+      "step": 763
+    },
+    {
+      "epoch": 0.2597755865351921,
+      "grad_norm": 1.6359696437957223,
+      "learning_rate": 8.680386019865253e-06,
+      "loss": 0.8736,
+      "step": 764
+    },
+    {
+      "epoch": 0.26011560693641617,
+      "grad_norm": 1.9275763227542202,
+      "learning_rate": 8.676655638199068e-06,
+      "loss": 0.7778,
+      "step": 765
+    },
+    {
+      "epoch": 0.26045562733764027,
+      "grad_norm": 1.5111168632740775,
+      "learning_rate": 8.67292079531315e-06,
+      "loss": 0.7518,
+      "step": 766
+    },
+    {
+      "epoch": 0.2607956477388643,
+      "grad_norm": 1.9868081030493614,
+      "learning_rate": 8.669181495739332e-06,
+      "loss": 0.876,
+      "step": 767
+    },
+    {
+      "epoch": 0.26113566814008843,
+      "grad_norm": 1.9632367709448835,
+      "learning_rate": 8.665437744014838e-06,
+      "loss": 0.7469,
+      "step": 768
+    },
+    {
+      "epoch": 0.2614756885413125,
+      "grad_norm": 2.658790741994479,
+      "learning_rate": 8.661689544682301e-06,
+      "loss": 0.8102,
+      "step": 769
+    },
+    {
+      "epoch": 0.26181570894253653,
+      "grad_norm": 1.9709245415214305,
+      "learning_rate": 8.657936902289756e-06,
+      "loss": 0.8966,
+      "step": 770
+    },
+    {
+      "epoch": 0.26215572934376064,
+      "grad_norm": 3.0701001889258515,
+      "learning_rate": 8.65417982139062e-06,
+      "loss": 0.9841,
+      "step": 771
+    },
+    {
+      "epoch": 0.2624957497449847,
+      "grad_norm": 4.54013202807214,
+      "learning_rate": 8.650418306543704e-06,
+      "loss": 0.8277,
+      "step": 772
+    },
+    {
+      "epoch": 0.2628357701462088,
+      "grad_norm": 1.6031790587096684,
+      "learning_rate": 8.646652362313193e-06,
+      "loss": 0.8168,
+      "step": 773
+    },
+    {
+      "epoch": 0.26317579054743284,
+      "grad_norm": 2.2725192507554857,
+      "learning_rate": 8.642881993268647e-06,
+      "loss": 0.8552,
+      "step": 774
+    },
+    {
+      "epoch": 0.2635158109486569,
+      "grad_norm": 1.9449496721499624,
+      "learning_rate": 8.639107203985e-06,
+      "loss": 0.8014,
+      "step": 775
+    },
+    {
+      "epoch": 0.263855831349881,
+      "grad_norm": 4.227240205226276,
+      "learning_rate": 8.635327999042543e-06,
+      "loss": 0.9003,
+      "step": 776
+    },
+    {
+      "epoch": 0.26419585175110505,
+      "grad_norm": 1.7770839323226375,
+      "learning_rate": 8.63154438302693e-06,
+      "loss": 0.8669,
+      "step": 777
+    },
+    {
+      "epoch": 0.26453587215232915,
+      "grad_norm": 4.668426873038303,
+      "learning_rate": 8.627756360529166e-06,
+      "loss": 0.861,
+      "step": 778
+    },
+    {
+      "epoch": 0.2648758925535532,
+      "grad_norm": 2.0048269343626663,
+      "learning_rate": 8.6239639361456e-06,
+      "loss": 0.7886,
+      "step": 779
+    },
+    {
+      "epoch": 0.2652159129547773,
+      "grad_norm": 2.793168861569981,
+      "learning_rate": 8.620167114477926e-06,
+      "loss": 0.8552,
+      "step": 780
+    },
+    {
+      "epoch": 0.26555593335600136,
+      "grad_norm": 1.832488110710129,
+      "learning_rate": 8.616365900133175e-06,
+      "loss": 0.8196,
+      "step": 781
+    },
+    {
+      "epoch": 0.2658959537572254,
+      "grad_norm": 4.023379445825273,
+      "learning_rate": 8.612560297723697e-06,
+      "loss": 0.7989,
+      "step": 782
+    },
+    {
+      "epoch": 0.2662359741584495,
+      "grad_norm": 2.332651801821611,
+      "learning_rate": 8.608750311867182e-06,
+      "loss": 0.7508,
+      "step": 783
+    },
+    {
+      "epoch": 0.26657599455967357,
+      "grad_norm": 6.696709614360918,
+      "learning_rate": 8.60493594718663e-06,
+      "loss": 0.8147,
+      "step": 784
+    },
+    {
+      "epoch": 0.2669160149608977,
+      "grad_norm": 1.8834108862545373,
+      "learning_rate": 8.601117208310351e-06,
+      "loss": 0.9059,
+      "step": 785
+    },
+    {
+      "epoch": 0.2672560353621217,
+      "grad_norm": 2.077771875039454,
+      "learning_rate": 8.597294099871974e-06,
+      "loss": 0.7673,
+      "step": 786
+    },
+    {
+      "epoch": 0.2675960557633458,
+      "grad_norm": 6.9046174801442675,
+      "learning_rate": 8.59346662651042e-06,
+      "loss": 0.8263,
+      "step": 787
+    },
+    {
+      "epoch": 0.2679360761645699,
+      "grad_norm": 2.77437081702886,
+      "learning_rate": 8.589634792869908e-06,
+      "loss": 0.8334,
+      "step": 788
+    },
+    {
+      "epoch": 0.26827609656579393,
+      "grad_norm": 2.308063436284089,
+      "learning_rate": 8.58579860359995e-06,
+      "loss": 0.8516,
+      "step": 789
+    },
+    {
+      "epoch": 0.26861611696701804,
+      "grad_norm": 1.5049106697591117,
+      "learning_rate": 8.581958063355344e-06,
+      "loss": 0.7896,
+      "step": 790
+    },
+    {
+      "epoch": 0.2689561373682421,
+      "grad_norm": 2.273632666328108,
+      "learning_rate": 8.578113176796165e-06,
+      "loss": 0.9209,
+      "step": 791
+    },
+    {
+      "epoch": 0.2692961577694662,
+      "grad_norm": 2.223581774886736,
+      "learning_rate": 8.574263948587762e-06,
+      "loss": 0.7586,
+      "step": 792
+    },
+    {
+      "epoch": 0.26963617817069024,
+      "grad_norm": 2.234320213681529,
+      "learning_rate": 8.570410383400754e-06,
+      "loss": 0.9106,
+      "step": 793
+    },
+    {
+      "epoch": 0.2699761985719143,
+      "grad_norm": 1.8073070815618781,
+      "learning_rate": 8.56655248591102e-06,
+      "loss": 0.8563,
+      "step": 794
+    },
+    {
+      "epoch": 0.2703162189731384,
+      "grad_norm": 2.216862016544106,
+      "learning_rate": 8.562690260799696e-06,
+      "loss": 0.8404,
+      "step": 795
+    },
+    {
+      "epoch": 0.27065623937436245,
+      "grad_norm": 3.3719814382949944,
+      "learning_rate": 8.558823712753171e-06,
+      "loss": 0.8676,
+      "step": 796
+    },
+    {
+      "epoch": 0.27099625977558656,
+      "grad_norm": 2.196633013307635,
+      "learning_rate": 8.554952846463081e-06,
+      "loss": 0.8648,
+      "step": 797
+    },
+    {
+      "epoch": 0.2713362801768106,
+      "grad_norm": 2.5165242077281595,
+      "learning_rate": 8.551077666626292e-06,
+      "loss": 0.7004,
+      "step": 798
+    },
+    {
+      "epoch": 0.27167630057803466,
+      "grad_norm": 1.7249240592256398,
+      "learning_rate": 8.54719817794492e-06,
+      "loss": 0.7373,
+      "step": 799
+    },
+    {
+      "epoch": 0.27201632097925876,
+      "grad_norm": 1.677617648661045,
+      "learning_rate": 8.543314385126296e-06,
+      "loss": 0.8333,
+      "step": 800
+    },
+    {
+      "epoch": 0.2723563413804828,
+      "grad_norm": 2.7130063773245223,
+      "learning_rate": 8.539426292882976e-06,
+      "loss": 0.7646,
+      "step": 801
+    },
+    {
+      "epoch": 0.2726963617817069,
+      "grad_norm": 3.5342476236653084,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.747,
+      "step": 802
+    },
+    {
+      "epoch": 0.27303638218293097,
+      "grad_norm": 2.6210750163057357,
+      "learning_rate": 8.531637228998569e-06,
+      "loss": 0.8778,
+      "step": 803
+    },
+    {
+      "epoch": 0.2733764025841551,
+      "grad_norm": 2.0347303242471853,
+      "learning_rate": 8.527736266808658e-06,
+      "loss": 0.769,
+      "step": 804
+    },
+    {
+      "epoch": 0.2737164229853791,
+      "grad_norm": 2.4027128478891915,
+      "learning_rate": 8.523831024096396e-06,
+      "loss": 0.8585,
+      "step": 805
+    },
+    {
+      "epoch": 0.2740564433866032,
+      "grad_norm": 1.8350089573984862,
+      "learning_rate": 8.519921505600368e-06,
+      "loss": 0.8113,
+      "step": 806
+    },
+    {
+      "epoch": 0.2743964637878273,
+      "grad_norm": 2.2589489484204908,
+      "learning_rate": 8.516007716064352e-06,
+      "loss": 0.8187,
+      "step": 807
+    },
+    {
+      "epoch": 0.27473648418905133,
+      "grad_norm": 1.858296448626169,
+      "learning_rate": 8.5120896602373e-06,
+      "loss": 0.9453,
+      "step": 808
+    },
+    {
+      "epoch": 0.27507650459027544,
+      "grad_norm": 2.355100329431512,
+      "learning_rate": 8.508167342873342e-06,
+      "loss": 0.8078,
+      "step": 809
+    },
+    {
+      "epoch": 0.2754165249914995,
+      "grad_norm": 1.9899667485080101,
+      "learning_rate": 8.504240768731787e-06,
+      "loss": 0.8554,
+      "step": 810
+    },
+    {
+      "epoch": 0.27575654539272354,
+      "grad_norm": 1.8787714779536697,
+      "learning_rate": 8.500309942577098e-06,
+      "loss": 0.8568,
+      "step": 811
+    },
+    {
+      "epoch": 0.27609656579394765,
+      "grad_norm": 2.068637813773282,
+      "learning_rate": 8.496374869178908e-06,
+      "loss": 0.848,
+      "step": 812
+    },
+    {
+      "epoch": 0.2764365861951717,
+      "grad_norm": 1.7791526260663866,
+      "learning_rate": 8.492435553311995e-06,
+      "loss": 0.8251,
+      "step": 813
+    },
+    {
+      "epoch": 0.2767766065963958,
+      "grad_norm": 1.8869896486229023,
+      "learning_rate": 8.48849199975629e-06,
+      "loss": 0.778,
+      "step": 814
+    },
+    {
+      "epoch": 0.27711662699761985,
+      "grad_norm": 2.2096789072585414,
+      "learning_rate": 8.484544213296864e-06,
+      "loss": 0.8346,
+      "step": 815
+    },
+    {
+      "epoch": 0.2774566473988439,
+      "grad_norm": 2.170554684057787,
+      "learning_rate": 8.480592198723922e-06,
+      "loss": 0.9079,
+      "step": 816
+    },
+    {
+      "epoch": 0.277796667800068,
+      "grad_norm": 1.9605343168925984,
+      "learning_rate": 8.476635960832805e-06,
+      "loss": 0.9024,
+      "step": 817
+    },
+    {
+      "epoch": 0.27813668820129206,
+      "grad_norm": 2.5300613494335638,
+      "learning_rate": 8.472675504423972e-06,
+      "loss": 0.7871,
+      "step": 818
+    },
+    {
+      "epoch": 0.27847670860251617,
+      "grad_norm": 2.8042077194040287,
+      "learning_rate": 8.468710834303007e-06,
+      "loss": 0.7785,
+      "step": 819
+    },
+    {
+      "epoch": 0.2788167290037402,
+      "grad_norm": 1.697967343522963,
+      "learning_rate": 8.464741955280603e-06,
+      "loss": 0.8535,
+      "step": 820
+    },
+    {
+      "epoch": 0.2791567494049643,
+      "grad_norm": 2.2291137255399303,
+      "learning_rate": 8.460768872172558e-06,
+      "loss": 0.8406,
+      "step": 821
+    },
+    {
+      "epoch": 0.27949676980618837,
+      "grad_norm": 1.919715258452679,
+      "learning_rate": 8.456791589799777e-06,
+      "loss": 0.8334,
+      "step": 822
+    },
+    {
+      "epoch": 0.2798367902074124,
+      "grad_norm": 1.9257825337008065,
+      "learning_rate": 8.45281011298826e-06,
+      "loss": 0.7674,
+      "step": 823
+    },
+    {
+      "epoch": 0.28017681060863653,
+      "grad_norm": 1.9491101623001321,
+      "learning_rate": 8.448824446569087e-06,
+      "loss": 0.8832,
+      "step": 824
+    },
+    {
+      "epoch": 0.2805168310098606,
+      "grad_norm": 1.870418445256147,
+      "learning_rate": 8.444834595378434e-06,
+      "loss": 0.8243,
+      "step": 825
+    },
+    {
+      "epoch": 0.2808568514110847,
+      "grad_norm": 15.898061851643817,
+      "learning_rate": 8.440840564257547e-06,
+      "loss": 0.9136,
+      "step": 826
+    },
+    {
+      "epoch": 0.28119687181230874,
+      "grad_norm": 4.472135160620738,
+      "learning_rate": 8.436842358052746e-06,
+      "loss": 0.7969,
+      "step": 827
+    },
+    {
+      "epoch": 0.2815368922135328,
+      "grad_norm": 1.761895889926857,
+      "learning_rate": 8.432839981615419e-06,
+      "loss": 0.7631,
+      "step": 828
+    },
+    {
+      "epoch": 0.2818769126147569,
+      "grad_norm": 2.3826293642920735,
+      "learning_rate": 8.428833439802012e-06,
+      "loss": 0.8369,
+      "step": 829
+    },
+    {
+      "epoch": 0.28221693301598094,
+      "grad_norm": 2.027771199922908,
+      "learning_rate": 8.424822737474023e-06,
+      "loss": 0.752,
+      "step": 830
+    },
+    {
+      "epoch": 0.28255695341720505,
+      "grad_norm": 2.5040444225047596,
+      "learning_rate": 8.420807879498002e-06,
+      "loss": 0.9132,
+      "step": 831
+    },
+    {
+      "epoch": 0.2828969738184291,
+      "grad_norm": 1.8159022349945535,
+      "learning_rate": 8.416788870745544e-06,
+      "loss": 0.8259,
+      "step": 832
+    },
+    {
+      "epoch": 0.2832369942196532,
+      "grad_norm": 2.620947283954682,
+      "learning_rate": 8.412765716093273e-06,
+      "loss": 0.8616,
+      "step": 833
+    },
+    {
+      "epoch": 0.28357701462087725,
+      "grad_norm": 1.985024515267911,
+      "learning_rate": 8.408738420422847e-06,
+      "loss": 0.8538,
+      "step": 834
+    },
+    {
+      "epoch": 0.2839170350221013,
+      "grad_norm": 1.7903361247800387,
+      "learning_rate": 8.40470698862095e-06,
+      "loss": 0.8478,
+      "step": 835
+    },
+    {
+      "epoch": 0.2842570554233254,
+      "grad_norm": 2.1338723210061974,
+      "learning_rate": 8.400671425579283e-06,
+      "loss": 0.7906,
+      "step": 836
+    },
+    {
+      "epoch": 0.28459707582454946,
+      "grad_norm": 1.9295239837495932,
+      "learning_rate": 8.396631736194563e-06,
+      "loss": 0.8481,
+      "step": 837
+    },
+    {
+      "epoch": 0.28493709622577357,
+      "grad_norm": 1.983034000002347,
+      "learning_rate": 8.39258792536851e-06,
+      "loss": 0.8847,
+      "step": 838
+    },
+    {
+      "epoch": 0.2852771166269976,
+      "grad_norm": 2.2803142495667035,
+      "learning_rate": 8.388539998007847e-06,
+      "loss": 0.9007,
+      "step": 839
+    },
+    {
+      "epoch": 0.28561713702822167,
+      "grad_norm": 2.7645969730004807,
+      "learning_rate": 8.384487959024293e-06,
+      "loss": 0.7356,
+      "step": 840
+    },
+    {
+      "epoch": 0.2859571574294458,
+      "grad_norm": 3.149778399577589,
+      "learning_rate": 8.380431813334548e-06,
+      "loss": 0.7855,
+      "step": 841
+    },
+    {
+      "epoch": 0.2862971778306698,
+      "grad_norm": 1.830554950516933,
+      "learning_rate": 8.37637156586031e-06,
+      "loss": 0.8831,
+      "step": 842
+    },
+    {
+      "epoch": 0.28663719823189393,
+      "grad_norm": 1.9793055946594367,
+      "learning_rate": 8.372307221528239e-06,
+      "loss": 0.8116,
+      "step": 843
+    },
+    {
+      "epoch": 0.286977218633118,
+      "grad_norm": 2.062954991657379,
+      "learning_rate": 8.368238785269976e-06,
+      "loss": 0.8563,
+      "step": 844
+    },
+    {
+      "epoch": 0.2873172390343421,
+      "grad_norm": 2.6403272415419834,
+      "learning_rate": 8.36416626202212e-06,
+      "loss": 0.8033,
+      "step": 845
+    },
+    {
+      "epoch": 0.28765725943556614,
+      "grad_norm": 2.1424920280150506,
+      "learning_rate": 8.360089656726238e-06,
+      "loss": 0.9417,
+      "step": 846
+    },
+    {
+      "epoch": 0.2879972798367902,
+      "grad_norm": 1.8495657823428482,
+      "learning_rate": 8.356008974328843e-06,
+      "loss": 0.8778,
+      "step": 847
+    },
+    {
+      "epoch": 0.2883373002380143,
+      "grad_norm": 1.964102211636596,
+      "learning_rate": 8.351924219781393e-06,
+      "loss": 0.8762,
+      "step": 848
+    },
+    {
+      "epoch": 0.28867732063923834,
+      "grad_norm": 5.329745330260959,
+      "learning_rate": 8.347835398040297e-06,
+      "loss": 0.8703,
+      "step": 849
+    },
+    {
+      "epoch": 0.28901734104046245,
+      "grad_norm": 6.872745204669564,
+      "learning_rate": 8.34374251406689e-06,
+      "loss": 0.9126,
+      "step": 850
+    },
+    {
+      "epoch": 0.2893573614416865,
+      "grad_norm": 3.021940703140289,
+      "learning_rate": 8.339645572827439e-06,
+      "loss": 0.8435,
+      "step": 851
+    },
+    {
+      "epoch": 0.28969738184291055,
+      "grad_norm": 1.7359605774084226,
+      "learning_rate": 8.335544579293138e-06,
+      "loss": 0.8956,
+      "step": 852
+    },
+    {
+      "epoch": 0.29003740224413466,
+      "grad_norm": 2.0495366847155645,
+      "learning_rate": 8.331439538440089e-06,
+      "loss": 0.8737,
+      "step": 853
+    },
+    {
+      "epoch": 0.2903774226453587,
+      "grad_norm": 1.7403332283198236,
+      "learning_rate": 8.327330455249316e-06,
+      "loss": 0.836,
+      "step": 854
+    },
+    {
+      "epoch": 0.2907174430465828,
+      "grad_norm": 1.765839927053788,
+      "learning_rate": 8.323217334706736e-06,
+      "loss": 0.7708,
+      "step": 855
+    },
+    {
+      "epoch": 0.29105746344780686,
+      "grad_norm": 2.561707026442392,
+      "learning_rate": 8.319100181803177e-06,
+      "loss": 0.8048,
+      "step": 856
+    },
+    {
+      "epoch": 0.29139748384903097,
+      "grad_norm": 1.893355173621553,
+      "learning_rate": 8.314979001534351e-06,
+      "loss": 0.8355,
+      "step": 857
+    },
+    {
+      "epoch": 0.291737504250255,
+      "grad_norm": 1.6626617126300058,
+      "learning_rate": 8.310853798900861e-06,
+      "loss": 0.8117,
+      "step": 858
+    },
+    {
+      "epoch": 0.29207752465147907,
+      "grad_norm": 2.2181642572480404,
+      "learning_rate": 8.306724578908187e-06,
+      "loss": 0.8809,
+      "step": 859
+    },
+    {
+      "epoch": 0.2924175450527032,
+      "grad_norm": 2.1608263105904237,
+      "learning_rate": 8.302591346566691e-06,
+      "loss": 0.9428,
+      "step": 860
+    },
+    {
+      "epoch": 0.2927575654539272,
+      "grad_norm": 2.2361606195465,
+      "learning_rate": 8.298454106891593e-06,
+      "loss": 0.8456,
+      "step": 861
+    },
+    {
+      "epoch": 0.29309758585515133,
+      "grad_norm": 2.161708305297874,
+      "learning_rate": 8.294312864902985e-06,
+      "loss": 0.7702,
+      "step": 862
+    },
+    {
+      "epoch": 0.2934376062563754,
+      "grad_norm": 1.9911157415642,
+      "learning_rate": 8.290167625625811e-06,
+      "loss": 0.8566,
+      "step": 863
+    },
+    {
+      "epoch": 0.29377762665759943,
+      "grad_norm": 2.204248530981356,
+      "learning_rate": 8.286018394089864e-06,
+      "loss": 0.785,
+      "step": 864
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 2.54115940873232,
+      "learning_rate": 8.281865175329783e-06,
+      "loss": 0.8669,
+      "step": 865
+    },
+    {
+      "epoch": 0.2944576674600476,
+      "grad_norm": 1.6985894379936504,
+      "learning_rate": 8.277707974385047e-06,
+      "loss": 0.8809,
+      "step": 866
+    },
+    {
+      "epoch": 0.2947976878612717,
+      "grad_norm": 1.9914761180754428,
+      "learning_rate": 8.273546796299962e-06,
+      "loss": 0.868,
+      "step": 867
+    },
+    {
+      "epoch": 0.29513770826249575,
+      "grad_norm": 1.9848345218936125,
+      "learning_rate": 8.269381646123666e-06,
+      "loss": 0.8266,
+      "step": 868
+    },
+    {
+      "epoch": 0.2954777286637198,
+      "grad_norm": 2.0822000899070674,
+      "learning_rate": 8.265212528910113e-06,
+      "loss": 0.9115,
+      "step": 869
+    },
+    {
+      "epoch": 0.2958177490649439,
+      "grad_norm": 1.9979737857871827,
+      "learning_rate": 8.261039449718068e-06,
+      "loss": 0.7968,
+      "step": 870
+    },
+    {
+      "epoch": 0.29615776946616795,
+      "grad_norm": 2.0218833894280532,
+      "learning_rate": 8.256862413611113e-06,
+      "loss": 0.8031,
+      "step": 871
+    },
+    {
+      "epoch": 0.29649778986739206,
+      "grad_norm": 2.006168397097048,
+      "learning_rate": 8.252681425657617e-06,
+      "loss": 0.8669,
+      "step": 872
+    },
+    {
+      "epoch": 0.2968378102686161,
+      "grad_norm": 2.2355962957542377,
+      "learning_rate": 8.248496490930753e-06,
+      "loss": 0.8274,
+      "step": 873
+    },
+    {
+      "epoch": 0.2971778306698402,
+      "grad_norm": 1.7904006258629988,
+      "learning_rate": 8.244307614508487e-06,
+      "loss": 0.7554,
+      "step": 874
+    },
+    {
+      "epoch": 0.29751785107106427,
+      "grad_norm": 2.0320747257565444,
+      "learning_rate": 8.240114801473558e-06,
+      "loss": 0.7651,
+      "step": 875
+    },
+    {
+      "epoch": 0.2978578714722883,
+      "grad_norm": 2.6665182314923412,
+      "learning_rate": 8.23591805691349e-06,
+      "loss": 0.8223,
+      "step": 876
+    },
+    {
+      "epoch": 0.2981978918735124,
+      "grad_norm": 1.8328124167485744,
+      "learning_rate": 8.23171738592057e-06,
+      "loss": 0.9082,
+      "step": 877
+    },
+    {
+      "epoch": 0.29853791227473647,
+      "grad_norm": 2.3699720185830757,
+      "learning_rate": 8.227512793591855e-06,
+      "loss": 0.9096,
+      "step": 878
+    },
+    {
+      "epoch": 0.2988779326759606,
+      "grad_norm": 1.9090567074503153,
+      "learning_rate": 8.223304285029159e-06,
+      "loss": 0.7705,
+      "step": 879
+    },
+    {
+      "epoch": 0.29921795307718463,
+      "grad_norm": 2.2190907511890368,
+      "learning_rate": 8.219091865339045e-06,
+      "loss": 0.7971,
+      "step": 880
+    },
+    {
+      "epoch": 0.2995579734784087,
+      "grad_norm": 2.253264189984432,
+      "learning_rate": 8.214875539632825e-06,
+      "loss": 0.7269,
+      "step": 881
+    },
+    {
+      "epoch": 0.2998979938796328,
+      "grad_norm": 2.0376522816245934,
+      "learning_rate": 8.21065531302655e-06,
+      "loss": 0.8329,
+      "step": 882
+    },
+    {
+      "epoch": 0.30023801428085684,
+      "grad_norm": 3.6223570808965007,
+      "learning_rate": 8.206431190641002e-06,
+      "loss": 0.8321,
+      "step": 883
+    },
+    {
+      "epoch": 0.30057803468208094,
+      "grad_norm": 3.8252210397062694,
+      "learning_rate": 8.202203177601693e-06,
+      "loss": 0.8164,
+      "step": 884
+    },
+    {
+      "epoch": 0.300918055083305,
+      "grad_norm": 1.776370483766253,
+      "learning_rate": 8.197971279038854e-06,
+      "loss": 0.8426,
+      "step": 885
+    },
+    {
+      "epoch": 0.3012580754845291,
+      "grad_norm": 2.385354175900532,
+      "learning_rate": 8.193735500087432e-06,
+      "loss": 0.7418,
+      "step": 886
+    },
+    {
+      "epoch": 0.30159809588575315,
+      "grad_norm": 1.6779774511826855,
+      "learning_rate": 8.189495845887083e-06,
+      "loss": 0.7568,
+      "step": 887
+    },
+    {
+      "epoch": 0.3019381162869772,
+      "grad_norm": 2.4863457173840544,
+      "learning_rate": 8.185252321582162e-06,
+      "loss": 0.8176,
+      "step": 888
+    },
+    {
+      "epoch": 0.3022781366882013,
+      "grad_norm": 4.0386363547881485,
+      "learning_rate": 8.18100493232172e-06,
+      "loss": 0.9485,
+      "step": 889
+    },
+    {
+      "epoch": 0.30261815708942535,
+      "grad_norm": 1.7173326803227138,
+      "learning_rate": 8.176753683259506e-06,
+      "loss": 0.7396,
+      "step": 890
+    },
+    {
+      "epoch": 0.30295817749064946,
+      "grad_norm": 2.3498732585677202,
+      "learning_rate": 8.172498579553939e-06,
+      "loss": 0.7183,
+      "step": 891
+    },
+    {
+      "epoch": 0.3032981978918735,
+      "grad_norm": 2.44411432379618,
+      "learning_rate": 8.168239626368126e-06,
+      "loss": 0.7807,
+      "step": 892
+    },
+    {
+      "epoch": 0.30363821829309756,
+      "grad_norm": 2.478384366935357,
+      "learning_rate": 8.16397682886984e-06,
+      "loss": 0.8315,
+      "step": 893
+    },
+    {
+      "epoch": 0.30397823869432167,
+      "grad_norm": 2.5774348066125894,
+      "learning_rate": 8.15971019223152e-06,
+      "loss": 0.8123,
+      "step": 894
+    },
+    {
+      "epoch": 0.3043182590955457,
+      "grad_norm": 1.7111325488707947,
+      "learning_rate": 8.155439721630265e-06,
+      "loss": 0.8263,
+      "step": 895
+    },
+    {
+      "epoch": 0.3046582794967698,
+      "grad_norm": 2.7448751652607553,
+      "learning_rate": 8.151165422247822e-06,
+      "loss": 0.8248,
+      "step": 896
+    },
+    {
+      "epoch": 0.3049982998979939,
+      "grad_norm": 2.1485101781392877,
+      "learning_rate": 8.146887299270585e-06,
+      "loss": 0.8035,
+      "step": 897
+    },
+    {
+      "epoch": 0.305338320299218,
+      "grad_norm": 1.776591747388704,
+      "learning_rate": 8.142605357889592e-06,
+      "loss": 0.8089,
+      "step": 898
+    },
+    {
+      "epoch": 0.30567834070044203,
+      "grad_norm": 3.2632560104335173,
+      "learning_rate": 8.13831960330051e-06,
+      "loss": 0.8202,
+      "step": 899
+    },
+    {
+      "epoch": 0.3060183611016661,
+      "grad_norm": 2.8739582827981347,
+      "learning_rate": 8.13403004070363e-06,
+      "loss": 0.9092,
+      "step": 900
+    },
+    {
+      "epoch": 0.3063583815028902,
+      "grad_norm": 2.1943148602179994,
+      "learning_rate": 8.129736675303873e-06,
+      "loss": 0.8322,
+      "step": 901
+    },
+    {
+      "epoch": 0.30669840190411424,
+      "grad_norm": 1.8531479477302115,
+      "learning_rate": 8.125439512310765e-06,
+      "loss": 0.7566,
+      "step": 902
+    },
+    {
+      "epoch": 0.30703842230533834,
+      "grad_norm": 1.7228875957473064,
+      "learning_rate": 8.121138556938444e-06,
+      "loss": 0.8078,
+      "step": 903
+    },
+    {
+      "epoch": 0.3073784427065624,
+      "grad_norm": 2.3898144961502745,
+      "learning_rate": 8.116833814405648e-06,
+      "loss": 0.8067,
+      "step": 904
+    },
+    {
+      "epoch": 0.30771846310778644,
+      "grad_norm": 1.802933531637354,
+      "learning_rate": 8.112525289935716e-06,
+      "loss": 0.7799,
+      "step": 905
+    },
+    {
+      "epoch": 0.30805848350901055,
+      "grad_norm": 2.5139323313707673,
+      "learning_rate": 8.108212988756568e-06,
+      "loss": 0.9037,
+      "step": 906
+    },
+    {
+      "epoch": 0.3083985039102346,
+      "grad_norm": 1.9251179471419289,
+      "learning_rate": 8.10389691610071e-06,
+      "loss": 0.8635,
+      "step": 907
+    },
+    {
+      "epoch": 0.3087385243114587,
+      "grad_norm": 2.193694058263112,
+      "learning_rate": 8.099577077205225e-06,
+      "loss": 0.8323,
+      "step": 908
+    },
+    {
+      "epoch": 0.30907854471268276,
+      "grad_norm": 2.0048196549770885,
+      "learning_rate": 8.095253477311765e-06,
+      "loss": 0.7756,
+      "step": 909
+    },
+    {
+      "epoch": 0.3094185651139068,
+      "grad_norm": 1.980230499045498,
+      "learning_rate": 8.090926121666547e-06,
+      "loss": 0.7977,
+      "step": 910
+    },
+    {
+      "epoch": 0.3097585855151309,
+      "grad_norm": 2.4883937143671564,
+      "learning_rate": 8.086595015520345e-06,
+      "loss": 0.8233,
+      "step": 911
+    },
+    {
+      "epoch": 0.31009860591635496,
+      "grad_norm": 2.1286197933597584,
+      "learning_rate": 8.08226016412848e-06,
+      "loss": 0.9729,
+      "step": 912
+    },
+    {
+      "epoch": 0.31043862631757907,
+      "grad_norm": 2.086146264470014,
+      "learning_rate": 8.07792157275082e-06,
+      "loss": 0.8914,
+      "step": 913
+    },
+    {
+      "epoch": 0.3107786467188031,
+      "grad_norm": 2.0772062973899423,
+      "learning_rate": 8.073579246651775e-06,
+      "loss": 0.945,
+      "step": 914
+    },
+    {
+      "epoch": 0.3111186671200272,
+      "grad_norm": 2.0695504749457143,
+      "learning_rate": 8.069233191100278e-06,
+      "loss": 0.8634,
+      "step": 915
+    },
+    {
+      "epoch": 0.3114586875212513,
+      "grad_norm": 2.0384939167389393,
+      "learning_rate": 8.064883411369799e-06,
+      "loss": 0.7785,
+      "step": 916
+    },
+    {
+      "epoch": 0.3117987079224753,
+      "grad_norm": 2.1885792727969138,
+      "learning_rate": 8.060529912738316e-06,
+      "loss": 0.8655,
+      "step": 917
+    },
+    {
+      "epoch": 0.31213872832369943,
+      "grad_norm": 2.1403065502782406,
+      "learning_rate": 8.056172700488324e-06,
+      "loss": 0.8965,
+      "step": 918
+    },
+    {
+      "epoch": 0.3124787487249235,
+      "grad_norm": 2.1339209234822647,
+      "learning_rate": 8.051811779906823e-06,
+      "loss": 0.7545,
+      "step": 919
+    },
+    {
+      "epoch": 0.3128187691261476,
+      "grad_norm": 1.8344955563510745,
+      "learning_rate": 8.047447156285314e-06,
+      "loss": 0.8804,
+      "step": 920
+    },
+    {
+      "epoch": 0.31315878952737164,
+      "grad_norm": 2.423743963035901,
+      "learning_rate": 8.043078834919792e-06,
+      "loss": 0.8068,
+      "step": 921
+    },
+    {
+      "epoch": 0.3134988099285957,
+      "grad_norm": 1.787239252052783,
+      "learning_rate": 8.038706821110738e-06,
+      "loss": 0.9271,
+      "step": 922
+    },
+    {
+      "epoch": 0.3138388303298198,
+      "grad_norm": 2.0084809664473684,
+      "learning_rate": 8.03433112016311e-06,
+      "loss": 0.8244,
+      "step": 923
+    },
+    {
+      "epoch": 0.31417885073104385,
+      "grad_norm": 2.0471692656266085,
+      "learning_rate": 8.029951737386345e-06,
+      "loss": 0.7478,
+      "step": 924
+    },
+    {
+      "epoch": 0.31451887113226795,
+      "grad_norm": 2.143535486588015,
+      "learning_rate": 8.025568678094346e-06,
+      "loss": 0.7579,
+      "step": 925
+    },
+    {
+      "epoch": 0.314858891533492,
+      "grad_norm": 1.835302258844517,
+      "learning_rate": 8.021181947605474e-06,
+      "loss": 0.771,
+      "step": 926
+    },
+    {
+      "epoch": 0.3151989119347161,
+      "grad_norm": 2.109820781590098,
+      "learning_rate": 8.016791551242548e-06,
+      "loss": 0.8985,
+      "step": 927
+    },
+    {
+      "epoch": 0.31553893233594016,
+      "grad_norm": 1.9465940317912624,
+      "learning_rate": 8.012397494332832e-06,
+      "loss": 0.9183,
+      "step": 928
+    },
+    {
+      "epoch": 0.3158789527371642,
+      "grad_norm": 1.9499330380070024,
+      "learning_rate": 8.00799978220804e-06,
+      "loss": 0.8158,
+      "step": 929
+    },
+    {
+      "epoch": 0.3162189731383883,
+      "grad_norm": 1.8510195207977735,
+      "learning_rate": 8.003598420204307e-06,
+      "loss": 0.8287,
+      "step": 930
+    },
+    {
+      "epoch": 0.31655899353961237,
+      "grad_norm": 2.2106314828328895,
+      "learning_rate": 7.99919341366221e-06,
+      "loss": 0.8159,
+      "step": 931
+    },
+    {
+      "epoch": 0.31689901394083647,
+      "grad_norm": 2.132400411746793,
+      "learning_rate": 7.994784767926743e-06,
+      "loss": 0.8686,
+      "step": 932
+    },
+    {
+      "epoch": 0.3172390343420605,
+      "grad_norm": 1.9835217982818234,
+      "learning_rate": 7.99037248834731e-06,
+      "loss": 0.7661,
+      "step": 933
+    },
+    {
+      "epoch": 0.3175790547432846,
+      "grad_norm": 1.8257541034651736,
+      "learning_rate": 7.985956580277738e-06,
+      "loss": 0.8968,
+      "step": 934
+    },
+    {
+      "epoch": 0.3179190751445087,
+      "grad_norm": 1.7806392840797605,
+      "learning_rate": 7.981537049076243e-06,
+      "loss": 0.8334,
+      "step": 935
+    },
+    {
+      "epoch": 0.31825909554573273,
+      "grad_norm": 1.946318339623462,
+      "learning_rate": 7.977113900105444e-06,
+      "loss": 0.8255,
+      "step": 936
+    },
+    {
+      "epoch": 0.31859911594695683,
+      "grad_norm": 2.5853186802795856,
+      "learning_rate": 7.972687138732352e-06,
+      "loss": 0.8669,
+      "step": 937
+    },
+    {
+      "epoch": 0.3189391363481809,
+      "grad_norm": 2.2143025012416913,
+      "learning_rate": 7.968256770328353e-06,
+      "loss": 0.7807,
+      "step": 938
+    },
+    {
+      "epoch": 0.319279156749405,
+      "grad_norm": 2.025109539800048,
+      "learning_rate": 7.96382280026922e-06,
+      "loss": 0.7668,
+      "step": 939
+    },
+    {
+      "epoch": 0.31961917715062904,
+      "grad_norm": 2.309014049320171,
+      "learning_rate": 7.959385233935087e-06,
+      "loss": 0.7586,
+      "step": 940
+    },
+    {
+      "epoch": 0.3199591975518531,
+      "grad_norm": 2.297274599241439,
+      "learning_rate": 7.954944076710457e-06,
+      "loss": 0.8962,
+      "step": 941
+    },
+    {
+      "epoch": 0.3202992179530772,
+      "grad_norm": 2.054541472734675,
+      "learning_rate": 7.95049933398419e-06,
+      "loss": 0.8203,
+      "step": 942
+    },
+    {
+      "epoch": 0.32063923835430125,
+      "grad_norm": 2.6215309252037873,
+      "learning_rate": 7.946051011149494e-06,
+      "loss": 0.8248,
+      "step": 943
+    },
+    {
+      "epoch": 0.32097925875552535,
+      "grad_norm": 2.711737030629169,
+      "learning_rate": 7.941599113603923e-06,
+      "loss": 0.8764,
+      "step": 944
+    },
+    {
+      "epoch": 0.3213192791567494,
+      "grad_norm": 2.2912349790018633,
+      "learning_rate": 7.937143646749367e-06,
+      "loss": 0.7335,
+      "step": 945
+    },
+    {
+      "epoch": 0.32165929955797345,
+      "grad_norm": 1.7908168705820537,
+      "learning_rate": 7.93268461599205e-06,
+      "loss": 0.8435,
+      "step": 946
+    },
+    {
+      "epoch": 0.32199931995919756,
+      "grad_norm": 4.946848177327164,
+      "learning_rate": 7.928222026742517e-06,
+      "loss": 0.8039,
+      "step": 947
+    },
+    {
+      "epoch": 0.3223393403604216,
+      "grad_norm": 2.0622056965822932,
+      "learning_rate": 7.923755884415634e-06,
+      "loss": 0.9067,
+      "step": 948
+    },
+    {
+      "epoch": 0.3226793607616457,
+      "grad_norm": 2.6667796950375715,
+      "learning_rate": 7.919286194430573e-06,
+      "loss": 0.7022,
+      "step": 949
+    },
+    {
+      "epoch": 0.32301938116286977,
+      "grad_norm": 1.9317481210027303,
+      "learning_rate": 7.914812962210819e-06,
+      "loss": 0.8264,
+      "step": 950
+    },
+    {
+      "epoch": 0.3233594015640938,
+      "grad_norm": 1.8020949558699724,
+      "learning_rate": 7.910336193184146e-06,
+      "loss": 0.7472,
+      "step": 951
+    },
+    {
+      "epoch": 0.3236994219653179,
+      "grad_norm": 1.9275396744198368,
+      "learning_rate": 7.905855892782625e-06,
+      "loss": 0.7309,
+      "step": 952
+    },
+    {
+      "epoch": 0.324039442366542,
+      "grad_norm": 1.8182685008859338,
+      "learning_rate": 7.901372066442615e-06,
+      "loss": 0.7625,
+      "step": 953
+    },
+    {
+      "epoch": 0.3243794627677661,
+      "grad_norm": 2.010794684369215,
+      "learning_rate": 7.89688471960474e-06,
+      "loss": 0.8687,
+      "step": 954
+    },
+    {
+      "epoch": 0.32471948316899013,
+      "grad_norm": 2.0540586819091997,
+      "learning_rate": 7.892393857713914e-06,
+      "loss": 0.8335,
+      "step": 955
+    },
+    {
+      "epoch": 0.32505950357021424,
+      "grad_norm": 2.325367093840662,
+      "learning_rate": 7.887899486219304e-06,
+      "loss": 0.783,
+      "step": 956
+    },
+    {
+      "epoch": 0.3253995239714383,
+      "grad_norm": 2.1873077332560156,
+      "learning_rate": 7.883401610574338e-06,
+      "loss": 0.8885,
+      "step": 957
+    },
+    {
+      "epoch": 0.32573954437266234,
+      "grad_norm": 2.177598891005239,
+      "learning_rate": 7.878900236236693e-06,
+      "loss": 0.763,
+      "step": 958
+    },
+    {
+      "epoch": 0.32607956477388644,
+      "grad_norm": 1.9372352426800454,
+      "learning_rate": 7.874395368668302e-06,
+      "loss": 0.8097,
+      "step": 959
+    },
+    {
+      "epoch": 0.3264195851751105,
+      "grad_norm": 1.8413146998215235,
+      "learning_rate": 7.869887013335324e-06,
+      "loss": 0.7083,
+      "step": 960
+    },
+    {
+      "epoch": 0.3267596055763346,
+      "grad_norm": 1.704925469624369,
+      "learning_rate": 7.865375175708158e-06,
+      "loss": 0.6822,
+      "step": 961
+    },
+    {
+      "epoch": 0.32709962597755865,
+      "grad_norm": 3.100437896445037,
+      "learning_rate": 7.860859861261423e-06,
+      "loss": 0.7932,
+      "step": 962
+    },
+    {
+      "epoch": 0.3274396463787827,
+      "grad_norm": 1.6975108379354062,
+      "learning_rate": 7.856341075473963e-06,
+      "loss": 0.7636,
+      "step": 963
+    },
+    {
+      "epoch": 0.3277796667800068,
+      "grad_norm": 1.7501297373086775,
+      "learning_rate": 7.851818823828828e-06,
+      "loss": 0.7754,
+      "step": 964
+    },
+    {
+      "epoch": 0.32811968718123086,
+      "grad_norm": 2.545140448497676,
+      "learning_rate": 7.847293111813276e-06,
+      "loss": 0.9082,
+      "step": 965
+    },
+    {
+      "epoch": 0.32845970758245496,
+      "grad_norm": 1.8008653754801884,
+      "learning_rate": 7.842763944918766e-06,
+      "loss": 0.83,
+      "step": 966
+    },
+    {
+      "epoch": 0.328799727983679,
+      "grad_norm": 2.6708500385698484,
+      "learning_rate": 7.838231328640945e-06,
+      "loss": 0.8698,
+      "step": 967
+    },
+    {
+      "epoch": 0.3291397483849031,
+      "grad_norm": 1.8002730480278057,
+      "learning_rate": 7.83369526847965e-06,
+      "loss": 0.8098,
+      "step": 968
+    },
+    {
+      "epoch": 0.32947976878612717,
+      "grad_norm": 2.032934392213337,
+      "learning_rate": 7.82915576993889e-06,
+      "loss": 0.841,
+      "step": 969
+    },
+    {
+      "epoch": 0.3298197891873512,
+      "grad_norm": 2.1268627260893473,
+      "learning_rate": 7.824612838526853e-06,
+      "loss": 0.8791,
+      "step": 970
+    },
+    {
+      "epoch": 0.3301598095885753,
+      "grad_norm": 1.9943092833782559,
+      "learning_rate": 7.82006647975589e-06,
+      "loss": 0.8746,
+      "step": 971
+    },
+    {
+      "epoch": 0.3304998299897994,
+      "grad_norm": 12.28140454443901,
+      "learning_rate": 7.81551669914251e-06,
+      "loss": 0.8952,
+      "step": 972
+    },
+    {
+      "epoch": 0.3308398503910235,
+      "grad_norm": 2.0328755642218055,
+      "learning_rate": 7.810963502207373e-06,
+      "loss": 0.7673,
+      "step": 973
+    },
+    {
+      "epoch": 0.33117987079224753,
+      "grad_norm": 1.7777944287029475,
+      "learning_rate": 7.806406894475286e-06,
+      "loss": 0.8826,
+      "step": 974
+    },
+    {
+      "epoch": 0.3315198911934716,
+      "grad_norm": 1.89053310342559,
+      "learning_rate": 7.801846881475199e-06,
+      "loss": 0.8305,
+      "step": 975
+    },
+    {
+      "epoch": 0.3318599115946957,
+      "grad_norm": 1.7916446125501533,
+      "learning_rate": 7.797283468740184e-06,
+      "loss": 0.7707,
+      "step": 976
+    },
+    {
+      "epoch": 0.33219993199591974,
+      "grad_norm": 1.583383697703338,
+      "learning_rate": 7.792716661807443e-06,
+      "loss": 0.7796,
+      "step": 977
+    },
+    {
+      "epoch": 0.33253995239714385,
+      "grad_norm": 2.4183910263443087,
+      "learning_rate": 7.788146466218301e-06,
+      "loss": 0.7304,
+      "step": 978
+    },
+    {
+      "epoch": 0.3328799727983679,
+      "grad_norm": 2.3658727843683907,
+      "learning_rate": 7.78357288751819e-06,
+      "loss": 0.7426,
+      "step": 979
+    },
+    {
+      "epoch": 0.333219993199592,
+      "grad_norm": 1.828116229828553,
+      "learning_rate": 7.778995931256646e-06,
+      "loss": 0.8078,
+      "step": 980
+    },
+    {
+      "epoch": 0.33356001360081605,
+      "grad_norm": 3.597726067768484,
+      "learning_rate": 7.774415602987304e-06,
+      "loss": 0.6857,
+      "step": 981
+    },
+    {
+      "epoch": 0.3339000340020401,
+      "grad_norm": 3.166303108924718,
+      "learning_rate": 7.769831908267896e-06,
+      "loss": 0.7904,
+      "step": 982
+    },
+    {
+      "epoch": 0.3342400544032642,
+      "grad_norm": 2.8933855627551304,
+      "learning_rate": 7.765244852660233e-06,
+      "loss": 0.8998,
+      "step": 983
+    },
+    {
+      "epoch": 0.33458007480448826,
+      "grad_norm": 1.9338110662323469,
+      "learning_rate": 7.760654441730202e-06,
+      "loss": 0.9007,
+      "step": 984
+    },
+    {
+      "epoch": 0.33492009520571236,
+      "grad_norm": 1.6591607534919344,
+      "learning_rate": 7.756060681047769e-06,
+      "loss": 0.8238,
+      "step": 985
+    },
+    {
+      "epoch": 0.3352601156069364,
+      "grad_norm": 3.243453226362474,
+      "learning_rate": 7.751463576186957e-06,
+      "loss": 0.7642,
+      "step": 986
+    },
+    {
+      "epoch": 0.33560013600816047,
+      "grad_norm": 2.7428152535139607,
+      "learning_rate": 7.746863132725856e-06,
+      "loss": 0.7282,
+      "step": 987
+    },
+    {
+      "epoch": 0.33594015640938457,
+      "grad_norm": 2.1236456873501144,
+      "learning_rate": 7.742259356246594e-06,
+      "loss": 0.7627,
+      "step": 988
+    },
+    {
+      "epoch": 0.3362801768106086,
+      "grad_norm": 1.6743079956506057,
+      "learning_rate": 7.737652252335356e-06,
+      "loss": 0.8406,
+      "step": 989
+    },
+    {
+      "epoch": 0.33662019721183273,
+      "grad_norm": 1.87268843836185,
+      "learning_rate": 7.733041826582357e-06,
+      "loss": 0.8455,
+      "step": 990
+    },
+    {
+      "epoch": 0.3369602176130568,
+      "grad_norm": 2.030664659551619,
+      "learning_rate": 7.728428084581844e-06,
+      "loss": 0.7965,
+      "step": 991
+    },
+    {
+      "epoch": 0.3373002380142809,
+      "grad_norm": 3.5752732157045215,
+      "learning_rate": 7.72381103193209e-06,
+      "loss": 0.7149,
+      "step": 992
+    },
+    {
+      "epoch": 0.33764025841550493,
+      "grad_norm": 2.8550170098863323,
+      "learning_rate": 7.719190674235383e-06,
+      "loss": 0.8308,
+      "step": 993
+    },
+    {
+      "epoch": 0.337980278816729,
+      "grad_norm": 2.590955515466377,
+      "learning_rate": 7.714567017098023e-06,
+      "loss": 0.902,
+      "step": 994
+    },
+    {
+      "epoch": 0.3383202992179531,
+      "grad_norm": 2.0307936229873467,
+      "learning_rate": 7.709940066130312e-06,
+      "loss": 0.8208,
+      "step": 995
+    },
+    {
+      "epoch": 0.33866031961917714,
+      "grad_norm": 1.4578787186534492,
+      "learning_rate": 7.705309826946547e-06,
+      "loss": 0.8051,
+      "step": 996
+    },
+    {
+      "epoch": 0.33900034002040125,
+      "grad_norm": 2.07437464133569,
+      "learning_rate": 7.70067630516502e-06,
+      "loss": 0.7707,
+      "step": 997
+    },
+    {
+      "epoch": 0.3393403604216253,
+      "grad_norm": 2.15530600462403,
+      "learning_rate": 7.696039506408001e-06,
+      "loss": 0.7745,
+      "step": 998
+    },
+    {
+      "epoch": 0.33968038082284935,
+      "grad_norm": 1.9609888444625139,
+      "learning_rate": 7.691399436301743e-06,
+      "loss": 0.7726,
+      "step": 999
+    },
+    {
+      "epoch": 0.34002040122407345,
+      "grad_norm": 1.7131939444884539,
+      "learning_rate": 7.686756100476458e-06,
+      "loss": 0.8546,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3403604216252975,
+      "grad_norm": 2.0565210562712397,
+      "learning_rate": 7.68210950456633e-06,
+      "loss": 0.7028,
+      "step": 1001
+    },
+    {
+      "epoch": 0.3407004420265216,
+      "grad_norm": 2.1703917499812664,
+      "learning_rate": 7.677459654209493e-06,
+      "loss": 0.83,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34104046242774566,
+      "grad_norm": 2.047040344782625,
+      "learning_rate": 7.672806555048034e-06,
+      "loss": 0.949,
+      "step": 1003
+    },
+    {
+      "epoch": 0.3413804828289697,
+      "grad_norm": 2.938822311667402,
+      "learning_rate": 7.66815021272798e-06,
+      "loss": 1.0131,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3417205032301938,
+      "grad_norm": 2.5006641954997257,
+      "learning_rate": 7.663490632899293e-06,
+      "loss": 0.7971,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34206052363141787,
+      "grad_norm": 2.1529578288278817,
+      "learning_rate": 7.658827821215863e-06,
+      "loss": 0.8715,
+      "step": 1006
+    },
+    {
+      "epoch": 0.342400544032642,
+      "grad_norm": 3.1033304500720442,
+      "learning_rate": 7.654161783335506e-06,
+      "loss": 0.7939,
+      "step": 1007
+    },
+    {
+      "epoch": 0.342740564433866,
+      "grad_norm": 2.484067747922095,
+      "learning_rate": 7.649492524919944e-06,
+      "loss": 0.8513,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34308058483509013,
+      "grad_norm": 2.1827092693175256,
+      "learning_rate": 7.644820051634813e-06,
+      "loss": 0.8447,
+      "step": 1009
+    },
+    {
+      "epoch": 0.3434206052363142,
+      "grad_norm": 1.9307893267754808,
+      "learning_rate": 7.64014436914965e-06,
+      "loss": 0.7384,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34376062563753823,
+      "grad_norm": 1.9144570087459527,
+      "learning_rate": 7.635465483137885e-06,
+      "loss": 0.8265,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34410064603876234,
+      "grad_norm": 1.9292779787587635,
+      "learning_rate": 7.63078339927683e-06,
+      "loss": 0.7725,
+      "step": 1012
+    },
+    {
+      "epoch": 0.3444406664399864,
+      "grad_norm": 2.532262819572805,
+      "learning_rate": 7.626098123247691e-06,
+      "loss": 0.8484,
+      "step": 1013
+    },
+    {
+      "epoch": 0.3447806868412105,
+      "grad_norm": 1.516985451302082,
+      "learning_rate": 7.621409660735531e-06,
+      "loss": 0.7265,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34512070724243454,
+      "grad_norm": 2.4387700274364335,
+      "learning_rate": 7.616718017429288e-06,
+      "loss": 0.8759,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3454607276436586,
+      "grad_norm": 1.7038256750677205,
+      "learning_rate": 7.612023199021759e-06,
+      "loss": 0.8983,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3458007480448827,
+      "grad_norm": 1.915028296278682,
+      "learning_rate": 7.607325211209593e-06,
+      "loss": 0.8207,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34614076844610675,
+      "grad_norm": 5.327535008237301,
+      "learning_rate": 7.6026240596932854e-06,
+      "loss": 0.8199,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34648078884733086,
+      "grad_norm": 2.0485629266374596,
+      "learning_rate": 7.597919750177168e-06,
+      "loss": 0.7972,
+      "step": 1019
+    },
+    {
+      "epoch": 0.3468208092485549,
+      "grad_norm": 2.5900475236404215,
+      "learning_rate": 7.593212288369408e-06,
+      "loss": 0.8245,
+      "step": 1020
+    },
+    {
+      "epoch": 0.347160829649779,
+      "grad_norm": 1.71454309402805,
+      "learning_rate": 7.588501679981997e-06,
+      "loss": 0.7877,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34750085005100306,
+      "grad_norm": 1.7173724399881714,
+      "learning_rate": 7.583787930730737e-06,
+      "loss": 0.8098,
+      "step": 1022
+    },
+    {
+      "epoch": 0.3478408704522271,
+      "grad_norm": 1.3352281724662016,
+      "learning_rate": 7.579071046335256e-06,
+      "loss": 0.8151,
+      "step": 1023
+    },
+    {
+      "epoch": 0.3481808908534512,
+      "grad_norm": 2.2809433135453565,
+      "learning_rate": 7.57435103251897e-06,
+      "loss": 0.8194,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34852091125467527,
+      "grad_norm": 1.776619177140561,
+      "learning_rate": 7.569627895009104e-06,
+      "loss": 0.7774,
+      "step": 1025
+    },
+    {
+      "epoch": 0.3488609316558994,
+      "grad_norm": 1.8490485181067193,
+      "learning_rate": 7.564901639536671e-06,
+      "loss": 0.7279,
+      "step": 1026
+    },
+    {
+      "epoch": 0.3492009520571234,
+      "grad_norm": 1.9994188958218952,
+      "learning_rate": 7.560172271836459e-06,
+      "loss": 0.8714,
+      "step": 1027
+    },
+    {
+      "epoch": 0.3495409724583475,
+      "grad_norm": 2.4785018885474903,
+      "learning_rate": 7.555439797647044e-06,
+      "loss": 0.7691,
+      "step": 1028
+    },
+    {
+      "epoch": 0.3498809928595716,
+      "grad_norm": 2.3412120228462427,
+      "learning_rate": 7.5507042227107655e-06,
+      "loss": 0.7584,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35022101326079563,
+      "grad_norm": 2.165022439020042,
+      "learning_rate": 7.545965552773724e-06,
+      "loss": 0.8901,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35056103366201974,
+      "grad_norm": 2.006216307009843,
+      "learning_rate": 7.54122379358578e-06,
+      "loss": 0.9007,
+      "step": 1031
+    },
+    {
+      "epoch": 0.3509010540632438,
+      "grad_norm": 4.543392686075566,
+      "learning_rate": 7.536478950900537e-06,
+      "loss": 0.8423,
+      "step": 1032
+    },
+    {
+      "epoch": 0.3512410744644679,
+      "grad_norm": 1.976663114216136,
+      "learning_rate": 7.531731030475345e-06,
+      "loss": 0.8405,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35158109486569195,
+      "grad_norm": 2.117430140673597,
+      "learning_rate": 7.526980038071288e-06,
+      "loss": 0.8223,
+      "step": 1034
+    },
+    {
+      "epoch": 0.351921115266916,
+      "grad_norm": 1.9728797359425245,
+      "learning_rate": 7.52222597945317e-06,
+      "loss": 0.9271,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3522611356681401,
+      "grad_norm": 3.175594333388401,
+      "learning_rate": 7.517468860389528e-06,
+      "loss": 0.6723,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35260115606936415,
+      "grad_norm": 1.9729371741404098,
+      "learning_rate": 7.512708686652603e-06,
+      "loss": 0.8227,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 1.4699410732834897,
+      "learning_rate": 7.507945464018347e-06,
+      "loss": 0.85,
+      "step": 1038
+    },
+    {
+      "epoch": 0.3532811968718123,
+      "grad_norm": 2.4025141879021743,
+      "learning_rate": 7.5031791982664085e-06,
+      "loss": 0.8379,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35362121727303636,
+      "grad_norm": 1.5826438634608264,
+      "learning_rate": 7.49840989518013e-06,
+      "loss": 0.8426,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35396123767426046,
+      "grad_norm": 1.8143996630470258,
+      "learning_rate": 7.493637560546539e-06,
+      "loss": 0.8218,
+      "step": 1041
+    },
+    {
+      "epoch": 0.3543012580754845,
+      "grad_norm": 2.9819746909856875,
+      "learning_rate": 7.488862200156345e-06,
+      "loss": 0.8116,
+      "step": 1042
+    },
+    {
+      "epoch": 0.3546412784767086,
+      "grad_norm": 1.6025956087863886,
+      "learning_rate": 7.484083819803922e-06,
+      "loss": 0.8996,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35498129887793267,
+      "grad_norm": 2.2569463838656394,
+      "learning_rate": 7.479302425287314e-06,
+      "loss": 0.793,
+      "step": 1044
+    },
+    {
+      "epoch": 0.3553213192791567,
+      "grad_norm": 1.7640317823457161,
+      "learning_rate": 7.4745180224082205e-06,
+      "loss": 0.9197,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35566133968038083,
+      "grad_norm": 2.0978448016759796,
+      "learning_rate": 7.469730616971992e-06,
+      "loss": 0.8864,
+      "step": 1046
+    },
+    {
+      "epoch": 0.3560013600816049,
+      "grad_norm": 1.815417919798777,
+      "learning_rate": 7.464940214787622e-06,
+      "loss": 0.8489,
+      "step": 1047
+    },
+    {
+      "epoch": 0.356341380482829,
+      "grad_norm": 2.5179528675012306,
+      "learning_rate": 7.4601468216677375e-06,
+      "loss": 0.7985,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35668140088405303,
+      "grad_norm": 2.0737400366436978,
+      "learning_rate": 7.455350443428598e-06,
+      "loss": 0.8621,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35702142128527714,
+      "grad_norm": 1.9500143741002205,
+      "learning_rate": 7.450551085890087e-06,
+      "loss": 0.7252,
+      "step": 1050
+    },
+    {
+      "epoch": 0.3573614416865012,
+      "grad_norm": 2.487299951341808,
+      "learning_rate": 7.445748754875697e-06,
+      "loss": 0.8219,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35770146208772524,
+      "grad_norm": 1.9293676887234388,
+      "learning_rate": 7.440943456212533e-06,
+      "loss": 0.8055,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35804148248894935,
+      "grad_norm": 1.630631399721771,
+      "learning_rate": 7.4361351957313e-06,
+      "loss": 0.8086,
+      "step": 1053
+    },
+    {
+      "epoch": 0.3583815028901734,
+      "grad_norm": 1.993396157634338,
+      "learning_rate": 7.431323979266296e-06,
+      "loss": 0.8358,
+      "step": 1054
+    },
+    {
+      "epoch": 0.3587215232913975,
+      "grad_norm": 2.1096399659613434,
+      "learning_rate": 7.4265098126554065e-06,
+      "loss": 0.6622,
+      "step": 1055
+    },
+    {
+      "epoch": 0.35906154369262155,
+      "grad_norm": 2.0957274147043568,
+      "learning_rate": 7.421692701740095e-06,
+      "loss": 0.8535,
+      "step": 1056
+    },
+    {
+      "epoch": 0.3594015640938456,
+      "grad_norm": 2.280331567819238,
+      "learning_rate": 7.416872652365401e-06,
+      "loss": 0.7613,
+      "step": 1057
+    },
+    {
+      "epoch": 0.3597415844950697,
+      "grad_norm": 1.7088762504236967,
+      "learning_rate": 7.412049670379927e-06,
+      "loss": 0.8598,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36008160489629376,
+      "grad_norm": 1.8455695752782821,
+      "learning_rate": 7.4072237616358356e-06,
+      "loss": 0.9097,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36042162529751787,
+      "grad_norm": 2.4513346866686483,
+      "learning_rate": 7.402394931988836e-06,
+      "loss": 0.8521,
+      "step": 1060
+    },
+    {
+      "epoch": 0.3607616456987419,
+      "grad_norm": 1.7993496451007047,
+      "learning_rate": 7.397563187298189e-06,
+      "loss": 0.8043,
+      "step": 1061
+    },
+    {
+      "epoch": 0.361101666099966,
+      "grad_norm": 2.5033019796675777,
+      "learning_rate": 7.392728533426687e-06,
+      "loss": 0.6882,
+      "step": 1062
+    },
+    {
+      "epoch": 0.3614416865011901,
+      "grad_norm": 1.6959114046452552,
+      "learning_rate": 7.387890976240655e-06,
+      "loss": 0.7466,
+      "step": 1063
+    },
+    {
+      "epoch": 0.3617817069024141,
+      "grad_norm": 1.7439198675611804,
+      "learning_rate": 7.383050521609938e-06,
+      "loss": 0.8818,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36212172730363823,
+      "grad_norm": 2.053598063714424,
+      "learning_rate": 7.378207175407899e-06,
+      "loss": 0.7483,
+      "step": 1065
+    },
+    {
+      "epoch": 0.3624617477048623,
+      "grad_norm": 2.781756547823472,
+      "learning_rate": 7.3733609435114096e-06,
+      "loss": 0.8445,
+      "step": 1066
+    },
+    {
+      "epoch": 0.3628017681060864,
+      "grad_norm": 2.09779855425177,
+      "learning_rate": 7.368511831800844e-06,
+      "loss": 0.851,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36314178850731044,
+      "grad_norm": 3.020910053469312,
+      "learning_rate": 7.363659846160066e-06,
+      "loss": 0.8892,
+      "step": 1068
+    },
+    {
+      "epoch": 0.3634818089085345,
+      "grad_norm": 2.4787490690903944,
+      "learning_rate": 7.358804992476432e-06,
+      "loss": 0.8432,
+      "step": 1069
+    },
+    {
+      "epoch": 0.3638218293097586,
+      "grad_norm": 2.8989448273348573,
+      "learning_rate": 7.353947276640776e-06,
+      "loss": 0.6853,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36416184971098264,
+      "grad_norm": 1.726759255900026,
+      "learning_rate": 7.349086704547408e-06,
+      "loss": 0.8831,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36450187011220675,
+      "grad_norm": 1.8399420040429915,
+      "learning_rate": 7.344223282094095e-06,
+      "loss": 0.7882,
+      "step": 1072
+    },
+    {
+      "epoch": 0.3648418905134308,
+      "grad_norm": 1.806073228885492,
+      "learning_rate": 7.3393570151820714e-06,
+      "loss": 0.8412,
+      "step": 1073
+    },
+    {
+      "epoch": 0.3651819109146549,
+      "grad_norm": 1.6710296660184638,
+      "learning_rate": 7.334487909716021e-06,
+      "loss": 0.9138,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36552193131587896,
+      "grad_norm": 2.1624300932227682,
+      "learning_rate": 7.329615971604071e-06,
+      "loss": 0.8227,
+      "step": 1075
+    },
+    {
+      "epoch": 0.365861951717103,
+      "grad_norm": 1.7009224007949533,
+      "learning_rate": 7.324741206757785e-06,
+      "loss": 0.8129,
+      "step": 1076
+    },
+    {
+      "epoch": 0.3662019721183271,
+      "grad_norm": 2.408114357062906,
+      "learning_rate": 7.3198636210921556e-06,
+      "loss": 0.8746,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36654199251955116,
+      "grad_norm": 1.963920816993082,
+      "learning_rate": 7.314983220525604e-06,
+      "loss": 0.8986,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36688201292077527,
+      "grad_norm": 2.625997972868134,
+      "learning_rate": 7.3101000109799616e-06,
+      "loss": 0.8639,
+      "step": 1079
+    },
+    {
+      "epoch": 0.3672220333219993,
+      "grad_norm": 2.089484971251662,
+      "learning_rate": 7.305213998380466e-06,
+      "loss": 0.9753,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36756205372322337,
+      "grad_norm": 1.879765672023159,
+      "learning_rate": 7.300325188655762e-06,
+      "loss": 0.8868,
+      "step": 1081
+    },
+    {
+      "epoch": 0.3679020741244475,
+      "grad_norm": 2.742242310525079,
+      "learning_rate": 7.295433587737885e-06,
+      "loss": 0.8266,
+      "step": 1082
+    },
+    {
+      "epoch": 0.3682420945256715,
+      "grad_norm": 2.110886719188207,
+      "learning_rate": 7.29053920156226e-06,
+      "loss": 0.8568,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36858211492689563,
+      "grad_norm": 1.6424790092242403,
+      "learning_rate": 7.285642036067687e-06,
+      "loss": 0.803,
+      "step": 1084
+    },
+    {
+      "epoch": 0.3689221353281197,
+      "grad_norm": 1.6786948404232462,
+      "learning_rate": 7.280742097196342e-06,
+      "loss": 0.7703,
+      "step": 1085
+    },
+    {
+      "epoch": 0.36926215572934373,
+      "grad_norm": 3.0606693350336154,
+      "learning_rate": 7.275839390893766e-06,
+      "loss": 0.7358,
+      "step": 1086
+    },
+    {
+      "epoch": 0.36960217613056784,
+      "grad_norm": 2.1162203963053146,
+      "learning_rate": 7.270933923108857e-06,
+      "loss": 0.7855,
+      "step": 1087
+    },
+    {
+      "epoch": 0.3699421965317919,
+      "grad_norm": 1.8016862422684354,
+      "learning_rate": 7.2660256997938635e-06,
+      "loss": 0.8466,
+      "step": 1088
+    },
+    {
+      "epoch": 0.370282216933016,
+      "grad_norm": 1.6606430113607151,
+      "learning_rate": 7.261114726904379e-06,
+      "loss": 0.8152,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37062223733424005,
+      "grad_norm": 2.761851717407974,
+      "learning_rate": 7.25620101039933e-06,
+      "loss": 0.9042,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37096225773546415,
+      "grad_norm": 1.9771042284445832,
+      "learning_rate": 7.2512845562409764e-06,
+      "loss": 0.7056,
+      "step": 1091
+    },
+    {
+      "epoch": 0.3713022781366882,
+      "grad_norm": 1.653746373879253,
+      "learning_rate": 7.246365370394896e-06,
+      "loss": 0.8067,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37164229853791225,
+      "grad_norm": 1.8811326242800515,
+      "learning_rate": 7.241443458829985e-06,
+      "loss": 0.9338,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37198231893913636,
+      "grad_norm": 2.7501158364403637,
+      "learning_rate": 7.236518827518444e-06,
+      "loss": 0.8809,
+      "step": 1094
+    },
+    {
+      "epoch": 0.3723223393403604,
+      "grad_norm": 2.5384080733406074,
+      "learning_rate": 7.231591482435777e-06,
+      "loss": 0.7763,
+      "step": 1095
+    },
+    {
+      "epoch": 0.3726623597415845,
+      "grad_norm": 2.228534473485439,
+      "learning_rate": 7.226661429560776e-06,
+      "loss": 0.7819,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37300238014280857,
+      "grad_norm": 1.9926490104103836,
+      "learning_rate": 7.221728674875522e-06,
+      "loss": 0.886,
+      "step": 1097
+    },
+    {
+      "epoch": 0.3733424005440326,
+      "grad_norm": 2.362100344429819,
+      "learning_rate": 7.216793224365373e-06,
+      "loss": 0.8621,
+      "step": 1098
+    },
+    {
+      "epoch": 0.3736824209452567,
+      "grad_norm": 2.0797006061886103,
+      "learning_rate": 7.2118550840189605e-06,
+      "loss": 0.8351,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37402244134648077,
+      "grad_norm": 1.9013645029122852,
+      "learning_rate": 7.206914259828177e-06,
+      "loss": 0.7349,
+      "step": 1100
+    },
+    {
+      "epoch": 0.3743624617477049,
+      "grad_norm": 2.096526511001182,
+      "learning_rate": 7.201970757788172e-06,
+      "loss": 0.8096,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37470248214892893,
+      "grad_norm": 1.9487792698034627,
+      "learning_rate": 7.197024583897345e-06,
+      "loss": 0.7688,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37504250255015303,
+      "grad_norm": 1.7616136379351248,
+      "learning_rate": 7.19207574415734e-06,
+      "loss": 0.877,
+      "step": 1103
+    },
+    {
+      "epoch": 0.3753825229513771,
+      "grad_norm": 2.216518466261683,
+      "learning_rate": 7.187124244573029e-06,
+      "loss": 0.7425,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37572254335260113,
+      "grad_norm": 1.9420595220834904,
+      "learning_rate": 7.182170091152518e-06,
+      "loss": 0.8859,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37606256375382524,
+      "grad_norm": 3.1843147726618892,
+      "learning_rate": 7.17721328990713e-06,
+      "loss": 0.7846,
+      "step": 1106
+    },
+    {
+      "epoch": 0.3764025841550493,
+      "grad_norm": 3.5406950144124134,
+      "learning_rate": 7.1722538468514015e-06,
+      "loss": 0.7288,
+      "step": 1107
+    },
+    {
+      "epoch": 0.3767426045562734,
+      "grad_norm": 1.9027244103388907,
+      "learning_rate": 7.167291768003075e-06,
+      "loss": 0.8369,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37708262495749745,
+      "grad_norm": 2.229506384560749,
+      "learning_rate": 7.162327059383089e-06,
+      "loss": 0.8006,
+      "step": 1109
+    },
+    {
+      "epoch": 0.3774226453587215,
+      "grad_norm": 2.483789935208878,
+      "learning_rate": 7.157359727015578e-06,
+      "loss": 0.8531,
+      "step": 1110
+    },
+    {
+      "epoch": 0.3777626657599456,
+      "grad_norm": 2.0952967393905597,
+      "learning_rate": 7.152389776927855e-06,
+      "loss": 0.7728,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37810268616116965,
+      "grad_norm": 2.8696175226987135,
+      "learning_rate": 7.147417215150411e-06,
+      "loss": 0.8312,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37844270656239376,
+      "grad_norm": 2.0770110520679115,
+      "learning_rate": 7.142442047716905e-06,
+      "loss": 0.7291,
+      "step": 1113
+    },
+    {
+      "epoch": 0.3787827269636178,
+      "grad_norm": 4.9493875756484345,
+      "learning_rate": 7.13746428066416e-06,
+      "loss": 0.8679,
+      "step": 1114
+    },
+    {
+      "epoch": 0.3791227473648419,
+      "grad_norm": 1.4736356648123277,
+      "learning_rate": 7.132483920032154e-06,
+      "loss": 0.8668,
+      "step": 1115
+    },
+    {
+      "epoch": 0.37946276776606597,
+      "grad_norm": 2.1871010916988407,
+      "learning_rate": 7.127500971864008e-06,
+      "loss": 0.8253,
+      "step": 1116
+    },
+    {
+      "epoch": 0.37980278816729,
+      "grad_norm": 2.7048260870598395,
+      "learning_rate": 7.122515442205985e-06,
+      "loss": 0.8072,
+      "step": 1117
+    },
+    {
+      "epoch": 0.3801428085685141,
+      "grad_norm": 1.6568599508148707,
+      "learning_rate": 7.117527337107481e-06,
+      "loss": 0.883,
+      "step": 1118
+    },
+    {
+      "epoch": 0.3804828289697382,
+      "grad_norm": 1.7744379748328984,
+      "learning_rate": 7.112536662621017e-06,
+      "loss": 0.8028,
+      "step": 1119
+    },
+    {
+      "epoch": 0.3808228493709623,
+      "grad_norm": 1.9075136504794952,
+      "learning_rate": 7.10754342480223e-06,
+      "loss": 0.7285,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38116286977218633,
+      "grad_norm": 1.9011664282216376,
+      "learning_rate": 7.102547629709867e-06,
+      "loss": 0.8961,
+      "step": 1121
+    },
+    {
+      "epoch": 0.3815028901734104,
+      "grad_norm": 3.3324355205350096,
+      "learning_rate": 7.097549283405782e-06,
+      "loss": 0.8518,
+      "step": 1122
+    },
+    {
+      "epoch": 0.3818429105746345,
+      "grad_norm": 1.7672003604856446,
+      "learning_rate": 7.092548391954919e-06,
+      "loss": 0.8808,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38218293097585854,
+      "grad_norm": 2.013141247061694,
+      "learning_rate": 7.087544961425317e-06,
+      "loss": 0.725,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38252295137708264,
+      "grad_norm": 1.985634648817913,
+      "learning_rate": 7.082538997888087e-06,
+      "loss": 0.75,
+      "step": 1125
+    },
+    {
+      "epoch": 0.3828629717783067,
+      "grad_norm": 1.9538380256881642,
+      "learning_rate": 7.077530507417423e-06,
+      "loss": 0.7453,
+      "step": 1126
+    },
+    {
+      "epoch": 0.3832029921795308,
+      "grad_norm": 1.4275684400693776,
+      "learning_rate": 7.07251949609058e-06,
+      "loss": 0.8418,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38354301258075485,
+      "grad_norm": 1.876464428838775,
+      "learning_rate": 7.067505969987869e-06,
+      "loss": 0.8856,
+      "step": 1128
+    },
+    {
+      "epoch": 0.3838830329819789,
+      "grad_norm": 2.4676650596958134,
+      "learning_rate": 7.06248993519266e-06,
+      "loss": 0.7442,
+      "step": 1129
+    },
+    {
+      "epoch": 0.384223053383203,
+      "grad_norm": 1.9772400387901186,
+      "learning_rate": 7.05747139779136e-06,
+      "loss": 0.7476,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38456307378442706,
+      "grad_norm": 1.7114896839198552,
+      "learning_rate": 7.0524503638734175e-06,
+      "loss": 0.7586,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38490309418565116,
+      "grad_norm": 1.979719268713037,
+      "learning_rate": 7.047426839531308e-06,
+      "loss": 0.7806,
+      "step": 1132
+    },
+    {
+      "epoch": 0.3852431145868752,
+      "grad_norm": 2.573830551015435,
+      "learning_rate": 7.04240083086053e-06,
+      "loss": 0.804,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38558313498809926,
+      "grad_norm": 2.0092570242754286,
+      "learning_rate": 7.037372343959592e-06,
+      "loss": 0.784,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38592315538932337,
+      "grad_norm": 1.7021704495221963,
+      "learning_rate": 7.032341384930018e-06,
+      "loss": 0.8437,
+      "step": 1135
+    },
+    {
+      "epoch": 0.3862631757905474,
+      "grad_norm": 2.005711746099124,
+      "learning_rate": 7.0273079598763236e-06,
+      "loss": 0.8695,
+      "step": 1136
+    },
+    {
+      "epoch": 0.3866031961917715,
+      "grad_norm": 3.7689109362135897,
+      "learning_rate": 7.022272074906021e-06,
+      "loss": 0.8509,
+      "step": 1137
+    },
+    {
+      "epoch": 0.3869432165929956,
+      "grad_norm": 2.040973796481224,
+      "learning_rate": 7.017233736129606e-06,
+      "loss": 0.7938,
+      "step": 1138
+    },
+    {
+      "epoch": 0.3872832369942196,
+      "grad_norm": 1.6275779088061617,
+      "learning_rate": 7.012192949660552e-06,
+      "loss": 0.7431,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38762325739544373,
+      "grad_norm": 3.289438767702211,
+      "learning_rate": 7.007149721615303e-06,
+      "loss": 0.8054,
+      "step": 1140
+    },
+    {
+      "epoch": 0.3879632777966678,
+      "grad_norm": 1.5783269725777442,
+      "learning_rate": 7.002104058113264e-06,
+      "loss": 0.7602,
+      "step": 1141
+    },
+    {
+      "epoch": 0.3883032981978919,
+      "grad_norm": 2.401935957785332,
+      "learning_rate": 6.997055965276796e-06,
+      "loss": 0.8494,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38864331859911594,
+      "grad_norm": 2.2155438256160447,
+      "learning_rate": 6.9920054492312086e-06,
+      "loss": 0.8322,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38898333900034004,
+      "grad_norm": 2.5303097895478426,
+      "learning_rate": 6.98695251610475e-06,
+      "loss": 0.7219,
+      "step": 1144
+    },
+    {
+      "epoch": 0.3893233594015641,
+      "grad_norm": 1.9492455847007166,
+      "learning_rate": 6.981897172028605e-06,
+      "loss": 0.7452,
+      "step": 1145
+    },
+    {
+      "epoch": 0.38966337980278815,
+      "grad_norm": 1.8888613087300397,
+      "learning_rate": 6.9768394231368765e-06,
+      "loss": 0.7079,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39000340020401225,
+      "grad_norm": 1.8614707671392436,
+      "learning_rate": 6.971779275566593e-06,
+      "loss": 0.8869,
+      "step": 1147
+    },
+    {
+      "epoch": 0.3903434206052363,
+      "grad_norm": 2.196139754184654,
+      "learning_rate": 6.96671673545769e-06,
+      "loss": 0.8182,
+      "step": 1148
+    },
+    {
+      "epoch": 0.3906834410064604,
+      "grad_norm": 2.015645605594665,
+      "learning_rate": 6.961651808953008e-06,
+      "loss": 0.788,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39102346140768446,
+      "grad_norm": 2.118358018353812,
+      "learning_rate": 6.956584502198278e-06,
+      "loss": 0.7944,
+      "step": 1150
+    },
+    {
+      "epoch": 0.3913634818089085,
+      "grad_norm": 2.5297802062767505,
+      "learning_rate": 6.9515148213421265e-06,
+      "loss": 0.7594,
+      "step": 1151
+    },
+    {
+      "epoch": 0.3917035022101326,
+      "grad_norm": 1.7235690561643628,
+      "learning_rate": 6.946442772536055e-06,
+      "loss": 0.8006,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39204352261135667,
+      "grad_norm": 2.0605815364417723,
+      "learning_rate": 6.941368361934442e-06,
+      "loss": 0.7571,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39238354301258077,
+      "grad_norm": 1.802107862787346,
+      "learning_rate": 6.9362915956945264e-06,
+      "loss": 0.7694,
+      "step": 1154
+    },
+    {
+      "epoch": 0.3927235634138048,
+      "grad_norm": 3.0394698418460373,
+      "learning_rate": 6.931212479976413e-06,
+      "loss": 0.8776,
+      "step": 1155
+    },
+    {
+      "epoch": 0.3930635838150289,
+      "grad_norm": 2.344468637665646,
+      "learning_rate": 6.9261310209430525e-06,
+      "loss": 0.7557,
+      "step": 1156
+    },
+    {
+      "epoch": 0.393403604216253,
+      "grad_norm": 2.1496119244458693,
+      "learning_rate": 6.921047224760239e-06,
+      "loss": 0.8703,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39374362461747703,
+      "grad_norm": 1.8022051744909748,
+      "learning_rate": 6.9159610975966044e-06,
+      "loss": 0.9033,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39408364501870113,
+      "grad_norm": 1.7062404002111817,
+      "learning_rate": 6.910872645623608e-06,
+      "loss": 0.8358,
+      "step": 1159
+    },
+    {
+      "epoch": 0.3944236654199252,
+      "grad_norm": 2.807991980988546,
+      "learning_rate": 6.905781875015529e-06,
+      "loss": 0.732,
+      "step": 1160
+    },
+    {
+      "epoch": 0.3947636858211493,
+      "grad_norm": 2.010080594688907,
+      "learning_rate": 6.900688791949463e-06,
+      "loss": 0.8275,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39510370622237334,
+      "grad_norm": 2.543059620006072,
+      "learning_rate": 6.895593402605308e-06,
+      "loss": 0.8672,
+      "step": 1162
+    },
+    {
+      "epoch": 0.3954437266235974,
+      "grad_norm": 2.323489162154479,
+      "learning_rate": 6.890495713165761e-06,
+      "loss": 0.8112,
+      "step": 1163
+    },
+    {
+      "epoch": 0.3957837470248215,
+      "grad_norm": 2.181543784406478,
+      "learning_rate": 6.885395729816313e-06,
+      "loss": 0.9508,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39612376742604555,
+      "grad_norm": 2.226593260830025,
+      "learning_rate": 6.880293458745237e-06,
+      "loss": 0.6444,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39646378782726965,
+      "grad_norm": 1.634156363859306,
+      "learning_rate": 6.87518890614358e-06,
+      "loss": 0.8444,
+      "step": 1166
+    },
+    {
+      "epoch": 0.3968038082284937,
+      "grad_norm": 2.5459595291596493,
+      "learning_rate": 6.870082078205158e-06,
+      "loss": 0.8996,
+      "step": 1167
+    },
+    {
+      "epoch": 0.3971438286297178,
+      "grad_norm": 3.9234543111468567,
+      "learning_rate": 6.86497298112655e-06,
+      "loss": 0.9022,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39748384903094186,
+      "grad_norm": 1.9173410435574607,
+      "learning_rate": 6.859861621107084e-06,
+      "loss": 0.8068,
+      "step": 1169
+    },
+    {
+      "epoch": 0.3978238694321659,
+      "grad_norm": 3.035334796056879,
+      "learning_rate": 6.85474800434884e-06,
+      "loss": 0.8787,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39816388983339,
+      "grad_norm": 1.996336800913808,
+      "learning_rate": 6.849632137056631e-06,
+      "loss": 0.8218,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39850391023461407,
+      "grad_norm": 2.4997036230867877,
+      "learning_rate": 6.844514025438003e-06,
+      "loss": 0.8944,
+      "step": 1172
+    },
+    {
+      "epoch": 0.3988439306358382,
+      "grad_norm": 2.058620980233996,
+      "learning_rate": 6.8393936757032255e-06,
+      "loss": 0.8202,
+      "step": 1173
+    },
+    {
+      "epoch": 0.3991839510370622,
+      "grad_norm": 2.0680642784567764,
+      "learning_rate": 6.834271094065284e-06,
+      "loss": 0.8465,
+      "step": 1174
+    },
+    {
+      "epoch": 0.3995239714382863,
+      "grad_norm": 1.8727199079644916,
+      "learning_rate": 6.82914628673987e-06,
+      "loss": 0.8527,
+      "step": 1175
+    },
+    {
+      "epoch": 0.3998639918395104,
+      "grad_norm": 1.590081791651527,
+      "learning_rate": 6.824019259945376e-06,
+      "loss": 0.8613,
+      "step": 1176
+    },
+    {
+      "epoch": 0.40020401224073443,
+      "grad_norm": 2.0260284044840278,
+      "learning_rate": 6.818890019902891e-06,
+      "loss": 0.7465,
+      "step": 1177
+    },
+    {
+      "epoch": 0.40054403264195854,
+      "grad_norm": 1.8398479387280182,
+      "learning_rate": 6.813758572836187e-06,
+      "loss": 0.7806,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4008840530431826,
+      "grad_norm": 2.105917280765373,
+      "learning_rate": 6.808624924971711e-06,
+      "loss": 0.7141,
+      "step": 1179
+    },
+    {
+      "epoch": 0.40122407344440664,
+      "grad_norm": 1.9634355488466153,
+      "learning_rate": 6.803489082538586e-06,
+      "loss": 0.8055,
+      "step": 1180
+    },
+    {
+      "epoch": 0.40156409384563074,
+      "grad_norm": 1.8526489425120056,
+      "learning_rate": 6.798351051768597e-06,
+      "loss": 0.8832,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4019041142468548,
+      "grad_norm": 2.3356716947930316,
+      "learning_rate": 6.79321083889618e-06,
+      "loss": 0.7484,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4022441346480789,
+      "grad_norm": 2.0837152056282964,
+      "learning_rate": 6.788068450158422e-06,
+      "loss": 0.718,
+      "step": 1183
+    },
+    {
+      "epoch": 0.40258415504930295,
+      "grad_norm": 1.8999186943331179,
+      "learning_rate": 6.78292389179505e-06,
+      "loss": 0.7811,
+      "step": 1184
+    },
+    {
+      "epoch": 0.40292417545052706,
+      "grad_norm": 1.6658111499434904,
+      "learning_rate": 6.777777170048423e-06,
+      "loss": 0.8201,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4032641958517511,
+      "grad_norm": 1.680917916696707,
+      "learning_rate": 6.772628291163527e-06,
+      "loss": 0.807,
+      "step": 1186
+    },
+    {
+      "epoch": 0.40360421625297516,
+      "grad_norm": 1.7407790818800217,
+      "learning_rate": 6.76747726138796e-06,
+      "loss": 0.8313,
+      "step": 1187
+    },
+    {
+      "epoch": 0.40394423665419926,
+      "grad_norm": 1.5617043074412897,
+      "learning_rate": 6.762324086971936e-06,
+      "loss": 0.9455,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4042842570554233,
+      "grad_norm": 4.146596153131383,
+      "learning_rate": 6.75716877416827e-06,
+      "loss": 0.7997,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4046242774566474,
+      "grad_norm": 2.0617653843196884,
+      "learning_rate": 6.752011329232369e-06,
+      "loss": 0.8153,
+      "step": 1190
+    },
+    {
+      "epoch": 0.40496429785787147,
+      "grad_norm": 2.076737017009885,
+      "learning_rate": 6.746851758422228e-06,
+      "loss": 0.8002,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4053043182590955,
+      "grad_norm": 1.813071998464279,
+      "learning_rate": 6.741690067998423e-06,
+      "loss": 0.8347,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4056443386603196,
+      "grad_norm": 1.9981326256931067,
+      "learning_rate": 6.736526264224101e-06,
+      "loss": 0.9294,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4059843590615437,
+      "grad_norm": 1.8827868546011934,
+      "learning_rate": 6.731360353364975e-06,
+      "loss": 0.867,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4063243794627678,
+      "grad_norm": 1.8243418052617972,
+      "learning_rate": 6.726192341689311e-06,
+      "loss": 0.8223,
+      "step": 1195
+    },
+    {
+      "epoch": 0.40666439986399183,
+      "grad_norm": 1.770350872789149,
+      "learning_rate": 6.721022235467926e-06,
+      "loss": 0.8619,
+      "step": 1196
+    },
+    {
+      "epoch": 0.40700442026521594,
+      "grad_norm": 1.8639174786964454,
+      "learning_rate": 6.7158500409741815e-06,
+      "loss": 0.9201,
+      "step": 1197
+    },
+    {
+      "epoch": 0.40734444066644,
+      "grad_norm": 2.0254058760681803,
+      "learning_rate": 6.710675764483968e-06,
+      "loss": 0.7695,
+      "step": 1198
+    },
+    {
+      "epoch": 0.40768446106766404,
+      "grad_norm": 1.760831957768078,
+      "learning_rate": 6.7054994122757046e-06,
+      "loss": 0.819,
+      "step": 1199
+    },
+    {
+      "epoch": 0.40802448146888814,
+      "grad_norm": 2.3849603429891997,
+      "learning_rate": 6.700320990630329e-06,
+      "loss": 0.8816,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4083645018701122,
+      "grad_norm": 3.24791611718312,
+      "learning_rate": 6.69514050583129e-06,
+      "loss": 0.9234,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4087045222713363,
+      "grad_norm": 1.7471455004845486,
+      "learning_rate": 6.689957964164539e-06,
+      "loss": 0.7623,
+      "step": 1202
+    },
+    {
+      "epoch": 0.40904454267256035,
+      "grad_norm": 2.785514967554598,
+      "learning_rate": 6.684773371918526e-06,
+      "loss": 0.7937,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4093845630737844,
+      "grad_norm": 1.8030855427647954,
+      "learning_rate": 6.679586735384184e-06,
+      "loss": 0.8442,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4097245834750085,
+      "grad_norm": 3.721825112571208,
+      "learning_rate": 6.674398060854931e-06,
+      "loss": 0.7539,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41006460387623256,
+      "grad_norm": 1.80798373103189,
+      "learning_rate": 6.669207354626657e-06,
+      "loss": 0.8992,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41040462427745666,
+      "grad_norm": 2.0935781969101352,
+      "learning_rate": 6.664014622997717e-06,
+      "loss": 0.8665,
+      "step": 1207
+    },
+    {
+      "epoch": 0.4107446446786807,
+      "grad_norm": 1.77781409293517,
+      "learning_rate": 6.65881987226892e-06,
+      "loss": 0.9314,
+      "step": 1208
+    },
+    {
+      "epoch": 0.4110846650799048,
+      "grad_norm": 2.0253755293143105,
+      "learning_rate": 6.65362310874353e-06,
+      "loss": 0.8807,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41142468548112887,
+      "grad_norm": 3.8985316382602444,
+      "learning_rate": 6.648424338727254e-06,
+      "loss": 0.7557,
+      "step": 1210
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 2.9177556160015317,
+      "learning_rate": 6.643223568528228e-06,
+      "loss": 0.7773,
+      "step": 1211
+    },
+    {
+      "epoch": 0.412104726283577,
+      "grad_norm": 1.7364907122378959,
+      "learning_rate": 6.638020804457017e-06,
+      "loss": 0.7708,
+      "step": 1212
+    },
+    {
+      "epoch": 0.4124447466848011,
+      "grad_norm": 2.5902233868083107,
+      "learning_rate": 6.632816052826611e-06,
+      "loss": 0.8803,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4127847670860252,
+      "grad_norm": 2.196962080702439,
+      "learning_rate": 6.627609319952404e-06,
+      "loss": 0.8413,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41312478748724923,
+      "grad_norm": 1.7517859998624659,
+      "learning_rate": 6.622400612152199e-06,
+      "loss": 0.7581,
+      "step": 1215
+    },
+    {
+      "epoch": 0.4134648078884733,
+      "grad_norm": 2.725187748694304,
+      "learning_rate": 6.617189935746191e-06,
+      "loss": 0.8616,
+      "step": 1216
+    },
+    {
+      "epoch": 0.4138048282896974,
+      "grad_norm": 2.186121004596113,
+      "learning_rate": 6.6119772970569686e-06,
+      "loss": 0.867,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41414484869092144,
+      "grad_norm": 1.7550877153581408,
+      "learning_rate": 6.606762702409499e-06,
+      "loss": 0.8189,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41448486909214555,
+      "grad_norm": 1.911639257053032,
+      "learning_rate": 6.60154615813112e-06,
+      "loss": 0.8447,
+      "step": 1219
+    },
+    {
+      "epoch": 0.4148248894933696,
+      "grad_norm": 2.2359482219667566,
+      "learning_rate": 6.596327670551541e-06,
+      "loss": 0.7098,
+      "step": 1220
+    },
+    {
+      "epoch": 0.4151649098945937,
+      "grad_norm": 1.8270789562811462,
+      "learning_rate": 6.591107246002825e-06,
+      "loss": 0.8242,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41550493029581775,
+      "grad_norm": 1.4119348213396106,
+      "learning_rate": 6.585884890819388e-06,
+      "loss": 0.9456,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4158449506970418,
+      "grad_norm": 2.4831333073302115,
+      "learning_rate": 6.5806606113379855e-06,
+      "loss": 0.843,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4161849710982659,
+      "grad_norm": 2.0195834149272907,
+      "learning_rate": 6.57543441389771e-06,
+      "loss": 0.7616,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41652499149948996,
+      "grad_norm": 1.8560990970772,
+      "learning_rate": 6.570206304839979e-06,
+      "loss": 0.772,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41686501190071407,
+      "grad_norm": 3.4777502532844364,
+      "learning_rate": 6.564976290508535e-06,
+      "loss": 0.7694,
+      "step": 1226
+    },
+    {
+      "epoch": 0.4172050323019381,
+      "grad_norm": 2.1141524879270337,
+      "learning_rate": 6.559744377249426e-06,
+      "loss": 0.8212,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41754505270316217,
+      "grad_norm": 1.9995986640884873,
+      "learning_rate": 6.554510571411009e-06,
+      "loss": 0.7696,
+      "step": 1228
+    },
+    {
+      "epoch": 0.4178850731043863,
+      "grad_norm": 4.388048606323549,
+      "learning_rate": 6.549274879343932e-06,
+      "loss": 0.8521,
+      "step": 1229
+    },
+    {
+      "epoch": 0.4182250935056103,
+      "grad_norm": 2.512566436457032,
+      "learning_rate": 6.54403730740114e-06,
+      "loss": 0.8375,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41856511390683443,
+      "grad_norm": 2.022875278093896,
+      "learning_rate": 6.53879786193785e-06,
+      "loss": 0.8287,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4189051343080585,
+      "grad_norm": 2.1514715191325338,
+      "learning_rate": 6.533556549311557e-06,
+      "loss": 0.8248,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41924515470928253,
+      "grad_norm": 1.903140049004442,
+      "learning_rate": 6.52831337588202e-06,
+      "loss": 0.8134,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41958517511050664,
+      "grad_norm": 3.9459515916276815,
+      "learning_rate": 6.52306834801126e-06,
+      "loss": 0.8674,
+      "step": 1234
+    },
+    {
+      "epoch": 0.4199251955117307,
+      "grad_norm": 1.9243723260268402,
+      "learning_rate": 6.517821472063543e-06,
+      "loss": 0.8009,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4202652159129548,
+      "grad_norm": 2.1622315952308484,
+      "learning_rate": 6.51257275440538e-06,
+      "loss": 0.8302,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42060523631417884,
+      "grad_norm": 2.246688116810149,
+      "learning_rate": 6.507322201405515e-06,
+      "loss": 0.8518,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42094525671540295,
+      "grad_norm": 1.7866795660553714,
+      "learning_rate": 6.502069819434921e-06,
+      "loss": 0.7996,
+      "step": 1238
+    },
+    {
+      "epoch": 0.421285277116627,
+      "grad_norm": 2.483859305848629,
+      "learning_rate": 6.496815614866792e-06,
+      "loss": 0.798,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42162529751785105,
+      "grad_norm": 2.6273097801827743,
+      "learning_rate": 6.491559594076526e-06,
+      "loss": 0.7717,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42196531791907516,
+      "grad_norm": 1.9705933045927748,
+      "learning_rate": 6.486301763441732e-06,
+      "loss": 0.8437,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4223053383202992,
+      "grad_norm": 1.6987794156789002,
+      "learning_rate": 6.4810421293422124e-06,
+      "loss": 0.7846,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4226453587215233,
+      "grad_norm": 1.7952873310252566,
+      "learning_rate": 6.475780698159959e-06,
+      "loss": 0.8228,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42298537912274736,
+      "grad_norm": 2.3781383897255357,
+      "learning_rate": 6.470517476279143e-06,
+      "loss": 0.9275,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4233253995239714,
+      "grad_norm": 1.7388598039077234,
+      "learning_rate": 6.465252470086109e-06,
+      "loss": 0.7543,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4236654199251955,
+      "grad_norm": 3.847734904824095,
+      "learning_rate": 6.459985685969365e-06,
+      "loss": 0.7801,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42400544032641957,
+      "grad_norm": 2.0015735326442123,
+      "learning_rate": 6.454717130319583e-06,
+      "loss": 0.8404,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4243454607276437,
+      "grad_norm": 1.676873018907222,
+      "learning_rate": 6.449446809529573e-06,
+      "loss": 0.7616,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4246854811288677,
+      "grad_norm": 1.9476313763300463,
+      "learning_rate": 6.444174729994295e-06,
+      "loss": 0.8572,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42502550153009183,
+      "grad_norm": 3.039778353662543,
+      "learning_rate": 6.438900898110843e-06,
+      "loss": 0.6842,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4253655219313159,
+      "grad_norm": 2.212503519532677,
+      "learning_rate": 6.433625320278435e-06,
+      "loss": 0.7895,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42570554233253993,
+      "grad_norm": 3.696419960495062,
+      "learning_rate": 6.4283480028984065e-06,
+      "loss": 0.7889,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42604556273376404,
+      "grad_norm": 1.8504613276269528,
+      "learning_rate": 6.423068952374208e-06,
+      "loss": 0.6952,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4263855831349881,
+      "grad_norm": 2.011583690777513,
+      "learning_rate": 6.4177881751113854e-06,
+      "loss": 0.7343,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4267256035362122,
+      "grad_norm": 1.9658764564932893,
+      "learning_rate": 6.412505677517592e-06,
+      "loss": 0.8955,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42706562393743625,
+      "grad_norm": 1.7771550054870846,
+      "learning_rate": 6.4072214660025555e-06,
+      "loss": 0.794,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4274056443386603,
+      "grad_norm": 2.170566029231825,
+      "learning_rate": 6.401935546978091e-06,
+      "loss": 0.8307,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4277456647398844,
+      "grad_norm": 2.795606918335267,
+      "learning_rate": 6.396647926858082e-06,
+      "loss": 0.7408,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42808568514110845,
+      "grad_norm": 1.9857919239067945,
+      "learning_rate": 6.391358612058479e-06,
+      "loss": 0.7435,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42842570554233256,
+      "grad_norm": 4.183650214686305,
+      "learning_rate": 6.386067608997286e-06,
+      "loss": 0.8171,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4287657259435566,
+      "grad_norm": 2.2866953692840517,
+      "learning_rate": 6.3807749240945594e-06,
+      "loss": 0.9282,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4291057463447807,
+      "grad_norm": 1.723522832107471,
+      "learning_rate": 6.375480563772391e-06,
+      "loss": 0.8644,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42944576674600476,
+      "grad_norm": 1.9376266728840439,
+      "learning_rate": 6.3701845344549105e-06,
+      "loss": 0.7975,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4297857871472288,
+      "grad_norm": 2.0320058786563884,
+      "learning_rate": 6.3648868425682695e-06,
+      "loss": 0.8404,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4301258075484529,
+      "grad_norm": 1.833242518517846,
+      "learning_rate": 6.359587494540638e-06,
+      "loss": 0.9727,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43046582794967697,
+      "grad_norm": 2.0656866508193237,
+      "learning_rate": 6.354286496802195e-06,
+      "loss": 0.7088,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4308058483509011,
+      "grad_norm": 2.333059652715563,
+      "learning_rate": 6.348983855785122e-06,
+      "loss": 0.7784,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4311458687521251,
+      "grad_norm": 1.8812951830256721,
+      "learning_rate": 6.343679577923596e-06,
+      "loss": 0.8082,
+      "step": 1268
+    },
+    {
+      "epoch": 0.4314858891533492,
+      "grad_norm": 2.2757541276249103,
+      "learning_rate": 6.338373669653777e-06,
+      "loss": 0.8048,
+      "step": 1269
+    },
+    {
+      "epoch": 0.4318259095545733,
+      "grad_norm": 1.7616508457249394,
+      "learning_rate": 6.333066137413803e-06,
+      "loss": 0.7967,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43216592995579733,
+      "grad_norm": 2.5839639191285912,
+      "learning_rate": 6.327756987643788e-06,
+      "loss": 0.8475,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43250595035702144,
+      "grad_norm": 1.983977392970035,
+      "learning_rate": 6.322446226785803e-06,
+      "loss": 0.7688,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4328459707582455,
+      "grad_norm": 1.920278640728409,
+      "learning_rate": 6.317133861283876e-06,
+      "loss": 0.8112,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43318599115946954,
+      "grad_norm": 2.175785687192607,
+      "learning_rate": 6.311819897583981e-06,
+      "loss": 0.8807,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43352601156069365,
+      "grad_norm": 2.1329693868490156,
+      "learning_rate": 6.306504342134032e-06,
+      "loss": 0.7646,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4338660319619177,
+      "grad_norm": 2.7197062834600643,
+      "learning_rate": 6.301187201383876e-06,
+      "loss": 0.8924,
+      "step": 1276
+    },
+    {
+      "epoch": 0.4342060523631418,
+      "grad_norm": 2.0900606207430976,
+      "learning_rate": 6.295868481785281e-06,
+      "loss": 0.8063,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43454607276436585,
+      "grad_norm": 1.7320405000379613,
+      "learning_rate": 6.290548189791932e-06,
+      "loss": 0.7871,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43488609316558996,
+      "grad_norm": 3.246424780026875,
+      "learning_rate": 6.285226331859423e-06,
+      "loss": 0.7022,
+      "step": 1279
+    },
+    {
+      "epoch": 0.435226113566814,
+      "grad_norm": 3.2299354049530558,
+      "learning_rate": 6.279902914445246e-06,
+      "loss": 0.8512,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43556613396803806,
+      "grad_norm": 1.7974856940773503,
+      "learning_rate": 6.274577944008785e-06,
+      "loss": 0.7445,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43590615436926217,
+      "grad_norm": 1.7902884652613178,
+      "learning_rate": 6.26925142701131e-06,
+      "loss": 0.7549,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4362461747704862,
+      "grad_norm": 1.7575937886544872,
+      "learning_rate": 6.263923369915968e-06,
+      "loss": 0.7033,
+      "step": 1283
+    },
+    {
+      "epoch": 0.4365861951717103,
+      "grad_norm": 2.003652418101978,
+      "learning_rate": 6.258593779187774e-06,
+      "loss": 0.7226,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4369262155729344,
+      "grad_norm": 1.586460473657157,
+      "learning_rate": 6.2532626612936035e-06,
+      "loss": 0.7918,
+      "step": 1285
+    },
+    {
+      "epoch": 0.4372662359741584,
+      "grad_norm": 1.7790120795884707,
+      "learning_rate": 6.247930022702184e-06,
+      "loss": 0.7426,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43760625637538253,
+      "grad_norm": 1.4845861438510766,
+      "learning_rate": 6.242595869884093e-06,
+      "loss": 0.75,
+      "step": 1287
+    },
+    {
+      "epoch": 0.4379462767766066,
+      "grad_norm": 2.586395804452128,
+      "learning_rate": 6.237260209311738e-06,
+      "loss": 0.7247,
+      "step": 1288
+    },
+    {
+      "epoch": 0.4382862971778307,
+      "grad_norm": 1.7488826727955935,
+      "learning_rate": 6.231923047459362e-06,
+      "loss": 0.7819,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43862631757905474,
+      "grad_norm": 2.7476753864785306,
+      "learning_rate": 6.2265843908030255e-06,
+      "loss": 0.8755,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43896633798027884,
+      "grad_norm": 2.04843978154324,
+      "learning_rate": 6.2212442458206065e-06,
+      "loss": 0.845,
+      "step": 1291
+    },
+    {
+      "epoch": 0.4393063583815029,
+      "grad_norm": 2.056402370171357,
+      "learning_rate": 6.215902618991789e-06,
+      "loss": 0.6932,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43964637878272694,
+      "grad_norm": 2.3809614347105814,
+      "learning_rate": 6.21055951679805e-06,
+      "loss": 0.8301,
+      "step": 1293
+    },
+    {
+      "epoch": 0.43998639918395105,
+      "grad_norm": 2.216045492126213,
+      "learning_rate": 6.20521494572266e-06,
+      "loss": 0.8544,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4403264195851751,
+      "grad_norm": 1.7506779190930466,
+      "learning_rate": 6.1998689122506765e-06,
+      "loss": 0.8289,
+      "step": 1295
+    },
+    {
+      "epoch": 0.4406664399863992,
+      "grad_norm": 1.904093376273434,
+      "learning_rate": 6.19452142286892e-06,
+      "loss": 0.7709,
+      "step": 1296
+    },
+    {
+      "epoch": 0.44100646038762326,
+      "grad_norm": 2.7089730576700664,
+      "learning_rate": 6.1891724840659895e-06,
+      "loss": 0.8263,
+      "step": 1297
+    },
+    {
+      "epoch": 0.4413464807888473,
+      "grad_norm": 1.7011604570765408,
+      "learning_rate": 6.183822102332234e-06,
+      "loss": 0.7318,
+      "step": 1298
+    },
+    {
+      "epoch": 0.4416865011900714,
+      "grad_norm": 1.7063499626312666,
+      "learning_rate": 6.17847028415976e-06,
+      "loss": 0.7845,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44202652159129546,
+      "grad_norm": 2.5512420126857998,
+      "learning_rate": 6.1731170360424116e-06,
+      "loss": 0.8297,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44236654199251957,
+      "grad_norm": 2.3308930708868614,
+      "learning_rate": 6.1677623644757715e-06,
+      "loss": 0.7281,
+      "step": 1301
+    },
+    {
+      "epoch": 0.4427065623937436,
+      "grad_norm": 1.9814345181679105,
+      "learning_rate": 6.162406275957147e-06,
+      "loss": 0.6841,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4430465827949677,
+      "grad_norm": 1.8308664790589082,
+      "learning_rate": 6.157048776985568e-06,
+      "loss": 0.7597,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4433866031961918,
+      "grad_norm": 1.9450975948070095,
+      "learning_rate": 6.151689874061773e-06,
+      "loss": 0.8809,
+      "step": 1304
+    },
+    {
+      "epoch": 0.4437266235974158,
+      "grad_norm": 2.783986050727857,
+      "learning_rate": 6.1463295736882045e-06,
+      "loss": 0.7678,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44406664399863993,
+      "grad_norm": 2.654115532064757,
+      "learning_rate": 6.140967882369001e-06,
+      "loss": 0.7656,
+      "step": 1306
+    },
+    {
+      "epoch": 0.444406664399864,
+      "grad_norm": 2.1154238353595938,
+      "learning_rate": 6.135604806609988e-06,
+      "loss": 0.7393,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4447466848010881,
+      "grad_norm": 1.8316694323729121,
+      "learning_rate": 6.130240352918675e-06,
+      "loss": 0.7955,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44508670520231214,
+      "grad_norm": 1.7923597848780743,
+      "learning_rate": 6.1248745278042375e-06,
+      "loss": 0.7902,
+      "step": 1309
+    },
+    {
+      "epoch": 0.4454267256035362,
+      "grad_norm": 2.1469603822475882,
+      "learning_rate": 6.119507337777517e-06,
+      "loss": 0.8111,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4457667460047603,
+      "grad_norm": 3.204925579816377,
+      "learning_rate": 6.114138789351015e-06,
+      "loss": 0.898,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44610676640598435,
+      "grad_norm": 2.4583222181061735,
+      "learning_rate": 6.108768889038875e-06,
+      "loss": 0.8401,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44644678680720845,
+      "grad_norm": 1.6306721028852273,
+      "learning_rate": 6.103397643356888e-06,
+      "loss": 0.8261,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4467868072084325,
+      "grad_norm": 1.6997011922153806,
+      "learning_rate": 6.098025058822467e-06,
+      "loss": 0.8157,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44712682760965655,
+      "grad_norm": 1.894873083473829,
+      "learning_rate": 6.092651141954663e-06,
+      "loss": 0.818,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44746684801088066,
+      "grad_norm": 1.8152984091344468,
+      "learning_rate": 6.087275899274132e-06,
+      "loss": 0.846,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4478068684121047,
+      "grad_norm": 1.8804247156065967,
+      "learning_rate": 6.081899337303148e-06,
+      "loss": 0.8775,
+      "step": 1317
+    },
+    {
+      "epoch": 0.4481468888133288,
+      "grad_norm": 2.2811320840226874,
+      "learning_rate": 6.076521462565575e-06,
+      "loss": 0.8405,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44848690921455286,
+      "grad_norm": 1.9014030398430317,
+      "learning_rate": 6.071142281586883e-06,
+      "loss": 0.6665,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44882692961577697,
+      "grad_norm": 1.7598473537668629,
+      "learning_rate": 6.0657618008941135e-06,
+      "loss": 0.8114,
+      "step": 1320
+    },
+    {
+      "epoch": 0.449166950017001,
+      "grad_norm": 1.701348824209668,
+      "learning_rate": 6.060380027015897e-06,
+      "loss": 0.9063,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44950697041822507,
+      "grad_norm": 2.0151604743954192,
+      "learning_rate": 6.054996966482425e-06,
+      "loss": 0.7727,
+      "step": 1322
+    },
+    {
+      "epoch": 0.4498469908194492,
+      "grad_norm": 1.8240997834179458,
+      "learning_rate": 6.049612625825454e-06,
+      "loss": 0.6151,
+      "step": 1323
+    },
+    {
+      "epoch": 0.4501870112206732,
+      "grad_norm": 1.7621003478299089,
+      "learning_rate": 6.044227011578292e-06,
+      "loss": 0.8248,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45052703162189733,
+      "grad_norm": 2.3449253350037647,
+      "learning_rate": 6.038840130275795e-06,
+      "loss": 0.8094,
+      "step": 1325
+    },
+    {
+      "epoch": 0.4508670520231214,
+      "grad_norm": 2.186940711291871,
+      "learning_rate": 6.033451988454352e-06,
+      "loss": 0.8526,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45120707242434543,
+      "grad_norm": 1.8098545794134784,
+      "learning_rate": 6.0280625926518865e-06,
+      "loss": 0.8167,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45154709282556954,
+      "grad_norm": 2.11938751373049,
+      "learning_rate": 6.02267194940784e-06,
+      "loss": 0.8615,
+      "step": 1328
+    },
+    {
+      "epoch": 0.4518871132267936,
+      "grad_norm": 5.798776280397436,
+      "learning_rate": 6.0172800652631706e-06,
+      "loss": 0.8126,
+      "step": 1329
+    },
+    {
+      "epoch": 0.4522271336280177,
+      "grad_norm": 1.9362659319180398,
+      "learning_rate": 6.011886946760337e-06,
+      "loss": 0.8515,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45256715402924175,
+      "grad_norm": 1.9240359926336854,
+      "learning_rate": 6.006492600443301e-06,
+      "loss": 0.795,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45290717443046585,
+      "grad_norm": 2.6312493040402223,
+      "learning_rate": 6.001097032857513e-06,
+      "loss": 0.9005,
+      "step": 1332
+    },
+    {
+      "epoch": 0.4532471948316899,
+      "grad_norm": 2.4588002159411975,
+      "learning_rate": 5.995700250549903e-06,
+      "loss": 0.9122,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45358721523291395,
+      "grad_norm": 1.8102464748281866,
+      "learning_rate": 5.990302260068877e-06,
+      "loss": 0.7861,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45392723563413806,
+      "grad_norm": 2.4499379979673654,
+      "learning_rate": 5.9849030679643075e-06,
+      "loss": 0.8793,
+      "step": 1335
+    },
+    {
+      "epoch": 0.4542672560353621,
+      "grad_norm": 1.643369045844105,
+      "learning_rate": 5.97950268078752e-06,
+      "loss": 0.9176,
+      "step": 1336
+    },
+    {
+      "epoch": 0.4546072764365862,
+      "grad_norm": 2.94241862218008,
+      "learning_rate": 5.9741011050913e-06,
+      "loss": 0.7631,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45494729683781027,
+      "grad_norm": 2.260745557245212,
+      "learning_rate": 5.968698347429864e-06,
+      "loss": 0.8574,
+      "step": 1338
+    },
+    {
+      "epoch": 0.4552873172390343,
+      "grad_norm": 1.9189222697412902,
+      "learning_rate": 5.96329441435887e-06,
+      "loss": 0.8627,
+      "step": 1339
+    },
+    {
+      "epoch": 0.4556273376402584,
+      "grad_norm": 2.3145970729902725,
+      "learning_rate": 5.9578893124354e-06,
+      "loss": 0.8203,
+      "step": 1340
+    },
+    {
+      "epoch": 0.4559673580414825,
+      "grad_norm": 1.7097699263826742,
+      "learning_rate": 5.9524830482179565e-06,
+      "loss": 0.8143,
+      "step": 1341
+    },
+    {
+      "epoch": 0.4563073784427066,
+      "grad_norm": 1.7644511385918709,
+      "learning_rate": 5.9470756282664455e-06,
+      "loss": 0.8428,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45664739884393063,
+      "grad_norm": 2.0834007860288835,
+      "learning_rate": 5.941667059142184e-06,
+      "loss": 0.8975,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45698741924515474,
+      "grad_norm": 1.9264637214129008,
+      "learning_rate": 5.936257347407877e-06,
+      "loss": 0.7147,
+      "step": 1344
+    },
+    {
+      "epoch": 0.4573274396463788,
+      "grad_norm": 1.7108620469281093,
+      "learning_rate": 5.9308464996276195e-06,
+      "loss": 0.8773,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45766746004760284,
+      "grad_norm": 1.8740444271707064,
+      "learning_rate": 5.925434522366884e-06,
+      "loss": 0.8765,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45800748044882694,
+      "grad_norm": 2.9159889860260706,
+      "learning_rate": 5.920021422192512e-06,
+      "loss": 0.7429,
+      "step": 1347
+    },
+    {
+      "epoch": 0.458347500850051,
+      "grad_norm": 2.2683984119230964,
+      "learning_rate": 5.914607205672711e-06,
+      "loss": 0.8265,
+      "step": 1348
+    },
+    {
+      "epoch": 0.4586875212512751,
+      "grad_norm": 2.229706808343337,
+      "learning_rate": 5.909191879377041e-06,
+      "loss": 0.8355,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45902754165249915,
+      "grad_norm": 2.284917272323551,
+      "learning_rate": 5.903775449876406e-06,
+      "loss": 0.706,
+      "step": 1350
+    },
+    {
+      "epoch": 0.4593675620537232,
+      "grad_norm": 1.9399879628948615,
+      "learning_rate": 5.898357923743052e-06,
+      "loss": 0.6978,
+      "step": 1351
+    },
+    {
+      "epoch": 0.4597075824549473,
+      "grad_norm": 2.0206889656325777,
+      "learning_rate": 5.892939307550556e-06,
+      "loss": 0.7937,
+      "step": 1352
+    },
+    {
+      "epoch": 0.46004760285617136,
+      "grad_norm": 2.0270151745480107,
+      "learning_rate": 5.887519607873815e-06,
+      "loss": 0.801,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46038762325739546,
+      "grad_norm": 1.8581224024582177,
+      "learning_rate": 5.882098831289044e-06,
+      "loss": 0.8618,
+      "step": 1354
+    },
+    {
+      "epoch": 0.4607276436586195,
+      "grad_norm": 1.9755282021333846,
+      "learning_rate": 5.8766769843737604e-06,
+      "loss": 0.7721,
+      "step": 1355
+    },
+    {
+      "epoch": 0.4610676640598436,
+      "grad_norm": 1.8373059374517786,
+      "learning_rate": 5.8712540737067835e-06,
+      "loss": 0.7952,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46140768446106767,
+      "grad_norm": 1.5959326369230662,
+      "learning_rate": 5.865830105868226e-06,
+      "loss": 0.7782,
+      "step": 1357
+    },
+    {
+      "epoch": 0.4617477048622917,
+      "grad_norm": 1.9988446870321313,
+      "learning_rate": 5.860405087439475e-06,
+      "loss": 0.8748,
+      "step": 1358
+    },
+    {
+      "epoch": 0.4620877252635158,
+      "grad_norm": 4.024311573353348,
+      "learning_rate": 5.8549790250032e-06,
+      "loss": 0.7804,
+      "step": 1359
+    },
+    {
+      "epoch": 0.4624277456647399,
+      "grad_norm": 1.7924867446440473,
+      "learning_rate": 5.849551925143334e-06,
+      "loss": 0.7366,
+      "step": 1360
+    },
+    {
+      "epoch": 0.462767766065964,
+      "grad_norm": 1.9333933529106293,
+      "learning_rate": 5.84412379444507e-06,
+      "loss": 0.7634,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46310778646718803,
+      "grad_norm": 2.239027398676504,
+      "learning_rate": 5.838694639494852e-06,
+      "loss": 0.7516,
+      "step": 1362
+    },
+    {
+      "epoch": 0.4634478068684121,
+      "grad_norm": 2.564558822755495,
+      "learning_rate": 5.833264466880363e-06,
+      "loss": 0.7493,
+      "step": 1363
+    },
+    {
+      "epoch": 0.4637878272696362,
+      "grad_norm": 2.406933971641201,
+      "learning_rate": 5.827833283190527e-06,
+      "loss": 0.7643,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46412784767086024,
+      "grad_norm": 1.6911599873633802,
+      "learning_rate": 5.8224010950154895e-06,
+      "loss": 0.8361,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46446786807208434,
+      "grad_norm": 1.8143704851616336,
+      "learning_rate": 5.81696790894662e-06,
+      "loss": 0.8781,
+      "step": 1366
+    },
+    {
+      "epoch": 0.4648078884733084,
+      "grad_norm": 1.9049878875916997,
+      "learning_rate": 5.811533731576494e-06,
+      "loss": 0.883,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46514790887453245,
+      "grad_norm": 1.6427893085515717,
+      "learning_rate": 5.806098569498892e-06,
+      "loss": 0.7631,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46548792927575655,
+      "grad_norm": 1.950478915992292,
+      "learning_rate": 5.800662429308787e-06,
+      "loss": 0.7777,
+      "step": 1369
+    },
+    {
+      "epoch": 0.4658279496769806,
+      "grad_norm": 2.3387814639730995,
+      "learning_rate": 5.795225317602344e-06,
+      "loss": 0.7839,
+      "step": 1370
+    },
+    {
+      "epoch": 0.4661679700782047,
+      "grad_norm": 2.3136990487853306,
+      "learning_rate": 5.789787240976903e-06,
+      "loss": 0.8801,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46650799047942876,
+      "grad_norm": 1.883331409095713,
+      "learning_rate": 5.784348206030974e-06,
+      "loss": 0.7718,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46684801088065286,
+      "grad_norm": 1.5552364347893213,
+      "learning_rate": 5.778908219364234e-06,
+      "loss": 0.7953,
+      "step": 1373
+    },
+    {
+      "epoch": 0.4671880312818769,
+      "grad_norm": 1.8758755215294272,
+      "learning_rate": 5.77346728757751e-06,
+      "loss": 0.9304,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46752805168310096,
+      "grad_norm": 2.8864202166172857,
+      "learning_rate": 5.768025417272779e-06,
+      "loss": 0.8601,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46786807208432507,
+      "grad_norm": 1.7477299549548073,
+      "learning_rate": 5.762582615053155e-06,
+      "loss": 0.8618,
+      "step": 1376
+    },
+    {
+      "epoch": 0.4682080924855491,
+      "grad_norm": 1.6359643840822298,
+      "learning_rate": 5.757138887522884e-06,
+      "loss": 0.8735,
+      "step": 1377
+    },
+    {
+      "epoch": 0.4685481128867732,
+      "grad_norm": 1.9885997278079388,
+      "learning_rate": 5.751694241287336e-06,
+      "loss": 0.7201,
+      "step": 1378
+    },
+    {
+      "epoch": 0.4688881332879973,
+      "grad_norm": 2.024775175272147,
+      "learning_rate": 5.7462486829529895e-06,
+      "loss": 0.9019,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46922815368922133,
+      "grad_norm": 2.1517611217388164,
+      "learning_rate": 5.7408022191274385e-06,
+      "loss": 0.7558,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46956817409044543,
+      "grad_norm": 1.6071554576538785,
+      "learning_rate": 5.735354856419371e-06,
+      "loss": 0.7544,
+      "step": 1381
+    },
+    {
+      "epoch": 0.4699081944916695,
+      "grad_norm": 3.0679645169810588,
+      "learning_rate": 5.729906601438564e-06,
+      "loss": 0.6876,
+      "step": 1382
+    },
+    {
+      "epoch": 0.4702482148928936,
+      "grad_norm": 1.7186871226619356,
+      "learning_rate": 5.724457460795883e-06,
+      "loss": 0.9415,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 2.8408348818159648,
+      "learning_rate": 5.71900744110326e-06,
+      "loss": 0.7498,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47092825569534175,
+      "grad_norm": 1.5730885516835262,
+      "learning_rate": 5.713556548973701e-06,
+      "loss": 0.8499,
+      "step": 1385
+    },
+    {
+      "epoch": 0.4712682760965658,
+      "grad_norm": 2.4054974564917146,
+      "learning_rate": 5.708104791021267e-06,
+      "loss": 0.7346,
+      "step": 1386
+    },
+    {
+      "epoch": 0.47160829649778985,
+      "grad_norm": 7.6135309739245605,
+      "learning_rate": 5.702652173861073e-06,
+      "loss": 0.9721,
+      "step": 1387
+    },
+    {
+      "epoch": 0.47194831689901395,
+      "grad_norm": 2.157238376626225,
+      "learning_rate": 5.697198704109269e-06,
+      "loss": 0.8068,
+      "step": 1388
+    },
+    {
+      "epoch": 0.472288337300238,
+      "grad_norm": 2.0366968869428206,
+      "learning_rate": 5.691744388383047e-06,
+      "loss": 0.8825,
+      "step": 1389
+    },
+    {
+      "epoch": 0.4726283577014621,
+      "grad_norm": 3.332007974752036,
+      "learning_rate": 5.686289233300625e-06,
+      "loss": 0.8573,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47296837810268616,
+      "grad_norm": 2.996038777755443,
+      "learning_rate": 5.680833245481234e-06,
+      "loss": 0.7937,
+      "step": 1391
+    },
+    {
+      "epoch": 0.4733083985039102,
+      "grad_norm": 1.876848717220222,
+      "learning_rate": 5.6753764315451196e-06,
+      "loss": 0.8256,
+      "step": 1392
+    },
+    {
+      "epoch": 0.4736484189051343,
+      "grad_norm": 1.7456685592190304,
+      "learning_rate": 5.669918798113531e-06,
+      "loss": 0.781,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47398843930635837,
+      "grad_norm": 2.1150837869236243,
+      "learning_rate": 5.664460351808706e-06,
+      "loss": 0.7863,
+      "step": 1394
+    },
+    {
+      "epoch": 0.4743284597075825,
+      "grad_norm": 2.106099742367351,
+      "learning_rate": 5.659001099253878e-06,
+      "loss": 0.8522,
+      "step": 1395
+    },
+    {
+      "epoch": 0.4746684801088065,
+      "grad_norm": 1.6430769745388225,
+      "learning_rate": 5.653541047073248e-06,
+      "loss": 0.8509,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47500850051003063,
+      "grad_norm": 2.7934049226507893,
+      "learning_rate": 5.648080201891994e-06,
+      "loss": 0.6624,
+      "step": 1397
+    },
+    {
+      "epoch": 0.4753485209112547,
+      "grad_norm": 1.8064873975135194,
+      "learning_rate": 5.642618570336258e-06,
+      "loss": 0.7733,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47568854131247873,
+      "grad_norm": 1.7201690606896272,
+      "learning_rate": 5.637156159033129e-06,
+      "loss": 0.7874,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47602856171370284,
+      "grad_norm": 2.377775914178791,
+      "learning_rate": 5.631692974610647e-06,
+      "loss": 0.9048,
+      "step": 1400
+    },
+    {
+      "epoch": 0.4763685821149269,
+      "grad_norm": 1.9190322261379347,
+      "learning_rate": 5.626229023697789e-06,
+      "loss": 0.8154,
+      "step": 1401
+    },
+    {
+      "epoch": 0.476708602516151,
+      "grad_norm": 1.846727656609046,
+      "learning_rate": 5.6207643129244625e-06,
+      "loss": 0.7141,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47704862291737504,
+      "grad_norm": 1.6992506612080924,
+      "learning_rate": 5.6152988489214985e-06,
+      "loss": 0.8167,
+      "step": 1403
+    },
+    {
+      "epoch": 0.4773886433185991,
+      "grad_norm": 1.8237274064366962,
+      "learning_rate": 5.609832638320637e-06,
+      "loss": 0.8415,
+      "step": 1404
+    },
+    {
+      "epoch": 0.4777286637198232,
+      "grad_norm": 2.046569580094362,
+      "learning_rate": 5.604365687754528e-06,
+      "loss": 0.8745,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47806868412104725,
+      "grad_norm": 2.576331253597654,
+      "learning_rate": 5.59889800385672e-06,
+      "loss": 0.7876,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47840870452227136,
+      "grad_norm": 1.696243517927065,
+      "learning_rate": 5.59342959326165e-06,
+      "loss": 0.6787,
+      "step": 1407
+    },
+    {
+      "epoch": 0.4787487249234954,
+      "grad_norm": 1.7387459270264907,
+      "learning_rate": 5.587960462604634e-06,
+      "loss": 0.7973,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47908874532471946,
+      "grad_norm": 2.1302859398907934,
+      "learning_rate": 5.582490618521864e-06,
+      "loss": 0.7099,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47942876572594356,
+      "grad_norm": 1.7380121057600302,
+      "learning_rate": 5.5770200676504e-06,
+      "loss": 0.877,
+      "step": 1410
+    },
+    {
+      "epoch": 0.4797687861271676,
+      "grad_norm": 2.327507521749447,
+      "learning_rate": 5.571548816628159e-06,
+      "loss": 0.7612,
+      "step": 1411
+    },
+    {
+      "epoch": 0.4801088065283917,
+      "grad_norm": 1.9212379412038696,
+      "learning_rate": 5.5660768720939e-06,
+      "loss": 0.8138,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48044882692961577,
+      "grad_norm": 1.6433392321151539,
+      "learning_rate": 5.560604240687235e-06,
+      "loss": 0.8439,
+      "step": 1413
+    },
+    {
+      "epoch": 0.4807888473308399,
+      "grad_norm": 1.6964960554479955,
+      "learning_rate": 5.555130929048603e-06,
+      "loss": 0.8821,
+      "step": 1414
+    },
+    {
+      "epoch": 0.4811288677320639,
+      "grad_norm": 1.8194872677059573,
+      "learning_rate": 5.5496569438192695e-06,
+      "loss": 0.7899,
+      "step": 1415
+    },
+    {
+      "epoch": 0.481468888133288,
+      "grad_norm": 1.8072104696328097,
+      "learning_rate": 5.544182291641317e-06,
+      "loss": 0.7687,
+      "step": 1416
+    },
+    {
+      "epoch": 0.4818089085345121,
+      "grad_norm": 1.7790457934236203,
+      "learning_rate": 5.538706979157635e-06,
+      "loss": 0.7862,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48214892893573613,
+      "grad_norm": 1.7951850721079083,
+      "learning_rate": 5.533231013011919e-06,
+      "loss": 0.7515,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48248894933696024,
+      "grad_norm": 1.5561662456874903,
+      "learning_rate": 5.527754399848657e-06,
+      "loss": 0.8133,
+      "step": 1419
+    },
+    {
+      "epoch": 0.4828289697381843,
+      "grad_norm": 2.157936588540733,
+      "learning_rate": 5.522277146313117e-06,
+      "loss": 0.7538,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48316899013940834,
+      "grad_norm": 2.4290999302724288,
+      "learning_rate": 5.51679925905135e-06,
+      "loss": 0.7814,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48350901054063244,
+      "grad_norm": 1.841958993342082,
+      "learning_rate": 5.511320744710171e-06,
+      "loss": 0.8118,
+      "step": 1422
+    },
+    {
+      "epoch": 0.4838490309418565,
+      "grad_norm": 2.09445138995467,
+      "learning_rate": 5.505841609937162e-06,
+      "loss": 0.771,
+      "step": 1423
+    },
+    {
+      "epoch": 0.4841890513430806,
+      "grad_norm": 1.9025737655981094,
+      "learning_rate": 5.500361861380651e-06,
+      "loss": 0.8158,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48452907174430465,
+      "grad_norm": 1.593238831589755,
+      "learning_rate": 5.494881505689714e-06,
+      "loss": 0.845,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48486909214552876,
+      "grad_norm": 2.1406933443102902,
+      "learning_rate": 5.489400549514165e-06,
+      "loss": 0.7092,
+      "step": 1426
+    },
+    {
+      "epoch": 0.4852091125467528,
+      "grad_norm": 2.5531087262152354,
+      "learning_rate": 5.483918999504544e-06,
+      "loss": 0.7776,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48554913294797686,
+      "grad_norm": 1.919105692583851,
+      "learning_rate": 5.478436862312113e-06,
+      "loss": 0.719,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48588915334920096,
+      "grad_norm": 1.8829313213513676,
+      "learning_rate": 5.472954144588847e-06,
+      "loss": 0.7841,
+      "step": 1429
+    },
+    {
+      "epoch": 0.486229173750425,
+      "grad_norm": 1.4881374527192293,
+      "learning_rate": 5.467470852987424e-06,
+      "loss": 0.7724,
+      "step": 1430
+    },
+    {
+      "epoch": 0.4865691941516491,
+      "grad_norm": 1.924768602698392,
+      "learning_rate": 5.4619869941612204e-06,
+      "loss": 0.7726,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48690921455287317,
+      "grad_norm": 1.5858390673137608,
+      "learning_rate": 5.456502574764299e-06,
+      "loss": 0.8339,
+      "step": 1432
+    },
+    {
+      "epoch": 0.4872492349540972,
+      "grad_norm": 2.009244140179676,
+      "learning_rate": 5.4510176014514e-06,
+      "loss": 0.8099,
+      "step": 1433
+    },
+    {
+      "epoch": 0.4875892553553213,
+      "grad_norm": 1.7556534967172963,
+      "learning_rate": 5.445532080877942e-06,
+      "loss": 0.8161,
+      "step": 1434
+    },
+    {
+      "epoch": 0.4879292757565454,
+      "grad_norm": 2.1544729558220315,
+      "learning_rate": 5.440046019700004e-06,
+      "loss": 0.822,
+      "step": 1435
+    },
+    {
+      "epoch": 0.4882692961577695,
+      "grad_norm": 2.050182429485166,
+      "learning_rate": 5.434559424574323e-06,
+      "loss": 0.7798,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48860931655899353,
+      "grad_norm": 2.250279830505212,
+      "learning_rate": 5.429072302158279e-06,
+      "loss": 0.772,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48894933696021764,
+      "grad_norm": 1.7152397535373827,
+      "learning_rate": 5.4235846591098995e-06,
+      "loss": 0.7366,
+      "step": 1438
+    },
+    {
+      "epoch": 0.4892893573614417,
+      "grad_norm": 1.99011884513494,
+      "learning_rate": 5.4180965020878365e-06,
+      "loss": 0.7173,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48962937776266574,
+      "grad_norm": 2.173655615360162,
+      "learning_rate": 5.41260783775137e-06,
+      "loss": 0.7406,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48996939816388985,
+      "grad_norm": 2.2971002978470576,
+      "learning_rate": 5.407118672760393e-06,
+      "loss": 0.9206,
+      "step": 1441
+    },
+    {
+      "epoch": 0.4903094185651139,
+      "grad_norm": 1.5023371228055133,
+      "learning_rate": 5.401629013775408e-06,
+      "loss": 0.8379,
+      "step": 1442
+    },
+    {
+      "epoch": 0.490649438966338,
+      "grad_norm": 2.0910378047987375,
+      "learning_rate": 5.396138867457517e-06,
+      "loss": 0.829,
+      "step": 1443
+    },
+    {
+      "epoch": 0.49098945936756205,
+      "grad_norm": 3.7490164898531675,
+      "learning_rate": 5.39064824046841e-06,
+      "loss": 0.9118,
+      "step": 1444
+    },
+    {
+      "epoch": 0.4913294797687861,
+      "grad_norm": 2.126644865691582,
+      "learning_rate": 5.385157139470365e-06,
+      "loss": 0.865,
+      "step": 1445
+    },
+    {
+      "epoch": 0.4916695001700102,
+      "grad_norm": 1.8516246611445681,
+      "learning_rate": 5.379665571126232e-06,
+      "loss": 0.8226,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49200952057123426,
+      "grad_norm": 2.3484056426904467,
+      "learning_rate": 5.374173542099429e-06,
+      "loss": 0.7315,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49234954097245837,
+      "grad_norm": 1.5288170614125434,
+      "learning_rate": 5.368681059053934e-06,
+      "loss": 0.791,
+      "step": 1448
+    },
+    {
+      "epoch": 0.4926895613736824,
+      "grad_norm": 1.9647101647201137,
+      "learning_rate": 5.363188128654272e-06,
+      "loss": 0.7512,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49302958177490647,
+      "grad_norm": 1.5696810158458874,
+      "learning_rate": 5.357694757565515e-06,
+      "loss": 0.8299,
+      "step": 1450
+    },
+    {
+      "epoch": 0.4933696021761306,
+      "grad_norm": 2.0603657799295165,
+      "learning_rate": 5.352200952453268e-06,
+      "loss": 0.8204,
+      "step": 1451
+    },
+    {
+      "epoch": 0.4937096225773546,
+      "grad_norm": 1.9159683428169025,
+      "learning_rate": 5.3467067199836665e-06,
+      "loss": 0.832,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49404964297857873,
+      "grad_norm": 2.2261677908581343,
+      "learning_rate": 5.341212066823356e-06,
+      "loss": 0.8217,
+      "step": 1453
+    },
+    {
+      "epoch": 0.4943896633798028,
+      "grad_norm": 1.81536044760645,
+      "learning_rate": 5.335716999639499e-06,
+      "loss": 0.7984,
+      "step": 1454
+    },
+    {
+      "epoch": 0.4947296837810269,
+      "grad_norm": 2.9109596077188447,
+      "learning_rate": 5.330221525099761e-06,
+      "loss": 0.7177,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49506970418225094,
+      "grad_norm": 2.0524828104111554,
+      "learning_rate": 5.3247256498722985e-06,
+      "loss": 0.8618,
+      "step": 1456
+    },
+    {
+      "epoch": 0.495409724583475,
+      "grad_norm": 1.7519967928504512,
+      "learning_rate": 5.319229380625754e-06,
+      "loss": 0.8823,
+      "step": 1457
+    },
+    {
+      "epoch": 0.4957497449846991,
+      "grad_norm": 1.8637116727130303,
+      "learning_rate": 5.31373272402925e-06,
+      "loss": 0.7384,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49608976538592314,
+      "grad_norm": 2.152147896428193,
+      "learning_rate": 5.308235686752379e-06,
+      "loss": 0.8812,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49642978578714725,
+      "grad_norm": 2.078235845036691,
+      "learning_rate": 5.302738275465196e-06,
+      "loss": 0.7727,
+      "step": 1460
+    },
+    {
+      "epoch": 0.4967698061883713,
+      "grad_norm": 2.102266420017895,
+      "learning_rate": 5.297240496838206e-06,
+      "loss": 0.8562,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49710982658959535,
+      "grad_norm": 1.8174835697796294,
+      "learning_rate": 5.291742357542364e-06,
+      "loss": 0.8048,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49744984699081946,
+      "grad_norm": 1.754386383349789,
+      "learning_rate": 5.2862438642490634e-06,
+      "loss": 0.7872,
+      "step": 1463
+    },
+    {
+      "epoch": 0.4977898673920435,
+      "grad_norm": 4.158515391970638,
+      "learning_rate": 5.280745023630119e-06,
+      "loss": 0.7779,
+      "step": 1464
+    },
+    {
+      "epoch": 0.4981298877932676,
+      "grad_norm": 2.4054339081607767,
+      "learning_rate": 5.275245842357778e-06,
+      "loss": 0.7462,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49846990819449166,
+      "grad_norm": 2.689734918133284,
+      "learning_rate": 5.269746327104693e-06,
+      "loss": 0.8174,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49880992859571577,
+      "grad_norm": 2.295691045707937,
+      "learning_rate": 5.264246484543926e-06,
+      "loss": 0.7969,
+      "step": 1467
+    },
+    {
+      "epoch": 0.4991499489969398,
+      "grad_norm": 2.107973616659225,
+      "learning_rate": 5.258746321348934e-06,
+      "loss": 0.7944,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49948996939816387,
+      "grad_norm": 1.930350616477822,
+      "learning_rate": 5.253245844193564e-06,
+      "loss": 0.8304,
+      "step": 1469
+    },
+    {
+      "epoch": 0.499829989799388,
+      "grad_norm": 1.7525142127819853,
+      "learning_rate": 5.247745059752044e-06,
+      "loss": 0.7762,
+      "step": 1470
+    },
+    {
+      "epoch": 0.500170010200612,
+      "grad_norm": 1.8529587101036085,
+      "learning_rate": 5.242243974698975e-06,
+      "loss": 0.8314,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5005100306018361,
+      "grad_norm": 2.3903901423865457,
+      "learning_rate": 5.236742595709321e-06,
+      "loss": 0.7822,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5008500510030602,
+      "grad_norm": 2.0348324811278777,
+      "learning_rate": 5.231240929458406e-06,
+      "loss": 0.7494,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5011900714042843,
+      "grad_norm": 1.7162050480913733,
+      "learning_rate": 5.225738982621898e-06,
+      "loss": 0.9737,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5015300918055083,
+      "grad_norm": 2.2064804559522093,
+      "learning_rate": 5.220236761875811e-06,
+      "loss": 0.7815,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5018701122067324,
+      "grad_norm": 2.2251124219486798,
+      "learning_rate": 5.214734273896488e-06,
+      "loss": 0.7881,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5022101326079564,
+      "grad_norm": 2.981440295612256,
+      "learning_rate": 5.209231525360594e-06,
+      "loss": 0.8,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5025501530091806,
+      "grad_norm": 7.45515078847437,
+      "learning_rate": 5.203728522945115e-06,
+      "loss": 0.7911,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5028901734104047,
+      "grad_norm": 1.92941189117845,
+      "learning_rate": 5.198225273327343e-06,
+      "loss": 0.7445,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5032301938116287,
+      "grad_norm": 1.8975113113950242,
+      "learning_rate": 5.1927217831848685e-06,
+      "loss": 0.843,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5035702142128528,
+      "grad_norm": 2.1094734040883893,
+      "learning_rate": 5.187218059195578e-06,
+      "loss": 0.8277,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5039102346140768,
+      "grad_norm": 2.1248192008186964,
+      "learning_rate": 5.181714108037635e-06,
+      "loss": 0.7933,
+      "step": 1482
+    },
+    {
+      "epoch": 0.504250255015301,
+      "grad_norm": 1.8280812151104824,
+      "learning_rate": 5.176209936389485e-06,
+      "loss": 0.7447,
+      "step": 1483
+    },
+    {
+      "epoch": 0.504590275416525,
+      "grad_norm": 3.305850569207107,
+      "learning_rate": 5.17070555092984e-06,
+      "loss": 0.7644,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5049302958177491,
+      "grad_norm": 2.714270205426286,
+      "learning_rate": 5.1652009583376676e-06,
+      "loss": 0.7827,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5052703162189731,
+      "grad_norm": 3.091352082861896,
+      "learning_rate": 5.159696165292189e-06,
+      "loss": 0.8001,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5056103366201972,
+      "grad_norm": 1.593603634259395,
+      "learning_rate": 5.154191178472873e-06,
+      "loss": 0.8329,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5059503570214213,
+      "grad_norm": 2.1224743879515633,
+      "learning_rate": 5.148686004559412e-06,
+      "loss": 0.7409,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5062903774226454,
+      "grad_norm": 3.01066198517414,
+      "learning_rate": 5.143180650231741e-06,
+      "loss": 0.8388,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5066303978238694,
+      "grad_norm": 1.7136792637059917,
+      "learning_rate": 5.13767512217e-06,
+      "loss": 0.7341,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5069704182250935,
+      "grad_norm": 2.0631670268496096,
+      "learning_rate": 5.1321694270545455e-06,
+      "loss": 0.7773,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5073104386263175,
+      "grad_norm": 1.874567686624954,
+      "learning_rate": 5.12666357156594e-06,
+      "loss": 0.7862,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5076504590275417,
+      "grad_norm": 6.715516555383307,
+      "learning_rate": 5.121157562384936e-06,
+      "loss": 0.8309,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5079904794287657,
+      "grad_norm": 1.8974326591409743,
+      "learning_rate": 5.115651406192473e-06,
+      "loss": 0.8229,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5083304998299898,
+      "grad_norm": 1.7753191132738406,
+      "learning_rate": 5.110145109669671e-06,
+      "loss": 0.7212,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5086705202312138,
+      "grad_norm": 1.8871431993037888,
+      "learning_rate": 5.104638679497818e-06,
+      "loss": 0.7695,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5090105406324379,
+      "grad_norm": 2.1209005798272993,
+      "learning_rate": 5.0991321223583655e-06,
+      "loss": 0.8439,
+      "step": 1497
+    },
+    {
+      "epoch": 0.509350561033662,
+      "grad_norm": 1.8469772744160644,
+      "learning_rate": 5.093625444932917e-06,
+      "loss": 0.806,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5096905814348861,
+      "grad_norm": 2.281929279661747,
+      "learning_rate": 5.088118653903225e-06,
+      "loss": 0.8326,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5100306018361102,
+      "grad_norm": 1.9613864164506285,
+      "learning_rate": 5.08261175595118e-06,
+      "loss": 0.6774,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5103706222373342,
+      "grad_norm": 3.9527172939988366,
+      "learning_rate": 5.0771047577587995e-06,
+      "loss": 0.7849,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5107106426385584,
+      "grad_norm": 3.9507534634382218,
+      "learning_rate": 5.071597666008223e-06,
+      "loss": 0.8327,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5110506630397824,
+      "grad_norm": 2.3477801057322587,
+      "learning_rate": 5.066090487381705e-06,
+      "loss": 0.7659,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5113906834410065,
+      "grad_norm": 1.6637921910839917,
+      "learning_rate": 5.060583228561604e-06,
+      "loss": 0.7807,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5117307038422305,
+      "grad_norm": 1.9869572742030395,
+      "learning_rate": 5.055075896230379e-06,
+      "loss": 0.8009,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5120707242434546,
+      "grad_norm": 2.7596350421531493,
+      "learning_rate": 5.0495684970705725e-06,
+      "loss": 0.8015,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5124107446446787,
+      "grad_norm": 2.5992292669330306,
+      "learning_rate": 5.044061037764814e-06,
+      "loss": 0.7465,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5127507650459028,
+      "grad_norm": 1.7505681170663439,
+      "learning_rate": 5.0385535249958015e-06,
+      "loss": 0.8182,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5130907854471268,
+      "grad_norm": 2.1632485122473404,
+      "learning_rate": 5.033045965446303e-06,
+      "loss": 0.7183,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5134308058483509,
+      "grad_norm": 1.753555699009244,
+      "learning_rate": 5.027538365799135e-06,
+      "loss": 0.7862,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5137708262495749,
+      "grad_norm": 1.9243297591580018,
+      "learning_rate": 5.022030732737172e-06,
+      "loss": 0.8458,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5141108466507991,
+      "grad_norm": 2.177617314164665,
+      "learning_rate": 5.016523072943321e-06,
+      "loss": 0.8359,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5144508670520231,
+      "grad_norm": 1.4739366045699038,
+      "learning_rate": 5.011015393100529e-06,
+      "loss": 0.7725,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5147908874532472,
+      "grad_norm": 1.9809330712521047,
+      "learning_rate": 5.00550769989176e-06,
+      "loss": 0.8364,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5151309078544712,
+      "grad_norm": 1.7781499734908552,
+      "learning_rate": 5e-06,
+      "loss": 0.7835,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5154709282556953,
+      "grad_norm": 1.8799570009883524,
+      "learning_rate": 4.994492300108241e-06,
+      "loss": 0.7751,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5158109486569195,
+      "grad_norm": 3.073785329201592,
+      "learning_rate": 4.988984606899473e-06,
+      "loss": 0.7751,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5161509690581435,
+      "grad_norm": 1.776054822698832,
+      "learning_rate": 4.9834769270566805e-06,
+      "loss": 0.853,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5164909894593676,
+      "grad_norm": 1.7634367926127321,
+      "learning_rate": 4.977969267262829e-06,
+      "loss": 0.8076,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5168310098605916,
+      "grad_norm": 4.086423843722984,
+      "learning_rate": 4.972461634200866e-06,
+      "loss": 0.9011,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5171710302618157,
+      "grad_norm": 1.6421236623641557,
+      "learning_rate": 4.966954034553699e-06,
+      "loss": 0.8642,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5175110506630398,
+      "grad_norm": 1.8584454764781853,
+      "learning_rate": 4.961446475004199e-06,
+      "loss": 0.7653,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5178510710642639,
+      "grad_norm": 1.4656599468459788,
+      "learning_rate": 4.955938962235186e-06,
+      "loss": 0.8445,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5181910914654879,
+      "grad_norm": 1.8519620362169094,
+      "learning_rate": 4.950431502929428e-06,
+      "loss": 0.7378,
+      "step": 1524
+    },
+    {
+      "epoch": 0.518531111866712,
+      "grad_norm": 2.3725424976917413,
+      "learning_rate": 4.944924103769623e-06,
+      "loss": 0.7779,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5188711322679361,
+      "grad_norm": 1.7226943911959502,
+      "learning_rate": 4.939416771438397e-06,
+      "loss": 0.7654,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5192111526691602,
+      "grad_norm": 1.8557184031747187,
+      "learning_rate": 4.933909512618298e-06,
+      "loss": 0.863,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5195511730703842,
+      "grad_norm": 2.115844790979398,
+      "learning_rate": 4.928402333991777e-06,
+      "loss": 0.7592,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5198911934716083,
+      "grad_norm": 1.6348805097468087,
+      "learning_rate": 4.922895242241202e-06,
+      "loss": 0.8855,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5202312138728323,
+      "grad_norm": 2.3404047261939858,
+      "learning_rate": 4.91738824404882e-06,
+      "loss": 0.8327,
+      "step": 1530
+    },
+    {
+      "epoch": 0.5205712342740565,
+      "grad_norm": 1.8623557309255268,
+      "learning_rate": 4.9118813460967754e-06,
+      "loss": 0.8303,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5209112546752805,
+      "grad_norm": 2.2490241499645847,
+      "learning_rate": 4.906374555067085e-06,
+      "loss": 0.8482,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5212512750765046,
+      "grad_norm": 2.331643434530923,
+      "learning_rate": 4.900867877641636e-06,
+      "loss": 0.839,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5215912954777286,
+      "grad_norm": 2.095247817090777,
+      "learning_rate": 4.895361320502185e-06,
+      "loss": 0.7988,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5219313158789527,
+      "grad_norm": 1.723635906341325,
+      "learning_rate": 4.88985489033033e-06,
+      "loss": 0.7934,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5222713362801769,
+      "grad_norm": 2.046788329469224,
+      "learning_rate": 4.8843485938075286e-06,
+      "loss": 0.817,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5226113566814009,
+      "grad_norm": 2.0488449285989003,
+      "learning_rate": 4.878842437615065e-06,
+      "loss": 0.7112,
+      "step": 1537
+    },
+    {
+      "epoch": 0.522951377082625,
+      "grad_norm": 1.9845826095458905,
+      "learning_rate": 4.873336428434062e-06,
+      "loss": 0.759,
+      "step": 1538
+    },
+    {
+      "epoch": 0.523291397483849,
+      "grad_norm": 4.138763157740066,
+      "learning_rate": 4.8678305729454545e-06,
+      "loss": 0.8152,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5236314178850731,
+      "grad_norm": 3.647861718118265,
+      "learning_rate": 4.862324877830003e-06,
+      "loss": 0.8438,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5239714382862972,
+      "grad_norm": 2.2618049383095196,
+      "learning_rate": 4.856819349768262e-06,
+      "loss": 0.7159,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5243114586875213,
+      "grad_norm": 1.5426259829995164,
+      "learning_rate": 4.851313995440589e-06,
+      "loss": 0.7474,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5246514790887453,
+      "grad_norm": 1.9326266116251898,
+      "learning_rate": 4.845808821527131e-06,
+      "loss": 0.7739,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5249914994899694,
+      "grad_norm": 1.838847556325622,
+      "learning_rate": 4.840303834707811e-06,
+      "loss": 0.7753,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5253315198911934,
+      "grad_norm": 1.7052452591835734,
+      "learning_rate": 4.834799041662333e-06,
+      "loss": 0.6825,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5256715402924176,
+      "grad_norm": 2.2889041372273056,
+      "learning_rate": 4.829294449070161e-06,
+      "loss": 0.8191,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5260115606936416,
+      "grad_norm": 3.1528847052040416,
+      "learning_rate": 4.8237900636105154e-06,
+      "loss": 0.8092,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5263515810948657,
+      "grad_norm": 4.090914456990795,
+      "learning_rate": 4.818285891962367e-06,
+      "loss": 0.8098,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5266916014960897,
+      "grad_norm": 1.659291969185622,
+      "learning_rate": 4.812781940804424e-06,
+      "loss": 0.8033,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5270316218973138,
+      "grad_norm": 2.134430251801482,
+      "learning_rate": 4.807278216815132e-06,
+      "loss": 0.8078,
+      "step": 1550
+    },
+    {
+      "epoch": 0.527371642298538,
+      "grad_norm": 3.1931811524180778,
+      "learning_rate": 4.801774726672658e-06,
+      "loss": 0.9237,
+      "step": 1551
+    },
+    {
+      "epoch": 0.527711662699762,
+      "grad_norm": 2.616206525270748,
+      "learning_rate": 4.796271477054887e-06,
+      "loss": 0.7764,
+      "step": 1552
+    },
+    {
+      "epoch": 0.528051683100986,
+      "grad_norm": 2.221145968614602,
+      "learning_rate": 4.790768474639407e-06,
+      "loss": 0.8206,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5283917035022101,
+      "grad_norm": 2.892055480983333,
+      "learning_rate": 4.785265726103514e-06,
+      "loss": 0.7451,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5287317239034343,
+      "grad_norm": 1.6659559597323008,
+      "learning_rate": 4.77976323812419e-06,
+      "loss": 0.8594,
+      "step": 1555
+    },
+    {
+      "epoch": 0.5290717443046583,
+      "grad_norm": 1.9584708522977028,
+      "learning_rate": 4.7742610173781025e-06,
+      "loss": 0.7449,
+      "step": 1556
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 3.613058182805375,
+      "learning_rate": 4.768759070541596e-06,
+      "loss": 0.8322,
+      "step": 1557
+    },
+    {
+      "epoch": 0.5297517851071064,
+      "grad_norm": 1.9177605042321149,
+      "learning_rate": 4.76325740429068e-06,
+      "loss": 0.8372,
+      "step": 1558
+    },
+    {
+      "epoch": 0.5300918055083305,
+      "grad_norm": 1.9899246030541402,
+      "learning_rate": 4.7577560253010275e-06,
+      "loss": 0.7641,
+      "step": 1559
+    },
+    {
+      "epoch": 0.5304318259095546,
+      "grad_norm": 2.3476977109929975,
+      "learning_rate": 4.752254940247956e-06,
+      "loss": 0.8484,
+      "step": 1560
+    },
+    {
+      "epoch": 0.5307718463107787,
+      "grad_norm": 1.8469551861867957,
+      "learning_rate": 4.746754155806437e-06,
+      "loss": 0.8196,
+      "step": 1561
+    },
+    {
+      "epoch": 0.5311118667120027,
+      "grad_norm": 2.1425767308127495,
+      "learning_rate": 4.741253678651067e-06,
+      "loss": 0.86,
+      "step": 1562
+    },
+    {
+      "epoch": 0.5314518871132268,
+      "grad_norm": 2.4488270324011054,
+      "learning_rate": 4.735753515456076e-06,
+      "loss": 0.801,
+      "step": 1563
+    },
+    {
+      "epoch": 0.5317919075144508,
+      "grad_norm": 2.5625039316596947,
+      "learning_rate": 4.7302536728953095e-06,
+      "loss": 0.7215,
+      "step": 1564
+    },
+    {
+      "epoch": 0.532131927915675,
+      "grad_norm": 2.330648518728265,
+      "learning_rate": 4.724754157642223e-06,
+      "loss": 0.8298,
+      "step": 1565
+    },
+    {
+      "epoch": 0.532471948316899,
+      "grad_norm": 1.6442430241473305,
+      "learning_rate": 4.719254976369882e-06,
+      "loss": 0.8346,
+      "step": 1566
+    },
+    {
+      "epoch": 0.5328119687181231,
+      "grad_norm": 2.06375287355313,
+      "learning_rate": 4.713756135750939e-06,
+      "loss": 0.8094,
+      "step": 1567
+    },
+    {
+      "epoch": 0.5331519891193471,
+      "grad_norm": 1.8032858583233626,
+      "learning_rate": 4.708257642457637e-06,
+      "loss": 0.7847,
+      "step": 1568
+    },
+    {
+      "epoch": 0.5334920095205712,
+      "grad_norm": 2.9072319400722106,
+      "learning_rate": 4.702759503161794e-06,
+      "loss": 0.7787,
+      "step": 1569
+    },
+    {
+      "epoch": 0.5338320299217953,
+      "grad_norm": 2.737668124200652,
+      "learning_rate": 4.697261724534805e-06,
+      "loss": 0.9145,
+      "step": 1570
+    },
+    {
+      "epoch": 0.5341720503230194,
+      "grad_norm": 1.5899160859318942,
+      "learning_rate": 4.691764313247621e-06,
+      "loss": 0.806,
+      "step": 1571
+    },
+    {
+      "epoch": 0.5345120707242434,
+      "grad_norm": 2.128696822075121,
+      "learning_rate": 4.686267275970751e-06,
+      "loss": 0.8027,
+      "step": 1572
+    },
+    {
+      "epoch": 0.5348520911254675,
+      "grad_norm": 6.327392776694456,
+      "learning_rate": 4.680770619374248e-06,
+      "loss": 0.9375,
+      "step": 1573
+    },
+    {
+      "epoch": 0.5351921115266915,
+      "grad_norm": 1.8507733306360983,
+      "learning_rate": 4.675274350127702e-06,
+      "loss": 0.6373,
+      "step": 1574
+    },
+    {
+      "epoch": 0.5355321319279157,
+      "grad_norm": 3.2014249362155764,
+      "learning_rate": 4.669778474900241e-06,
+      "loss": 0.7396,
+      "step": 1575
+    },
+    {
+      "epoch": 0.5358721523291398,
+      "grad_norm": 1.8220955607716283,
+      "learning_rate": 4.664283000360501e-06,
+      "loss": 0.7536,
+      "step": 1576
+    },
+    {
+      "epoch": 0.5362121727303638,
+      "grad_norm": 2.0103763713686202,
+      "learning_rate": 4.6587879331766465e-06,
+      "loss": 0.766,
+      "step": 1577
+    },
+    {
+      "epoch": 0.5365521931315879,
+      "grad_norm": 1.9882582274236145,
+      "learning_rate": 4.653293280016335e-06,
+      "loss": 0.7164,
+      "step": 1578
+    },
+    {
+      "epoch": 0.536892213532812,
+      "grad_norm": 2.332642295450942,
+      "learning_rate": 4.647799047546733e-06,
+      "loss": 0.804,
+      "step": 1579
+    },
+    {
+      "epoch": 0.5372322339340361,
+      "grad_norm": 2.1481887049056367,
+      "learning_rate": 4.642305242434488e-06,
+      "loss": 0.8621,
+      "step": 1580
+    },
+    {
+      "epoch": 0.5375722543352601,
+      "grad_norm": 2.233927473620768,
+      "learning_rate": 4.63681187134573e-06,
+      "loss": 0.8682,
+      "step": 1581
+    },
+    {
+      "epoch": 0.5379122747364842,
+      "grad_norm": 1.8595284687765963,
+      "learning_rate": 4.6313189409460694e-06,
+      "loss": 0.8078,
+      "step": 1582
+    },
+    {
+      "epoch": 0.5382522951377082,
+      "grad_norm": 1.6189064806741513,
+      "learning_rate": 4.625826457900573e-06,
+      "loss": 0.7825,
+      "step": 1583
+    },
+    {
+      "epoch": 0.5385923155389324,
+      "grad_norm": 1.7109496499151693,
+      "learning_rate": 4.62033442887377e-06,
+      "loss": 0.8204,
+      "step": 1584
+    },
+    {
+      "epoch": 0.5389323359401564,
+      "grad_norm": 1.9628012655876577,
+      "learning_rate": 4.614842860529636e-06,
+      "loss": 0.7718,
+      "step": 1585
+    },
+    {
+      "epoch": 0.5392723563413805,
+      "grad_norm": 2.215853939912509,
+      "learning_rate": 4.6093517595315906e-06,
+      "loss": 0.8478,
+      "step": 1586
+    },
+    {
+      "epoch": 0.5396123767426045,
+      "grad_norm": 1.7285326667444412,
+      "learning_rate": 4.603861132542484e-06,
+      "loss": 0.7447,
+      "step": 1587
+    },
+    {
+      "epoch": 0.5399523971438286,
+      "grad_norm": 1.8991399112924394,
+      "learning_rate": 4.598370986224594e-06,
+      "loss": 0.804,
+      "step": 1588
+    },
+    {
+      "epoch": 0.5402924175450527,
+      "grad_norm": 2.1876690348212477,
+      "learning_rate": 4.59288132723961e-06,
+      "loss": 0.827,
+      "step": 1589
+    },
+    {
+      "epoch": 0.5406324379462768,
+      "grad_norm": 3.388267144912403,
+      "learning_rate": 4.587392162248631e-06,
+      "loss": 0.9509,
+      "step": 1590
+    },
+    {
+      "epoch": 0.5409724583475009,
+      "grad_norm": 2.75901823504875,
+      "learning_rate": 4.581903497912164e-06,
+      "loss": 0.8255,
+      "step": 1591
+    },
+    {
+      "epoch": 0.5413124787487249,
+      "grad_norm": 3.0797491504488193,
+      "learning_rate": 4.576415340890101e-06,
+      "loss": 0.9066,
+      "step": 1592
+    },
+    {
+      "epoch": 0.541652499149949,
+      "grad_norm": 1.7281255669918878,
+      "learning_rate": 4.570927697841722e-06,
+      "loss": 0.8885,
+      "step": 1593
+    },
+    {
+      "epoch": 0.5419925195511731,
+      "grad_norm": 1.780260195431119,
+      "learning_rate": 4.565440575425678e-06,
+      "loss": 0.8186,
+      "step": 1594
+    },
+    {
+      "epoch": 0.5423325399523972,
+      "grad_norm": 1.698715980373987,
+      "learning_rate": 4.559953980299998e-06,
+      "loss": 0.7423,
+      "step": 1595
+    },
+    {
+      "epoch": 0.5426725603536212,
+      "grad_norm": 1.8233055697908436,
+      "learning_rate": 4.554467919122061e-06,
+      "loss": 0.7461,
+      "step": 1596
+    },
+    {
+      "epoch": 0.5430125807548453,
+      "grad_norm": 2.0732693883965747,
+      "learning_rate": 4.548982398548601e-06,
+      "loss": 0.8519,
+      "step": 1597
+    },
+    {
+      "epoch": 0.5433526011560693,
+      "grad_norm": 4.076606566555027,
+      "learning_rate": 4.543497425235705e-06,
+      "loss": 0.8375,
+      "step": 1598
+    },
+    {
+      "epoch": 0.5436926215572935,
+      "grad_norm": 3.726800785573923,
+      "learning_rate": 4.538013005838781e-06,
+      "loss": 0.8457,
+      "step": 1599
+    },
+    {
+      "epoch": 0.5440326419585175,
+      "grad_norm": 1.640942051570328,
+      "learning_rate": 4.532529147012578e-06,
+      "loss": 0.7555,
+      "step": 1600
+    },
+    {
+      "epoch": 0.5443726623597416,
+      "grad_norm": 2.1236371326115004,
+      "learning_rate": 4.527045855411153e-06,
+      "loss": 0.7701,
+      "step": 1601
+    },
+    {
+      "epoch": 0.5447126827609656,
+      "grad_norm": 1.7181777314185185,
+      "learning_rate": 4.521563137687889e-06,
+      "loss": 0.8164,
+      "step": 1602
+    },
+    {
+      "epoch": 0.5450527031621897,
+      "grad_norm": 2.448664581131007,
+      "learning_rate": 4.516081000495458e-06,
+      "loss": 0.8668,
+      "step": 1603
+    },
+    {
+      "epoch": 0.5453927235634138,
+      "grad_norm": 2.494208677148641,
+      "learning_rate": 4.510599450485838e-06,
+      "loss": 0.8405,
+      "step": 1604
+    },
+    {
+      "epoch": 0.5457327439646379,
+      "grad_norm": 1.9946946250715516,
+      "learning_rate": 4.505118494310289e-06,
+      "loss": 0.8654,
+      "step": 1605
+    },
+    {
+      "epoch": 0.5460727643658619,
+      "grad_norm": 2.016493132082253,
+      "learning_rate": 4.499638138619351e-06,
+      "loss": 0.7986,
+      "step": 1606
+    },
+    {
+      "epoch": 0.546412784767086,
+      "grad_norm": 5.5575317498533705,
+      "learning_rate": 4.49415839006284e-06,
+      "loss": 0.8583,
+      "step": 1607
+    },
+    {
+      "epoch": 0.5467528051683102,
+      "grad_norm": 1.5958424468455283,
+      "learning_rate": 4.488679255289829e-06,
+      "loss": 0.7993,
+      "step": 1608
+    },
+    {
+      "epoch": 0.5470928255695342,
+      "grad_norm": 3.075588169517648,
+      "learning_rate": 4.483200740948652e-06,
+      "loss": 0.6526,
+      "step": 1609
+    },
+    {
+      "epoch": 0.5474328459707583,
+      "grad_norm": 2.27438391686465,
+      "learning_rate": 4.477722853686883e-06,
+      "loss": 0.7749,
+      "step": 1610
+    },
+    {
+      "epoch": 0.5477728663719823,
+      "grad_norm": 1.899436303639678,
+      "learning_rate": 4.472245600151344e-06,
+      "loss": 0.7449,
+      "step": 1611
+    },
+    {
+      "epoch": 0.5481128867732064,
+      "grad_norm": 1.8159424708385177,
+      "learning_rate": 4.466768986988082e-06,
+      "loss": 0.7725,
+      "step": 1612
+    },
+    {
+      "epoch": 0.5484529071744305,
+      "grad_norm": 2.3206076275584118,
+      "learning_rate": 4.461293020842366e-06,
+      "loss": 0.8011,
+      "step": 1613
+    },
+    {
+      "epoch": 0.5487929275756546,
+      "grad_norm": 2.5334818132018735,
+      "learning_rate": 4.4558177083586855e-06,
+      "loss": 0.8291,
+      "step": 1614
+    },
+    {
+      "epoch": 0.5491329479768786,
+      "grad_norm": 1.824370986639498,
+      "learning_rate": 4.450343056180731e-06,
+      "loss": 0.8763,
+      "step": 1615
+    },
+    {
+      "epoch": 0.5494729683781027,
+      "grad_norm": 2.265085787284215,
+      "learning_rate": 4.444869070951398e-06,
+      "loss": 0.7383,
+      "step": 1616
+    },
+    {
+      "epoch": 0.5498129887793267,
+      "grad_norm": 1.5706339248830496,
+      "learning_rate": 4.439395759312765e-06,
+      "loss": 0.7321,
+      "step": 1617
+    },
+    {
+      "epoch": 0.5501530091805509,
+      "grad_norm": 1.8450960380829842,
+      "learning_rate": 4.433923127906101e-06,
+      "loss": 0.8253,
+      "step": 1618
+    },
+    {
+      "epoch": 0.5504930295817749,
+      "grad_norm": 1.8895954088436864,
+      "learning_rate": 4.428451183371844e-06,
+      "loss": 0.7584,
+      "step": 1619
+    },
+    {
+      "epoch": 0.550833049982999,
+      "grad_norm": 1.7647869933641376,
+      "learning_rate": 4.422979932349601e-06,
+      "loss": 0.8461,
+      "step": 1620
+    },
+    {
+      "epoch": 0.551173070384223,
+      "grad_norm": 1.726859257880695,
+      "learning_rate": 4.417509381478139e-06,
+      "loss": 0.9478,
+      "step": 1621
+    },
+    {
+      "epoch": 0.5515130907854471,
+      "grad_norm": 1.7830292603282358,
+      "learning_rate": 4.412039537395369e-06,
+      "loss": 0.8192,
+      "step": 1622
+    },
+    {
+      "epoch": 0.5518531111866712,
+      "grad_norm": 2.135528104500953,
+      "learning_rate": 4.4065704067383526e-06,
+      "loss": 0.789,
+      "step": 1623
+    },
+    {
+      "epoch": 0.5521931315878953,
+      "grad_norm": 2.2835744809721947,
+      "learning_rate": 4.401101996143281e-06,
+      "loss": 0.7897,
+      "step": 1624
+    },
+    {
+      "epoch": 0.5525331519891193,
+      "grad_norm": 1.9968779612406193,
+      "learning_rate": 4.395634312245473e-06,
+      "loss": 0.8017,
+      "step": 1625
+    },
+    {
+      "epoch": 0.5528731723903434,
+      "grad_norm": 1.7386489203169355,
+      "learning_rate": 4.390167361679363e-06,
+      "loss": 0.8258,
+      "step": 1626
+    },
+    {
+      "epoch": 0.5532131927915674,
+      "grad_norm": 2.3880896647187813,
+      "learning_rate": 4.384701151078502e-06,
+      "loss": 0.6548,
+      "step": 1627
+    },
+    {
+      "epoch": 0.5535532131927916,
+      "grad_norm": 2.3052787463523248,
+      "learning_rate": 4.379235687075538e-06,
+      "loss": 0.8939,
+      "step": 1628
+    },
+    {
+      "epoch": 0.5538932335940157,
+      "grad_norm": 1.9456406943708848,
+      "learning_rate": 4.373770976302212e-06,
+      "loss": 0.7207,
+      "step": 1629
+    },
+    {
+      "epoch": 0.5542332539952397,
+      "grad_norm": 2.233470695480692,
+      "learning_rate": 4.368307025389355e-06,
+      "loss": 0.9426,
+      "step": 1630
+    },
+    {
+      "epoch": 0.5545732743964638,
+      "grad_norm": 1.9001780195038485,
+      "learning_rate": 4.362843840966872e-06,
+      "loss": 0.7396,
+      "step": 1631
+    },
+    {
+      "epoch": 0.5549132947976878,
+      "grad_norm": 1.6170663156516558,
+      "learning_rate": 4.357381429663744e-06,
+      "loss": 0.7398,
+      "step": 1632
+    },
+    {
+      "epoch": 0.555253315198912,
+      "grad_norm": 1.7903568747147653,
+      "learning_rate": 4.351919798108006e-06,
+      "loss": 0.7973,
+      "step": 1633
+    },
+    {
+      "epoch": 0.555593335600136,
+      "grad_norm": 1.8519949395264552,
+      "learning_rate": 4.346458952926754e-06,
+      "loss": 0.7845,
+      "step": 1634
+    },
+    {
+      "epoch": 0.5559333560013601,
+      "grad_norm": 2.312852525979355,
+      "learning_rate": 4.340998900746123e-06,
+      "loss": 0.7661,
+      "step": 1635
+    },
+    {
+      "epoch": 0.5562733764025841,
+      "grad_norm": 1.9538845003355985,
+      "learning_rate": 4.335539648191295e-06,
+      "loss": 0.8089,
+      "step": 1636
+    },
+    {
+      "epoch": 0.5566133968038083,
+      "grad_norm": 2.0572388515661495,
+      "learning_rate": 4.330081201886473e-06,
+      "loss": 0.8594,
+      "step": 1637
+    },
+    {
+      "epoch": 0.5569534172050323,
+      "grad_norm": 1.7570956508285698,
+      "learning_rate": 4.324623568454881e-06,
+      "loss": 0.7019,
+      "step": 1638
+    },
+    {
+      "epoch": 0.5572934376062564,
+      "grad_norm": 2.6666409487156297,
+      "learning_rate": 4.319166754518768e-06,
+      "loss": 0.8802,
+      "step": 1639
+    },
+    {
+      "epoch": 0.5576334580074804,
+      "grad_norm": 8.161817414104679,
+      "learning_rate": 4.313710766699377e-06,
+      "loss": 0.8173,
+      "step": 1640
+    },
+    {
+      "epoch": 0.5579734784087045,
+      "grad_norm": 1.5517265216589782,
+      "learning_rate": 4.308255611616954e-06,
+      "loss": 0.7627,
+      "step": 1641
+    },
+    {
+      "epoch": 0.5583134988099286,
+      "grad_norm": 2.276892346390744,
+      "learning_rate": 4.302801295890731e-06,
+      "loss": 0.8266,
+      "step": 1642
+    },
+    {
+      "epoch": 0.5586535192111527,
+      "grad_norm": 1.5667650901676113,
+      "learning_rate": 4.297347826138929e-06,
+      "loss": 0.7707,
+      "step": 1643
+    },
+    {
+      "epoch": 0.5589935396123767,
+      "grad_norm": 1.6949899190570468,
+      "learning_rate": 4.291895208978734e-06,
+      "loss": 0.7413,
+      "step": 1644
+    },
+    {
+      "epoch": 0.5593335600136008,
+      "grad_norm": 2.006877875279477,
+      "learning_rate": 4.2864434510263e-06,
+      "loss": 0.7829,
+      "step": 1645
+    },
+    {
+      "epoch": 0.5596735804148248,
+      "grad_norm": 3.095263458414816,
+      "learning_rate": 4.280992558896742e-06,
+      "loss": 0.7722,
+      "step": 1646
+    },
+    {
+      "epoch": 0.560013600816049,
+      "grad_norm": 2.240127359014377,
+      "learning_rate": 4.275542539204118e-06,
+      "loss": 0.7562,
+      "step": 1647
+    },
+    {
+      "epoch": 0.5603536212172731,
+      "grad_norm": 2.1541576931196107,
+      "learning_rate": 4.270093398561437e-06,
+      "loss": 0.7223,
+      "step": 1648
+    },
+    {
+      "epoch": 0.5606936416184971,
+      "grad_norm": 1.890091367677857,
+      "learning_rate": 4.26464514358063e-06,
+      "loss": 0.7961,
+      "step": 1649
+    },
+    {
+      "epoch": 0.5610336620197212,
+      "grad_norm": 2.1168877878836514,
+      "learning_rate": 4.259197780872562e-06,
+      "loss": 0.8332,
+      "step": 1650
+    },
+    {
+      "epoch": 0.5613736824209452,
+      "grad_norm": 1.7140754905216478,
+      "learning_rate": 4.2537513170470105e-06,
+      "loss": 0.8327,
+      "step": 1651
+    },
+    {
+      "epoch": 0.5617137028221694,
+      "grad_norm": 2.441707848688923,
+      "learning_rate": 4.248305758712666e-06,
+      "loss": 0.7136,
+      "step": 1652
+    },
+    {
+      "epoch": 0.5620537232233934,
+      "grad_norm": 2.2663018210778483,
+      "learning_rate": 4.2428611124771184e-06,
+      "loss": 0.7338,
+      "step": 1653
+    },
+    {
+      "epoch": 0.5623937436246175,
+      "grad_norm": 1.7042736119872506,
+      "learning_rate": 4.237417384946846e-06,
+      "loss": 0.8221,
+      "step": 1654
+    },
+    {
+      "epoch": 0.5627337640258415,
+      "grad_norm": 1.7857448847407418,
+      "learning_rate": 4.231974582727223e-06,
+      "loss": 0.8938,
+      "step": 1655
+    },
+    {
+      "epoch": 0.5630737844270656,
+      "grad_norm": 8.534990207003116,
+      "learning_rate": 4.226532712422492e-06,
+      "loss": 0.8593,
+      "step": 1656
+    },
+    {
+      "epoch": 0.5634138048282897,
+      "grad_norm": 2.0702826945140456,
+      "learning_rate": 4.221091780635768e-06,
+      "loss": 0.8043,
+      "step": 1657
+    },
+    {
+      "epoch": 0.5637538252295138,
+      "grad_norm": 1.880149571201394,
+      "learning_rate": 4.215651793969026e-06,
+      "loss": 0.7408,
+      "step": 1658
+    },
+    {
+      "epoch": 0.5640938456307378,
+      "grad_norm": 1.9094946983165033,
+      "learning_rate": 4.210212759023099e-06,
+      "loss": 0.85,
+      "step": 1659
+    },
+    {
+      "epoch": 0.5644338660319619,
+      "grad_norm": 2.0826267533303144,
+      "learning_rate": 4.204774682397658e-06,
+      "loss": 0.7968,
+      "step": 1660
+    },
+    {
+      "epoch": 0.564773886433186,
+      "grad_norm": 1.9576152950783854,
+      "learning_rate": 4.199337570691214e-06,
+      "loss": 0.7934,
+      "step": 1661
+    },
+    {
+      "epoch": 0.5651139068344101,
+      "grad_norm": 1.8605000019574227,
+      "learning_rate": 4.1939014305011116e-06,
+      "loss": 0.7489,
+      "step": 1662
+    },
+    {
+      "epoch": 0.5654539272356341,
+      "grad_norm": 1.7452937853609298,
+      "learning_rate": 4.188466268423507e-06,
+      "loss": 0.798,
+      "step": 1663
+    },
+    {
+      "epoch": 0.5657939476368582,
+      "grad_norm": 2.2351881221107233,
+      "learning_rate": 4.183032091053381e-06,
+      "loss": 0.7977,
+      "step": 1664
+    },
+    {
+      "epoch": 0.5661339680380822,
+      "grad_norm": 1.949488664977891,
+      "learning_rate": 4.1775989049845105e-06,
+      "loss": 0.7882,
+      "step": 1665
+    },
+    {
+      "epoch": 0.5664739884393064,
+      "grad_norm": 1.9412918778769286,
+      "learning_rate": 4.172166716809475e-06,
+      "loss": 0.8033,
+      "step": 1666
+    },
+    {
+      "epoch": 0.5668140088405305,
+      "grad_norm": 1.8838839464107233,
+      "learning_rate": 4.166735533119638e-06,
+      "loss": 0.7347,
+      "step": 1667
+    },
+    {
+      "epoch": 0.5671540292417545,
+      "grad_norm": 2.4094539519885823,
+      "learning_rate": 4.16130536050515e-06,
+      "loss": 0.8985,
+      "step": 1668
+    },
+    {
+      "epoch": 0.5674940496429786,
+      "grad_norm": 1.832393433721923,
+      "learning_rate": 4.155876205554931e-06,
+      "loss": 0.7948,
+      "step": 1669
+    },
+    {
+      "epoch": 0.5678340700442026,
+      "grad_norm": 1.7842845158543639,
+      "learning_rate": 4.150448074856667e-06,
+      "loss": 0.856,
+      "step": 1670
+    },
+    {
+      "epoch": 0.5681740904454268,
+      "grad_norm": 5.093298965869656,
+      "learning_rate": 4.145020974996802e-06,
+      "loss": 0.8544,
+      "step": 1671
+    },
+    {
+      "epoch": 0.5685141108466508,
+      "grad_norm": 1.8288966692154494,
+      "learning_rate": 4.139594912560526e-06,
+      "loss": 0.7695,
+      "step": 1672
+    },
+    {
+      "epoch": 0.5688541312478749,
+      "grad_norm": 2.5087803047490196,
+      "learning_rate": 4.134169894131776e-06,
+      "loss": 0.8,
+      "step": 1673
+    },
+    {
+      "epoch": 0.5691941516490989,
+      "grad_norm": 3.9759397871617006,
+      "learning_rate": 4.1287459262932164e-06,
+      "loss": 0.8681,
+      "step": 1674
+    },
+    {
+      "epoch": 0.569534172050323,
+      "grad_norm": 1.7556306149575895,
+      "learning_rate": 4.123323015626241e-06,
+      "loss": 0.9425,
+      "step": 1675
+    },
+    {
+      "epoch": 0.5698741924515471,
+      "grad_norm": 1.822886620788618,
+      "learning_rate": 4.11790116871096e-06,
+      "loss": 0.8339,
+      "step": 1676
+    },
+    {
+      "epoch": 0.5702142128527712,
+      "grad_norm": 2.32577771957017,
+      "learning_rate": 4.112480392126187e-06,
+      "loss": 0.7799,
+      "step": 1677
+    },
+    {
+      "epoch": 0.5705542332539952,
+      "grad_norm": 2.3654748461284267,
+      "learning_rate": 4.107060692449447e-06,
+      "loss": 0.7794,
+      "step": 1678
+    },
+    {
+      "epoch": 0.5708942536552193,
+      "grad_norm": 2.1326059990159,
+      "learning_rate": 4.1016420762569496e-06,
+      "loss": 0.6922,
+      "step": 1679
+    },
+    {
+      "epoch": 0.5712342740564433,
+      "grad_norm": 1.6500811511500117,
+      "learning_rate": 4.096224550123597e-06,
+      "loss": 0.9321,
+      "step": 1680
+    },
+    {
+      "epoch": 0.5715742944576675,
+      "grad_norm": 2.288324199496334,
+      "learning_rate": 4.090808120622961e-06,
+      "loss": 0.8088,
+      "step": 1681
+    },
+    {
+      "epoch": 0.5719143148588915,
+      "grad_norm": 3.335932302163143,
+      "learning_rate": 4.08539279432729e-06,
+      "loss": 0.7918,
+      "step": 1682
+    },
+    {
+      "epoch": 0.5722543352601156,
+      "grad_norm": 2.018206897778278,
+      "learning_rate": 4.079978577807487e-06,
+      "loss": 0.8091,
+      "step": 1683
+    },
+    {
+      "epoch": 0.5725943556613396,
+      "grad_norm": 1.8917366828566053,
+      "learning_rate": 4.074565477633117e-06,
+      "loss": 0.8174,
+      "step": 1684
+    },
+    {
+      "epoch": 0.5729343760625637,
+      "grad_norm": 1.6840308709452019,
+      "learning_rate": 4.069153500372382e-06,
+      "loss": 0.794,
+      "step": 1685
+    },
+    {
+      "epoch": 0.5732743964637879,
+      "grad_norm": 2.0560413539850972,
+      "learning_rate": 4.063742652592125e-06,
+      "loss": 0.8338,
+      "step": 1686
+    },
+    {
+      "epoch": 0.5736144168650119,
+      "grad_norm": 2.715704335230041,
+      "learning_rate": 4.0583329408578185e-06,
+      "loss": 0.8608,
+      "step": 1687
+    },
+    {
+      "epoch": 0.573954437266236,
+      "grad_norm": 1.8308222638540965,
+      "learning_rate": 4.052924371733555e-06,
+      "loss": 0.7391,
+      "step": 1688
+    },
+    {
+      "epoch": 0.57429445766746,
+      "grad_norm": 3.93217327404691,
+      "learning_rate": 4.047516951782046e-06,
+      "loss": 0.8336,
+      "step": 1689
+    },
+    {
+      "epoch": 0.5746344780686842,
+      "grad_norm": 1.7155299358553424,
+      "learning_rate": 4.0421106875646e-06,
+      "loss": 0.7387,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5749744984699082,
+      "grad_norm": 2.924796486558408,
+      "learning_rate": 4.036705585641131e-06,
+      "loss": 0.8656,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5753145188711323,
+      "grad_norm": 2.1154010899917015,
+      "learning_rate": 4.031301652570139e-06,
+      "loss": 0.8103,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5756545392723563,
+      "grad_norm": 1.7593975839358962,
+      "learning_rate": 4.0258988949087015e-06,
+      "loss": 0.7343,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5759945596735804,
+      "grad_norm": 7.277324615448209,
+      "learning_rate": 4.020497319212482e-06,
+      "loss": 0.9342,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5763345800748045,
+      "grad_norm": 1.8762066741865282,
+      "learning_rate": 4.015096932035695e-06,
+      "loss": 0.8569,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5766746004760286,
+      "grad_norm": 2.5094661743116227,
+      "learning_rate": 4.009697739931125e-06,
+      "loss": 0.7803,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5770146208772526,
+      "grad_norm": 2.368718744294282,
+      "learning_rate": 4.004299749450099e-06,
+      "loss": 0.7593,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5773546412784767,
+      "grad_norm": 1.7379622792872598,
+      "learning_rate": 3.99890296714249e-06,
+      "loss": 0.8102,
+      "step": 1698
+    },
+    {
+      "epoch": 0.5776946616797007,
+      "grad_norm": 1.7950661253826894,
+      "learning_rate": 3.993507399556699e-06,
+      "loss": 0.8261,
+      "step": 1699
+    },
+    {
+      "epoch": 0.5780346820809249,
+      "grad_norm": 1.7895033857380052,
+      "learning_rate": 3.988113053239664e-06,
+      "loss": 0.7831,
+      "step": 1700
+    },
+    {
+      "epoch": 0.578374702482149,
+      "grad_norm": 2.513724608663744,
+      "learning_rate": 3.982719934736832e-06,
+      "loss": 0.7863,
+      "step": 1701
+    },
+    {
+      "epoch": 0.578714722883373,
+      "grad_norm": 1.7341110425502526,
+      "learning_rate": 3.977328050592161e-06,
+      "loss": 0.9247,
+      "step": 1702
+    },
+    {
+      "epoch": 0.579054743284597,
+      "grad_norm": 1.8845071679357839,
+      "learning_rate": 3.971937407348115e-06,
+      "loss": 0.8488,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5793947636858211,
+      "grad_norm": 1.8072089808348282,
+      "learning_rate": 3.966548011545648e-06,
+      "loss": 0.7179,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5797347840870453,
+      "grad_norm": 3.0702338796288307,
+      "learning_rate": 3.961159869724207e-06,
+      "loss": 0.7202,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5800748044882693,
+      "grad_norm": 1.759787279998497,
+      "learning_rate": 3.955772988421709e-06,
+      "loss": 0.768,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5804148248894934,
+      "grad_norm": 2.3433623345553034,
+      "learning_rate": 3.950387374174548e-06,
+      "loss": 0.6933,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5807548452907174,
+      "grad_norm": 1.7657990260540302,
+      "learning_rate": 3.945003033517578e-06,
+      "loss": 0.7882,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5810948656919415,
+      "grad_norm": 1.6247662467902115,
+      "learning_rate": 3.9396199729841044e-06,
+      "loss": 0.7497,
+      "step": 1709
+    },
+    {
+      "epoch": 0.5814348860931656,
+      "grad_norm": 4.31861063498843,
+      "learning_rate": 3.934238199105887e-06,
+      "loss": 0.6626,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5817749064943897,
+      "grad_norm": 1.9596435537578243,
+      "learning_rate": 3.928857718413119e-06,
+      "loss": 0.6802,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5821149268956137,
+      "grad_norm": 2.133099365831387,
+      "learning_rate": 3.9234785374344264e-06,
+      "loss": 0.8929,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5824549472968378,
+      "grad_norm": 1.6595599877288754,
+      "learning_rate": 3.918100662696853e-06,
+      "loss": 0.9019,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5827949676980619,
+      "grad_norm": 4.129797231985507,
+      "learning_rate": 3.9127241007258695e-06,
+      "loss": 0.704,
+      "step": 1714
+    },
+    {
+      "epoch": 0.583134988099286,
+      "grad_norm": 2.2096678154176854,
+      "learning_rate": 3.907348858045338e-06,
+      "loss": 0.767,
+      "step": 1715
+    },
+    {
+      "epoch": 0.58347500850051,
+      "grad_norm": 2.050103065477324,
+      "learning_rate": 3.9019749411775336e-06,
+      "loss": 0.7629,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5838150289017341,
+      "grad_norm": 2.224362598207352,
+      "learning_rate": 3.8966023566431154e-06,
+      "loss": 0.7301,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5841550493029581,
+      "grad_norm": 3.471505729243675,
+      "learning_rate": 3.891231110961126e-06,
+      "loss": 0.8771,
+      "step": 1718
+    },
+    {
+      "epoch": 0.5844950697041823,
+      "grad_norm": 2.2233454405320017,
+      "learning_rate": 3.885861210648987e-06,
+      "loss": 0.8412,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5848350901054064,
+      "grad_norm": 1.8614586671959485,
+      "learning_rate": 3.880492662222483e-06,
+      "loss": 0.7183,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5851751105066304,
+      "grad_norm": 3.3042359139392645,
+      "learning_rate": 3.875125472195764e-06,
+      "loss": 0.7574,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5855151309078545,
+      "grad_norm": 1.6934396731563182,
+      "learning_rate": 3.869759647081326e-06,
+      "loss": 0.7454,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5858551513090785,
+      "grad_norm": 1.9354501363208503,
+      "learning_rate": 3.8643951933900125e-06,
+      "loss": 0.8003,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5861951717103027,
+      "grad_norm": 2.033785372249832,
+      "learning_rate": 3.859032117631002e-06,
+      "loss": 0.9099,
+      "step": 1724
+    },
+    {
+      "epoch": 0.5865351921115267,
+      "grad_norm": 1.6684392122823892,
+      "learning_rate": 3.853670426311797e-06,
+      "loss": 0.7391,
+      "step": 1725
+    },
+    {
+      "epoch": 0.5868752125127508,
+      "grad_norm": 2.023118010181095,
+      "learning_rate": 3.848310125938229e-06,
+      "loss": 0.8358,
+      "step": 1726
+    },
+    {
+      "epoch": 0.5872152329139748,
+      "grad_norm": 2.3269706588518604,
+      "learning_rate": 3.842951223014433e-06,
+      "loss": 0.8102,
+      "step": 1727
+    },
+    {
+      "epoch": 0.5875552533151989,
+      "grad_norm": 1.9400944184932607,
+      "learning_rate": 3.837593724042854e-06,
+      "loss": 0.7688,
+      "step": 1728
+    },
+    {
+      "epoch": 0.587895273716423,
+      "grad_norm": 1.6651797388262217,
+      "learning_rate": 3.832237635524229e-06,
+      "loss": 0.7588,
+      "step": 1729
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 1.5455784587197665,
+      "learning_rate": 3.826882963957589e-06,
+      "loss": 0.7464,
+      "step": 1730
+    },
+    {
+      "epoch": 0.5885753145188711,
+      "grad_norm": 1.85747121655971,
+      "learning_rate": 3.821529715840241e-06,
+      "loss": 0.7595,
+      "step": 1731
+    },
+    {
+      "epoch": 0.5889153349200952,
+      "grad_norm": 2.6341826537925495,
+      "learning_rate": 3.816177897667767e-06,
+      "loss": 0.7596,
+      "step": 1732
+    },
+    {
+      "epoch": 0.5892553553213192,
+      "grad_norm": 2.3495118501585717,
+      "learning_rate": 3.810827515934013e-06,
+      "loss": 0.7428,
+      "step": 1733
+    },
+    {
+      "epoch": 0.5895953757225434,
+      "grad_norm": 4.349211175598281,
+      "learning_rate": 3.8054785771310817e-06,
+      "loss": 0.7613,
+      "step": 1734
+    },
+    {
+      "epoch": 0.5899353961237674,
+      "grad_norm": 2.1363024851103876,
+      "learning_rate": 3.8001310877493265e-06,
+      "loss": 0.844,
+      "step": 1735
+    },
+    {
+      "epoch": 0.5902754165249915,
+      "grad_norm": 4.0769838432188505,
+      "learning_rate": 3.7947850542773396e-06,
+      "loss": 0.7463,
+      "step": 1736
+    },
+    {
+      "epoch": 0.5906154369262155,
+      "grad_norm": 1.6554083285238113,
+      "learning_rate": 3.7894404832019514e-06,
+      "loss": 0.8421,
+      "step": 1737
+    },
+    {
+      "epoch": 0.5909554573274396,
+      "grad_norm": 1.503047933162913,
+      "learning_rate": 3.784097381008212e-06,
+      "loss": 0.7792,
+      "step": 1738
+    },
+    {
+      "epoch": 0.5912954777286638,
+      "grad_norm": 5.221808674383001,
+      "learning_rate": 3.778755754179394e-06,
+      "loss": 0.9246,
+      "step": 1739
+    },
+    {
+      "epoch": 0.5916354981298878,
+      "grad_norm": 1.679707592266952,
+      "learning_rate": 3.7734156091969766e-06,
+      "loss": 0.8391,
+      "step": 1740
+    },
+    {
+      "epoch": 0.5919755185311119,
+      "grad_norm": 2.1866548043016145,
+      "learning_rate": 3.7680769525406398e-06,
+      "loss": 0.8404,
+      "step": 1741
+    },
+    {
+      "epoch": 0.5923155389323359,
+      "grad_norm": 1.9227767898016401,
+      "learning_rate": 3.762739790688264e-06,
+      "loss": 0.6675,
+      "step": 1742
+    },
+    {
+      "epoch": 0.5926555593335601,
+      "grad_norm": 2.2649866796850584,
+      "learning_rate": 3.757404130115909e-06,
+      "loss": 0.7401,
+      "step": 1743
+    },
+    {
+      "epoch": 0.5929955797347841,
+      "grad_norm": 1.972603607998956,
+      "learning_rate": 3.752069977297817e-06,
+      "loss": 0.7905,
+      "step": 1744
+    },
+    {
+      "epoch": 0.5933356001360082,
+      "grad_norm": 2.2111615655437995,
+      "learning_rate": 3.7467373387063973e-06,
+      "loss": 0.7023,
+      "step": 1745
+    },
+    {
+      "epoch": 0.5936756205372322,
+      "grad_norm": 1.9245235611037521,
+      "learning_rate": 3.741406220812227e-06,
+      "loss": 0.9047,
+      "step": 1746
+    },
+    {
+      "epoch": 0.5940156409384563,
+      "grad_norm": 1.733840029867424,
+      "learning_rate": 3.7360766300840323e-06,
+      "loss": 0.7679,
+      "step": 1747
+    },
+    {
+      "epoch": 0.5943556613396804,
+      "grad_norm": 1.679265103876112,
+      "learning_rate": 3.7307485729886917e-06,
+      "loss": 0.835,
+      "step": 1748
+    },
+    {
+      "epoch": 0.5946956817409045,
+      "grad_norm": 1.751220728165291,
+      "learning_rate": 3.725422055991218e-06,
+      "loss": 0.7547,
+      "step": 1749
+    },
+    {
+      "epoch": 0.5950357021421285,
+      "grad_norm": 1.8592254513227962,
+      "learning_rate": 3.720097085554756e-06,
+      "loss": 0.789,
+      "step": 1750
+    },
+    {
+      "epoch": 0.5953757225433526,
+      "grad_norm": 1.5578672257415482,
+      "learning_rate": 3.7147736681405784e-06,
+      "loss": 0.7403,
+      "step": 1751
+    },
+    {
+      "epoch": 0.5957157429445766,
+      "grad_norm": 1.7162665067510214,
+      "learning_rate": 3.709451810208068e-06,
+      "loss": 0.7607,
+      "step": 1752
+    },
+    {
+      "epoch": 0.5960557633458008,
+      "grad_norm": 1.8084933447047609,
+      "learning_rate": 3.7041315182147203e-06,
+      "loss": 0.7423,
+      "step": 1753
+    },
+    {
+      "epoch": 0.5963957837470248,
+      "grad_norm": 1.6360987508622518,
+      "learning_rate": 3.6988127986161247e-06,
+      "loss": 0.7364,
+      "step": 1754
+    },
+    {
+      "epoch": 0.5967358041482489,
+      "grad_norm": 1.620889477822385,
+      "learning_rate": 3.6934956578659697e-06,
+      "loss": 0.7419,
+      "step": 1755
+    },
+    {
+      "epoch": 0.5970758245494729,
+      "grad_norm": 2.005053928359656,
+      "learning_rate": 3.688180102416022e-06,
+      "loss": 0.8189,
+      "step": 1756
+    },
+    {
+      "epoch": 0.597415844950697,
+      "grad_norm": 2.49591536599233,
+      "learning_rate": 3.682866138716126e-06,
+      "loss": 0.7555,
+      "step": 1757
+    },
+    {
+      "epoch": 0.5977558653519212,
+      "grad_norm": 1.9099333716331244,
+      "learning_rate": 3.6775537732141986e-06,
+      "loss": 0.6334,
+      "step": 1758
+    },
+    {
+      "epoch": 0.5980958857531452,
+      "grad_norm": 1.6443659221282751,
+      "learning_rate": 3.6722430123562124e-06,
+      "loss": 0.8068,
+      "step": 1759
+    },
+    {
+      "epoch": 0.5984359061543693,
+      "grad_norm": 2.623550934402015,
+      "learning_rate": 3.6669338625861983e-06,
+      "loss": 0.8823,
+      "step": 1760
+    },
+    {
+      "epoch": 0.5987759265555933,
+      "grad_norm": 2.7847663723361222,
+      "learning_rate": 3.661626330346224e-06,
+      "loss": 0.7644,
+      "step": 1761
+    },
+    {
+      "epoch": 0.5991159469568174,
+      "grad_norm": 1.978249284886559,
+      "learning_rate": 3.656320422076406e-06,
+      "loss": 0.6911,
+      "step": 1762
+    },
+    {
+      "epoch": 0.5994559673580415,
+      "grad_norm": 1.6370605423957132,
+      "learning_rate": 3.6510161442148783e-06,
+      "loss": 0.7779,
+      "step": 1763
+    },
+    {
+      "epoch": 0.5997959877592656,
+      "grad_norm": 1.9618067800993437,
+      "learning_rate": 3.6457135031978077e-06,
+      "loss": 0.8111,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6001360081604896,
+      "grad_norm": 2.0871896457836785,
+      "learning_rate": 3.6404125054593653e-06,
+      "loss": 0.6968,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6004760285617137,
+      "grad_norm": 3.6449233310433375,
+      "learning_rate": 3.635113157431732e-06,
+      "loss": 0.9092,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6008160489629377,
+      "grad_norm": 2.43520006058048,
+      "learning_rate": 3.629815465545091e-06,
+      "loss": 0.8429,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6011560693641619,
+      "grad_norm": 2.1109713075027723,
+      "learning_rate": 3.62451943622761e-06,
+      "loss": 0.6774,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6014960897653859,
+      "grad_norm": 1.7035409015911598,
+      "learning_rate": 3.6192250759054427e-06,
+      "loss": 0.8139,
+      "step": 1769
+    },
+    {
+      "epoch": 0.60183611016661,
+      "grad_norm": 2.7589386792158628,
+      "learning_rate": 3.6139323910027136e-06,
+      "loss": 0.7972,
+      "step": 1770
+    },
+    {
+      "epoch": 0.602176130567834,
+      "grad_norm": 2.02117501459737,
+      "learning_rate": 3.608641387941523e-06,
+      "loss": 0.8181,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6025161509690582,
+      "grad_norm": 1.9303717752979608,
+      "learning_rate": 3.6033520731419214e-06,
+      "loss": 0.8203,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6028561713702822,
+      "grad_norm": 1.884365094703858,
+      "learning_rate": 3.598064453021911e-06,
+      "loss": 0.7987,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6031961917715063,
+      "grad_norm": 2.7068608465393287,
+      "learning_rate": 3.592778533997446e-06,
+      "loss": 0.7508,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6035362121727303,
+      "grad_norm": 1.987215906609273,
+      "learning_rate": 3.5874943224824097e-06,
+      "loss": 0.6987,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6038762325739544,
+      "grad_norm": 2.2665245253039443,
+      "learning_rate": 3.582211824888615e-06,
+      "loss": 0.7874,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6042162529751786,
+      "grad_norm": 2.079144255350973,
+      "learning_rate": 3.5769310476257935e-06,
+      "loss": 0.8801,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6045562733764026,
+      "grad_norm": 1.684740476192313,
+      "learning_rate": 3.5716519971015947e-06,
+      "loss": 0.8109,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6048962937776267,
+      "grad_norm": 2.356591268153879,
+      "learning_rate": 3.5663746797215658e-06,
+      "loss": 0.7333,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6052363141788507,
+      "grad_norm": 1.9527315149213702,
+      "learning_rate": 3.561099101889158e-06,
+      "loss": 0.8158,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6055763345800748,
+      "grad_norm": 2.0298727319213175,
+      "learning_rate": 3.555825270005707e-06,
+      "loss": 0.8055,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6059163549812989,
+      "grad_norm": 2.1789885450127557,
+      "learning_rate": 3.5505531904704287e-06,
+      "loss": 0.8846,
+      "step": 1782
+    },
+    {
+      "epoch": 0.606256375382523,
+      "grad_norm": 2.190682206775641,
+      "learning_rate": 3.5452828696804196e-06,
+      "loss": 0.8113,
+      "step": 1783
+    },
+    {
+      "epoch": 0.606596395783747,
+      "grad_norm": 1.7727471010909939,
+      "learning_rate": 3.5400143140306355e-06,
+      "loss": 0.8189,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6069364161849711,
+      "grad_norm": 1.9034533372715055,
+      "learning_rate": 3.5347475299138932e-06,
+      "loss": 0.8361,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6072764365861951,
+      "grad_norm": 1.7827761115461676,
+      "learning_rate": 3.5294825237208573e-06,
+      "loss": 0.7705,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6076164569874193,
+      "grad_norm": 2.130570518035593,
+      "learning_rate": 3.524219301840043e-06,
+      "loss": 0.8345,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6079564773886433,
+      "grad_norm": 2.377972509331428,
+      "learning_rate": 3.5189578706577896e-06,
+      "loss": 0.8651,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6082964977898674,
+      "grad_norm": 2.0170332375599647,
+      "learning_rate": 3.5136982365582704e-06,
+      "loss": 0.7541,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6086365181910914,
+      "grad_norm": 2.0487060631425535,
+      "learning_rate": 3.5084404059234773e-06,
+      "loss": 0.747,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6089765385923155,
+      "grad_norm": 1.5497246080487372,
+      "learning_rate": 3.5031843851332105e-06,
+      "loss": 0.7551,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6093165589935396,
+      "grad_norm": 1.8661412507465704,
+      "learning_rate": 3.4979301805650805e-06,
+      "loss": 0.7471,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6096565793947637,
+      "grad_norm": 1.9081392430588895,
+      "learning_rate": 3.492677798594486e-06,
+      "loss": 0.6867,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6099965997959877,
+      "grad_norm": 3.953731191700172,
+      "learning_rate": 3.4874272455946217e-06,
+      "loss": 0.863,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6103366201972118,
+      "grad_norm": 1.7838153201393292,
+      "learning_rate": 3.4821785279364585e-06,
+      "loss": 0.9178,
+      "step": 1795
+    },
+    {
+      "epoch": 0.610676640598436,
+      "grad_norm": 1.9609099043585159,
+      "learning_rate": 3.476931651988742e-06,
+      "loss": 0.7292,
+      "step": 1796
+    },
+    {
+      "epoch": 0.61101666099966,
+      "grad_norm": 1.7849795981254395,
+      "learning_rate": 3.471686624117982e-06,
+      "loss": 0.8199,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6113566814008841,
+      "grad_norm": 2.1270736856013337,
+      "learning_rate": 3.466443450688445e-06,
+      "loss": 0.875,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6116967018021081,
+      "grad_norm": 2.2091957079882945,
+      "learning_rate": 3.461202138062153e-06,
+      "loss": 0.6927,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6120367222033322,
+      "grad_norm": 1.5834963127345454,
+      "learning_rate": 3.4559626925988623e-06,
+      "loss": 0.7318,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6123767426045563,
+      "grad_norm": 2.0496819681593506,
+      "learning_rate": 3.450725120656069e-06,
+      "loss": 0.7506,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6127167630057804,
+      "grad_norm": 2.2013789066829,
+      "learning_rate": 3.4454894285889916e-06,
+      "loss": 0.8603,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6130567834070044,
+      "grad_norm": 2.09843751252011,
+      "learning_rate": 3.4402556227505746e-06,
+      "loss": 0.7768,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6133968038082285,
+      "grad_norm": 2.3451379947690607,
+      "learning_rate": 3.435023709491467e-06,
+      "loss": 0.8272,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6137368242094525,
+      "grad_norm": 3.5327166657343456,
+      "learning_rate": 3.4297936951600217e-06,
+      "loss": 0.7466,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6140768446106767,
+      "grad_norm": 1.905762345831794,
+      "learning_rate": 3.424565586102293e-06,
+      "loss": 0.8158,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6144168650119007,
+      "grad_norm": 2.3123581203476804,
+      "learning_rate": 3.4193393886620153e-06,
+      "loss": 0.8462,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6147568854131248,
+      "grad_norm": 1.946148455776188,
+      "learning_rate": 3.4141151091806134e-06,
+      "loss": 0.704,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6150969058143488,
+      "grad_norm": 1.3443343547386855,
+      "learning_rate": 3.408892753997175e-06,
+      "loss": 0.693,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6154369262155729,
+      "grad_norm": 2.704265394408822,
+      "learning_rate": 3.40367232944846e-06,
+      "loss": 0.7594,
+      "step": 1810
+    },
+    {
+      "epoch": 0.615776946616797,
+      "grad_norm": 1.8211554347276468,
+      "learning_rate": 3.3984538418688795e-06,
+      "loss": 0.669,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6161169670180211,
+      "grad_norm": 2.0413514509876602,
+      "learning_rate": 3.3932372975905027e-06,
+      "loss": 0.868,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6164569874192452,
+      "grad_norm": 2.121098812311723,
+      "learning_rate": 3.3880227029430335e-06,
+      "loss": 0.7582,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6167970078204692,
+      "grad_norm": 1.6909069939174246,
+      "learning_rate": 3.3828100642538097e-06,
+      "loss": 0.7327,
+      "step": 1814
+    },
+    {
+      "epoch": 0.6171370282216933,
+      "grad_norm": 1.6334734116518446,
+      "learning_rate": 3.377599387847803e-06,
+      "loss": 0.8684,
+      "step": 1815
+    },
+    {
+      "epoch": 0.6174770486229174,
+      "grad_norm": 2.1432970505365043,
+      "learning_rate": 3.372390680047597e-06,
+      "loss": 0.8207,
+      "step": 1816
+    },
+    {
+      "epoch": 0.6178170690241415,
+      "grad_norm": 1.8165425685515983,
+      "learning_rate": 3.3671839471733906e-06,
+      "loss": 0.797,
+      "step": 1817
+    },
+    {
+      "epoch": 0.6181570894253655,
+      "grad_norm": 2.1615351800791354,
+      "learning_rate": 3.3619791955429826e-06,
+      "loss": 0.773,
+      "step": 1818
+    },
+    {
+      "epoch": 0.6184971098265896,
+      "grad_norm": 2.6904615455880614,
+      "learning_rate": 3.3567764314717744e-06,
+      "loss": 0.7907,
+      "step": 1819
+    },
+    {
+      "epoch": 0.6188371302278136,
+      "grad_norm": 2.834095748254144,
+      "learning_rate": 3.351575661272749e-06,
+      "loss": 0.8717,
+      "step": 1820
+    },
+    {
+      "epoch": 0.6191771506290378,
+      "grad_norm": 2.145877874306492,
+      "learning_rate": 3.346376891256471e-06,
+      "loss": 0.8086,
+      "step": 1821
+    },
+    {
+      "epoch": 0.6195171710302618,
+      "grad_norm": 2.023497317198109,
+      "learning_rate": 3.341180127731083e-06,
+      "loss": 0.8679,
+      "step": 1822
+    },
+    {
+      "epoch": 0.6198571914314859,
+      "grad_norm": 2.1023255103173564,
+      "learning_rate": 3.335985377002285e-06,
+      "loss": 0.8146,
+      "step": 1823
+    },
+    {
+      "epoch": 0.6201972118327099,
+      "grad_norm": 1.731210236073491,
+      "learning_rate": 3.330792645373344e-06,
+      "loss": 0.7683,
+      "step": 1824
+    },
+    {
+      "epoch": 0.6205372322339341,
+      "grad_norm": 1.9233595089290563,
+      "learning_rate": 3.3256019391450696e-06,
+      "loss": 0.7869,
+      "step": 1825
+    },
+    {
+      "epoch": 0.6208772526351581,
+      "grad_norm": 2.208075651046932,
+      "learning_rate": 3.320413264615817e-06,
+      "loss": 0.6999,
+      "step": 1826
+    },
+    {
+      "epoch": 0.6212172730363822,
+      "grad_norm": 1.8428253882643195,
+      "learning_rate": 3.315226628081475e-06,
+      "loss": 0.724,
+      "step": 1827
+    },
+    {
+      "epoch": 0.6215572934376062,
+      "grad_norm": 2.5874398711101665,
+      "learning_rate": 3.3100420358354614e-06,
+      "loss": 0.8652,
+      "step": 1828
+    },
+    {
+      "epoch": 0.6218973138388303,
+      "grad_norm": 1.673959608619458,
+      "learning_rate": 3.3048594941687117e-06,
+      "loss": 0.8139,
+      "step": 1829
+    },
+    {
+      "epoch": 0.6222373342400545,
+      "grad_norm": 1.8058623104229596,
+      "learning_rate": 3.299679009369672e-06,
+      "loss": 0.7287,
+      "step": 1830
+    },
+    {
+      "epoch": 0.6225773546412785,
+      "grad_norm": 1.7517624272915204,
+      "learning_rate": 3.2945005877242975e-06,
+      "loss": 0.767,
+      "step": 1831
+    },
+    {
+      "epoch": 0.6229173750425026,
+      "grad_norm": 1.9637001633251892,
+      "learning_rate": 3.2893242355160327e-06,
+      "loss": 0.7267,
+      "step": 1832
+    },
+    {
+      "epoch": 0.6232573954437266,
+      "grad_norm": 2.0442734243489378,
+      "learning_rate": 3.28414995902582e-06,
+      "loss": 0.7311,
+      "step": 1833
+    },
+    {
+      "epoch": 0.6235974158449507,
+      "grad_norm": 4.43268536813694,
+      "learning_rate": 3.2789777645320736e-06,
+      "loss": 0.6776,
+      "step": 1834
+    },
+    {
+      "epoch": 0.6239374362461748,
+      "grad_norm": 2.120174804949898,
+      "learning_rate": 3.2738076583106903e-06,
+      "loss": 0.8519,
+      "step": 1835
+    },
+    {
+      "epoch": 0.6242774566473989,
+      "grad_norm": 3.2551292864590056,
+      "learning_rate": 3.268639646635027e-06,
+      "loss": 0.9032,
+      "step": 1836
+    },
+    {
+      "epoch": 0.6246174770486229,
+      "grad_norm": 1.601086184729794,
+      "learning_rate": 3.2634737357758994e-06,
+      "loss": 0.8248,
+      "step": 1837
+    },
+    {
+      "epoch": 0.624957497449847,
+      "grad_norm": 2.240230510804796,
+      "learning_rate": 3.2583099320015787e-06,
+      "loss": 0.7506,
+      "step": 1838
+    },
+    {
+      "epoch": 0.625297517851071,
+      "grad_norm": 1.9608041776165144,
+      "learning_rate": 3.253148241577773e-06,
+      "loss": 0.7333,
+      "step": 1839
+    },
+    {
+      "epoch": 0.6256375382522952,
+      "grad_norm": 1.9849085294573612,
+      "learning_rate": 3.2479886707676323e-06,
+      "loss": 0.7508,
+      "step": 1840
+    },
+    {
+      "epoch": 0.6259775586535192,
+      "grad_norm": 1.8228816637234933,
+      "learning_rate": 3.2428312258317306e-06,
+      "loss": 0.7946,
+      "step": 1841
+    },
+    {
+      "epoch": 0.6263175790547433,
+      "grad_norm": 1.7804154421261742,
+      "learning_rate": 3.2376759130280644e-06,
+      "loss": 0.7698,
+      "step": 1842
+    },
+    {
+      "epoch": 0.6266575994559673,
+      "grad_norm": 1.8927973749913074,
+      "learning_rate": 3.23252273861204e-06,
+      "loss": 0.8284,
+      "step": 1843
+    },
+    {
+      "epoch": 0.6269976198571914,
+      "grad_norm": 1.8550081010363115,
+      "learning_rate": 3.2273717088364743e-06,
+      "loss": 0.7924,
+      "step": 1844
+    },
+    {
+      "epoch": 0.6273376402584155,
+      "grad_norm": 2.1922949801484966,
+      "learning_rate": 3.222222829951578e-06,
+      "loss": 0.8388,
+      "step": 1845
+    },
+    {
+      "epoch": 0.6276776606596396,
+      "grad_norm": 2.1626087410744264,
+      "learning_rate": 3.2170761082049504e-06,
+      "loss": 0.7447,
+      "step": 1846
+    },
+    {
+      "epoch": 0.6280176810608636,
+      "grad_norm": 1.956904781654036,
+      "learning_rate": 3.21193154984158e-06,
+      "loss": 0.6497,
+      "step": 1847
+    },
+    {
+      "epoch": 0.6283577014620877,
+      "grad_norm": 5.531479827665114,
+      "learning_rate": 3.2067891611038203e-06,
+      "loss": 0.8239,
+      "step": 1848
+    },
+    {
+      "epoch": 0.6286977218633119,
+      "grad_norm": 2.0361393715810956,
+      "learning_rate": 3.201648948231404e-06,
+      "loss": 0.805,
+      "step": 1849
+    },
+    {
+      "epoch": 0.6290377422645359,
+      "grad_norm": 2.4082749710623523,
+      "learning_rate": 3.196510917461414e-06,
+      "loss": 0.7744,
+      "step": 1850
+    },
+    {
+      "epoch": 0.62937776266576,
+      "grad_norm": 1.8187471571807,
+      "learning_rate": 3.191375075028291e-06,
+      "loss": 0.7981,
+      "step": 1851
+    },
+    {
+      "epoch": 0.629717783066984,
+      "grad_norm": 1.8526733968838733,
+      "learning_rate": 3.1862414271638163e-06,
+      "loss": 0.7936,
+      "step": 1852
+    },
+    {
+      "epoch": 0.630057803468208,
+      "grad_norm": 2.1219438112421853,
+      "learning_rate": 3.181109980097111e-06,
+      "loss": 0.8523,
+      "step": 1853
+    },
+    {
+      "epoch": 0.6303978238694322,
+      "grad_norm": 2.3455237931427178,
+      "learning_rate": 3.1759807400546266e-06,
+      "loss": 0.7498,
+      "step": 1854
+    },
+    {
+      "epoch": 0.6307378442706563,
+      "grad_norm": 2.126668312041449,
+      "learning_rate": 3.1708537132601324e-06,
+      "loss": 0.8679,
+      "step": 1855
+    },
+    {
+      "epoch": 0.6310778646718803,
+      "grad_norm": 2.298564439126398,
+      "learning_rate": 3.1657289059347184e-06,
+      "loss": 0.7885,
+      "step": 1856
+    },
+    {
+      "epoch": 0.6314178850731044,
+      "grad_norm": 1.7168007294839274,
+      "learning_rate": 3.1606063242967753e-06,
+      "loss": 0.866,
+      "step": 1857
+    },
+    {
+      "epoch": 0.6317579054743284,
+      "grad_norm": 1.8392023959375048,
+      "learning_rate": 3.1554859745619986e-06,
+      "loss": 0.7636,
+      "step": 1858
+    },
+    {
+      "epoch": 0.6320979258755526,
+      "grad_norm": 2.2699088048706235,
+      "learning_rate": 3.15036786294337e-06,
+      "loss": 0.837,
+      "step": 1859
+    },
+    {
+      "epoch": 0.6324379462767766,
+      "grad_norm": 2.9330773051419263,
+      "learning_rate": 3.145251995651162e-06,
+      "loss": 0.8315,
+      "step": 1860
+    },
+    {
+      "epoch": 0.6327779666780007,
+      "grad_norm": 2.1837196368742133,
+      "learning_rate": 3.1401383788929175e-06,
+      "loss": 0.7574,
+      "step": 1861
+    },
+    {
+      "epoch": 0.6331179870792247,
+      "grad_norm": 1.9146680400998761,
+      "learning_rate": 3.1350270188734523e-06,
+      "loss": 0.7177,
+      "step": 1862
+    },
+    {
+      "epoch": 0.6334580074804488,
+      "grad_norm": 1.3903306685323171,
+      "learning_rate": 3.129917921794844e-06,
+      "loss": 0.693,
+      "step": 1863
+    },
+    {
+      "epoch": 0.6337980278816729,
+      "grad_norm": 2.1306437683234325,
+      "learning_rate": 3.1248110938564202e-06,
+      "loss": 0.7523,
+      "step": 1864
+    },
+    {
+      "epoch": 0.634138048282897,
+      "grad_norm": 1.9181652277534635,
+      "learning_rate": 3.1197065412547632e-06,
+      "loss": 0.8886,
+      "step": 1865
+    },
+    {
+      "epoch": 0.634478068684121,
+      "grad_norm": 1.7562861259048224,
+      "learning_rate": 3.1146042701836865e-06,
+      "loss": 0.7553,
+      "step": 1866
+    },
+    {
+      "epoch": 0.6348180890853451,
+      "grad_norm": 1.7325396538375666,
+      "learning_rate": 3.10950428683424e-06,
+      "loss": 0.8295,
+      "step": 1867
+    },
+    {
+      "epoch": 0.6351581094865691,
+      "grad_norm": 1.8009663516914018,
+      "learning_rate": 3.1044065973946945e-06,
+      "loss": 0.7395,
+      "step": 1868
+    },
+    {
+      "epoch": 0.6354981298877933,
+      "grad_norm": 1.6392428927823481,
+      "learning_rate": 3.0993112080505383e-06,
+      "loss": 0.872,
+      "step": 1869
+    },
+    {
+      "epoch": 0.6358381502890174,
+      "grad_norm": 1.5989553503629714,
+      "learning_rate": 3.0942181249844726e-06,
+      "loss": 0.6544,
+      "step": 1870
+    },
+    {
+      "epoch": 0.6361781706902414,
+      "grad_norm": 2.6768868064774374,
+      "learning_rate": 3.089127354376393e-06,
+      "loss": 0.7539,
+      "step": 1871
+    },
+    {
+      "epoch": 0.6365181910914655,
+      "grad_norm": 4.5238927522554535,
+      "learning_rate": 3.084038902403398e-06,
+      "loss": 0.7274,
+      "step": 1872
+    },
+    {
+      "epoch": 0.6368582114926895,
+      "grad_norm": 1.6009647281884545,
+      "learning_rate": 3.0789527752397624e-06,
+      "loss": 0.7217,
+      "step": 1873
+    },
+    {
+      "epoch": 0.6371982318939137,
+      "grad_norm": 2.7730101673025573,
+      "learning_rate": 3.07386897905695e-06,
+      "loss": 0.8029,
+      "step": 1874
+    },
+    {
+      "epoch": 0.6375382522951377,
+      "grad_norm": 2.6935224464307757,
+      "learning_rate": 3.068787520023587e-06,
+      "loss": 0.8331,
+      "step": 1875
+    },
+    {
+      "epoch": 0.6378782726963618,
+      "grad_norm": 2.128963112443544,
+      "learning_rate": 3.0637084043054744e-06,
+      "loss": 0.8188,
+      "step": 1876
+    },
+    {
+      "epoch": 0.6382182930975858,
+      "grad_norm": 2.21461662243339,
+      "learning_rate": 3.058631638065561e-06,
+      "loss": 0.8867,
+      "step": 1877
+    },
+    {
+      "epoch": 0.63855831349881,
+      "grad_norm": 2.0004837364007018,
+      "learning_rate": 3.0535572274639456e-06,
+      "loss": 0.7891,
+      "step": 1878
+    },
+    {
+      "epoch": 0.638898333900034,
+      "grad_norm": 2.645633375201975,
+      "learning_rate": 3.048485178657875e-06,
+      "loss": 0.7086,
+      "step": 1879
+    },
+    {
+      "epoch": 0.6392383543012581,
+      "grad_norm": 1.9850643249437119,
+      "learning_rate": 3.0434154978017215e-06,
+      "loss": 0.797,
+      "step": 1880
+    },
+    {
+      "epoch": 0.6395783747024821,
+      "grad_norm": 1.4653698856711659,
+      "learning_rate": 3.0383481910469936e-06,
+      "loss": 0.793,
+      "step": 1881
+    },
+    {
+      "epoch": 0.6399183951037062,
+      "grad_norm": 2.4247940024767525,
+      "learning_rate": 3.03328326454231e-06,
+      "loss": 0.7956,
+      "step": 1882
+    },
+    {
+      "epoch": 0.6402584155049303,
+      "grad_norm": 1.7278252319233067,
+      "learning_rate": 3.0282207244334084e-06,
+      "loss": 0.7589,
+      "step": 1883
+    },
+    {
+      "epoch": 0.6405984359061544,
+      "grad_norm": 1.7440612715525123,
+      "learning_rate": 3.0231605768631256e-06,
+      "loss": 0.8077,
+      "step": 1884
+    },
+    {
+      "epoch": 0.6409384563073784,
+      "grad_norm": 1.7115699043404717,
+      "learning_rate": 3.018102827971397e-06,
+      "loss": 0.7597,
+      "step": 1885
+    },
+    {
+      "epoch": 0.6412784767086025,
+      "grad_norm": 1.5585975933489125,
+      "learning_rate": 3.0130474838952518e-06,
+      "loss": 0.77,
+      "step": 1886
+    },
+    {
+      "epoch": 0.6416184971098265,
+      "grad_norm": 1.6776789416972602,
+      "learning_rate": 3.007994550768793e-06,
+      "loss": 0.8699,
+      "step": 1887
+    },
+    {
+      "epoch": 0.6419585175110507,
+      "grad_norm": 2.0656741398519305,
+      "learning_rate": 3.0029440347232064e-06,
+      "loss": 0.7675,
+      "step": 1888
+    },
+    {
+      "epoch": 0.6422985379122748,
+      "grad_norm": 2.423344521347543,
+      "learning_rate": 2.997895941886737e-06,
+      "loss": 0.7608,
+      "step": 1889
+    },
+    {
+      "epoch": 0.6426385583134988,
+      "grad_norm": 3.3517642400306653,
+      "learning_rate": 2.9928502783846987e-06,
+      "loss": 0.7183,
+      "step": 1890
+    },
+    {
+      "epoch": 0.6429785787147229,
+      "grad_norm": 4.199360112371653,
+      "learning_rate": 2.9878070503394484e-06,
+      "loss": 0.8042,
+      "step": 1891
+    },
+    {
+      "epoch": 0.6433185991159469,
+      "grad_norm": 2.348727029739639,
+      "learning_rate": 2.982766263870395e-06,
+      "loss": 0.8536,
+      "step": 1892
+    },
+    {
+      "epoch": 0.6436586195171711,
+      "grad_norm": 2.3407286077425815,
+      "learning_rate": 2.977727925093981e-06,
+      "loss": 0.7965,
+      "step": 1893
+    },
+    {
+      "epoch": 0.6439986399183951,
+      "grad_norm": 2.2628555536887243,
+      "learning_rate": 2.972692040123678e-06,
+      "loss": 0.8354,
+      "step": 1894
+    },
+    {
+      "epoch": 0.6443386603196192,
+      "grad_norm": 1.9098209839122338,
+      "learning_rate": 2.9676586150699843e-06,
+      "loss": 0.7583,
+      "step": 1895
+    },
+    {
+      "epoch": 0.6446786807208432,
+      "grad_norm": 2.030613992719097,
+      "learning_rate": 2.962627656040408e-06,
+      "loss": 0.6792,
+      "step": 1896
+    },
+    {
+      "epoch": 0.6450187011220673,
+      "grad_norm": 2.206570139271304,
+      "learning_rate": 2.957599169139472e-06,
+      "loss": 0.7251,
+      "step": 1897
+    },
+    {
+      "epoch": 0.6453587215232914,
+      "grad_norm": 2.2690884747822615,
+      "learning_rate": 2.9525731604686925e-06,
+      "loss": 0.7452,
+      "step": 1898
+    },
+    {
+      "epoch": 0.6456987419245155,
+      "grad_norm": 1.6036554282102649,
+      "learning_rate": 2.9475496361265834e-06,
+      "loss": 0.8009,
+      "step": 1899
+    },
+    {
+      "epoch": 0.6460387623257395,
+      "grad_norm": 1.5691756233346612,
+      "learning_rate": 2.942528602208642e-06,
+      "loss": 0.7995,
+      "step": 1900
+    },
+    {
+      "epoch": 0.6463787827269636,
+      "grad_norm": 2.5262095313197785,
+      "learning_rate": 2.9375100648073413e-06,
+      "loss": 0.7977,
+      "step": 1901
+    },
+    {
+      "epoch": 0.6467188031281876,
+      "grad_norm": 2.329442430159881,
+      "learning_rate": 2.9324940300121325e-06,
+      "loss": 0.7225,
+      "step": 1902
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 2.2795978824555316,
+      "learning_rate": 2.9274805039094225e-06,
+      "loss": 0.7248,
+      "step": 1903
+    },
+    {
+      "epoch": 0.6473988439306358,
+      "grad_norm": 1.8928729552824115,
+      "learning_rate": 2.922469492582578e-06,
+      "loss": 0.7438,
+      "step": 1904
+    },
+    {
+      "epoch": 0.6477388643318599,
+      "grad_norm": 1.645826518622666,
+      "learning_rate": 2.9174610021119136e-06,
+      "loss": 0.7018,
+      "step": 1905
+    },
+    {
+      "epoch": 0.648078884733084,
+      "grad_norm": 1.887335955741179,
+      "learning_rate": 2.912455038574686e-06,
+      "loss": 0.7666,
+      "step": 1906
+    },
+    {
+      "epoch": 0.6484189051343081,
+      "grad_norm": 2.6577609754031966,
+      "learning_rate": 2.907451608045081e-06,
+      "loss": 0.7754,
+      "step": 1907
+    },
+    {
+      "epoch": 0.6487589255355322,
+      "grad_norm": 1.424048835180028,
+      "learning_rate": 2.9024507165942196e-06,
+      "loss": 0.8108,
+      "step": 1908
+    },
+    {
+      "epoch": 0.6490989459367562,
+      "grad_norm": 2.3645854173941014,
+      "learning_rate": 2.8974523702901346e-06,
+      "loss": 0.9007,
+      "step": 1909
+    },
+    {
+      "epoch": 0.6494389663379803,
+      "grad_norm": 1.9228751348732287,
+      "learning_rate": 2.892456575197771e-06,
+      "loss": 0.8843,
+      "step": 1910
+    },
+    {
+      "epoch": 0.6497789867392043,
+      "grad_norm": 1.7424474860546162,
+      "learning_rate": 2.8874633373789848e-06,
+      "loss": 0.812,
+      "step": 1911
+    },
+    {
+      "epoch": 0.6501190071404285,
+      "grad_norm": 2.183670810574612,
+      "learning_rate": 2.8824726628925204e-06,
+      "loss": 0.844,
+      "step": 1912
+    },
+    {
+      "epoch": 0.6504590275416525,
+      "grad_norm": 3.2905500209946386,
+      "learning_rate": 2.877484557794017e-06,
+      "loss": 0.7829,
+      "step": 1913
+    },
+    {
+      "epoch": 0.6507990479428766,
+      "grad_norm": 1.8504014727826574,
+      "learning_rate": 2.872499028135993e-06,
+      "loss": 0.8476,
+      "step": 1914
+    },
+    {
+      "epoch": 0.6511390683441006,
+      "grad_norm": 1.786637343753024,
+      "learning_rate": 2.8675160799678483e-06,
+      "loss": 0.7481,
+      "step": 1915
+    },
+    {
+      "epoch": 0.6514790887453247,
+      "grad_norm": 1.7737603638634836,
+      "learning_rate": 2.8625357193358416e-06,
+      "loss": 0.6805,
+      "step": 1916
+    },
+    {
+      "epoch": 0.6518191091465488,
+      "grad_norm": 2.3513830613944897,
+      "learning_rate": 2.8575579522830965e-06,
+      "loss": 0.8911,
+      "step": 1917
+    },
+    {
+      "epoch": 0.6521591295477729,
+      "grad_norm": 2.308686347074699,
+      "learning_rate": 2.8525827848495912e-06,
+      "loss": 0.7006,
+      "step": 1918
+    },
+    {
+      "epoch": 0.6524991499489969,
+      "grad_norm": 3.0666789675701036,
+      "learning_rate": 2.847610223072145e-06,
+      "loss": 0.7716,
+      "step": 1919
+    },
+    {
+      "epoch": 0.652839170350221,
+      "grad_norm": 1.6176969102311178,
+      "learning_rate": 2.842640272984422e-06,
+      "loss": 0.7157,
+      "step": 1920
+    },
+    {
+      "epoch": 0.653179190751445,
+      "grad_norm": 1.9114007467129297,
+      "learning_rate": 2.837672940616911e-06,
+      "loss": 0.8591,
+      "step": 1921
+    },
+    {
+      "epoch": 0.6535192111526692,
+      "grad_norm": 2.673597039865379,
+      "learning_rate": 2.8327082319969268e-06,
+      "loss": 0.7577,
+      "step": 1922
+    },
+    {
+      "epoch": 0.6538592315538932,
+      "grad_norm": 1.6371578518735068,
+      "learning_rate": 2.8277461531485985e-06,
+      "loss": 0.7447,
+      "step": 1923
+    },
+    {
+      "epoch": 0.6541992519551173,
+      "grad_norm": 1.6541414287928828,
+      "learning_rate": 2.8227867100928706e-06,
+      "loss": 0.7508,
+      "step": 1924
+    },
+    {
+      "epoch": 0.6545392723563414,
+      "grad_norm": 2.6084708691300764,
+      "learning_rate": 2.8178299088474836e-06,
+      "loss": 0.7826,
+      "step": 1925
+    },
+    {
+      "epoch": 0.6548792927575654,
+      "grad_norm": 2.8225766202041194,
+      "learning_rate": 2.8128757554269716e-06,
+      "loss": 0.9343,
+      "step": 1926
+    },
+    {
+      "epoch": 0.6552193131587896,
+      "grad_norm": 1.8679165197646086,
+      "learning_rate": 2.8079242558426612e-06,
+      "loss": 0.7858,
+      "step": 1927
+    },
+    {
+      "epoch": 0.6555593335600136,
+      "grad_norm": 1.6626905717574614,
+      "learning_rate": 2.8029754161026535e-06,
+      "loss": 0.8268,
+      "step": 1928
+    },
+    {
+      "epoch": 0.6558993539612377,
+      "grad_norm": 1.7902883623187187,
+      "learning_rate": 2.7980292422118282e-06,
+      "loss": 0.7367,
+      "step": 1929
+    },
+    {
+      "epoch": 0.6562393743624617,
+      "grad_norm": 2.28189192357985,
+      "learning_rate": 2.7930857401718244e-06,
+      "loss": 0.8161,
+      "step": 1930
+    },
+    {
+      "epoch": 0.6565793947636859,
+      "grad_norm": 1.776198319138567,
+      "learning_rate": 2.7881449159810416e-06,
+      "loss": 0.6802,
+      "step": 1931
+    },
+    {
+      "epoch": 0.6569194151649099,
+      "grad_norm": 2.2447524877488156,
+      "learning_rate": 2.7832067756346293e-06,
+      "loss": 0.765,
+      "step": 1932
+    },
+    {
+      "epoch": 0.657259435566134,
+      "grad_norm": 1.626806606726708,
+      "learning_rate": 2.7782713251244797e-06,
+      "loss": 0.786,
+      "step": 1933
+    },
+    {
+      "epoch": 0.657599455967358,
+      "grad_norm": 2.2354711587791236,
+      "learning_rate": 2.7733385704392257e-06,
+      "loss": 0.832,
+      "step": 1934
+    },
+    {
+      "epoch": 0.6579394763685821,
+      "grad_norm": 2.0096403282944912,
+      "learning_rate": 2.768408517564224e-06,
+      "loss": 0.7716,
+      "step": 1935
+    },
+    {
+      "epoch": 0.6582794967698062,
+      "grad_norm": 1.9832226393988026,
+      "learning_rate": 2.763481172481556e-06,
+      "loss": 0.7496,
+      "step": 1936
+    },
+    {
+      "epoch": 0.6586195171710303,
+      "grad_norm": 1.9113293186754516,
+      "learning_rate": 2.7585565411700164e-06,
+      "loss": 0.7712,
+      "step": 1937
+    },
+    {
+      "epoch": 0.6589595375722543,
+      "grad_norm": 1.800190920309849,
+      "learning_rate": 2.7536346296051063e-06,
+      "loss": 0.775,
+      "step": 1938
+    },
+    {
+      "epoch": 0.6592995579734784,
+      "grad_norm": 2.2869209794289835,
+      "learning_rate": 2.7487154437590252e-06,
+      "loss": 0.7114,
+      "step": 1939
+    },
+    {
+      "epoch": 0.6596395783747024,
+      "grad_norm": 1.752548967169787,
+      "learning_rate": 2.743798989600672e-06,
+      "loss": 0.8391,
+      "step": 1940
+    },
+    {
+      "epoch": 0.6599795987759266,
+      "grad_norm": 1.9696733758583753,
+      "learning_rate": 2.738885273095624e-06,
+      "loss": 0.6906,
+      "step": 1941
+    },
+    {
+      "epoch": 0.6603196191771507,
+      "grad_norm": 1.756956191108405,
+      "learning_rate": 2.733974300206137e-06,
+      "loss": 0.8137,
+      "step": 1942
+    },
+    {
+      "epoch": 0.6606596395783747,
+      "grad_norm": 2.5904794588208433,
+      "learning_rate": 2.7290660768911435e-06,
+      "loss": 0.8262,
+      "step": 1943
+    },
+    {
+      "epoch": 0.6609996599795988,
+      "grad_norm": 2.718207473803795,
+      "learning_rate": 2.7241606091062334e-06,
+      "loss": 0.7992,
+      "step": 1944
+    },
+    {
+      "epoch": 0.6613396803808228,
+      "grad_norm": 1.9225267669432748,
+      "learning_rate": 2.719257902803658e-06,
+      "loss": 0.8342,
+      "step": 1945
+    },
+    {
+      "epoch": 0.661679700782047,
+      "grad_norm": 1.8175234136298053,
+      "learning_rate": 2.7143579639323146e-06,
+      "loss": 0.7721,
+      "step": 1946
+    },
+    {
+      "epoch": 0.662019721183271,
+      "grad_norm": 2.2267444521145365,
+      "learning_rate": 2.7094607984377423e-06,
+      "loss": 0.7256,
+      "step": 1947
+    },
+    {
+      "epoch": 0.6623597415844951,
+      "grad_norm": 1.6676057305902945,
+      "learning_rate": 2.7045664122621173e-06,
+      "loss": 0.7588,
+      "step": 1948
+    },
+    {
+      "epoch": 0.6626997619857191,
+      "grad_norm": 1.9495109160790398,
+      "learning_rate": 2.6996748113442397e-06,
+      "loss": 0.7012,
+      "step": 1949
+    },
+    {
+      "epoch": 0.6630397823869432,
+      "grad_norm": 2.275323709986825,
+      "learning_rate": 2.6947860016195372e-06,
+      "loss": 0.809,
+      "step": 1950
+    },
+    {
+      "epoch": 0.6633798027881673,
+      "grad_norm": 1.7705415744996646,
+      "learning_rate": 2.6898999890200405e-06,
+      "loss": 0.7813,
+      "step": 1951
+    },
+    {
+      "epoch": 0.6637198231893914,
+      "grad_norm": 1.8201284706907614,
+      "learning_rate": 2.6850167794743966e-06,
+      "loss": 0.7378,
+      "step": 1952
+    },
+    {
+      "epoch": 0.6640598435906154,
+      "grad_norm": 1.7763246698039323,
+      "learning_rate": 2.680136378907845e-06,
+      "loss": 0.8054,
+      "step": 1953
+    },
+    {
+      "epoch": 0.6643998639918395,
+      "grad_norm": 2.1067341590101787,
+      "learning_rate": 2.6752587932422175e-06,
+      "loss": 0.8473,
+      "step": 1954
+    },
+    {
+      "epoch": 0.6647398843930635,
+      "grad_norm": 2.1826612345959284,
+      "learning_rate": 2.67038402839593e-06,
+      "loss": 0.8311,
+      "step": 1955
+    },
+    {
+      "epoch": 0.6650799047942877,
+      "grad_norm": 2.151119086744043,
+      "learning_rate": 2.6655120902839802e-06,
+      "loss": 0.7625,
+      "step": 1956
+    },
+    {
+      "epoch": 0.6654199251955117,
+      "grad_norm": 2.5514025172235963,
+      "learning_rate": 2.6606429848179306e-06,
+      "loss": 0.7488,
+      "step": 1957
+    },
+    {
+      "epoch": 0.6657599455967358,
+      "grad_norm": 2.4317656380313477,
+      "learning_rate": 2.655776717905906e-06,
+      "loss": 0.7954,
+      "step": 1958
+    },
+    {
+      "epoch": 0.6660999659979598,
+      "grad_norm": 2.3399416449946364,
+      "learning_rate": 2.6509132954525946e-06,
+      "loss": 0.7008,
+      "step": 1959
+    },
+    {
+      "epoch": 0.666439986399184,
+      "grad_norm": 2.7631747409481386,
+      "learning_rate": 2.6460527233592225e-06,
+      "loss": 0.7061,
+      "step": 1960
+    },
+    {
+      "epoch": 0.666780006800408,
+      "grad_norm": 1.9223571690932921,
+      "learning_rate": 2.641195007523568e-06,
+      "loss": 0.8037,
+      "step": 1961
+    },
+    {
+      "epoch": 0.6671200272016321,
+      "grad_norm": 2.0570084331599454,
+      "learning_rate": 2.636340153839935e-06,
+      "loss": 0.7771,
+      "step": 1962
+    },
+    {
+      "epoch": 0.6674600476028562,
+      "grad_norm": 1.73207328780135,
+      "learning_rate": 2.631488168199159e-06,
+      "loss": 0.8048,
+      "step": 1963
+    },
+    {
+      "epoch": 0.6678000680040802,
+      "grad_norm": 1.7955944649193327,
+      "learning_rate": 2.626639056488593e-06,
+      "loss": 0.7144,
+      "step": 1964
+    },
+    {
+      "epoch": 0.6681400884053044,
+      "grad_norm": 1.899052873587027,
+      "learning_rate": 2.621792824592103e-06,
+      "loss": 0.8188,
+      "step": 1965
+    },
+    {
+      "epoch": 0.6684801088065284,
+      "grad_norm": 1.669623057694321,
+      "learning_rate": 2.616949478390065e-06,
+      "loss": 0.7515,
+      "step": 1966
+    },
+    {
+      "epoch": 0.6688201292077525,
+      "grad_norm": 2.6012423751773817,
+      "learning_rate": 2.612109023759346e-06,
+      "loss": 0.724,
+      "step": 1967
+    },
+    {
+      "epoch": 0.6691601496089765,
+      "grad_norm": 1.7503290521049435,
+      "learning_rate": 2.6072714665733135e-06,
+      "loss": 0.7963,
+      "step": 1968
+    },
+    {
+      "epoch": 0.6695001700102006,
+      "grad_norm": 1.7646392017997337,
+      "learning_rate": 2.60243681270181e-06,
+      "loss": 0.7704,
+      "step": 1969
+    },
+    {
+      "epoch": 0.6698401904114247,
+      "grad_norm": 2.23434001820653,
+      "learning_rate": 2.597605068011163e-06,
+      "loss": 0.7679,
+      "step": 1970
+    },
+    {
+      "epoch": 0.6701802108126488,
+      "grad_norm": 2.2030201912789242,
+      "learning_rate": 2.5927762383641657e-06,
+      "loss": 0.8307,
+      "step": 1971
+    },
+    {
+      "epoch": 0.6705202312138728,
+      "grad_norm": 1.4053187069136137,
+      "learning_rate": 2.5879503296200736e-06,
+      "loss": 0.6733,
+      "step": 1972
+    },
+    {
+      "epoch": 0.6708602516150969,
+      "grad_norm": 2.0200673415876076,
+      "learning_rate": 2.583127347634601e-06,
+      "loss": 0.7548,
+      "step": 1973
+    },
+    {
+      "epoch": 0.6712002720163209,
+      "grad_norm": 7.747179478216865,
+      "learning_rate": 2.5783072982599057e-06,
+      "loss": 0.8986,
+      "step": 1974
+    },
+    {
+      "epoch": 0.6715402924175451,
+      "grad_norm": 4.347151217698582,
+      "learning_rate": 2.573490187344596e-06,
+      "loss": 0.8158,
+      "step": 1975
+    },
+    {
+      "epoch": 0.6718803128187691,
+      "grad_norm": 1.5960191169683489,
+      "learning_rate": 2.5686760207337045e-06,
+      "loss": 0.6319,
+      "step": 1976
+    },
+    {
+      "epoch": 0.6722203332199932,
+      "grad_norm": 2.0210012377694633,
+      "learning_rate": 2.563864804268701e-06,
+      "loss": 0.8267,
+      "step": 1977
+    },
+    {
+      "epoch": 0.6725603536212172,
+      "grad_norm": 2.121999429293373,
+      "learning_rate": 2.559056543787468e-06,
+      "loss": 0.7567,
+      "step": 1978
+    },
+    {
+      "epoch": 0.6729003740224413,
+      "grad_norm": 2.5273540123412044,
+      "learning_rate": 2.554251245124305e-06,
+      "loss": 0.6441,
+      "step": 1979
+    },
+    {
+      "epoch": 0.6732403944236655,
+      "grad_norm": 2.0741394840030853,
+      "learning_rate": 2.5494489141099155e-06,
+      "loss": 0.8274,
+      "step": 1980
+    },
+    {
+      "epoch": 0.6735804148248895,
+      "grad_norm": 1.67166625283495,
+      "learning_rate": 2.5446495565714024e-06,
+      "loss": 0.7647,
+      "step": 1981
+    },
+    {
+      "epoch": 0.6739204352261136,
+      "grad_norm": 1.915541986064764,
+      "learning_rate": 2.539853178332265e-06,
+      "loss": 0.8623,
+      "step": 1982
+    },
+    {
+      "epoch": 0.6742604556273376,
+      "grad_norm": 2.0838754278048075,
+      "learning_rate": 2.5350597852123798e-06,
+      "loss": 0.9025,
+      "step": 1983
+    },
+    {
+      "epoch": 0.6746004760285618,
+      "grad_norm": 1.8777603204555824,
+      "learning_rate": 2.530269383028009e-06,
+      "loss": 0.805,
+      "step": 1984
+    },
+    {
+      "epoch": 0.6749404964297858,
+      "grad_norm": 1.6370793848887677,
+      "learning_rate": 2.5254819775917795e-06,
+      "loss": 0.7331,
+      "step": 1985
+    },
+    {
+      "epoch": 0.6752805168310099,
+      "grad_norm": 1.9510489857076145,
+      "learning_rate": 2.5206975747126873e-06,
+      "loss": 0.6924,
+      "step": 1986
+    },
+    {
+      "epoch": 0.6756205372322339,
+      "grad_norm": 1.8840509441418891,
+      "learning_rate": 2.51591618019608e-06,
+      "loss": 0.7736,
+      "step": 1987
+    },
+    {
+      "epoch": 0.675960557633458,
+      "grad_norm": 2.1944785850385866,
+      "learning_rate": 2.511137799843658e-06,
+      "loss": 0.7507,
+      "step": 1988
+    },
+    {
+      "epoch": 0.6763005780346821,
+      "grad_norm": 2.563444560543499,
+      "learning_rate": 2.506362439453463e-06,
+      "loss": 0.8487,
+      "step": 1989
+    },
+    {
+      "epoch": 0.6766405984359062,
+      "grad_norm": 1.8896121725043236,
+      "learning_rate": 2.5015901048198716e-06,
+      "loss": 0.7212,
+      "step": 1990
+    },
+    {
+      "epoch": 0.6769806188371302,
+      "grad_norm": 1.6981617236104471,
+      "learning_rate": 2.4968208017335936e-06,
+      "loss": 0.8074,
+      "step": 1991
+    },
+    {
+      "epoch": 0.6773206392383543,
+      "grad_norm": 2.1743389190119093,
+      "learning_rate": 2.4920545359816533e-06,
+      "loss": 0.8572,
+      "step": 1992
+    },
+    {
+      "epoch": 0.6776606596395783,
+      "grad_norm": 1.7914808391383161,
+      "learning_rate": 2.487291313347397e-06,
+      "loss": 0.8223,
+      "step": 1993
+    },
+    {
+      "epoch": 0.6780006800408025,
+      "grad_norm": 1.715883330092319,
+      "learning_rate": 2.4825311396104727e-06,
+      "loss": 0.8229,
+      "step": 1994
+    },
+    {
+      "epoch": 0.6783407004420265,
+      "grad_norm": 1.9340151796766085,
+      "learning_rate": 2.477774020546831e-06,
+      "loss": 0.7705,
+      "step": 1995
+    },
+    {
+      "epoch": 0.6786807208432506,
+      "grad_norm": 2.1553476660518376,
+      "learning_rate": 2.473019961928716e-06,
+      "loss": 0.8944,
+      "step": 1996
+    },
+    {
+      "epoch": 0.6790207412444746,
+      "grad_norm": 2.3845638574577306,
+      "learning_rate": 2.4682689695246557e-06,
+      "loss": 0.7879,
+      "step": 1997
+    },
+    {
+      "epoch": 0.6793607616456987,
+      "grad_norm": 2.8669737688434775,
+      "learning_rate": 2.4635210490994648e-06,
+      "loss": 0.7056,
+      "step": 1998
+    },
+    {
+      "epoch": 0.6797007820469229,
+      "grad_norm": 3.049192996998203,
+      "learning_rate": 2.458776206414221e-06,
+      "loss": 0.8073,
+      "step": 1999
+    },
+    {
+      "epoch": 0.6800408024481469,
+      "grad_norm": 1.8798273806791783,
+      "learning_rate": 2.4540344472262766e-06,
+      "loss": 0.7979,
+      "step": 2000
+    },
+    {
+      "epoch": 0.680380822849371,
+      "grad_norm": 2.8826970776464167,
+      "learning_rate": 2.4492957772892345e-06,
+      "loss": 0.7671,
+      "step": 2001
+    },
+    {
+      "epoch": 0.680720843250595,
+      "grad_norm": 1.7248005566280082,
+      "learning_rate": 2.4445602023529558e-06,
+      "loss": 0.7626,
+      "step": 2002
+    },
+    {
+      "epoch": 0.6810608636518191,
+      "grad_norm": 1.6157884725720708,
+      "learning_rate": 2.439827728163542e-06,
+      "loss": 0.6729,
+      "step": 2003
+    },
+    {
+      "epoch": 0.6814008840530432,
+      "grad_norm": 2.015980236780887,
+      "learning_rate": 2.4350983604633323e-06,
+      "loss": 0.7427,
+      "step": 2004
+    },
+    {
+      "epoch": 0.6817409044542673,
+      "grad_norm": 2.213874167082533,
+      "learning_rate": 2.4303721049908973e-06,
+      "loss": 0.8243,
+      "step": 2005
+    },
+    {
+      "epoch": 0.6820809248554913,
+      "grad_norm": 1.8548248745686642,
+      "learning_rate": 2.425648967481031e-06,
+      "loss": 0.7634,
+      "step": 2006
+    },
+    {
+      "epoch": 0.6824209452567154,
+      "grad_norm": 1.652684543766916,
+      "learning_rate": 2.4209289536647467e-06,
+      "loss": 0.8613,
+      "step": 2007
+    },
+    {
+      "epoch": 0.6827609656579394,
+      "grad_norm": 1.9971828674188044,
+      "learning_rate": 2.4162120692692623e-06,
+      "loss": 0.7493,
+      "step": 2008
+    },
+    {
+      "epoch": 0.6831009860591636,
+      "grad_norm": 1.599253674855469,
+      "learning_rate": 2.4114983200180053e-06,
+      "loss": 0.7948,
+      "step": 2009
+    },
+    {
+      "epoch": 0.6834410064603876,
+      "grad_norm": 2.2583508234748653,
+      "learning_rate": 2.406787711630591e-06,
+      "loss": 0.7357,
+      "step": 2010
+    },
+    {
+      "epoch": 0.6837810268616117,
+      "grad_norm": 2.0791026003084396,
+      "learning_rate": 2.4020802498228333e-06,
+      "loss": 0.8317,
+      "step": 2011
+    },
+    {
+      "epoch": 0.6841210472628357,
+      "grad_norm": 1.9114254449996253,
+      "learning_rate": 2.3973759403067175e-06,
+      "loss": 0.8558,
+      "step": 2012
+    },
+    {
+      "epoch": 0.6844610676640599,
+      "grad_norm": 1.6266872694061658,
+      "learning_rate": 2.3926747887904084e-06,
+      "loss": 0.8107,
+      "step": 2013
+    },
+    {
+      "epoch": 0.684801088065284,
+      "grad_norm": 2.2202773218114706,
+      "learning_rate": 2.3879768009782434e-06,
+      "loss": 0.8187,
+      "step": 2014
+    },
+    {
+      "epoch": 0.685141108466508,
+      "grad_norm": 1.8724141832862042,
+      "learning_rate": 2.3832819825707136e-06,
+      "loss": 0.7582,
+      "step": 2015
+    },
+    {
+      "epoch": 0.685481128867732,
+      "grad_norm": 2.6048840426082043,
+      "learning_rate": 2.3785903392644714e-06,
+      "loss": 0.7355,
+      "step": 2016
+    },
+    {
+      "epoch": 0.6858211492689561,
+      "grad_norm": 2.4069817743883695,
+      "learning_rate": 2.37390187675231e-06,
+      "loss": 0.8101,
+      "step": 2017
+    },
+    {
+      "epoch": 0.6861611696701803,
+      "grad_norm": 2.3122636622751997,
+      "learning_rate": 2.3692166007231686e-06,
+      "loss": 0.796,
+      "step": 2018
+    },
+    {
+      "epoch": 0.6865011900714043,
+      "grad_norm": 2.7148809354845103,
+      "learning_rate": 2.364534516862117e-06,
+      "loss": 0.7821,
+      "step": 2019
+    },
+    {
+      "epoch": 0.6868412104726284,
+      "grad_norm": 1.7589729603840554,
+      "learning_rate": 2.359855630850352e-06,
+      "loss": 0.805,
+      "step": 2020
+    },
+    {
+      "epoch": 0.6871812308738524,
+      "grad_norm": 2.188045522331186,
+      "learning_rate": 2.3551799483651894e-06,
+      "loss": 0.7042,
+      "step": 2021
+    },
+    {
+      "epoch": 0.6875212512750765,
+      "grad_norm": 1.6884288952943125,
+      "learning_rate": 2.3505074750800585e-06,
+      "loss": 0.7188,
+      "step": 2022
+    },
+    {
+      "epoch": 0.6878612716763006,
+      "grad_norm": 1.6892834429534136,
+      "learning_rate": 2.3458382166644967e-06,
+      "loss": 0.6986,
+      "step": 2023
+    },
+    {
+      "epoch": 0.6882012920775247,
+      "grad_norm": 2.0889129257298324,
+      "learning_rate": 2.3411721787841363e-06,
+      "loss": 0.671,
+      "step": 2024
+    },
+    {
+      "epoch": 0.6885413124787487,
+      "grad_norm": 1.9926930177927726,
+      "learning_rate": 2.3365093671007078e-06,
+      "loss": 0.7946,
+      "step": 2025
+    },
+    {
+      "epoch": 0.6888813328799728,
+      "grad_norm": 2.050530946497787,
+      "learning_rate": 2.3318497872720193e-06,
+      "loss": 0.7665,
+      "step": 2026
+    },
+    {
+      "epoch": 0.6892213532811968,
+      "grad_norm": 3.331008521125105,
+      "learning_rate": 2.327193444951966e-06,
+      "loss": 0.7251,
+      "step": 2027
+    },
+    {
+      "epoch": 0.689561373682421,
+      "grad_norm": 1.558219400732492,
+      "learning_rate": 2.322540345790508e-06,
+      "loss": 0.8328,
+      "step": 2028
+    },
+    {
+      "epoch": 0.689901394083645,
+      "grad_norm": 1.6259596450082816,
+      "learning_rate": 2.3178904954336718e-06,
+      "loss": 0.7147,
+      "step": 2029
+    },
+    {
+      "epoch": 0.6902414144848691,
+      "grad_norm": 1.8488785938546937,
+      "learning_rate": 2.313243899523544e-06,
+      "loss": 0.8313,
+      "step": 2030
+    },
+    {
+      "epoch": 0.6905814348860931,
+      "grad_norm": 1.8394321602635277,
+      "learning_rate": 2.3086005636982582e-06,
+      "loss": 0.8232,
+      "step": 2031
+    },
+    {
+      "epoch": 0.6909214552873172,
+      "grad_norm": 1.9494231310016805,
+      "learning_rate": 2.303960493591999e-06,
+      "loss": 0.6783,
+      "step": 2032
+    },
+    {
+      "epoch": 0.6912614756885413,
+      "grad_norm": 2.0407917593034863,
+      "learning_rate": 2.29932369483498e-06,
+      "loss": 0.8164,
+      "step": 2033
+    },
+    {
+      "epoch": 0.6916014960897654,
+      "grad_norm": 1.5657183822956182,
+      "learning_rate": 2.2946901730534533e-06,
+      "loss": 0.8238,
+      "step": 2034
+    },
+    {
+      "epoch": 0.6919415164909895,
+      "grad_norm": 2.1179360693491556,
+      "learning_rate": 2.29005993386969e-06,
+      "loss": 0.6922,
+      "step": 2035
+    },
+    {
+      "epoch": 0.6922815368922135,
+      "grad_norm": 2.838283362200117,
+      "learning_rate": 2.285432982901979e-06,
+      "loss": 0.7736,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6926215572934376,
+      "grad_norm": 1.9673971851860206,
+      "learning_rate": 2.2808093257646184e-06,
+      "loss": 0.8444,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6929615776946617,
+      "grad_norm": 2.112143965612654,
+      "learning_rate": 2.2761889680679106e-06,
+      "loss": 0.7465,
+      "step": 2038
+    },
+    {
+      "epoch": 0.6933015980958858,
+      "grad_norm": 3.5658767108987313,
+      "learning_rate": 2.271571915418157e-06,
+      "loss": 0.7382,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6936416184971098,
+      "grad_norm": 1.7890203918330567,
+      "learning_rate": 2.266958173417644e-06,
+      "loss": 0.7754,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6939816388983339,
+      "grad_norm": 1.879135172415941,
+      "learning_rate": 2.2623477476646447e-06,
+      "loss": 0.9036,
+      "step": 2041
+    },
+    {
+      "epoch": 0.694321659299558,
+      "grad_norm": 2.469200512596644,
+      "learning_rate": 2.2577406437534055e-06,
+      "loss": 0.7346,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6946616797007821,
+      "grad_norm": 1.7547922743933462,
+      "learning_rate": 2.253136867274146e-06,
+      "loss": 0.837,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6950017001020061,
+      "grad_norm": 2.0761435680953837,
+      "learning_rate": 2.2485364238130435e-06,
+      "loss": 0.7821,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6953417205032302,
+      "grad_norm": 1.868569645752557,
+      "learning_rate": 2.243939318952234e-06,
+      "loss": 0.8159,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6956817409044542,
+      "grad_norm": 2.2407320960892743,
+      "learning_rate": 2.239345558269801e-06,
+      "loss": 0.8396,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6960217613056784,
+      "grad_norm": 1.7022760601328137,
+      "learning_rate": 2.23475514733977e-06,
+      "loss": 0.817,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6963617817069024,
+      "grad_norm": 1.6301569903177133,
+      "learning_rate": 2.230168091732106e-06,
+      "loss": 0.615,
+      "step": 2048
+    },
+    {
+      "epoch": 0.6967018021081265,
+      "grad_norm": 1.9046707956588385,
+      "learning_rate": 2.2255843970126957e-06,
+      "loss": 0.7858,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6970418225093505,
+      "grad_norm": 1.9673929880121703,
+      "learning_rate": 2.221004068743356e-06,
+      "loss": 0.845,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6973818429105746,
+      "grad_norm": 1.782595694166562,
+      "learning_rate": 2.2164271124818103e-06,
+      "loss": 0.7656,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6977218633117988,
+      "grad_norm": 1.6386330711294346,
+      "learning_rate": 2.2118535337817003e-06,
+      "loss": 0.7537,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6980618837130228,
+      "grad_norm": 1.4316705818253959,
+      "learning_rate": 2.207283338192559e-06,
+      "loss": 0.6975,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6984019041142469,
+      "grad_norm": 1.5611707150455605,
+      "learning_rate": 2.2027165312598185e-06,
+      "loss": 0.6761,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6987419245154709,
+      "grad_norm": 1.5355495259096092,
+      "learning_rate": 2.1981531185248034e-06,
+      "loss": 0.6972,
+      "step": 2055
+    },
+    {
+      "epoch": 0.699081944916695,
+      "grad_norm": 1.9515735201013045,
+      "learning_rate": 2.1935931055247127e-06,
+      "loss": 0.7739,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6994219653179191,
+      "grad_norm": 1.707681607294952,
+      "learning_rate": 2.1890364977926283e-06,
+      "loss": 0.8014,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6997619857191432,
+      "grad_norm": 1.9416088276270818,
+      "learning_rate": 2.18448330085749e-06,
+      "loss": 0.7066,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7001020061203672,
+      "grad_norm": 1.816549981430375,
+      "learning_rate": 2.1799335202441104e-06,
+      "loss": 0.8464,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7004420265215913,
+      "grad_norm": 1.9376066291343164,
+      "learning_rate": 2.1753871614731474e-06,
+      "loss": 0.7222,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7007820469228153,
+      "grad_norm": 2.118492372862069,
+      "learning_rate": 2.1708442300611115e-06,
+      "loss": 0.7918,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7011220673240395,
+      "grad_norm": 2.0839760418604376,
+      "learning_rate": 2.1663047315203533e-06,
+      "loss": 0.8174,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7014620877252635,
+      "grad_norm": 2.3407523269538313,
+      "learning_rate": 2.1617686713590557e-06,
+      "loss": 0.7331,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7018021081264876,
+      "grad_norm": 2.3029622189762757,
+      "learning_rate": 2.1572360550812354e-06,
+      "loss": 0.8031,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7021421285277116,
+      "grad_norm": 2.1267697856713657,
+      "learning_rate": 2.1527068881867243e-06,
+      "loss": 0.7973,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7024821489289358,
+      "grad_norm": 1.8215858343618574,
+      "learning_rate": 2.148181176171174e-06,
+      "loss": 0.8117,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7028221693301598,
+      "grad_norm": 1.4157717285401026,
+      "learning_rate": 2.1436589245260375e-06,
+      "loss": 0.9047,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7031621897313839,
+      "grad_norm": 1.7979159155184197,
+      "learning_rate": 2.1391401387385773e-06,
+      "loss": 0.8326,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7035022101326079,
+      "grad_norm": 2.0914051675300835,
+      "learning_rate": 2.134624824291846e-06,
+      "loss": 0.8622,
+      "step": 2069
+    },
+    {
+      "epoch": 0.703842230533832,
+      "grad_norm": 2.1070541206963056,
+      "learning_rate": 2.1301129866646774e-06,
+      "loss": 0.8943,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7041822509350562,
+      "grad_norm": 2.13463784164494,
+      "learning_rate": 2.1256046313317002e-06,
+      "loss": 0.8321,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7045222713362802,
+      "grad_norm": 2.5136301018463754,
+      "learning_rate": 2.1210997637633067e-06,
+      "loss": 0.7691,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7048622917375043,
+      "grad_norm": 2.2541601501295783,
+      "learning_rate": 2.1165983894256647e-06,
+      "loss": 0.7222,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7052023121387283,
+      "grad_norm": 2.98901817786412,
+      "learning_rate": 2.1121005137806964e-06,
+      "loss": 0.7528,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7055423325399524,
+      "grad_norm": 1.8580295018583073,
+      "learning_rate": 2.1076061422860862e-06,
+      "loss": 0.7779,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 2.2207128533601983,
+      "learning_rate": 2.1031152803952605e-06,
+      "loss": 0.7673,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7062223733424006,
+      "grad_norm": 1.6735932994012201,
+      "learning_rate": 2.098627933557389e-06,
+      "loss": 0.8363,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7065623937436246,
+      "grad_norm": 1.5680321057110584,
+      "learning_rate": 2.0941441072173766e-06,
+      "loss": 0.7077,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7069024141448487,
+      "grad_norm": 1.8545485537631567,
+      "learning_rate": 2.089663806815856e-06,
+      "loss": 0.7437,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7072424345460727,
+      "grad_norm": 1.700070379165538,
+      "learning_rate": 2.085187037789184e-06,
+      "loss": 0.7529,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7075824549472969,
+      "grad_norm": 2.092387228234501,
+      "learning_rate": 2.080713805569427e-06,
+      "loss": 0.8072,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7079224753485209,
+      "grad_norm": 1.6576023640763133,
+      "learning_rate": 2.0762441155843678e-06,
+      "loss": 0.8356,
+      "step": 2082
+    },
+    {
+      "epoch": 0.708262495749745,
+      "grad_norm": 2.096722406636028,
+      "learning_rate": 2.071777973257482e-06,
+      "loss": 0.6847,
+      "step": 2083
+    },
+    {
+      "epoch": 0.708602516150969,
+      "grad_norm": 2.647186031515588,
+      "learning_rate": 2.0673153840079502e-06,
+      "loss": 0.8209,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7089425365521931,
+      "grad_norm": 1.772423654028428,
+      "learning_rate": 2.0628563532506334e-06,
+      "loss": 0.7487,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7092825569534172,
+      "grad_norm": 2.006215820360962,
+      "learning_rate": 2.058400886396079e-06,
+      "loss": 0.7881,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7096225773546413,
+      "grad_norm": 1.5471023818172427,
+      "learning_rate": 2.053948988850508e-06,
+      "loss": 0.8533,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7099625977558653,
+      "grad_norm": 1.765248777545704,
+      "learning_rate": 2.0495006660158113e-06,
+      "loss": 0.6848,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7103026181570894,
+      "grad_norm": 1.6747092561465118,
+      "learning_rate": 2.045055923289544e-06,
+      "loss": 0.7101,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7106426385583134,
+      "grad_norm": 1.9042522058118778,
+      "learning_rate": 2.040614766064913e-06,
+      "loss": 0.7763,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7109826589595376,
+      "grad_norm": 6.255027159940595,
+      "learning_rate": 2.036177199730781e-06,
+      "loss": 0.85,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7113226793607617,
+      "grad_norm": 1.622002495881479,
+      "learning_rate": 2.0317432296716475e-06,
+      "loss": 0.727,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7116626997619857,
+      "grad_norm": 1.691104238150303,
+      "learning_rate": 2.0273128612676506e-06,
+      "loss": 0.7537,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7120027201632098,
+      "grad_norm": 1.7231893078223932,
+      "learning_rate": 2.0228860998945577e-06,
+      "loss": 0.7642,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7123427405644339,
+      "grad_norm": 2.1757614848576345,
+      "learning_rate": 2.0184629509237583e-06,
+      "loss": 0.8258,
+      "step": 2095
+    },
+    {
+      "epoch": 0.712682760965658,
+      "grad_norm": 1.7874598523749379,
+      "learning_rate": 2.0140434197222647e-06,
+      "loss": 0.8447,
+      "step": 2096
+    },
+    {
+      "epoch": 0.713022781366882,
+      "grad_norm": 1.4437605429731617,
+      "learning_rate": 2.00962751165269e-06,
+      "loss": 0.8244,
+      "step": 2097
+    },
+    {
+      "epoch": 0.7133628017681061,
+      "grad_norm": 2.0064545785073773,
+      "learning_rate": 2.00521523207326e-06,
+      "loss": 0.7231,
+      "step": 2098
+    },
+    {
+      "epoch": 0.7137028221693301,
+      "grad_norm": 1.8662037760559647,
+      "learning_rate": 2.0008065863377903e-06,
+      "loss": 0.7298,
+      "step": 2099
+    },
+    {
+      "epoch": 0.7140428425705543,
+      "grad_norm": 1.864146642080098,
+      "learning_rate": 1.996401579795693e-06,
+      "loss": 0.8252,
+      "step": 2100
+    },
+    {
+      "epoch": 0.7143828629717783,
+      "grad_norm": 2.131314441947248,
+      "learning_rate": 1.9920002177919622e-06,
+      "loss": 0.8641,
+      "step": 2101
+    },
+    {
+      "epoch": 0.7147228833730024,
+      "grad_norm": 1.6478848825280967,
+      "learning_rate": 1.987602505667169e-06,
+      "loss": 0.8314,
+      "step": 2102
+    },
+    {
+      "epoch": 0.7150629037742264,
+      "grad_norm": 1.6965083554911622,
+      "learning_rate": 1.983208448757455e-06,
+      "loss": 0.7798,
+      "step": 2103
+    },
+    {
+      "epoch": 0.7154029241754505,
+      "grad_norm": 2.2399140677663514,
+      "learning_rate": 1.978818052394528e-06,
+      "loss": 0.7183,
+      "step": 2104
+    },
+    {
+      "epoch": 0.7157429445766746,
+      "grad_norm": 2.052486306229749,
+      "learning_rate": 1.974431321905656e-06,
+      "loss": 0.6849,
+      "step": 2105
+    },
+    {
+      "epoch": 0.7160829649778987,
+      "grad_norm": 2.3591721996625266,
+      "learning_rate": 1.9700482626136548e-06,
+      "loss": 0.8216,
+      "step": 2106
+    },
+    {
+      "epoch": 0.7164229853791227,
+      "grad_norm": 1.7604095436525182,
+      "learning_rate": 1.9656688798368905e-06,
+      "loss": 0.6536,
+      "step": 2107
+    },
+    {
+      "epoch": 0.7167630057803468,
+      "grad_norm": 1.9615351287468117,
+      "learning_rate": 1.9612931788892637e-06,
+      "loss": 0.7843,
+      "step": 2108
+    },
+    {
+      "epoch": 0.7171030261815708,
+      "grad_norm": 2.0592818898356997,
+      "learning_rate": 1.956921165080208e-06,
+      "loss": 0.7863,
+      "step": 2109
+    },
+    {
+      "epoch": 0.717443046582795,
+      "grad_norm": 2.8058446206181253,
+      "learning_rate": 1.9525528437146886e-06,
+      "loss": 0.7241,
+      "step": 2110
+    },
+    {
+      "epoch": 0.7177830669840191,
+      "grad_norm": 1.9945581002621289,
+      "learning_rate": 1.9481882200931794e-06,
+      "loss": 0.903,
+      "step": 2111
+    },
+    {
+      "epoch": 0.7181230873852431,
+      "grad_norm": 2.180468609375338,
+      "learning_rate": 1.94382729951168e-06,
+      "loss": 0.8265,
+      "step": 2112
+    },
+    {
+      "epoch": 0.7184631077864672,
+      "grad_norm": 2.623414185405911,
+      "learning_rate": 1.9394700872616856e-06,
+      "loss": 0.7801,
+      "step": 2113
+    },
+    {
+      "epoch": 0.7188031281876912,
+      "grad_norm": 1.7582604702565359,
+      "learning_rate": 1.9351165886302026e-06,
+      "loss": 0.802,
+      "step": 2114
+    },
+    {
+      "epoch": 0.7191431485889154,
+      "grad_norm": 2.5152283185802533,
+      "learning_rate": 1.9307668088997206e-06,
+      "loss": 0.8063,
+      "step": 2115
+    },
+    {
+      "epoch": 0.7194831689901394,
+      "grad_norm": 2.7666858070153437,
+      "learning_rate": 1.9264207533482264e-06,
+      "loss": 0.7285,
+      "step": 2116
+    },
+    {
+      "epoch": 0.7198231893913635,
+      "grad_norm": 2.069299953932944,
+      "learning_rate": 1.922078427249181e-06,
+      "loss": 0.6232,
+      "step": 2117
+    },
+    {
+      "epoch": 0.7201632097925875,
+      "grad_norm": 1.9669276644036997,
+      "learning_rate": 1.917739835871523e-06,
+      "loss": 0.8435,
+      "step": 2118
+    },
+    {
+      "epoch": 0.7205032301938117,
+      "grad_norm": 2.352513634957127,
+      "learning_rate": 1.9134049844796583e-06,
+      "loss": 0.7437,
+      "step": 2119
+    },
+    {
+      "epoch": 0.7208432505950357,
+      "grad_norm": 1.7010870421085014,
+      "learning_rate": 1.9090738783334535e-06,
+      "loss": 0.8334,
+      "step": 2120
+    },
+    {
+      "epoch": 0.7211832709962598,
+      "grad_norm": 1.8286429902827022,
+      "learning_rate": 1.904746522688236e-06,
+      "loss": 0.8278,
+      "step": 2121
+    },
+    {
+      "epoch": 0.7215232913974838,
+      "grad_norm": 2.204105230258611,
+      "learning_rate": 1.9004229227947752e-06,
+      "loss": 0.7401,
+      "step": 2122
+    },
+    {
+      "epoch": 0.7218633117987079,
+      "grad_norm": 2.1051613276517003,
+      "learning_rate": 1.896103083899291e-06,
+      "loss": 0.7424,
+      "step": 2123
+    },
+    {
+      "epoch": 0.722203332199932,
+      "grad_norm": 1.913948718818119,
+      "learning_rate": 1.891787011243434e-06,
+      "loss": 0.9045,
+      "step": 2124
+    },
+    {
+      "epoch": 0.7225433526011561,
+      "grad_norm": 1.5026786148699889,
+      "learning_rate": 1.8874747100642844e-06,
+      "loss": 0.7094,
+      "step": 2125
+    },
+    {
+      "epoch": 0.7228833730023801,
+      "grad_norm": 2.46070363197272,
+      "learning_rate": 1.8831661855943517e-06,
+      "loss": 0.7417,
+      "step": 2126
+    },
+    {
+      "epoch": 0.7232233934036042,
+      "grad_norm": 1.7241794429729518,
+      "learning_rate": 1.8788614430615582e-06,
+      "loss": 0.7184,
+      "step": 2127
+    },
+    {
+      "epoch": 0.7235634138048282,
+      "grad_norm": 1.7709850180647637,
+      "learning_rate": 1.8745604876892376e-06,
+      "loss": 0.8362,
+      "step": 2128
+    },
+    {
+      "epoch": 0.7239034342060524,
+      "grad_norm": 2.062995631686198,
+      "learning_rate": 1.8702633246961282e-06,
+      "loss": 0.7107,
+      "step": 2129
+    },
+    {
+      "epoch": 0.7242434546072765,
+      "grad_norm": 1.7639145613774367,
+      "learning_rate": 1.8659699592963705e-06,
+      "loss": 0.7337,
+      "step": 2130
+    },
+    {
+      "epoch": 0.7245834750085005,
+      "grad_norm": 1.6485206528522929,
+      "learning_rate": 1.8616803966994912e-06,
+      "loss": 0.7445,
+      "step": 2131
+    },
+    {
+      "epoch": 0.7249234954097246,
+      "grad_norm": 2.2432367551430485,
+      "learning_rate": 1.8573946421104082e-06,
+      "loss": 0.7921,
+      "step": 2132
+    },
+    {
+      "epoch": 0.7252635158109486,
+      "grad_norm": 2.4352161818113176,
+      "learning_rate": 1.8531127007294159e-06,
+      "loss": 0.8438,
+      "step": 2133
+    },
+    {
+      "epoch": 0.7256035362121728,
+      "grad_norm": 1.7595604425268574,
+      "learning_rate": 1.8488345777521804e-06,
+      "loss": 0.7116,
+      "step": 2134
+    },
+    {
+      "epoch": 0.7259435566133968,
+      "grad_norm": 1.9847525687521061,
+      "learning_rate": 1.8445602783697375e-06,
+      "loss": 0.7425,
+      "step": 2135
+    },
+    {
+      "epoch": 0.7262835770146209,
+      "grad_norm": 2.6735934681474363,
+      "learning_rate": 1.8402898077684806e-06,
+      "loss": 0.7244,
+      "step": 2136
+    },
+    {
+      "epoch": 0.7266235974158449,
+      "grad_norm": 1.9110660509672215,
+      "learning_rate": 1.8360231711301618e-06,
+      "loss": 0.8195,
+      "step": 2137
+    },
+    {
+      "epoch": 0.726963617817069,
+      "grad_norm": 2.047331681525153,
+      "learning_rate": 1.8317603736318746e-06,
+      "loss": 0.8365,
+      "step": 2138
+    },
+    {
+      "epoch": 0.7273036382182931,
+      "grad_norm": 1.6690303524368486,
+      "learning_rate": 1.8275014204460623e-06,
+      "loss": 0.7402,
+      "step": 2139
+    },
+    {
+      "epoch": 0.7276436586195172,
+      "grad_norm": 3.9059200473743965,
+      "learning_rate": 1.8232463167404968e-06,
+      "loss": 0.7426,
+      "step": 2140
+    },
+    {
+      "epoch": 0.7279836790207412,
+      "grad_norm": 1.932200172688164,
+      "learning_rate": 1.818995067678279e-06,
+      "loss": 0.7984,
+      "step": 2141
+    },
+    {
+      "epoch": 0.7283236994219653,
+      "grad_norm": 1.666195338393341,
+      "learning_rate": 1.8147476784178398e-06,
+      "loss": 0.8126,
+      "step": 2142
+    },
+    {
+      "epoch": 0.7286637198231893,
+      "grad_norm": 1.3626844223285068,
+      "learning_rate": 1.8105041541129187e-06,
+      "loss": 0.7733,
+      "step": 2143
+    },
+    {
+      "epoch": 0.7290037402244135,
+      "grad_norm": 2.5378369721423577,
+      "learning_rate": 1.8062644999125694e-06,
+      "loss": 0.7947,
+      "step": 2144
+    },
+    {
+      "epoch": 0.7293437606256375,
+      "grad_norm": 2.071953765717228,
+      "learning_rate": 1.8020287209611464e-06,
+      "loss": 0.7677,
+      "step": 2145
+    },
+    {
+      "epoch": 0.7296837810268616,
+      "grad_norm": 2.161502417367569,
+      "learning_rate": 1.7977968223983089e-06,
+      "loss": 0.8209,
+      "step": 2146
+    },
+    {
+      "epoch": 0.7300238014280857,
+      "grad_norm": 1.736684016690494,
+      "learning_rate": 1.7935688093589987e-06,
+      "loss": 0.7406,
+      "step": 2147
+    },
+    {
+      "epoch": 0.7303638218293098,
+      "grad_norm": 1.6090760426244202,
+      "learning_rate": 1.789344686973452e-06,
+      "loss": 0.7293,
+      "step": 2148
+    },
+    {
+      "epoch": 0.7307038422305339,
+      "grad_norm": 1.5720198527469798,
+      "learning_rate": 1.785124460367177e-06,
+      "loss": 0.7335,
+      "step": 2149
+    },
+    {
+      "epoch": 0.7310438626317579,
+      "grad_norm": 2.8678996816857842,
+      "learning_rate": 1.7809081346609574e-06,
+      "loss": 0.7563,
+      "step": 2150
+    },
+    {
+      "epoch": 0.731383883032982,
+      "grad_norm": 1.7011052505024924,
+      "learning_rate": 1.7766957149708442e-06,
+      "loss": 0.805,
+      "step": 2151
+    },
+    {
+      "epoch": 0.731723903434206,
+      "grad_norm": 2.179333898822292,
+      "learning_rate": 1.7724872064081461e-06,
+      "loss": 0.7768,
+      "step": 2152
+    },
+    {
+      "epoch": 0.7320639238354302,
+      "grad_norm": 1.9214766322552161,
+      "learning_rate": 1.768282614079432e-06,
+      "loss": 0.8926,
+      "step": 2153
+    },
+    {
+      "epoch": 0.7324039442366542,
+      "grad_norm": 1.6016132177586981,
+      "learning_rate": 1.7640819430865113e-06,
+      "loss": 0.7477,
+      "step": 2154
+    },
+    {
+      "epoch": 0.7327439646378783,
+      "grad_norm": 1.7429702206342037,
+      "learning_rate": 1.7598851985264426e-06,
+      "loss": 0.7465,
+      "step": 2155
+    },
+    {
+      "epoch": 0.7330839850391023,
+      "grad_norm": 3.780369551041426,
+      "learning_rate": 1.7556923854915148e-06,
+      "loss": 0.8328,
+      "step": 2156
+    },
+    {
+      "epoch": 0.7334240054403264,
+      "grad_norm": 5.582776086223021,
+      "learning_rate": 1.7515035090692466e-06,
+      "loss": 0.7591,
+      "step": 2157
+    },
+    {
+      "epoch": 0.7337640258415505,
+      "grad_norm": 1.75407838644776,
+      "learning_rate": 1.7473185743423853e-06,
+      "loss": 0.8702,
+      "step": 2158
+    },
+    {
+      "epoch": 0.7341040462427746,
+      "grad_norm": 3.88944154897513,
+      "learning_rate": 1.74313758638889e-06,
+      "loss": 0.7773,
+      "step": 2159
+    },
+    {
+      "epoch": 0.7344440666439986,
+      "grad_norm": 1.7976911792803083,
+      "learning_rate": 1.7389605502819324e-06,
+      "loss": 0.7521,
+      "step": 2160
+    },
+    {
+      "epoch": 0.7347840870452227,
+      "grad_norm": 2.645410770587803,
+      "learning_rate": 1.734787471089887e-06,
+      "loss": 0.7529,
+      "step": 2161
+    },
+    {
+      "epoch": 0.7351241074464467,
+      "grad_norm": 2.2156116956693133,
+      "learning_rate": 1.730618353876334e-06,
+      "loss": 0.7916,
+      "step": 2162
+    },
+    {
+      "epoch": 0.7354641278476709,
+      "grad_norm": 2.6170696483726616,
+      "learning_rate": 1.726453203700037e-06,
+      "loss": 0.8291,
+      "step": 2163
+    },
+    {
+      "epoch": 0.735804148248895,
+      "grad_norm": 1.7461843428392605,
+      "learning_rate": 1.7222920256149544e-06,
+      "loss": 0.8014,
+      "step": 2164
+    },
+    {
+      "epoch": 0.736144168650119,
+      "grad_norm": 2.1471971587915606,
+      "learning_rate": 1.7181348246702184e-06,
+      "loss": 0.908,
+      "step": 2165
+    },
+    {
+      "epoch": 0.736484189051343,
+      "grad_norm": 1.6110057508169202,
+      "learning_rate": 1.7139816059101372e-06,
+      "loss": 0.8735,
+      "step": 2166
+    },
+    {
+      "epoch": 0.7368242094525671,
+      "grad_norm": 1.7252936307832583,
+      "learning_rate": 1.7098323743741906e-06,
+      "loss": 0.8038,
+      "step": 2167
+    },
+    {
+      "epoch": 0.7371642298537913,
+      "grad_norm": 1.4751364286474202,
+      "learning_rate": 1.705687135097016e-06,
+      "loss": 0.7546,
+      "step": 2168
+    },
+    {
+      "epoch": 0.7375042502550153,
+      "grad_norm": 1.8792991681150337,
+      "learning_rate": 1.7015458931084084e-06,
+      "loss": 0.7799,
+      "step": 2169
+    },
+    {
+      "epoch": 0.7378442706562394,
+      "grad_norm": 1.9139596611161407,
+      "learning_rate": 1.69740865343331e-06,
+      "loss": 0.7647,
+      "step": 2170
+    },
+    {
+      "epoch": 0.7381842910574634,
+      "grad_norm": 2.90776955170561,
+      "learning_rate": 1.6932754210918133e-06,
+      "loss": 0.8511,
+      "step": 2171
+    },
+    {
+      "epoch": 0.7385243114586875,
+      "grad_norm": 1.593958716915834,
+      "learning_rate": 1.689146201099141e-06,
+      "loss": 0.7145,
+      "step": 2172
+    },
+    {
+      "epoch": 0.7388643318599116,
+      "grad_norm": 2.13248225387608,
+      "learning_rate": 1.6850209984656497e-06,
+      "loss": 0.8358,
+      "step": 2173
+    },
+    {
+      "epoch": 0.7392043522611357,
+      "grad_norm": 1.5208790512501984,
+      "learning_rate": 1.6808998181968238e-06,
+      "loss": 0.8051,
+      "step": 2174
+    },
+    {
+      "epoch": 0.7395443726623597,
+      "grad_norm": 1.6847664492358727,
+      "learning_rate": 1.6767826652932651e-06,
+      "loss": 0.6941,
+      "step": 2175
+    },
+    {
+      "epoch": 0.7398843930635838,
+      "grad_norm": 2.169697482882676,
+      "learning_rate": 1.6726695447506873e-06,
+      "loss": 0.6935,
+      "step": 2176
+    },
+    {
+      "epoch": 0.7402244134648079,
+      "grad_norm": 2.4175688054476634,
+      "learning_rate": 1.6685604615599117e-06,
+      "loss": 0.7608,
+      "step": 2177
+    },
+    {
+      "epoch": 0.740564433866032,
+      "grad_norm": 1.9499646893695743,
+      "learning_rate": 1.6644554207068642e-06,
+      "loss": 0.6843,
+      "step": 2178
+    },
+    {
+      "epoch": 0.740904454267256,
+      "grad_norm": 3.3649722434795937,
+      "learning_rate": 1.6603544271725607e-06,
+      "loss": 0.7127,
+      "step": 2179
+    },
+    {
+      "epoch": 0.7412444746684801,
+      "grad_norm": 1.9340583992381386,
+      "learning_rate": 1.656257485933111e-06,
+      "loss": 0.6918,
+      "step": 2180
+    },
+    {
+      "epoch": 0.7415844950697041,
+      "grad_norm": 2.1990875141504214,
+      "learning_rate": 1.652164601959705e-06,
+      "loss": 0.7244,
+      "step": 2181
+    },
+    {
+      "epoch": 0.7419245154709283,
+      "grad_norm": 1.6090430714552022,
+      "learning_rate": 1.648075780218607e-06,
+      "loss": 0.7215,
+      "step": 2182
+    },
+    {
+      "epoch": 0.7422645358721524,
+      "grad_norm": 2.1536726028566986,
+      "learning_rate": 1.6439910256711595e-06,
+      "loss": 0.7827,
+      "step": 2183
+    },
+    {
+      "epoch": 0.7426045562733764,
+      "grad_norm": 2.073708232272958,
+      "learning_rate": 1.6399103432737635e-06,
+      "loss": 0.7036,
+      "step": 2184
+    },
+    {
+      "epoch": 0.7429445766746005,
+      "grad_norm": 2.593918036362238,
+      "learning_rate": 1.635833737977881e-06,
+      "loss": 0.8041,
+      "step": 2185
+    },
+    {
+      "epoch": 0.7432845970758245,
+      "grad_norm": 2.958691737480463,
+      "learning_rate": 1.631761214730026e-06,
+      "loss": 0.7823,
+      "step": 2186
+    },
+    {
+      "epoch": 0.7436246174770487,
+      "grad_norm": 2.511544350856816,
+      "learning_rate": 1.6276927784717628e-06,
+      "loss": 0.8576,
+      "step": 2187
+    },
+    {
+      "epoch": 0.7439646378782727,
+      "grad_norm": 2.9344433102702556,
+      "learning_rate": 1.623628434139693e-06,
+      "loss": 0.8594,
+      "step": 2188
+    },
+    {
+      "epoch": 0.7443046582794968,
+      "grad_norm": 1.8620840535756518,
+      "learning_rate": 1.6195681866654517e-06,
+      "loss": 0.7979,
+      "step": 2189
+    },
+    {
+      "epoch": 0.7446446786807208,
+      "grad_norm": 2.3243589724742857,
+      "learning_rate": 1.6155120409757096e-06,
+      "loss": 0.8446,
+      "step": 2190
+    },
+    {
+      "epoch": 0.7449846990819449,
+      "grad_norm": 3.875803179972586,
+      "learning_rate": 1.6114600019921538e-06,
+      "loss": 0.8572,
+      "step": 2191
+    },
+    {
+      "epoch": 0.745324719483169,
+      "grad_norm": 1.7545550229086833,
+      "learning_rate": 1.6074120746314915e-06,
+      "loss": 0.8357,
+      "step": 2192
+    },
+    {
+      "epoch": 0.7456647398843931,
+      "grad_norm": 2.7660506150712343,
+      "learning_rate": 1.6033682638054376e-06,
+      "loss": 0.7566,
+      "step": 2193
+    },
+    {
+      "epoch": 0.7460047602856171,
+      "grad_norm": 2.208599221707356,
+      "learning_rate": 1.5993285744207183e-06,
+      "loss": 0.8451,
+      "step": 2194
+    },
+    {
+      "epoch": 0.7463447806868412,
+      "grad_norm": 1.6557633542709016,
+      "learning_rate": 1.5952930113790516e-06,
+      "loss": 0.7773,
+      "step": 2195
+    },
+    {
+      "epoch": 0.7466848010880652,
+      "grad_norm": 1.992062183353188,
+      "learning_rate": 1.5912615795771557e-06,
+      "loss": 0.8025,
+      "step": 2196
+    },
+    {
+      "epoch": 0.7470248214892894,
+      "grad_norm": 2.6192247388362597,
+      "learning_rate": 1.5872342839067305e-06,
+      "loss": 0.7836,
+      "step": 2197
+    },
+    {
+      "epoch": 0.7473648418905134,
+      "grad_norm": 1.6643325962419526,
+      "learning_rate": 1.5832111292544571e-06,
+      "loss": 0.6351,
+      "step": 2198
+    },
+    {
+      "epoch": 0.7477048622917375,
+      "grad_norm": 1.7740267079319336,
+      "learning_rate": 1.5791921205019984e-06,
+      "loss": 0.7505,
+      "step": 2199
+    },
+    {
+      "epoch": 0.7480448826929615,
+      "grad_norm": 3.3050676756035293,
+      "learning_rate": 1.5751772625259787e-06,
+      "loss": 0.871,
+      "step": 2200
+    },
+    {
+      "epoch": 0.7483849030941857,
+      "grad_norm": 2.1038805512690875,
+      "learning_rate": 1.571166560197991e-06,
+      "loss": 0.7385,
+      "step": 2201
+    },
+    {
+      "epoch": 0.7487249234954098,
+      "grad_norm": 2.7781533096347144,
+      "learning_rate": 1.567160018384582e-06,
+      "loss": 0.6283,
+      "step": 2202
+    },
+    {
+      "epoch": 0.7490649438966338,
+      "grad_norm": 1.8302184287905139,
+      "learning_rate": 1.563157641947255e-06,
+      "loss": 0.7385,
+      "step": 2203
+    },
+    {
+      "epoch": 0.7494049642978579,
+      "grad_norm": 2.5596522162134363,
+      "learning_rate": 1.5591594357424555e-06,
+      "loss": 0.8839,
+      "step": 2204
+    },
+    {
+      "epoch": 0.7497449846990819,
+      "grad_norm": 2.120816300981687,
+      "learning_rate": 1.555165404621567e-06,
+      "loss": 0.7999,
+      "step": 2205
+    },
+    {
+      "epoch": 0.7500850051003061,
+      "grad_norm": 2.3376616281310407,
+      "learning_rate": 1.5511755534309143e-06,
+      "loss": 0.8791,
+      "step": 2206
+    },
+    {
+      "epoch": 0.7504250255015301,
+      "grad_norm": 1.9094661848594028,
+      "learning_rate": 1.5471898870117414e-06,
+      "loss": 0.8485,
+      "step": 2207
+    },
+    {
+      "epoch": 0.7507650459027542,
+      "grad_norm": 1.7917052477602944,
+      "learning_rate": 1.5432084102002243e-06,
+      "loss": 0.7979,
+      "step": 2208
+    },
+    {
+      "epoch": 0.7511050663039782,
+      "grad_norm": 2.0422812529030425,
+      "learning_rate": 1.539231127827443e-06,
+      "loss": 0.789,
+      "step": 2209
+    },
+    {
+      "epoch": 0.7514450867052023,
+      "grad_norm": 2.671426125801568,
+      "learning_rate": 1.5352580447194e-06,
+      "loss": 0.7645,
+      "step": 2210
+    },
+    {
+      "epoch": 0.7517851071064264,
+      "grad_norm": 2.107626624470792,
+      "learning_rate": 1.5312891656969936e-06,
+      "loss": 0.7658,
+      "step": 2211
+    },
+    {
+      "epoch": 0.7521251275076505,
+      "grad_norm": 1.8158680120697543,
+      "learning_rate": 1.5273244955760286e-06,
+      "loss": 0.7953,
+      "step": 2212
+    },
+    {
+      "epoch": 0.7524651479088745,
+      "grad_norm": 1.978438021845548,
+      "learning_rate": 1.5233640391671973e-06,
+      "loss": 0.8217,
+      "step": 2213
+    },
+    {
+      "epoch": 0.7528051683100986,
+      "grad_norm": 2.2654036793743955,
+      "learning_rate": 1.5194078012760781e-06,
+      "loss": 0.7727,
+      "step": 2214
+    },
+    {
+      "epoch": 0.7531451887113226,
+      "grad_norm": 1.654479252128216,
+      "learning_rate": 1.5154557867031378e-06,
+      "loss": 0.7005,
+      "step": 2215
+    },
+    {
+      "epoch": 0.7534852091125468,
+      "grad_norm": 1.9670315140819592,
+      "learning_rate": 1.511508000243711e-06,
+      "loss": 0.8233,
+      "step": 2216
+    },
+    {
+      "epoch": 0.7538252295137708,
+      "grad_norm": 1.6308197393473027,
+      "learning_rate": 1.5075644466880063e-06,
+      "loss": 0.7718,
+      "step": 2217
+    },
+    {
+      "epoch": 0.7541652499149949,
+      "grad_norm": 1.9162779769547262,
+      "learning_rate": 1.5036251308210926e-06,
+      "loss": 0.7515,
+      "step": 2218
+    },
+    {
+      "epoch": 0.754505270316219,
+      "grad_norm": 2.9000813161046404,
+      "learning_rate": 1.4996900574229022e-06,
+      "loss": 0.8187,
+      "step": 2219
+    },
+    {
+      "epoch": 0.754845290717443,
+      "grad_norm": 2.216886994384,
+      "learning_rate": 1.4957592312682157e-06,
+      "loss": 0.7672,
+      "step": 2220
+    },
+    {
+      "epoch": 0.7551853111186672,
+      "grad_norm": 1.586989789116563,
+      "learning_rate": 1.4918326571266584e-06,
+      "loss": 0.7531,
+      "step": 2221
+    },
+    {
+      "epoch": 0.7555253315198912,
+      "grad_norm": 1.8489815625024457,
+      "learning_rate": 1.4879103397627027e-06,
+      "loss": 0.7646,
+      "step": 2222
+    },
+    {
+      "epoch": 0.7558653519211153,
+      "grad_norm": 1.8544337375478892,
+      "learning_rate": 1.4839922839356484e-06,
+      "loss": 0.7514,
+      "step": 2223
+    },
+    {
+      "epoch": 0.7562053723223393,
+      "grad_norm": 2.344521934054793,
+      "learning_rate": 1.4800784943996316e-06,
+      "loss": 0.7807,
+      "step": 2224
+    },
+    {
+      "epoch": 0.7565453927235634,
+      "grad_norm": 2.653629676503456,
+      "learning_rate": 1.4761689759036058e-06,
+      "loss": 0.8042,
+      "step": 2225
+    },
+    {
+      "epoch": 0.7568854131247875,
+      "grad_norm": 1.6279456912028547,
+      "learning_rate": 1.4722637331913447e-06,
+      "loss": 0.6855,
+      "step": 2226
+    },
+    {
+      "epoch": 0.7572254335260116,
+      "grad_norm": 2.235296588127983,
+      "learning_rate": 1.4683627710014325e-06,
+      "loss": 0.7996,
+      "step": 2227
+    },
+    {
+      "epoch": 0.7575654539272356,
+      "grad_norm": 1.9873774972236244,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.7425,
+      "step": 2228
+    },
+    {
+      "epoch": 0.7579054743284597,
+      "grad_norm": 1.8195331106101174,
+      "learning_rate": 1.4605737071170257e-06,
+      "loss": 0.7902,
+      "step": 2229
+    },
+    {
+      "epoch": 0.7582454947296838,
+      "grad_norm": 2.0299601573336705,
+      "learning_rate": 1.4566856148737057e-06,
+      "loss": 0.6815,
+      "step": 2230
+    },
+    {
+      "epoch": 0.7585855151309079,
+      "grad_norm": 1.400935242806698,
+      "learning_rate": 1.452801822055081e-06,
+      "loss": 0.7916,
+      "step": 2231
+    },
+    {
+      "epoch": 0.7589255355321319,
+      "grad_norm": 1.7386391395292276,
+      "learning_rate": 1.4489223333737084e-06,
+      "loss": 0.8002,
+      "step": 2232
+    },
+    {
+      "epoch": 0.759265555933356,
+      "grad_norm": 1.73667800084222,
+      "learning_rate": 1.4450471535369225e-06,
+      "loss": 0.7085,
+      "step": 2233
+    },
+    {
+      "epoch": 0.75960557633458,
+      "grad_norm": 1.739383492547159,
+      "learning_rate": 1.44117628724683e-06,
+      "loss": 0.8655,
+      "step": 2234
+    },
+    {
+      "epoch": 0.7599455967358042,
+      "grad_norm": 1.8392317788494839,
+      "learning_rate": 1.437309739200306e-06,
+      "loss": 0.7253,
+      "step": 2235
+    },
+    {
+      "epoch": 0.7602856171370282,
+      "grad_norm": 1.546424794533531,
+      "learning_rate": 1.4334475140889813e-06,
+      "loss": 0.7947,
+      "step": 2236
+    },
+    {
+      "epoch": 0.7606256375382523,
+      "grad_norm": 2.339707672874242,
+      "learning_rate": 1.4295896165992473e-06,
+      "loss": 0.8063,
+      "step": 2237
+    },
+    {
+      "epoch": 0.7609656579394763,
+      "grad_norm": 1.6335877702583506,
+      "learning_rate": 1.4257360514122393e-06,
+      "loss": 0.7126,
+      "step": 2238
+    },
+    {
+      "epoch": 0.7613056783407004,
+      "grad_norm": 1.9927357472606342,
+      "learning_rate": 1.4218868232038351e-06,
+      "loss": 0.8169,
+      "step": 2239
+    },
+    {
+      "epoch": 0.7616456987419246,
+      "grad_norm": 1.9336783631243173,
+      "learning_rate": 1.4180419366446568e-06,
+      "loss": 0.8019,
+      "step": 2240
+    },
+    {
+      "epoch": 0.7619857191431486,
+      "grad_norm": 1.8763655884472532,
+      "learning_rate": 1.4142013964000513e-06,
+      "loss": 0.8054,
+      "step": 2241
+    },
+    {
+      "epoch": 0.7623257395443727,
+      "grad_norm": 2.025901859532698,
+      "learning_rate": 1.4103652071300945e-06,
+      "loss": 0.8657,
+      "step": 2242
+    },
+    {
+      "epoch": 0.7626657599455967,
+      "grad_norm": 1.9142987689118731,
+      "learning_rate": 1.4065333734895815e-06,
+      "loss": 0.8067,
+      "step": 2243
+    },
+    {
+      "epoch": 0.7630057803468208,
+      "grad_norm": 1.8673978585090811,
+      "learning_rate": 1.4027059001280269e-06,
+      "loss": 0.7602,
+      "step": 2244
+    },
+    {
+      "epoch": 0.7633458007480449,
+      "grad_norm": 1.5244016754272622,
+      "learning_rate": 1.3988827916896491e-06,
+      "loss": 0.761,
+      "step": 2245
+    },
+    {
+      "epoch": 0.763685821149269,
+      "grad_norm": 1.6762856131051267,
+      "learning_rate": 1.3950640528133713e-06,
+      "loss": 0.8457,
+      "step": 2246
+    },
+    {
+      "epoch": 0.764025841550493,
+      "grad_norm": 1.696468220368342,
+      "learning_rate": 1.3912496881328185e-06,
+      "loss": 0.6888,
+      "step": 2247
+    },
+    {
+      "epoch": 0.7643658619517171,
+      "grad_norm": 1.807989274352168,
+      "learning_rate": 1.3874397022763024e-06,
+      "loss": 0.7174,
+      "step": 2248
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 2.1049766319752674,
+      "learning_rate": 1.3836340998668284e-06,
+      "loss": 0.8443,
+      "step": 2249
+    },
+    {
+      "epoch": 0.7650459027541653,
+      "grad_norm": 4.147430162429283,
+      "learning_rate": 1.379832885522074e-06,
+      "loss": 0.8293,
+      "step": 2250
+    },
+    {
+      "epoch": 0.7653859231553893,
+      "grad_norm": 2.203536974887446,
+      "learning_rate": 1.3760360638544012e-06,
+      "loss": 0.766,
+      "step": 2251
+    },
+    {
+      "epoch": 0.7657259435566134,
+      "grad_norm": 2.84334808411539,
+      "learning_rate": 1.3722436394708349e-06,
+      "loss": 0.7397,
+      "step": 2252
+    },
+    {
+      "epoch": 0.7660659639578374,
+      "grad_norm": 2.4606094795939875,
+      "learning_rate": 1.3684556169730706e-06,
+      "loss": 0.8327,
+      "step": 2253
+    },
+    {
+      "epoch": 0.7664059843590616,
+      "grad_norm": 3.801030903497411,
+      "learning_rate": 1.3646720009574582e-06,
+      "loss": 0.7486,
+      "step": 2254
+    },
+    {
+      "epoch": 0.7667460047602856,
+      "grad_norm": 2.2370175205395,
+      "learning_rate": 1.3608927960150008e-06,
+      "loss": 0.8567,
+      "step": 2255
+    },
+    {
+      "epoch": 0.7670860251615097,
+      "grad_norm": 1.8816253225209865,
+      "learning_rate": 1.3571180067313539e-06,
+      "loss": 0.8999,
+      "step": 2256
+    },
+    {
+      "epoch": 0.7674260455627337,
+      "grad_norm": 2.383535780477955,
+      "learning_rate": 1.3533476376868088e-06,
+      "loss": 0.7714,
+      "step": 2257
+    },
+    {
+      "epoch": 0.7677660659639578,
+      "grad_norm": 1.672812334875347,
+      "learning_rate": 1.3495816934562976e-06,
+      "loss": 0.7594,
+      "step": 2258
+    },
+    {
+      "epoch": 0.768106086365182,
+      "grad_norm": 1.771870165986916,
+      "learning_rate": 1.3458201786093795e-06,
+      "loss": 0.8326,
+      "step": 2259
+    },
+    {
+      "epoch": 0.768446106766406,
+      "grad_norm": 1.681066649638471,
+      "learning_rate": 1.3420630977102455e-06,
+      "loss": 0.6953,
+      "step": 2260
+    },
+    {
+      "epoch": 0.7687861271676301,
+      "grad_norm": 2.0034594737419833,
+      "learning_rate": 1.3383104553177001e-06,
+      "loss": 0.8098,
+      "step": 2261
+    },
+    {
+      "epoch": 0.7691261475688541,
+      "grad_norm": 1.9711809712880015,
+      "learning_rate": 1.334562255985164e-06,
+      "loss": 0.6649,
+      "step": 2262
+    },
+    {
+      "epoch": 0.7694661679700782,
+      "grad_norm": 1.7828734904005346,
+      "learning_rate": 1.3308185042606698e-06,
+      "loss": 0.8103,
+      "step": 2263
+    },
+    {
+      "epoch": 0.7698061883713023,
+      "grad_norm": 1.8609944720601692,
+      "learning_rate": 1.3270792046868486e-06,
+      "loss": 0.8238,
+      "step": 2264
+    },
+    {
+      "epoch": 0.7701462087725264,
+      "grad_norm": 1.7934456884377525,
+      "learning_rate": 1.323344361800934e-06,
+      "loss": 0.7776,
+      "step": 2265
+    },
+    {
+      "epoch": 0.7704862291737504,
+      "grad_norm": 1.7275631998115004,
+      "learning_rate": 1.3196139801347485e-06,
+      "loss": 0.8547,
+      "step": 2266
+    },
+    {
+      "epoch": 0.7708262495749745,
+      "grad_norm": 2.7409306203098707,
+      "learning_rate": 1.3158880642147026e-06,
+      "loss": 0.7145,
+      "step": 2267
+    },
+    {
+      "epoch": 0.7711662699761985,
+      "grad_norm": 2.087778768589785,
+      "learning_rate": 1.3121666185617859e-06,
+      "loss": 0.726,
+      "step": 2268
+    },
+    {
+      "epoch": 0.7715062903774227,
+      "grad_norm": 1.780483085521547,
+      "learning_rate": 1.3084496476915698e-06,
+      "loss": 0.8217,
+      "step": 2269
+    },
+    {
+      "epoch": 0.7718463107786467,
+      "grad_norm": 2.8825219877755415,
+      "learning_rate": 1.3047371561141903e-06,
+      "loss": 0.8109,
+      "step": 2270
+    },
+    {
+      "epoch": 0.7721863311798708,
+      "grad_norm": 1.9288735242516986,
+      "learning_rate": 1.3010291483343478e-06,
+      "loss": 0.812,
+      "step": 2271
+    },
+    {
+      "epoch": 0.7725263515810948,
+      "grad_norm": 2.431431619077142,
+      "learning_rate": 1.2973256288513086e-06,
+      "loss": 0.8052,
+      "step": 2272
+    },
+    {
+      "epoch": 0.7728663719823189,
+      "grad_norm": 1.7830394043812279,
+      "learning_rate": 1.2936266021588872e-06,
+      "loss": 0.6701,
+      "step": 2273
+    },
+    {
+      "epoch": 0.773206392383543,
+      "grad_norm": 2.105756285208052,
+      "learning_rate": 1.2899320727454472e-06,
+      "loss": 0.82,
+      "step": 2274
+    },
+    {
+      "epoch": 0.7735464127847671,
+      "grad_norm": 5.425208376062677,
+      "learning_rate": 1.2862420450938955e-06,
+      "loss": 0.743,
+      "step": 2275
+    },
+    {
+      "epoch": 0.7738864331859912,
+      "grad_norm": 1.827529432109907,
+      "learning_rate": 1.28255652368168e-06,
+      "loss": 0.7528,
+      "step": 2276
+    },
+    {
+      "epoch": 0.7742264535872152,
+      "grad_norm": 2.0046945708038524,
+      "learning_rate": 1.2788755129807767e-06,
+      "loss": 0.8446,
+      "step": 2277
+    },
+    {
+      "epoch": 0.7745664739884393,
+      "grad_norm": 1.947473671747887,
+      "learning_rate": 1.2751990174576883e-06,
+      "loss": 0.7801,
+      "step": 2278
+    },
+    {
+      "epoch": 0.7749064943896634,
+      "grad_norm": 1.730669713034862,
+      "learning_rate": 1.2715270415734425e-06,
+      "loss": 0.7308,
+      "step": 2279
+    },
+    {
+      "epoch": 0.7752465147908875,
+      "grad_norm": 2.0448368740503335,
+      "learning_rate": 1.2678595897835788e-06,
+      "loss": 0.7239,
+      "step": 2280
+    },
+    {
+      "epoch": 0.7755865351921115,
+      "grad_norm": 1.7996589672894716,
+      "learning_rate": 1.2641966665381517e-06,
+      "loss": 0.8383,
+      "step": 2281
+    },
+    {
+      "epoch": 0.7759265555933356,
+      "grad_norm": 1.9501431444735777,
+      "learning_rate": 1.2605382762817164e-06,
+      "loss": 0.7416,
+      "step": 2282
+    },
+    {
+      "epoch": 0.7762665759945597,
+      "grad_norm": 2.44528318223147,
+      "learning_rate": 1.2568844234533294e-06,
+      "loss": 0.867,
+      "step": 2283
+    },
+    {
+      "epoch": 0.7766065963957838,
+      "grad_norm": 1.5135812296404374,
+      "learning_rate": 1.253235112486541e-06,
+      "loss": 0.7896,
+      "step": 2284
+    },
+    {
+      "epoch": 0.7769466167970078,
+      "grad_norm": 1.6910504402544813,
+      "learning_rate": 1.249590347809393e-06,
+      "loss": 0.8092,
+      "step": 2285
+    },
+    {
+      "epoch": 0.7772866371982319,
+      "grad_norm": 2.0907001787083623,
+      "learning_rate": 1.2459501338444085e-06,
+      "loss": 0.7999,
+      "step": 2286
+    },
+    {
+      "epoch": 0.7776266575994559,
+      "grad_norm": 1.7084265208677907,
+      "learning_rate": 1.2423144750085875e-06,
+      "loss": 0.8109,
+      "step": 2287
+    },
+    {
+      "epoch": 0.7779666780006801,
+      "grad_norm": 1.7912086159226683,
+      "learning_rate": 1.2386833757134076e-06,
+      "loss": 0.7468,
+      "step": 2288
+    },
+    {
+      "epoch": 0.7783066984019041,
+      "grad_norm": 1.9542261253674484,
+      "learning_rate": 1.2350568403648088e-06,
+      "loss": 0.7268,
+      "step": 2289
+    },
+    {
+      "epoch": 0.7786467188031282,
+      "grad_norm": 1.4777106097697732,
+      "learning_rate": 1.2314348733631958e-06,
+      "loss": 0.7642,
+      "step": 2290
+    },
+    {
+      "epoch": 0.7789867392043522,
+      "grad_norm": 2.1056584699669814,
+      "learning_rate": 1.2278174791034281e-06,
+      "loss": 0.8599,
+      "step": 2291
+    },
+    {
+      "epoch": 0.7793267596055763,
+      "grad_norm": 2.1244138537440156,
+      "learning_rate": 1.224204661974821e-06,
+      "loss": 0.7469,
+      "step": 2292
+    },
+    {
+      "epoch": 0.7796667800068005,
+      "grad_norm": 1.842651124665579,
+      "learning_rate": 1.2205964263611325e-06,
+      "loss": 0.7238,
+      "step": 2293
+    },
+    {
+      "epoch": 0.7800068004080245,
+      "grad_norm": 9.972026665603295,
+      "learning_rate": 1.2169927766405598e-06,
+      "loss": 0.753,
+      "step": 2294
+    },
+    {
+      "epoch": 0.7803468208092486,
+      "grad_norm": 1.8250005499709825,
+      "learning_rate": 1.2133937171857406e-06,
+      "loss": 0.7459,
+      "step": 2295
+    },
+    {
+      "epoch": 0.7806868412104726,
+      "grad_norm": 1.4938745812322698,
+      "learning_rate": 1.2097992523637387e-06,
+      "loss": 0.741,
+      "step": 2296
+    },
+    {
+      "epoch": 0.7810268616116967,
+      "grad_norm": 1.8847996493887464,
+      "learning_rate": 1.2062093865360458e-06,
+      "loss": 0.8275,
+      "step": 2297
+    },
+    {
+      "epoch": 0.7813668820129208,
+      "grad_norm": 1.7228031928755019,
+      "learning_rate": 1.2026241240585702e-06,
+      "loss": 0.75,
+      "step": 2298
+    },
+    {
+      "epoch": 0.7817069024141449,
+      "grad_norm": 2.501687535032944,
+      "learning_rate": 1.1990434692816367e-06,
+      "loss": 0.7823,
+      "step": 2299
+    },
+    {
+      "epoch": 0.7820469228153689,
+      "grad_norm": 2.7653925468074614,
+      "learning_rate": 1.1954674265499773e-06,
+      "loss": 0.681,
+      "step": 2300
+    },
+    {
+      "epoch": 0.782386943216593,
+      "grad_norm": 2.3143819049767864,
+      "learning_rate": 1.1918960002027308e-06,
+      "loss": 0.8237,
+      "step": 2301
+    },
+    {
+      "epoch": 0.782726963617817,
+      "grad_norm": 1.5824914893100979,
+      "learning_rate": 1.1883291945734315e-06,
+      "loss": 0.7691,
+      "step": 2302
+    },
+    {
+      "epoch": 0.7830669840190412,
+      "grad_norm": 2.2301539841962708,
+      "learning_rate": 1.1847670139900074e-06,
+      "loss": 0.7281,
+      "step": 2303
+    },
+    {
+      "epoch": 0.7834070044202652,
+      "grad_norm": 2.4915256532784738,
+      "learning_rate": 1.1812094627747777e-06,
+      "loss": 0.7732,
+      "step": 2304
+    },
+    {
+      "epoch": 0.7837470248214893,
+      "grad_norm": 2.2299044959118848,
+      "learning_rate": 1.1776565452444389e-06,
+      "loss": 0.7285,
+      "step": 2305
+    },
+    {
+      "epoch": 0.7840870452227133,
+      "grad_norm": 1.8798158077309424,
+      "learning_rate": 1.174108265710071e-06,
+      "loss": 0.8002,
+      "step": 2306
+    },
+    {
+      "epoch": 0.7844270656239375,
+      "grad_norm": 2.035534060132518,
+      "learning_rate": 1.1705646284771227e-06,
+      "loss": 0.729,
+      "step": 2307
+    },
+    {
+      "epoch": 0.7847670860251615,
+      "grad_norm": 1.7825151695794803,
+      "learning_rate": 1.1670256378454093e-06,
+      "loss": 0.7919,
+      "step": 2308
+    },
+    {
+      "epoch": 0.7851071064263856,
+      "grad_norm": 2.0755250446855404,
+      "learning_rate": 1.1634912981091096e-06,
+      "loss": 0.801,
+      "step": 2309
+    },
+    {
+      "epoch": 0.7854471268276096,
+      "grad_norm": 1.9145697285689294,
+      "learning_rate": 1.159961613556757e-06,
+      "loss": 0.7888,
+      "step": 2310
+    },
+    {
+      "epoch": 0.7857871472288337,
+      "grad_norm": 2.071282232646433,
+      "learning_rate": 1.1564365884712409e-06,
+      "loss": 0.8008,
+      "step": 2311
+    },
+    {
+      "epoch": 0.7861271676300579,
+      "grad_norm": 1.8475994726429972,
+      "learning_rate": 1.1529162271297912e-06,
+      "loss": 0.7505,
+      "step": 2312
+    },
+    {
+      "epoch": 0.7864671880312819,
+      "grad_norm": 1.8531724346083438,
+      "learning_rate": 1.1494005338039839e-06,
+      "loss": 0.7435,
+      "step": 2313
+    },
+    {
+      "epoch": 0.786807208432506,
+      "grad_norm": 2.362438597585831,
+      "learning_rate": 1.1458895127597275e-06,
+      "loss": 0.7681,
+      "step": 2314
+    },
+    {
+      "epoch": 0.78714722883373,
+      "grad_norm": 2.2288635557885463,
+      "learning_rate": 1.1423831682572623e-06,
+      "loss": 0.7871,
+      "step": 2315
+    },
+    {
+      "epoch": 0.7874872492349541,
+      "grad_norm": 1.5469503925739951,
+      "learning_rate": 1.1388815045511525e-06,
+      "loss": 0.7279,
+      "step": 2316
+    },
+    {
+      "epoch": 0.7878272696361782,
+      "grad_norm": 2.0731404202763626,
+      "learning_rate": 1.1353845258902867e-06,
+      "loss": 0.788,
+      "step": 2317
+    },
+    {
+      "epoch": 0.7881672900374023,
+      "grad_norm": 2.262319333024703,
+      "learning_rate": 1.131892236517866e-06,
+      "loss": 0.6889,
+      "step": 2318
+    },
+    {
+      "epoch": 0.7885073104386263,
+      "grad_norm": 2.157562568854031,
+      "learning_rate": 1.1284046406713994e-06,
+      "loss": 0.6274,
+      "step": 2319
+    },
+    {
+      "epoch": 0.7888473308398504,
+      "grad_norm": 1.8322331540862034,
+      "learning_rate": 1.1249217425827063e-06,
+      "loss": 0.7697,
+      "step": 2320
+    },
+    {
+      "epoch": 0.7891873512410744,
+      "grad_norm": 1.9705120789647137,
+      "learning_rate": 1.1214435464779006e-06,
+      "loss": 0.8014,
+      "step": 2321
+    },
+    {
+      "epoch": 0.7895273716422986,
+      "grad_norm": 1.620994185215544,
+      "learning_rate": 1.117970056577395e-06,
+      "loss": 0.7806,
+      "step": 2322
+    },
+    {
+      "epoch": 0.7898673920435226,
+      "grad_norm": 2.1814551902341357,
+      "learning_rate": 1.1145012770958885e-06,
+      "loss": 0.7728,
+      "step": 2323
+    },
+    {
+      "epoch": 0.7902074124447467,
+      "grad_norm": 2.2535509074314417,
+      "learning_rate": 1.1110372122423663e-06,
+      "loss": 0.7814,
+      "step": 2324
+    },
+    {
+      "epoch": 0.7905474328459707,
+      "grad_norm": 2.814674945307446,
+      "learning_rate": 1.107577866220092e-06,
+      "loss": 0.7463,
+      "step": 2325
+    },
+    {
+      "epoch": 0.7908874532471948,
+      "grad_norm": 1.9527766493519834,
+      "learning_rate": 1.104123243226603e-06,
+      "loss": 0.7806,
+      "step": 2326
+    },
+    {
+      "epoch": 0.7912274736484189,
+      "grad_norm": 2.6070147128679895,
+      "learning_rate": 1.1006733474537095e-06,
+      "loss": 0.802,
+      "step": 2327
+    },
+    {
+      "epoch": 0.791567494049643,
+      "grad_norm": 1.9013207094703433,
+      "learning_rate": 1.0972281830874794e-06,
+      "loss": 0.8648,
+      "step": 2328
+    },
+    {
+      "epoch": 0.791907514450867,
+      "grad_norm": 2.0738357697149494,
+      "learning_rate": 1.0937877543082464e-06,
+      "loss": 0.6966,
+      "step": 2329
+    },
+    {
+      "epoch": 0.7922475348520911,
+      "grad_norm": 3.3351856836112437,
+      "learning_rate": 1.090352065290593e-06,
+      "loss": 0.7704,
+      "step": 2330
+    },
+    {
+      "epoch": 0.7925875552533151,
+      "grad_norm": 2.4139686701986545,
+      "learning_rate": 1.086921120203353e-06,
+      "loss": 0.8781,
+      "step": 2331
+    },
+    {
+      "epoch": 0.7929275756545393,
+      "grad_norm": 2.0367586641793785,
+      "learning_rate": 1.0834949232096008e-06,
+      "loss": 0.7859,
+      "step": 2332
+    },
+    {
+      "epoch": 0.7932675960557634,
+      "grad_norm": 1.731703445699507,
+      "learning_rate": 1.0800734784666556e-06,
+      "loss": 0.654,
+      "step": 2333
+    },
+    {
+      "epoch": 0.7936076164569874,
+      "grad_norm": 1.9761986748758829,
+      "learning_rate": 1.076656790126065e-06,
+      "loss": 0.8221,
+      "step": 2334
+    },
+    {
+      "epoch": 0.7939476368582115,
+      "grad_norm": 2.089346557643535,
+      "learning_rate": 1.0732448623336057e-06,
+      "loss": 0.7591,
+      "step": 2335
+    },
+    {
+      "epoch": 0.7942876572594356,
+      "grad_norm": 2.0490768682382074,
+      "learning_rate": 1.0698376992292808e-06,
+      "loss": 0.8476,
+      "step": 2336
+    },
+    {
+      "epoch": 0.7946276776606597,
+      "grad_norm": 1.7404822303267078,
+      "learning_rate": 1.0664353049473085e-06,
+      "loss": 0.8059,
+      "step": 2337
+    },
+    {
+      "epoch": 0.7949676980618837,
+      "grad_norm": 1.9516697066879696,
+      "learning_rate": 1.0630376836161248e-06,
+      "loss": 0.7247,
+      "step": 2338
+    },
+    {
+      "epoch": 0.7953077184631078,
+      "grad_norm": 1.7179222033684476,
+      "learning_rate": 1.0596448393583709e-06,
+      "loss": 0.7071,
+      "step": 2339
+    },
+    {
+      "epoch": 0.7956477388643318,
+      "grad_norm": 2.089155897647102,
+      "learning_rate": 1.0562567762908915e-06,
+      "loss": 0.7622,
+      "step": 2340
+    },
+    {
+      "epoch": 0.795987759265556,
+      "grad_norm": 2.0146843017178906,
+      "learning_rate": 1.052873498524732e-06,
+      "loss": 0.7588,
+      "step": 2341
+    },
+    {
+      "epoch": 0.79632777966678,
+      "grad_norm": 1.8914540998106915,
+      "learning_rate": 1.0494950101651274e-06,
+      "loss": 0.7959,
+      "step": 2342
+    },
+    {
+      "epoch": 0.7966678000680041,
+      "grad_norm": 2.392413934717188,
+      "learning_rate": 1.046121315311508e-06,
+      "loss": 0.8566,
+      "step": 2343
+    },
+    {
+      "epoch": 0.7970078204692281,
+      "grad_norm": 1.763253874255017,
+      "learning_rate": 1.04275241805748e-06,
+      "loss": 0.7474,
+      "step": 2344
+    },
+    {
+      "epoch": 0.7973478408704522,
+      "grad_norm": 1.558308712843251,
+      "learning_rate": 1.0393883224908358e-06,
+      "loss": 0.6914,
+      "step": 2345
+    },
+    {
+      "epoch": 0.7976878612716763,
+      "grad_norm": 1.7605529498524175,
+      "learning_rate": 1.036029032693534e-06,
+      "loss": 0.837,
+      "step": 2346
+    },
+    {
+      "epoch": 0.7980278816729004,
+      "grad_norm": 1.78998260592103,
+      "learning_rate": 1.0326745527417098e-06,
+      "loss": 0.8327,
+      "step": 2347
+    },
+    {
+      "epoch": 0.7983679020741244,
+      "grad_norm": 1.7542085033348727,
+      "learning_rate": 1.0293248867056527e-06,
+      "loss": 0.8285,
+      "step": 2348
+    },
+    {
+      "epoch": 0.7987079224753485,
+      "grad_norm": 1.9678558489079983,
+      "learning_rate": 1.0259800386498204e-06,
+      "loss": 0.8867,
+      "step": 2349
+    },
+    {
+      "epoch": 0.7990479428765725,
+      "grad_norm": 2.214701849578641,
+      "learning_rate": 1.022640012632819e-06,
+      "loss": 0.9296,
+      "step": 2350
+    },
+    {
+      "epoch": 0.7993879632777967,
+      "grad_norm": 1.9468124426612539,
+      "learning_rate": 1.0193048127074034e-06,
+      "loss": 0.8632,
+      "step": 2351
+    },
+    {
+      "epoch": 0.7997279836790208,
+      "grad_norm": 1.5760592103068731,
+      "learning_rate": 1.0159744429204776e-06,
+      "loss": 0.8049,
+      "step": 2352
+    },
+    {
+      "epoch": 0.8000680040802448,
+      "grad_norm": 1.950530829583191,
+      "learning_rate": 1.0126489073130779e-06,
+      "loss": 0.6512,
+      "step": 2353
+    },
+    {
+      "epoch": 0.8004080244814689,
+      "grad_norm": 1.7143537928592565,
+      "learning_rate": 1.0093282099203805e-06,
+      "loss": 0.7408,
+      "step": 2354
+    },
+    {
+      "epoch": 0.8007480448826929,
+      "grad_norm": 3.702934075061963,
+      "learning_rate": 1.0060123547716888e-06,
+      "loss": 0.7784,
+      "step": 2355
+    },
+    {
+      "epoch": 0.8010880652839171,
+      "grad_norm": 1.599124103938154,
+      "learning_rate": 1.0027013458904288e-06,
+      "loss": 0.8521,
+      "step": 2356
+    },
+    {
+      "epoch": 0.8014280856851411,
+      "grad_norm": 1.713620640939179,
+      "learning_rate": 9.993951872941493e-07,
+      "loss": 0.8589,
+      "step": 2357
+    },
+    {
+      "epoch": 0.8017681060863652,
+      "grad_norm": 1.9033657273086837,
+      "learning_rate": 9.960938829945104e-07,
+      "loss": 0.7361,
+      "step": 2358
+    },
+    {
+      "epoch": 0.8021081264875892,
+      "grad_norm": 1.8996990852991276,
+      "learning_rate": 9.927974369972871e-07,
+      "loss": 0.7452,
+      "step": 2359
+    },
+    {
+      "epoch": 0.8024481468888133,
+      "grad_norm": 2.0379615163395193,
+      "learning_rate": 9.895058533023532e-07,
+      "loss": 0.5995,
+      "step": 2360
+    },
+    {
+      "epoch": 0.8027881672900374,
+      "grad_norm": 1.7890665810043467,
+      "learning_rate": 9.862191359036883e-07,
+      "loss": 0.9003,
+      "step": 2361
+    },
+    {
+      "epoch": 0.8031281876912615,
+      "grad_norm": 2.6456228617289748,
+      "learning_rate": 9.829372887893624e-07,
+      "loss": 0.7455,
+      "step": 2362
+    },
+    {
+      "epoch": 0.8034682080924855,
+      "grad_norm": 1.6805195436291074,
+      "learning_rate": 9.796603159415407e-07,
+      "loss": 0.7163,
+      "step": 2363
+    },
+    {
+      "epoch": 0.8038082284937096,
+      "grad_norm": 2.042414294670765,
+      "learning_rate": 9.763882213364705e-07,
+      "loss": 0.6174,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8041482488949337,
+      "grad_norm": 1.8519040933711384,
+      "learning_rate": 9.731210089444803e-07,
+      "loss": 0.7669,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8044882692961578,
+      "grad_norm": 2.2531030397503464,
+      "learning_rate": 9.69858682729976e-07,
+      "loss": 0.8395,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8048282896973818,
+      "grad_norm": 2.0748617126240516,
+      "learning_rate": 9.66601246651432e-07,
+      "loss": 0.7717,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8051683100986059,
+      "grad_norm": 3.841576998570947,
+      "learning_rate": 9.633487046613932e-07,
+      "loss": 0.8345,
+      "step": 2368
+    },
+    {
+      "epoch": 0.80550833049983,
+      "grad_norm": 1.7489227453665286,
+      "learning_rate": 9.60101060706462e-07,
+      "loss": 0.8575,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8058483509010541,
+      "grad_norm": 1.7114506077871587,
+      "learning_rate": 9.568583187273018e-07,
+      "loss": 0.8861,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8061883713022782,
+      "grad_norm": 2.629625241075618,
+      "learning_rate": 9.536204826586243e-07,
+      "loss": 0.707,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8065283917035022,
+      "grad_norm": 2.451325517306834,
+      "learning_rate": 9.503875564291886e-07,
+      "loss": 0.7568,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8068684121047263,
+      "grad_norm": 1.865454750034287,
+      "learning_rate": 9.471595439617986e-07,
+      "loss": 0.8517,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8072084325059503,
+      "grad_norm": 2.0078901844899053,
+      "learning_rate": 9.439364491732927e-07,
+      "loss": 0.7792,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8075484529071745,
+      "grad_norm": 2.1884086746809506,
+      "learning_rate": 9.407182759745464e-07,
+      "loss": 0.7711,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8078884733083985,
+      "grad_norm": 2.359962377985305,
+      "learning_rate": 9.375050282704596e-07,
+      "loss": 0.7623,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8082284937096226,
+      "grad_norm": 1.965811233572265,
+      "learning_rate": 9.342967099599587e-07,
+      "loss": 0.7636,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8085685141108466,
+      "grad_norm": 2.1355877396954557,
+      "learning_rate": 9.31093324935985e-07,
+      "loss": 0.8835,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8089085345120707,
+      "grad_norm": 1.725291600691821,
+      "learning_rate": 9.278948770854984e-07,
+      "loss": 0.8575,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8092485549132948,
+      "grad_norm": 2.1970920436039725,
+      "learning_rate": 9.247013702894653e-07,
+      "loss": 0.7891,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8095885753145189,
+      "grad_norm": 2.1911323854498304,
+      "learning_rate": 9.215128084228564e-07,
+      "loss": 0.7819,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8099285957157429,
+      "grad_norm": 1.9062475395377723,
+      "learning_rate": 9.183291953546425e-07,
+      "loss": 0.7573,
+      "step": 2382
+    },
+    {
+      "epoch": 0.810268616116967,
+      "grad_norm": 1.5938084300323458,
+      "learning_rate": 9.151505349477901e-07,
+      "loss": 0.744,
+      "step": 2383
+    },
+    {
+      "epoch": 0.810608636518191,
+      "grad_norm": 2.4169596699907103,
+      "learning_rate": 9.11976831059258e-07,
+      "loss": 0.7026,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8109486569194152,
+      "grad_norm": 2.6219709533237903,
+      "learning_rate": 9.088080875399862e-07,
+      "loss": 0.6643,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8112886773206393,
+      "grad_norm": 2.01219370432533,
+      "learning_rate": 9.056443082349015e-07,
+      "loss": 0.7425,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8116286977218633,
+      "grad_norm": 2.066167061827305,
+      "learning_rate": 9.024854969829016e-07,
+      "loss": 0.6546,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8119687181230874,
+      "grad_norm": 1.4300615570476107,
+      "learning_rate": 8.993316576168626e-07,
+      "loss": 0.7899,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8123087385243115,
+      "grad_norm": 1.7707829875888732,
+      "learning_rate": 8.961827939636198e-07,
+      "loss": 0.8382,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8126487589255356,
+      "grad_norm": 2.1704787690301597,
+      "learning_rate": 8.930389098439751e-07,
+      "loss": 0.7779,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8129887793267596,
+      "grad_norm": 1.6521448642583239,
+      "learning_rate": 8.899000090726905e-07,
+      "loss": 0.788,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8133287997279837,
+      "grad_norm": 1.9493462394819676,
+      "learning_rate": 8.867660954584773e-07,
+      "loss": 0.8392,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8136688201292077,
+      "grad_norm": 1.8331269676018516,
+      "learning_rate": 8.836371728039989e-07,
+      "loss": 0.78,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8140088405304319,
+      "grad_norm": 2.026606682272262,
+      "learning_rate": 8.80513244905859e-07,
+      "loss": 0.8935,
+      "step": 2394
+    },
+    {
+      "epoch": 0.8143488609316559,
+      "grad_norm": 1.7620149620203838,
+      "learning_rate": 8.773943155546044e-07,
+      "loss": 0.6249,
+      "step": 2395
+    },
+    {
+      "epoch": 0.81468888133288,
+      "grad_norm": 2.7206868334859506,
+      "learning_rate": 8.74280388534714e-07,
+      "loss": 0.7804,
+      "step": 2396
+    },
+    {
+      "epoch": 0.815028901734104,
+      "grad_norm": 1.7308067133219402,
+      "learning_rate": 8.711714676245975e-07,
+      "loss": 0.7325,
+      "step": 2397
+    },
+    {
+      "epoch": 0.8153689221353281,
+      "grad_norm": 2.51943783841568,
+      "learning_rate": 8.680675565965918e-07,
+      "loss": 0.752,
+      "step": 2398
+    },
+    {
+      "epoch": 0.8157089425365522,
+      "grad_norm": 1.549980412078022,
+      "learning_rate": 8.64968659216951e-07,
+      "loss": 0.9601,
+      "step": 2399
+    },
+    {
+      "epoch": 0.8160489629377763,
+      "grad_norm": 2.0254215121489194,
+      "learning_rate": 8.618747792458515e-07,
+      "loss": 0.8119,
+      "step": 2400
+    },
+    {
+      "epoch": 0.8163889833390003,
+      "grad_norm": 1.9202778211704874,
+      "learning_rate": 8.58785920437376e-07,
+      "loss": 0.7911,
+      "step": 2401
+    },
+    {
+      "epoch": 0.8167290037402244,
+      "grad_norm": 1.6548864238372174,
+      "learning_rate": 8.557020865395194e-07,
+      "loss": 0.7711,
+      "step": 2402
+    },
+    {
+      "epoch": 0.8170690241414484,
+      "grad_norm": 1.5676328354601332,
+      "learning_rate": 8.526232812941748e-07,
+      "loss": 0.6984,
+      "step": 2403
+    },
+    {
+      "epoch": 0.8174090445426726,
+      "grad_norm": 1.7193145665504181,
+      "learning_rate": 8.49549508437138e-07,
+      "loss": 0.8111,
+      "step": 2404
+    },
+    {
+      "epoch": 0.8177490649438967,
+      "grad_norm": 1.68834796706292,
+      "learning_rate": 8.464807716980961e-07,
+      "loss": 0.7438,
+      "step": 2405
+    },
+    {
+      "epoch": 0.8180890853451207,
+      "grad_norm": 2.349332554637134,
+      "learning_rate": 8.434170748006226e-07,
+      "loss": 0.8144,
+      "step": 2406
+    },
+    {
+      "epoch": 0.8184291057463448,
+      "grad_norm": 7.841596102340736,
+      "learning_rate": 8.403584214621823e-07,
+      "loss": 0.7929,
+      "step": 2407
+    },
+    {
+      "epoch": 0.8187691261475688,
+      "grad_norm": 2.141322231062195,
+      "learning_rate": 8.373048153941144e-07,
+      "loss": 0.8196,
+      "step": 2408
+    },
+    {
+      "epoch": 0.819109146548793,
+      "grad_norm": 2.0397457850733884,
+      "learning_rate": 8.34256260301638e-07,
+      "loss": 0.7033,
+      "step": 2409
+    },
+    {
+      "epoch": 0.819449166950017,
+      "grad_norm": 2.205136821953891,
+      "learning_rate": 8.312127598838387e-07,
+      "loss": 0.7234,
+      "step": 2410
+    },
+    {
+      "epoch": 0.8197891873512411,
+      "grad_norm": 4.337708830724585,
+      "learning_rate": 8.281743178336754e-07,
+      "loss": 0.7171,
+      "step": 2411
+    },
+    {
+      "epoch": 0.8201292077524651,
+      "grad_norm": 1.6721477091051125,
+      "learning_rate": 8.251409378379638e-07,
+      "loss": 0.8007,
+      "step": 2412
+    },
+    {
+      "epoch": 0.8204692281536892,
+      "grad_norm": 3.8418421223750685,
+      "learning_rate": 8.22112623577378e-07,
+      "loss": 0.7635,
+      "step": 2413
+    },
+    {
+      "epoch": 0.8208092485549133,
+      "grad_norm": 1.6182376861006265,
+      "learning_rate": 8.19089378726447e-07,
+      "loss": 0.7876,
+      "step": 2414
+    },
+    {
+      "epoch": 0.8211492689561374,
+      "grad_norm": 2.2512602655558376,
+      "learning_rate": 8.160712069535464e-07,
+      "loss": 0.7364,
+      "step": 2415
+    },
+    {
+      "epoch": 0.8214892893573614,
+      "grad_norm": 1.5613182342547798,
+      "learning_rate": 8.130581119209008e-07,
+      "loss": 0.7997,
+      "step": 2416
+    },
+    {
+      "epoch": 0.8218293097585855,
+      "grad_norm": 2.123606402304467,
+      "learning_rate": 8.100500972845688e-07,
+      "loss": 0.7256,
+      "step": 2417
+    },
+    {
+      "epoch": 0.8221693301598096,
+      "grad_norm": 2.5565946650892846,
+      "learning_rate": 8.070471666944496e-07,
+      "loss": 0.7453,
+      "step": 2418
+    },
+    {
+      "epoch": 0.8225093505610337,
+      "grad_norm": 2.2217115142287667,
+      "learning_rate": 8.040493237942698e-07,
+      "loss": 0.8128,
+      "step": 2419
+    },
+    {
+      "epoch": 0.8228493709622577,
+      "grad_norm": 1.8879938400884209,
+      "learning_rate": 8.010565722215851e-07,
+      "loss": 0.7291,
+      "step": 2420
+    },
+    {
+      "epoch": 0.8231893913634818,
+      "grad_norm": 1.8601116173360819,
+      "learning_rate": 7.98068915607772e-07,
+      "loss": 0.801,
+      "step": 2421
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 1.7254034450571911,
+      "learning_rate": 7.950863575780249e-07,
+      "loss": 0.7592,
+      "step": 2422
+    },
+    {
+      "epoch": 0.82386943216593,
+      "grad_norm": 2.6122625161288586,
+      "learning_rate": 7.921089017513522e-07,
+      "loss": 0.8019,
+      "step": 2423
+    },
+    {
+      "epoch": 0.824209452567154,
+      "grad_norm": 1.6973528915543705,
+      "learning_rate": 7.891365517405702e-07,
+      "loss": 0.8974,
+      "step": 2424
+    },
+    {
+      "epoch": 0.8245494729683781,
+      "grad_norm": 1.7759681048749987,
+      "learning_rate": 7.861693111523022e-07,
+      "loss": 0.7917,
+      "step": 2425
+    },
+    {
+      "epoch": 0.8248894933696022,
+      "grad_norm": 1.7705878213685702,
+      "learning_rate": 7.832071835869687e-07,
+      "loss": 0.8071,
+      "step": 2426
+    },
+    {
+      "epoch": 0.8252295137708262,
+      "grad_norm": 2.156963508346104,
+      "learning_rate": 7.802501726387901e-07,
+      "loss": 0.7664,
+      "step": 2427
+    },
+    {
+      "epoch": 0.8255695341720504,
+      "grad_norm": 1.7276432522564236,
+      "learning_rate": 7.772982818957742e-07,
+      "loss": 0.7373,
+      "step": 2428
+    },
+    {
+      "epoch": 0.8259095545732744,
+      "grad_norm": 5.112931675706146,
+      "learning_rate": 7.743515149397185e-07,
+      "loss": 0.777,
+      "step": 2429
+    },
+    {
+      "epoch": 0.8262495749744985,
+      "grad_norm": 1.859215069146444,
+      "learning_rate": 7.714098753462018e-07,
+      "loss": 0.7991,
+      "step": 2430
+    },
+    {
+      "epoch": 0.8265895953757225,
+      "grad_norm": 2.0011677462274737,
+      "learning_rate": 7.684733666845812e-07,
+      "loss": 0.7925,
+      "step": 2431
+    },
+    {
+      "epoch": 0.8269296157769466,
+      "grad_norm": 2.7022442512396987,
+      "learning_rate": 7.655419925179919e-07,
+      "loss": 0.6235,
+      "step": 2432
+    },
+    {
+      "epoch": 0.8272696361781707,
+      "grad_norm": 2.0241160886895293,
+      "learning_rate": 7.626157564033332e-07,
+      "loss": 0.6865,
+      "step": 2433
+    },
+    {
+      "epoch": 0.8276096565793948,
+      "grad_norm": 1.906792937890856,
+      "learning_rate": 7.596946618912754e-07,
+      "loss": 0.8559,
+      "step": 2434
+    },
+    {
+      "epoch": 0.8279496769806188,
+      "grad_norm": 1.8797923348430752,
+      "learning_rate": 7.567787125262449e-07,
+      "loss": 0.6898,
+      "step": 2435
+    },
+    {
+      "epoch": 0.8282896973818429,
+      "grad_norm": 1.8757719817795944,
+      "learning_rate": 7.538679118464298e-07,
+      "loss": 0.7356,
+      "step": 2436
+    },
+    {
+      "epoch": 0.8286297177830669,
+      "grad_norm": 2.699911415427047,
+      "learning_rate": 7.509622633837671e-07,
+      "loss": 0.6198,
+      "step": 2437
+    },
+    {
+      "epoch": 0.8289697381842911,
+      "grad_norm": 1.8535353558494398,
+      "learning_rate": 7.480617706639442e-07,
+      "loss": 0.6603,
+      "step": 2438
+    },
+    {
+      "epoch": 0.8293097585855151,
+      "grad_norm": 2.0077754858561025,
+      "learning_rate": 7.451664372063916e-07,
+      "loss": 0.8192,
+      "step": 2439
+    },
+    {
+      "epoch": 0.8296497789867392,
+      "grad_norm": 2.3842511987435064,
+      "learning_rate": 7.422762665242788e-07,
+      "loss": 0.8319,
+      "step": 2440
+    },
+    {
+      "epoch": 0.8299897993879632,
+      "grad_norm": 2.3986998450988835,
+      "learning_rate": 7.393912621245142e-07,
+      "loss": 0.798,
+      "step": 2441
+    },
+    {
+      "epoch": 0.8303298197891874,
+      "grad_norm": 1.761347248138979,
+      "learning_rate": 7.365114275077334e-07,
+      "loss": 0.7448,
+      "step": 2442
+    },
+    {
+      "epoch": 0.8306698401904115,
+      "grad_norm": 1.9120511252582268,
+      "learning_rate": 7.33636766168303e-07,
+      "loss": 0.8883,
+      "step": 2443
+    },
+    {
+      "epoch": 0.8310098605916355,
+      "grad_norm": 1.8267075244116617,
+      "learning_rate": 7.307672815943084e-07,
+      "loss": 0.7732,
+      "step": 2444
+    },
+    {
+      "epoch": 0.8313498809928596,
+      "grad_norm": 1.9441639104439716,
+      "learning_rate": 7.279029772675572e-07,
+      "loss": 0.8854,
+      "step": 2445
+    },
+    {
+      "epoch": 0.8316899013940836,
+      "grad_norm": 2.1103415209340604,
+      "learning_rate": 7.250438566635692e-07,
+      "loss": 0.8216,
+      "step": 2446
+    },
+    {
+      "epoch": 0.8320299217953078,
+      "grad_norm": 1.8163467229280887,
+      "learning_rate": 7.221899232515727e-07,
+      "loss": 0.863,
+      "step": 2447
+    },
+    {
+      "epoch": 0.8323699421965318,
+      "grad_norm": 2.1150765935601474,
+      "learning_rate": 7.193411804945061e-07,
+      "loss": 0.6834,
+      "step": 2448
+    },
+    {
+      "epoch": 0.8327099625977559,
+      "grad_norm": 1.9172782529773438,
+      "learning_rate": 7.164976318490058e-07,
+      "loss": 0.8915,
+      "step": 2449
+    },
+    {
+      "epoch": 0.8330499829989799,
+      "grad_norm": 2.302226063344335,
+      "learning_rate": 7.136592807654085e-07,
+      "loss": 0.7917,
+      "step": 2450
+    },
+    {
+      "epoch": 0.833390003400204,
+      "grad_norm": 3.323395060515028,
+      "learning_rate": 7.108261306877423e-07,
+      "loss": 0.7571,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8337300238014281,
+      "grad_norm": 2.044271938418409,
+      "learning_rate": 7.079981850537266e-07,
+      "loss": 0.8017,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8340700442026522,
+      "grad_norm": 2.1409274203005193,
+      "learning_rate": 7.051754472947625e-07,
+      "loss": 0.7459,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8344100646038762,
+      "grad_norm": 2.95578414657139,
+      "learning_rate": 7.023579208359349e-07,
+      "loss": 0.8399,
+      "step": 2454
+    },
+    {
+      "epoch": 0.8347500850051003,
+      "grad_norm": 1.7525330525619445,
+      "learning_rate": 6.995456090960034e-07,
+      "loss": 0.8179,
+      "step": 2455
+    },
+    {
+      "epoch": 0.8350901054063243,
+      "grad_norm": 3.5586869346230676,
+      "learning_rate": 6.967385154874001e-07,
+      "loss": 0.9779,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8354301258075485,
+      "grad_norm": 1.7571946774112055,
+      "learning_rate": 6.939366434162287e-07,
+      "loss": 0.8006,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8357701462087725,
+      "grad_norm": 2.4033098083942446,
+      "learning_rate": 6.911399962822518e-07,
+      "loss": 0.7554,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8361101666099966,
+      "grad_norm": 1.869358589109052,
+      "learning_rate": 6.883485774788973e-07,
+      "loss": 0.7259,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8364501870112206,
+      "grad_norm": 1.91414508963133,
+      "learning_rate": 6.855623903932457e-07,
+      "loss": 0.6757,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8367902074124447,
+      "grad_norm": 1.7629021010412422,
+      "learning_rate": 6.82781438406031e-07,
+      "loss": 0.6845,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8371302278136689,
+      "grad_norm": 1.733843016954561,
+      "learning_rate": 6.800057248916347e-07,
+      "loss": 0.7731,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8374702482148929,
+      "grad_norm": 1.6233499660099797,
+      "learning_rate": 6.772352532180815e-07,
+      "loss": 0.7542,
+      "step": 2463
+    },
+    {
+      "epoch": 0.837810268616117,
+      "grad_norm": 1.9854817972998593,
+      "learning_rate": 6.74470026747035e-07,
+      "loss": 0.7532,
+      "step": 2464
+    },
+    {
+      "epoch": 0.838150289017341,
+      "grad_norm": 2.0817806024786454,
+      "learning_rate": 6.717100488337952e-07,
+      "loss": 0.7815,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8384903094185651,
+      "grad_norm": 2.966599408068768,
+      "learning_rate": 6.689553228272955e-07,
+      "loss": 0.7962,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8388303298197892,
+      "grad_norm": 2.0775168349055577,
+      "learning_rate": 6.662058520700926e-07,
+      "loss": 0.7808,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8391703502210133,
+      "grad_norm": 2.4039263789618595,
+      "learning_rate": 6.634616398983712e-07,
+      "loss": 0.8221,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8395103706222373,
+      "grad_norm": 1.622008521580856,
+      "learning_rate": 6.607226896419305e-07,
+      "loss": 0.7502,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8398503910234614,
+      "grad_norm": 2.1962585919549564,
+      "learning_rate": 6.579890046241888e-07,
+      "loss": 0.7449,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8401904114246855,
+      "grad_norm": 1.819758576227155,
+      "learning_rate": 6.552605881621732e-07,
+      "loss": 0.7057,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8405304318259096,
+      "grad_norm": 1.9641795678417464,
+      "learning_rate": 6.525374435665183e-07,
+      "loss": 0.73,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8408704522271336,
+      "grad_norm": 1.6323001552109164,
+      "learning_rate": 6.498195741414637e-07,
+      "loss": 0.7322,
+      "step": 2473
+    },
+    {
+      "epoch": 0.8412104726283577,
+      "grad_norm": 1.8961797022845666,
+      "learning_rate": 6.471069831848453e-07,
+      "loss": 0.721,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8415504930295817,
+      "grad_norm": 1.811172379243824,
+      "learning_rate": 6.443996739880981e-07,
+      "loss": 0.7265,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8418905134308059,
+      "grad_norm": 1.7705443438552309,
+      "learning_rate": 6.416976498362432e-07,
+      "loss": 0.641,
+      "step": 2476
+    },
+    {
+      "epoch": 0.84223053383203,
+      "grad_norm": 2.5591611017829186,
+      "learning_rate": 6.39000914007894e-07,
+      "loss": 0.8185,
+      "step": 2477
+    },
+    {
+      "epoch": 0.842570554233254,
+      "grad_norm": 1.500821178484566,
+      "learning_rate": 6.363094697752436e-07,
+      "loss": 0.8445,
+      "step": 2478
+    },
+    {
+      "epoch": 0.842910574634478,
+      "grad_norm": 1.8045464912453517,
+      "learning_rate": 6.336233204040654e-07,
+      "loss": 0.8186,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8432505950357021,
+      "grad_norm": 2.1257740354966024,
+      "learning_rate": 6.309424691537075e-07,
+      "loss": 0.7636,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8435906154369263,
+      "grad_norm": 2.1805061395433025,
+      "learning_rate": 6.282669192770896e-07,
+      "loss": 0.7369,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8439306358381503,
+      "grad_norm": 2.794846227145433,
+      "learning_rate": 6.255966740207003e-07,
+      "loss": 0.7512,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8442706562393744,
+      "grad_norm": 2.139417416344505,
+      "learning_rate": 6.229317366245891e-07,
+      "loss": 0.858,
+      "step": 2483
+    },
+    {
+      "epoch": 0.8446106766405984,
+      "grad_norm": 1.6794324817896975,
+      "learning_rate": 6.20272110322368e-07,
+      "loss": 0.7415,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8449506970418225,
+      "grad_norm": 2.557876957181143,
+      "learning_rate": 6.176177983412013e-07,
+      "loss": 0.7493,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8452907174430466,
+      "grad_norm": 1.6147587726820207,
+      "learning_rate": 6.14968803901807e-07,
+      "loss": 0.7066,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8456307378442707,
+      "grad_norm": 6.264968025708578,
+      "learning_rate": 6.123251302184502e-07,
+      "loss": 0.7846,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8459707582454947,
+      "grad_norm": 2.0720939669815297,
+      "learning_rate": 6.096867804989387e-07,
+      "loss": 0.8005,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8463107786467188,
+      "grad_norm": 1.7818861934834522,
+      "learning_rate": 6.07053757944624e-07,
+      "loss": 0.8083,
+      "step": 2489
+    },
+    {
+      "epoch": 0.8466507990479428,
+      "grad_norm": 1.9663397712493058,
+      "learning_rate": 6.044260657503881e-07,
+      "loss": 0.7888,
+      "step": 2490
+    },
+    {
+      "epoch": 0.846990819449167,
+      "grad_norm": 1.7007116920435754,
+      "learning_rate": 6.018037071046518e-07,
+      "loss": 0.727,
+      "step": 2491
+    },
+    {
+      "epoch": 0.847330839850391,
+      "grad_norm": 2.01433634361555,
+      "learning_rate": 5.991866851893569e-07,
+      "loss": 0.7841,
+      "step": 2492
+    },
+    {
+      "epoch": 0.8476708602516151,
+      "grad_norm": 2.0056879931109104,
+      "learning_rate": 5.965750031799772e-07,
+      "loss": 0.7634,
+      "step": 2493
+    },
+    {
+      "epoch": 0.8480108806528391,
+      "grad_norm": 1.7817072170081534,
+      "learning_rate": 5.939686642455012e-07,
+      "loss": 0.7755,
+      "step": 2494
+    },
+    {
+      "epoch": 0.8483509010540632,
+      "grad_norm": 1.771192620612961,
+      "learning_rate": 5.913676715484363e-07,
+      "loss": 0.8514,
+      "step": 2495
+    },
+    {
+      "epoch": 0.8486909214552874,
+      "grad_norm": 1.9377408759282009,
+      "learning_rate": 5.887720282448034e-07,
+      "loss": 0.7875,
+      "step": 2496
+    },
+    {
+      "epoch": 0.8490309418565114,
+      "grad_norm": 1.875586498487033,
+      "learning_rate": 5.861817374841311e-07,
+      "loss": 0.7402,
+      "step": 2497
+    },
+    {
+      "epoch": 0.8493709622577355,
+      "grad_norm": 1.896813734307177,
+      "learning_rate": 5.835968024094551e-07,
+      "loss": 0.7494,
+      "step": 2498
+    },
+    {
+      "epoch": 0.8497109826589595,
+      "grad_norm": 1.746600133481283,
+      "learning_rate": 5.810172261573099e-07,
+      "loss": 0.7486,
+      "step": 2499
+    },
+    {
+      "epoch": 0.8500510030601837,
+      "grad_norm": 1.701793210043788,
+      "learning_rate": 5.784430118577322e-07,
+      "loss": 0.7742,
+      "step": 2500
+    },
+    {
+      "epoch": 0.8503910234614077,
+      "grad_norm": 1.8039181908755018,
+      "learning_rate": 5.758741626342479e-07,
+      "loss": 0.8416,
+      "step": 2501
+    },
+    {
+      "epoch": 0.8507310438626318,
+      "grad_norm": 1.7572654084027113,
+      "learning_rate": 5.733106816038736e-07,
+      "loss": 0.6848,
+      "step": 2502
+    },
+    {
+      "epoch": 0.8510710642638558,
+      "grad_norm": 2.018493313007424,
+      "learning_rate": 5.707525718771151e-07,
+      "loss": 0.8917,
+      "step": 2503
+    },
+    {
+      "epoch": 0.8514110846650799,
+      "grad_norm": 2.0723674784682644,
+      "learning_rate": 5.681998365579594e-07,
+      "loss": 0.8585,
+      "step": 2504
+    },
+    {
+      "epoch": 0.851751105066304,
+      "grad_norm": 1.8580629745251314,
+      "learning_rate": 5.6565247874387e-07,
+      "loss": 0.7809,
+      "step": 2505
+    },
+    {
+      "epoch": 0.8520911254675281,
+      "grad_norm": 1.8712383894476246,
+      "learning_rate": 5.631105015257871e-07,
+      "loss": 0.7901,
+      "step": 2506
+    },
+    {
+      "epoch": 0.8524311458687521,
+      "grad_norm": 1.574135923996909,
+      "learning_rate": 5.60573907988124e-07,
+      "loss": 0.7791,
+      "step": 2507
+    },
+    {
+      "epoch": 0.8527711662699762,
+      "grad_norm": 2.195303216051158,
+      "learning_rate": 5.58042701208758e-07,
+      "loss": 0.6425,
+      "step": 2508
+    },
+    {
+      "epoch": 0.8531111866712002,
+      "grad_norm": 2.0222539828068418,
+      "learning_rate": 5.55516884259033e-07,
+      "loss": 0.8305,
+      "step": 2509
+    },
+    {
+      "epoch": 0.8534512070724244,
+      "grad_norm": 2.0135511435332427,
+      "learning_rate": 5.529964602037519e-07,
+      "loss": 0.7716,
+      "step": 2510
+    },
+    {
+      "epoch": 0.8537912274736484,
+      "grad_norm": 1.7236389648693269,
+      "learning_rate": 5.504814321011732e-07,
+      "loss": 0.6894,
+      "step": 2511
+    },
+    {
+      "epoch": 0.8541312478748725,
+      "grad_norm": 1.8873379975752618,
+      "learning_rate": 5.479718030030084e-07,
+      "loss": 0.7636,
+      "step": 2512
+    },
+    {
+      "epoch": 0.8544712682760965,
+      "grad_norm": 1.7426057926099683,
+      "learning_rate": 5.454675759544176e-07,
+      "loss": 0.8053,
+      "step": 2513
+    },
+    {
+      "epoch": 0.8548112886773206,
+      "grad_norm": 1.8753643537027582,
+      "learning_rate": 5.429687539940076e-07,
+      "loss": 0.723,
+      "step": 2514
+    },
+    {
+      "epoch": 0.8551513090785448,
+      "grad_norm": 1.7277156698902645,
+      "learning_rate": 5.404753401538249e-07,
+      "loss": 0.7989,
+      "step": 2515
+    },
+    {
+      "epoch": 0.8554913294797688,
+      "grad_norm": 2.2677017882113675,
+      "learning_rate": 5.379873374593563e-07,
+      "loss": 0.7536,
+      "step": 2516
+    },
+    {
+      "epoch": 0.8558313498809929,
+      "grad_norm": 1.622977811890628,
+      "learning_rate": 5.355047489295195e-07,
+      "loss": 0.7579,
+      "step": 2517
+    },
+    {
+      "epoch": 0.8561713702822169,
+      "grad_norm": 1.8045814672152072,
+      "learning_rate": 5.330275775766642e-07,
+      "loss": 0.7795,
+      "step": 2518
+    },
+    {
+      "epoch": 0.856511390683441,
+      "grad_norm": 1.8483802620891205,
+      "learning_rate": 5.30555826406568e-07,
+      "loss": 0.8143,
+      "step": 2519
+    },
+    {
+      "epoch": 0.8568514110846651,
+      "grad_norm": 2.0653176883961075,
+      "learning_rate": 5.28089498418431e-07,
+      "loss": 0.8275,
+      "step": 2520
+    },
+    {
+      "epoch": 0.8571914314858892,
+      "grad_norm": 2.3816394206381752,
+      "learning_rate": 5.256285966048719e-07,
+      "loss": 0.7278,
+      "step": 2521
+    },
+    {
+      "epoch": 0.8575314518871132,
+      "grad_norm": 1.6020707124660172,
+      "learning_rate": 5.23173123951925e-07,
+      "loss": 0.8589,
+      "step": 2522
+    },
+    {
+      "epoch": 0.8578714722883373,
+      "grad_norm": 2.2904803070652684,
+      "learning_rate": 5.207230834390403e-07,
+      "loss": 0.7793,
+      "step": 2523
+    },
+    {
+      "epoch": 0.8582114926895614,
+      "grad_norm": 2.2504916928798666,
+      "learning_rate": 5.182784780390721e-07,
+      "loss": 0.7643,
+      "step": 2524
+    },
+    {
+      "epoch": 0.8585515130907855,
+      "grad_norm": 2.1813636901240074,
+      "learning_rate": 5.158393107182835e-07,
+      "loss": 0.7989,
+      "step": 2525
+    },
+    {
+      "epoch": 0.8588915334920095,
+      "grad_norm": 2.17526125079606,
+      "learning_rate": 5.134055844363367e-07,
+      "loss": 0.7287,
+      "step": 2526
+    },
+    {
+      "epoch": 0.8592315538932336,
+      "grad_norm": 2.0068173957653364,
+      "learning_rate": 5.109773021462921e-07,
+      "loss": 0.8449,
+      "step": 2527
+    },
+    {
+      "epoch": 0.8595715742944576,
+      "grad_norm": 1.9711345745234496,
+      "learning_rate": 5.085544667946057e-07,
+      "loss": 0.8109,
+      "step": 2528
+    },
+    {
+      "epoch": 0.8599115946956818,
+      "grad_norm": 2.3317026622050583,
+      "learning_rate": 5.061370813211219e-07,
+      "loss": 0.7172,
+      "step": 2529
+    },
+    {
+      "epoch": 0.8602516150969058,
+      "grad_norm": 1.757253994535814,
+      "learning_rate": 5.037251486590755e-07,
+      "loss": 0.7579,
+      "step": 2530
+    },
+    {
+      "epoch": 0.8605916354981299,
+      "grad_norm": 2.0285732960091543,
+      "learning_rate": 5.013186717350815e-07,
+      "loss": 0.796,
+      "step": 2531
+    },
+    {
+      "epoch": 0.8609316558993539,
+      "grad_norm": 1.659656431274867,
+      "learning_rate": 4.989176534691381e-07,
+      "loss": 0.7392,
+      "step": 2532
+    },
+    {
+      "epoch": 0.861271676300578,
+      "grad_norm": 2.0089159264540553,
+      "learning_rate": 4.965220967746181e-07,
+      "loss": 0.7919,
+      "step": 2533
+    },
+    {
+      "epoch": 0.8616116967018022,
+      "grad_norm": 2.473425505190949,
+      "learning_rate": 4.94132004558266e-07,
+      "loss": 0.7572,
+      "step": 2534
+    },
+    {
+      "epoch": 0.8619517171030262,
+      "grad_norm": 2.151920557303182,
+      "learning_rate": 4.917473797202005e-07,
+      "loss": 0.7254,
+      "step": 2535
+    },
+    {
+      "epoch": 0.8622917375042503,
+      "grad_norm": 2.232388197647897,
+      "learning_rate": 4.893682251539012e-07,
+      "loss": 0.6701,
+      "step": 2536
+    },
+    {
+      "epoch": 0.8626317579054743,
+      "grad_norm": 1.8999196855734106,
+      "learning_rate": 4.869945437462126e-07,
+      "loss": 0.7422,
+      "step": 2537
+    },
+    {
+      "epoch": 0.8629717783066984,
+      "grad_norm": 1.8645892417506236,
+      "learning_rate": 4.846263383773364e-07,
+      "loss": 0.7827,
+      "step": 2538
+    },
+    {
+      "epoch": 0.8633117987079225,
+      "grad_norm": 1.9633226448791292,
+      "learning_rate": 4.822636119208335e-07,
+      "loss": 0.8252,
+      "step": 2539
+    },
+    {
+      "epoch": 0.8636518191091466,
+      "grad_norm": 2.412498853567893,
+      "learning_rate": 4.799063672436111e-07,
+      "loss": 0.6881,
+      "step": 2540
+    },
+    {
+      "epoch": 0.8639918395103706,
+      "grad_norm": 2.0646181111873116,
+      "learning_rate": 4.775546072059311e-07,
+      "loss": 0.7813,
+      "step": 2541
+    },
+    {
+      "epoch": 0.8643318599115947,
+      "grad_norm": 1.811749527105386,
+      "learning_rate": 4.752083346613956e-07,
+      "loss": 0.804,
+      "step": 2542
+    },
+    {
+      "epoch": 0.8646718803128187,
+      "grad_norm": 2.0205742904901003,
+      "learning_rate": 4.728675524569487e-07,
+      "loss": 0.7478,
+      "step": 2543
+    },
+    {
+      "epoch": 0.8650119007140429,
+      "grad_norm": 1.6948710939652791,
+      "learning_rate": 4.7053226343287626e-07,
+      "loss": 0.7354,
+      "step": 2544
+    },
+    {
+      "epoch": 0.8653519211152669,
+      "grad_norm": 1.8571787878044181,
+      "learning_rate": 4.68202470422795e-07,
+      "loss": 0.7866,
+      "step": 2545
+    },
+    {
+      "epoch": 0.865691941516491,
+      "grad_norm": 4.332069466957595,
+      "learning_rate": 4.6587817625365406e-07,
+      "loss": 0.9335,
+      "step": 2546
+    },
+    {
+      "epoch": 0.866031961917715,
+      "grad_norm": 2.1059679315712083,
+      "learning_rate": 4.6355938374572975e-07,
+      "loss": 0.7504,
+      "step": 2547
+    },
+    {
+      "epoch": 0.8663719823189391,
+      "grad_norm": 2.411521023379306,
+      "learning_rate": 4.612460957126247e-07,
+      "loss": 0.7945,
+      "step": 2548
+    },
+    {
+      "epoch": 0.8667120027201632,
+      "grad_norm": 3.5331071294473904,
+      "learning_rate": 4.589383149612603e-07,
+      "loss": 0.7663,
+      "step": 2549
+    },
+    {
+      "epoch": 0.8670520231213873,
+      "grad_norm": 1.8408470267130377,
+      "learning_rate": 4.5663604429187547e-07,
+      "loss": 0.7752,
+      "step": 2550
+    },
+    {
+      "epoch": 0.8673920435226113,
+      "grad_norm": 1.8173297059452884,
+      "learning_rate": 4.543392864980256e-07,
+      "loss": 0.734,
+      "step": 2551
+    },
+    {
+      "epoch": 0.8677320639238354,
+      "grad_norm": 1.6548861190043966,
+      "learning_rate": 4.5204804436657423e-07,
+      "loss": 0.7518,
+      "step": 2552
+    },
+    {
+      "epoch": 0.8680720843250596,
+      "grad_norm": 1.8814419154796098,
+      "learning_rate": 4.4976232067769356e-07,
+      "loss": 0.8335,
+      "step": 2553
+    },
+    {
+      "epoch": 0.8684121047262836,
+      "grad_norm": 2.4824574969789017,
+      "learning_rate": 4.474821182048583e-07,
+      "loss": 0.7759,
+      "step": 2554
+    },
+    {
+      "epoch": 0.8687521251275077,
+      "grad_norm": 1.929420733997793,
+      "learning_rate": 4.45207439714847e-07,
+      "loss": 0.6974,
+      "step": 2555
+    },
+    {
+      "epoch": 0.8690921455287317,
+      "grad_norm": 1.810989892213861,
+      "learning_rate": 4.4293828796773133e-07,
+      "loss": 0.7086,
+      "step": 2556
+    },
+    {
+      "epoch": 0.8694321659299558,
+      "grad_norm": 3.143595670737321,
+      "learning_rate": 4.406746657168809e-07,
+      "loss": 0.8016,
+      "step": 2557
+    },
+    {
+      "epoch": 0.8697721863311799,
+      "grad_norm": 2.015317063610089,
+      "learning_rate": 4.384165757089526e-07,
+      "loss": 0.7969,
+      "step": 2558
+    },
+    {
+      "epoch": 0.870112206732404,
+      "grad_norm": 2.7863828548912277,
+      "learning_rate": 4.361640206838913e-07,
+      "loss": 0.7793,
+      "step": 2559
+    },
+    {
+      "epoch": 0.870452227133628,
+      "grad_norm": 1.9725121522194389,
+      "learning_rate": 4.339170033749279e-07,
+      "loss": 0.6607,
+      "step": 2560
+    },
+    {
+      "epoch": 0.8707922475348521,
+      "grad_norm": 1.7743293301120258,
+      "learning_rate": 4.316755265085715e-07,
+      "loss": 0.7992,
+      "step": 2561
+    },
+    {
+      "epoch": 0.8711322679360761,
+      "grad_norm": 4.284073115166657,
+      "learning_rate": 4.294395928046091e-07,
+      "loss": 0.6972,
+      "step": 2562
+    },
+    {
+      "epoch": 0.8714722883373003,
+      "grad_norm": 1.7376821947334347,
+      "learning_rate": 4.272092049761012e-07,
+      "loss": 0.7081,
+      "step": 2563
+    },
+    {
+      "epoch": 0.8718123087385243,
+      "grad_norm": 2.4003264376544524,
+      "learning_rate": 4.2498436572938117e-07,
+      "loss": 0.7366,
+      "step": 2564
+    },
+    {
+      "epoch": 0.8721523291397484,
+      "grad_norm": 1.7900283874803544,
+      "learning_rate": 4.227650777640474e-07,
+      "loss": 0.7543,
+      "step": 2565
+    },
+    {
+      "epoch": 0.8724923495409724,
+      "grad_norm": 2.121565874741983,
+      "learning_rate": 4.2055134377296245e-07,
+      "loss": 0.8625,
+      "step": 2566
+    },
+    {
+      "epoch": 0.8728323699421965,
+      "grad_norm": 2.065896891684966,
+      "learning_rate": 4.183431664422527e-07,
+      "loss": 0.8362,
+      "step": 2567
+    },
+    {
+      "epoch": 0.8731723903434206,
+      "grad_norm": 3.5516283553936154,
+      "learning_rate": 4.1614054845129814e-07,
+      "loss": 0.7923,
+      "step": 2568
+    },
+    {
+      "epoch": 0.8735124107446447,
+      "grad_norm": 1.7586409614005896,
+      "learning_rate": 4.139434924727359e-07,
+      "loss": 0.8087,
+      "step": 2569
+    },
+    {
+      "epoch": 0.8738524311458687,
+      "grad_norm": 1.763698578164718,
+      "learning_rate": 4.1175200117245127e-07,
+      "loss": 0.7511,
+      "step": 2570
+    },
+    {
+      "epoch": 0.8741924515470928,
+      "grad_norm": 1.9405876321607063,
+      "learning_rate": 4.095660772095822e-07,
+      "loss": 0.7895,
+      "step": 2571
+    },
+    {
+      "epoch": 0.8745324719483168,
+      "grad_norm": 2.2201530125382662,
+      "learning_rate": 4.0738572323650636e-07,
+      "loss": 0.7936,
+      "step": 2572
+    },
+    {
+      "epoch": 0.874872492349541,
+      "grad_norm": 2.698353558807428,
+      "learning_rate": 4.05210941898847e-07,
+      "loss": 0.7766,
+      "step": 2573
+    },
+    {
+      "epoch": 0.8752125127507651,
+      "grad_norm": 1.8695910167521532,
+      "learning_rate": 4.0304173583546214e-07,
+      "loss": 0.7827,
+      "step": 2574
+    },
+    {
+      "epoch": 0.8755525331519891,
+      "grad_norm": 1.787702184500189,
+      "learning_rate": 4.008781076784457e-07,
+      "loss": 0.8141,
+      "step": 2575
+    },
+    {
+      "epoch": 0.8758925535532132,
+      "grad_norm": 1.9350367343497943,
+      "learning_rate": 3.9872006005312545e-07,
+      "loss": 0.8147,
+      "step": 2576
+    },
+    {
+      "epoch": 0.8762325739544373,
+      "grad_norm": 1.9411047058406645,
+      "learning_rate": 3.965675955780551e-07,
+      "loss": 0.8205,
+      "step": 2577
+    },
+    {
+      "epoch": 0.8765725943556614,
+      "grad_norm": 2.083484099729529,
+      "learning_rate": 3.9442071686501605e-07,
+      "loss": 0.7374,
+      "step": 2578
+    },
+    {
+      "epoch": 0.8769126147568854,
+      "grad_norm": 1.484808045925216,
+      "learning_rate": 3.9227942651900943e-07,
+      "loss": 0.7934,
+      "step": 2579
+    },
+    {
+      "epoch": 0.8772526351581095,
+      "grad_norm": 1.6403262654823596,
+      "learning_rate": 3.901437271382591e-07,
+      "loss": 0.75,
+      "step": 2580
+    },
+    {
+      "epoch": 0.8775926555593335,
+      "grad_norm": 2.139199109393669,
+      "learning_rate": 3.8801362131420105e-07,
+      "loss": 0.7095,
+      "step": 2581
+    },
+    {
+      "epoch": 0.8779326759605577,
+      "grad_norm": 2.283437669529752,
+      "learning_rate": 3.858891116314861e-07,
+      "loss": 0.8167,
+      "step": 2582
+    },
+    {
+      "epoch": 0.8782726963617817,
+      "grad_norm": 1.777253272527263,
+      "learning_rate": 3.8377020066797557e-07,
+      "loss": 0.6707,
+      "step": 2583
+    },
+    {
+      "epoch": 0.8786127167630058,
+      "grad_norm": 2.032396863440268,
+      "learning_rate": 3.8165689099473436e-07,
+      "loss": 0.7875,
+      "step": 2584
+    },
+    {
+      "epoch": 0.8789527371642298,
+      "grad_norm": 2.1496667940512424,
+      "learning_rate": 3.7954918517603636e-07,
+      "loss": 0.7843,
+      "step": 2585
+    },
+    {
+      "epoch": 0.8792927575654539,
+      "grad_norm": 2.0953430909454838,
+      "learning_rate": 3.7744708576934795e-07,
+      "loss": 0.7498,
+      "step": 2586
+    },
+    {
+      "epoch": 0.879632777966678,
+      "grad_norm": 2.6645355484935,
+      "learning_rate": 3.7535059532533945e-07,
+      "loss": 0.7451,
+      "step": 2587
+    },
+    {
+      "epoch": 0.8799727983679021,
+      "grad_norm": 2.7910264296663567,
+      "learning_rate": 3.732597163878715e-07,
+      "loss": 0.761,
+      "step": 2588
+    },
+    {
+      "epoch": 0.8803128187691261,
+      "grad_norm": 2.1309852876618893,
+      "learning_rate": 3.711744514939991e-07,
+      "loss": 0.7839,
+      "step": 2589
+    },
+    {
+      "epoch": 0.8806528391703502,
+      "grad_norm": 2.0023508759747273,
+      "learning_rate": 3.690948031739622e-07,
+      "loss": 0.7626,
+      "step": 2590
+    },
+    {
+      "epoch": 0.8809928595715742,
+      "grad_norm": 2.1149106283441266,
+      "learning_rate": 3.67020773951185e-07,
+      "loss": 0.7549,
+      "step": 2591
+    },
+    {
+      "epoch": 0.8813328799727984,
+      "grad_norm": 1.785186211034344,
+      "learning_rate": 3.649523663422783e-07,
+      "loss": 0.7699,
+      "step": 2592
+    },
+    {
+      "epoch": 0.8816729003740225,
+      "grad_norm": 2.53236936356743,
+      "learning_rate": 3.6288958285702726e-07,
+      "loss": 0.7464,
+      "step": 2593
+    },
+    {
+      "epoch": 0.8820129207752465,
+      "grad_norm": 2.2182667953687516,
+      "learning_rate": 3.6083242599839365e-07,
+      "loss": 0.7926,
+      "step": 2594
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 2.0193533863123316,
+      "learning_rate": 3.587808982625124e-07,
+      "loss": 0.7586,
+      "step": 2595
+    },
+    {
+      "epoch": 0.8826929615776946,
+      "grad_norm": 1.8994718280339355,
+      "learning_rate": 3.567350021386895e-07,
+      "loss": 0.7463,
+      "step": 2596
+    },
+    {
+      "epoch": 0.8830329819789188,
+      "grad_norm": 1.9243015384876234,
+      "learning_rate": 3.546947401093953e-07,
+      "loss": 0.8557,
+      "step": 2597
+    },
+    {
+      "epoch": 0.8833730023801428,
+      "grad_norm": 1.8567845441632016,
+      "learning_rate": 3.5266011465026394e-07,
+      "loss": 0.8092,
+      "step": 2598
+    },
+    {
+      "epoch": 0.8837130227813669,
+      "grad_norm": 1.946576177975856,
+      "learning_rate": 3.506311282300934e-07,
+      "loss": 0.7336,
+      "step": 2599
+    },
+    {
+      "epoch": 0.8840530431825909,
+      "grad_norm": 1.7654478089333205,
+      "learning_rate": 3.486077833108342e-07,
+      "loss": 0.7989,
+      "step": 2600
+    },
+    {
+      "epoch": 0.884393063583815,
+      "grad_norm": 2.3364268494283444,
+      "learning_rate": 3.4659008234759597e-07,
+      "loss": 0.6956,
+      "step": 2601
+    },
+    {
+      "epoch": 0.8847330839850391,
+      "grad_norm": 2.1363703287698383,
+      "learning_rate": 3.4457802778863846e-07,
+      "loss": 0.7131,
+      "step": 2602
+    },
+    {
+      "epoch": 0.8850731043862632,
+      "grad_norm": 3.273937151667049,
+      "learning_rate": 3.4257162207536887e-07,
+      "loss": 0.821,
+      "step": 2603
+    },
+    {
+      "epoch": 0.8854131247874872,
+      "grad_norm": 1.6417747965293206,
+      "learning_rate": 3.405708676423408e-07,
+      "loss": 0.8703,
+      "step": 2604
+    },
+    {
+      "epoch": 0.8857531451887113,
+      "grad_norm": 1.7182489892417279,
+      "learning_rate": 3.3857576691725346e-07,
+      "loss": 0.7239,
+      "step": 2605
+    },
+    {
+      "epoch": 0.8860931655899354,
+      "grad_norm": 1.7767635345343673,
+      "learning_rate": 3.365863223209409e-07,
+      "loss": 0.7327,
+      "step": 2606
+    },
+    {
+      "epoch": 0.8864331859911595,
+      "grad_norm": 1.7793048886202218,
+      "learning_rate": 3.3460253626737774e-07,
+      "loss": 0.7237,
+      "step": 2607
+    },
+    {
+      "epoch": 0.8867732063923836,
+      "grad_norm": 2.404768196088338,
+      "learning_rate": 3.3262441116367174e-07,
+      "loss": 0.7197,
+      "step": 2608
+    },
+    {
+      "epoch": 0.8871132267936076,
+      "grad_norm": 2.5845357263659863,
+      "learning_rate": 3.306519494100618e-07,
+      "loss": 0.7361,
+      "step": 2609
+    },
+    {
+      "epoch": 0.8874532471948317,
+      "grad_norm": 1.7048562949785628,
+      "learning_rate": 3.286851533999136e-07,
+      "loss": 0.8217,
+      "step": 2610
+    },
+    {
+      "epoch": 0.8877932675960558,
+      "grad_norm": 1.8598354259713392,
+      "learning_rate": 3.2672402551971903e-07,
+      "loss": 0.7256,
+      "step": 2611
+    },
+    {
+      "epoch": 0.8881332879972799,
+      "grad_norm": 2.0484876787203614,
+      "learning_rate": 3.2476856814909364e-07,
+      "loss": 0.7733,
+      "step": 2612
+    },
+    {
+      "epoch": 0.8884733083985039,
+      "grad_norm": 4.050508948351152,
+      "learning_rate": 3.2281878366077046e-07,
+      "loss": 0.7087,
+      "step": 2613
+    },
+    {
+      "epoch": 0.888813328799728,
+      "grad_norm": 1.7444011978485319,
+      "learning_rate": 3.208746744205998e-07,
+      "loss": 0.8651,
+      "step": 2614
+    },
+    {
+      "epoch": 0.889153349200952,
+      "grad_norm": 1.7629638699985104,
+      "learning_rate": 3.1893624278754587e-07,
+      "loss": 0.7781,
+      "step": 2615
+    },
+    {
+      "epoch": 0.8894933696021762,
+      "grad_norm": 1.8022588695624528,
+      "learning_rate": 3.170034911136832e-07,
+      "loss": 0.8746,
+      "step": 2616
+    },
+    {
+      "epoch": 0.8898333900034002,
+      "grad_norm": 1.8390591568479497,
+      "learning_rate": 3.150764217441954e-07,
+      "loss": 0.6708,
+      "step": 2617
+    },
+    {
+      "epoch": 0.8901734104046243,
+      "grad_norm": 1.7855141453060075,
+      "learning_rate": 3.131550370173703e-07,
+      "loss": 0.7825,
+      "step": 2618
+    },
+    {
+      "epoch": 0.8905134308058483,
+      "grad_norm": 1.8373816915953056,
+      "learning_rate": 3.112393392645985e-07,
+      "loss": 0.7392,
+      "step": 2619
+    },
+    {
+      "epoch": 0.8908534512070724,
+      "grad_norm": 1.999434435687263,
+      "learning_rate": 3.093293308103679e-07,
+      "loss": 0.89,
+      "step": 2620
+    },
+    {
+      "epoch": 0.8911934716082965,
+      "grad_norm": 1.5593372875987372,
+      "learning_rate": 3.074250139722679e-07,
+      "loss": 0.7572,
+      "step": 2621
+    },
+    {
+      "epoch": 0.8915334920095206,
+      "grad_norm": 2.1392476418230184,
+      "learning_rate": 3.0552639106097684e-07,
+      "loss": 0.7994,
+      "step": 2622
+    },
+    {
+      "epoch": 0.8918735124107446,
+      "grad_norm": 2.4268837548387157,
+      "learning_rate": 3.0363346438026633e-07,
+      "loss": 0.8267,
+      "step": 2623
+    },
+    {
+      "epoch": 0.8922135328119687,
+      "grad_norm": 2.521048048719284,
+      "learning_rate": 3.0174623622699685e-07,
+      "loss": 0.7818,
+      "step": 2624
+    },
+    {
+      "epoch": 0.8925535532131927,
+      "grad_norm": 1.7313419670899555,
+      "learning_rate": 2.998647088911127e-07,
+      "loss": 0.7824,
+      "step": 2625
+    },
+    {
+      "epoch": 0.8928935736144169,
+      "grad_norm": 3.2464092926902364,
+      "learning_rate": 2.9798888465564226e-07,
+      "loss": 0.7654,
+      "step": 2626
+    },
+    {
+      "epoch": 0.893233594015641,
+      "grad_norm": 1.5731620759656288,
+      "learning_rate": 2.961187657966919e-07,
+      "loss": 0.8325,
+      "step": 2627
+    },
+    {
+      "epoch": 0.893573614416865,
+      "grad_norm": 2.6025793303197804,
+      "learning_rate": 2.942543545834475e-07,
+      "loss": 0.7288,
+      "step": 2628
+    },
+    {
+      "epoch": 0.893913634818089,
+      "grad_norm": 1.829882281307138,
+      "learning_rate": 2.923956532781691e-07,
+      "loss": 0.7506,
+      "step": 2629
+    },
+    {
+      "epoch": 0.8942536552193131,
+      "grad_norm": 2.3041651665798337,
+      "learning_rate": 2.9054266413618525e-07,
+      "loss": 0.7911,
+      "step": 2630
+    },
+    {
+      "epoch": 0.8945936756205373,
+      "grad_norm": 1.6349830871902187,
+      "learning_rate": 2.88695389405898e-07,
+      "loss": 0.7504,
+      "step": 2631
+    },
+    {
+      "epoch": 0.8949336960217613,
+      "grad_norm": 3.3922771842440587,
+      "learning_rate": 2.8685383132877163e-07,
+      "loss": 0.787,
+      "step": 2632
+    },
+    {
+      "epoch": 0.8952737164229854,
+      "grad_norm": 2.0614678915265525,
+      "learning_rate": 2.8501799213933646e-07,
+      "loss": 0.7534,
+      "step": 2633
+    },
+    {
+      "epoch": 0.8956137368242094,
+      "grad_norm": 1.5441939097294124,
+      "learning_rate": 2.831878740651833e-07,
+      "loss": 0.8937,
+      "step": 2634
+    },
+    {
+      "epoch": 0.8959537572254336,
+      "grad_norm": 2.0399976051493565,
+      "learning_rate": 2.8136347932695926e-07,
+      "loss": 0.6901,
+      "step": 2635
+    },
+    {
+      "epoch": 0.8962937776266576,
+      "grad_norm": 1.9641640372424263,
+      "learning_rate": 2.7954481013836744e-07,
+      "loss": 0.8211,
+      "step": 2636
+    },
+    {
+      "epoch": 0.8966337980278817,
+      "grad_norm": 2.7068467503326112,
+      "learning_rate": 2.7773186870616585e-07,
+      "loss": 0.8513,
+      "step": 2637
+    },
+    {
+      "epoch": 0.8969738184291057,
+      "grad_norm": 1.999992912921896,
+      "learning_rate": 2.759246572301599e-07,
+      "loss": 0.7835,
+      "step": 2638
+    },
+    {
+      "epoch": 0.8973138388303298,
+      "grad_norm": 1.4208853502746028,
+      "learning_rate": 2.741231779032022e-07,
+      "loss": 0.7349,
+      "step": 2639
+    },
+    {
+      "epoch": 0.8976538592315539,
+      "grad_norm": 1.803477923520495,
+      "learning_rate": 2.72327432911193e-07,
+      "loss": 0.744,
+      "step": 2640
+    },
+    {
+      "epoch": 0.897993879632778,
+      "grad_norm": 6.723981213587554,
+      "learning_rate": 2.7053742443307054e-07,
+      "loss": 0.718,
+      "step": 2641
+    },
+    {
+      "epoch": 0.898333900034002,
+      "grad_norm": 2.2565609405076814,
+      "learning_rate": 2.6875315464081566e-07,
+      "loss": 0.7945,
+      "step": 2642
+    },
+    {
+      "epoch": 0.8986739204352261,
+      "grad_norm": 2.046096256829784,
+      "learning_rate": 2.669746256994449e-07,
+      "loss": 0.73,
+      "step": 2643
+    },
+    {
+      "epoch": 0.8990139408364501,
+      "grad_norm": 1.656402743730646,
+      "learning_rate": 2.652018397670081e-07,
+      "loss": 0.7564,
+      "step": 2644
+    },
+    {
+      "epoch": 0.8993539612376743,
+      "grad_norm": 2.1387537466555,
+      "learning_rate": 2.6343479899458737e-07,
+      "loss": 0.7734,
+      "step": 2645
+    },
+    {
+      "epoch": 0.8996939816388984,
+      "grad_norm": 4.0058075533193325,
+      "learning_rate": 2.616735055262931e-07,
+      "loss": 0.7913,
+      "step": 2646
+    },
+    {
+      "epoch": 0.9000340020401224,
+      "grad_norm": 1.8790063917737332,
+      "learning_rate": 2.5991796149926306e-07,
+      "loss": 0.7609,
+      "step": 2647
+    },
+    {
+      "epoch": 0.9003740224413465,
+      "grad_norm": 1.7616231603038723,
+      "learning_rate": 2.5816816904365715e-07,
+      "loss": 0.6813,
+      "step": 2648
+    },
+    {
+      "epoch": 0.9007140428425705,
+      "grad_norm": 1.6567138767856624,
+      "learning_rate": 2.5642413028265867e-07,
+      "loss": 0.7752,
+      "step": 2649
+    },
+    {
+      "epoch": 0.9010540632437947,
+      "grad_norm": 1.5576675294760438,
+      "learning_rate": 2.546858473324676e-07,
+      "loss": 0.6574,
+      "step": 2650
+    },
+    {
+      "epoch": 0.9013940836450187,
+      "grad_norm": 1.8413136738169835,
+      "learning_rate": 2.529533223022995e-07,
+      "loss": 0.7272,
+      "step": 2651
+    },
+    {
+      "epoch": 0.9017341040462428,
+      "grad_norm": 1.7136660665303882,
+      "learning_rate": 2.5122655729438393e-07,
+      "loss": 0.8226,
+      "step": 2652
+    },
+    {
+      "epoch": 0.9020741244474668,
+      "grad_norm": 2.2558521762249355,
+      "learning_rate": 2.495055544039632e-07,
+      "loss": 0.812,
+      "step": 2653
+    },
+    {
+      "epoch": 0.9024141448486909,
+      "grad_norm": 1.7798515566519915,
+      "learning_rate": 2.477903157192846e-07,
+      "loss": 0.741,
+      "step": 2654
+    },
+    {
+      "epoch": 0.902754165249915,
+      "grad_norm": 2.098481810020344,
+      "learning_rate": 2.4608084332160277e-07,
+      "loss": 0.8253,
+      "step": 2655
+    },
+    {
+      "epoch": 0.9030941856511391,
+      "grad_norm": 1.6779439834154712,
+      "learning_rate": 2.443771392851768e-07,
+      "loss": 0.7023,
+      "step": 2656
+    },
+    {
+      "epoch": 0.9034342060523631,
+      "grad_norm": 1.9775008264088612,
+      "learning_rate": 2.4267920567726364e-07,
+      "loss": 0.7944,
+      "step": 2657
+    },
+    {
+      "epoch": 0.9037742264535872,
+      "grad_norm": 2.5593382037495216,
+      "learning_rate": 2.409870445581225e-07,
+      "loss": 0.7293,
+      "step": 2658
+    },
+    {
+      "epoch": 0.9041142468548113,
+      "grad_norm": 2.0166338772224086,
+      "learning_rate": 2.393006579810037e-07,
+      "loss": 0.7021,
+      "step": 2659
+    },
+    {
+      "epoch": 0.9044542672560354,
+      "grad_norm": 1.6435413275557214,
+      "learning_rate": 2.3762004799215422e-07,
+      "loss": 0.7309,
+      "step": 2660
+    },
+    {
+      "epoch": 0.9047942876572594,
+      "grad_norm": 2.4504668750280008,
+      "learning_rate": 2.3594521663081072e-07,
+      "loss": 0.7637,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9051343080584835,
+      "grad_norm": 1.6120218788123009,
+      "learning_rate": 2.3427616592919587e-07,
+      "loss": 0.7751,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9054743284597075,
+      "grad_norm": 2.086946712646895,
+      "learning_rate": 2.3261289791252306e-07,
+      "loss": 0.6903,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9058143488609317,
+      "grad_norm": 1.7397600903656776,
+      "learning_rate": 2.3095541459898452e-07,
+      "loss": 0.7838,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9061543692621558,
+      "grad_norm": 2.2776303771036366,
+      "learning_rate": 2.2930371799975593e-07,
+      "loss": 0.8619,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9064943896633798,
+      "grad_norm": 1.6304572281291108,
+      "learning_rate": 2.2765781011899025e-07,
+      "loss": 0.8539,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9068344100646039,
+      "grad_norm": 2.8418896975885586,
+      "learning_rate": 2.260176929538166e-07,
+      "loss": 0.9118,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9071744304658279,
+      "grad_norm": 1.7963190573302756,
+      "learning_rate": 2.243833684943375e-07,
+      "loss": 0.8397,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9075144508670521,
+      "grad_norm": 1.6299943334815419,
+      "learning_rate": 2.2275483872362835e-07,
+      "loss": 0.7385,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9078544712682761,
+      "grad_norm": 1.9271317601269167,
+      "learning_rate": 2.2113210561773124e-07,
+      "loss": 0.7455,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9081944916695002,
+      "grad_norm": 1.6366781921840725,
+      "learning_rate": 2.1951517114565446e-07,
+      "loss": 0.7428,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9085345120707242,
+      "grad_norm": 1.6460432599583494,
+      "learning_rate": 2.179040372693736e-07,
+      "loss": 0.6801,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9088745324719483,
+      "grad_norm": 3.0862578766801514,
+      "learning_rate": 2.162987059438204e-07,
+      "loss": 0.7899,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9092145528731724,
+      "grad_norm": 2.0209153785335814,
+      "learning_rate": 2.1469917911689232e-07,
+      "loss": 0.8979,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9095545732743965,
+      "grad_norm": 2.019796094061022,
+      "learning_rate": 2.1310545872943788e-07,
+      "loss": 0.7872,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9098945936756205,
+      "grad_norm": 2.4953968703645515,
+      "learning_rate": 2.115175467152636e-07,
+      "loss": 0.7581,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9102346140768446,
+      "grad_norm": 1.9678752909443469,
+      "learning_rate": 2.0993544500112706e-07,
+      "loss": 0.7204,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9105746344780686,
+      "grad_norm": 2.4724784821104935,
+      "learning_rate": 2.0835915550673492e-07,
+      "loss": 0.8005,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9109146548792928,
+      "grad_norm": 2.3974865270538066,
+      "learning_rate": 2.0678868014474328e-07,
+      "loss": 0.8121,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9112546752805168,
+      "grad_norm": 2.626755117379274,
+      "learning_rate": 2.0522402082075121e-07,
+      "loss": 0.615,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9115946956817409,
+      "grad_norm": 1.7814776861792805,
+      "learning_rate": 2.0366517943330278e-07,
+      "loss": 0.845,
+      "step": 2681
+    },
+    {
+      "epoch": 0.911934716082965,
+      "grad_norm": 1.9241174576916087,
+      "learning_rate": 2.0211215787388105e-07,
+      "loss": 0.8233,
+      "step": 2682
+    },
+    {
+      "epoch": 0.912274736484189,
+      "grad_norm": 1.6974694912568216,
+      "learning_rate": 2.0056495802690923e-07,
+      "loss": 0.8282,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9126147568854132,
+      "grad_norm": 3.7787758396910336,
+      "learning_rate": 1.9902358176974335e-07,
+      "loss": 0.7483,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9129547772866372,
+      "grad_norm": 2.0551170310858593,
+      "learning_rate": 1.974880309726762e-07,
+      "loss": 0.781,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9132947976878613,
+      "grad_norm": 1.7924440273829534,
+      "learning_rate": 1.959583074989302e-07,
+      "loss": 0.7122,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9136348180890853,
+      "grad_norm": 2.225503399226498,
+      "learning_rate": 1.9443441320465716e-07,
+      "loss": 0.7122,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9139748384903095,
+      "grad_norm": 1.8911761329043806,
+      "learning_rate": 1.9291634993893803e-07,
+      "loss": 0.6713,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9143148588915335,
+      "grad_norm": 1.8502644569783382,
+      "learning_rate": 1.9140411954377437e-07,
+      "loss": 0.6624,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9146548792927576,
+      "grad_norm": 2.653961623513032,
+      "learning_rate": 1.8989772385409445e-07,
+      "loss": 0.8623,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9149948996939816,
+      "grad_norm": 2.0652612784577316,
+      "learning_rate": 1.883971646977434e-07,
+      "loss": 0.7011,
+      "step": 2691
+    },
+    {
+      "epoch": 0.9153349200952057,
+      "grad_norm": 2.568246822202351,
+      "learning_rate": 1.8690244389548694e-07,
+      "loss": 0.6886,
+      "step": 2692
+    },
+    {
+      "epoch": 0.9156749404964298,
+      "grad_norm": 2.1817257448676104,
+      "learning_rate": 1.8541356326100436e-07,
+      "loss": 0.7835,
+      "step": 2693
+    },
+    {
+      "epoch": 0.9160149608976539,
+      "grad_norm": 1.5833475097932033,
+      "learning_rate": 1.8393052460088877e-07,
+      "loss": 0.7628,
+      "step": 2694
+    },
+    {
+      "epoch": 0.9163549812988779,
+      "grad_norm": 2.0794302266744555,
+      "learning_rate": 1.8245332971464803e-07,
+      "loss": 0.7234,
+      "step": 2695
+    },
+    {
+      "epoch": 0.916695001700102,
+      "grad_norm": 1.891383361699711,
+      "learning_rate": 1.8098198039469438e-07,
+      "loss": 0.8679,
+      "step": 2696
+    },
+    {
+      "epoch": 0.917035022101326,
+      "grad_norm": 1.847279800591813,
+      "learning_rate": 1.7951647842635035e-07,
+      "loss": 0.7993,
+      "step": 2697
+    },
+    {
+      "epoch": 0.9173750425025502,
+      "grad_norm": 1.672424819741688,
+      "learning_rate": 1.780568255878423e-07,
+      "loss": 0.7778,
+      "step": 2698
+    },
+    {
+      "epoch": 0.9177150629037742,
+      "grad_norm": 2.6871925509688985,
+      "learning_rate": 1.7660302365029969e-07,
+      "loss": 0.69,
+      "step": 2699
+    },
+    {
+      "epoch": 0.9180550833049983,
+      "grad_norm": 1.6768457562216357,
+      "learning_rate": 1.7515507437775193e-07,
+      "loss": 0.7657,
+      "step": 2700
+    },
+    {
+      "epoch": 0.9183951037062223,
+      "grad_norm": 1.809600285567411,
+      "learning_rate": 1.7371297952712752e-07,
+      "loss": 0.7147,
+      "step": 2701
+    },
+    {
+      "epoch": 0.9187351241074464,
+      "grad_norm": 1.9059976949501458,
+      "learning_rate": 1.722767408482501e-07,
+      "loss": 0.7172,
+      "step": 2702
+    },
+    {
+      "epoch": 0.9190751445086706,
+      "grad_norm": 1.9399888604391888,
+      "learning_rate": 1.7084636008383837e-07,
+      "loss": 0.7127,
+      "step": 2703
+    },
+    {
+      "epoch": 0.9194151649098946,
+      "grad_norm": 2.6413221231221455,
+      "learning_rate": 1.6942183896950458e-07,
+      "loss": 0.8735,
+      "step": 2704
+    },
+    {
+      "epoch": 0.9197551853111187,
+      "grad_norm": 2.052359841123048,
+      "learning_rate": 1.680031792337472e-07,
+      "loss": 0.7513,
+      "step": 2705
+    },
+    {
+      "epoch": 0.9200952057123427,
+      "grad_norm": 1.7351319515737746,
+      "learning_rate": 1.6659038259795644e-07,
+      "loss": 0.7591,
+      "step": 2706
+    },
+    {
+      "epoch": 0.9204352261135668,
+      "grad_norm": 3.5928987788596505,
+      "learning_rate": 1.6518345077640606e-07,
+      "loss": 0.865,
+      "step": 2707
+    },
+    {
+      "epoch": 0.9207752465147909,
+      "grad_norm": 2.475516242774065,
+      "learning_rate": 1.6378238547625436e-07,
+      "loss": 0.7076,
+      "step": 2708
+    },
+    {
+      "epoch": 0.921115266916015,
+      "grad_norm": 2.757378010289697,
+      "learning_rate": 1.6238718839753975e-07,
+      "loss": 0.8151,
+      "step": 2709
+    },
+    {
+      "epoch": 0.921455287317239,
+      "grad_norm": 2.839365711351715,
+      "learning_rate": 1.609978612331825e-07,
+      "loss": 0.7181,
+      "step": 2710
+    },
+    {
+      "epoch": 0.9217953077184631,
+      "grad_norm": 1.8003430537941647,
+      "learning_rate": 1.5961440566897913e-07,
+      "loss": 0.8018,
+      "step": 2711
+    },
+    {
+      "epoch": 0.9221353281196872,
+      "grad_norm": 2.3010251129273405,
+      "learning_rate": 1.582368233836007e-07,
+      "loss": 0.7241,
+      "step": 2712
+    },
+    {
+      "epoch": 0.9224753485209113,
+      "grad_norm": 1.7723077920834232,
+      "learning_rate": 1.5686511604859456e-07,
+      "loss": 0.8193,
+      "step": 2713
+    },
+    {
+      "epoch": 0.9228153689221353,
+      "grad_norm": 2.2601407328290364,
+      "learning_rate": 1.5549928532837544e-07,
+      "loss": 0.855,
+      "step": 2714
+    },
+    {
+      "epoch": 0.9231553893233594,
+      "grad_norm": 1.949904865719386,
+      "learning_rate": 1.5413933288023207e-07,
+      "loss": 0.8482,
+      "step": 2715
+    },
+    {
+      "epoch": 0.9234954097245834,
+      "grad_norm": 1.3958294128404651,
+      "learning_rate": 1.5278526035431673e-07,
+      "loss": 0.7979,
+      "step": 2716
+    },
+    {
+      "epoch": 0.9238354301258076,
+      "grad_norm": 3.1940088213187745,
+      "learning_rate": 1.5143706939364844e-07,
+      "loss": 0.7152,
+      "step": 2717
+    },
+    {
+      "epoch": 0.9241754505270316,
+      "grad_norm": 2.3293084372099138,
+      "learning_rate": 1.5009476163410975e-07,
+      "loss": 0.7087,
+      "step": 2718
+    },
+    {
+      "epoch": 0.9245154709282557,
+      "grad_norm": 1.724579154051739,
+      "learning_rate": 1.4875833870444334e-07,
+      "loss": 0.8299,
+      "step": 2719
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "grad_norm": 1.8586372981224857,
+      "learning_rate": 1.474278022262543e-07,
+      "loss": 0.7033,
+      "step": 2720
+    },
+    {
+      "epoch": 0.9251955117307038,
+      "grad_norm": 1.53996008803897,
+      "learning_rate": 1.4610315381400175e-07,
+      "loss": 0.7594,
+      "step": 2721
+    },
+    {
+      "epoch": 0.925535532131928,
+      "grad_norm": 2.2916687247168483,
+      "learning_rate": 1.4478439507500218e-07,
+      "loss": 0.7009,
+      "step": 2722
+    },
+    {
+      "epoch": 0.925875552533152,
+      "grad_norm": 1.9641042638955648,
+      "learning_rate": 1.4347152760942507e-07,
+      "loss": 0.6479,
+      "step": 2723
+    },
+    {
+      "epoch": 0.9262155729343761,
+      "grad_norm": 1.9501129366148464,
+      "learning_rate": 1.4216455301029274e-07,
+      "loss": 0.7925,
+      "step": 2724
+    },
+    {
+      "epoch": 0.9265555933356001,
+      "grad_norm": 2.198382597904698,
+      "learning_rate": 1.4086347286347502e-07,
+      "loss": 0.857,
+      "step": 2725
+    },
+    {
+      "epoch": 0.9268956137368242,
+      "grad_norm": 2.132363720106857,
+      "learning_rate": 1.3956828874768901e-07,
+      "loss": 0.6655,
+      "step": 2726
+    },
+    {
+      "epoch": 0.9272356341380483,
+      "grad_norm": 1.8802823835704232,
+      "learning_rate": 1.3827900223450152e-07,
+      "loss": 0.8498,
+      "step": 2727
+    },
+    {
+      "epoch": 0.9275756545392724,
+      "grad_norm": 2.4695615148666104,
+      "learning_rate": 1.3699561488831892e-07,
+      "loss": 0.767,
+      "step": 2728
+    },
+    {
+      "epoch": 0.9279156749404964,
+      "grad_norm": 1.547770137207598,
+      "learning_rate": 1.357181282663933e-07,
+      "loss": 0.8417,
+      "step": 2729
+    },
+    {
+      "epoch": 0.9282556953417205,
+      "grad_norm": 1.8562537232375893,
+      "learning_rate": 1.3444654391881306e-07,
+      "loss": 0.7578,
+      "step": 2730
+    },
+    {
+      "epoch": 0.9285957157429445,
+      "grad_norm": 2.3415279377546647,
+      "learning_rate": 1.3318086338850843e-07,
+      "loss": 0.6844,
+      "step": 2731
+    },
+    {
+      "epoch": 0.9289357361441687,
+      "grad_norm": 2.056878813634059,
+      "learning_rate": 1.3192108821124428e-07,
+      "loss": 0.8104,
+      "step": 2732
+    },
+    {
+      "epoch": 0.9292757565453927,
+      "grad_norm": 1.800632904887497,
+      "learning_rate": 1.3066721991561891e-07,
+      "loss": 0.7732,
+      "step": 2733
+    },
+    {
+      "epoch": 0.9296157769466168,
+      "grad_norm": 1.801402555086326,
+      "learning_rate": 1.2941926002306536e-07,
+      "loss": 0.754,
+      "step": 2734
+    },
+    {
+      "epoch": 0.9299557973478408,
+      "grad_norm": 2.1441261210894607,
+      "learning_rate": 1.2817721004784568e-07,
+      "loss": 0.7945,
+      "step": 2735
+    },
+    {
+      "epoch": 0.9302958177490649,
+      "grad_norm": 2.5741151332940677,
+      "learning_rate": 1.2694107149705258e-07,
+      "loss": 0.7383,
+      "step": 2736
+    },
+    {
+      "epoch": 0.930635838150289,
+      "grad_norm": 1.5286432889504458,
+      "learning_rate": 1.2571084587060466e-07,
+      "loss": 0.6856,
+      "step": 2737
+    },
+    {
+      "epoch": 0.9309758585515131,
+      "grad_norm": 11.326187707959681,
+      "learning_rate": 1.2448653466124672e-07,
+      "loss": 0.8106,
+      "step": 2738
+    },
+    {
+      "epoch": 0.9313158789527372,
+      "grad_norm": 1.7511130965052701,
+      "learning_rate": 1.2326813935454596e-07,
+      "loss": 0.7444,
+      "step": 2739
+    },
+    {
+      "epoch": 0.9316558993539612,
+      "grad_norm": 1.7617651287605105,
+      "learning_rate": 1.2205566142889257e-07,
+      "loss": 0.854,
+      "step": 2740
+    },
+    {
+      "epoch": 0.9319959197551854,
+      "grad_norm": 2.145302434001505,
+      "learning_rate": 1.2084910235549586e-07,
+      "loss": 0.8164,
+      "step": 2741
+    },
+    {
+      "epoch": 0.9323359401564094,
+      "grad_norm": 1.5133580779256988,
+      "learning_rate": 1.19648463598383e-07,
+      "loss": 0.7909,
+      "step": 2742
+    },
+    {
+      "epoch": 0.9326759605576335,
+      "grad_norm": 1.7694718002515495,
+      "learning_rate": 1.1845374661439813e-07,
+      "loss": 0.7474,
+      "step": 2743
+    },
+    {
+      "epoch": 0.9330159809588575,
+      "grad_norm": 1.624165239983796,
+      "learning_rate": 1.1726495285319883e-07,
+      "loss": 0.7366,
+      "step": 2744
+    },
+    {
+      "epoch": 0.9333560013600816,
+      "grad_norm": 2.2415272140749227,
+      "learning_rate": 1.1608208375725794e-07,
+      "loss": 0.7531,
+      "step": 2745
+    },
+    {
+      "epoch": 0.9336960217613057,
+      "grad_norm": 1.806699065485192,
+      "learning_rate": 1.1490514076185621e-07,
+      "loss": 0.7165,
+      "step": 2746
+    },
+    {
+      "epoch": 0.9340360421625298,
+      "grad_norm": 2.1552408637827267,
+      "learning_rate": 1.1373412529508687e-07,
+      "loss": 0.7309,
+      "step": 2747
+    },
+    {
+      "epoch": 0.9343760625637538,
+      "grad_norm": 1.8671167963972226,
+      "learning_rate": 1.1256903877784886e-07,
+      "loss": 0.7929,
+      "step": 2748
+    },
+    {
+      "epoch": 0.9347160829649779,
+      "grad_norm": 2.3625092627025497,
+      "learning_rate": 1.1140988262384633e-07,
+      "loss": 0.7205,
+      "step": 2749
+    },
+    {
+      "epoch": 0.9350561033662019,
+      "grad_norm": 1.8238588793473012,
+      "learning_rate": 1.1025665823958975e-07,
+      "loss": 0.78,
+      "step": 2750
+    },
+    {
+      "epoch": 0.9353961237674261,
+      "grad_norm": 2.234751620091944,
+      "learning_rate": 1.0910936702438924e-07,
+      "loss": 0.7863,
+      "step": 2751
+    },
+    {
+      "epoch": 0.9357361441686501,
+      "grad_norm": 2.5286309526580717,
+      "learning_rate": 1.0796801037035898e-07,
+      "loss": 0.7947,
+      "step": 2752
+    },
+    {
+      "epoch": 0.9360761645698742,
+      "grad_norm": 1.4596449824820212,
+      "learning_rate": 1.068325896624095e-07,
+      "loss": 0.7242,
+      "step": 2753
+    },
+    {
+      "epoch": 0.9364161849710982,
+      "grad_norm": 1.8586921437706274,
+      "learning_rate": 1.0570310627825042e-07,
+      "loss": 0.7237,
+      "step": 2754
+    },
+    {
+      "epoch": 0.9367562053723223,
+      "grad_norm": 2.1973293560111355,
+      "learning_rate": 1.0457956158838545e-07,
+      "loss": 0.7914,
+      "step": 2755
+    },
+    {
+      "epoch": 0.9370962257735465,
+      "grad_norm": 1.9486083197205395,
+      "learning_rate": 1.0346195695611461e-07,
+      "loss": 0.6833,
+      "step": 2756
+    },
+    {
+      "epoch": 0.9374362461747705,
+      "grad_norm": 2.0127688563988406,
+      "learning_rate": 1.0235029373752758e-07,
+      "loss": 0.8055,
+      "step": 2757
+    },
+    {
+      "epoch": 0.9377762665759946,
+      "grad_norm": 1.9434824986065704,
+      "learning_rate": 1.0124457328150705e-07,
+      "loss": 0.7631,
+      "step": 2758
+    },
+    {
+      "epoch": 0.9381162869772186,
+      "grad_norm": 1.8861521568684736,
+      "learning_rate": 1.0014479692972368e-07,
+      "loss": 0.8138,
+      "step": 2759
+    },
+    {
+      "epoch": 0.9384563073784427,
+      "grad_norm": 2.4202833010762235,
+      "learning_rate": 9.905096601663556e-08,
+      "loss": 0.8974,
+      "step": 2760
+    },
+    {
+      "epoch": 0.9387963277796668,
+      "grad_norm": 1.7851590249860223,
+      "learning_rate": 9.796308186948711e-08,
+      "loss": 0.8358,
+      "step": 2761
+    },
+    {
+      "epoch": 0.9391363481808909,
+      "grad_norm": 2.4290115845610325,
+      "learning_rate": 9.688114580830688e-08,
+      "loss": 0.7979,
+      "step": 2762
+    },
+    {
+      "epoch": 0.9394763685821149,
+      "grad_norm": 1.5838175734575497,
+      "learning_rate": 9.580515914590637e-08,
+      "loss": 0.8471,
+      "step": 2763
+    },
+    {
+      "epoch": 0.939816388983339,
+      "grad_norm": 2.027267479543276,
+      "learning_rate": 9.473512318787681e-08,
+      "loss": 0.6754,
+      "step": 2764
+    },
+    {
+      "epoch": 0.940156409384563,
+      "grad_norm": 1.809277908733106,
+      "learning_rate": 9.367103923259124e-08,
+      "loss": 0.7902,
+      "step": 2765
+    },
+    {
+      "epoch": 0.9404964297857872,
+      "grad_norm": 4.2105282282100225,
+      "learning_rate": 9.261290857119853e-08,
+      "loss": 0.7979,
+      "step": 2766
+    },
+    {
+      "epoch": 0.9408364501870112,
+      "grad_norm": 2.1122972531258655,
+      "learning_rate": 9.156073248762387e-08,
+      "loss": 0.7509,
+      "step": 2767
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 1.9840880876989972,
+      "learning_rate": 9.051451225856877e-08,
+      "loss": 0.6946,
+      "step": 2768
+    },
+    {
+      "epoch": 0.9415164909894593,
+      "grad_norm": 2.524068083748069,
+      "learning_rate": 8.947424915350723e-08,
+      "loss": 0.7643,
+      "step": 2769
+    },
+    {
+      "epoch": 0.9418565113906835,
+      "grad_norm": 2.133668114970484,
+      "learning_rate": 8.843994443468451e-08,
+      "loss": 0.8265,
+      "step": 2770
+    },
+    {
+      "epoch": 0.9421965317919075,
+      "grad_norm": 1.4893159029193104,
+      "learning_rate": 8.741159935711563e-08,
+      "loss": 0.8069,
+      "step": 2771
+    },
+    {
+      "epoch": 0.9425365521931316,
+      "grad_norm": 2.1763296510811716,
+      "learning_rate": 8.638921516858634e-08,
+      "loss": 0.8413,
+      "step": 2772
+    },
+    {
+      "epoch": 0.9428765725943556,
+      "grad_norm": 2.498518002319564,
+      "learning_rate": 8.537279310964763e-08,
+      "loss": 0.8262,
+      "step": 2773
+    },
+    {
+      "epoch": 0.9432165929955797,
+      "grad_norm": 8.993724154572014,
+      "learning_rate": 8.436233441361629e-08,
+      "loss": 0.8378,
+      "step": 2774
+    },
+    {
+      "epoch": 0.9435566133968039,
+      "grad_norm": 1.576329182227026,
+      "learning_rate": 8.335784030657324e-08,
+      "loss": 0.8379,
+      "step": 2775
+    },
+    {
+      "epoch": 0.9438966337980279,
+      "grad_norm": 2.424881688164331,
+      "learning_rate": 8.235931200736235e-08,
+      "loss": 0.7671,
+      "step": 2776
+    },
+    {
+      "epoch": 0.944236654199252,
+      "grad_norm": 2.807110671874095,
+      "learning_rate": 8.136675072758948e-08,
+      "loss": 0.7003,
+      "step": 2777
+    },
+    {
+      "epoch": 0.944576674600476,
+      "grad_norm": 4.154937564787449,
+      "learning_rate": 8.038015767161789e-08,
+      "loss": 0.8277,
+      "step": 2778
+    },
+    {
+      "epoch": 0.9449166950017001,
+      "grad_norm": 1.674791036585208,
+      "learning_rate": 7.939953403657164e-08,
+      "loss": 0.7474,
+      "step": 2779
+    },
+    {
+      "epoch": 0.9452567154029242,
+      "grad_norm": 2.0417610679261635,
+      "learning_rate": 7.842488101232893e-08,
+      "loss": 0.7489,
+      "step": 2780
+    },
+    {
+      "epoch": 0.9455967358041483,
+      "grad_norm": 2.355200646804559,
+      "learning_rate": 7.745619978152653e-08,
+      "loss": 0.8185,
+      "step": 2781
+    },
+    {
+      "epoch": 0.9459367562053723,
+      "grad_norm": 2.1071996152685863,
+      "learning_rate": 7.649349151955199e-08,
+      "loss": 0.7427,
+      "step": 2782
+    },
+    {
+      "epoch": 0.9462767766065964,
+      "grad_norm": 2.5068513815147413,
+      "learning_rate": 7.553675739454647e-08,
+      "loss": 0.7599,
+      "step": 2783
+    },
+    {
+      "epoch": 0.9466167970078204,
+      "grad_norm": 2.137090570866946,
+      "learning_rate": 7.4585998567403e-08,
+      "loss": 0.7743,
+      "step": 2784
+    },
+    {
+      "epoch": 0.9469568174090446,
+      "grad_norm": 1.8432992230207976,
+      "learning_rate": 7.364121619176213e-08,
+      "loss": 0.7191,
+      "step": 2785
+    },
+    {
+      "epoch": 0.9472968378102686,
+      "grad_norm": 1.9437651826974969,
+      "learning_rate": 7.270241141401568e-08,
+      "loss": 0.8466,
+      "step": 2786
+    },
+    {
+      "epoch": 0.9476368582114927,
+      "grad_norm": 1.7799878186616265,
+      "learning_rate": 7.17695853732997e-08,
+      "loss": 0.77,
+      "step": 2787
+    },
+    {
+      "epoch": 0.9479768786127167,
+      "grad_norm": 1.7842807831458771,
+      "learning_rate": 7.084273920149654e-08,
+      "loss": 0.9092,
+      "step": 2788
+    },
+    {
+      "epoch": 0.9483168990139408,
+      "grad_norm": 1.7117275262466325,
+      "learning_rate": 6.99218740232338e-08,
+      "loss": 0.8042,
+      "step": 2789
+    },
+    {
+      "epoch": 0.948656919415165,
+      "grad_norm": 1.6622479904527474,
+      "learning_rate": 6.900699095587937e-08,
+      "loss": 0.7579,
+      "step": 2790
+    },
+    {
+      "epoch": 0.948996939816389,
+      "grad_norm": 1.8529558113196523,
+      "learning_rate": 6.809809110954413e-08,
+      "loss": 0.8636,
+      "step": 2791
+    },
+    {
+      "epoch": 0.949336960217613,
+      "grad_norm": 2.163551138095109,
+      "learning_rate": 6.719517558707922e-08,
+      "loss": 0.892,
+      "step": 2792
+    },
+    {
+      "epoch": 0.9496769806188371,
+      "grad_norm": 1.8487943962662825,
+      "learning_rate": 6.629824548407381e-08,
+      "loss": 0.7987,
+      "step": 2793
+    },
+    {
+      "epoch": 0.9500170010200613,
+      "grad_norm": 2.0154438939507497,
+      "learning_rate": 6.540730188885347e-08,
+      "loss": 0.7981,
+      "step": 2794
+    },
+    {
+      "epoch": 0.9503570214212853,
+      "grad_norm": 1.9570984130674787,
+      "learning_rate": 6.452234588248285e-08,
+      "loss": 0.7771,
+      "step": 2795
+    },
+    {
+      "epoch": 0.9506970418225094,
+      "grad_norm": 1.6604363995442626,
+      "learning_rate": 6.364337853875745e-08,
+      "loss": 0.654,
+      "step": 2796
+    },
+    {
+      "epoch": 0.9510370622237334,
+      "grad_norm": 1.8616496731827172,
+      "learning_rate": 6.277040092420916e-08,
+      "loss": 0.7682,
+      "step": 2797
+    },
+    {
+      "epoch": 0.9513770826249575,
+      "grad_norm": 1.4864670570007181,
+      "learning_rate": 6.190341409810063e-08,
+      "loss": 0.7729,
+      "step": 2798
+    },
+    {
+      "epoch": 0.9517171030261816,
+      "grad_norm": 2.167348761585454,
+      "learning_rate": 6.104241911242592e-08,
+      "loss": 0.8381,
+      "step": 2799
+    },
+    {
+      "epoch": 0.9520571234274057,
+      "grad_norm": 2.0214824467624606,
+      "learning_rate": 6.018741701190767e-08,
+      "loss": 0.8774,
+      "step": 2800
+    },
+    {
+      "epoch": 0.9523971438286297,
+      "grad_norm": 2.046012942264478,
+      "learning_rate": 5.933840883399766e-08,
+      "loss": 0.8185,
+      "step": 2801
+    },
+    {
+      "epoch": 0.9527371642298538,
+      "grad_norm": 1.6744820889133898,
+      "learning_rate": 5.8495395608874625e-08,
+      "loss": 0.7855,
+      "step": 2802
+    },
+    {
+      "epoch": 0.9530771846310778,
+      "grad_norm": 1.9991594031620603,
+      "learning_rate": 5.7658378359443104e-08,
+      "loss": 0.7734,
+      "step": 2803
+    },
+    {
+      "epoch": 0.953417205032302,
+      "grad_norm": 1.8896007551548046,
+      "learning_rate": 5.6827358101331774e-08,
+      "loss": 0.753,
+      "step": 2804
+    },
+    {
+      "epoch": 0.953757225433526,
+      "grad_norm": 2.499578946675649,
+      "learning_rate": 5.600233584289294e-08,
+      "loss": 0.7645,
+      "step": 2805
+    },
+    {
+      "epoch": 0.9540972458347501,
+      "grad_norm": 2.0291966720592898,
+      "learning_rate": 5.518331258520138e-08,
+      "loss": 0.7063,
+      "step": 2806
+    },
+    {
+      "epoch": 0.9544372662359741,
+      "grad_norm": 2.115533353024901,
+      "learning_rate": 5.437028932205213e-08,
+      "loss": 0.7353,
+      "step": 2807
+    },
+    {
+      "epoch": 0.9547772866371982,
+      "grad_norm": 2.230634267024501,
+      "learning_rate": 5.356326703995884e-08,
+      "loss": 0.7527,
+      "step": 2808
+    },
+    {
+      "epoch": 0.9551173070384223,
+      "grad_norm": 1.9500468263820696,
+      "learning_rate": 5.276224671815655e-08,
+      "loss": 0.7196,
+      "step": 2809
+    },
+    {
+      "epoch": 0.9554573274396464,
+      "grad_norm": 1.9933718903005513,
+      "learning_rate": 5.196722932859499e-08,
+      "loss": 0.7947,
+      "step": 2810
+    },
+    {
+      "epoch": 0.9557973478408704,
+      "grad_norm": 3.9589726805428658,
+      "learning_rate": 5.117821583594085e-08,
+      "loss": 0.8693,
+      "step": 2811
+    },
+    {
+      "epoch": 0.9561373682420945,
+      "grad_norm": 1.697109589336965,
+      "learning_rate": 5.0395207197575516e-08,
+      "loss": 0.8559,
+      "step": 2812
+    },
+    {
+      "epoch": 0.9564773886433185,
+      "grad_norm": 1.7784029867244318,
+      "learning_rate": 4.9618204363595656e-08,
+      "loss": 0.7196,
+      "step": 2813
+    },
+    {
+      "epoch": 0.9568174090445427,
+      "grad_norm": 1.7135140822695198,
+      "learning_rate": 4.8847208276808224e-08,
+      "loss": 0.7667,
+      "step": 2814
+    },
+    {
+      "epoch": 0.9571574294457668,
+      "grad_norm": 2.0635249556468676,
+      "learning_rate": 4.808221987273265e-08,
+      "loss": 0.7779,
+      "step": 2815
+    },
+    {
+      "epoch": 0.9574974498469908,
+      "grad_norm": 1.7283710035195046,
+      "learning_rate": 4.732324007959921e-08,
+      "loss": 0.7874,
+      "step": 2816
+    },
+    {
+      "epoch": 0.9578374702482149,
+      "grad_norm": 2.1716890834730247,
+      "learning_rate": 4.657026981834623e-08,
+      "loss": 0.776,
+      "step": 2817
+    },
+    {
+      "epoch": 0.9581774906494389,
+      "grad_norm": 1.8440991529889135,
+      "learning_rate": 4.5823310002621745e-08,
+      "loss": 0.7675,
+      "step": 2818
+    },
+    {
+      "epoch": 0.9585175110506631,
+      "grad_norm": 1.9867632028220694,
+      "learning_rate": 4.5082361538779095e-08,
+      "loss": 0.8159,
+      "step": 2819
+    },
+    {
+      "epoch": 0.9588575314518871,
+      "grad_norm": 1.6735880869924653,
+      "learning_rate": 4.434742532587855e-08,
+      "loss": 0.8242,
+      "step": 2820
+    },
+    {
+      "epoch": 0.9591975518531112,
+      "grad_norm": 1.9489756840557746,
+      "learning_rate": 4.3618502255684533e-08,
+      "loss": 0.7493,
+      "step": 2821
+    },
+    {
+      "epoch": 0.9595375722543352,
+      "grad_norm": 1.8908904594168388,
+      "learning_rate": 4.289559321266623e-08,
+      "loss": 0.7569,
+      "step": 2822
+    },
+    {
+      "epoch": 0.9598775926555594,
+      "grad_norm": 2.0307318453443663,
+      "learning_rate": 4.2178699073994744e-08,
+      "loss": 0.8468,
+      "step": 2823
+    },
+    {
+      "epoch": 0.9602176130567834,
+      "grad_norm": 2.155622909096085,
+      "learning_rate": 4.1467820709541474e-08,
+      "loss": 0.7998,
+      "step": 2824
+    },
+    {
+      "epoch": 0.9605576334580075,
+      "grad_norm": 3.045935542421655,
+      "learning_rate": 4.0762958981880876e-08,
+      "loss": 0.6656,
+      "step": 2825
+    },
+    {
+      "epoch": 0.9608976538592315,
+      "grad_norm": 1.462100425448403,
+      "learning_rate": 4.006411474628491e-08,
+      "loss": 0.7495,
+      "step": 2826
+    },
+    {
+      "epoch": 0.9612376742604556,
+      "grad_norm": 2.0395632850495873,
+      "learning_rate": 3.937128885072528e-08,
+      "loss": 0.7343,
+      "step": 2827
+    },
+    {
+      "epoch": 0.9615776946616797,
+      "grad_norm": 1.4943247088061522,
+      "learning_rate": 3.868448213587006e-08,
+      "loss": 0.7365,
+      "step": 2828
+    },
+    {
+      "epoch": 0.9619177150629038,
+      "grad_norm": 1.6462875432934096,
+      "learning_rate": 3.800369543508431e-08,
+      "loss": 0.7184,
+      "step": 2829
+    },
+    {
+      "epoch": 0.9622577354641279,
+      "grad_norm": 1.7317699680459382,
+      "learning_rate": 3.7328929574428354e-08,
+      "loss": 0.7887,
+      "step": 2830
+    },
+    {
+      "epoch": 0.9625977558653519,
+      "grad_norm": 2.0339479815422683,
+      "learning_rate": 3.6660185372656144e-08,
+      "loss": 0.6979,
+      "step": 2831
+    },
+    {
+      "epoch": 0.962937776266576,
+      "grad_norm": 2.024001193970488,
+      "learning_rate": 3.5997463641216925e-08,
+      "loss": 0.7664,
+      "step": 2832
+    },
+    {
+      "epoch": 0.9632777966678001,
+      "grad_norm": 2.7918239147503123,
+      "learning_rate": 3.534076518424967e-08,
+      "loss": 0.7498,
+      "step": 2833
+    },
+    {
+      "epoch": 0.9636178170690242,
+      "grad_norm": 1.8084785912714854,
+      "learning_rate": 3.469009079858698e-08,
+      "loss": 0.5801,
+      "step": 2834
+    },
+    {
+      "epoch": 0.9639578374702482,
+      "grad_norm": 1.6988220138175392,
+      "learning_rate": 3.404544127375064e-08,
+      "loss": 0.7183,
+      "step": 2835
+    },
+    {
+      "epoch": 0.9642978578714723,
+      "grad_norm": 1.7500355302390025,
+      "learning_rate": 3.340681739195328e-08,
+      "loss": 0.8591,
+      "step": 2836
+    },
+    {
+      "epoch": 0.9646378782726963,
+      "grad_norm": 2.0915983335526542,
+      "learning_rate": 3.277421992809448e-08,
+      "loss": 0.7898,
+      "step": 2837
+    },
+    {
+      "epoch": 0.9649778986739205,
+      "grad_norm": 1.9905149738010572,
+      "learning_rate": 3.2147649649761914e-08,
+      "loss": 0.7486,
+      "step": 2838
+    },
+    {
+      "epoch": 0.9653179190751445,
+      "grad_norm": 1.8204792814385986,
+      "learning_rate": 3.152710731723019e-08,
+      "loss": 0.7907,
+      "step": 2839
+    },
+    {
+      "epoch": 0.9656579394763686,
+      "grad_norm": 2.2038325461595125,
+      "learning_rate": 3.0912593683460336e-08,
+      "loss": 0.8221,
+      "step": 2840
+    },
+    {
+      "epoch": 0.9659979598775926,
+      "grad_norm": 2.432522698561901,
+      "learning_rate": 3.030410949409701e-08,
+      "loss": 0.8842,
+      "step": 2841
+    },
+    {
+      "epoch": 0.9663379802788167,
+      "grad_norm": 1.7618523901465664,
+      "learning_rate": 2.9701655487469062e-08,
+      "loss": 0.7653,
+      "step": 2842
+    },
+    {
+      "epoch": 0.9666780006800408,
+      "grad_norm": 2.0644672247287055,
+      "learning_rate": 2.9105232394588955e-08,
+      "loss": 0.7707,
+      "step": 2843
+    },
+    {
+      "epoch": 0.9670180210812649,
+      "grad_norm": 1.63818941154093,
+      "learning_rate": 2.8514840939150023e-08,
+      "loss": 0.8413,
+      "step": 2844
+    },
+    {
+      "epoch": 0.9673580414824889,
+      "grad_norm": 1.973884901127744,
+      "learning_rate": 2.793048183752922e-08,
+      "loss": 0.7836,
+      "step": 2845
+    },
+    {
+      "epoch": 0.967698061883713,
+      "grad_norm": 2.4371181181523354,
+      "learning_rate": 2.735215579878159e-08,
+      "loss": 0.7305,
+      "step": 2846
+    },
+    {
+      "epoch": 0.9680380822849372,
+      "grad_norm": 1.7853562497323494,
+      "learning_rate": 2.6779863524642458e-08,
+      "loss": 0.7943,
+      "step": 2847
+    },
+    {
+      "epoch": 0.9683781026861612,
+      "grad_norm": 1.6858828008976603,
+      "learning_rate": 2.6213605709525803e-08,
+      "loss": 0.7232,
+      "step": 2848
+    },
+    {
+      "epoch": 0.9687181230873853,
+      "grad_norm": 2.158947668013324,
+      "learning_rate": 2.5653383040524228e-08,
+      "loss": 0.8468,
+      "step": 2849
+    },
+    {
+      "epoch": 0.9690581434886093,
+      "grad_norm": 1.9632625315112604,
+      "learning_rate": 2.509919619740675e-08,
+      "loss": 0.7163,
+      "step": 2850
+    },
+    {
+      "epoch": 0.9693981638898334,
+      "grad_norm": 2.1695249086671757,
+      "learning_rate": 2.4551045852617694e-08,
+      "loss": 0.6472,
+      "step": 2851
+    },
+    {
+      "epoch": 0.9697381842910575,
+      "grad_norm": 2.7117847345646107,
+      "learning_rate": 2.4008932671277795e-08,
+      "loss": 0.8502,
+      "step": 2852
+    },
+    {
+      "epoch": 0.9700782046922816,
+      "grad_norm": 6.101355782708873,
+      "learning_rate": 2.3472857311183095e-08,
+      "loss": 0.8528,
+      "step": 2853
+    },
+    {
+      "epoch": 0.9704182250935056,
+      "grad_norm": 2.254936461781147,
+      "learning_rate": 2.294282042280105e-08,
+      "loss": 0.7906,
+      "step": 2854
+    },
+    {
+      "epoch": 0.9707582454947297,
+      "grad_norm": 1.8108204204000768,
+      "learning_rate": 2.2418822649274974e-08,
+      "loss": 0.7968,
+      "step": 2855
+    },
+    {
+      "epoch": 0.9710982658959537,
+      "grad_norm": 2.6951434301051402,
+      "learning_rate": 2.1900864626417385e-08,
+      "loss": 0.776,
+      "step": 2856
+    },
+    {
+      "epoch": 0.9714382862971779,
+      "grad_norm": 1.6020976275419296,
+      "learning_rate": 2.1388946982714986e-08,
+      "loss": 0.7704,
+      "step": 2857
+    },
+    {
+      "epoch": 0.9717783066984019,
+      "grad_norm": 2.4901873723353556,
+      "learning_rate": 2.088307033932313e-08,
+      "loss": 0.7724,
+      "step": 2858
+    },
+    {
+      "epoch": 0.972118327099626,
+      "grad_norm": 1.9510473315864627,
+      "learning_rate": 2.0383235310068027e-08,
+      "loss": 0.7161,
+      "step": 2859
+    },
+    {
+      "epoch": 0.97245834750085,
+      "grad_norm": 1.7537820967663633,
+      "learning_rate": 1.9889442501444533e-08,
+      "loss": 0.7845,
+      "step": 2860
+    },
+    {
+      "epoch": 0.9727983679020741,
+      "grad_norm": 2.770379816331149,
+      "learning_rate": 1.9401692512617254e-08,
+      "loss": 0.7648,
+      "step": 2861
+    },
+    {
+      "epoch": 0.9731383883032982,
+      "grad_norm": 1.559159771690986,
+      "learning_rate": 1.891998593541611e-08,
+      "loss": 0.8496,
+      "step": 2862
+    },
+    {
+      "epoch": 0.9734784087045223,
+      "grad_norm": 1.8239680223638182,
+      "learning_rate": 1.8444323354340765e-08,
+      "loss": 0.8466,
+      "step": 2863
+    },
+    {
+      "epoch": 0.9738184291057463,
+      "grad_norm": 2.097607953123935,
+      "learning_rate": 1.7974705346554543e-08,
+      "loss": 0.7707,
+      "step": 2864
+    },
+    {
+      "epoch": 0.9741584495069704,
+      "grad_norm": 15.595515977393774,
+      "learning_rate": 1.7511132481888293e-08,
+      "loss": 0.8303,
+      "step": 2865
+    },
+    {
+      "epoch": 0.9744984699081944,
+      "grad_norm": 2.2442397263635923,
+      "learning_rate": 1.7053605322837064e-08,
+      "loss": 0.8331,
+      "step": 2866
+    },
+    {
+      "epoch": 0.9748384903094186,
+      "grad_norm": 1.7730283098392012,
+      "learning_rate": 1.6602124424558998e-08,
+      "loss": 0.8179,
+      "step": 2867
+    },
+    {
+      "epoch": 0.9751785107106427,
+      "grad_norm": 1.9544407790331693,
+      "learning_rate": 1.6156690334878655e-08,
+      "loss": 0.8524,
+      "step": 2868
+    },
+    {
+      "epoch": 0.9755185311118667,
+      "grad_norm": 1.8253723434879328,
+      "learning_rate": 1.571730359427981e-08,
+      "loss": 0.7918,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9758585515130908,
+      "grad_norm": 2.0309536185191175,
+      "learning_rate": 1.5283964735911537e-08,
+      "loss": 0.6513,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9761985719143148,
+      "grad_norm": 1.838769072190465,
+      "learning_rate": 1.4856674285582128e-08,
+      "loss": 0.8031,
+      "step": 2871
+    },
+    {
+      "epoch": 0.976538592315539,
+      "grad_norm": 2.0950822778973692,
+      "learning_rate": 1.4435432761762958e-08,
+      "loss": 0.7676,
+      "step": 2872
+    },
+    {
+      "epoch": 0.976878612716763,
+      "grad_norm": 2.0893793425469873,
+      "learning_rate": 1.4020240675583496e-08,
+      "loss": 0.7125,
+      "step": 2873
+    },
+    {
+      "epoch": 0.9772186331179871,
+      "grad_norm": 2.4146069116338076,
+      "learning_rate": 1.3611098530834643e-08,
+      "loss": 0.7995,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9775586535192111,
+      "grad_norm": 1.6888893378242318,
+      "learning_rate": 1.3208006823965391e-08,
+      "loss": 0.7872,
+      "step": 2875
+    },
+    {
+      "epoch": 0.9778986739204353,
+      "grad_norm": 1.8282106707284866,
+      "learning_rate": 1.2810966044083384e-08,
+      "loss": 0.7569,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9782386943216593,
+      "grad_norm": 3.0888469815054056,
+      "learning_rate": 1.241997667295436e-08,
+      "loss": 0.7845,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9785787147228834,
+      "grad_norm": 2.6085632259240135,
+      "learning_rate": 1.2035039185001595e-08,
+      "loss": 0.6891,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9789187351241074,
+      "grad_norm": 1.6548350534770722,
+      "learning_rate": 1.1656154047303691e-08,
+      "loss": 0.776,
+      "step": 2879
+    },
+    {
+      "epoch": 0.9792587555253315,
+      "grad_norm": 2.2479317743891163,
+      "learning_rate": 1.128332171959734e-08,
+      "loss": 0.6986,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9795987759265556,
+      "grad_norm": 1.5207555564257944,
+      "learning_rate": 1.0916542654273443e-08,
+      "loss": 0.7656,
+      "step": 2881
+    },
+    {
+      "epoch": 0.9799387963277797,
+      "grad_norm": 1.7179799866128536,
+      "learning_rate": 1.0555817296378223e-08,
+      "loss": 0.7789,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9802788167290037,
+      "grad_norm": 3.5457830250161146,
+      "learning_rate": 1.0201146083612113e-08,
+      "loss": 0.6575,
+      "step": 2883
+    },
+    {
+      "epoch": 0.9806188371302278,
+      "grad_norm": 1.66796935458302,
+      "learning_rate": 9.852529446330306e-09,
+      "loss": 0.8078,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9809588575314518,
+      "grad_norm": 1.6131158456304904,
+      "learning_rate": 9.509967807541098e-09,
+      "loss": 0.8186,
+      "step": 2885
+    },
+    {
+      "epoch": 0.981298877932676,
+      "grad_norm": 1.82468180774407,
+      "learning_rate": 9.17346158290533e-09,
+      "loss": 0.739,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9816388983339001,
+      "grad_norm": 1.9688243210497778,
+      "learning_rate": 8.843011180736383e-09,
+      "loss": 0.7168,
+      "step": 2887
+    },
+    {
+      "epoch": 0.9819789187351241,
+      "grad_norm": 1.5022510538028564,
+      "learning_rate": 8.518617002000184e-09,
+      "loss": 0.7572,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9823189391363482,
+      "grad_norm": 1.727052969368207,
+      "learning_rate": 8.200279440313541e-09,
+      "loss": 0.8521,
+      "step": 2889
+    },
+    {
+      "epoch": 0.9826589595375722,
+      "grad_norm": 2.2067244626842633,
+      "learning_rate": 7.88799888194358e-09,
+      "loss": 0.8,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9829989799387964,
+      "grad_norm": 1.7065083718462186,
+      "learning_rate": 7.581775705809424e-09,
+      "loss": 0.7378,
+      "step": 2891
+    },
+    {
+      "epoch": 0.9833390003400204,
+      "grad_norm": 2.9586498581307064,
+      "learning_rate": 7.281610283479401e-09,
+      "loss": 0.6801,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9836790207412445,
+      "grad_norm": 5.171238278863206,
+      "learning_rate": 6.987502979170502e-09,
+      "loss": 0.7629,
+      "step": 2893
+    },
+    {
+      "epoch": 0.9840190411424685,
+      "grad_norm": 1.90520126690055,
+      "learning_rate": 6.69945414975115e-09,
+      "loss": 0.8113,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9843590615436926,
+      "grad_norm": 2.2503726394658616,
+      "learning_rate": 6.417464144736208e-09,
+      "loss": 0.7513,
+      "step": 2895
+    },
+    {
+      "epoch": 0.9846990819449167,
+      "grad_norm": 2.153027725123987,
+      "learning_rate": 6.141533306289749e-09,
+      "loss": 0.8205,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9850391023461408,
+      "grad_norm": 1.736801672039933,
+      "learning_rate": 5.871661969223951e-09,
+      "loss": 0.7942,
+      "step": 2897
+    },
+    {
+      "epoch": 0.9853791227473648,
+      "grad_norm": 2.3675066320350475,
+      "learning_rate": 5.6078504609979874e-09,
+      "loss": 0.7259,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9857191431485889,
+      "grad_norm": 2.017743519753625,
+      "learning_rate": 5.350099101718575e-09,
+      "loss": 0.7131,
+      "step": 2899
+    },
+    {
+      "epoch": 0.9860591635498129,
+      "grad_norm": 2.7453562229219703,
+      "learning_rate": 5.098408204138872e-09,
+      "loss": 0.7977,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9863991839510371,
+      "grad_norm": 1.6134295512276415,
+      "learning_rate": 4.852778073657361e-09,
+      "loss": 0.7135,
+      "step": 2901
+    },
+    {
+      "epoch": 0.9867392043522611,
+      "grad_norm": 2.0405012351886276,
+      "learning_rate": 4.613209008320629e-09,
+      "loss": 0.7537,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9870792247534852,
+      "grad_norm": 1.4871414756238235,
+      "learning_rate": 4.379701298818928e-09,
+      "loss": 0.7353,
+      "step": 2903
+    },
+    {
+      "epoch": 0.9874192451547092,
+      "grad_norm": 2.093132630125787,
+      "learning_rate": 4.152255228487834e-09,
+      "loss": 0.7546,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9877592655559334,
+      "grad_norm": 1.7225547695335006,
+      "learning_rate": 3.9308710733093616e-09,
+      "loss": 0.7336,
+      "step": 2905
+    },
+    {
+      "epoch": 0.9880992859571575,
+      "grad_norm": 2.05442384710289,
+      "learning_rate": 3.715549101908633e-09,
+      "loss": 0.7587,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9884393063583815,
+      "grad_norm": 5.175498096828514,
+      "learning_rate": 3.5062895755544337e-09,
+      "loss": 0.7976,
+      "step": 2907
+    },
+    {
+      "epoch": 0.9887793267596056,
+      "grad_norm": 2.1043805063081678,
+      "learning_rate": 3.3030927481614294e-09,
+      "loss": 0.8033,
+      "step": 2908
+    },
+    {
+      "epoch": 0.9891193471608296,
+      "grad_norm": 2.0514526782549884,
+      "learning_rate": 3.10595886628684e-09,
+      "loss": 0.7745,
+      "step": 2909
+    },
+    {
+      "epoch": 0.9894593675620538,
+      "grad_norm": 1.8077379168303007,
+      "learning_rate": 2.9148881691298812e-09,
+      "loss": 0.8329,
+      "step": 2910
+    },
+    {
+      "epoch": 0.9897993879632778,
+      "grad_norm": 2.7491338803401852,
+      "learning_rate": 2.7298808885350968e-09,
+      "loss": 0.7437,
+      "step": 2911
+    },
+    {
+      "epoch": 0.9901394083645019,
+      "grad_norm": 1.8239821715791795,
+      "learning_rate": 2.550937248987917e-09,
+      "loss": 0.752,
+      "step": 2912
+    },
+    {
+      "epoch": 0.9904794287657259,
+      "grad_norm": 1.9332024110813362,
+      "learning_rate": 2.378057467617434e-09,
+      "loss": 0.7285,
+      "step": 2913
+    },
+    {
+      "epoch": 0.99081944916695,
+      "grad_norm": 2.1222346021287284,
+      "learning_rate": 2.211241754193627e-09,
+      "loss": 0.7529,
+      "step": 2914
+    },
+    {
+      "epoch": 0.9911594695681741,
+      "grad_norm": 2.949500732515013,
+      "learning_rate": 2.050490311130138e-09,
+      "loss": 0.7769,
+      "step": 2915
+    },
+    {
+      "epoch": 0.9914994899693982,
+      "grad_norm": 2.2876307190277387,
+      "learning_rate": 1.8958033334803837e-09,
+      "loss": 0.7332,
+      "step": 2916
+    },
+    {
+      "epoch": 0.9918395103706222,
+      "grad_norm": 1.8659398649610075,
+      "learning_rate": 1.7471810089403352e-09,
+      "loss": 0.7737,
+      "step": 2917
+    },
+    {
+      "epoch": 0.9921795307718463,
+      "grad_norm": 4.208386889103378,
+      "learning_rate": 1.6046235178474034e-09,
+      "loss": 0.6483,
+      "step": 2918
+    },
+    {
+      "epoch": 0.9925195511730703,
+      "grad_norm": 2.7900149310075437,
+      "learning_rate": 1.4681310331787767e-09,
+      "loss": 0.7181,
+      "step": 2919
+    },
+    {
+      "epoch": 0.9928595715742945,
+      "grad_norm": 3.2570105193883836,
+      "learning_rate": 1.3377037205541954e-09,
+      "loss": 0.745,
+      "step": 2920
+    },
+    {
+      "epoch": 0.9931995919755185,
+      "grad_norm": 2.253063484591565,
+      "learning_rate": 1.2133417382320656e-09,
+      "loss": 0.7365,
+      "step": 2921
+    },
+    {
+      "epoch": 0.9935396123767426,
+      "grad_norm": 1.6986142527464747,
+      "learning_rate": 1.0950452371116805e-09,
+      "loss": 0.7249,
+      "step": 2922
+    },
+    {
+      "epoch": 0.9938796327779666,
+      "grad_norm": 2.1769362876424783,
+      "learning_rate": 9.828143607343298e-10,
+      "loss": 0.675,
+      "step": 2923
+    },
+    {
+      "epoch": 0.9942196531791907,
+      "grad_norm": 1.8623993129979777,
+      "learning_rate": 8.766492452783048e-10,
+      "loss": 0.7031,
+      "step": 2924
+    },
+    {
+      "epoch": 0.9945596735804149,
+      "grad_norm": 2.8223920826039577,
+      "learning_rate": 7.765500195650034e-10,
+      "loss": 0.7948,
+      "step": 2925
+    },
+    {
+      "epoch": 0.9948996939816389,
+      "grad_norm": 1.8686593078473168,
+      "learning_rate": 6.825168050528241e-10,
+      "loss": 0.727,
+      "step": 2926
+    },
+    {
+      "epoch": 0.995239714382863,
+      "grad_norm": 2.162089701353925,
+      "learning_rate": 5.945497158404979e-10,
+      "loss": 0.7878,
+      "step": 2927
+    },
+    {
+      "epoch": 0.995579734784087,
+      "grad_norm": 2.0618636520713265,
+      "learning_rate": 5.126488586676414e-10,
+      "loss": 0.681,
+      "step": 2928
+    },
+    {
+      "epoch": 0.9959197551853112,
+      "grad_norm": 2.0577534585653656,
+      "learning_rate": 4.368143329114283e-10,
+      "loss": 0.7572,
+      "step": 2929
+    },
+    {
+      "epoch": 0.9962597755865352,
+      "grad_norm": 1.9033317775907852,
+      "learning_rate": 3.6704623058825275e-10,
+      "loss": 0.725,
+      "step": 2930
+    },
+    {
+      "epoch": 0.9965997959877593,
+      "grad_norm": 1.7966473182708045,
+      "learning_rate": 3.033446363548409e-10,
+      "loss": 0.7165,
+      "step": 2931
+    },
+    {
+      "epoch": 0.9969398163889833,
+      "grad_norm": 1.9439610520049038,
+      "learning_rate": 2.4570962750547487e-10,
+      "loss": 0.806,
+      "step": 2932
+    },
+    {
+      "epoch": 0.9972798367902074,
+      "grad_norm": 1.8447978817356736,
+      "learning_rate": 1.9414127397476834e-10,
+      "loss": 0.6776,
+      "step": 2933
+    },
+    {
+      "epoch": 0.9976198571914315,
+      "grad_norm": 1.8689492566426245,
+      "learning_rate": 1.486396383343358e-10,
+      "loss": 0.7369,
+      "step": 2934
+    },
+    {
+      "epoch": 0.9979598775926556,
+      "grad_norm": 2.3593280079362984,
+      "learning_rate": 1.0920477579612342e-10,
+      "loss": 0.737,
+      "step": 2935
+    },
+    {
+      "epoch": 0.9982998979938796,
+      "grad_norm": 1.8130326623567172,
+      "learning_rate": 7.583673420963333e-11,
+      "loss": 0.7728,
+      "step": 2936
+    },
+    {
+      "epoch": 0.9986399183951037,
+      "grad_norm": 1.8422384775818308,
+      "learning_rate": 4.8535554063589006e-11,
+      "loss": 0.7864,
+      "step": 2937
+    },
+    {
+      "epoch": 0.9989799387963277,
+      "grad_norm": 2.6330920946601504,
+      "learning_rate": 2.7301268484825062e-11,
+      "loss": 0.7965,
+      "step": 2938
+    },
+    {
+      "epoch": 0.9993199591975519,
+      "grad_norm": 1.4728667367747035,
+      "learning_rate": 1.2133903238842337e-11,
+      "loss": 0.6569,
+      "step": 2939
+    },
+    {
+      "epoch": 0.999659979598776,
+      "grad_norm": 2.449775970829956,
+      "learning_rate": 3.033476729807916e-12,
+      "loss": 0.7428,
+      "step": 2940
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.9956841641574536,
+      "learning_rate": 0.0,
+      "loss": 0.7876,
+      "step": 2941
+    },
+    {
+      "epoch": 1.0,
+      "step": 2941,
+      "total_flos": 3322796039995392.0,
+      "train_loss": 0.8056760676374892,
+      "train_runtime": 111614.1149,
+      "train_samples_per_second": 0.843,
+      "train_steps_per_second": 0.026
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2941,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3322796039995392.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}