diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12698 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.999815668202765,
+  "eval_steps": 500,
+  "global_step": 1808,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0005529953917050691,
+      "grad_norm": 5.9790191650390625,
+      "learning_rate": 5.524861878453039e-08,
+      "loss": 0.8575,
+      "step": 1
+    },
+    {
+      "epoch": 0.0011059907834101382,
+      "grad_norm": 5.738432884216309,
+      "learning_rate": 1.1049723756906078e-07,
+      "loss": 0.8535,
+      "step": 2
+    },
+    {
+      "epoch": 0.0016589861751152074,
+      "grad_norm": 5.905203342437744,
+      "learning_rate": 1.6574585635359117e-07,
+      "loss": 0.8907,
+      "step": 3
+    },
+    {
+      "epoch": 0.0022119815668202765,
+      "grad_norm": 5.97217321395874,
+      "learning_rate": 2.2099447513812156e-07,
+      "loss": 0.8695,
+      "step": 4
+    },
+    {
+      "epoch": 0.0027649769585253456,
+      "grad_norm": 6.003969669342041,
+      "learning_rate": 2.7624309392265196e-07,
+      "loss": 0.8746,
+      "step": 5
+    },
+    {
+      "epoch": 0.0033179723502304147,
+      "grad_norm": 5.878486156463623,
+      "learning_rate": 3.3149171270718233e-07,
+      "loss": 0.866,
+      "step": 6
+    },
+    {
+      "epoch": 0.003870967741935484,
+      "grad_norm": 5.74072265625,
+      "learning_rate": 3.867403314917127e-07,
+      "loss": 0.8351,
+      "step": 7
+    },
+    {
+      "epoch": 0.004423963133640553,
+      "grad_norm": 5.6800150871276855,
+      "learning_rate": 4.419889502762431e-07,
+      "loss": 0.8534,
+      "step": 8
+    },
+    {
+      "epoch": 0.004976958525345622,
+      "grad_norm": 5.634627819061279,
+      "learning_rate": 4.972375690607735e-07,
+      "loss": 0.8592,
+      "step": 9
+    },
+    {
+      "epoch": 0.005529953917050691,
+      "grad_norm": 5.71673059463501,
+      "learning_rate": 5.524861878453039e-07,
+      "loss": 0.8825,
+      "step": 10
+    },
+    {
+      "epoch": 0.00608294930875576,
+      "grad_norm": 5.533100605010986,
+      "learning_rate": 6.077348066298343e-07,
+      "loss": 0.8662,
+      "step": 11
+    },
+    {
+      "epoch": 0.0066359447004608295,
+      "grad_norm": 5.378372669219971,
+      "learning_rate": 6.629834254143647e-07,
+      "loss": 0.8688,
+      "step": 12
+    },
+    {
+      "epoch": 0.007188940092165899,
+      "grad_norm": 5.282576560974121,
+      "learning_rate": 7.18232044198895e-07,
+      "loss": 0.8179,
+      "step": 13
+    },
+    {
+      "epoch": 0.007741935483870968,
+      "grad_norm": 4.443999290466309,
+      "learning_rate": 7.734806629834254e-07,
+      "loss": 0.829,
+      "step": 14
+    },
+    {
+      "epoch": 0.008294930875576038,
+      "grad_norm": 4.481590270996094,
+      "learning_rate": 8.287292817679559e-07,
+      "loss": 0.8441,
+      "step": 15
+    },
+    {
+      "epoch": 0.008847926267281106,
+      "grad_norm": 4.04002571105957,
+      "learning_rate": 8.839779005524863e-07,
+      "loss": 0.7884,
+      "step": 16
+    },
+    {
+      "epoch": 0.009400921658986176,
+      "grad_norm": 3.9912734031677246,
+      "learning_rate": 9.392265193370166e-07,
+      "loss": 0.7851,
+      "step": 17
+    },
+    {
+      "epoch": 0.009953917050691244,
+      "grad_norm": 3.7548842430114746,
+      "learning_rate": 9.94475138121547e-07,
+      "loss": 0.7836,
+      "step": 18
+    },
+    {
+      "epoch": 0.010506912442396314,
+      "grad_norm": 2.2732059955596924,
+      "learning_rate": 1.0497237569060774e-06,
+      "loss": 0.7572,
+      "step": 19
+    },
+    {
+      "epoch": 0.011059907834101382,
+      "grad_norm": 2.244140625,
+      "learning_rate": 1.1049723756906078e-06,
+      "loss": 0.7637,
+      "step": 20
+    },
+    {
+      "epoch": 0.011612903225806452,
+      "grad_norm": 2.1460845470428467,
+      "learning_rate": 1.160220994475138e-06,
+      "loss": 0.7849,
+      "step": 21
+    },
+    {
+      "epoch": 0.01216589861751152,
+      "grad_norm": 2.0333800315856934,
+      "learning_rate": 1.2154696132596686e-06,
+      "loss": 0.8008,
+      "step": 22
+    },
+    {
+      "epoch": 0.01271889400921659,
+      "grad_norm": 1.962172508239746,
+      "learning_rate": 1.270718232044199e-06,
+      "loss": 0.7914,
+      "step": 23
+    },
+    {
+      "epoch": 0.013271889400921659,
+      "grad_norm": 1.9014993906021118,
+      "learning_rate": 1.3259668508287293e-06,
+      "loss": 0.754,
+      "step": 24
+    },
+    {
+      "epoch": 0.013824884792626729,
+      "grad_norm": 1.5539250373840332,
+      "learning_rate": 1.3812154696132598e-06,
+      "loss": 0.7218,
+      "step": 25
+    },
+    {
+      "epoch": 0.014377880184331797,
+      "grad_norm": 2.1412103176116943,
+      "learning_rate": 1.43646408839779e-06,
+      "loss": 0.7083,
+      "step": 26
+    },
+    {
+      "epoch": 0.014930875576036867,
+      "grad_norm": 2.793086290359497,
+      "learning_rate": 1.4917127071823205e-06,
+      "loss": 0.7505,
+      "step": 27
+    },
+    {
+      "epoch": 0.015483870967741935,
+      "grad_norm": 2.779872179031372,
+      "learning_rate": 1.5469613259668508e-06,
+      "loss": 0.7609,
+      "step": 28
+    },
+    {
+      "epoch": 0.016036866359447004,
+      "grad_norm": 2.9033734798431396,
+      "learning_rate": 1.6022099447513815e-06,
+      "loss": 0.7388,
+      "step": 29
+    },
+    {
+      "epoch": 0.016589861751152075,
+      "grad_norm": 2.6718666553497314,
+      "learning_rate": 1.6574585635359118e-06,
+      "loss": 0.7395,
+      "step": 30
+    },
+    {
+      "epoch": 0.017142857142857144,
+      "grad_norm": 2.3672306537628174,
+      "learning_rate": 1.7127071823204422e-06,
+      "loss": 0.736,
+      "step": 31
+    },
+    {
+      "epoch": 0.017695852534562212,
+      "grad_norm": 2.0016157627105713,
+      "learning_rate": 1.7679558011049725e-06,
+      "loss": 0.7169,
+      "step": 32
+    },
+    {
+      "epoch": 0.01824884792626728,
+      "grad_norm": 1.755001425743103,
+      "learning_rate": 1.823204419889503e-06,
+      "loss": 0.7358,
+      "step": 33
+    },
+    {
+      "epoch": 0.018801843317972352,
+      "grad_norm": 1.084032416343689,
+      "learning_rate": 1.8784530386740332e-06,
+      "loss": 0.654,
+      "step": 34
+    },
+    {
+      "epoch": 0.01935483870967742,
+      "grad_norm": 1.0225807428359985,
+      "learning_rate": 1.933701657458564e-06,
+      "loss": 0.6793,
+      "step": 35
+    },
+    {
+      "epoch": 0.01990783410138249,
+      "grad_norm": 1.0972223281860352,
+      "learning_rate": 1.988950276243094e-06,
+      "loss": 0.6625,
+      "step": 36
+    },
+    {
+      "epoch": 0.020460829493087557,
+      "grad_norm": 1.1796579360961914,
+      "learning_rate": 2.0441988950276245e-06,
+      "loss": 0.6905,
+      "step": 37
+    },
+    {
+      "epoch": 0.02101382488479263,
+      "grad_norm": 1.1285828351974487,
+      "learning_rate": 2.0994475138121547e-06,
+      "loss": 0.6433,
+      "step": 38
+    },
+    {
+      "epoch": 0.021566820276497697,
+      "grad_norm": 1.0303235054016113,
+      "learning_rate": 2.1546961325966854e-06,
+      "loss": 0.637,
+      "step": 39
+    },
+    {
+      "epoch": 0.022119815668202765,
+      "grad_norm": 0.9121658205986023,
+      "learning_rate": 2.2099447513812157e-06,
+      "loss": 0.6506,
+      "step": 40
+    },
+    {
+      "epoch": 0.022672811059907833,
+      "grad_norm": 0.8387944102287292,
+      "learning_rate": 2.265193370165746e-06,
+      "loss": 0.6217,
+      "step": 41
+    },
+    {
+      "epoch": 0.023225806451612905,
+      "grad_norm": 0.762744665145874,
+      "learning_rate": 2.320441988950276e-06,
+      "loss": 0.6131,
+      "step": 42
+    },
+    {
+      "epoch": 0.023778801843317973,
+      "grad_norm": 0.8466107845306396,
+      "learning_rate": 2.375690607734807e-06,
+      "loss": 0.6313,
+      "step": 43
+    },
+    {
+      "epoch": 0.02433179723502304,
+      "grad_norm": 0.7411535382270813,
+      "learning_rate": 2.430939226519337e-06,
+      "loss": 0.6178,
+      "step": 44
+    },
+    {
+      "epoch": 0.02488479262672811,
+      "grad_norm": 0.7712140083312988,
+      "learning_rate": 2.486187845303868e-06,
+      "loss": 0.6218,
+      "step": 45
+    },
+    {
+      "epoch": 0.02543778801843318,
+      "grad_norm": 1.0630971193313599,
+      "learning_rate": 2.541436464088398e-06,
+      "loss": 0.6497,
+      "step": 46
+    },
+    {
+      "epoch": 0.02599078341013825,
+      "grad_norm": 0.8186689615249634,
+      "learning_rate": 2.5966850828729284e-06,
+      "loss": 0.6493,
+      "step": 47
+    },
+    {
+      "epoch": 0.026543778801843318,
+      "grad_norm": 0.7290191650390625,
+      "learning_rate": 2.6519337016574586e-06,
+      "loss": 0.6278,
+      "step": 48
+    },
+    {
+      "epoch": 0.027096774193548386,
+      "grad_norm": 0.6419755816459656,
+      "learning_rate": 2.707182320441989e-06,
+      "loss": 0.6495,
+      "step": 49
+    },
+    {
+      "epoch": 0.027649769585253458,
+      "grad_norm": 0.6034302711486816,
+      "learning_rate": 2.7624309392265196e-06,
+      "loss": 0.6262,
+      "step": 50
+    },
+    {
+      "epoch": 0.028202764976958526,
+      "grad_norm": 0.7188774943351746,
+      "learning_rate": 2.81767955801105e-06,
+      "loss": 0.6416,
+      "step": 51
+    },
+    {
+      "epoch": 0.028755760368663594,
+      "grad_norm": 0.7371018528938293,
+      "learning_rate": 2.87292817679558e-06,
+      "loss": 0.6181,
+      "step": 52
+    },
+    {
+      "epoch": 0.029308755760368663,
+      "grad_norm": 0.7526310086250305,
+      "learning_rate": 2.9281767955801104e-06,
+      "loss": 0.6644,
+      "step": 53
+    },
+    {
+      "epoch": 0.029861751152073734,
+      "grad_norm": 0.6002934575080872,
+      "learning_rate": 2.983425414364641e-06,
+      "loss": 0.57,
+      "step": 54
+    },
+    {
+      "epoch": 0.030414746543778803,
+      "grad_norm": 0.535647451877594,
+      "learning_rate": 3.0386740331491713e-06,
+      "loss": 0.6158,
+      "step": 55
+    },
+    {
+      "epoch": 0.03096774193548387,
+      "grad_norm": 0.5333572030067444,
+      "learning_rate": 3.0939226519337016e-06,
+      "loss": 0.6237,
+      "step": 56
+    },
+    {
+      "epoch": 0.03152073732718894,
+      "grad_norm": 0.6033058166503906,
+      "learning_rate": 3.149171270718232e-06,
+      "loss": 0.5905,
+      "step": 57
+    },
+    {
+      "epoch": 0.03207373271889401,
+      "grad_norm": 0.5824403166770935,
+      "learning_rate": 3.204419889502763e-06,
+      "loss": 0.5934,
+      "step": 58
+    },
+    {
+      "epoch": 0.03262672811059908,
+      "grad_norm": 0.5389731526374817,
+      "learning_rate": 3.2596685082872933e-06,
+      "loss": 0.6149,
+      "step": 59
+    },
+    {
+      "epoch": 0.03317972350230415,
+      "grad_norm": 0.5317662954330444,
+      "learning_rate": 3.3149171270718235e-06,
+      "loss": 0.5953,
+      "step": 60
+    },
+    {
+      "epoch": 0.033732718894009216,
+      "grad_norm": 0.5464670658111572,
+      "learning_rate": 3.370165745856354e-06,
+      "loss": 0.5848,
+      "step": 61
+    },
+    {
+      "epoch": 0.03428571428571429,
+      "grad_norm": 0.5087230205535889,
+      "learning_rate": 3.4254143646408845e-06,
+      "loss": 0.5912,
+      "step": 62
+    },
+    {
+      "epoch": 0.03483870967741935,
+      "grad_norm": 0.4756256937980652,
+      "learning_rate": 3.4806629834254147e-06,
+      "loss": 0.5842,
+      "step": 63
+    },
+    {
+      "epoch": 0.035391705069124424,
+      "grad_norm": 0.49870121479034424,
+      "learning_rate": 3.535911602209945e-06,
+      "loss": 0.6009,
+      "step": 64
+    },
+    {
+      "epoch": 0.035944700460829496,
+      "grad_norm": 0.44439196586608887,
+      "learning_rate": 3.5911602209944757e-06,
+      "loss": 0.5832,
+      "step": 65
+    },
+    {
+      "epoch": 0.03649769585253456,
+      "grad_norm": 0.4260464608669281,
+      "learning_rate": 3.646408839779006e-06,
+      "loss": 0.5896,
+      "step": 66
+    },
+    {
+      "epoch": 0.03705069124423963,
+      "grad_norm": 0.4569395184516907,
+      "learning_rate": 3.7016574585635362e-06,
+      "loss": 0.6184,
+      "step": 67
+    },
+    {
+      "epoch": 0.037603686635944704,
+      "grad_norm": 0.4692537486553192,
+      "learning_rate": 3.7569060773480665e-06,
+      "loss": 0.5623,
+      "step": 68
+    },
+    {
+      "epoch": 0.03815668202764977,
+      "grad_norm": 0.40080568194389343,
+      "learning_rate": 3.812154696132597e-06,
+      "loss": 0.5832,
+      "step": 69
+    },
+    {
+      "epoch": 0.03870967741935484,
+      "grad_norm": 0.47561773657798767,
+      "learning_rate": 3.867403314917128e-06,
+      "loss": 0.5801,
+      "step": 70
+    },
+    {
+      "epoch": 0.039262672811059905,
+      "grad_norm": 0.47063136100769043,
+      "learning_rate": 3.922651933701658e-06,
+      "loss": 0.5913,
+      "step": 71
+    },
+    {
+      "epoch": 0.03981566820276498,
+      "grad_norm": 0.435703843832016,
+      "learning_rate": 3.977900552486188e-06,
+      "loss": 0.583,
+      "step": 72
+    },
+    {
+      "epoch": 0.04036866359447005,
+      "grad_norm": 0.44961071014404297,
+      "learning_rate": 4.033149171270719e-06,
+      "loss": 0.6015,
+      "step": 73
+    },
+    {
+      "epoch": 0.04092165898617511,
+      "grad_norm": 0.43612730503082275,
+      "learning_rate": 4.088397790055249e-06,
+      "loss": 0.5916,
+      "step": 74
+    },
+    {
+      "epoch": 0.041474654377880185,
+      "grad_norm": 0.4045799970626831,
+      "learning_rate": 4.143646408839779e-06,
+      "loss": 0.5365,
+      "step": 75
+    },
+    {
+      "epoch": 0.04202764976958526,
+      "grad_norm": 0.47303467988967896,
+      "learning_rate": 4.1988950276243095e-06,
+      "loss": 0.587,
+      "step": 76
+    },
+    {
+      "epoch": 0.04258064516129032,
+      "grad_norm": 0.3942665159702301,
+      "learning_rate": 4.2541436464088406e-06,
+      "loss": 0.5682,
+      "step": 77
+    },
+    {
+      "epoch": 0.04313364055299539,
+      "grad_norm": 0.40599432587623596,
+      "learning_rate": 4.309392265193371e-06,
+      "loss": 0.5686,
+      "step": 78
+    },
+    {
+      "epoch": 0.04368663594470046,
+      "grad_norm": 0.41502925753593445,
+      "learning_rate": 4.364640883977901e-06,
+      "loss": 0.5644,
+      "step": 79
+    },
+    {
+      "epoch": 0.04423963133640553,
+      "grad_norm": 0.4402730166912079,
+      "learning_rate": 4.419889502762431e-06,
+      "loss": 0.5458,
+      "step": 80
+    },
+    {
+      "epoch": 0.0447926267281106,
+      "grad_norm": 0.4110409915447235,
+      "learning_rate": 4.475138121546962e-06,
+      "loss": 0.5447,
+      "step": 81
+    },
+    {
+      "epoch": 0.045345622119815666,
+      "grad_norm": 0.46120530366897583,
+      "learning_rate": 4.530386740331492e-06,
+      "loss": 0.584,
+      "step": 82
+    },
+    {
+      "epoch": 0.04589861751152074,
+      "grad_norm": 0.4089435040950775,
+      "learning_rate": 4.585635359116022e-06,
+      "loss": 0.555,
+      "step": 83
+    },
+    {
+      "epoch": 0.04645161290322581,
+      "grad_norm": 0.4719107449054718,
+      "learning_rate": 4.640883977900552e-06,
+      "loss": 0.5714,
+      "step": 84
+    },
+    {
+      "epoch": 0.047004608294930875,
+      "grad_norm": 0.45129814743995667,
+      "learning_rate": 4.6961325966850835e-06,
+      "loss": 0.5443,
+      "step": 85
+    },
+    {
+      "epoch": 0.047557603686635946,
+      "grad_norm": 0.4321492612361908,
+      "learning_rate": 4.751381215469614e-06,
+      "loss": 0.5624,
+      "step": 86
+    },
+    {
+      "epoch": 0.04811059907834101,
+      "grad_norm": 0.4300953149795532,
+      "learning_rate": 4.806629834254144e-06,
+      "loss": 0.5497,
+      "step": 87
+    },
+    {
+      "epoch": 0.04866359447004608,
+      "grad_norm": 0.4590497314929962,
+      "learning_rate": 4.861878453038674e-06,
+      "loss": 0.5652,
+      "step": 88
+    },
+    {
+      "epoch": 0.049216589861751155,
+      "grad_norm": 0.47058796882629395,
+      "learning_rate": 4.9171270718232054e-06,
+      "loss": 0.5768,
+      "step": 89
+    },
+    {
+      "epoch": 0.04976958525345622,
+      "grad_norm": 0.43738317489624023,
+      "learning_rate": 4.972375690607736e-06,
+      "loss": 0.58,
+      "step": 90
+    },
+    {
+      "epoch": 0.05032258064516129,
+      "grad_norm": 0.4055286943912506,
+      "learning_rate": 5.027624309392266e-06,
+      "loss": 0.5503,
+      "step": 91
+    },
+    {
+      "epoch": 0.05087557603686636,
+      "grad_norm": 0.44679558277130127,
+      "learning_rate": 5.082872928176796e-06,
+      "loss": 0.5347,
+      "step": 92
+    },
+    {
+      "epoch": 0.05142857142857143,
+      "grad_norm": 0.46003517508506775,
+      "learning_rate": 5.1381215469613265e-06,
+      "loss": 0.5854,
+      "step": 93
+    },
+    {
+      "epoch": 0.0519815668202765,
+      "grad_norm": 0.44815686345100403,
+      "learning_rate": 5.193370165745857e-06,
+      "loss": 0.5316,
+      "step": 94
+    },
+    {
+      "epoch": 0.052534562211981564,
+      "grad_norm": 0.4559434950351715,
+      "learning_rate": 5.248618784530387e-06,
+      "loss": 0.5724,
+      "step": 95
+    },
+    {
+      "epoch": 0.053087557603686636,
+      "grad_norm": 0.45906394720077515,
+      "learning_rate": 5.303867403314917e-06,
+      "loss": 0.565,
+      "step": 96
+    },
+    {
+      "epoch": 0.05364055299539171,
+      "grad_norm": 0.3892553448677063,
+      "learning_rate": 5.3591160220994476e-06,
+      "loss": 0.5397,
+      "step": 97
+    },
+    {
+      "epoch": 0.05419354838709677,
+      "grad_norm": 0.4995177090167999,
+      "learning_rate": 5.414364640883978e-06,
+      "loss": 0.5509,
+      "step": 98
+    },
+    {
+      "epoch": 0.054746543778801844,
+      "grad_norm": 0.4675018787384033,
+      "learning_rate": 5.469613259668509e-06,
+      "loss": 0.5432,
+      "step": 99
+    },
+    {
+      "epoch": 0.055299539170506916,
+      "grad_norm": 0.4121154844760895,
+      "learning_rate": 5.524861878453039e-06,
+      "loss": 0.5463,
+      "step": 100
+    },
+    {
+      "epoch": 0.05585253456221198,
+      "grad_norm": 0.4763623774051666,
+      "learning_rate": 5.5801104972375695e-06,
+      "loss": 0.5673,
+      "step": 101
+    },
+    {
+      "epoch": 0.05640552995391705,
+      "grad_norm": 0.44080275297164917,
+      "learning_rate": 5.6353591160221e-06,
+      "loss": 0.5684,
+      "step": 102
+    },
+    {
+      "epoch": 0.05695852534562212,
+      "grad_norm": 0.4623638391494751,
+      "learning_rate": 5.69060773480663e-06,
+      "loss": 0.5738,
+      "step": 103
+    },
+    {
+      "epoch": 0.05751152073732719,
+      "grad_norm": 0.379423588514328,
+      "learning_rate": 5.74585635359116e-06,
+      "loss": 0.5503,
+      "step": 104
+    },
+    {
+      "epoch": 0.05806451612903226,
+      "grad_norm": 0.45095521211624146,
+      "learning_rate": 5.8011049723756905e-06,
+      "loss": 0.5683,
+      "step": 105
+    },
+    {
+      "epoch": 0.058617511520737325,
+      "grad_norm": 0.4157678484916687,
+      "learning_rate": 5.856353591160221e-06,
+      "loss": 0.5459,
+      "step": 106
+    },
+    {
+      "epoch": 0.0591705069124424,
+      "grad_norm": 0.4484108090400696,
+      "learning_rate": 5.911602209944752e-06,
+      "loss": 0.5465,
+      "step": 107
+    },
+    {
+      "epoch": 0.05972350230414747,
+      "grad_norm": 0.41227927803993225,
+      "learning_rate": 5.966850828729282e-06,
+      "loss": 0.5737,
+      "step": 108
+    },
+    {
+      "epoch": 0.060276497695852534,
+      "grad_norm": 0.4771617352962494,
+      "learning_rate": 6.0220994475138124e-06,
+      "loss": 0.5295,
+      "step": 109
+    },
+    {
+      "epoch": 0.060829493087557605,
+      "grad_norm": 0.42870959639549255,
+      "learning_rate": 6.077348066298343e-06,
+      "loss": 0.5432,
+      "step": 110
+    },
+    {
+      "epoch": 0.06138248847926267,
+      "grad_norm": 0.47716712951660156,
+      "learning_rate": 6.132596685082873e-06,
+      "loss": 0.5695,
+      "step": 111
+    },
+    {
+      "epoch": 0.06193548387096774,
+      "grad_norm": 0.43211421370506287,
+      "learning_rate": 6.187845303867403e-06,
+      "loss": 0.5408,
+      "step": 112
+    },
+    {
+      "epoch": 0.062488479262672814,
+      "grad_norm": 0.49489325284957886,
+      "learning_rate": 6.2430939226519335e-06,
+      "loss": 0.5488,
+      "step": 113
+    },
+    {
+      "epoch": 0.06304147465437789,
+      "grad_norm": 0.4600902199745178,
+      "learning_rate": 6.298342541436464e-06,
+      "loss": 0.5445,
+      "step": 114
+    },
+    {
+      "epoch": 0.06359447004608294,
+      "grad_norm": 0.41128382086753845,
+      "learning_rate": 6.353591160220996e-06,
+      "loss": 0.5507,
+      "step": 115
+    },
+    {
+      "epoch": 0.06414746543778801,
+      "grad_norm": 0.38756921887397766,
+      "learning_rate": 6.408839779005526e-06,
+      "loss": 0.5072,
+      "step": 116
+    },
+    {
+      "epoch": 0.06470046082949309,
+      "grad_norm": 0.4521966278553009,
+      "learning_rate": 6.464088397790056e-06,
+      "loss": 0.5597,
+      "step": 117
+    },
+    {
+      "epoch": 0.06525345622119816,
+      "grad_norm": 0.39012429118156433,
+      "learning_rate": 6.5193370165745865e-06,
+      "loss": 0.5017,
+      "step": 118
+    },
+    {
+      "epoch": 0.06580645161290323,
+      "grad_norm": 0.4543527364730835,
+      "learning_rate": 6.574585635359117e-06,
+      "loss": 0.5342,
+      "step": 119
+    },
+    {
+      "epoch": 0.0663594470046083,
+      "grad_norm": 0.4338400065898895,
+      "learning_rate": 6.629834254143647e-06,
+      "loss": 0.5489,
+      "step": 120
+    },
+    {
+      "epoch": 0.06691244239631336,
+      "grad_norm": 0.461480975151062,
+      "learning_rate": 6.685082872928177e-06,
+      "loss": 0.5327,
+      "step": 121
+    },
+    {
+      "epoch": 0.06746543778801843,
+      "grad_norm": 0.4241288900375366,
+      "learning_rate": 6.740331491712708e-06,
+      "loss": 0.5297,
+      "step": 122
+    },
+    {
+      "epoch": 0.0680184331797235,
+      "grad_norm": 0.5540463924407959,
+      "learning_rate": 6.795580110497239e-06,
+      "loss": 0.5262,
+      "step": 123
+    },
+    {
+      "epoch": 0.06857142857142857,
+      "grad_norm": 0.4794096350669861,
+      "learning_rate": 6.850828729281769e-06,
+      "loss": 0.5638,
+      "step": 124
+    },
+    {
+      "epoch": 0.06912442396313365,
+      "grad_norm": 0.43071404099464417,
+      "learning_rate": 6.906077348066299e-06,
+      "loss": 0.5254,
+      "step": 125
+    },
+    {
+      "epoch": 0.0696774193548387,
+      "grad_norm": 0.4853145480155945,
+      "learning_rate": 6.9613259668508295e-06,
+      "loss": 0.5524,
+      "step": 126
+    },
+    {
+      "epoch": 0.07023041474654378,
+      "grad_norm": 0.47362276911735535,
+      "learning_rate": 7.01657458563536e-06,
+      "loss": 0.5469,
+      "step": 127
+    },
+    {
+      "epoch": 0.07078341013824885,
+      "grad_norm": 0.45325443148612976,
+      "learning_rate": 7.07182320441989e-06,
+      "loss": 0.5476,
+      "step": 128
+    },
+    {
+      "epoch": 0.07133640552995392,
+      "grad_norm": 0.43244412541389465,
+      "learning_rate": 7.12707182320442e-06,
+      "loss": 0.5367,
+      "step": 129
+    },
+    {
+      "epoch": 0.07188940092165899,
+      "grad_norm": 0.5289068222045898,
+      "learning_rate": 7.182320441988951e-06,
+      "loss": 0.5796,
+      "step": 130
+    },
+    {
+      "epoch": 0.07244239631336405,
+      "grad_norm": 0.4485274851322174,
+      "learning_rate": 7.237569060773482e-06,
+      "loss": 0.5396,
+      "step": 131
+    },
+    {
+      "epoch": 0.07299539170506912,
+      "grad_norm": 0.42793214321136475,
+      "learning_rate": 7.292817679558012e-06,
+      "loss": 0.5532,
+      "step": 132
+    },
+    {
+      "epoch": 0.07354838709677419,
+      "grad_norm": 0.4250226318836212,
+      "learning_rate": 7.348066298342542e-06,
+      "loss": 0.5847,
+      "step": 133
+    },
+    {
+      "epoch": 0.07410138248847926,
+      "grad_norm": 0.44791412353515625,
+      "learning_rate": 7.4033149171270724e-06,
+      "loss": 0.5216,
+      "step": 134
+    },
+    {
+      "epoch": 0.07465437788018434,
+      "grad_norm": 0.48473551869392395,
+      "learning_rate": 7.458563535911603e-06,
+      "loss": 0.5434,
+      "step": 135
+    },
+    {
+      "epoch": 0.07520737327188941,
+      "grad_norm": 0.5042223334312439,
+      "learning_rate": 7.513812154696133e-06,
+      "loss": 0.5292,
+      "step": 136
+    },
+    {
+      "epoch": 0.07576036866359447,
+      "grad_norm": 0.5112206935882568,
+      "learning_rate": 7.569060773480663e-06,
+      "loss": 0.573,
+      "step": 137
+    },
+    {
+      "epoch": 0.07631336405529954,
+      "grad_norm": 0.4625641107559204,
+      "learning_rate": 7.624309392265194e-06,
+      "loss": 0.5656,
+      "step": 138
+    },
+    {
+      "epoch": 0.07686635944700461,
+      "grad_norm": 0.507746160030365,
+      "learning_rate": 7.679558011049725e-06,
+      "loss": 0.5352,
+      "step": 139
+    },
+    {
+      "epoch": 0.07741935483870968,
+      "grad_norm": 0.4666261672973633,
+      "learning_rate": 7.734806629834256e-06,
+      "loss": 0.5465,
+      "step": 140
+    },
+    {
+      "epoch": 0.07797235023041475,
+      "grad_norm": 0.44455286860466003,
+      "learning_rate": 7.790055248618785e-06,
+      "loss": 0.5395,
+      "step": 141
+    },
+    {
+      "epoch": 0.07852534562211981,
+      "grad_norm": 0.4286342263221741,
+      "learning_rate": 7.845303867403316e-06,
+      "loss": 0.5045,
+      "step": 142
+    },
+    {
+      "epoch": 0.07907834101382488,
+      "grad_norm": 0.4748035669326782,
+      "learning_rate": 7.900552486187846e-06,
+      "loss": 0.552,
+      "step": 143
+    },
+    {
+      "epoch": 0.07963133640552995,
+      "grad_norm": 0.44316938519477844,
+      "learning_rate": 7.955801104972377e-06,
+      "loss": 0.5394,
+      "step": 144
+    },
+    {
+      "epoch": 0.08018433179723503,
+      "grad_norm": 0.4846290647983551,
+      "learning_rate": 8.011049723756906e-06,
+      "loss": 0.5267,
+      "step": 145
+    },
+    {
+      "epoch": 0.0807373271889401,
+      "grad_norm": 0.4428083300590515,
+      "learning_rate": 8.066298342541437e-06,
+      "loss": 0.4941,
+      "step": 146
+    },
+    {
+      "epoch": 0.08129032258064516,
+      "grad_norm": 0.44677111506462097,
+      "learning_rate": 8.121546961325968e-06,
+      "loss": 0.4885,
+      "step": 147
+    },
+    {
+      "epoch": 0.08184331797235023,
+      "grad_norm": 0.4421268105506897,
+      "learning_rate": 8.176795580110498e-06,
+      "loss": 0.5765,
+      "step": 148
+    },
+    {
+      "epoch": 0.0823963133640553,
+      "grad_norm": 0.4853808581829071,
+      "learning_rate": 8.232044198895029e-06,
+      "loss": 0.5297,
+      "step": 149
+    },
+    {
+      "epoch": 0.08294930875576037,
+      "grad_norm": 0.47924932837486267,
+      "learning_rate": 8.287292817679558e-06,
+      "loss": 0.5192,
+      "step": 150
+    },
+    {
+      "epoch": 0.08350230414746544,
+      "grad_norm": 0.5630611181259155,
+      "learning_rate": 8.34254143646409e-06,
+      "loss": 0.5439,
+      "step": 151
+    },
+    {
+      "epoch": 0.08405529953917051,
+      "grad_norm": 0.6081127524375916,
+      "learning_rate": 8.397790055248619e-06,
+      "loss": 0.5012,
+      "step": 152
+    },
+    {
+      "epoch": 0.08460829493087557,
+      "grad_norm": 0.44527262449264526,
+      "learning_rate": 8.45303867403315e-06,
+      "loss": 0.5305,
+      "step": 153
+    },
+    {
+      "epoch": 0.08516129032258064,
+      "grad_norm": 0.4948616623878479,
+      "learning_rate": 8.508287292817681e-06,
+      "loss": 0.5109,
+      "step": 154
+    },
+    {
+      "epoch": 0.08571428571428572,
+      "grad_norm": 0.5024743676185608,
+      "learning_rate": 8.56353591160221e-06,
+      "loss": 0.5632,
+      "step": 155
+    },
+    {
+      "epoch": 0.08626728110599079,
+      "grad_norm": 0.45424172282218933,
+      "learning_rate": 8.618784530386742e-06,
+      "loss": 0.5029,
+      "step": 156
+    },
+    {
+      "epoch": 0.08682027649769586,
+      "grad_norm": 0.5445595979690552,
+      "learning_rate": 8.674033149171271e-06,
+      "loss": 0.5451,
+      "step": 157
+    },
+    {
+      "epoch": 0.08737327188940092,
+      "grad_norm": 0.495613157749176,
+      "learning_rate": 8.729281767955802e-06,
+      "loss": 0.527,
+      "step": 158
+    },
+    {
+      "epoch": 0.08792626728110599,
+      "grad_norm": 0.6915062665939331,
+      "learning_rate": 8.784530386740332e-06,
+      "loss": 0.5378,
+      "step": 159
+    },
+    {
+      "epoch": 0.08847926267281106,
+      "grad_norm": 0.4711109697818756,
+      "learning_rate": 8.839779005524863e-06,
+      "loss": 0.5105,
+      "step": 160
+    },
+    {
+      "epoch": 0.08903225806451613,
+      "grad_norm": 0.5774613618850708,
+      "learning_rate": 8.895027624309392e-06,
+      "loss": 0.5268,
+      "step": 161
+    },
+    {
+      "epoch": 0.0895852534562212,
+      "grad_norm": 0.534370481967926,
+      "learning_rate": 8.950276243093923e-06,
+      "loss": 0.5275,
+      "step": 162
+    },
+    {
+      "epoch": 0.09013824884792626,
+      "grad_norm": 0.5167669057846069,
+      "learning_rate": 9.005524861878454e-06,
+      "loss": 0.5173,
+      "step": 163
+    },
+    {
+      "epoch": 0.09069124423963133,
+      "grad_norm": 0.5220374464988708,
+      "learning_rate": 9.060773480662984e-06,
+      "loss": 0.5117,
+      "step": 164
+    },
+    {
+      "epoch": 0.0912442396313364,
+      "grad_norm": 0.6151928305625916,
+      "learning_rate": 9.116022099447515e-06,
+      "loss": 0.5236,
+      "step": 165
+    },
+    {
+      "epoch": 0.09179723502304148,
+      "grad_norm": 0.54198157787323,
+      "learning_rate": 9.171270718232044e-06,
+      "loss": 0.5113,
+      "step": 166
+    },
+    {
+      "epoch": 0.09235023041474655,
+      "grad_norm": 0.540729820728302,
+      "learning_rate": 9.226519337016575e-06,
+      "loss": 0.5293,
+      "step": 167
+    },
+    {
+      "epoch": 0.09290322580645162,
+      "grad_norm": 0.4339202642440796,
+      "learning_rate": 9.281767955801105e-06,
+      "loss": 0.5105,
+      "step": 168
+    },
+    {
+      "epoch": 0.09345622119815668,
+      "grad_norm": 0.5427307486534119,
+      "learning_rate": 9.337016574585636e-06,
+      "loss": 0.5442,
+      "step": 169
+    },
+    {
+      "epoch": 0.09400921658986175,
+      "grad_norm": 0.5696178674697876,
+      "learning_rate": 9.392265193370167e-06,
+      "loss": 0.5198,
+      "step": 170
+    },
+    {
+      "epoch": 0.09456221198156682,
+      "grad_norm": 0.475676029920578,
+      "learning_rate": 9.447513812154696e-06,
+      "loss": 0.5275,
+      "step": 171
+    },
+    {
+      "epoch": 0.09511520737327189,
+      "grad_norm": 0.6084780693054199,
+      "learning_rate": 9.502762430939228e-06,
+      "loss": 0.5344,
+      "step": 172
+    },
+    {
+      "epoch": 0.09566820276497696,
+      "grad_norm": 0.45912471413612366,
+      "learning_rate": 9.558011049723757e-06,
+      "loss": 0.5128,
+      "step": 173
+    },
+    {
+      "epoch": 0.09622119815668202,
+      "grad_norm": 0.5522315502166748,
+      "learning_rate": 9.613259668508288e-06,
+      "loss": 0.4985,
+      "step": 174
+    },
+    {
+      "epoch": 0.0967741935483871,
+      "grad_norm": 0.5598754286766052,
+      "learning_rate": 9.668508287292818e-06,
+      "loss": 0.5438,
+      "step": 175
+    },
+    {
+      "epoch": 0.09732718894009217,
+      "grad_norm": 0.4842788577079773,
+      "learning_rate": 9.723756906077349e-06,
+      "loss": 0.5098,
+      "step": 176
+    },
+    {
+      "epoch": 0.09788018433179724,
+      "grad_norm": 0.5201513171195984,
+      "learning_rate": 9.779005524861878e-06,
+      "loss": 0.5155,
+      "step": 177
+    },
+    {
+      "epoch": 0.09843317972350231,
+      "grad_norm": 0.4839669466018677,
+      "learning_rate": 9.834254143646411e-06,
+      "loss": 0.5201,
+      "step": 178
+    },
+    {
+      "epoch": 0.09898617511520737,
+      "grad_norm": 0.5859594941139221,
+      "learning_rate": 9.88950276243094e-06,
+      "loss": 0.5603,
+      "step": 179
+    },
+    {
+      "epoch": 0.09953917050691244,
+      "grad_norm": 0.48163536190986633,
+      "learning_rate": 9.944751381215471e-06,
+      "loss": 0.5279,
+      "step": 180
+    },
+    {
+      "epoch": 0.10009216589861751,
+      "grad_norm": 0.508065938949585,
+      "learning_rate": 1e-05,
+      "loss": 0.5186,
+      "step": 181
+    },
+    {
+      "epoch": 0.10064516129032258,
+      "grad_norm": 0.47829869389533997,
+      "learning_rate": 9.999990678956964e-06,
+      "loss": 0.5224,
+      "step": 182
+    },
+    {
+      "epoch": 0.10119815668202765,
+      "grad_norm": 0.5300391912460327,
+      "learning_rate": 9.999962715862601e-06,
+      "loss": 0.5128,
+      "step": 183
+    },
+    {
+      "epoch": 0.10175115207373273,
+      "grad_norm": 0.47900837659835815,
+      "learning_rate": 9.999916110821174e-06,
+      "loss": 0.5232,
+      "step": 184
+    },
+    {
+      "epoch": 0.10230414746543778,
+      "grad_norm": 0.4351784288883209,
+      "learning_rate": 9.999850864006444e-06,
+      "loss": 0.5274,
+      "step": 185
+    },
+    {
+      "epoch": 0.10285714285714286,
+      "grad_norm": 0.48379045724868774,
+      "learning_rate": 9.99976697566168e-06,
+      "loss": 0.4964,
+      "step": 186
+    },
+    {
+      "epoch": 0.10341013824884793,
+      "grad_norm": 0.4367005228996277,
+      "learning_rate": 9.999664446099651e-06,
+      "loss": 0.5311,
+      "step": 187
+    },
+    {
+      "epoch": 0.103963133640553,
+      "grad_norm": 0.4944308400154114,
+      "learning_rate": 9.999543275702632e-06,
+      "loss": 0.5093,
+      "step": 188
+    },
+    {
+      "epoch": 0.10451612903225807,
+      "grad_norm": 0.44800275564193726,
+      "learning_rate": 9.999403464922393e-06,
+      "loss": 0.5124,
+      "step": 189
+    },
+    {
+      "epoch": 0.10506912442396313,
+      "grad_norm": 0.5029193758964539,
+      "learning_rate": 9.99924501428021e-06,
+      "loss": 0.5538,
+      "step": 190
+    },
+    {
+      "epoch": 0.1056221198156682,
+      "grad_norm": 0.4435194730758667,
+      "learning_rate": 9.999067924366854e-06,
+      "loss": 0.5291,
+      "step": 191
+    },
+    {
+      "epoch": 0.10617511520737327,
+      "grad_norm": 0.45577532052993774,
+      "learning_rate": 9.998872195842588e-06,
+      "loss": 0.5107,
+      "step": 192
+    },
+    {
+      "epoch": 0.10672811059907834,
+      "grad_norm": 0.4373863637447357,
+      "learning_rate": 9.998657829437171e-06,
+      "loss": 0.5137,
+      "step": 193
+    },
+    {
+      "epoch": 0.10728110599078342,
+      "grad_norm": 0.4408453404903412,
+      "learning_rate": 9.998424825949848e-06,
+      "loss": 0.5061,
+      "step": 194
+    },
+    {
+      "epoch": 0.10783410138248847,
+      "grad_norm": 0.5477350950241089,
+      "learning_rate": 9.998173186249357e-06,
+      "loss": 0.5363,
+      "step": 195
+    },
+    {
+      "epoch": 0.10838709677419354,
+      "grad_norm": 0.43939393758773804,
+      "learning_rate": 9.99790291127391e-06,
+      "loss": 0.5095,
+      "step": 196
+    },
+    {
+      "epoch": 0.10894009216589862,
+      "grad_norm": 0.47420209646224976,
+      "learning_rate": 9.997614002031211e-06,
+      "loss": 0.5297,
+      "step": 197
+    },
+    {
+      "epoch": 0.10949308755760369,
+      "grad_norm": 0.45546746253967285,
+      "learning_rate": 9.99730645959843e-06,
+      "loss": 0.5139,
+      "step": 198
+    },
+    {
+      "epoch": 0.11004608294930876,
+      "grad_norm": 0.4920537769794464,
+      "learning_rate": 9.996980285122218e-06,
+      "loss": 0.5344,
+      "step": 199
+    },
+    {
+      "epoch": 0.11059907834101383,
+      "grad_norm": 0.562940776348114,
+      "learning_rate": 9.996635479818683e-06,
+      "loss": 0.535,
+      "step": 200
+    },
+    {
+      "epoch": 0.11115207373271889,
+      "grad_norm": 0.5914273262023926,
+      "learning_rate": 9.99627204497341e-06,
+      "loss": 0.5134,
+      "step": 201
+    },
+    {
+      "epoch": 0.11170506912442396,
+      "grad_norm": 0.4358191192150116,
+      "learning_rate": 9.995889981941432e-06,
+      "loss": 0.4886,
+      "step": 202
+    },
+    {
+      "epoch": 0.11225806451612903,
+      "grad_norm": 0.569436252117157,
+      "learning_rate": 9.995489292147238e-06,
+      "loss": 0.5302,
+      "step": 203
+    },
+    {
+      "epoch": 0.1128110599078341,
+      "grad_norm": 0.5561317801475525,
+      "learning_rate": 9.995069977084769e-06,
+      "loss": 0.5214,
+      "step": 204
+    },
+    {
+      "epoch": 0.11336405529953918,
+      "grad_norm": 0.4526347219944,
+      "learning_rate": 9.994632038317407e-06,
+      "loss": 0.5106,
+      "step": 205
+    },
+    {
+      "epoch": 0.11391705069124423,
+      "grad_norm": 0.6103693246841431,
+      "learning_rate": 9.99417547747797e-06,
+      "loss": 0.5187,
+      "step": 206
+    },
+    {
+      "epoch": 0.1144700460829493,
+      "grad_norm": 0.5255654454231262,
+      "learning_rate": 9.993700296268705e-06,
+      "loss": 0.5253,
+      "step": 207
+    },
+    {
+      "epoch": 0.11502304147465438,
+      "grad_norm": 0.48559436202049255,
+      "learning_rate": 9.993206496461287e-06,
+      "loss": 0.5127,
+      "step": 208
+    },
+    {
+      "epoch": 0.11557603686635945,
+      "grad_norm": 0.5230923891067505,
+      "learning_rate": 9.992694079896812e-06,
+      "loss": 0.5078,
+      "step": 209
+    },
+    {
+      "epoch": 0.11612903225806452,
+      "grad_norm": 0.45681560039520264,
+      "learning_rate": 9.992163048485776e-06,
+      "loss": 0.522,
+      "step": 210
+    },
+    {
+      "epoch": 0.11668202764976958,
+      "grad_norm": 0.510287344455719,
+      "learning_rate": 9.99161340420809e-06,
+      "loss": 0.509,
+      "step": 211
+    },
+    {
+      "epoch": 0.11723502304147465,
+      "grad_norm": 0.4897903501987457,
+      "learning_rate": 9.991045149113055e-06,
+      "loss": 0.5118,
+      "step": 212
+    },
+    {
+      "epoch": 0.11778801843317972,
+      "grad_norm": 0.516163170337677,
+      "learning_rate": 9.990458285319362e-06,
+      "loss": 0.4915,
+      "step": 213
+    },
+    {
+      "epoch": 0.1183410138248848,
+      "grad_norm": 0.5389583110809326,
+      "learning_rate": 9.98985281501509e-06,
+      "loss": 0.5192,
+      "step": 214
+    },
+    {
+      "epoch": 0.11889400921658987,
+      "grad_norm": 0.6268092393875122,
+      "learning_rate": 9.989228740457679e-06,
+      "loss": 0.524,
+      "step": 215
+    },
+    {
+      "epoch": 0.11944700460829494,
+      "grad_norm": 0.447966605424881,
+      "learning_rate": 9.988586063973942e-06,
+      "loss": 0.4864,
+      "step": 216
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6052084565162659,
+      "learning_rate": 9.987924787960043e-06,
+      "loss": 0.5005,
+      "step": 217
+    },
+    {
+      "epoch": 0.12055299539170507,
+      "grad_norm": 0.5076256990432739,
+      "learning_rate": 9.987244914881498e-06,
+      "loss": 0.5219,
+      "step": 218
+    },
+    {
+      "epoch": 0.12110599078341014,
+      "grad_norm": 0.5539697408676147,
+      "learning_rate": 9.986546447273153e-06,
+      "loss": 0.5024,
+      "step": 219
+    },
+    {
+      "epoch": 0.12165898617511521,
+      "grad_norm": 0.4377414584159851,
+      "learning_rate": 9.985829387739192e-06,
+      "loss": 0.5291,
+      "step": 220
+    },
+    {
+      "epoch": 0.12221198156682028,
+      "grad_norm": 0.5455598831176758,
+      "learning_rate": 9.985093738953108e-06,
+      "loss": 0.5126,
+      "step": 221
+    },
+    {
+      "epoch": 0.12276497695852534,
+      "grad_norm": 0.462424099445343,
+      "learning_rate": 9.98433950365771e-06,
+      "loss": 0.5225,
+      "step": 222
+    },
+    {
+      "epoch": 0.12331797235023041,
+      "grad_norm": 0.49225398898124695,
+      "learning_rate": 9.983566684665097e-06,
+      "loss": 0.5328,
+      "step": 223
+    },
+    {
+      "epoch": 0.12387096774193548,
+      "grad_norm": 0.5304319262504578,
+      "learning_rate": 9.982775284856665e-06,
+      "loss": 0.5212,
+      "step": 224
+    },
+    {
+      "epoch": 0.12442396313364056,
+      "grad_norm": 0.4655149579048157,
+      "learning_rate": 9.981965307183081e-06,
+      "loss": 0.5161,
+      "step": 225
+    },
+    {
+      "epoch": 0.12497695852534563,
+      "grad_norm": 0.6474348902702332,
+      "learning_rate": 9.98113675466428e-06,
+      "loss": 0.4983,
+      "step": 226
+    },
+    {
+      "epoch": 0.12552995391705069,
+      "grad_norm": 0.43538525700569153,
+      "learning_rate": 9.980289630389453e-06,
+      "loss": 0.5286,
+      "step": 227
+    },
+    {
+      "epoch": 0.12608294930875577,
+      "grad_norm": 0.5731073617935181,
+      "learning_rate": 9.97942393751703e-06,
+      "loss": 0.5414,
+      "step": 228
+    },
+    {
+      "epoch": 0.12663594470046083,
+      "grad_norm": 0.4664866626262665,
+      "learning_rate": 9.978539679274675e-06,
+      "loss": 0.5052,
+      "step": 229
+    },
+    {
+      "epoch": 0.1271889400921659,
+      "grad_norm": 0.5314398407936096,
+      "learning_rate": 9.977636858959274e-06,
+      "loss": 0.527,
+      "step": 230
+    },
+    {
+      "epoch": 0.12774193548387097,
+      "grad_norm": 0.538694441318512,
+      "learning_rate": 9.976715479936916e-06,
+      "loss": 0.5094,
+      "step": 231
+    },
+    {
+      "epoch": 0.12829493087557603,
+      "grad_norm": 0.4854016900062561,
+      "learning_rate": 9.975775545642889e-06,
+      "loss": 0.5078,
+      "step": 232
+    },
+    {
+      "epoch": 0.12884792626728112,
+      "grad_norm": 0.5078600645065308,
+      "learning_rate": 9.974817059581656e-06,
+      "loss": 0.5236,
+      "step": 233
+    },
+    {
+      "epoch": 0.12940092165898617,
+      "grad_norm": 0.455322802066803,
+      "learning_rate": 9.973840025326858e-06,
+      "loss": 0.5137,
+      "step": 234
+    },
+    {
+      "epoch": 0.12995391705069123,
+      "grad_norm": 0.5216564536094666,
+      "learning_rate": 9.972844446521281e-06,
+      "loss": 0.5289,
+      "step": 235
+    },
+    {
+      "epoch": 0.13050691244239632,
+      "grad_norm": 0.4569193422794342,
+      "learning_rate": 9.971830326876864e-06,
+      "loss": 0.5137,
+      "step": 236
+    },
+    {
+      "epoch": 0.13105990783410137,
+      "grad_norm": 0.5937883853912354,
+      "learning_rate": 9.970797670174663e-06,
+      "loss": 0.5142,
+      "step": 237
+    },
+    {
+      "epoch": 0.13161290322580646,
+      "grad_norm": 0.48625534772872925,
+      "learning_rate": 9.969746480264855e-06,
+      "loss": 0.5279,
+      "step": 238
+    },
+    {
+      "epoch": 0.13216589861751152,
+      "grad_norm": 0.47141098976135254,
+      "learning_rate": 9.968676761066714e-06,
+      "loss": 0.5042,
+      "step": 239
+    },
+    {
+      "epoch": 0.1327188940092166,
+      "grad_norm": 0.4889548122882843,
+      "learning_rate": 9.967588516568601e-06,
+      "loss": 0.5154,
+      "step": 240
+    },
+    {
+      "epoch": 0.13327188940092166,
+      "grad_norm": 0.4120868444442749,
+      "learning_rate": 9.966481750827943e-06,
+      "loss": 0.504,
+      "step": 241
+    },
+    {
+      "epoch": 0.13382488479262672,
+      "grad_norm": 0.5358414053916931,
+      "learning_rate": 9.965356467971228e-06,
+      "loss": 0.5264,
+      "step": 242
+    },
+    {
+      "epoch": 0.1343778801843318,
+      "grad_norm": 0.472572922706604,
+      "learning_rate": 9.964212672193978e-06,
+      "loss": 0.5176,
+      "step": 243
+    },
+    {
+      "epoch": 0.13493087557603686,
+      "grad_norm": 0.4870736598968506,
+      "learning_rate": 9.96305036776074e-06,
+      "loss": 0.5071,
+      "step": 244
+    },
+    {
+      "epoch": 0.13548387096774195,
+      "grad_norm": 0.5704755783081055,
+      "learning_rate": 9.96186955900507e-06,
+      "loss": 0.5198,
+      "step": 245
+    },
+    {
+      "epoch": 0.136036866359447,
+      "grad_norm": 0.5287049412727356,
+      "learning_rate": 9.960670250329517e-06,
+      "loss": 0.5216,
+      "step": 246
+    },
+    {
+      "epoch": 0.13658986175115206,
+      "grad_norm": 0.5611394643783569,
+      "learning_rate": 9.959452446205603e-06,
+      "loss": 0.5188,
+      "step": 247
+    },
+    {
+      "epoch": 0.13714285714285715,
+      "grad_norm": 0.4512503445148468,
+      "learning_rate": 9.958216151173812e-06,
+      "loss": 0.4979,
+      "step": 248
+    },
+    {
+      "epoch": 0.1376958525345622,
+      "grad_norm": 0.47202053666114807,
+      "learning_rate": 9.956961369843565e-06,
+      "loss": 0.4834,
+      "step": 249
+    },
+    {
+      "epoch": 0.1382488479262673,
+      "grad_norm": 0.5006586909294128,
+      "learning_rate": 9.95568810689321e-06,
+      "loss": 0.5285,
+      "step": 250
+    },
+    {
+      "epoch": 0.13880184331797235,
+      "grad_norm": 0.4660753011703491,
+      "learning_rate": 9.954396367070006e-06,
+      "loss": 0.5192,
+      "step": 251
+    },
+    {
+      "epoch": 0.1393548387096774,
+      "grad_norm": 0.4885999262332916,
+      "learning_rate": 9.953086155190095e-06,
+      "loss": 0.5028,
+      "step": 252
+    },
+    {
+      "epoch": 0.1399078341013825,
+      "grad_norm": 0.49867549538612366,
+      "learning_rate": 9.951757476138495e-06,
+      "loss": 0.5039,
+      "step": 253
+    },
+    {
+      "epoch": 0.14046082949308755,
+      "grad_norm": 0.5823776125907898,
+      "learning_rate": 9.950410334869075e-06,
+      "loss": 0.535,
+      "step": 254
+    },
+    {
+      "epoch": 0.14101382488479264,
+      "grad_norm": 0.47015827894210815,
+      "learning_rate": 9.949044736404538e-06,
+      "loss": 0.528,
+      "step": 255
+    },
+    {
+      "epoch": 0.1415668202764977,
+      "grad_norm": 0.5392245054244995,
+      "learning_rate": 9.94766068583641e-06,
+      "loss": 0.4855,
+      "step": 256
+    },
+    {
+      "epoch": 0.14211981566820275,
+      "grad_norm": 0.5323054194450378,
+      "learning_rate": 9.946258188325003e-06,
+      "loss": 0.5269,
+      "step": 257
+    },
+    {
+      "epoch": 0.14267281105990784,
+      "grad_norm": 0.5459584593772888,
+      "learning_rate": 9.944837249099418e-06,
+      "loss": 0.4874,
+      "step": 258
+    },
+    {
+      "epoch": 0.1432258064516129,
+      "grad_norm": 0.4532824158668518,
+      "learning_rate": 9.943397873457503e-06,
+      "loss": 0.5111,
+      "step": 259
+    },
+    {
+      "epoch": 0.14377880184331798,
+      "grad_norm": 0.46974876523017883,
+      "learning_rate": 9.94194006676586e-06,
+      "loss": 0.5261,
+      "step": 260
+    },
+    {
+      "epoch": 0.14433179723502304,
+      "grad_norm": 0.5061238408088684,
+      "learning_rate": 9.94046383445979e-06,
+      "loss": 0.5217,
+      "step": 261
+    },
+    {
+      "epoch": 0.1448847926267281,
+      "grad_norm": 0.4556806981563568,
+      "learning_rate": 9.938969182043312e-06,
+      "loss": 0.4969,
+      "step": 262
+    },
+    {
+      "epoch": 0.14543778801843318,
+      "grad_norm": 0.48571473360061646,
+      "learning_rate": 9.937456115089108e-06,
+      "loss": 0.5057,
+      "step": 263
+    },
+    {
+      "epoch": 0.14599078341013824,
+      "grad_norm": 0.497122198343277,
+      "learning_rate": 9.935924639238526e-06,
+      "loss": 0.5205,
+      "step": 264
+    },
+    {
+      "epoch": 0.14654377880184333,
+      "grad_norm": 0.5786699652671814,
+      "learning_rate": 9.934374760201546e-06,
+      "loss": 0.5122,
+      "step": 265
+    },
+    {
+      "epoch": 0.14709677419354839,
+      "grad_norm": 0.5201807022094727,
+      "learning_rate": 9.932806483756763e-06,
+      "loss": 0.5328,
+      "step": 266
+    },
+    {
+      "epoch": 0.14764976958525344,
+      "grad_norm": 0.5433037877082825,
+      "learning_rate": 9.931219815751368e-06,
+      "loss": 0.5153,
+      "step": 267
+    },
+    {
+      "epoch": 0.14820276497695853,
+      "grad_norm": 0.47332048416137695,
+      "learning_rate": 9.929614762101117e-06,
+      "loss": 0.5004,
+      "step": 268
+    },
+    {
+      "epoch": 0.1487557603686636,
+      "grad_norm": 0.5838091373443604,
+      "learning_rate": 9.927991328790324e-06,
+      "loss": 0.4889,
+      "step": 269
+    },
+    {
+      "epoch": 0.14930875576036867,
+      "grad_norm": 0.4618660807609558,
+      "learning_rate": 9.926349521871824e-06,
+      "loss": 0.5265,
+      "step": 270
+    },
+    {
+      "epoch": 0.14986175115207373,
+      "grad_norm": 0.5086191892623901,
+      "learning_rate": 9.924689347466959e-06,
+      "loss": 0.5106,
+      "step": 271
+    },
+    {
+      "epoch": 0.15041474654377882,
+      "grad_norm": 0.457042932510376,
+      "learning_rate": 9.92301081176555e-06,
+      "loss": 0.518,
+      "step": 272
+    },
+    {
+      "epoch": 0.15096774193548387,
+      "grad_norm": 0.5256728529930115,
+      "learning_rate": 9.92131392102588e-06,
+      "loss": 0.531,
+      "step": 273
+    },
+    {
+      "epoch": 0.15152073732718893,
+      "grad_norm": 0.4683758318424225,
+      "learning_rate": 9.919598681574665e-06,
+      "loss": 0.4896,
+      "step": 274
+    },
+    {
+      "epoch": 0.15207373271889402,
+      "grad_norm": 0.4681558609008789,
+      "learning_rate": 9.917865099807034e-06,
+      "loss": 0.5089,
+      "step": 275
+    },
+    {
+      "epoch": 0.15262672811059907,
+      "grad_norm": 0.4770297706127167,
+      "learning_rate": 9.916113182186503e-06,
+      "loss": 0.4973,
+      "step": 276
+    },
+    {
+      "epoch": 0.15317972350230416,
+      "grad_norm": 0.4939088225364685,
+      "learning_rate": 9.91434293524495e-06,
+      "loss": 0.5388,
+      "step": 277
+    },
+    {
+      "epoch": 0.15373271889400922,
+      "grad_norm": 0.5165495276451111,
+      "learning_rate": 9.912554365582596e-06,
+      "loss": 0.5046,
+      "step": 278
+    },
+    {
+      "epoch": 0.15428571428571428,
+      "grad_norm": 0.4685964584350586,
+      "learning_rate": 9.910747479867975e-06,
+      "loss": 0.5373,
+      "step": 279
+    },
+    {
+      "epoch": 0.15483870967741936,
+      "grad_norm": 0.5078836679458618,
+      "learning_rate": 9.908922284837911e-06,
+      "loss": 0.5105,
+      "step": 280
+    },
+    {
+      "epoch": 0.15539170506912442,
+      "grad_norm": 0.4230501651763916,
+      "learning_rate": 9.90707878729749e-06,
+      "loss": 0.4749,
+      "step": 281
+    },
+    {
+      "epoch": 0.1559447004608295,
+      "grad_norm": 0.483981728553772,
+      "learning_rate": 9.905216994120044e-06,
+      "loss": 0.4954,
+      "step": 282
+    },
+    {
+      "epoch": 0.15649769585253456,
+      "grad_norm": 0.4270246922969818,
+      "learning_rate": 9.90333691224711e-06,
+      "loss": 0.5007,
+      "step": 283
+    },
+    {
+      "epoch": 0.15705069124423962,
+      "grad_norm": 0.4967236816883087,
+      "learning_rate": 9.901438548688423e-06,
+      "loss": 0.5308,
+      "step": 284
+    },
+    {
+      "epoch": 0.1576036866359447,
+      "grad_norm": 0.4965408444404602,
+      "learning_rate": 9.89952191052187e-06,
+      "loss": 0.5093,
+      "step": 285
+    },
+    {
+      "epoch": 0.15815668202764976,
+      "grad_norm": 0.45975926518440247,
+      "learning_rate": 9.89758700489348e-06,
+      "loss": 0.5122,
+      "step": 286
+    },
+    {
+      "epoch": 0.15870967741935485,
+      "grad_norm": 0.5180716514587402,
+      "learning_rate": 9.895633839017387e-06,
+      "loss": 0.5515,
+      "step": 287
+    },
+    {
+      "epoch": 0.1592626728110599,
+      "grad_norm": 0.43234091997146606,
+      "learning_rate": 9.893662420175809e-06,
+      "loss": 0.4901,
+      "step": 288
+    },
+    {
+      "epoch": 0.15981566820276497,
+      "grad_norm": 0.4894983768463135,
+      "learning_rate": 9.89167275571902e-06,
+      "loss": 0.5052,
+      "step": 289
+    },
+    {
+      "epoch": 0.16036866359447005,
+      "grad_norm": 0.4710214138031006,
+      "learning_rate": 9.889664853065315e-06,
+      "loss": 0.5,
+      "step": 290
+    },
+    {
+      "epoch": 0.1609216589861751,
+      "grad_norm": 0.5677446126937866,
+      "learning_rate": 9.887638719700996e-06,
+      "loss": 0.5169,
+      "step": 291
+    },
+    {
+      "epoch": 0.1614746543778802,
+      "grad_norm": 0.45073893666267395,
+      "learning_rate": 9.88559436318033e-06,
+      "loss": 0.5258,
+      "step": 292
+    },
+    {
+      "epoch": 0.16202764976958525,
+      "grad_norm": 0.5394282341003418,
+      "learning_rate": 9.883531791125538e-06,
+      "loss": 0.4799,
+      "step": 293
+    },
+    {
+      "epoch": 0.1625806451612903,
+      "grad_norm": 0.4784766435623169,
+      "learning_rate": 9.881451011226742e-06,
+      "loss": 0.4693,
+      "step": 294
+    },
+    {
+      "epoch": 0.1631336405529954,
+      "grad_norm": 0.435619980096817,
+      "learning_rate": 9.87935203124196e-06,
+      "loss": 0.5196,
+      "step": 295
+    },
+    {
+      "epoch": 0.16368663594470045,
+      "grad_norm": 0.5438254475593567,
+      "learning_rate": 9.877234858997066e-06,
+      "loss": 0.5078,
+      "step": 296
+    },
+    {
+      "epoch": 0.16423963133640554,
+      "grad_norm": 0.4646199643611908,
+      "learning_rate": 9.875099502385761e-06,
+      "loss": 0.5239,
+      "step": 297
+    },
+    {
+      "epoch": 0.1647926267281106,
+      "grad_norm": 0.549284040927887,
+      "learning_rate": 9.872945969369546e-06,
+      "loss": 0.5029,
+      "step": 298
+    },
+    {
+      "epoch": 0.16534562211981566,
+      "grad_norm": 0.5235300064086914,
+      "learning_rate": 9.87077426797769e-06,
+      "loss": 0.5225,
+      "step": 299
+    },
+    {
+      "epoch": 0.16589861751152074,
+      "grad_norm": 0.525598406791687,
+      "learning_rate": 9.8685844063072e-06,
+      "loss": 0.5366,
+      "step": 300
+    },
+    {
+      "epoch": 0.1664516129032258,
+      "grad_norm": 0.5001644492149353,
+      "learning_rate": 9.866376392522798e-06,
+      "loss": 0.5025,
+      "step": 301
+    },
+    {
+      "epoch": 0.16700460829493088,
+      "grad_norm": 0.4980228841304779,
+      "learning_rate": 9.864150234856876e-06,
+      "loss": 0.5332,
+      "step": 302
+    },
+    {
+      "epoch": 0.16755760368663594,
+      "grad_norm": 0.44216716289520264,
+      "learning_rate": 9.861905941609482e-06,
+      "loss": 0.4831,
+      "step": 303
+    },
+    {
+      "epoch": 0.16811059907834103,
+      "grad_norm": 0.45178863406181335,
+      "learning_rate": 9.859643521148275e-06,
+      "loss": 0.5164,
+      "step": 304
+    },
+    {
+      "epoch": 0.16866359447004609,
+      "grad_norm": 0.4833186864852905,
+      "learning_rate": 9.857362981908505e-06,
+      "loss": 0.5128,
+      "step": 305
+    },
+    {
+      "epoch": 0.16921658986175114,
+      "grad_norm": 0.5167268514633179,
+      "learning_rate": 9.855064332392972e-06,
+      "loss": 0.5149,
+      "step": 306
+    },
+    {
+      "epoch": 0.16976958525345623,
+      "grad_norm": 0.4465697109699249,
+      "learning_rate": 9.852747581172002e-06,
+      "loss": 0.5016,
+      "step": 307
+    },
+    {
+      "epoch": 0.1703225806451613,
+      "grad_norm": 0.5105251669883728,
+      "learning_rate": 9.850412736883408e-06,
+      "loss": 0.5021,
+      "step": 308
+    },
+    {
+      "epoch": 0.17087557603686637,
+      "grad_norm": 0.5230706334114075,
+      "learning_rate": 9.848059808232464e-06,
+      "loss": 0.4857,
+      "step": 309
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 0.5510496497154236,
+      "learning_rate": 9.845688803991873e-06,
+      "loss": 0.5305,
+      "step": 310
+    },
+    {
+      "epoch": 0.1719815668202765,
+      "grad_norm": 0.4802294671535492,
+      "learning_rate": 9.843299733001723e-06,
+      "loss": 0.5084,
+      "step": 311
+    },
+    {
+      "epoch": 0.17253456221198157,
+      "grad_norm": 0.5729020833969116,
+      "learning_rate": 9.840892604169473e-06,
+      "loss": 0.4797,
+      "step": 312
+    },
+    {
+      "epoch": 0.17308755760368663,
+      "grad_norm": 0.6035499572753906,
+      "learning_rate": 9.838467426469897e-06,
+      "loss": 0.5069,
+      "step": 313
+    },
+    {
+      "epoch": 0.17364055299539172,
+      "grad_norm": 0.4519449770450592,
+      "learning_rate": 9.836024208945074e-06,
+      "loss": 0.4959,
+      "step": 314
+    },
+    {
+      "epoch": 0.17419354838709677,
+      "grad_norm": 0.5990521907806396,
+      "learning_rate": 9.833562960704336e-06,
+      "loss": 0.522,
+      "step": 315
+    },
+    {
+      "epoch": 0.17474654377880183,
+      "grad_norm": 0.4257313013076782,
+      "learning_rate": 9.831083690924246e-06,
+      "loss": 0.4769,
+      "step": 316
+    },
+    {
+      "epoch": 0.17529953917050692,
+      "grad_norm": 0.5456443428993225,
+      "learning_rate": 9.828586408848553e-06,
+      "loss": 0.5129,
+      "step": 317
+    },
+    {
+      "epoch": 0.17585253456221198,
+      "grad_norm": 0.4802759885787964,
+      "learning_rate": 9.82607112378817e-06,
+      "loss": 0.5043,
+      "step": 318
+    },
+    {
+      "epoch": 0.17640552995391706,
+      "grad_norm": 0.6043391823768616,
+      "learning_rate": 9.823537845121126e-06,
+      "loss": 0.5063,
+      "step": 319
+    },
+    {
+      "epoch": 0.17695852534562212,
+      "grad_norm": 0.4696696698665619,
+      "learning_rate": 9.82098658229254e-06,
+      "loss": 0.5164,
+      "step": 320
+    },
+    {
+      "epoch": 0.17751152073732718,
+      "grad_norm": 0.5530909299850464,
+      "learning_rate": 9.818417344814587e-06,
+      "loss": 0.5275,
+      "step": 321
+    },
+    {
+      "epoch": 0.17806451612903226,
+      "grad_norm": 0.609788179397583,
+      "learning_rate": 9.815830142266457e-06,
+      "loss": 0.4716,
+      "step": 322
+    },
+    {
+      "epoch": 0.17861751152073732,
+      "grad_norm": 0.46327316761016846,
+      "learning_rate": 9.813224984294318e-06,
+      "loss": 0.4887,
+      "step": 323
+    },
+    {
+      "epoch": 0.1791705069124424,
+      "grad_norm": 0.5943720936775208,
+      "learning_rate": 9.810601880611286e-06,
+      "loss": 0.5009,
+      "step": 324
+    },
+    {
+      "epoch": 0.17972350230414746,
+      "grad_norm": 0.4262332320213318,
+      "learning_rate": 9.807960840997387e-06,
+      "loss": 0.4747,
+      "step": 325
+    },
+    {
+      "epoch": 0.18027649769585252,
+      "grad_norm": 0.47295740246772766,
+      "learning_rate": 9.805301875299518e-06,
+      "loss": 0.5092,
+      "step": 326
+    },
+    {
+      "epoch": 0.1808294930875576,
+      "grad_norm": 0.4910508394241333,
+      "learning_rate": 9.802624993431414e-06,
+      "loss": 0.5195,
+      "step": 327
+    },
+    {
+      "epoch": 0.18138248847926267,
+      "grad_norm": 0.5312582850456238,
+      "learning_rate": 9.799930205373605e-06,
+      "loss": 0.5077,
+      "step": 328
+    },
+    {
+      "epoch": 0.18193548387096775,
+      "grad_norm": 0.49236226081848145,
+      "learning_rate": 9.797217521173385e-06,
+      "loss": 0.5228,
+      "step": 329
+    },
+    {
+      "epoch": 0.1824884792626728,
+      "grad_norm": 0.5442107319831848,
+      "learning_rate": 9.794486950944775e-06,
+      "loss": 0.4909,
+      "step": 330
+    },
+    {
+      "epoch": 0.18304147465437787,
+      "grad_norm": 0.4408930242061615,
+      "learning_rate": 9.79173850486848e-06,
+      "loss": 0.4919,
+      "step": 331
+    },
+    {
+      "epoch": 0.18359447004608295,
+      "grad_norm": 0.4888366460800171,
+      "learning_rate": 9.78897219319185e-06,
+      "loss": 0.5247,
+      "step": 332
+    },
+    {
+      "epoch": 0.184147465437788,
+      "grad_norm": 0.48040100932121277,
+      "learning_rate": 9.786188026228853e-06,
+      "loss": 0.4746,
+      "step": 333
+    },
+    {
+      "epoch": 0.1847004608294931,
+      "grad_norm": 0.46840646862983704,
+      "learning_rate": 9.783386014360024e-06,
+      "loss": 0.5174,
+      "step": 334
+    },
+    {
+      "epoch": 0.18525345622119815,
+      "grad_norm": 0.4824281930923462,
+      "learning_rate": 9.780566168032432e-06,
+      "loss": 0.4944,
+      "step": 335
+    },
+    {
+      "epoch": 0.18580645161290324,
+      "grad_norm": 0.4401717483997345,
+      "learning_rate": 9.77772849775964e-06,
+      "loss": 0.5074,
+      "step": 336
+    },
+    {
+      "epoch": 0.1863594470046083,
+      "grad_norm": 0.5038358569145203,
+      "learning_rate": 9.774873014121667e-06,
+      "loss": 0.5052,
+      "step": 337
+    },
+    {
+      "epoch": 0.18691244239631336,
+      "grad_norm": 0.47577494382858276,
+      "learning_rate": 9.771999727764946e-06,
+      "loss": 0.4856,
+      "step": 338
+    },
+    {
+      "epoch": 0.18746543778801844,
+      "grad_norm": 0.4989863634109497,
+      "learning_rate": 9.76910864940229e-06,
+      "loss": 0.4839,
+      "step": 339
+    },
+    {
+      "epoch": 0.1880184331797235,
+      "grad_norm": 0.49821236729621887,
+      "learning_rate": 9.766199789812845e-06,
+      "loss": 0.498,
+      "step": 340
+    },
+    {
+      "epoch": 0.18857142857142858,
+      "grad_norm": 0.44591957330703735,
+      "learning_rate": 9.763273159842052e-06,
+      "loss": 0.4832,
+      "step": 341
+    },
+    {
+      "epoch": 0.18912442396313364,
+      "grad_norm": 0.4908744990825653,
+      "learning_rate": 9.76032877040161e-06,
+      "loss": 0.4927,
+      "step": 342
+    },
+    {
+      "epoch": 0.1896774193548387,
+      "grad_norm": 0.55243319272995,
+      "learning_rate": 9.757366632469427e-06,
+      "loss": 0.5005,
+      "step": 343
+    },
+    {
+      "epoch": 0.19023041474654379,
+      "grad_norm": 0.4234575629234314,
+      "learning_rate": 9.754386757089596e-06,
+      "loss": 0.4799,
+      "step": 344
+    },
+    {
+      "epoch": 0.19078341013824884,
+      "grad_norm": 0.49401965737342834,
+      "learning_rate": 9.751389155372329e-06,
+      "loss": 0.4942,
+      "step": 345
+    },
+    {
+      "epoch": 0.19133640552995393,
+      "grad_norm": 0.4622778594493866,
+      "learning_rate": 9.74837383849394e-06,
+      "loss": 0.4644,
+      "step": 346
+    },
+    {
+      "epoch": 0.191889400921659,
+      "grad_norm": 0.4600273072719574,
+      "learning_rate": 9.745340817696787e-06,
+      "loss": 0.5019,
+      "step": 347
+    },
+    {
+      "epoch": 0.19244239631336404,
+      "grad_norm": 0.4513567388057709,
+      "learning_rate": 9.742290104289237e-06,
+      "loss": 0.4704,
+      "step": 348
+    },
+    {
+      "epoch": 0.19299539170506913,
+      "grad_norm": 0.5180230736732483,
+      "learning_rate": 9.739221709645622e-06,
+      "loss": 0.4917,
+      "step": 349
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 0.44200998544692993,
+      "learning_rate": 9.736135645206198e-06,
+      "loss": 0.5047,
+      "step": 350
+    },
+    {
+      "epoch": 0.19410138248847927,
+      "grad_norm": 0.5660358667373657,
+      "learning_rate": 9.7330319224771e-06,
+      "loss": 0.5023,
+      "step": 351
+    },
+    {
+      "epoch": 0.19465437788018433,
+      "grad_norm": 0.44904786348342896,
+      "learning_rate": 9.729910553030304e-06,
+      "loss": 0.4998,
+      "step": 352
+    },
+    {
+      "epoch": 0.1952073732718894,
+      "grad_norm": 0.523826003074646,
+      "learning_rate": 9.726771548503575e-06,
+      "loss": 0.4966,
+      "step": 353
+    },
+    {
+      "epoch": 0.19576036866359448,
+      "grad_norm": 0.4515267312526703,
+      "learning_rate": 9.72361492060043e-06,
+      "loss": 0.4996,
+      "step": 354
+    },
+    {
+      "epoch": 0.19631336405529953,
+      "grad_norm": 0.4984499216079712,
+      "learning_rate": 9.7204406810901e-06,
+      "loss": 0.4985,
+      "step": 355
+    },
+    {
+      "epoch": 0.19686635944700462,
+      "grad_norm": 0.4689009189605713,
+      "learning_rate": 9.71724884180747e-06,
+      "loss": 0.4936,
+      "step": 356
+    },
+    {
+      "epoch": 0.19741935483870968,
+      "grad_norm": 0.46954217553138733,
+      "learning_rate": 9.714039414653047e-06,
+      "loss": 0.491,
+      "step": 357
+    },
+    {
+      "epoch": 0.19797235023041473,
+      "grad_norm": 0.5114353895187378,
+      "learning_rate": 9.71081241159292e-06,
+      "loss": 0.488,
+      "step": 358
+    },
+    {
+      "epoch": 0.19852534562211982,
+      "grad_norm": 0.4611765742301941,
+      "learning_rate": 9.707567844658698e-06,
+      "loss": 0.4749,
+      "step": 359
+    },
+    {
+      "epoch": 0.19907834101382488,
+      "grad_norm": 0.5052332878112793,
+      "learning_rate": 9.704305725947483e-06,
+      "loss": 0.5234,
+      "step": 360
+    },
+    {
+      "epoch": 0.19963133640552996,
+      "grad_norm": 0.5130250453948975,
+      "learning_rate": 9.701026067621813e-06,
+      "loss": 0.5006,
+      "step": 361
+    },
+    {
+      "epoch": 0.20018433179723502,
+      "grad_norm": 0.47562843561172485,
+      "learning_rate": 9.697728881909622e-06,
+      "loss": 0.529,
+      "step": 362
+    },
+    {
+      "epoch": 0.20073732718894008,
+      "grad_norm": 0.48228660225868225,
+      "learning_rate": 9.694414181104197e-06,
+      "loss": 0.5057,
+      "step": 363
+    },
+    {
+      "epoch": 0.20129032258064516,
+      "grad_norm": 0.501700222492218,
+      "learning_rate": 9.691081977564124e-06,
+      "loss": 0.4949,
+      "step": 364
+    },
+    {
+      "epoch": 0.20184331797235022,
+      "grad_norm": 0.4578889012336731,
+      "learning_rate": 9.687732283713247e-06,
+      "loss": 0.4989,
+      "step": 365
+    },
+    {
+      "epoch": 0.2023963133640553,
+      "grad_norm": 0.5223492980003357,
+      "learning_rate": 9.684365112040622e-06,
+      "loss": 0.4772,
+      "step": 366
+    },
+    {
+      "epoch": 0.20294930875576037,
+      "grad_norm": 0.4984261393547058,
+      "learning_rate": 9.680980475100471e-06,
+      "loss": 0.5132,
+      "step": 367
+    },
+    {
+      "epoch": 0.20350230414746545,
+      "grad_norm": 0.4805150032043457,
+      "learning_rate": 9.677578385512135e-06,
+      "loss": 0.5069,
+      "step": 368
+    },
+    {
+      "epoch": 0.2040552995391705,
+      "grad_norm": 0.46725666522979736,
+      "learning_rate": 9.674158855960017e-06,
+      "loss": 0.4806,
+      "step": 369
+    },
+    {
+      "epoch": 0.20460829493087557,
+      "grad_norm": 0.47330862283706665,
+      "learning_rate": 9.670721899193556e-06,
+      "loss": 0.5127,
+      "step": 370
+    },
+    {
+      "epoch": 0.20516129032258065,
+      "grad_norm": 0.5010029077529907,
+      "learning_rate": 9.667267528027157e-06,
+      "loss": 0.5189,
+      "step": 371
+    },
+    {
+      "epoch": 0.2057142857142857,
+      "grad_norm": 0.5415194034576416,
+      "learning_rate": 9.663795755340159e-06,
+      "loss": 0.5053,
+      "step": 372
+    },
+    {
+      "epoch": 0.2062672811059908,
+      "grad_norm": 0.4000757932662964,
+      "learning_rate": 9.660306594076779e-06,
+      "loss": 0.4939,
+      "step": 373
+    },
+    {
+      "epoch": 0.20682027649769585,
+      "grad_norm": 0.5064148306846619,
+      "learning_rate": 9.656800057246065e-06,
+      "loss": 0.4937,
+      "step": 374
+    },
+    {
+      "epoch": 0.2073732718894009,
+      "grad_norm": 0.467369943857193,
+      "learning_rate": 9.65327615792185e-06,
+      "loss": 0.5215,
+      "step": 375
+    },
+    {
+      "epoch": 0.207926267281106,
+      "grad_norm": 0.4753814935684204,
+      "learning_rate": 9.649734909242699e-06,
+      "loss": 0.4903,
+      "step": 376
+    },
+    {
+      "epoch": 0.20847926267281106,
+      "grad_norm": 0.4203657805919647,
+      "learning_rate": 9.646176324411867e-06,
+      "loss": 0.4874,
+      "step": 377
+    },
+    {
+      "epoch": 0.20903225806451614,
+      "grad_norm": 0.4930543601512909,
+      "learning_rate": 9.642600416697242e-06,
+      "loss": 0.4816,
+      "step": 378
+    },
+    {
+      "epoch": 0.2095852534562212,
+      "grad_norm": 0.4189959764480591,
+      "learning_rate": 9.639007199431298e-06,
+      "loss": 0.491,
+      "step": 379
+    },
+    {
+      "epoch": 0.21013824884792626,
+      "grad_norm": 0.48571497201919556,
+      "learning_rate": 9.635396686011052e-06,
+      "loss": 0.5005,
+      "step": 380
+    },
+    {
+      "epoch": 0.21069124423963134,
+      "grad_norm": 0.4344756007194519,
+      "learning_rate": 9.631768889898004e-06,
+      "loss": 0.5113,
+      "step": 381
+    },
+    {
+      "epoch": 0.2112442396313364,
+      "grad_norm": 0.5239787101745605,
+      "learning_rate": 9.628123824618087e-06,
+      "loss": 0.5013,
+      "step": 382
+    },
+    {
+      "epoch": 0.21179723502304149,
+      "grad_norm": 0.40745410323143005,
+      "learning_rate": 9.624461503761628e-06,
+      "loss": 0.4824,
+      "step": 383
+    },
+    {
+      "epoch": 0.21235023041474654,
+      "grad_norm": 0.5618027448654175,
+      "learning_rate": 9.620781940983288e-06,
+      "loss": 0.4852,
+      "step": 384
+    },
+    {
+      "epoch": 0.2129032258064516,
+      "grad_norm": 0.5030412673950195,
+      "learning_rate": 9.617085150002012e-06,
+      "loss": 0.4868,
+      "step": 385
+    },
+    {
+      "epoch": 0.2134562211981567,
+      "grad_norm": 0.5133850574493408,
+      "learning_rate": 9.613371144600976e-06,
+      "loss": 0.5076,
+      "step": 386
+    },
+    {
+      "epoch": 0.21400921658986174,
+      "grad_norm": 0.47164881229400635,
+      "learning_rate": 9.609639938627546e-06,
+      "loss": 0.5076,
+      "step": 387
+    },
+    {
+      "epoch": 0.21456221198156683,
+      "grad_norm": 0.4972430467605591,
+      "learning_rate": 9.605891545993214e-06,
+      "loss": 0.4843,
+      "step": 388
+    },
+    {
+      "epoch": 0.2151152073732719,
+      "grad_norm": 0.4716038405895233,
+      "learning_rate": 9.60212598067355e-06,
+      "loss": 0.4782,
+      "step": 389
+    },
+    {
+      "epoch": 0.21566820276497695,
+      "grad_norm": 0.4822791814804077,
+      "learning_rate": 9.59834325670815e-06,
+      "loss": 0.4874,
+      "step": 390
+    },
+    {
+      "epoch": 0.21622119815668203,
+      "grad_norm": 0.4538308382034302,
+      "learning_rate": 9.594543388200592e-06,
+      "loss": 0.501,
+      "step": 391
+    },
+    {
+      "epoch": 0.2167741935483871,
+      "grad_norm": 0.4863637387752533,
+      "learning_rate": 9.59072638931837e-06,
+      "loss": 0.5115,
+      "step": 392
+    },
+    {
+      "epoch": 0.21732718894009218,
+      "grad_norm": 0.4747362434864044,
+      "learning_rate": 9.586892274292846e-06,
+      "loss": 0.5262,
+      "step": 393
+    },
+    {
+      "epoch": 0.21788018433179723,
+      "grad_norm": 0.4688088595867157,
+      "learning_rate": 9.583041057419203e-06,
+      "loss": 0.517,
+      "step": 394
+    },
+    {
+      "epoch": 0.2184331797235023,
+      "grad_norm": 0.49005988240242004,
+      "learning_rate": 9.579172753056383e-06,
+      "loss": 0.4737,
+      "step": 395
+    },
+    {
+      "epoch": 0.21898617511520738,
+      "grad_norm": 0.4679119288921356,
+      "learning_rate": 9.575287375627037e-06,
+      "loss": 0.4978,
+      "step": 396
+    },
+    {
+      "epoch": 0.21953917050691243,
+      "grad_norm": 0.47708967328071594,
+      "learning_rate": 9.571384939617478e-06,
+      "loss": 0.5005,
+      "step": 397
+    },
+    {
+      "epoch": 0.22009216589861752,
+      "grad_norm": 0.47183945775032043,
+      "learning_rate": 9.567465459577613e-06,
+      "loss": 0.4708,
+      "step": 398
+    },
+    {
+      "epoch": 0.22064516129032258,
+      "grad_norm": 0.47582703828811646,
+      "learning_rate": 9.563528950120895e-06,
+      "loss": 0.4983,
+      "step": 399
+    },
+    {
+      "epoch": 0.22119815668202766,
+      "grad_norm": 0.48260581493377686,
+      "learning_rate": 9.559575425924279e-06,
+      "loss": 0.4894,
+      "step": 400
+    },
+    {
+      "epoch": 0.22175115207373272,
+      "grad_norm": 0.5306166410446167,
+      "learning_rate": 9.55560490172815e-06,
+      "loss": 0.5068,
+      "step": 401
+    },
+    {
+      "epoch": 0.22230414746543778,
+      "grad_norm": 0.49763691425323486,
+      "learning_rate": 9.551617392336281e-06,
+      "loss": 0.5028,
+      "step": 402
+    },
+    {
+      "epoch": 0.22285714285714286,
+      "grad_norm": 0.5285375118255615,
+      "learning_rate": 9.547612912615769e-06,
+      "loss": 0.484,
+      "step": 403
+    },
+    {
+      "epoch": 0.22341013824884792,
+      "grad_norm": 0.5528775453567505,
+      "learning_rate": 9.543591477496985e-06,
+      "loss": 0.4774,
+      "step": 404
+    },
+    {
+      "epoch": 0.223963133640553,
+      "grad_norm": 0.48812106251716614,
+      "learning_rate": 9.539553101973516e-06,
+      "loss": 0.5042,
+      "step": 405
+    },
+    {
+      "epoch": 0.22451612903225807,
+      "grad_norm": 0.5272664427757263,
+      "learning_rate": 9.535497801102114e-06,
+      "loss": 0.4987,
+      "step": 406
+    },
+    {
+      "epoch": 0.22506912442396312,
+      "grad_norm": 0.5162638425827026,
+      "learning_rate": 9.531425590002629e-06,
+      "loss": 0.4942,
+      "step": 407
+    },
+    {
+      "epoch": 0.2256221198156682,
+      "grad_norm": 0.45863449573516846,
+      "learning_rate": 9.527336483857965e-06,
+      "loss": 0.5165,
+      "step": 408
+    },
+    {
+      "epoch": 0.22617511520737327,
+      "grad_norm": 0.5142653584480286,
+      "learning_rate": 9.523230497914016e-06,
+      "loss": 0.4883,
+      "step": 409
+    },
+    {
+      "epoch": 0.22672811059907835,
+      "grad_norm": 0.4717908203601837,
+      "learning_rate": 9.519107647479609e-06,
+      "loss": 0.4859,
+      "step": 410
+    },
+    {
+      "epoch": 0.2272811059907834,
+      "grad_norm": 0.4761084020137787,
+      "learning_rate": 9.514967947926453e-06,
+      "loss": 0.5022,
+      "step": 411
+    },
+    {
+      "epoch": 0.22783410138248847,
+      "grad_norm": 0.4796617925167084,
+      "learning_rate": 9.510811414689073e-06,
+      "loss": 0.5058,
+      "step": 412
+    },
+    {
+      "epoch": 0.22838709677419355,
+      "grad_norm": 0.5060638189315796,
+      "learning_rate": 9.506638063264759e-06,
+      "loss": 0.4859,
+      "step": 413
+    },
+    {
+      "epoch": 0.2289400921658986,
+      "grad_norm": 0.42853257060050964,
+      "learning_rate": 9.502447909213508e-06,
+      "loss": 0.5004,
+      "step": 414
+    },
+    {
+      "epoch": 0.2294930875576037,
+      "grad_norm": 0.45876848697662354,
+      "learning_rate": 9.498240968157962e-06,
+      "loss": 0.4898,
+      "step": 415
+    },
+    {
+      "epoch": 0.23004608294930876,
+      "grad_norm": 0.4455052614212036,
+      "learning_rate": 9.49401725578335e-06,
+      "loss": 0.4972,
+      "step": 416
+    },
+    {
+      "epoch": 0.2305990783410138,
+      "grad_norm": 0.41394197940826416,
+      "learning_rate": 9.489776787837438e-06,
+      "loss": 0.5006,
+      "step": 417
+    },
+    {
+      "epoch": 0.2311520737327189,
+      "grad_norm": 0.44113320112228394,
+      "learning_rate": 9.485519580130456e-06,
+      "loss": 0.4644,
+      "step": 418
+    },
+    {
+      "epoch": 0.23170506912442396,
+      "grad_norm": 0.4882364869117737,
+      "learning_rate": 9.481245648535053e-06,
+      "loss": 0.4726,
+      "step": 419
+    },
+    {
+      "epoch": 0.23225806451612904,
+      "grad_norm": 0.4658440053462982,
+      "learning_rate": 9.476955008986228e-06,
+      "loss": 0.497,
+      "step": 420
+    },
+    {
+      "epoch": 0.2328110599078341,
+      "grad_norm": 0.4944906532764435,
+      "learning_rate": 9.472647677481275e-06,
+      "loss": 0.4743,
+      "step": 421
+    },
+    {
+      "epoch": 0.23336405529953916,
+      "grad_norm": 0.4508844316005707,
+      "learning_rate": 9.468323670079725e-06,
+      "loss": 0.4884,
+      "step": 422
+    },
+    {
+      "epoch": 0.23391705069124424,
+      "grad_norm": 0.470225065946579,
+      "learning_rate": 9.463983002903278e-06,
+      "loss": 0.4929,
+      "step": 423
+    },
+    {
+      "epoch": 0.2344700460829493,
+      "grad_norm": 0.5286823511123657,
+      "learning_rate": 9.459625692135756e-06,
+      "loss": 0.4938,
+      "step": 424
+    },
+    {
+      "epoch": 0.2350230414746544,
+      "grad_norm": 0.45246848464012146,
+      "learning_rate": 9.45525175402303e-06,
+      "loss": 0.4834,
+      "step": 425
+    },
+    {
+      "epoch": 0.23557603686635945,
+      "grad_norm": 0.48045429587364197,
+      "learning_rate": 9.450861204872965e-06,
+      "loss": 0.4877,
+      "step": 426
+    },
+    {
+      "epoch": 0.2361290322580645,
+      "grad_norm": 0.5548929572105408,
+      "learning_rate": 9.44645406105536e-06,
+      "loss": 0.5185,
+      "step": 427
+    },
+    {
+      "epoch": 0.2366820276497696,
+      "grad_norm": 0.4929436147212982,
+      "learning_rate": 9.44203033900189e-06,
+      "loss": 0.5048,
+      "step": 428
+    },
+    {
+      "epoch": 0.23723502304147465,
+      "grad_norm": 0.46492716670036316,
+      "learning_rate": 9.437590055206032e-06,
+      "loss": 0.4939,
+      "step": 429
+    },
+    {
+      "epoch": 0.23778801843317973,
+      "grad_norm": 0.4560331404209137,
+      "learning_rate": 9.433133226223018e-06,
+      "loss": 0.5016,
+      "step": 430
+    },
+    {
+      "epoch": 0.2383410138248848,
+      "grad_norm": 0.4622935652732849,
+      "learning_rate": 9.428659868669765e-06,
+      "loss": 0.4748,
+      "step": 431
+    },
+    {
+      "epoch": 0.23889400921658988,
+      "grad_norm": 0.4686061441898346,
+      "learning_rate": 9.424169999224819e-06,
+      "loss": 0.5012,
+      "step": 432
+    },
+    {
+      "epoch": 0.23944700460829493,
+      "grad_norm": 0.47005850076675415,
+      "learning_rate": 9.419663634628283e-06,
+      "loss": 0.4902,
+      "step": 433
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5422276854515076,
+      "learning_rate": 9.415140791681767e-06,
+      "loss": 0.4923,
+      "step": 434
+    },
+    {
+      "epoch": 0.24055299539170508,
+      "grad_norm": 0.42137962579727173,
+      "learning_rate": 9.410601487248315e-06,
+      "loss": 0.4762,
+      "step": 435
+    },
+    {
+      "epoch": 0.24110599078341013,
+      "grad_norm": 0.518340528011322,
+      "learning_rate": 9.406045738252349e-06,
+      "loss": 0.5257,
+      "step": 436
+    },
+    {
+      "epoch": 0.24165898617511522,
+      "grad_norm": 0.5296422243118286,
+      "learning_rate": 9.4014735616796e-06,
+      "loss": 0.4959,
+      "step": 437
+    },
+    {
+      "epoch": 0.24221198156682028,
+      "grad_norm": 0.4660092890262604,
+      "learning_rate": 9.396884974577052e-06,
+      "loss": 0.4902,
+      "step": 438
+    },
+    {
+      "epoch": 0.24276497695852534,
+      "grad_norm": 0.5939272046089172,
+      "learning_rate": 9.392279994052872e-06,
+      "loss": 0.4859,
+      "step": 439
+    },
+    {
+      "epoch": 0.24331797235023042,
+      "grad_norm": 0.4680648446083069,
+      "learning_rate": 9.387658637276348e-06,
+      "loss": 0.4856,
+      "step": 440
+    },
+    {
+      "epoch": 0.24387096774193548,
+      "grad_norm": 0.4640378952026367,
+      "learning_rate": 9.383020921477824e-06,
+      "loss": 0.4976,
+      "step": 441
+    },
+    {
+      "epoch": 0.24442396313364056,
+      "grad_norm": 0.5372658371925354,
+      "learning_rate": 9.378366863948644e-06,
+      "loss": 0.4731,
+      "step": 442
+    },
+    {
+      "epoch": 0.24497695852534562,
+      "grad_norm": 0.48611563444137573,
+      "learning_rate": 9.373696482041072e-06,
+      "loss": 0.474,
+      "step": 443
+    },
+    {
+      "epoch": 0.24552995391705068,
+      "grad_norm": 0.4766850173473358,
+      "learning_rate": 9.369009793168243e-06,
+      "loss": 0.4878,
+      "step": 444
+    },
+    {
+      "epoch": 0.24608294930875577,
+      "grad_norm": 0.503322184085846,
+      "learning_rate": 9.364306814804086e-06,
+      "loss": 0.49,
+      "step": 445
+    },
+    {
+      "epoch": 0.24663594470046082,
+      "grad_norm": 0.5735645294189453,
+      "learning_rate": 9.35958756448327e-06,
+      "loss": 0.5357,
+      "step": 446
+    },
+    {
+      "epoch": 0.2471889400921659,
+      "grad_norm": 0.5782787799835205,
+      "learning_rate": 9.354852059801127e-06,
+      "loss": 0.5052,
+      "step": 447
+    },
+    {
+      "epoch": 0.24774193548387097,
+      "grad_norm": 0.48679542541503906,
+      "learning_rate": 9.350100318413594e-06,
+      "loss": 0.4824,
+      "step": 448
+    },
+    {
+      "epoch": 0.24829493087557603,
+      "grad_norm": 0.535241961479187,
+      "learning_rate": 9.345332358037147e-06,
+      "loss": 0.4814,
+      "step": 449
+    },
+    {
+      "epoch": 0.2488479262672811,
+      "grad_norm": 0.5126506090164185,
+      "learning_rate": 9.340548196448729e-06,
+      "loss": 0.4903,
+      "step": 450
+    },
+    {
+      "epoch": 0.24940092165898617,
+      "grad_norm": 0.459888219833374,
+      "learning_rate": 9.33574785148569e-06,
+      "loss": 0.4764,
+      "step": 451
+    },
+    {
+      "epoch": 0.24995391705069125,
+      "grad_norm": 0.47858306765556335,
+      "learning_rate": 9.330931341045723e-06,
+      "loss": 0.4759,
+      "step": 452
+    },
+    {
+      "epoch": 0.2505069124423963,
+      "grad_norm": 0.4925948977470398,
+      "learning_rate": 9.326098683086786e-06,
+      "loss": 0.5092,
+      "step": 453
+    },
+    {
+      "epoch": 0.25105990783410137,
+      "grad_norm": 0.4197613000869751,
+      "learning_rate": 9.321249895627043e-06,
+      "loss": 0.4855,
+      "step": 454
+    },
+    {
+      "epoch": 0.25161290322580643,
+      "grad_norm": 0.4368072748184204,
+      "learning_rate": 9.316384996744798e-06,
+      "loss": 0.4731,
+      "step": 455
+    },
+    {
+      "epoch": 0.25216589861751154,
+      "grad_norm": 0.4712340533733368,
+      "learning_rate": 9.311504004578425e-06,
+      "loss": 0.4864,
+      "step": 456
+    },
+    {
+      "epoch": 0.2527188940092166,
+      "grad_norm": 0.4518265426158905,
+      "learning_rate": 9.306606937326295e-06,
+      "loss": 0.5119,
+      "step": 457
+    },
+    {
+      "epoch": 0.25327188940092166,
+      "grad_norm": 0.48842400312423706,
+      "learning_rate": 9.301693813246721e-06,
+      "loss": 0.4732,
+      "step": 458
+    },
+    {
+      "epoch": 0.2538248847926267,
+      "grad_norm": 0.5565005540847778,
+      "learning_rate": 9.296764650657881e-06,
+      "loss": 0.5018,
+      "step": 459
+    },
+    {
+      "epoch": 0.2543778801843318,
+      "grad_norm": 0.4874143600463867,
+      "learning_rate": 9.291819467937746e-06,
+      "loss": 0.4712,
+      "step": 460
+    },
+    {
+      "epoch": 0.2549308755760369,
+      "grad_norm": 0.5408375859260559,
+      "learning_rate": 9.286858283524022e-06,
+      "loss": 0.4785,
+      "step": 461
+    },
+    {
+      "epoch": 0.25548387096774194,
+      "grad_norm": 0.4784450829029083,
+      "learning_rate": 9.281881115914075e-06,
+      "loss": 0.4893,
+      "step": 462
+    },
+    {
+      "epoch": 0.256036866359447,
+      "grad_norm": 0.4540750980377197,
+      "learning_rate": 9.276887983664861e-06,
+      "loss": 0.4873,
+      "step": 463
+    },
+    {
+      "epoch": 0.25658986175115206,
+      "grad_norm": 0.5902926921844482,
+      "learning_rate": 9.271878905392863e-06,
+      "loss": 0.4903,
+      "step": 464
+    },
+    {
+      "epoch": 0.2571428571428571,
+      "grad_norm": 0.44395557045936584,
+      "learning_rate": 9.26685389977401e-06,
+      "loss": 0.4862,
+      "step": 465
+    },
+    {
+      "epoch": 0.25769585253456223,
+      "grad_norm": 0.44266945123672485,
+      "learning_rate": 9.261812985543625e-06,
+      "loss": 0.4875,
+      "step": 466
+    },
+    {
+      "epoch": 0.2582488479262673,
+      "grad_norm": 0.5877528190612793,
+      "learning_rate": 9.256756181496334e-06,
+      "loss": 0.498,
+      "step": 467
+    },
+    {
+      "epoch": 0.25880184331797235,
+      "grad_norm": 0.43955346941947937,
+      "learning_rate": 9.251683506486016e-06,
+      "loss": 0.4782,
+      "step": 468
+    },
+    {
+      "epoch": 0.2593548387096774,
+      "grad_norm": 0.5338640809059143,
+      "learning_rate": 9.246594979425719e-06,
+      "loss": 0.4834,
+      "step": 469
+    },
+    {
+      "epoch": 0.25990783410138246,
+      "grad_norm": 0.4647727608680725,
+      "learning_rate": 9.241490619287593e-06,
+      "loss": 0.4957,
+      "step": 470
+    },
+    {
+      "epoch": 0.2604608294930876,
+      "grad_norm": 0.3950575590133667,
+      "learning_rate": 9.236370445102825e-06,
+      "loss": 0.4864,
+      "step": 471
+    },
+    {
+      "epoch": 0.26101382488479263,
+      "grad_norm": 0.5759503245353699,
+      "learning_rate": 9.231234475961559e-06,
+      "loss": 0.5049,
+      "step": 472
+    },
+    {
+      "epoch": 0.2615668202764977,
+      "grad_norm": 0.49982547760009766,
+      "learning_rate": 9.22608273101283e-06,
+      "loss": 0.5099,
+      "step": 473
+    },
+    {
+      "epoch": 0.26211981566820275,
+      "grad_norm": 0.45275378227233887,
+      "learning_rate": 9.220915229464496e-06,
+      "loss": 0.4768,
+      "step": 474
+    },
+    {
+      "epoch": 0.2626728110599078,
+      "grad_norm": 0.4712145924568176,
+      "learning_rate": 9.215731990583155e-06,
+      "loss": 0.4418,
+      "step": 475
+    },
+    {
+      "epoch": 0.2632258064516129,
+      "grad_norm": 0.4405660927295685,
+      "learning_rate": 9.210533033694084e-06,
+      "loss": 0.4938,
+      "step": 476
+    },
+    {
+      "epoch": 0.263778801843318,
+      "grad_norm": 0.4714353382587433,
+      "learning_rate": 9.205318378181167e-06,
+      "loss": 0.4806,
+      "step": 477
+    },
+    {
+      "epoch": 0.26433179723502304,
+      "grad_norm": 0.5084551572799683,
+      "learning_rate": 9.200088043486813e-06,
+      "loss": 0.487,
+      "step": 478
+    },
+    {
+      "epoch": 0.2648847926267281,
+      "grad_norm": 0.4339804947376251,
+      "learning_rate": 9.194842049111889e-06,
+      "loss": 0.4703,
+      "step": 479
+    },
+    {
+      "epoch": 0.2654377880184332,
+      "grad_norm": 0.5167462825775146,
+      "learning_rate": 9.189580414615658e-06,
+      "loss": 0.495,
+      "step": 480
+    },
+    {
+      "epoch": 0.26599078341013827,
+      "grad_norm": 0.5027214288711548,
+      "learning_rate": 9.184303159615682e-06,
+      "loss": 0.487,
+      "step": 481
+    },
+    {
+      "epoch": 0.2665437788018433,
+      "grad_norm": 0.5376026034355164,
+      "learning_rate": 9.179010303787772e-06,
+      "loss": 0.4974,
+      "step": 482
+    },
+    {
+      "epoch": 0.2670967741935484,
+      "grad_norm": 0.4745848774909973,
+      "learning_rate": 9.173701866865905e-06,
+      "loss": 0.4761,
+      "step": 483
+    },
+    {
+      "epoch": 0.26764976958525344,
+      "grad_norm": 0.615476667881012,
+      "learning_rate": 9.168377868642142e-06,
+      "loss": 0.4903,
+      "step": 484
+    },
+    {
+      "epoch": 0.26820276497695855,
+      "grad_norm": 0.43914616107940674,
+      "learning_rate": 9.163038328966578e-06,
+      "loss": 0.5018,
+      "step": 485
+    },
+    {
+      "epoch": 0.2687557603686636,
+      "grad_norm": 0.5516188144683838,
+      "learning_rate": 9.157683267747239e-06,
+      "loss": 0.5042,
+      "step": 486
+    },
+    {
+      "epoch": 0.26930875576036867,
+      "grad_norm": 0.5231643319129944,
+      "learning_rate": 9.152312704950028e-06,
+      "loss": 0.4745,
+      "step": 487
+    },
+    {
+      "epoch": 0.2698617511520737,
+      "grad_norm": 0.4931274354457855,
+      "learning_rate": 9.146926660598646e-06,
+      "loss": 0.4891,
+      "step": 488
+    },
+    {
+      "epoch": 0.2704147465437788,
+      "grad_norm": 0.5022376179695129,
+      "learning_rate": 9.141525154774513e-06,
+      "loss": 0.4707,
+      "step": 489
+    },
+    {
+      "epoch": 0.2709677419354839,
+      "grad_norm": 0.4848520755767822,
+      "learning_rate": 9.136108207616694e-06,
+      "loss": 0.5176,
+      "step": 490
+    },
+    {
+      "epoch": 0.27152073732718895,
+      "grad_norm": 0.5185691118240356,
+      "learning_rate": 9.130675839321829e-06,
+      "loss": 0.4951,
+      "step": 491
+    },
+    {
+      "epoch": 0.272073732718894,
+      "grad_norm": 0.4987480938434601,
+      "learning_rate": 9.125228070144056e-06,
+      "loss": 0.4865,
+      "step": 492
+    },
+    {
+      "epoch": 0.27262672811059907,
+      "grad_norm": 0.47905007004737854,
+      "learning_rate": 9.119764920394928e-06,
+      "loss": 0.4931,
+      "step": 493
+    },
+    {
+      "epoch": 0.27317972350230413,
+      "grad_norm": 0.5060631632804871,
+      "learning_rate": 9.114286410443349e-06,
+      "loss": 0.4912,
+      "step": 494
+    },
+    {
+      "epoch": 0.27373271889400924,
+      "grad_norm": 0.5127936005592346,
+      "learning_rate": 9.108792560715487e-06,
+      "loss": 0.5004,
+      "step": 495
+    },
+    {
+      "epoch": 0.2742857142857143,
+      "grad_norm": 0.5518050193786621,
+      "learning_rate": 9.103283391694711e-06,
+      "loss": 0.5051,
+      "step": 496
+    },
+    {
+      "epoch": 0.27483870967741936,
+      "grad_norm": 0.5080561637878418,
+      "learning_rate": 9.097758923921495e-06,
+      "loss": 0.4679,
+      "step": 497
+    },
+    {
+      "epoch": 0.2753917050691244,
+      "grad_norm": 0.5666584372520447,
+      "learning_rate": 9.092219177993365e-06,
+      "loss": 0.4861,
+      "step": 498
+    },
+    {
+      "epoch": 0.2759447004608295,
+      "grad_norm": 0.48917266726493835,
+      "learning_rate": 9.086664174564804e-06,
+      "loss": 0.4616,
+      "step": 499
+    },
+    {
+      "epoch": 0.2764976958525346,
+      "grad_norm": 0.4538862407207489,
+      "learning_rate": 9.081093934347178e-06,
+      "loss": 0.4917,
+      "step": 500
+    },
+    {
+      "epoch": 0.27705069124423964,
+      "grad_norm": 0.6453654170036316,
+      "learning_rate": 9.075508478108674e-06,
+      "loss": 0.477,
+      "step": 501
+    },
+    {
+      "epoch": 0.2776036866359447,
+      "grad_norm": 0.5007541179656982,
+      "learning_rate": 9.069907826674199e-06,
+      "loss": 0.4748,
+      "step": 502
+    },
+    {
+      "epoch": 0.27815668202764976,
+      "grad_norm": 0.5134044289588928,
+      "learning_rate": 9.064292000925316e-06,
+      "loss": 0.5122,
+      "step": 503
+    },
+    {
+      "epoch": 0.2787096774193548,
+      "grad_norm": 0.5208799242973328,
+      "learning_rate": 9.05866102180017e-06,
+      "loss": 0.4908,
+      "step": 504
+    },
+    {
+      "epoch": 0.27926267281105993,
+      "grad_norm": 0.4694979786872864,
+      "learning_rate": 9.053014910293399e-06,
+      "loss": 0.5006,
+      "step": 505
+    },
+    {
+      "epoch": 0.279815668202765,
+      "grad_norm": 0.4957534670829773,
+      "learning_rate": 9.047353687456064e-06,
+      "loss": 0.4901,
+      "step": 506
+    },
+    {
+      "epoch": 0.28036866359447005,
+      "grad_norm": 0.5231026411056519,
+      "learning_rate": 9.041677374395563e-06,
+      "loss": 0.5062,
+      "step": 507
+    },
+    {
+      "epoch": 0.2809216589861751,
+      "grad_norm": 0.5049005150794983,
+      "learning_rate": 9.035985992275562e-06,
+      "loss": 0.501,
+      "step": 508
+    },
+    {
+      "epoch": 0.28147465437788016,
+      "grad_norm": 0.48721855878829956,
+      "learning_rate": 9.030279562315907e-06,
+      "loss": 0.4881,
+      "step": 509
+    },
+    {
+      "epoch": 0.2820276497695853,
+      "grad_norm": 0.47729742527008057,
+      "learning_rate": 9.02455810579255e-06,
+      "loss": 0.4734,
+      "step": 510
+    },
+    {
+      "epoch": 0.28258064516129033,
+      "grad_norm": 0.5377198457717896,
+      "learning_rate": 9.018821644037466e-06,
+      "loss": 0.4655,
+      "step": 511
+    },
+    {
+      "epoch": 0.2831336405529954,
+      "grad_norm": 0.5237789154052734,
+      "learning_rate": 9.01307019843858e-06,
+      "loss": 0.5081,
+      "step": 512
+    },
+    {
+      "epoch": 0.28368663594470045,
+      "grad_norm": 0.6095771193504333,
+      "learning_rate": 9.00730379043968e-06,
+      "loss": 0.4938,
+      "step": 513
+    },
+    {
+      "epoch": 0.2842396313364055,
+      "grad_norm": 0.5196386575698853,
+      "learning_rate": 9.001522441540342e-06,
+      "loss": 0.4901,
+      "step": 514
+    },
+    {
+      "epoch": 0.2847926267281106,
+      "grad_norm": 0.5163764357566833,
+      "learning_rate": 8.995726173295845e-06,
+      "loss": 0.4901,
+      "step": 515
+    },
+    {
+      "epoch": 0.2853456221198157,
+      "grad_norm": 0.5501850247383118,
+      "learning_rate": 8.989915007317095e-06,
+      "loss": 0.4884,
+      "step": 516
+    },
+    {
+      "epoch": 0.28589861751152074,
+      "grad_norm": 0.4907573461532593,
+      "learning_rate": 8.984088965270547e-06,
+      "loss": 0.4803,
+      "step": 517
+    },
+    {
+      "epoch": 0.2864516129032258,
+      "grad_norm": 0.4490540325641632,
+      "learning_rate": 8.978248068878113e-06,
+      "loss": 0.5025,
+      "step": 518
+    },
+    {
+      "epoch": 0.28700460829493085,
+      "grad_norm": 0.541654109954834,
+      "learning_rate": 8.97239233991709e-06,
+      "loss": 0.4844,
+      "step": 519
+    },
+    {
+      "epoch": 0.28755760368663597,
+      "grad_norm": 0.4671088457107544,
+      "learning_rate": 8.966521800220084e-06,
+      "loss": 0.4775,
+      "step": 520
+    },
+    {
+      "epoch": 0.288110599078341,
+      "grad_norm": 0.5631542801856995,
+      "learning_rate": 8.960636471674913e-06,
+      "loss": 0.4723,
+      "step": 521
+    },
+    {
+      "epoch": 0.2886635944700461,
+      "grad_norm": 0.5776397585868835,
+      "learning_rate": 8.954736376224536e-06,
+      "loss": 0.4949,
+      "step": 522
+    },
+    {
+      "epoch": 0.28921658986175114,
+      "grad_norm": 0.5042744874954224,
+      "learning_rate": 8.948821535866973e-06,
+      "loss": 0.4934,
+      "step": 523
+    },
+    {
+      "epoch": 0.2897695852534562,
+      "grad_norm": 0.5939375162124634,
+      "learning_rate": 8.942891972655217e-06,
+      "loss": 0.4884,
+      "step": 524
+    },
+    {
+      "epoch": 0.2903225806451613,
+      "grad_norm": 0.4696662127971649,
+      "learning_rate": 8.93694770869715e-06,
+      "loss": 0.5052,
+      "step": 525
+    },
+    {
+      "epoch": 0.29087557603686637,
+      "grad_norm": 0.45712941884994507,
+      "learning_rate": 8.930988766155472e-06,
+      "loss": 0.5488,
+      "step": 526
+    },
+    {
+      "epoch": 0.2914285714285714,
+      "grad_norm": 0.5518166422843933,
+      "learning_rate": 8.925015167247604e-06,
+      "loss": 0.5129,
+      "step": 527
+    },
+    {
+      "epoch": 0.2919815668202765,
+      "grad_norm": 0.446274071931839,
+      "learning_rate": 8.919026934245618e-06,
+      "loss": 0.504,
+      "step": 528
+    },
+    {
+      "epoch": 0.29253456221198154,
+      "grad_norm": 0.5146806836128235,
+      "learning_rate": 8.91302408947614e-06,
+      "loss": 0.4779,
+      "step": 529
+    },
+    {
+      "epoch": 0.29308755760368665,
+      "grad_norm": 0.4867715537548065,
+      "learning_rate": 8.907006655320287e-06,
+      "loss": 0.4743,
+      "step": 530
+    },
+    {
+      "epoch": 0.2936405529953917,
+      "grad_norm": 0.47047367691993713,
+      "learning_rate": 8.90097465421356e-06,
+      "loss": 0.4881,
+      "step": 531
+    },
+    {
+      "epoch": 0.29419354838709677,
+      "grad_norm": 0.46602970361709595,
+      "learning_rate": 8.894928108645772e-06,
+      "loss": 0.486,
+      "step": 532
+    },
+    {
+      "epoch": 0.29474654377880183,
+      "grad_norm": 0.45530185103416443,
+      "learning_rate": 8.888867041160975e-06,
+      "loss": 0.4726,
+      "step": 533
+    },
+    {
+      "epoch": 0.2952995391705069,
+      "grad_norm": 0.4860335886478424,
+      "learning_rate": 8.882791474357354e-06,
+      "loss": 0.4818,
+      "step": 534
+    },
+    {
+      "epoch": 0.295852534562212,
+      "grad_norm": 0.4780627191066742,
+      "learning_rate": 8.876701430887156e-06,
+      "loss": 0.4937,
+      "step": 535
+    },
+    {
+      "epoch": 0.29640552995391706,
+      "grad_norm": 0.488271027803421,
+      "learning_rate": 8.870596933456603e-06,
+      "loss": 0.4988,
+      "step": 536
+    },
+    {
+      "epoch": 0.2969585253456221,
+      "grad_norm": 0.48193061351776123,
+      "learning_rate": 8.864478004825812e-06,
+      "loss": 0.4925,
+      "step": 537
+    },
+    {
+      "epoch": 0.2975115207373272,
+      "grad_norm": 0.5294972658157349,
+      "learning_rate": 8.8583446678087e-06,
+      "loss": 0.4787,
+      "step": 538
+    },
+    {
+      "epoch": 0.29806451612903223,
+      "grad_norm": 0.48833131790161133,
+      "learning_rate": 8.852196945272906e-06,
+      "loss": 0.5055,
+      "step": 539
+    },
+    {
+      "epoch": 0.29861751152073734,
+      "grad_norm": 0.4671526551246643,
+      "learning_rate": 8.846034860139706e-06,
+      "loss": 0.4775,
+      "step": 540
+    },
+    {
+      "epoch": 0.2991705069124424,
+      "grad_norm": 0.6074989438056946,
+      "learning_rate": 8.839858435383918e-06,
+      "loss": 0.4894,
+      "step": 541
+    },
+    {
+      "epoch": 0.29972350230414746,
+      "grad_norm": 0.46650996804237366,
+      "learning_rate": 8.833667694033841e-06,
+      "loss": 0.4913,
+      "step": 542
+    },
+    {
+      "epoch": 0.3002764976958525,
+      "grad_norm": 0.7234087586402893,
+      "learning_rate": 8.827462659171134e-06,
+      "loss": 0.4794,
+      "step": 543
+    },
+    {
+      "epoch": 0.30082949308755763,
+      "grad_norm": 0.5737038850784302,
+      "learning_rate": 8.821243353930756e-06,
+      "loss": 0.5169,
+      "step": 544
+    },
+    {
+      "epoch": 0.3013824884792627,
+      "grad_norm": 0.6461325883865356,
+      "learning_rate": 8.815009801500875e-06,
+      "loss": 0.4846,
+      "step": 545
+    },
+    {
+      "epoch": 0.30193548387096775,
+      "grad_norm": 0.4608705937862396,
+      "learning_rate": 8.808762025122774e-06,
+      "loss": 0.4763,
+      "step": 546
+    },
+    {
+      "epoch": 0.3024884792626728,
+      "grad_norm": 0.6168321371078491,
+      "learning_rate": 8.802500048090769e-06,
+      "loss": 0.4906,
+      "step": 547
+    },
+    {
+      "epoch": 0.30304147465437786,
+      "grad_norm": 0.5032297968864441,
+      "learning_rate": 8.796223893752125e-06,
+      "loss": 0.4809,
+      "step": 548
+    },
+    {
+      "epoch": 0.303594470046083,
+      "grad_norm": 0.48676756024360657,
+      "learning_rate": 8.789933585506962e-06,
+      "loss": 0.4844,
+      "step": 549
+    },
+    {
+      "epoch": 0.30414746543778803,
+      "grad_norm": 0.5273764133453369,
+      "learning_rate": 8.783629146808175e-06,
+      "loss": 0.4979,
+      "step": 550
+    },
+    {
+      "epoch": 0.3047004608294931,
+      "grad_norm": 0.47446364164352417,
+      "learning_rate": 8.77731060116134e-06,
+      "loss": 0.4633,
+      "step": 551
+    },
+    {
+      "epoch": 0.30525345622119815,
+      "grad_norm": 0.5045877695083618,
+      "learning_rate": 8.77097797212463e-06,
+      "loss": 0.4907,
+      "step": 552
+    },
+    {
+      "epoch": 0.3058064516129032,
+      "grad_norm": 0.4806043803691864,
+      "learning_rate": 8.764631283308733e-06,
+      "loss": 0.4794,
+      "step": 553
+    },
+    {
+      "epoch": 0.3063594470046083,
+      "grad_norm": 0.5195322036743164,
+      "learning_rate": 8.758270558376752e-06,
+      "loss": 0.4921,
+      "step": 554
+    },
+    {
+      "epoch": 0.3069124423963134,
+      "grad_norm": 0.5429467558860779,
+      "learning_rate": 8.751895821044118e-06,
+      "loss": 0.4825,
+      "step": 555
+    },
+    {
+      "epoch": 0.30746543778801844,
+      "grad_norm": 0.48403802514076233,
+      "learning_rate": 8.745507095078515e-06,
+      "loss": 0.4997,
+      "step": 556
+    },
+    {
+      "epoch": 0.3080184331797235,
+      "grad_norm": 0.5179433822631836,
+      "learning_rate": 8.739104404299781e-06,
+      "loss": 0.4444,
+      "step": 557
+    },
+    {
+      "epoch": 0.30857142857142855,
+      "grad_norm": 0.5344632863998413,
+      "learning_rate": 8.732687772579816e-06,
+      "loss": 0.4695,
+      "step": 558
+    },
+    {
+      "epoch": 0.30912442396313367,
+      "grad_norm": 0.5367107391357422,
+      "learning_rate": 8.7262572238425e-06,
+      "loss": 0.453,
+      "step": 559
+    },
+    {
+      "epoch": 0.3096774193548387,
+      "grad_norm": 0.5187183618545532,
+      "learning_rate": 8.719812782063603e-06,
+      "loss": 0.4975,
+      "step": 560
+    },
+    {
+      "epoch": 0.3102304147465438,
+      "grad_norm": 0.4541904330253601,
+      "learning_rate": 8.713354471270691e-06,
+      "loss": 0.5075,
+      "step": 561
+    },
+    {
+      "epoch": 0.31078341013824884,
+      "grad_norm": 0.4493202865123749,
+      "learning_rate": 8.706882315543043e-06,
+      "loss": 0.458,
+      "step": 562
+    },
+    {
+      "epoch": 0.3113364055299539,
+      "grad_norm": 0.571712076663971,
+      "learning_rate": 8.700396339011554e-06,
+      "loss": 0.5057,
+      "step": 563
+    },
+    {
+      "epoch": 0.311889400921659,
+      "grad_norm": 0.4367184042930603,
+      "learning_rate": 8.69389656585865e-06,
+      "loss": 0.4725,
+      "step": 564
+    },
+    {
+      "epoch": 0.31244239631336407,
+      "grad_norm": 0.5127226710319519,
+      "learning_rate": 8.6873830203182e-06,
+      "loss": 0.4615,
+      "step": 565
+    },
+    {
+      "epoch": 0.3129953917050691,
+      "grad_norm": 0.5464975833892822,
+      "learning_rate": 8.680855726675419e-06,
+      "loss": 0.4661,
+      "step": 566
+    },
+    {
+      "epoch": 0.3135483870967742,
+      "grad_norm": 0.5116366147994995,
+      "learning_rate": 8.674314709266778e-06,
+      "loss": 0.4789,
+      "step": 567
+    },
+    {
+      "epoch": 0.31410138248847924,
+      "grad_norm": 0.5597204566001892,
+      "learning_rate": 8.667759992479922e-06,
+      "loss": 0.4903,
+      "step": 568
+    },
+    {
+      "epoch": 0.31465437788018435,
+      "grad_norm": 0.4036201536655426,
+      "learning_rate": 8.661191600753567e-06,
+      "loss": 0.5093,
+      "step": 569
+    },
+    {
+      "epoch": 0.3152073732718894,
+      "grad_norm": 0.6131529808044434,
+      "learning_rate": 8.65460955857742e-06,
+      "loss": 0.4572,
+      "step": 570
+    },
+    {
+      "epoch": 0.31576036866359447,
+      "grad_norm": 0.532192587852478,
+      "learning_rate": 8.648013890492081e-06,
+      "loss": 0.4679,
+      "step": 571
+    },
+    {
+      "epoch": 0.31631336405529953,
+      "grad_norm": 0.4397170841693878,
+      "learning_rate": 8.641404621088951e-06,
+      "loss": 0.4841,
+      "step": 572
+    },
+    {
+      "epoch": 0.3168663594470046,
+      "grad_norm": 0.44879838824272156,
+      "learning_rate": 8.634781775010144e-06,
+      "loss": 0.4853,
+      "step": 573
+    },
+    {
+      "epoch": 0.3174193548387097,
+      "grad_norm": 0.5257536768913269,
+      "learning_rate": 8.628145376948392e-06,
+      "loss": 0.4915,
+      "step": 574
+    },
+    {
+      "epoch": 0.31797235023041476,
+      "grad_norm": 0.4205341637134552,
+      "learning_rate": 8.621495451646958e-06,
+      "loss": 0.4895,
+      "step": 575
+    },
+    {
+      "epoch": 0.3185253456221198,
+      "grad_norm": 0.4227660298347473,
+      "learning_rate": 8.614832023899536e-06,
+      "loss": 0.488,
+      "step": 576
+    },
+    {
+      "epoch": 0.3190783410138249,
+      "grad_norm": 0.441709965467453,
+      "learning_rate": 8.608155118550167e-06,
+      "loss": 0.4756,
+      "step": 577
+    },
+    {
+      "epoch": 0.31963133640552993,
+      "grad_norm": 0.4320451021194458,
+      "learning_rate": 8.601464760493138e-06,
+      "loss": 0.4726,
+      "step": 578
+    },
+    {
+      "epoch": 0.32018433179723504,
+      "grad_norm": 0.3868134617805481,
+      "learning_rate": 8.594760974672898e-06,
+      "loss": 0.475,
+      "step": 579
+    },
+    {
+      "epoch": 0.3207373271889401,
+      "grad_norm": 0.43438073992729187,
+      "learning_rate": 8.588043786083952e-06,
+      "loss": 0.4874,
+      "step": 580
+    },
+    {
+      "epoch": 0.32129032258064516,
+      "grad_norm": 0.40946924686431885,
+      "learning_rate": 8.581313219770788e-06,
+      "loss": 0.4843,
+      "step": 581
+    },
+    {
+      "epoch": 0.3218433179723502,
+      "grad_norm": 0.4077948331832886,
+      "learning_rate": 8.57456930082776e-06,
+      "loss": 0.4656,
+      "step": 582
+    },
+    {
+      "epoch": 0.3223963133640553,
+      "grad_norm": 0.4392508566379547,
+      "learning_rate": 8.567812054399016e-06,
+      "loss": 0.4911,
+      "step": 583
+    },
+    {
+      "epoch": 0.3229493087557604,
+      "grad_norm": 0.4225884974002838,
+      "learning_rate": 8.561041505678385e-06,
+      "loss": 0.4907,
+      "step": 584
+    },
+    {
+      "epoch": 0.32350230414746545,
+      "grad_norm": 0.4444757103919983,
+      "learning_rate": 8.5542576799093e-06,
+      "loss": 0.484,
+      "step": 585
+    },
+    {
+      "epoch": 0.3240552995391705,
+      "grad_norm": 0.4226742088794708,
+      "learning_rate": 8.547460602384697e-06,
+      "loss": 0.4947,
+      "step": 586
+    },
+    {
+      "epoch": 0.32460829493087556,
+      "grad_norm": 0.5053589344024658,
+      "learning_rate": 8.540650298446912e-06,
+      "loss": 0.4984,
+      "step": 587
+    },
+    {
+      "epoch": 0.3251612903225806,
+      "grad_norm": 0.4462142586708069,
+      "learning_rate": 8.533826793487601e-06,
+      "loss": 0.4624,
+      "step": 588
+    },
+    {
+      "epoch": 0.32571428571428573,
+      "grad_norm": 0.5019125938415527,
+      "learning_rate": 8.526990112947636e-06,
+      "loss": 0.4597,
+      "step": 589
+    },
+    {
+      "epoch": 0.3262672811059908,
+      "grad_norm": 0.4495023787021637,
+      "learning_rate": 8.520140282317018e-06,
+      "loss": 0.4619,
+      "step": 590
+    },
+    {
+      "epoch": 0.32682027649769585,
+      "grad_norm": 0.45633986592292786,
+      "learning_rate": 8.51327732713477e-06,
+      "loss": 0.4997,
+      "step": 591
+    },
+    {
+      "epoch": 0.3273732718894009,
+      "grad_norm": 0.45402443408966064,
+      "learning_rate": 8.506401272988854e-06,
+      "loss": 0.4726,
+      "step": 592
+    },
+    {
+      "epoch": 0.32792626728110597,
+      "grad_norm": 0.4144057631492615,
+      "learning_rate": 8.499512145516068e-06,
+      "loss": 0.483,
+      "step": 593
+    },
+    {
+      "epoch": 0.3284792626728111,
+      "grad_norm": 0.506801962852478,
+      "learning_rate": 8.492609970401955e-06,
+      "loss": 0.4936,
+      "step": 594
+    },
+    {
+      "epoch": 0.32903225806451614,
+      "grad_norm": 0.45201730728149414,
+      "learning_rate": 8.4856947733807e-06,
+      "loss": 0.5055,
+      "step": 595
+    },
+    {
+      "epoch": 0.3295852534562212,
+      "grad_norm": 0.4555196762084961,
+      "learning_rate": 8.478766580235044e-06,
+      "loss": 0.4847,
+      "step": 596
+    },
+    {
+      "epoch": 0.33013824884792625,
+      "grad_norm": 0.49041327834129333,
+      "learning_rate": 8.471825416796185e-06,
+      "loss": 0.4662,
+      "step": 597
+    },
+    {
+      "epoch": 0.3306912442396313,
+      "grad_norm": 0.466457724571228,
+      "learning_rate": 8.464871308943673e-06,
+      "loss": 0.4794,
+      "step": 598
+    },
+    {
+      "epoch": 0.3312442396313364,
+      "grad_norm": 0.5273913145065308,
+      "learning_rate": 8.457904282605324e-06,
+      "loss": 0.482,
+      "step": 599
+    },
+    {
+      "epoch": 0.3317972350230415,
+      "grad_norm": 0.42694202065467834,
+      "learning_rate": 8.45092436375712e-06,
+      "loss": 0.4663,
+      "step": 600
+    },
+    {
+      "epoch": 0.33235023041474654,
+      "grad_norm": 0.467692494392395,
+      "learning_rate": 8.443931578423108e-06,
+      "loss": 0.4606,
+      "step": 601
+    },
+    {
+      "epoch": 0.3329032258064516,
+      "grad_norm": 0.5536178946495056,
+      "learning_rate": 8.436925952675312e-06,
+      "loss": 0.49,
+      "step": 602
+    },
+    {
+      "epoch": 0.33345622119815665,
+      "grad_norm": 0.5051153898239136,
+      "learning_rate": 8.429907512633626e-06,
+      "loss": 0.5041,
+      "step": 603
+    },
+    {
+      "epoch": 0.33400921658986177,
+      "grad_norm": 0.4975510835647583,
+      "learning_rate": 8.422876284465722e-06,
+      "loss": 0.491,
+      "step": 604
+    },
+    {
+      "epoch": 0.3345622119815668,
+      "grad_norm": 0.5645290613174438,
+      "learning_rate": 8.415832294386956e-06,
+      "loss": 0.4942,
+      "step": 605
+    },
+    {
+      "epoch": 0.3351152073732719,
+      "grad_norm": 0.5082492828369141,
+      "learning_rate": 8.408775568660256e-06,
+      "loss": 0.4829,
+      "step": 606
+    },
+    {
+      "epoch": 0.33566820276497694,
+      "grad_norm": 0.4222676753997803,
+      "learning_rate": 8.401706133596047e-06,
+      "loss": 0.4837,
+      "step": 607
+    },
+    {
+      "epoch": 0.33622119815668206,
+      "grad_norm": 0.5251336693763733,
+      "learning_rate": 8.394624015552124e-06,
+      "loss": 0.492,
+      "step": 608
+    },
+    {
+      "epoch": 0.3367741935483871,
+      "grad_norm": 0.4806731641292572,
+      "learning_rate": 8.387529240933585e-06,
+      "loss": 0.4791,
+      "step": 609
+    },
+    {
+      "epoch": 0.33732718894009217,
+      "grad_norm": 0.5003568530082703,
+      "learning_rate": 8.380421836192705e-06,
+      "loss": 0.4768,
+      "step": 610
+    },
+    {
+      "epoch": 0.33788018433179723,
+      "grad_norm": 0.48693132400512695,
+      "learning_rate": 8.373301827828858e-06,
+      "loss": 0.5219,
+      "step": 611
+    },
+    {
+      "epoch": 0.3384331797235023,
+      "grad_norm": 0.43921712040901184,
+      "learning_rate": 8.366169242388402e-06,
+      "loss": 0.4668,
+      "step": 612
+    },
+    {
+      "epoch": 0.3389861751152074,
+      "grad_norm": 0.5484280586242676,
+      "learning_rate": 8.359024106464594e-06,
+      "loss": 0.4669,
+      "step": 613
+    },
+    {
+      "epoch": 0.33953917050691246,
+      "grad_norm": 0.4520418643951416,
+      "learning_rate": 8.351866446697483e-06,
+      "loss": 0.4973,
+      "step": 614
+    },
+    {
+      "epoch": 0.3400921658986175,
+      "grad_norm": 0.588077962398529,
+      "learning_rate": 8.344696289773806e-06,
+      "loss": 0.4749,
+      "step": 615
+    },
+    {
+      "epoch": 0.3406451612903226,
+      "grad_norm": 0.4226244390010834,
+      "learning_rate": 8.337513662426904e-06,
+      "loss": 0.482,
+      "step": 616
+    },
+    {
+      "epoch": 0.34119815668202763,
+      "grad_norm": 0.5818718671798706,
+      "learning_rate": 8.330318591436606e-06,
+      "loss": 0.4963,
+      "step": 617
+    },
+    {
+      "epoch": 0.34175115207373274,
+      "grad_norm": 0.4959838092327118,
+      "learning_rate": 8.32311110362914e-06,
+      "loss": 0.4928,
+      "step": 618
+    },
+    {
+      "epoch": 0.3423041474654378,
+      "grad_norm": 0.43233931064605713,
+      "learning_rate": 8.315891225877026e-06,
+      "loss": 0.4619,
+      "step": 619
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 0.506004273891449,
+      "learning_rate": 8.308658985098983e-06,
+      "loss": 0.4638,
+      "step": 620
+    },
+    {
+      "epoch": 0.3434101382488479,
+      "grad_norm": 0.5699809193611145,
+      "learning_rate": 8.301414408259821e-06,
+      "loss": 0.4666,
+      "step": 621
+    },
+    {
+      "epoch": 0.343963133640553,
+      "grad_norm": 0.4087035059928894,
+      "learning_rate": 8.294157522370343e-06,
+      "loss": 0.4792,
+      "step": 622
+    },
+    {
+      "epoch": 0.3445161290322581,
+      "grad_norm": 0.482768714427948,
+      "learning_rate": 8.28688835448725e-06,
+      "loss": 0.4943,
+      "step": 623
+    },
+    {
+      "epoch": 0.34506912442396315,
+      "grad_norm": 0.4292857348918915,
+      "learning_rate": 8.27960693171303e-06,
+      "loss": 0.4809,
+      "step": 624
+    },
+    {
+      "epoch": 0.3456221198156682,
+      "grad_norm": 0.43942004442214966,
+      "learning_rate": 8.272313281195868e-06,
+      "loss": 0.488,
+      "step": 625
+    },
+    {
+      "epoch": 0.34617511520737326,
+      "grad_norm": 0.5329846739768982,
+      "learning_rate": 8.265007430129535e-06,
+      "loss": 0.494,
+      "step": 626
+    },
+    {
+      "epoch": 0.3467281105990783,
+      "grad_norm": 0.41792014241218567,
+      "learning_rate": 8.25768940575329e-06,
+      "loss": 0.4838,
+      "step": 627
+    },
+    {
+      "epoch": 0.34728110599078343,
+      "grad_norm": 0.522320568561554,
+      "learning_rate": 8.250359235351783e-06,
+      "loss": 0.4778,
+      "step": 628
+    },
+    {
+      "epoch": 0.3478341013824885,
+      "grad_norm": 0.4983687996864319,
+      "learning_rate": 8.243016946254947e-06,
+      "loss": 0.4761,
+      "step": 629
+    },
+    {
+      "epoch": 0.34838709677419355,
+      "grad_norm": 0.5216838717460632,
+      "learning_rate": 8.235662565837901e-06,
+      "loss": 0.4949,
+      "step": 630
+    },
+    {
+      "epoch": 0.3489400921658986,
+      "grad_norm": 0.5560440421104431,
+      "learning_rate": 8.22829612152084e-06,
+      "loss": 0.4653,
+      "step": 631
+    },
+    {
+      "epoch": 0.34949308755760367,
+      "grad_norm": 0.505742073059082,
+      "learning_rate": 8.220917640768943e-06,
+      "loss": 0.4981,
+      "step": 632
+    },
+    {
+      "epoch": 0.3500460829493088,
+      "grad_norm": 0.4594072103500366,
+      "learning_rate": 8.213527151092267e-06,
+      "loss": 0.4522,
+      "step": 633
+    },
+    {
+      "epoch": 0.35059907834101384,
+      "grad_norm": 0.40718746185302734,
+      "learning_rate": 8.20612468004564e-06,
+      "loss": 0.4728,
+      "step": 634
+    },
+    {
+      "epoch": 0.3511520737327189,
+      "grad_norm": 0.500005304813385,
+      "learning_rate": 8.19871025522856e-06,
+      "loss": 0.4667,
+      "step": 635
+    },
+    {
+      "epoch": 0.35170506912442395,
+      "grad_norm": 0.4374920427799225,
+      "learning_rate": 8.191283904285098e-06,
+      "loss": 0.4801,
+      "step": 636
+    },
+    {
+      "epoch": 0.352258064516129,
+      "grad_norm": 0.4499951899051666,
+      "learning_rate": 8.183845654903789e-06,
+      "loss": 0.4588,
+      "step": 637
+    },
+    {
+      "epoch": 0.3528110599078341,
+      "grad_norm": 0.41557565331459045,
+      "learning_rate": 8.17639553481753e-06,
+      "loss": 0.4953,
+      "step": 638
+    },
+    {
+      "epoch": 0.3533640552995392,
+      "grad_norm": 0.44742459058761597,
+      "learning_rate": 8.168933571803475e-06,
+      "loss": 0.4711,
+      "step": 639
+    },
+    {
+      "epoch": 0.35391705069124424,
+      "grad_norm": 0.4027085304260254,
+      "learning_rate": 8.161459793682937e-06,
+      "loss": 0.4872,
+      "step": 640
+    },
+    {
+      "epoch": 0.3544700460829493,
+      "grad_norm": 0.4093347191810608,
+      "learning_rate": 8.153974228321279e-06,
+      "loss": 0.477,
+      "step": 641
+    },
+    {
+      "epoch": 0.35502304147465436,
+      "grad_norm": 0.45894891023635864,
+      "learning_rate": 8.146476903627813e-06,
+      "loss": 0.4844,
+      "step": 642
+    },
+    {
+      "epoch": 0.35557603686635947,
+      "grad_norm": 0.44424736499786377,
+      "learning_rate": 8.138967847555693e-06,
+      "loss": 0.4754,
+      "step": 643
+    },
+    {
+      "epoch": 0.3561290322580645,
+      "grad_norm": 0.4633028209209442,
+      "learning_rate": 8.13144708810181e-06,
+      "loss": 0.4873,
+      "step": 644
+    },
+    {
+      "epoch": 0.3566820276497696,
+      "grad_norm": 0.46212074160575867,
+      "learning_rate": 8.123914653306695e-06,
+      "loss": 0.4576,
+      "step": 645
+    },
+    {
+      "epoch": 0.35723502304147464,
+      "grad_norm": 0.5635020732879639,
+      "learning_rate": 8.116370571254407e-06,
+      "loss": 0.4651,
+      "step": 646
+    },
+    {
+      "epoch": 0.3577880184331797,
+      "grad_norm": 0.43892374634742737,
+      "learning_rate": 8.108814870072434e-06,
+      "loss": 0.5018,
+      "step": 647
+    },
+    {
+      "epoch": 0.3583410138248848,
+      "grad_norm": 0.4993932843208313,
+      "learning_rate": 8.10124757793158e-06,
+      "loss": 0.4848,
+      "step": 648
+    },
+    {
+      "epoch": 0.35889400921658987,
+      "grad_norm": 0.5206483006477356,
+      "learning_rate": 8.093668723045866e-06,
+      "loss": 0.4845,
+      "step": 649
+    },
+    {
+      "epoch": 0.35944700460829493,
+      "grad_norm": 0.4889640808105469,
+      "learning_rate": 8.08607833367243e-06,
+      "loss": 0.4731,
+      "step": 650
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5545540452003479,
+      "learning_rate": 8.078476438111404e-06,
+      "loss": 0.4919,
+      "step": 651
+    },
+    {
+      "epoch": 0.36055299539170504,
+      "grad_norm": 0.47018927335739136,
+      "learning_rate": 8.07086306470583e-06,
+      "loss": 0.4902,
+      "step": 652
+    },
+    {
+      "epoch": 0.36110599078341016,
+      "grad_norm": 0.5213498473167419,
+      "learning_rate": 8.06323824184154e-06,
+      "loss": 0.4918,
+      "step": 653
+    },
+    {
+      "epoch": 0.3616589861751152,
+      "grad_norm": 0.4931851923465729,
+      "learning_rate": 8.055601997947056e-06,
+      "loss": 0.4868,
+      "step": 654
+    },
+    {
+      "epoch": 0.3622119815668203,
+      "grad_norm": 0.4996345341205597,
+      "learning_rate": 8.04795436149348e-06,
+      "loss": 0.4908,
+      "step": 655
+    },
+    {
+      "epoch": 0.36276497695852533,
+      "grad_norm": 0.48268017172813416,
+      "learning_rate": 8.04029536099439e-06,
+      "loss": 0.4975,
+      "step": 656
+    },
+    {
+      "epoch": 0.3633179723502304,
+      "grad_norm": 0.5365729928016663,
+      "learning_rate": 8.032625025005736e-06,
+      "loss": 0.4891,
+      "step": 657
+    },
+    {
+      "epoch": 0.3638709677419355,
+      "grad_norm": 0.5524755716323853,
+      "learning_rate": 8.024943382125732e-06,
+      "loss": 0.4833,
+      "step": 658
+    },
+    {
+      "epoch": 0.36442396313364056,
+      "grad_norm": 0.4499988257884979,
+      "learning_rate": 8.017250460994748e-06,
+      "loss": 0.4724,
+      "step": 659
+    },
+    {
+      "epoch": 0.3649769585253456,
+      "grad_norm": 0.44743797183036804,
+      "learning_rate": 8.0095462902952e-06,
+      "loss": 0.503,
+      "step": 660
+    },
+    {
+      "epoch": 0.3655299539170507,
+      "grad_norm": 0.502610445022583,
+      "learning_rate": 8.001830898751455e-06,
+      "loss": 0.4813,
+      "step": 661
+    },
+    {
+      "epoch": 0.36608294930875573,
+      "grad_norm": 0.41622281074523926,
+      "learning_rate": 7.99410431512971e-06,
+      "loss": 0.5157,
+      "step": 662
+    },
+    {
+      "epoch": 0.36663594470046085,
+      "grad_norm": 0.4399860203266144,
+      "learning_rate": 7.986366568237893e-06,
+      "loss": 0.4741,
+      "step": 663
+    },
+    {
+      "epoch": 0.3671889400921659,
+      "grad_norm": 0.43299001455307007,
+      "learning_rate": 7.97861768692555e-06,
+      "loss": 0.4727,
+      "step": 664
+    },
+    {
+      "epoch": 0.36774193548387096,
+      "grad_norm": 0.44035786390304565,
+      "learning_rate": 7.970857700083747e-06,
+      "loss": 0.4662,
+      "step": 665
+    },
+    {
+      "epoch": 0.368294930875576,
+      "grad_norm": 0.422536164522171,
+      "learning_rate": 7.96308663664495e-06,
+      "loss": 0.4969,
+      "step": 666
+    },
+    {
+      "epoch": 0.3688479262672811,
+      "grad_norm": 0.44590070843696594,
+      "learning_rate": 7.955304525582928e-06,
+      "loss": 0.4591,
+      "step": 667
+    },
+    {
+      "epoch": 0.3694009216589862,
+      "grad_norm": 0.43369054794311523,
+      "learning_rate": 7.947511395912635e-06,
+      "loss": 0.4941,
+      "step": 668
+    },
+    {
+      "epoch": 0.36995391705069125,
+      "grad_norm": 0.40852341055870056,
+      "learning_rate": 7.939707276690111e-06,
+      "loss": 0.4859,
+      "step": 669
+    },
+    {
+      "epoch": 0.3705069124423963,
+      "grad_norm": 0.3855592906475067,
+      "learning_rate": 7.93189219701237e-06,
+      "loss": 0.4834,
+      "step": 670
+    },
+    {
+      "epoch": 0.37105990783410137,
+      "grad_norm": 0.40031880140304565,
+      "learning_rate": 7.924066186017288e-06,
+      "loss": 0.5146,
+      "step": 671
+    },
+    {
+      "epoch": 0.3716129032258065,
+      "grad_norm": 0.42813777923583984,
+      "learning_rate": 7.916229272883499e-06,
+      "loss": 0.4875,
+      "step": 672
+    },
+    {
+      "epoch": 0.37216589861751154,
+      "grad_norm": 0.4341396689414978,
+      "learning_rate": 7.908381486830286e-06,
+      "loss": 0.4787,
+      "step": 673
+    },
+    {
+      "epoch": 0.3727188940092166,
+      "grad_norm": 0.40330180525779724,
+      "learning_rate": 7.90052285711747e-06,
+      "loss": 0.4866,
+      "step": 674
+    },
+    {
+      "epoch": 0.37327188940092165,
+      "grad_norm": 0.4432542026042938,
+      "learning_rate": 7.892653413045298e-06,
+      "loss": 0.4574,
+      "step": 675
+    },
+    {
+      "epoch": 0.3738248847926267,
+      "grad_norm": 0.4223710298538208,
+      "learning_rate": 7.884773183954344e-06,
+      "loss": 0.4619,
+      "step": 676
+    },
+    {
+      "epoch": 0.3743778801843318,
+      "grad_norm": 0.39498183131217957,
+      "learning_rate": 7.876882199225388e-06,
+      "loss": 0.4717,
+      "step": 677
+    },
+    {
+      "epoch": 0.3749308755760369,
+      "grad_norm": 0.4838169813156128,
+      "learning_rate": 7.868980488279316e-06,
+      "loss": 0.4756,
+      "step": 678
+    },
+    {
+      "epoch": 0.37548387096774194,
+      "grad_norm": 0.3836471140384674,
+      "learning_rate": 7.861068080576998e-06,
+      "loss": 0.462,
+      "step": 679
+    },
+    {
+      "epoch": 0.376036866359447,
+      "grad_norm": 0.503296971321106,
+      "learning_rate": 7.853145005619199e-06,
+      "loss": 0.4963,
+      "step": 680
+    },
+    {
+      "epoch": 0.37658986175115206,
+      "grad_norm": 0.3729606568813324,
+      "learning_rate": 7.84521129294644e-06,
+      "loss": 0.4684,
+      "step": 681
+    },
+    {
+      "epoch": 0.37714285714285717,
+      "grad_norm": 0.5140215158462524,
+      "learning_rate": 7.837266972138918e-06,
+      "loss": 0.5041,
+      "step": 682
+    },
+    {
+      "epoch": 0.3776958525345622,
+      "grad_norm": 0.4366711676120758,
+      "learning_rate": 7.82931207281637e-06,
+      "loss": 0.5007,
+      "step": 683
+    },
+    {
+      "epoch": 0.3782488479262673,
+      "grad_norm": 0.43705347180366516,
+      "learning_rate": 7.821346624637984e-06,
+      "loss": 0.4988,
+      "step": 684
+    },
+    {
+      "epoch": 0.37880184331797234,
+      "grad_norm": 0.4190782308578491,
+      "learning_rate": 7.813370657302273e-06,
+      "loss": 0.4815,
+      "step": 685
+    },
+    {
+      "epoch": 0.3793548387096774,
+      "grad_norm": 0.43545421957969666,
+      "learning_rate": 7.80538420054697e-06,
+      "loss": 0.486,
+      "step": 686
+    },
+    {
+      "epoch": 0.3799078341013825,
+      "grad_norm": 0.4012760818004608,
+      "learning_rate": 7.797387284148919e-06,
+      "loss": 0.456,
+      "step": 687
+    },
+    {
+      "epoch": 0.38046082949308757,
+      "grad_norm": 0.45634332299232483,
+      "learning_rate": 7.789379937923958e-06,
+      "loss": 0.4827,
+      "step": 688
+    },
+    {
+      "epoch": 0.38101382488479263,
+      "grad_norm": 0.45957717299461365,
+      "learning_rate": 7.781362191726818e-06,
+      "loss": 0.4976,
+      "step": 689
+    },
+    {
+      "epoch": 0.3815668202764977,
+      "grad_norm": 0.4902954399585724,
+      "learning_rate": 7.773334075451e-06,
+      "loss": 0.4952,
+      "step": 690
+    },
+    {
+      "epoch": 0.38211981566820274,
+      "grad_norm": 0.4607703387737274,
+      "learning_rate": 7.765295619028672e-06,
+      "loss": 0.4995,
+      "step": 691
+    },
+    {
+      "epoch": 0.38267281105990786,
+      "grad_norm": 0.49557220935821533,
+      "learning_rate": 7.757246852430553e-06,
+      "loss": 0.4921,
+      "step": 692
+    },
+    {
+      "epoch": 0.3832258064516129,
+      "grad_norm": 0.4447820782661438,
+      "learning_rate": 7.749187805665801e-06,
+      "loss": 0.4589,
+      "step": 693
+    },
+    {
+      "epoch": 0.383778801843318,
+      "grad_norm": 0.4028923809528351,
+      "learning_rate": 7.741118508781906e-06,
+      "loss": 0.4732,
+      "step": 694
+    },
+    {
+      "epoch": 0.38433179723502303,
+      "grad_norm": 0.4806511700153351,
+      "learning_rate": 7.733038991864576e-06,
+      "loss": 0.4968,
+      "step": 695
+    },
+    {
+      "epoch": 0.3848847926267281,
+      "grad_norm": 0.4322826862335205,
+      "learning_rate": 7.724949285037619e-06,
+      "loss": 0.4904,
+      "step": 696
+    },
+    {
+      "epoch": 0.3854377880184332,
+      "grad_norm": 0.44291332364082336,
+      "learning_rate": 7.716849418462836e-06,
+      "loss": 0.4829,
+      "step": 697
+    },
+    {
+      "epoch": 0.38599078341013826,
+      "grad_norm": 0.4092819094657898,
+      "learning_rate": 7.70873942233991e-06,
+      "loss": 0.4809,
+      "step": 698
+    },
+    {
+      "epoch": 0.3865437788018433,
+      "grad_norm": 0.4542578160762787,
+      "learning_rate": 7.700619326906288e-06,
+      "loss": 0.4713,
+      "step": 699
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 0.4618943929672241,
+      "learning_rate": 7.69248916243708e-06,
+      "loss": 0.4443,
+      "step": 700
+    },
+    {
+      "epoch": 0.38764976958525343,
+      "grad_norm": 0.43175920844078064,
+      "learning_rate": 7.684348959244924e-06,
+      "loss": 0.4761,
+      "step": 701
+    },
+    {
+      "epoch": 0.38820276497695855,
+      "grad_norm": 0.4224050045013428,
+      "learning_rate": 7.676198747679898e-06,
+      "loss": 0.4943,
+      "step": 702
+    },
+    {
+      "epoch": 0.3887557603686636,
+      "grad_norm": 0.4592989683151245,
+      "learning_rate": 7.66803855812939e-06,
+      "loss": 0.4529,
+      "step": 703
+    },
+    {
+      "epoch": 0.38930875576036866,
+      "grad_norm": 0.4457801878452301,
+      "learning_rate": 7.659868421017989e-06,
+      "loss": 0.4785,
+      "step": 704
+    },
+    {
+      "epoch": 0.3898617511520737,
+      "grad_norm": 0.4617078900337219,
+      "learning_rate": 7.651688366807378e-06,
+      "loss": 0.4755,
+      "step": 705
+    },
+    {
+      "epoch": 0.3904147465437788,
+      "grad_norm": 0.43977999687194824,
+      "learning_rate": 7.64349842599621e-06,
+      "loss": 0.5098,
+      "step": 706
+    },
+    {
+      "epoch": 0.3909677419354839,
+      "grad_norm": 0.4340912699699402,
+      "learning_rate": 7.635298629120002e-06,
+      "loss": 0.4452,
+      "step": 707
+    },
+    {
+      "epoch": 0.39152073732718895,
+      "grad_norm": 0.5204182863235474,
+      "learning_rate": 7.6270890067510196e-06,
+      "loss": 0.4677,
+      "step": 708
+    },
+    {
+      "epoch": 0.392073732718894,
+      "grad_norm": 0.47488752007484436,
+      "learning_rate": 7.618869589498157e-06,
+      "loss": 0.465,
+      "step": 709
+    },
+    {
+      "epoch": 0.39262672811059907,
+      "grad_norm": 0.4327450692653656,
+      "learning_rate": 7.610640408006832e-06,
+      "loss": 0.4945,
+      "step": 710
+    },
+    {
+      "epoch": 0.3931797235023041,
+      "grad_norm": 0.5072281360626221,
+      "learning_rate": 7.602401492958868e-06,
+      "loss": 0.4987,
+      "step": 711
+    },
+    {
+      "epoch": 0.39373271889400924,
+      "grad_norm": 0.4239625632762909,
+      "learning_rate": 7.594152875072376e-06,
+      "loss": 0.4715,
+      "step": 712
+    },
+    {
+      "epoch": 0.3942857142857143,
+      "grad_norm": 0.4698258936405182,
+      "learning_rate": 7.5858945851016455e-06,
+      "loss": 0.4779,
+      "step": 713
+    },
+    {
+      "epoch": 0.39483870967741935,
+      "grad_norm": 0.42160850763320923,
+      "learning_rate": 7.577626653837028e-06,
+      "loss": 0.45,
+      "step": 714
+    },
+    {
+      "epoch": 0.3953917050691244,
+      "grad_norm": 0.4377395808696747,
+      "learning_rate": 7.5693491121048194e-06,
+      "loss": 0.5144,
+      "step": 715
+    },
+    {
+      "epoch": 0.39594470046082947,
+      "grad_norm": 0.4012657403945923,
+      "learning_rate": 7.561061990767149e-06,
+      "loss": 0.4812,
+      "step": 716
+    },
+    {
+      "epoch": 0.3964976958525346,
+      "grad_norm": 0.41947153210639954,
+      "learning_rate": 7.552765320721865e-06,
+      "loss": 0.4407,
+      "step": 717
+    },
+    {
+      "epoch": 0.39705069124423964,
+      "grad_norm": 0.48914283514022827,
+      "learning_rate": 7.544459132902409e-06,
+      "loss": 0.4689,
+      "step": 718
+    },
+    {
+      "epoch": 0.3976036866359447,
+      "grad_norm": 0.3954436480998993,
+      "learning_rate": 7.5361434582777205e-06,
+      "loss": 0.4866,
+      "step": 719
+    },
+    {
+      "epoch": 0.39815668202764976,
+      "grad_norm": 0.45554065704345703,
+      "learning_rate": 7.527818327852101e-06,
+      "loss": 0.4499,
+      "step": 720
+    },
+    {
+      "epoch": 0.3987096774193548,
+      "grad_norm": 0.3696426451206207,
+      "learning_rate": 7.519483772665113e-06,
+      "loss": 0.4963,
+      "step": 721
+    },
+    {
+      "epoch": 0.3992626728110599,
+      "grad_norm": 0.42578935623168945,
+      "learning_rate": 7.511139823791452e-06,
+      "loss": 0.4642,
+      "step": 722
+    },
+    {
+      "epoch": 0.399815668202765,
+      "grad_norm": 0.3758449852466583,
+      "learning_rate": 7.502786512340841e-06,
+      "loss": 0.516,
+      "step": 723
+    },
+    {
+      "epoch": 0.40036866359447004,
+      "grad_norm": 0.4170108139514923,
+      "learning_rate": 7.494423869457912e-06,
+      "loss": 0.4824,
+      "step": 724
+    },
+    {
+      "epoch": 0.4009216589861751,
+      "grad_norm": 0.3795914053916931,
+      "learning_rate": 7.4860519263220865e-06,
+      "loss": 0.4744,
+      "step": 725
+    },
+    {
+      "epoch": 0.40147465437788016,
+      "grad_norm": 0.389635294675827,
+      "learning_rate": 7.477670714147461e-06,
+      "loss": 0.4957,
+      "step": 726
+    },
+    {
+      "epoch": 0.40202764976958527,
+      "grad_norm": 0.44072335958480835,
+      "learning_rate": 7.469280264182689e-06,
+      "loss": 0.5172,
+      "step": 727
+    },
+    {
+      "epoch": 0.40258064516129033,
+      "grad_norm": 0.372860312461853,
+      "learning_rate": 7.460880607710872e-06,
+      "loss": 0.4465,
+      "step": 728
+    },
+    {
+      "epoch": 0.4031336405529954,
+      "grad_norm": 0.47300985455513,
+      "learning_rate": 7.452471776049432e-06,
+      "loss": 0.4636,
+      "step": 729
+    },
+    {
+      "epoch": 0.40368663594470044,
+      "grad_norm": 0.44356971979141235,
+      "learning_rate": 7.444053800550004e-06,
+      "loss": 0.4911,
+      "step": 730
+    },
+    {
+      "epoch": 0.4042396313364055,
+      "grad_norm": 0.4640626311302185,
+      "learning_rate": 7.435626712598312e-06,
+      "loss": 0.4873,
+      "step": 731
+    },
+    {
+      "epoch": 0.4047926267281106,
+      "grad_norm": 0.4652840197086334,
+      "learning_rate": 7.427190543614053e-06,
+      "loss": 0.4641,
+      "step": 732
+    },
+    {
+      "epoch": 0.4053456221198157,
+      "grad_norm": 0.46554264426231384,
+      "learning_rate": 7.418745325050787e-06,
+      "loss": 0.4797,
+      "step": 733
+    },
+    {
+      "epoch": 0.40589861751152073,
+      "grad_norm": 0.42912858724594116,
+      "learning_rate": 7.410291088395812e-06,
+      "loss": 0.5064,
+      "step": 734
+    },
+    {
+      "epoch": 0.4064516129032258,
+      "grad_norm": 0.41365447640419006,
+      "learning_rate": 7.401827865170047e-06,
+      "loss": 0.4873,
+      "step": 735
+    },
+    {
+      "epoch": 0.4070046082949309,
+      "grad_norm": 0.5490487813949585,
+      "learning_rate": 7.3933556869279235e-06,
+      "loss": 0.4583,
+      "step": 736
+    },
+    {
+      "epoch": 0.40755760368663596,
+      "grad_norm": 0.39917993545532227,
+      "learning_rate": 7.384874585257255e-06,
+      "loss": 0.4625,
+      "step": 737
+    },
+    {
+      "epoch": 0.408110599078341,
+      "grad_norm": 0.4784587621688843,
+      "learning_rate": 7.3763845917791245e-06,
+      "loss": 0.5073,
+      "step": 738
+    },
+    {
+      "epoch": 0.4086635944700461,
+      "grad_norm": 0.4989032447338104,
+      "learning_rate": 7.367885738147773e-06,
+      "loss": 0.4652,
+      "step": 739
+    },
+    {
+      "epoch": 0.40921658986175113,
+      "grad_norm": 0.4629876911640167,
+      "learning_rate": 7.359378056050472e-06,
+      "loss": 0.4601,
+      "step": 740
+    },
+    {
+      "epoch": 0.40976958525345625,
+      "grad_norm": 0.4878072440624237,
+      "learning_rate": 7.35086157720741e-06,
+      "loss": 0.4713,
+      "step": 741
+    },
+    {
+      "epoch": 0.4103225806451613,
+      "grad_norm": 0.49611517786979675,
+      "learning_rate": 7.3423363333715726e-06,
+      "loss": 0.4727,
+      "step": 742
+    },
+    {
+      "epoch": 0.41087557603686636,
+      "grad_norm": 0.5412954688072205,
+      "learning_rate": 7.333802356328624e-06,
+      "loss": 0.4795,
+      "step": 743
+    },
+    {
+      "epoch": 0.4114285714285714,
+      "grad_norm": 0.387992262840271,
+      "learning_rate": 7.3252596778967954e-06,
+      "loss": 0.492,
+      "step": 744
+    },
+    {
+      "epoch": 0.4119815668202765,
+      "grad_norm": 0.47084909677505493,
+      "learning_rate": 7.316708329926754e-06,
+      "loss": 0.4546,
+      "step": 745
+    },
+    {
+      "epoch": 0.4125345622119816,
+      "grad_norm": 0.5451803803443909,
+      "learning_rate": 7.308148344301491e-06,
+      "loss": 0.5053,
+      "step": 746
+    },
+    {
+      "epoch": 0.41308755760368665,
+      "grad_norm": 0.4724251329898834,
+      "learning_rate": 7.2995797529362075e-06,
+      "loss": 0.4937,
+      "step": 747
+    },
+    {
+      "epoch": 0.4136405529953917,
+      "grad_norm": 0.48225685954093933,
+      "learning_rate": 7.291002587778184e-06,
+      "loss": 0.4939,
+      "step": 748
+    },
+    {
+      "epoch": 0.41419354838709677,
+      "grad_norm": 0.4945637881755829,
+      "learning_rate": 7.282416880806673e-06,
+      "loss": 0.4934,
+      "step": 749
+    },
+    {
+      "epoch": 0.4147465437788018,
+      "grad_norm": 0.48181742429733276,
+      "learning_rate": 7.273822664032771e-06,
+      "loss": 0.4763,
+      "step": 750
+    },
+    {
+      "epoch": 0.41529953917050694,
+      "grad_norm": 0.41117456555366516,
+      "learning_rate": 7.265219969499302e-06,
+      "loss": 0.4928,
+      "step": 751
+    },
+    {
+      "epoch": 0.415852534562212,
+      "grad_norm": 0.5093626976013184,
+      "learning_rate": 7.256608829280705e-06,
+      "loss": 0.4563,
+      "step": 752
+    },
+    {
+      "epoch": 0.41640552995391705,
+      "grad_norm": 0.45837345719337463,
+      "learning_rate": 7.2479892754828995e-06,
+      "loss": 0.4736,
+      "step": 753
+    },
+    {
+      "epoch": 0.4169585253456221,
+      "grad_norm": 0.4117204546928406,
+      "learning_rate": 7.239361340243179e-06,
+      "loss": 0.4788,
+      "step": 754
+    },
+    {
+      "epoch": 0.41751152073732717,
+      "grad_norm": 0.44084909558296204,
+      "learning_rate": 7.230725055730088e-06,
+      "loss": 0.4781,
+      "step": 755
+    },
+    {
+      "epoch": 0.4180645161290323,
+      "grad_norm": 0.5877032279968262,
+      "learning_rate": 7.222080454143296e-06,
+      "loss": 0.4782,
+      "step": 756
+    },
+    {
+      "epoch": 0.41861751152073734,
+      "grad_norm": 0.3853990137577057,
+      "learning_rate": 7.213427567713485e-06,
+      "loss": 0.4498,
+      "step": 757
+    },
+    {
+      "epoch": 0.4191705069124424,
+      "grad_norm": 0.4723830819129944,
+      "learning_rate": 7.204766428702226e-06,
+      "loss": 0.4601,
+      "step": 758
+    },
+    {
+      "epoch": 0.41972350230414746,
+      "grad_norm": 0.43980076909065247,
+      "learning_rate": 7.196097069401857e-06,
+      "loss": 0.4783,
+      "step": 759
+    },
+    {
+      "epoch": 0.4202764976958525,
+      "grad_norm": 0.4152413606643677,
+      "learning_rate": 7.1874195221353706e-06,
+      "loss": 0.4873,
+      "step": 760
+    },
+    {
+      "epoch": 0.4208294930875576,
+      "grad_norm": 0.3723086714744568,
+      "learning_rate": 7.17873381925628e-06,
+      "loss": 0.4664,
+      "step": 761
+    },
+    {
+      "epoch": 0.4213824884792627,
+      "grad_norm": 0.44318974018096924,
+      "learning_rate": 7.17003999314851e-06,
+      "loss": 0.4865,
+      "step": 762
+    },
+    {
+      "epoch": 0.42193548387096774,
+      "grad_norm": 0.40795907378196716,
+      "learning_rate": 7.161338076226272e-06,
+      "loss": 0.4795,
+      "step": 763
+    },
+    {
+      "epoch": 0.4224884792626728,
+      "grad_norm": 0.4073677659034729,
+      "learning_rate": 7.1526281009339426e-06,
+      "loss": 0.4815,
+      "step": 764
+    },
+    {
+      "epoch": 0.42304147465437786,
+      "grad_norm": 0.4583486020565033,
+      "learning_rate": 7.143910099745944e-06,
+      "loss": 0.4872,
+      "step": 765
+    },
+    {
+      "epoch": 0.42359447004608297,
+      "grad_norm": 0.4036937654018402,
+      "learning_rate": 7.13518410516662e-06,
+      "loss": 0.4616,
+      "step": 766
+    },
+    {
+      "epoch": 0.42414746543778803,
+      "grad_norm": 0.42568910121917725,
+      "learning_rate": 7.126450149730122e-06,
+      "loss": 0.489,
+      "step": 767
+    },
+    {
+      "epoch": 0.4247004608294931,
+      "grad_norm": 0.42301973700523376,
+      "learning_rate": 7.1177082660002784e-06,
+      "loss": 0.4685,
+      "step": 768
+    },
+    {
+      "epoch": 0.42525345622119815,
+      "grad_norm": 0.45005002617836,
+      "learning_rate": 7.108958486570479e-06,
+      "loss": 0.4555,
+      "step": 769
+    },
+    {
+      "epoch": 0.4258064516129032,
+      "grad_norm": 0.45795372128486633,
+      "learning_rate": 7.1002008440635515e-06,
+      "loss": 0.4678,
+      "step": 770
+    },
+    {
+      "epoch": 0.4263594470046083,
+      "grad_norm": 0.41580426692962646,
+      "learning_rate": 7.091435371131642e-06,
+      "loss": 0.5058,
+      "step": 771
+    },
+    {
+      "epoch": 0.4269124423963134,
+      "grad_norm": 0.4396505355834961,
+      "learning_rate": 7.082662100456089e-06,
+      "loss": 0.492,
+      "step": 772
+    },
+    {
+      "epoch": 0.42746543778801843,
+      "grad_norm": 0.4275103211402893,
+      "learning_rate": 7.073881064747308e-06,
+      "loss": 0.4624,
+      "step": 773
+    },
+    {
+      "epoch": 0.4280184331797235,
+      "grad_norm": 0.37376993894577026,
+      "learning_rate": 7.0650922967446634e-06,
+      "loss": 0.4626,
+      "step": 774
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.40532317757606506,
+      "learning_rate": 7.056295829216347e-06,
+      "loss": 0.5042,
+      "step": 775
+    },
+    {
+      "epoch": 0.42912442396313366,
+      "grad_norm": 0.39862802624702454,
+      "learning_rate": 7.047491694959263e-06,
+      "loss": 0.4398,
+      "step": 776
+    },
+    {
+      "epoch": 0.4296774193548387,
+      "grad_norm": 0.4326246380805969,
+      "learning_rate": 7.038679926798895e-06,
+      "loss": 0.4902,
+      "step": 777
+    },
+    {
+      "epoch": 0.4302304147465438,
+      "grad_norm": 0.43664461374282837,
+      "learning_rate": 7.029860557589191e-06,
+      "loss": 0.4817,
+      "step": 778
+    },
+    {
+      "epoch": 0.43078341013824883,
+      "grad_norm": 0.4556635320186615,
+      "learning_rate": 7.02103362021244e-06,
+      "loss": 0.4825,
+      "step": 779
+    },
+    {
+      "epoch": 0.4313364055299539,
+      "grad_norm": 0.40778669714927673,
+      "learning_rate": 7.012199147579146e-06,
+      "loss": 0.496,
+      "step": 780
+    },
+    {
+      "epoch": 0.431889400921659,
+      "grad_norm": 0.41946834325790405,
+      "learning_rate": 7.00335717262791e-06,
+      "loss": 0.482,
+      "step": 781
+    },
+    {
+      "epoch": 0.43244239631336406,
+      "grad_norm": 0.4511401653289795,
+      "learning_rate": 6.994507728325304e-06,
+      "loss": 0.4763,
+      "step": 782
+    },
+    {
+      "epoch": 0.4329953917050691,
+      "grad_norm": 0.43290090560913086,
+      "learning_rate": 6.985650847665747e-06,
+      "loss": 0.4804,
+      "step": 783
+    },
+    {
+      "epoch": 0.4335483870967742,
+      "grad_norm": 0.4241175949573517,
+      "learning_rate": 6.976786563671386e-06,
+      "loss": 0.4776,
+      "step": 784
+    },
+    {
+      "epoch": 0.43410138248847924,
+      "grad_norm": 0.4226067364215851,
+      "learning_rate": 6.967914909391971e-06,
+      "loss": 0.4839,
+      "step": 785
+    },
+    {
+      "epoch": 0.43465437788018435,
+      "grad_norm": 0.39866408705711365,
+      "learning_rate": 6.959035917904728e-06,
+      "loss": 0.4596,
+      "step": 786
+    },
+    {
+      "epoch": 0.4352073732718894,
+      "grad_norm": 0.37730705738067627,
+      "learning_rate": 6.950149622314245e-06,
+      "loss": 0.4784,
+      "step": 787
+    },
+    {
+      "epoch": 0.43576036866359447,
+      "grad_norm": 0.41297316551208496,
+      "learning_rate": 6.941256055752337e-06,
+      "loss": 0.4887,
+      "step": 788
+    },
+    {
+      "epoch": 0.4363133640552995,
+      "grad_norm": 0.4532250761985779,
+      "learning_rate": 6.932355251377932e-06,
+      "loss": 0.4897,
+      "step": 789
+    },
+    {
+      "epoch": 0.4368663594470046,
+      "grad_norm": 0.4239371418952942,
+      "learning_rate": 6.923447242376942e-06,
+      "loss": 0.4873,
+      "step": 790
+    },
+    {
+      "epoch": 0.4374193548387097,
+      "grad_norm": 0.33959221839904785,
+      "learning_rate": 6.91453206196214e-06,
+      "loss": 0.4697,
+      "step": 791
+    },
+    {
+      "epoch": 0.43797235023041475,
+      "grad_norm": 0.4459823668003082,
+      "learning_rate": 6.90560974337304e-06,
+      "loss": 0.4799,
+      "step": 792
+    },
+    {
+      "epoch": 0.4385253456221198,
+      "grad_norm": 0.4143071174621582,
+      "learning_rate": 6.896680319875767e-06,
+      "loss": 0.4556,
+      "step": 793
+    },
+    {
+      "epoch": 0.43907834101382487,
+      "grad_norm": 0.4591521620750427,
+      "learning_rate": 6.887743824762937e-06,
+      "loss": 0.4841,
+      "step": 794
+    },
+    {
+      "epoch": 0.4396313364055299,
+      "grad_norm": 0.41245412826538086,
+      "learning_rate": 6.878800291353533e-06,
+      "loss": 0.5083,
+      "step": 795
+    },
+    {
+      "epoch": 0.44018433179723504,
+      "grad_norm": 0.4207186996936798,
+      "learning_rate": 6.8698497529927784e-06,
+      "loss": 0.4653,
+      "step": 796
+    },
+    {
+      "epoch": 0.4407373271889401,
+      "grad_norm": 0.42010927200317383,
+      "learning_rate": 6.860892243052015e-06,
+      "loss": 0.4595,
+      "step": 797
+    },
+    {
+      "epoch": 0.44129032258064516,
+      "grad_norm": 0.41488680243492126,
+      "learning_rate": 6.851927794928578e-06,
+      "loss": 0.4685,
+      "step": 798
+    },
+    {
+      "epoch": 0.4418433179723502,
+      "grad_norm": 0.43417924642562866,
+      "learning_rate": 6.842956442045668e-06,
+      "loss": 0.4753,
+      "step": 799
+    },
+    {
+      "epoch": 0.4423963133640553,
+      "grad_norm": 0.35748791694641113,
+      "learning_rate": 6.833978217852233e-06,
+      "loss": 0.4677,
+      "step": 800
+    },
+    {
+      "epoch": 0.4429493087557604,
+      "grad_norm": 0.4993172883987427,
+      "learning_rate": 6.824993155822838e-06,
+      "loss": 0.4729,
+      "step": 801
+    },
+    {
+      "epoch": 0.44350230414746544,
+      "grad_norm": 0.36592617630958557,
+      "learning_rate": 6.816001289457542e-06,
+      "loss": 0.4565,
+      "step": 802
+    },
+    {
+      "epoch": 0.4440552995391705,
+      "grad_norm": 0.3910996615886688,
+      "learning_rate": 6.807002652281777e-06,
+      "loss": 0.4619,
+      "step": 803
+    },
+    {
+      "epoch": 0.44460829493087556,
+      "grad_norm": 0.41243669390678406,
+      "learning_rate": 6.797997277846215e-06,
+      "loss": 0.46,
+      "step": 804
+    },
+    {
+      "epoch": 0.44516129032258067,
+      "grad_norm": 0.3570329546928406,
+      "learning_rate": 6.788985199726648e-06,
+      "loss": 0.4624,
+      "step": 805
+    },
+    {
+      "epoch": 0.44571428571428573,
+      "grad_norm": 0.4660005271434784,
+      "learning_rate": 6.779966451523866e-06,
+      "loss": 0.494,
+      "step": 806
+    },
+    {
+      "epoch": 0.4462672811059908,
+      "grad_norm": 0.38170328736305237,
+      "learning_rate": 6.770941066863523e-06,
+      "loss": 0.4848,
+      "step": 807
+    },
+    {
+      "epoch": 0.44682027649769585,
+      "grad_norm": 0.410656601190567,
+      "learning_rate": 6.761909079396018e-06,
+      "loss": 0.4783,
+      "step": 808
+    },
+    {
+      "epoch": 0.4473732718894009,
+      "grad_norm": 0.39199960231781006,
+      "learning_rate": 6.752870522796372e-06,
+      "loss": 0.4706,
+      "step": 809
+    },
+    {
+      "epoch": 0.447926267281106,
+      "grad_norm": 0.4048680067062378,
+      "learning_rate": 6.743825430764091e-06,
+      "loss": 0.4841,
+      "step": 810
+    },
+    {
+      "epoch": 0.4484792626728111,
+      "grad_norm": 0.3699415922164917,
+      "learning_rate": 6.734773837023054e-06,
+      "loss": 0.4479,
+      "step": 811
+    },
+    {
+      "epoch": 0.44903225806451613,
+      "grad_norm": 0.3935936391353607,
+      "learning_rate": 6.725715775321379e-06,
+      "loss": 0.4802,
+      "step": 812
+    },
+    {
+      "epoch": 0.4495852534562212,
+      "grad_norm": 0.39852485060691833,
+      "learning_rate": 6.7166512794312986e-06,
+      "loss": 0.4652,
+      "step": 813
+    },
+    {
+      "epoch": 0.45013824884792625,
+      "grad_norm": 0.38947340846061707,
+      "learning_rate": 6.707580383149035e-06,
+      "loss": 0.4925,
+      "step": 814
+    },
+    {
+      "epoch": 0.45069124423963136,
+      "grad_norm": 0.4168023467063904,
+      "learning_rate": 6.698503120294674e-06,
+      "loss": 0.4829,
+      "step": 815
+    },
+    {
+      "epoch": 0.4512442396313364,
+      "grad_norm": 0.39369869232177734,
+      "learning_rate": 6.6894195247120396e-06,
+      "loss": 0.4562,
+      "step": 816
+    },
+    {
+      "epoch": 0.4517972350230415,
+      "grad_norm": 0.3529163599014282,
+      "learning_rate": 6.680329630268565e-06,
+      "loss": 0.4628,
+      "step": 817
+    },
+    {
+      "epoch": 0.45235023041474653,
+      "grad_norm": 0.3897452652454376,
+      "learning_rate": 6.671233470855171e-06,
+      "loss": 0.4635,
+      "step": 818
+    },
+    {
+      "epoch": 0.4529032258064516,
+      "grad_norm": 0.4104883074760437,
+      "learning_rate": 6.662131080386132e-06,
+      "loss": 0.4716,
+      "step": 819
+    },
+    {
+      "epoch": 0.4534562211981567,
+      "grad_norm": 0.39856767654418945,
+      "learning_rate": 6.653022492798959e-06,
+      "loss": 0.4666,
+      "step": 820
+    },
+    {
+      "epoch": 0.45400921658986176,
+      "grad_norm": 0.3940970301628113,
+      "learning_rate": 6.643907742054267e-06,
+      "loss": 0.4899,
+      "step": 821
+    },
+    {
+      "epoch": 0.4545622119815668,
+      "grad_norm": 0.4091474413871765,
+      "learning_rate": 6.63478686213565e-06,
+      "loss": 0.4829,
+      "step": 822
+    },
+    {
+      "epoch": 0.4551152073732719,
+      "grad_norm": 0.4372076690196991,
+      "learning_rate": 6.625659887049554e-06,
+      "loss": 0.4872,
+      "step": 823
+    },
+    {
+      "epoch": 0.45566820276497694,
+      "grad_norm": 0.4298078417778015,
+      "learning_rate": 6.6165268508251455e-06,
+      "loss": 0.4748,
+      "step": 824
+    },
+    {
+      "epoch": 0.45622119815668205,
+      "grad_norm": 0.4123404622077942,
+      "learning_rate": 6.607387787514199e-06,
+      "loss": 0.4784,
+      "step": 825
+    },
+    {
+      "epoch": 0.4567741935483871,
+      "grad_norm": 0.48749467730522156,
+      "learning_rate": 6.598242731190955e-06,
+      "loss": 0.4617,
+      "step": 826
+    },
+    {
+      "epoch": 0.45732718894009217,
+      "grad_norm": 0.44510942697525024,
+      "learning_rate": 6.589091715951996e-06,
+      "loss": 0.5012,
+      "step": 827
+    },
+    {
+      "epoch": 0.4578801843317972,
+      "grad_norm": 0.4342193603515625,
+      "learning_rate": 6.579934775916128e-06,
+      "loss": 0.4686,
+      "step": 828
+    },
+    {
+      "epoch": 0.4584331797235023,
+      "grad_norm": 0.46518048644065857,
+      "learning_rate": 6.570771945224243e-06,
+      "loss": 0.4695,
+      "step": 829
+    },
+    {
+      "epoch": 0.4589861751152074,
+      "grad_norm": 0.4383966326713562,
+      "learning_rate": 6.561603258039195e-06,
+      "loss": 0.4719,
+      "step": 830
+    },
+    {
+      "epoch": 0.45953917050691245,
+      "grad_norm": 0.41184401512145996,
+      "learning_rate": 6.552428748545677e-06,
+      "loss": 0.4789,
+      "step": 831
+    },
+    {
+      "epoch": 0.4600921658986175,
+      "grad_norm": 0.4380422830581665,
+      "learning_rate": 6.543248450950087e-06,
+      "loss": 0.4907,
+      "step": 832
+    },
+    {
+      "epoch": 0.46064516129032257,
+      "grad_norm": 0.453457236289978,
+      "learning_rate": 6.534062399480405e-06,
+      "loss": 0.4747,
+      "step": 833
+    },
+    {
+      "epoch": 0.4611981566820276,
+      "grad_norm": 0.36671724915504456,
+      "learning_rate": 6.524870628386064e-06,
+      "loss": 0.4905,
+      "step": 834
+    },
+    {
+      "epoch": 0.46175115207373274,
+      "grad_norm": 0.42992615699768066,
+      "learning_rate": 6.515673171937822e-06,
+      "loss": 0.4634,
+      "step": 835
+    },
+    {
+      "epoch": 0.4623041474654378,
+      "grad_norm": 0.40917208790779114,
+      "learning_rate": 6.5064700644276315e-06,
+      "loss": 0.4661,
+      "step": 836
+    },
+    {
+      "epoch": 0.46285714285714286,
+      "grad_norm": 0.4510115385055542,
+      "learning_rate": 6.497261340168519e-06,
+      "loss": 0.4504,
+      "step": 837
+    },
+    {
+      "epoch": 0.4634101382488479,
+      "grad_norm": 0.45596301555633545,
+      "learning_rate": 6.4880470334944515e-06,
+      "loss": 0.4551,
+      "step": 838
+    },
+    {
+      "epoch": 0.46396313364055297,
+      "grad_norm": 0.4473353624343872,
+      "learning_rate": 6.478827178760205e-06,
+      "loss": 0.4834,
+      "step": 839
+    },
+    {
+      "epoch": 0.4645161290322581,
+      "grad_norm": 0.4285813271999359,
+      "learning_rate": 6.469601810341247e-06,
+      "loss": 0.498,
+      "step": 840
+    },
+    {
+      "epoch": 0.46506912442396314,
+      "grad_norm": 0.38439249992370605,
+      "learning_rate": 6.4603709626336e-06,
+      "loss": 0.4754,
+      "step": 841
+    },
+    {
+      "epoch": 0.4656221198156682,
+      "grad_norm": 0.3741195499897003,
+      "learning_rate": 6.451134670053716e-06,
+      "loss": 0.4887,
+      "step": 842
+    },
+    {
+      "epoch": 0.46617511520737326,
+      "grad_norm": 0.4147772789001465,
+      "learning_rate": 6.441892967038346e-06,
+      "loss": 0.5023,
+      "step": 843
+    },
+    {
+      "epoch": 0.4667281105990783,
+      "grad_norm": 0.3958702087402344,
+      "learning_rate": 6.4326458880444155e-06,
+      "loss": 0.507,
+      "step": 844
+    },
+    {
+      "epoch": 0.46728110599078343,
+      "grad_norm": 0.4042072892189026,
+      "learning_rate": 6.423393467548893e-06,
+      "loss": 0.444,
+      "step": 845
+    },
+    {
+      "epoch": 0.4678341013824885,
+      "grad_norm": 0.387441486120224,
+      "learning_rate": 6.414135740048662e-06,
+      "loss": 0.4446,
+      "step": 846
+    },
+    {
+      "epoch": 0.46838709677419355,
+      "grad_norm": 0.4227030277252197,
+      "learning_rate": 6.404872740060392e-06,
+      "loss": 0.5104,
+      "step": 847
+    },
+    {
+      "epoch": 0.4689400921658986,
+      "grad_norm": 0.3976932764053345,
+      "learning_rate": 6.395604502120412e-06,
+      "loss": 0.4686,
+      "step": 848
+    },
+    {
+      "epoch": 0.46949308755760366,
+      "grad_norm": 0.41935819387435913,
+      "learning_rate": 6.386331060784583e-06,
+      "loss": 0.4802,
+      "step": 849
+    },
+    {
+      "epoch": 0.4700460829493088,
+      "grad_norm": 0.3910142183303833,
+      "learning_rate": 6.377052450628159e-06,
+      "loss": 0.465,
+      "step": 850
+    },
+    {
+      "epoch": 0.47059907834101383,
+      "grad_norm": 0.3708650469779968,
+      "learning_rate": 6.367768706245674e-06,
+      "loss": 0.4844,
+      "step": 851
+    },
+    {
+      "epoch": 0.4711520737327189,
+      "grad_norm": 0.39117151498794556,
+      "learning_rate": 6.358479862250796e-06,
+      "loss": 0.4826,
+      "step": 852
+    },
+    {
+      "epoch": 0.47170506912442395,
+      "grad_norm": 0.36844393610954285,
+      "learning_rate": 6.349185953276216e-06,
+      "loss": 0.4531,
+      "step": 853
+    },
+    {
+      "epoch": 0.472258064516129,
+      "grad_norm": 0.3579118549823761,
+      "learning_rate": 6.339887013973498e-06,
+      "loss": 0.4579,
+      "step": 854
+    },
+    {
+      "epoch": 0.4728110599078341,
+      "grad_norm": 0.4207668900489807,
+      "learning_rate": 6.330583079012972e-06,
+      "loss": 0.4769,
+      "step": 855
+    },
+    {
+      "epoch": 0.4733640552995392,
+      "grad_norm": 0.37506356835365295,
+      "learning_rate": 6.321274183083588e-06,
+      "loss": 0.4592,
+      "step": 856
+    },
+    {
+      "epoch": 0.47391705069124423,
+      "grad_norm": 0.4109135568141937,
+      "learning_rate": 6.3119603608927914e-06,
+      "loss": 0.4721,
+      "step": 857
+    },
+    {
+      "epoch": 0.4744700460829493,
+      "grad_norm": 0.4288492202758789,
+      "learning_rate": 6.302641647166402e-06,
+      "loss": 0.4766,
+      "step": 858
+    },
+    {
+      "epoch": 0.47502304147465435,
+      "grad_norm": 0.37833884358406067,
+      "learning_rate": 6.293318076648469e-06,
+      "loss": 0.4779,
+      "step": 859
+    },
+    {
+      "epoch": 0.47557603686635946,
+      "grad_norm": 0.4019952714443207,
+      "learning_rate": 6.283989684101155e-06,
+      "loss": 0.487,
+      "step": 860
+    },
+    {
+      "epoch": 0.4761290322580645,
+      "grad_norm": 0.45245492458343506,
+      "learning_rate": 6.274656504304598e-06,
+      "loss": 0.4582,
+      "step": 861
+    },
+    {
+      "epoch": 0.4766820276497696,
+      "grad_norm": 0.4037781357765198,
+      "learning_rate": 6.265318572056787e-06,
+      "loss": 0.4536,
+      "step": 862
+    },
+    {
+      "epoch": 0.47723502304147464,
+      "grad_norm": 0.4167582392692566,
+      "learning_rate": 6.25597592217343e-06,
+      "loss": 0.4737,
+      "step": 863
+    },
+    {
+      "epoch": 0.47778801843317975,
+      "grad_norm": 0.38921552896499634,
+      "learning_rate": 6.2466285894878206e-06,
+      "loss": 0.4889,
+      "step": 864
+    },
+    {
+      "epoch": 0.4783410138248848,
+      "grad_norm": 0.3599966764450073,
+      "learning_rate": 6.2372766088507195e-06,
+      "loss": 0.4421,
+      "step": 865
+    },
+    {
+      "epoch": 0.47889400921658987,
+      "grad_norm": 0.440920889377594,
+      "learning_rate": 6.227920015130209e-06,
+      "loss": 0.4959,
+      "step": 866
+    },
+    {
+      "epoch": 0.4794470046082949,
+      "grad_norm": 0.39445987343788147,
+      "learning_rate": 6.218558843211576e-06,
+      "loss": 0.4784,
+      "step": 867
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4886539578437805,
+      "learning_rate": 6.209193127997173e-06,
+      "loss": 0.4702,
+      "step": 868
+    },
+    {
+      "epoch": 0.4805529953917051,
+      "grad_norm": 0.43818390369415283,
+      "learning_rate": 6.199822904406295e-06,
+      "loss": 0.4685,
+      "step": 869
+    },
+    {
+      "epoch": 0.48110599078341015,
+      "grad_norm": 0.4038926064968109,
+      "learning_rate": 6.190448207375046e-06,
+      "loss": 0.4572,
+      "step": 870
+    },
+    {
+      "epoch": 0.4816589861751152,
+      "grad_norm": 0.5002937912940979,
+      "learning_rate": 6.1810690718562065e-06,
+      "loss": 0.4899,
+      "step": 871
+    },
+    {
+      "epoch": 0.48221198156682027,
+      "grad_norm": 0.39266350865364075,
+      "learning_rate": 6.1716855328191064e-06,
+      "loss": 0.4682,
+      "step": 872
+    },
+    {
+      "epoch": 0.4827649769585253,
+      "grad_norm": 0.4712425470352173,
+      "learning_rate": 6.162297625249494e-06,
+      "loss": 0.473,
+      "step": 873
+    },
+    {
+      "epoch": 0.48331797235023044,
+      "grad_norm": 0.33758747577667236,
+      "learning_rate": 6.1529053841494065e-06,
+      "loss": 0.4771,
+      "step": 874
+    },
+    {
+      "epoch": 0.4838709677419355,
+      "grad_norm": 0.442381888628006,
+      "learning_rate": 6.143508844537038e-06,
+      "loss": 0.4834,
+      "step": 875
+    },
+    {
+      "epoch": 0.48442396313364056,
+      "grad_norm": 0.43109333515167236,
+      "learning_rate": 6.134108041446609e-06,
+      "loss": 0.4717,
+      "step": 876
+    },
+    {
+      "epoch": 0.4849769585253456,
+      "grad_norm": 0.4240846037864685,
+      "learning_rate": 6.124703009928233e-06,
+      "loss": 0.491,
+      "step": 877
+    },
+    {
+      "epoch": 0.48552995391705067,
+      "grad_norm": 0.41452300548553467,
+      "learning_rate": 6.115293785047793e-06,
+      "loss": 0.4789,
+      "step": 878
+    },
+    {
+      "epoch": 0.4860829493087558,
+      "grad_norm": 0.397856742143631,
+      "learning_rate": 6.105880401886804e-06,
+      "loss": 0.4655,
+      "step": 879
+    },
+    {
+      "epoch": 0.48663594470046084,
+      "grad_norm": 0.3984890878200531,
+      "learning_rate": 6.096462895542288e-06,
+      "loss": 0.4864,
+      "step": 880
+    },
+    {
+      "epoch": 0.4871889400921659,
+      "grad_norm": 0.4527179002761841,
+      "learning_rate": 6.087041301126636e-06,
+      "loss": 0.4884,
+      "step": 881
+    },
+    {
+      "epoch": 0.48774193548387096,
+      "grad_norm": 0.4393288791179657,
+      "learning_rate": 6.077615653767484e-06,
+      "loss": 0.45,
+      "step": 882
+    },
+    {
+      "epoch": 0.488294930875576,
+      "grad_norm": 0.4285862445831299,
+      "learning_rate": 6.068185988607575e-06,
+      "loss": 0.4858,
+      "step": 883
+    },
+    {
+      "epoch": 0.48884792626728113,
+      "grad_norm": 0.4243047833442688,
+      "learning_rate": 6.058752340804639e-06,
+      "loss": 0.4801,
+      "step": 884
+    },
+    {
+      "epoch": 0.4894009216589862,
+      "grad_norm": 0.37787267565727234,
+      "learning_rate": 6.0493147455312475e-06,
+      "loss": 0.4584,
+      "step": 885
+    },
+    {
+      "epoch": 0.48995391705069125,
+      "grad_norm": 0.4291735589504242,
+      "learning_rate": 6.039873237974695e-06,
+      "loss": 0.5038,
+      "step": 886
+    },
+    {
+      "epoch": 0.4905069124423963,
+      "grad_norm": 0.3942468464374542,
+      "learning_rate": 6.03042785333686e-06,
+      "loss": 0.4805,
+      "step": 887
+    },
+    {
+      "epoch": 0.49105990783410136,
+      "grad_norm": 0.42691728472709656,
+      "learning_rate": 6.020978626834077e-06,
+      "loss": 0.4587,
+      "step": 888
+    },
+    {
+      "epoch": 0.4916129032258065,
+      "grad_norm": 0.3891282379627228,
+      "learning_rate": 6.0115255936970056e-06,
+      "loss": 0.4756,
+      "step": 889
+    },
+    {
+      "epoch": 0.49216589861751153,
+      "grad_norm": 0.39589089155197144,
+      "learning_rate": 6.002068789170497e-06,
+      "loss": 0.4796,
+      "step": 890
+    },
+    {
+      "epoch": 0.4927188940092166,
+      "grad_norm": 0.4499064087867737,
+      "learning_rate": 5.992608248513465e-06,
+      "loss": 0.4667,
+      "step": 891
+    },
+    {
+      "epoch": 0.49327188940092165,
+      "grad_norm": 0.4281262457370758,
+      "learning_rate": 5.983144006998751e-06,
+      "loss": 0.4762,
+      "step": 892
+    },
+    {
+      "epoch": 0.4938248847926267,
+      "grad_norm": 0.4095862805843353,
+      "learning_rate": 5.973676099912996e-06,
+      "loss": 0.4691,
+      "step": 893
+    },
+    {
+      "epoch": 0.4943778801843318,
+      "grad_norm": 0.4401033818721771,
+      "learning_rate": 5.964204562556508e-06,
+      "loss": 0.4688,
+      "step": 894
+    },
+    {
+      "epoch": 0.4949308755760369,
+      "grad_norm": 0.41105031967163086,
+      "learning_rate": 5.954729430243129e-06,
+      "loss": 0.4877,
+      "step": 895
+    },
+    {
+      "epoch": 0.49548387096774194,
+      "grad_norm": 0.39041486382484436,
+      "learning_rate": 5.945250738300108e-06,
+      "loss": 0.4504,
+      "step": 896
+    },
+    {
+      "epoch": 0.496036866359447,
+      "grad_norm": 0.4423321485519409,
+      "learning_rate": 5.935768522067962e-06,
+      "loss": 0.4858,
+      "step": 897
+    },
+    {
+      "epoch": 0.49658986175115205,
+      "grad_norm": 0.4127327799797058,
+      "learning_rate": 5.9262828169003476e-06,
+      "loss": 0.4767,
+      "step": 898
+    },
+    {
+      "epoch": 0.49714285714285716,
+      "grad_norm": 0.41609784960746765,
+      "learning_rate": 5.9167936581639325e-06,
+      "loss": 0.4901,
+      "step": 899
+    },
+    {
+      "epoch": 0.4976958525345622,
+      "grad_norm": 0.3699811100959778,
+      "learning_rate": 5.9073010812382595e-06,
+      "loss": 0.4595,
+      "step": 900
+    },
+    {
+      "epoch": 0.4982488479262673,
+      "grad_norm": 0.45880934596061707,
+      "learning_rate": 5.897805121515616e-06,
+      "loss": 0.4514,
+      "step": 901
+    },
+    {
+      "epoch": 0.49880184331797234,
+      "grad_norm": 0.38559094071388245,
+      "learning_rate": 5.888305814400901e-06,
+      "loss": 0.4871,
+      "step": 902
+    },
+    {
+      "epoch": 0.4993548387096774,
+      "grad_norm": 0.40912917256355286,
+      "learning_rate": 5.878803195311496e-06,
+      "loss": 0.4726,
+      "step": 903
+    },
+    {
+      "epoch": 0.4999078341013825,
+      "grad_norm": 0.4119787812232971,
+      "learning_rate": 5.869297299677128e-06,
+      "loss": 0.459,
+      "step": 904
+    },
+    {
+      "epoch": 0.5004608294930876,
+      "grad_norm": 0.4300439953804016,
+      "learning_rate": 5.8597881629397435e-06,
+      "loss": 0.4764,
+      "step": 905
+    },
+    {
+      "epoch": 0.5010138248847926,
+      "grad_norm": 0.4216427803039551,
+      "learning_rate": 5.850275820553368e-06,
+      "loss": 0.4795,
+      "step": 906
+    },
+    {
+      "epoch": 0.5015668202764977,
+      "grad_norm": 0.578460693359375,
+      "learning_rate": 5.840760307983988e-06,
+      "loss": 0.4463,
+      "step": 907
+    },
+    {
+      "epoch": 0.5021198156682027,
+      "grad_norm": 0.38813382387161255,
+      "learning_rate": 5.831241660709402e-06,
+      "loss": 0.479,
+      "step": 908
+    },
+    {
+      "epoch": 0.5026728110599078,
+      "grad_norm": 0.4912916123867035,
+      "learning_rate": 5.8217199142190975e-06,
+      "loss": 0.4846,
+      "step": 909
+    },
+    {
+      "epoch": 0.5032258064516129,
+      "grad_norm": 0.4877387285232544,
+      "learning_rate": 5.812195104014119e-06,
+      "loss": 0.5051,
+      "step": 910
+    },
+    {
+      "epoch": 0.503778801843318,
+      "grad_norm": 0.4228856861591339,
+      "learning_rate": 5.802667265606933e-06,
+      "loss": 0.4853,
+      "step": 911
+    },
+    {
+      "epoch": 0.5043317972350231,
+      "grad_norm": 0.4995596706867218,
+      "learning_rate": 5.793136434521296e-06,
+      "loss": 0.4603,
+      "step": 912
+    },
+    {
+      "epoch": 0.5048847926267281,
+      "grad_norm": 0.5033881664276123,
+      "learning_rate": 5.783602646292123e-06,
+      "loss": 0.4793,
+      "step": 913
+    },
+    {
+      "epoch": 0.5054377880184332,
+      "grad_norm": 0.41643261909484863,
+      "learning_rate": 5.774065936465352e-06,
+      "loss": 0.4744,
+      "step": 914
+    },
+    {
+      "epoch": 0.5059907834101383,
+      "grad_norm": 0.4517223834991455,
+      "learning_rate": 5.764526340597818e-06,
+      "loss": 0.4659,
+      "step": 915
+    },
+    {
+      "epoch": 0.5065437788018433,
+      "grad_norm": 0.3723732531070709,
+      "learning_rate": 5.754983894257114e-06,
+      "loss": 0.4727,
+      "step": 916
+    },
+    {
+      "epoch": 0.5070967741935484,
+      "grad_norm": 0.46375709772109985,
+      "learning_rate": 5.745438633021462e-06,
+      "loss": 0.4977,
+      "step": 917
+    },
+    {
+      "epoch": 0.5076497695852534,
+      "grad_norm": 0.42543715238571167,
+      "learning_rate": 5.7358905924795775e-06,
+      "loss": 0.4931,
+      "step": 918
+    },
+    {
+      "epoch": 0.5082027649769585,
+      "grad_norm": 0.3639947772026062,
+      "learning_rate": 5.726339808230539e-06,
+      "loss": 0.4608,
+      "step": 919
+    },
+    {
+      "epoch": 0.5087557603686635,
+      "grad_norm": 0.47503146529197693,
+      "learning_rate": 5.716786315883657e-06,
+      "loss": 0.4746,
+      "step": 920
+    },
+    {
+      "epoch": 0.5093087557603687,
+      "grad_norm": 0.4439897835254669,
+      "learning_rate": 5.707230151058334e-06,
+      "loss": 0.47,
+      "step": 921
+    },
+    {
+      "epoch": 0.5098617511520738,
+      "grad_norm": 0.43655335903167725,
+      "learning_rate": 5.697671349383939e-06,
+      "loss": 0.4668,
+      "step": 922
+    },
+    {
+      "epoch": 0.5104147465437788,
+      "grad_norm": 0.4014508128166199,
+      "learning_rate": 5.688109946499676e-06,
+      "loss": 0.4537,
+      "step": 923
+    },
+    {
+      "epoch": 0.5109677419354839,
+      "grad_norm": 0.4049840271472931,
+      "learning_rate": 5.678545978054443e-06,
+      "loss": 0.4713,
+      "step": 924
+    },
+    {
+      "epoch": 0.511520737327189,
+      "grad_norm": 0.4141314923763275,
+      "learning_rate": 5.668979479706703e-06,
+      "loss": 0.4882,
+      "step": 925
+    },
+    {
+      "epoch": 0.512073732718894,
+      "grad_norm": 0.4375897943973541,
+      "learning_rate": 5.659410487124355e-06,
+      "loss": 0.4625,
+      "step": 926
+    },
+    {
+      "epoch": 0.5126267281105991,
+      "grad_norm": 0.3951698839664459,
+      "learning_rate": 5.649839035984597e-06,
+      "loss": 0.4841,
+      "step": 927
+    },
+    {
+      "epoch": 0.5131797235023041,
+      "grad_norm": 0.367379367351532,
+      "learning_rate": 5.6402651619737865e-06,
+      "loss": 0.4818,
+      "step": 928
+    },
+    {
+      "epoch": 0.5137327188940092,
+      "grad_norm": 0.3982902765274048,
+      "learning_rate": 5.630688900787326e-06,
+      "loss": 0.4835,
+      "step": 929
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 0.41130173206329346,
+      "learning_rate": 5.621110288129509e-06,
+      "loss": 0.4592,
+      "step": 930
+    },
+    {
+      "epoch": 0.5148387096774194,
+      "grad_norm": 0.4003121554851532,
+      "learning_rate": 5.6115293597134015e-06,
+      "loss": 0.4511,
+      "step": 931
+    },
+    {
+      "epoch": 0.5153917050691245,
+      "grad_norm": 0.3777535855770111,
+      "learning_rate": 5.601946151260702e-06,
+      "loss": 0.4943,
+      "step": 932
+    },
+    {
+      "epoch": 0.5159447004608295,
+      "grad_norm": 0.46429774165153503,
+      "learning_rate": 5.592360698501609e-06,
+      "loss": 0.4709,
+      "step": 933
+    },
+    {
+      "epoch": 0.5164976958525346,
+      "grad_norm": 0.38206857442855835,
+      "learning_rate": 5.58277303717469e-06,
+      "loss": 0.4817,
+      "step": 934
+    },
+    {
+      "epoch": 0.5170506912442396,
+      "grad_norm": 0.39985257387161255,
+      "learning_rate": 5.573183203026747e-06,
+      "loss": 0.4646,
+      "step": 935
+    },
+    {
+      "epoch": 0.5176036866359447,
+      "grad_norm": 0.4025905728340149,
+      "learning_rate": 5.563591231812683e-06,
+      "loss": 0.471,
+      "step": 936
+    },
+    {
+      "epoch": 0.5181566820276498,
+      "grad_norm": 0.41740676760673523,
+      "learning_rate": 5.553997159295366e-06,
+      "loss": 0.4798,
+      "step": 937
+    },
+    {
+      "epoch": 0.5187096774193548,
+      "grad_norm": 0.395632266998291,
+      "learning_rate": 5.544401021245505e-06,
+      "loss": 0.4905,
+      "step": 938
+    },
+    {
+      "epoch": 0.5192626728110599,
+      "grad_norm": 0.3581371605396271,
+      "learning_rate": 5.534802853441503e-06,
+      "loss": 0.4563,
+      "step": 939
+    },
+    {
+      "epoch": 0.5198156682027649,
+      "grad_norm": 0.3925972282886505,
+      "learning_rate": 5.525202691669335e-06,
+      "loss": 0.4553,
+      "step": 940
+    },
+    {
+      "epoch": 0.5203686635944701,
+      "grad_norm": 0.3491620421409607,
+      "learning_rate": 5.515600571722412e-06,
+      "loss": 0.4602,
+      "step": 941
+    },
+    {
+      "epoch": 0.5209216589861752,
+      "grad_norm": 0.37922897934913635,
+      "learning_rate": 5.505996529401442e-06,
+      "loss": 0.4588,
+      "step": 942
+    },
+    {
+      "epoch": 0.5214746543778802,
+      "grad_norm": 0.37720414996147156,
+      "learning_rate": 5.496390600514298e-06,
+      "loss": 0.4688,
+      "step": 943
+    },
+    {
+      "epoch": 0.5220276497695853,
+      "grad_norm": 0.38247910141944885,
+      "learning_rate": 5.486782820875895e-06,
+      "loss": 0.4619,
+      "step": 944
+    },
+    {
+      "epoch": 0.5225806451612903,
+      "grad_norm": 0.3531082570552826,
+      "learning_rate": 5.477173226308042e-06,
+      "loss": 0.4611,
+      "step": 945
+    },
+    {
+      "epoch": 0.5231336405529954,
+      "grad_norm": 0.35197713971138,
+      "learning_rate": 5.4675618526393185e-06,
+      "loss": 0.483,
+      "step": 946
+    },
+    {
+      "epoch": 0.5236866359447004,
+      "grad_norm": 0.42967379093170166,
+      "learning_rate": 5.457948735704933e-06,
+      "loss": 0.469,
+      "step": 947
+    },
+    {
+      "epoch": 0.5242396313364055,
+      "grad_norm": 0.41465815901756287,
+      "learning_rate": 5.448333911346598e-06,
+      "loss": 0.4792,
+      "step": 948
+    },
+    {
+      "epoch": 0.5247926267281106,
+      "grad_norm": 0.385215163230896,
+      "learning_rate": 5.43871741541239e-06,
+      "loss": 0.4789,
+      "step": 949
+    },
+    {
+      "epoch": 0.5253456221198156,
+      "grad_norm": 0.39359477162361145,
+      "learning_rate": 5.429099283756618e-06,
+      "loss": 0.4356,
+      "step": 950
+    },
+    {
+      "epoch": 0.5258986175115208,
+      "grad_norm": 0.37385135889053345,
+      "learning_rate": 5.419479552239689e-06,
+      "loss": 0.4713,
+      "step": 951
+    },
+    {
+      "epoch": 0.5264516129032258,
+      "grad_norm": 0.40417802333831787,
+      "learning_rate": 5.4098582567279755e-06,
+      "loss": 0.4714,
+      "step": 952
+    },
+    {
+      "epoch": 0.5270046082949309,
+      "grad_norm": 0.415487140417099,
+      "learning_rate": 5.400235433093682e-06,
+      "loss": 0.4696,
+      "step": 953
+    },
+    {
+      "epoch": 0.527557603686636,
+      "grad_norm": 0.40797826647758484,
+      "learning_rate": 5.39061111721471e-06,
+      "loss": 0.4861,
+      "step": 954
+    },
+    {
+      "epoch": 0.528110599078341,
+      "grad_norm": 0.4314478039741516,
+      "learning_rate": 5.380985344974524e-06,
+      "loss": 0.4885,
+      "step": 955
+    },
+    {
+      "epoch": 0.5286635944700461,
+      "grad_norm": 0.3722938895225525,
+      "learning_rate": 5.371358152262019e-06,
+      "loss": 0.4576,
+      "step": 956
+    },
+    {
+      "epoch": 0.5292165898617511,
+      "grad_norm": 0.4123469293117523,
+      "learning_rate": 5.361729574971387e-06,
+      "loss": 0.4657,
+      "step": 957
+    },
+    {
+      "epoch": 0.5297695852534562,
+      "grad_norm": 0.4114021956920624,
+      "learning_rate": 5.352099649001979e-06,
+      "loss": 0.4705,
+      "step": 958
+    },
+    {
+      "epoch": 0.5303225806451612,
+      "grad_norm": 0.38689500093460083,
+      "learning_rate": 5.3424684102581795e-06,
+      "loss": 0.4661,
+      "step": 959
+    },
+    {
+      "epoch": 0.5308755760368664,
+      "grad_norm": 0.4139274060726166,
+      "learning_rate": 5.3328358946492634e-06,
+      "loss": 0.4721,
+      "step": 960
+    },
+    {
+      "epoch": 0.5314285714285715,
+      "grad_norm": 0.3889627754688263,
+      "learning_rate": 5.323202138089266e-06,
+      "loss": 0.4632,
+      "step": 961
+    },
+    {
+      "epoch": 0.5319815668202765,
+      "grad_norm": 0.3822690546512604,
+      "learning_rate": 5.313567176496854e-06,
+      "loss": 0.4522,
+      "step": 962
+    },
+    {
+      "epoch": 0.5325345622119816,
+      "grad_norm": 0.4444742798805237,
+      "learning_rate": 5.303931045795184e-06,
+      "loss": 0.4627,
+      "step": 963
+    },
+    {
+      "epoch": 0.5330875576036866,
+      "grad_norm": 0.4218832552433014,
+      "learning_rate": 5.294293781911769e-06,
+      "loss": 0.4762,
+      "step": 964
+    },
+    {
+      "epoch": 0.5336405529953917,
+      "grad_norm": 0.40190473198890686,
+      "learning_rate": 5.284655420778353e-06,
+      "loss": 0.4399,
+      "step": 965
+    },
+    {
+      "epoch": 0.5341935483870968,
+      "grad_norm": 0.44916805624961853,
+      "learning_rate": 5.275015998330765e-06,
+      "loss": 0.4457,
+      "step": 966
+    },
+    {
+      "epoch": 0.5347465437788018,
+      "grad_norm": 0.4550216794013977,
+      "learning_rate": 5.265375550508793e-06,
+      "loss": 0.4719,
+      "step": 967
+    },
+    {
+      "epoch": 0.5352995391705069,
+      "grad_norm": 0.42324307560920715,
+      "learning_rate": 5.255734113256051e-06,
+      "loss": 0.46,
+      "step": 968
+    },
+    {
+      "epoch": 0.5358525345622119,
+      "grad_norm": 0.3758416473865509,
+      "learning_rate": 5.246091722519839e-06,
+      "loss": 0.4441,
+      "step": 969
+    },
+    {
+      "epoch": 0.5364055299539171,
+      "grad_norm": 0.37716051936149597,
+      "learning_rate": 5.236448414251012e-06,
+      "loss": 0.4643,
+      "step": 970
+    },
+    {
+      "epoch": 0.5369585253456222,
+      "grad_norm": 0.3788035809993744,
+      "learning_rate": 5.226804224403846e-06,
+      "loss": 0.4592,
+      "step": 971
+    },
+    {
+      "epoch": 0.5375115207373272,
+      "grad_norm": 0.38398197293281555,
+      "learning_rate": 5.217159188935907e-06,
+      "loss": 0.4744,
+      "step": 972
+    },
+    {
+      "epoch": 0.5380645161290323,
+      "grad_norm": 0.3767092823982239,
+      "learning_rate": 5.207513343807907e-06,
+      "loss": 0.4698,
+      "step": 973
+    },
+    {
+      "epoch": 0.5386175115207373,
+      "grad_norm": 0.39843782782554626,
+      "learning_rate": 5.1978667249835855e-06,
+      "loss": 0.4803,
+      "step": 974
+    },
+    {
+      "epoch": 0.5391705069124424,
+      "grad_norm": 0.38011598587036133,
+      "learning_rate": 5.18821936842956e-06,
+      "loss": 0.4398,
+      "step": 975
+    },
+    {
+      "epoch": 0.5397235023041475,
+      "grad_norm": 0.36936551332473755,
+      "learning_rate": 5.1785713101152e-06,
+      "loss": 0.4662,
+      "step": 976
+    },
+    {
+      "epoch": 0.5402764976958525,
+      "grad_norm": 0.3832260072231293,
+      "learning_rate": 5.168922586012495e-06,
+      "loss": 0.488,
+      "step": 977
+    },
+    {
+      "epoch": 0.5408294930875576,
+      "grad_norm": 0.38859879970550537,
+      "learning_rate": 5.159273232095911e-06,
+      "loss": 0.4982,
+      "step": 978
+    },
+    {
+      "epoch": 0.5413824884792626,
+      "grad_norm": 0.4032990038394928,
+      "learning_rate": 5.1496232843422665e-06,
+      "loss": 0.4919,
+      "step": 979
+    },
+    {
+      "epoch": 0.5419354838709678,
+      "grad_norm": 0.3422912061214447,
+      "learning_rate": 5.139972778730593e-06,
+      "loss": 0.4658,
+      "step": 980
+    },
+    {
+      "epoch": 0.5424884792626729,
+      "grad_norm": 0.38321739435195923,
+      "learning_rate": 5.130321751242002e-06,
+      "loss": 0.4738,
+      "step": 981
+    },
+    {
+      "epoch": 0.5430414746543779,
+      "grad_norm": 0.3506300151348114,
+      "learning_rate": 5.12067023785955e-06,
+      "loss": 0.4612,
+      "step": 982
+    },
+    {
+      "epoch": 0.543594470046083,
+      "grad_norm": 0.3691524565219879,
+      "learning_rate": 5.111018274568106e-06,
+      "loss": 0.4461,
+      "step": 983
+    },
+    {
+      "epoch": 0.544147465437788,
+      "grad_norm": 0.37267613410949707,
+      "learning_rate": 5.101365897354215e-06,
+      "loss": 0.4891,
+      "step": 984
+    },
+    {
+      "epoch": 0.5447004608294931,
+      "grad_norm": 0.3920471668243408,
+      "learning_rate": 5.091713142205967e-06,
+      "loss": 0.4645,
+      "step": 985
+    },
+    {
+      "epoch": 0.5452534562211981,
+      "grad_norm": 0.37700968980789185,
+      "learning_rate": 5.082060045112862e-06,
+      "loss": 0.4721,
+      "step": 986
+    },
+    {
+      "epoch": 0.5458064516129032,
+      "grad_norm": 0.4021558165550232,
+      "learning_rate": 5.072406642065671e-06,
+      "loss": 0.4825,
+      "step": 987
+    },
+    {
+      "epoch": 0.5463594470046083,
+      "grad_norm": 0.40953153371810913,
+      "learning_rate": 5.062752969056311e-06,
+      "loss": 0.4478,
+      "step": 988
+    },
+    {
+      "epoch": 0.5469124423963133,
+      "grad_norm": 0.4201180934906006,
+      "learning_rate": 5.0530990620777e-06,
+      "loss": 0.4405,
+      "step": 989
+    },
+    {
+      "epoch": 0.5474654377880185,
+      "grad_norm": 0.42839857935905457,
+      "learning_rate": 5.0434449571236314e-06,
+      "loss": 0.4774,
+      "step": 990
+    },
+    {
+      "epoch": 0.5480184331797235,
+      "grad_norm": 0.40306076407432556,
+      "learning_rate": 5.033790690188637e-06,
+      "loss": 0.475,
+      "step": 991
+    },
+    {
+      "epoch": 0.5485714285714286,
+      "grad_norm": 0.3931422829627991,
+      "learning_rate": 5.024136297267851e-06,
+      "loss": 0.4813,
+      "step": 992
+    },
+    {
+      "epoch": 0.5491244239631337,
+      "grad_norm": 0.41347619891166687,
+      "learning_rate": 5.01448181435688e-06,
+      "loss": 0.4726,
+      "step": 993
+    },
+    {
+      "epoch": 0.5496774193548387,
+      "grad_norm": 0.41298919916152954,
+      "learning_rate": 5.004827277451661e-06,
+      "loss": 0.4581,
+      "step": 994
+    },
+    {
+      "epoch": 0.5502304147465438,
+      "grad_norm": 0.3852965235710144,
+      "learning_rate": 4.9951727225483395e-06,
+      "loss": 0.4638,
+      "step": 995
+    },
+    {
+      "epoch": 0.5507834101382488,
+      "grad_norm": 0.37419193983078003,
+      "learning_rate": 4.985518185643122e-06,
+      "loss": 0.4675,
+      "step": 996
+    },
+    {
+      "epoch": 0.5513364055299539,
+      "grad_norm": 0.36649253964424133,
+      "learning_rate": 4.975863702732151e-06,
+      "loss": 0.4584,
+      "step": 997
+    },
+    {
+      "epoch": 0.551889400921659,
+      "grad_norm": 0.4342942237854004,
+      "learning_rate": 4.966209309811364e-06,
+      "loss": 0.4731,
+      "step": 998
+    },
+    {
+      "epoch": 0.552442396313364,
+      "grad_norm": 0.37662389874458313,
+      "learning_rate": 4.95655504287637e-06,
+      "loss": 0.4696,
+      "step": 999
+    },
+    {
+      "epoch": 0.5529953917050692,
+      "grad_norm": 0.40196794271469116,
+      "learning_rate": 4.946900937922302e-06,
+      "loss": 0.4496,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5535483870967742,
+      "grad_norm": 0.3595891296863556,
+      "learning_rate": 4.93724703094369e-06,
+      "loss": 0.4621,
+      "step": 1001
+    },
+    {
+      "epoch": 0.5541013824884793,
+      "grad_norm": 0.4029429256916046,
+      "learning_rate": 4.927593357934329e-06,
+      "loss": 0.4885,
+      "step": 1002
+    },
+    {
+      "epoch": 0.5546543778801843,
+      "grad_norm": 0.37034833431243896,
+      "learning_rate": 4.917939954887139e-06,
+      "loss": 0.462,
+      "step": 1003
+    },
+    {
+      "epoch": 0.5552073732718894,
+      "grad_norm": 0.4130437672138214,
+      "learning_rate": 4.908286857794034e-06,
+      "loss": 0.4646,
+      "step": 1004
+    },
+    {
+      "epoch": 0.5557603686635945,
+      "grad_norm": 0.39467036724090576,
+      "learning_rate": 4.8986341026457875e-06,
+      "loss": 0.4557,
+      "step": 1005
+    },
+    {
+      "epoch": 0.5563133640552995,
+      "grad_norm": 0.37382972240448,
+      "learning_rate": 4.888981725431897e-06,
+      "loss": 0.4589,
+      "step": 1006
+    },
+    {
+      "epoch": 0.5568663594470046,
+      "grad_norm": 0.3812313377857208,
+      "learning_rate": 4.879329762140453e-06,
+      "loss": 0.476,
+      "step": 1007
+    },
+    {
+      "epoch": 0.5574193548387096,
+      "grad_norm": 0.3706028461456299,
+      "learning_rate": 4.8696782487580005e-06,
+      "loss": 0.4534,
+      "step": 1008
+    },
+    {
+      "epoch": 0.5579723502304147,
+      "grad_norm": 0.40674856305122375,
+      "learning_rate": 4.8600272212694094e-06,
+      "loss": 0.4555,
+      "step": 1009
+    },
+    {
+      "epoch": 0.5585253456221199,
+      "grad_norm": 0.3949076235294342,
+      "learning_rate": 4.850376715657736e-06,
+      "loss": 0.4569,
+      "step": 1010
+    },
+    {
+      "epoch": 0.5590783410138249,
+      "grad_norm": 0.3829311430454254,
+      "learning_rate": 4.840726767904089e-06,
+      "loss": 0.4589,
+      "step": 1011
+    },
+    {
+      "epoch": 0.55963133640553,
+      "grad_norm": 0.36710280179977417,
+      "learning_rate": 4.831077413987505e-06,
+      "loss": 0.4763,
+      "step": 1012
+    },
+    {
+      "epoch": 0.560184331797235,
+      "grad_norm": 0.37851566076278687,
+      "learning_rate": 4.8214286898848e-06,
+      "loss": 0.4723,
+      "step": 1013
+    },
+    {
+      "epoch": 0.5607373271889401,
+      "grad_norm": 0.3961973488330841,
+      "learning_rate": 4.81178063157044e-06,
+      "loss": 0.456,
+      "step": 1014
+    },
+    {
+      "epoch": 0.5612903225806452,
+      "grad_norm": 0.38343876600265503,
+      "learning_rate": 4.802133275016415e-06,
+      "loss": 0.4626,
+      "step": 1015
+    },
+    {
+      "epoch": 0.5618433179723502,
+      "grad_norm": 0.3635197877883911,
+      "learning_rate": 4.792486656192094e-06,
+      "loss": 0.4692,
+      "step": 1016
+    },
+    {
+      "epoch": 0.5623963133640553,
+      "grad_norm": 0.3797365128993988,
+      "learning_rate": 4.782840811064095e-06,
+      "loss": 0.4587,
+      "step": 1017
+    },
+    {
+      "epoch": 0.5629493087557603,
+      "grad_norm": 0.35782304406166077,
+      "learning_rate": 4.773195775596155e-06,
+      "loss": 0.4503,
+      "step": 1018
+    },
+    {
+      "epoch": 0.5635023041474654,
+      "grad_norm": 0.370126336812973,
+      "learning_rate": 4.763551585748989e-06,
+      "loss": 0.4588,
+      "step": 1019
+    },
+    {
+      "epoch": 0.5640552995391706,
+      "grad_norm": 0.3596176207065582,
+      "learning_rate": 4.753908277480162e-06,
+      "loss": 0.4919,
+      "step": 1020
+    },
+    {
+      "epoch": 0.5646082949308756,
+      "grad_norm": 0.343191921710968,
+      "learning_rate": 4.74426588674395e-06,
+      "loss": 0.4667,
+      "step": 1021
+    },
+    {
+      "epoch": 0.5651612903225807,
+      "grad_norm": 0.3751341998577118,
+      "learning_rate": 4.734624449491208e-06,
+      "loss": 0.4529,
+      "step": 1022
+    },
+    {
+      "epoch": 0.5657142857142857,
+      "grad_norm": 0.3533148169517517,
+      "learning_rate": 4.724984001669237e-06,
+      "loss": 0.4491,
+      "step": 1023
+    },
+    {
+      "epoch": 0.5662672811059908,
+      "grad_norm": 0.38230666518211365,
+      "learning_rate": 4.715344579221649e-06,
+      "loss": 0.4961,
+      "step": 1024
+    },
+    {
+      "epoch": 0.5668202764976958,
+      "grad_norm": 0.40656420588493347,
+      "learning_rate": 4.705706218088232e-06,
+      "loss": 0.4773,
+      "step": 1025
+    },
+    {
+      "epoch": 0.5673732718894009,
+      "grad_norm": 0.3879065215587616,
+      "learning_rate": 4.696068954204817e-06,
+      "loss": 0.5097,
+      "step": 1026
+    },
+    {
+      "epoch": 0.567926267281106,
+      "grad_norm": 0.40097224712371826,
+      "learning_rate": 4.686432823503147e-06,
+      "loss": 0.4661,
+      "step": 1027
+    },
+    {
+      "epoch": 0.568479262672811,
+      "grad_norm": 0.37987810373306274,
+      "learning_rate": 4.676797861910735e-06,
+      "loss": 0.4429,
+      "step": 1028
+    },
+    {
+      "epoch": 0.5690322580645162,
+      "grad_norm": 0.38563069701194763,
+      "learning_rate": 4.667164105350739e-06,
+      "loss": 0.4719,
+      "step": 1029
+    },
+    {
+      "epoch": 0.5695852534562212,
+      "grad_norm": 0.3681628406047821,
+      "learning_rate": 4.657531589741822e-06,
+      "loss": 0.4757,
+      "step": 1030
+    },
+    {
+      "epoch": 0.5701382488479263,
+      "grad_norm": 0.4479837119579315,
+      "learning_rate": 4.647900350998022e-06,
+      "loss": 0.471,
+      "step": 1031
+    },
+    {
+      "epoch": 0.5706912442396314,
+      "grad_norm": 0.39742401242256165,
+      "learning_rate": 4.638270425028614e-06,
+      "loss": 0.4762,
+      "step": 1032
+    },
+    {
+      "epoch": 0.5712442396313364,
+      "grad_norm": 0.38390353322029114,
+      "learning_rate": 4.628641847737982e-06,
+      "loss": 0.4475,
+      "step": 1033
+    },
+    {
+      "epoch": 0.5717972350230415,
+      "grad_norm": 0.4764840602874756,
+      "learning_rate": 4.6190146550254775e-06,
+      "loss": 0.4827,
+      "step": 1034
+    },
+    {
+      "epoch": 0.5723502304147465,
+      "grad_norm": 0.41221946477890015,
+      "learning_rate": 4.609388882785291e-06,
+      "loss": 0.495,
+      "step": 1035
+    },
+    {
+      "epoch": 0.5729032258064516,
+      "grad_norm": 0.35299742221832275,
+      "learning_rate": 4.599764566906319e-06,
+      "loss": 0.4614,
+      "step": 1036
+    },
+    {
+      "epoch": 0.5734562211981566,
+      "grad_norm": 0.45965003967285156,
+      "learning_rate": 4.590141743272026e-06,
+      "loss": 0.462,
+      "step": 1037
+    },
+    {
+      "epoch": 0.5740092165898617,
+      "grad_norm": 0.37869590520858765,
+      "learning_rate": 4.5805204477603135e-06,
+      "loss": 0.4442,
+      "step": 1038
+    },
+    {
+      "epoch": 0.5745622119815669,
+      "grad_norm": 0.37475693225860596,
+      "learning_rate": 4.570900716243385e-06,
+      "loss": 0.4468,
+      "step": 1039
+    },
+    {
+      "epoch": 0.5751152073732719,
+      "grad_norm": 0.38735684752464294,
+      "learning_rate": 4.561282584587612e-06,
+      "loss": 0.4444,
+      "step": 1040
+    },
+    {
+      "epoch": 0.575668202764977,
+      "grad_norm": 0.3577326834201813,
+      "learning_rate": 4.551666088653404e-06,
+      "loss": 0.4864,
+      "step": 1041
+    },
+    {
+      "epoch": 0.576221198156682,
+      "grad_norm": 0.3724598288536072,
+      "learning_rate": 4.542051264295069e-06,
+      "loss": 0.4556,
+      "step": 1042
+    },
+    {
+      "epoch": 0.5767741935483871,
+      "grad_norm": 0.4160480797290802,
+      "learning_rate": 4.532438147360684e-06,
+      "loss": 0.4599,
+      "step": 1043
+    },
+    {
+      "epoch": 0.5773271889400922,
+      "grad_norm": 0.38303178548812866,
+      "learning_rate": 4.522826773691958e-06,
+      "loss": 0.4574,
+      "step": 1044
+    },
+    {
+      "epoch": 0.5778801843317972,
+      "grad_norm": 0.4100678861141205,
+      "learning_rate": 4.513217179124106e-06,
+      "loss": 0.4637,
+      "step": 1045
+    },
+    {
+      "epoch": 0.5784331797235023,
+      "grad_norm": 0.4295979142189026,
+      "learning_rate": 4.5036093994857025e-06,
+      "loss": 0.4747,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5789861751152073,
+      "grad_norm": 0.36032918095588684,
+      "learning_rate": 4.49400347059856e-06,
+      "loss": 0.4682,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5795391705069124,
+      "grad_norm": 0.4092552959918976,
+      "learning_rate": 4.484399428277589e-06,
+      "loss": 0.463,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5800921658986176,
+      "grad_norm": 0.3771011233329773,
+      "learning_rate": 4.474797308330665e-06,
+      "loss": 0.4865,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 0.37506765127182007,
+      "learning_rate": 4.465197146558498e-06,
+      "loss": 0.4558,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5811981566820277,
+      "grad_norm": 0.34741005301475525,
+      "learning_rate": 4.455598978754496e-06,
+      "loss": 0.445,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5817511520737327,
+      "grad_norm": 0.41416463255882263,
+      "learning_rate": 4.4460028407046344e-06,
+      "loss": 0.4755,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5823041474654378,
+      "grad_norm": 0.414931982755661,
+      "learning_rate": 4.436408768187319e-06,
+      "loss": 0.4879,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5828571428571429,
+      "grad_norm": 0.37698793411254883,
+      "learning_rate": 4.426816796973254e-06,
+      "loss": 0.4896,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5834101382488479,
+      "grad_norm": 0.4369506537914276,
+      "learning_rate": 4.417226962825311e-06,
+      "loss": 0.4723,
+      "step": 1055
+    },
+    {
+      "epoch": 0.583963133640553,
+      "grad_norm": 0.3790305256843567,
+      "learning_rate": 4.407639301498392e-06,
+      "loss": 0.4466,
+      "step": 1056
+    },
+    {
+      "epoch": 0.584516129032258,
+      "grad_norm": 0.3516313135623932,
+      "learning_rate": 4.398053848739299e-06,
+      "loss": 0.4483,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5850691244239631,
+      "grad_norm": 0.42550036311149597,
+      "learning_rate": 4.388470640286599e-06,
+      "loss": 0.4633,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5856221198156683,
+      "grad_norm": 0.3732847571372986,
+      "learning_rate": 4.378889711870492e-06,
+      "loss": 0.4909,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5861751152073733,
+      "grad_norm": 0.3723104000091553,
+      "learning_rate": 4.369311099212676e-06,
+      "loss": 0.4612,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5867281105990784,
+      "grad_norm": 0.4225739538669586,
+      "learning_rate": 4.359734838026214e-06,
+      "loss": 0.4576,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5872811059907834,
+      "grad_norm": 0.4037158489227295,
+      "learning_rate": 4.3501609640154056e-06,
+      "loss": 0.4803,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5878341013824885,
+      "grad_norm": 0.3930150866508484,
+      "learning_rate": 4.340589512875646e-06,
+      "loss": 0.4484,
+      "step": 1063
+    },
+    {
+      "epoch": 0.5883870967741935,
+      "grad_norm": 0.4097388684749603,
+      "learning_rate": 4.331020520293298e-06,
+      "loss": 0.4699,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5889400921658986,
+      "grad_norm": 0.3482785224914551,
+      "learning_rate": 4.321454021945559e-06,
+      "loss": 0.4632,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5894930875576037,
+      "grad_norm": 0.45652082562446594,
+      "learning_rate": 4.311890053500326e-06,
+      "loss": 0.4699,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5900460829493087,
+      "grad_norm": 0.3783150613307953,
+      "learning_rate": 4.302328650616062e-06,
+      "loss": 0.4662,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5905990783410138,
+      "grad_norm": 0.36994531750679016,
+      "learning_rate": 4.292769848941669e-06,
+      "loss": 0.4581,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5911520737327189,
+      "grad_norm": 0.3842531144618988,
+      "learning_rate": 4.283213684116346e-06,
+      "loss": 0.4631,
+      "step": 1069
+    },
+    {
+      "epoch": 0.591705069124424,
+      "grad_norm": 0.4432550370693207,
+      "learning_rate": 4.273660191769463e-06,
+      "loss": 0.4759,
+      "step": 1070
+    },
+    {
+      "epoch": 0.5922580645161291,
+      "grad_norm": 0.37499210238456726,
+      "learning_rate": 4.264109407520425e-06,
+      "loss": 0.4703,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5928110599078341,
+      "grad_norm": 0.383779376745224,
+      "learning_rate": 4.2545613669785405e-06,
+      "loss": 0.4793,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5933640552995392,
+      "grad_norm": 0.43077680468559265,
+      "learning_rate": 4.245016105742889e-06,
+      "loss": 0.447,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5939170506912442,
+      "grad_norm": 0.3452954888343811,
+      "learning_rate": 4.2354736594021854e-06,
+      "loss": 0.457,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5944700460829493,
+      "grad_norm": 0.3940063714981079,
+      "learning_rate": 4.2259340635346515e-06,
+      "loss": 0.4579,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5950230414746543,
+      "grad_norm": 0.37174075841903687,
+      "learning_rate": 4.21639735370788e-06,
+      "loss": 0.4718,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5955760368663594,
+      "grad_norm": 0.38617780804634094,
+      "learning_rate": 4.206863565478705e-06,
+      "loss": 0.4698,
+      "step": 1077
+    },
+    {
+      "epoch": 0.5961290322580645,
+      "grad_norm": 0.36345598101615906,
+      "learning_rate": 4.197332734393067e-06,
+      "loss": 0.4828,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5966820276497696,
+      "grad_norm": 0.37622806429862976,
+      "learning_rate": 4.187804895985881e-06,
+      "loss": 0.4584,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5972350230414747,
+      "grad_norm": 0.41539904475212097,
+      "learning_rate": 4.1782800857809025e-06,
+      "loss": 0.4606,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5977880184331797,
+      "grad_norm": 0.37827736139297485,
+      "learning_rate": 4.168758339290599e-06,
+      "loss": 0.4576,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5983410138248848,
+      "grad_norm": 0.36408886313438416,
+      "learning_rate": 4.159239692016012e-06,
+      "loss": 0.4792,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5988940092165899,
+      "grad_norm": 0.347615510225296,
+      "learning_rate": 4.149724179446631e-06,
+      "loss": 0.4627,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5994470046082949,
+      "grad_norm": 0.36233338713645935,
+      "learning_rate": 4.140211837060258e-06,
+      "loss": 0.4595,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.385028213262558,
+      "learning_rate": 4.130702700322873e-06,
+      "loss": 0.4469,
+      "step": 1085
+    },
+    {
+      "epoch": 0.600552995391705,
+      "grad_norm": 0.3965871036052704,
+      "learning_rate": 4.121196804688506e-06,
+      "loss": 0.4761,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6011059907834101,
+      "grad_norm": 0.36717289686203003,
+      "learning_rate": 4.111694185599101e-06,
+      "loss": 0.4401,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6016589861751153,
+      "grad_norm": 0.3657437264919281,
+      "learning_rate": 4.102194878484386e-06,
+      "loss": 0.4613,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6022119815668203,
+      "grad_norm": 0.4054916799068451,
+      "learning_rate": 4.092698918761742e-06,
+      "loss": 0.489,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6027649769585254,
+      "grad_norm": 0.40190011262893677,
+      "learning_rate": 4.083206341836069e-06,
+      "loss": 0.4703,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6033179723502304,
+      "grad_norm": 0.4301840662956238,
+      "learning_rate": 4.073717183099654e-06,
+      "loss": 0.4676,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6038709677419355,
+      "grad_norm": 0.3311327397823334,
+      "learning_rate": 4.06423147793204e-06,
+      "loss": 0.4695,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6044239631336406,
+      "grad_norm": 0.3920708894729614,
+      "learning_rate": 4.054749261699893e-06,
+      "loss": 0.4834,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6049769585253456,
+      "grad_norm": 0.33984097838401794,
+      "learning_rate": 4.045270569756872e-06,
+      "loss": 0.4633,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6055299539170507,
+      "grad_norm": 0.36689960956573486,
+      "learning_rate": 4.0357954374434936e-06,
+      "loss": 0.4778,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6060829493087557,
+      "grad_norm": 0.3129914402961731,
+      "learning_rate": 4.026323900087006e-06,
+      "loss": 0.4389,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6066359447004608,
+      "grad_norm": 0.3611215651035309,
+      "learning_rate": 4.016855993001251e-06,
+      "loss": 0.4687,
+      "step": 1097
+    },
+    {
+      "epoch": 0.607188940092166,
+      "grad_norm": 0.36361318826675415,
+      "learning_rate": 4.007391751486536e-06,
+      "loss": 0.4592,
+      "step": 1098
+    },
+    {
+      "epoch": 0.607741935483871,
+      "grad_norm": 0.3965631425380707,
+      "learning_rate": 3.997931210829503e-06,
+      "loss": 0.4403,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6082949308755761,
+      "grad_norm": 0.371737539768219,
+      "learning_rate": 3.988474406302995e-06,
+      "loss": 0.4433,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6088479262672811,
+      "grad_norm": 0.4065137207508087,
+      "learning_rate": 3.979021373165924e-06,
+      "loss": 0.4548,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6094009216589862,
+      "grad_norm": 0.36007609963417053,
+      "learning_rate": 3.9695721466631426e-06,
+      "loss": 0.4664,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6099539170506912,
+      "grad_norm": 0.39393535256385803,
+      "learning_rate": 3.9601267620253075e-06,
+      "loss": 0.4899,
+      "step": 1103
+    },
+    {
+      "epoch": 0.6105069124423963,
+      "grad_norm": 0.40391650795936584,
+      "learning_rate": 3.950685254468755e-06,
+      "loss": 0.4657,
+      "step": 1104
+    },
+    {
+      "epoch": 0.6110599078341014,
+      "grad_norm": 0.3698117434978485,
+      "learning_rate": 3.941247659195364e-06,
+      "loss": 0.4524,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6116129032258064,
+      "grad_norm": 0.35965868830680847,
+      "learning_rate": 3.931814011392427e-06,
+      "loss": 0.4629,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6121658986175115,
+      "grad_norm": 0.4399069547653198,
+      "learning_rate": 3.922384346232519e-06,
+      "loss": 0.491,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6127188940092166,
+      "grad_norm": 0.3660935163497925,
+      "learning_rate": 3.912958698873366e-06,
+      "loss": 0.474,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6132718894009217,
+      "grad_norm": 0.3902994990348816,
+      "learning_rate": 3.903537104457713e-06,
+      "loss": 0.4979,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6138248847926268,
+      "grad_norm": 0.4067245125770569,
+      "learning_rate": 3.894119598113196e-06,
+      "loss": 0.4724,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6143778801843318,
+      "grad_norm": 0.3835393190383911,
+      "learning_rate": 3.8847062149522075e-06,
+      "loss": 0.4558,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6149308755760369,
+      "grad_norm": 0.3831106126308441,
+      "learning_rate": 3.875296990071768e-06,
+      "loss": 0.4619,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6154838709677419,
+      "grad_norm": 0.3794800043106079,
+      "learning_rate": 3.8658919585533916e-06,
+      "loss": 0.4535,
+      "step": 1113
+    },
+    {
+      "epoch": 0.616036866359447,
+      "grad_norm": 0.3396313786506653,
+      "learning_rate": 3.8564911554629616e-06,
+      "loss": 0.4892,
+      "step": 1114
+    },
+    {
+      "epoch": 0.616589861751152,
+      "grad_norm": 0.37773552536964417,
+      "learning_rate": 3.847094615850593e-06,
+      "loss": 0.46,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6171428571428571,
+      "grad_norm": 0.38637781143188477,
+      "learning_rate": 3.837702374750507e-06,
+      "loss": 0.4701,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6176958525345622,
+      "grad_norm": 0.3425704538822174,
+      "learning_rate": 3.828314467180895e-06,
+      "loss": 0.4389,
+      "step": 1117
+    },
+    {
+      "epoch": 0.6182488479262673,
+      "grad_norm": 0.35482358932495117,
+      "learning_rate": 3.818930928143796e-06,
+      "loss": 0.477,
+      "step": 1118
+    },
+    {
+      "epoch": 0.6188018433179724,
+      "grad_norm": 0.3860636353492737,
+      "learning_rate": 3.8095517926249557e-06,
+      "loss": 0.4748,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6193548387096774,
+      "grad_norm": 0.412280797958374,
+      "learning_rate": 3.800177095593706e-06,
+      "loss": 0.4624,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6199078341013825,
+      "grad_norm": 0.3511303961277008,
+      "learning_rate": 3.790806872002828e-06,
+      "loss": 0.4867,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6204608294930876,
+      "grad_norm": 0.3653261065483093,
+      "learning_rate": 3.7814411567884256e-06,
+      "loss": 0.4737,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6210138248847926,
+      "grad_norm": 0.3349493145942688,
+      "learning_rate": 3.7720799848697924e-06,
+      "loss": 0.4705,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6215668202764977,
+      "grad_norm": 0.3643210828304291,
+      "learning_rate": 3.762723391149282e-06,
+      "loss": 0.4934,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6221198156682027,
+      "grad_norm": 0.3928675055503845,
+      "learning_rate": 3.75337141051218e-06,
+      "loss": 0.4726,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6226728110599078,
+      "grad_norm": 0.36943259835243225,
+      "learning_rate": 3.7440240778265725e-06,
+      "loss": 0.4557,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6232258064516129,
+      "grad_norm": 0.37895265221595764,
+      "learning_rate": 3.734681427943214e-06,
+      "loss": 0.4989,
+      "step": 1127
+    },
+    {
+      "epoch": 0.623778801843318,
+      "grad_norm": 0.34087231755256653,
+      "learning_rate": 3.725343495695404e-06,
+      "loss": 0.4418,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6243317972350231,
+      "grad_norm": 0.3866007626056671,
+      "learning_rate": 3.716010315898847e-06,
+      "loss": 0.4515,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6248847926267281,
+      "grad_norm": 0.4217049777507782,
+      "learning_rate": 3.706681923351533e-06,
+      "loss": 0.4962,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6254377880184332,
+      "grad_norm": 0.4117668867111206,
+      "learning_rate": 3.6973583528336e-06,
+      "loss": 0.4704,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6259907834101383,
+      "grad_norm": 0.3604739308357239,
+      "learning_rate": 3.68803963910721e-06,
+      "loss": 0.4839,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6265437788018433,
+      "grad_norm": 0.4002184569835663,
+      "learning_rate": 3.678725816916415e-06,
+      "loss": 0.4853,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6270967741935484,
+      "grad_norm": 0.3674345910549164,
+      "learning_rate": 3.6694169209870305e-06,
+      "loss": 0.4471,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6276497695852534,
+      "grad_norm": 0.37487921118736267,
+      "learning_rate": 3.660112986026504e-06,
+      "loss": 0.4646,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6282027649769585,
+      "grad_norm": 0.36266854405403137,
+      "learning_rate": 3.6508140467237873e-06,
+      "loss": 0.4802,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6287557603686635,
+      "grad_norm": 0.3480757176876068,
+      "learning_rate": 3.6415201377492048e-06,
+      "loss": 0.458,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6293087557603687,
+      "grad_norm": 0.3710945248603821,
+      "learning_rate": 3.632231293754328e-06,
+      "loss": 0.4648,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6298617511520738,
+      "grad_norm": 0.42196762561798096,
+      "learning_rate": 3.622947549371841e-06,
+      "loss": 0.4802,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6304147465437788,
+      "grad_norm": 0.3192515969276428,
+      "learning_rate": 3.6136689392154186e-06,
+      "loss": 0.4559,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6309677419354839,
+      "grad_norm": 0.37404048442840576,
+      "learning_rate": 3.6043954978795868e-06,
+      "loss": 0.4647,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6315207373271889,
+      "grad_norm": 0.37971508502960205,
+      "learning_rate": 3.5951272599396086e-06,
+      "loss": 0.4347,
+      "step": 1142
+    },
+    {
+      "epoch": 0.632073732718894,
+      "grad_norm": 0.410057932138443,
+      "learning_rate": 3.5858642599513394e-06,
+      "loss": 0.4759,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6326267281105991,
+      "grad_norm": 0.367079496383667,
+      "learning_rate": 3.576606532451108e-06,
+      "loss": 0.4864,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6331797235023041,
+      "grad_norm": 0.3866696357727051,
+      "learning_rate": 3.567354111955585e-06,
+      "loss": 0.4746,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6337327188940092,
+      "grad_norm": 0.3390635848045349,
+      "learning_rate": 3.5581070329616543e-06,
+      "loss": 0.4593,
+      "step": 1146
+    },
+    {
+      "epoch": 0.6342857142857142,
+      "grad_norm": 0.41744962334632874,
+      "learning_rate": 3.5488653299462844e-06,
+      "loss": 0.4514,
+      "step": 1147
+    },
+    {
+      "epoch": 0.6348387096774194,
+      "grad_norm": 0.3437979817390442,
+      "learning_rate": 3.5396290373664e-06,
+      "loss": 0.4689,
+      "step": 1148
+    },
+    {
+      "epoch": 0.6353917050691245,
+      "grad_norm": 0.32637640833854675,
+      "learning_rate": 3.5303981896587534e-06,
+      "loss": 0.4657,
+      "step": 1149
+    },
+    {
+      "epoch": 0.6359447004608295,
+      "grad_norm": 0.3600415885448456,
+      "learning_rate": 3.521172821239796e-06,
+      "loss": 0.4637,
+      "step": 1150
+    },
+    {
+      "epoch": 0.6364976958525346,
+      "grad_norm": 0.4324146807193756,
+      "learning_rate": 3.5119529665055506e-06,
+      "loss": 0.4715,
+      "step": 1151
+    },
+    {
+      "epoch": 0.6370506912442396,
+      "grad_norm": 0.3583220839500427,
+      "learning_rate": 3.502738659831483e-06,
+      "loss": 0.4666,
+      "step": 1152
+    },
+    {
+      "epoch": 0.6376036866359447,
+      "grad_norm": 0.35306817293167114,
+      "learning_rate": 3.49352993557237e-06,
+      "loss": 0.4554,
+      "step": 1153
+    },
+    {
+      "epoch": 0.6381566820276497,
+      "grad_norm": 0.39276352524757385,
+      "learning_rate": 3.48432682806218e-06,
+      "loss": 0.4492,
+      "step": 1154
+    },
+    {
+      "epoch": 0.6387096774193548,
+      "grad_norm": 0.3977315425872803,
+      "learning_rate": 3.4751293716139366e-06,
+      "loss": 0.4523,
+      "step": 1155
+    },
+    {
+      "epoch": 0.6392626728110599,
+      "grad_norm": 0.36172613501548767,
+      "learning_rate": 3.4659376005195956e-06,
+      "loss": 0.5053,
+      "step": 1156
+    },
+    {
+      "epoch": 0.639815668202765,
+      "grad_norm": 0.34491607546806335,
+      "learning_rate": 3.4567515490499135e-06,
+      "loss": 0.4633,
+      "step": 1157
+    },
+    {
+      "epoch": 0.6403686635944701,
+      "grad_norm": 0.3682193160057068,
+      "learning_rate": 3.4475712514543246e-06,
+      "loss": 0.4376,
+      "step": 1158
+    },
+    {
+      "epoch": 0.6409216589861751,
+      "grad_norm": 0.41406625509262085,
+      "learning_rate": 3.438396741960806e-06,
+      "loss": 0.4669,
+      "step": 1159
+    },
+    {
+      "epoch": 0.6414746543778802,
+      "grad_norm": 0.38740435242652893,
+      "learning_rate": 3.4292280547757586e-06,
+      "loss": 0.4516,
+      "step": 1160
+    },
+    {
+      "epoch": 0.6420276497695853,
+      "grad_norm": 0.35590702295303345,
+      "learning_rate": 3.420065224083873e-06,
+      "loss": 0.4406,
+      "step": 1161
+    },
+    {
+      "epoch": 0.6425806451612903,
+      "grad_norm": 0.3844543397426605,
+      "learning_rate": 3.4109082840480045e-06,
+      "loss": 0.4675,
+      "step": 1162
+    },
+    {
+      "epoch": 0.6431336405529954,
+      "grad_norm": 0.40717223286628723,
+      "learning_rate": 3.4017572688090467e-06,
+      "loss": 0.4862,
+      "step": 1163
+    },
+    {
+      "epoch": 0.6436866359447004,
+      "grad_norm": 0.37645992636680603,
+      "learning_rate": 3.392612212485802e-06,
+      "loss": 0.4813,
+      "step": 1164
+    },
+    {
+      "epoch": 0.6442396313364055,
+      "grad_norm": 0.33362942934036255,
+      "learning_rate": 3.3834731491748558e-06,
+      "loss": 0.4576,
+      "step": 1165
+    },
+    {
+      "epoch": 0.6447926267281106,
+      "grad_norm": 0.3458590805530548,
+      "learning_rate": 3.3743401129504496e-06,
+      "loss": 0.4664,
+      "step": 1166
+    },
+    {
+      "epoch": 0.6453456221198157,
+      "grad_norm": 0.39395076036453247,
+      "learning_rate": 3.3652131378643515e-06,
+      "loss": 0.4801,
+      "step": 1167
+    },
+    {
+      "epoch": 0.6458986175115208,
+      "grad_norm": 0.36739885807037354,
+      "learning_rate": 3.356092257945733e-06,
+      "loss": 0.4662,
+      "step": 1168
+    },
+    {
+      "epoch": 0.6464516129032258,
+      "grad_norm": 0.35936224460601807,
+      "learning_rate": 3.3469775072010417e-06,
+      "loss": 0.4678,
+      "step": 1169
+    },
+    {
+      "epoch": 0.6470046082949309,
+      "grad_norm": 0.35847219824790955,
+      "learning_rate": 3.337868919613869e-06,
+      "loss": 0.4595,
+      "step": 1170
+    },
+    {
+      "epoch": 0.647557603686636,
+      "grad_norm": 0.33475178480148315,
+      "learning_rate": 3.3287665291448323e-06,
+      "loss": 0.4631,
+      "step": 1171
+    },
+    {
+      "epoch": 0.648110599078341,
+      "grad_norm": 0.3567461669445038,
+      "learning_rate": 3.319670369731437e-06,
+      "loss": 0.475,
+      "step": 1172
+    },
+    {
+      "epoch": 0.6486635944700461,
+      "grad_norm": 0.4065379798412323,
+      "learning_rate": 3.310580475287963e-06,
+      "loss": 0.4596,
+      "step": 1173
+    },
+    {
+      "epoch": 0.6492165898617511,
+      "grad_norm": 0.34394586086273193,
+      "learning_rate": 3.3014968797053263e-06,
+      "loss": 0.4717,
+      "step": 1174
+    },
+    {
+      "epoch": 0.6497695852534562,
+      "grad_norm": 0.3613266944885254,
+      "learning_rate": 3.2924196168509657e-06,
+      "loss": 0.4933,
+      "step": 1175
+    },
+    {
+      "epoch": 0.6503225806451612,
+      "grad_norm": 0.38999873399734497,
+      "learning_rate": 3.283348720568702e-06,
+      "loss": 0.4756,
+      "step": 1176
+    },
+    {
+      "epoch": 0.6508755760368664,
+      "grad_norm": 0.3532698154449463,
+      "learning_rate": 3.274284224678621e-06,
+      "loss": 0.4575,
+      "step": 1177
+    },
+    {
+      "epoch": 0.6514285714285715,
+      "grad_norm": 0.33012717962265015,
+      "learning_rate": 3.2652261629769457e-06,
+      "loss": 0.4473,
+      "step": 1178
+    },
+    {
+      "epoch": 0.6519815668202765,
+      "grad_norm": 0.3971370458602905,
+      "learning_rate": 3.256174569235909e-06,
+      "loss": 0.5044,
+      "step": 1179
+    },
+    {
+      "epoch": 0.6525345622119816,
+      "grad_norm": 0.3590445816516876,
+      "learning_rate": 3.2471294772036287e-06,
+      "loss": 0.4675,
+      "step": 1180
+    },
+    {
+      "epoch": 0.6530875576036866,
+      "grad_norm": 0.3668496012687683,
+      "learning_rate": 3.238090920603981e-06,
+      "loss": 0.4419,
+      "step": 1181
+    },
+    {
+      "epoch": 0.6536405529953917,
+      "grad_norm": 0.3250981867313385,
+      "learning_rate": 3.2290589331364787e-06,
+      "loss": 0.4597,
+      "step": 1182
+    },
+    {
+      "epoch": 0.6541935483870968,
+      "grad_norm": 0.38982975482940674,
+      "learning_rate": 3.2200335484761352e-06,
+      "loss": 0.4638,
+      "step": 1183
+    },
+    {
+      "epoch": 0.6547465437788018,
+      "grad_norm": 0.38048577308654785,
+      "learning_rate": 3.2110148002733534e-06,
+      "loss": 0.473,
+      "step": 1184
+    },
+    {
+      "epoch": 0.6552995391705069,
+      "grad_norm": 0.35395413637161255,
+      "learning_rate": 3.2020027221537864e-06,
+      "loss": 0.4832,
+      "step": 1185
+    },
+    {
+      "epoch": 0.6558525345622119,
+      "grad_norm": 0.3707823157310486,
+      "learning_rate": 3.192997347718224e-06,
+      "loss": 0.4958,
+      "step": 1186
+    },
+    {
+      "epoch": 0.6564055299539171,
+      "grad_norm": 0.36011847853660583,
+      "learning_rate": 3.1839987105424586e-06,
+      "loss": 0.472,
+      "step": 1187
+    },
+    {
+      "epoch": 0.6569585253456222,
+      "grad_norm": 0.3693712055683136,
+      "learning_rate": 3.1750068441771637e-06,
+      "loss": 0.4649,
+      "step": 1188
+    },
+    {
+      "epoch": 0.6575115207373272,
+      "grad_norm": 0.34001612663269043,
+      "learning_rate": 3.1660217821477686e-06,
+      "loss": 0.4764,
+      "step": 1189
+    },
+    {
+      "epoch": 0.6580645161290323,
+      "grad_norm": 0.332956999540329,
+      "learning_rate": 3.1570435579543333e-06,
+      "loss": 0.4652,
+      "step": 1190
+    },
+    {
+      "epoch": 0.6586175115207373,
+      "grad_norm": 0.35701560974121094,
+      "learning_rate": 3.148072205071423e-06,
+      "loss": 0.4695,
+      "step": 1191
+    },
+    {
+      "epoch": 0.6591705069124424,
+      "grad_norm": 0.3385053277015686,
+      "learning_rate": 3.1391077569479856e-06,
+      "loss": 0.4524,
+      "step": 1192
+    },
+    {
+      "epoch": 0.6597235023041474,
+      "grad_norm": 0.35991233587265015,
+      "learning_rate": 3.130150247007222e-06,
+      "loss": 0.4322,
+      "step": 1193
+    },
+    {
+      "epoch": 0.6602764976958525,
+      "grad_norm": 0.36623886227607727,
+      "learning_rate": 3.1211997086464683e-06,
+      "loss": 0.4754,
+      "step": 1194
+    },
+    {
+      "epoch": 0.6608294930875576,
+      "grad_norm": 0.33661210536956787,
+      "learning_rate": 3.112256175237064e-06,
+      "loss": 0.464,
+      "step": 1195
+    },
+    {
+      "epoch": 0.6613824884792626,
+      "grad_norm": 0.3260611295700073,
+      "learning_rate": 3.103319680124235e-06,
+      "loss": 0.4608,
+      "step": 1196
+    },
+    {
+      "epoch": 0.6619354838709678,
+      "grad_norm": 0.38113391399383545,
+      "learning_rate": 3.0943902566269613e-06,
+      "loss": 0.4963,
+      "step": 1197
+    },
+    {
+      "epoch": 0.6624884792626728,
+      "grad_norm": 0.3196834623813629,
+      "learning_rate": 3.085467938037861e-06,
+      "loss": 0.4631,
+      "step": 1198
+    },
+    {
+      "epoch": 0.6630414746543779,
+      "grad_norm": 0.3713325262069702,
+      "learning_rate": 3.076552757623059e-06,
+      "loss": 0.4559,
+      "step": 1199
+    },
+    {
+      "epoch": 0.663594470046083,
+      "grad_norm": 0.35076281428337097,
+      "learning_rate": 3.0676447486220705e-06,
+      "loss": 0.4601,
+      "step": 1200
+    },
+    {
+      "epoch": 0.664147465437788,
+      "grad_norm": 0.34032174944877625,
+      "learning_rate": 3.058743944247665e-06,
+      "loss": 0.4702,
+      "step": 1201
+    },
+    {
+      "epoch": 0.6647004608294931,
+      "grad_norm": 0.37965843081474304,
+      "learning_rate": 3.0498503776857576e-06,
+      "loss": 0.4806,
+      "step": 1202
+    },
+    {
+      "epoch": 0.6652534562211981,
+      "grad_norm": 0.3258879780769348,
+      "learning_rate": 3.0409640820952735e-06,
+      "loss": 0.4331,
+      "step": 1203
+    },
+    {
+      "epoch": 0.6658064516129032,
+      "grad_norm": 0.3517114222049713,
+      "learning_rate": 3.0320850906080325e-06,
+      "loss": 0.4521,
+      "step": 1204
+    },
+    {
+      "epoch": 0.6663594470046083,
+      "grad_norm": 0.3746355473995209,
+      "learning_rate": 3.0232134363286163e-06,
+      "loss": 0.4702,
+      "step": 1205
+    },
+    {
+      "epoch": 0.6669124423963133,
+      "grad_norm": 0.3362243175506592,
+      "learning_rate": 3.0143491523342562e-06,
+      "loss": 0.4382,
+      "step": 1206
+    },
+    {
+      "epoch": 0.6674654377880185,
+      "grad_norm": 0.3588075041770935,
+      "learning_rate": 3.005492271674697e-06,
+      "loss": 0.452,
+      "step": 1207
+    },
+    {
+      "epoch": 0.6680184331797235,
+      "grad_norm": 0.4125935137271881,
+      "learning_rate": 2.9966428273720904e-06,
+      "loss": 0.4715,
+      "step": 1208
+    },
+    {
+      "epoch": 0.6685714285714286,
+      "grad_norm": 0.34157875180244446,
+      "learning_rate": 2.987800852420855e-06,
+      "loss": 0.4543,
+      "step": 1209
+    },
+    {
+      "epoch": 0.6691244239631337,
+      "grad_norm": 0.34906935691833496,
+      "learning_rate": 2.9789663797875614e-06,
+      "loss": 0.4535,
+      "step": 1210
+    },
+    {
+      "epoch": 0.6696774193548387,
+      "grad_norm": 0.3961377739906311,
+      "learning_rate": 2.970139442410811e-06,
+      "loss": 0.4992,
+      "step": 1211
+    },
+    {
+      "epoch": 0.6702304147465438,
+      "grad_norm": 0.3486752510070801,
+      "learning_rate": 2.961320073201107e-06,
+      "loss": 0.4603,
+      "step": 1212
+    },
+    {
+      "epoch": 0.6707834101382488,
+      "grad_norm": 0.3723919093608856,
+      "learning_rate": 2.952508305040739e-06,
+      "loss": 0.4735,
+      "step": 1213
+    },
+    {
+      "epoch": 0.6713364055299539,
+      "grad_norm": 0.33436325192451477,
+      "learning_rate": 2.9437041707836532e-06,
+      "loss": 0.4558,
+      "step": 1214
+    },
+    {
+      "epoch": 0.6718894009216589,
+      "grad_norm": 0.3649885058403015,
+      "learning_rate": 2.9349077032553387e-06,
+      "loss": 0.4734,
+      "step": 1215
+    },
+    {
+      "epoch": 0.6724423963133641,
+      "grad_norm": 0.4229625463485718,
+      "learning_rate": 2.9261189352526932e-06,
+      "loss": 0.4734,
+      "step": 1216
+    },
+    {
+      "epoch": 0.6729953917050692,
+      "grad_norm": 0.31533482670783997,
+      "learning_rate": 2.9173378995439107e-06,
+      "loss": 0.4596,
+      "step": 1217
+    },
+    {
+      "epoch": 0.6735483870967742,
+      "grad_norm": 0.39130347967147827,
+      "learning_rate": 2.9085646288683587e-06,
+      "loss": 0.4913,
+      "step": 1218
+    },
+    {
+      "epoch": 0.6741013824884793,
+      "grad_norm": 0.3557525873184204,
+      "learning_rate": 2.8997991559364493e-06,
+      "loss": 0.4708,
+      "step": 1219
+    },
+    {
+      "epoch": 0.6746543778801843,
+      "grad_norm": 0.36776086688041687,
+      "learning_rate": 2.8910415134295216e-06,
+      "loss": 0.4685,
+      "step": 1220
+    },
+    {
+      "epoch": 0.6752073732718894,
+      "grad_norm": 0.3936042785644531,
+      "learning_rate": 2.8822917339997237e-06,
+      "loss": 0.4667,
+      "step": 1221
+    },
+    {
+      "epoch": 0.6757603686635945,
+      "grad_norm": 0.3256804645061493,
+      "learning_rate": 2.87354985026988e-06,
+      "loss": 0.428,
+      "step": 1222
+    },
+    {
+      "epoch": 0.6763133640552995,
+      "grad_norm": 0.3750945031642914,
+      "learning_rate": 2.8648158948333817e-06,
+      "loss": 0.458,
+      "step": 1223
+    },
+    {
+      "epoch": 0.6768663594470046,
+      "grad_norm": 0.3525260090827942,
+      "learning_rate": 2.856089900254059e-06,
+      "loss": 0.4469,
+      "step": 1224
+    },
+    {
+      "epoch": 0.6774193548387096,
+      "grad_norm": 0.32836800813674927,
+      "learning_rate": 2.847371899066059e-06,
+      "loss": 0.4581,
+      "step": 1225
+    },
+    {
+      "epoch": 0.6779723502304148,
+      "grad_norm": 0.34020939469337463,
+      "learning_rate": 2.83866192377373e-06,
+      "loss": 0.444,
+      "step": 1226
+    },
+    {
+      "epoch": 0.6785253456221199,
+      "grad_norm": 0.31056728959083557,
+      "learning_rate": 2.829960006851492e-06,
+      "loss": 0.4748,
+      "step": 1227
+    },
+    {
+      "epoch": 0.6790783410138249,
+      "grad_norm": 0.3614257574081421,
+      "learning_rate": 2.8212661807437226e-06,
+      "loss": 0.487,
+      "step": 1228
+    },
+    {
+      "epoch": 0.67963133640553,
+      "grad_norm": 0.33884182572364807,
+      "learning_rate": 2.8125804778646315e-06,
+      "loss": 0.4871,
+      "step": 1229
+    },
+    {
+      "epoch": 0.680184331797235,
+      "grad_norm": 0.3698228597640991,
+      "learning_rate": 2.803902930598144e-06,
+      "loss": 0.4783,
+      "step": 1230
+    },
+    {
+      "epoch": 0.6807373271889401,
+      "grad_norm": 0.3055351674556732,
+      "learning_rate": 2.7952335712977764e-06,
+      "loss": 0.4488,
+      "step": 1231
+    },
+    {
+      "epoch": 0.6812903225806451,
+      "grad_norm": 0.3399738073348999,
+      "learning_rate": 2.7865724322865174e-06,
+      "loss": 0.4489,
+      "step": 1232
+    },
+    {
+      "epoch": 0.6818433179723502,
+      "grad_norm": 0.3493901193141937,
+      "learning_rate": 2.7779195458567067e-06,
+      "loss": 0.4508,
+      "step": 1233
+    },
+    {
+      "epoch": 0.6823963133640553,
+      "grad_norm": 0.38333308696746826,
+      "learning_rate": 2.7692749442699147e-06,
+      "loss": 0.4733,
+      "step": 1234
+    },
+    {
+      "epoch": 0.6829493087557603,
+      "grad_norm": 0.33638396859169006,
+      "learning_rate": 2.7606386597568223e-06,
+      "loss": 0.4521,
+      "step": 1235
+    },
+    {
+      "epoch": 0.6835023041474655,
+      "grad_norm": 0.33842137455940247,
+      "learning_rate": 2.7520107245171035e-06,
+      "loss": 0.4359,
+      "step": 1236
+    },
+    {
+      "epoch": 0.6840552995391705,
+      "grad_norm": 0.34261924028396606,
+      "learning_rate": 2.7433911707192976e-06,
+      "loss": 0.4713,
+      "step": 1237
+    },
+    {
+      "epoch": 0.6846082949308756,
+      "grad_norm": 0.33021679520606995,
+      "learning_rate": 2.7347800305007e-06,
+      "loss": 0.4563,
+      "step": 1238
+    },
+    {
+      "epoch": 0.6851612903225807,
+      "grad_norm": 0.33682435750961304,
+      "learning_rate": 2.7261773359672306e-06,
+      "loss": 0.4429,
+      "step": 1239
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 0.3547312319278717,
+      "learning_rate": 2.7175831191933275e-06,
+      "loss": 0.4846,
+      "step": 1240
+    },
+    {
+      "epoch": 0.6862672811059908,
+      "grad_norm": 0.3378848433494568,
+      "learning_rate": 2.7089974122218165e-06,
+      "loss": 0.4681,
+      "step": 1241
+    },
+    {
+      "epoch": 0.6868202764976958,
+      "grad_norm": 0.3516218066215515,
+      "learning_rate": 2.700420247063793e-06,
+      "loss": 0.4796,
+      "step": 1242
+    },
+    {
+      "epoch": 0.6873732718894009,
+      "grad_norm": 0.3513849973678589,
+      "learning_rate": 2.6918516556985088e-06,
+      "loss": 0.4732,
+      "step": 1243
+    },
+    {
+      "epoch": 0.687926267281106,
+      "grad_norm": 0.37055304646492004,
+      "learning_rate": 2.6832916700732466e-06,
+      "loss": 0.492,
+      "step": 1244
+    },
+    {
+      "epoch": 0.688479262672811,
+      "grad_norm": 0.3224850594997406,
+      "learning_rate": 2.674740322103204e-06,
+      "loss": 0.4618,
+      "step": 1245
+    },
+    {
+      "epoch": 0.6890322580645162,
+      "grad_norm": 0.3502301573753357,
+      "learning_rate": 2.666197643671375e-06,
+      "loss": 0.4562,
+      "step": 1246
+    },
+    {
+      "epoch": 0.6895852534562212,
+      "grad_norm": 0.37586522102355957,
+      "learning_rate": 2.657663666628428e-06,
+      "loss": 0.448,
+      "step": 1247
+    },
+    {
+      "epoch": 0.6901382488479263,
+      "grad_norm": 0.38795405626296997,
+      "learning_rate": 2.6491384227925903e-06,
+      "loss": 0.472,
+      "step": 1248
+    },
+    {
+      "epoch": 0.6906912442396314,
+      "grad_norm": 0.3674571216106415,
+      "learning_rate": 2.640621943949527e-06,
+      "loss": 0.4707,
+      "step": 1249
+    },
+    {
+      "epoch": 0.6912442396313364,
+      "grad_norm": 0.36492788791656494,
+      "learning_rate": 2.6321142618522288e-06,
+      "loss": 0.4644,
+      "step": 1250
+    },
+    {
+      "epoch": 0.6917972350230415,
+      "grad_norm": 0.345634400844574,
+      "learning_rate": 2.6236154082208776e-06,
+      "loss": 0.4556,
+      "step": 1251
+    },
+    {
+      "epoch": 0.6923502304147465,
+      "grad_norm": 0.3678940534591675,
+      "learning_rate": 2.6151254147427486e-06,
+      "loss": 0.4983,
+      "step": 1252
+    },
+    {
+      "epoch": 0.6929032258064516,
+      "grad_norm": 0.34437957406044006,
+      "learning_rate": 2.6066443130720786e-06,
+      "loss": 0.4613,
+      "step": 1253
+    },
+    {
+      "epoch": 0.6934562211981566,
+      "grad_norm": 0.3727521300315857,
+      "learning_rate": 2.598172134829955e-06,
+      "loss": 0.4771,
+      "step": 1254
+    },
+    {
+      "epoch": 0.6940092165898617,
+      "grad_norm": 0.3770442008972168,
+      "learning_rate": 2.5897089116041918e-06,
+      "loss": 0.4627,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6945622119815669,
+      "grad_norm": 0.3384653329849243,
+      "learning_rate": 2.5812546749492163e-06,
+      "loss": 0.4494,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6951152073732719,
+      "grad_norm": 0.3460019528865814,
+      "learning_rate": 2.5728094563859495e-06,
+      "loss": 0.4726,
+      "step": 1257
+    },
+    {
+      "epoch": 0.695668202764977,
+      "grad_norm": 0.328002005815506,
+      "learning_rate": 2.5643732874016903e-06,
+      "loss": 0.4437,
+      "step": 1258
+    },
+    {
+      "epoch": 0.696221198156682,
+      "grad_norm": 0.40374118089675903,
+      "learning_rate": 2.5559461994499968e-06,
+      "loss": 0.4766,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6967741935483871,
+      "grad_norm": 0.3571721017360687,
+      "learning_rate": 2.5475282239505685e-06,
+      "loss": 0.459,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6973271889400922,
+      "grad_norm": 0.42431801557540894,
+      "learning_rate": 2.5391193922891288e-06,
+      "loss": 0.4641,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6978801843317972,
+      "grad_norm": 0.35599109530448914,
+      "learning_rate": 2.5307197358173126e-06,
+      "loss": 0.4601,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6984331797235023,
+      "grad_norm": 0.3282775282859802,
+      "learning_rate": 2.5223292858525423e-06,
+      "loss": 0.4352,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6989861751152073,
+      "grad_norm": 0.3075070381164551,
+      "learning_rate": 2.5139480736779164e-06,
+      "loss": 0.4373,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6995391705069124,
+      "grad_norm": 0.3571772277355194,
+      "learning_rate": 2.5055761305420907e-06,
+      "loss": 0.4938,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7000921658986176,
+      "grad_norm": 0.3431309461593628,
+      "learning_rate": 2.4972134876591618e-06,
+      "loss": 0.4413,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7006451612903226,
+      "grad_norm": 0.3382079601287842,
+      "learning_rate": 2.4888601762085518e-06,
+      "loss": 0.502,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7011981566820277,
+      "grad_norm": 0.31066352128982544,
+      "learning_rate": 2.4805162273348905e-06,
+      "loss": 0.4596,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7017511520737327,
+      "grad_norm": 0.31966033577919006,
+      "learning_rate": 2.4721816721479007e-06,
+      "loss": 0.4769,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7023041474654378,
+      "grad_norm": 0.3357180058956146,
+      "learning_rate": 2.4638565417222816e-06,
+      "loss": 0.437,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7028571428571428,
+      "grad_norm": 0.335115522146225,
+      "learning_rate": 2.4555408670975928e-06,
+      "loss": 0.4498,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7034101382488479,
+      "grad_norm": 0.33428287506103516,
+      "learning_rate": 2.4472346792781366e-06,
+      "loss": 0.4661,
+      "step": 1272
+    },
+    {
+      "epoch": 0.703963133640553,
+      "grad_norm": 0.3672966957092285,
+      "learning_rate": 2.438938009232851e-06,
+      "loss": 0.4775,
+      "step": 1273
+    },
+    {
+      "epoch": 0.704516129032258,
+      "grad_norm": 0.3439885079860687,
+      "learning_rate": 2.4306508878951805e-06,
+      "loss": 0.4611,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7050691244239631,
+      "grad_norm": 0.3395749628543854,
+      "learning_rate": 2.4223733461629716e-06,
+      "loss": 0.4498,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7056221198156682,
+      "grad_norm": 0.35915130376815796,
+      "learning_rate": 2.4141054148983532e-06,
+      "loss": 0.4767,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7061751152073733,
+      "grad_norm": 0.393310546875,
+      "learning_rate": 2.4058471249276232e-06,
+      "loss": 0.4594,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7067281105990784,
+      "grad_norm": 0.3339405357837677,
+      "learning_rate": 2.397598507041132e-06,
+      "loss": 0.4732,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7072811059907834,
+      "grad_norm": 0.35175976157188416,
+      "learning_rate": 2.3893595919931673e-06,
+      "loss": 0.4608,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7078341013824885,
+      "grad_norm": 0.33776116371154785,
+      "learning_rate": 2.381130410501845e-06,
+      "loss": 0.4568,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7083870967741935,
+      "grad_norm": 0.34300360083580017,
+      "learning_rate": 2.3729109932489825e-06,
+      "loss": 0.4802,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7089400921658986,
+      "grad_norm": 0.36740466952323914,
+      "learning_rate": 2.3647013708799987e-06,
+      "loss": 0.4755,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7094930875576037,
+      "grad_norm": 0.3558812439441681,
+      "learning_rate": 2.3565015740037915e-06,
+      "loss": 0.4623,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7100460829493087,
+      "grad_norm": 0.4075135886669159,
+      "learning_rate": 2.3483116331926244e-06,
+      "loss": 0.4612,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7105990783410139,
+      "grad_norm": 0.3450377285480499,
+      "learning_rate": 2.340131578982013e-06,
+      "loss": 0.4695,
+      "step": 1285
+    },
+    {
+      "epoch": 0.7111520737327189,
+      "grad_norm": 0.3357861042022705,
+      "learning_rate": 2.3319614418706127e-06,
+      "loss": 0.4826,
+      "step": 1286
+    },
+    {
+      "epoch": 0.711705069124424,
+      "grad_norm": 0.3738630712032318,
+      "learning_rate": 2.3238012523201035e-06,
+      "loss": 0.4905,
+      "step": 1287
+    },
+    {
+      "epoch": 0.712258064516129,
+      "grad_norm": 0.3660827577114105,
+      "learning_rate": 2.315651040755077e-06,
+      "loss": 0.4754,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7128110599078341,
+      "grad_norm": 0.3476543426513672,
+      "learning_rate": 2.3075108375629218e-06,
+      "loss": 0.4728,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7133640552995392,
+      "grad_norm": 0.33124610781669617,
+      "learning_rate": 2.299380673093712e-06,
+      "loss": 0.474,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7139170506912442,
+      "grad_norm": 0.33316168189048767,
+      "learning_rate": 2.291260577660092e-06,
+      "loss": 0.4853,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7144700460829493,
+      "grad_norm": 0.3274078369140625,
+      "learning_rate": 2.2831505815371656e-06,
+      "loss": 0.4571,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7150230414746543,
+      "grad_norm": 0.3307998776435852,
+      "learning_rate": 2.275050714962383e-06,
+      "loss": 0.4498,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7155760368663594,
+      "grad_norm": 0.3073872923851013,
+      "learning_rate": 2.2669610081354254e-06,
+      "loss": 0.4661,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7161290322580646,
+      "grad_norm": 0.37436577677726746,
+      "learning_rate": 2.2588814912180944e-06,
+      "loss": 0.4746,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7166820276497696,
+      "grad_norm": 0.3833792507648468,
+      "learning_rate": 2.2508121943342008e-06,
+      "loss": 0.4502,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7172350230414747,
+      "grad_norm": 0.3415040075778961,
+      "learning_rate": 2.2427531475694496e-06,
+      "loss": 0.4652,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7177880184331797,
+      "grad_norm": 0.3346996605396271,
+      "learning_rate": 2.2347043809713306e-06,
+      "loss": 0.4474,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7183410138248848,
+      "grad_norm": 0.34259018301963806,
+      "learning_rate": 2.2266659245490017e-06,
+      "loss": 0.4832,
+      "step": 1299
+    },
+    {
+      "epoch": 0.7188940092165899,
+      "grad_norm": 0.3417535424232483,
+      "learning_rate": 2.218637808273184e-06,
+      "loss": 0.4674,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7194470046082949,
+      "grad_norm": 0.37401384115219116,
+      "learning_rate": 2.210620062076044e-06,
+      "loss": 0.4859,
+      "step": 1301
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.33238619565963745,
+      "learning_rate": 2.2026127158510843e-06,
+      "loss": 0.4441,
+      "step": 1302
+    },
+    {
+      "epoch": 0.720552995391705,
+      "grad_norm": 0.32708191871643066,
+      "learning_rate": 2.194615799453032e-06,
+      "loss": 0.4793,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7211059907834101,
+      "grad_norm": 0.30514681339263916,
+      "learning_rate": 2.186629342697727e-06,
+      "loss": 0.4409,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7216589861751153,
+      "grad_norm": 0.35654860734939575,
+      "learning_rate": 2.1786533753620155e-06,
+      "loss": 0.5136,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7222119815668203,
+      "grad_norm": 0.3565172255039215,
+      "learning_rate": 2.170687927183629e-06,
+      "loss": 0.4683,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7227649769585254,
+      "grad_norm": 0.3977636396884918,
+      "learning_rate": 2.1627330278610824e-06,
+      "loss": 0.4781,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7233179723502304,
+      "grad_norm": 0.37379372119903564,
+      "learning_rate": 2.154788707053559e-06,
+      "loss": 0.4713,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7238709677419355,
+      "grad_norm": 0.3560510575771332,
+      "learning_rate": 2.1468549943808033e-06,
+      "loss": 0.4725,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7244239631336405,
+      "grad_norm": 0.3701375424861908,
+      "learning_rate": 2.1389319194230017e-06,
+      "loss": 0.4887,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7249769585253456,
+      "grad_norm": 0.35575488209724426,
+      "learning_rate": 2.1310195117206863e-06,
+      "loss": 0.4628,
+      "step": 1311
+    },
+    {
+      "epoch": 0.7255299539170507,
+      "grad_norm": 0.34897711873054504,
+      "learning_rate": 2.1231178007746136e-06,
+      "loss": 0.443,
+      "step": 1312
+    },
+    {
+      "epoch": 0.7260829493087557,
+      "grad_norm": 0.36126595735549927,
+      "learning_rate": 2.1152268160456584e-06,
+      "loss": 0.4622,
+      "step": 1313
+    },
+    {
+      "epoch": 0.7266359447004608,
+      "grad_norm": 0.3314565420150757,
+      "learning_rate": 2.1073465869547043e-06,
+      "loss": 0.4418,
+      "step": 1314
+    },
+    {
+      "epoch": 0.727188940092166,
+      "grad_norm": 0.32749220728874207,
+      "learning_rate": 2.0994771428825332e-06,
+      "loss": 0.4325,
+      "step": 1315
+    },
+    {
+      "epoch": 0.727741935483871,
+      "grad_norm": 0.39038175344467163,
+      "learning_rate": 2.0916185131697163e-06,
+      "loss": 0.4928,
+      "step": 1316
+    },
+    {
+      "epoch": 0.7282949308755761,
+      "grad_norm": 0.3242311477661133,
+      "learning_rate": 2.083770727116503e-06,
+      "loss": 0.4708,
+      "step": 1317
+    },
+    {
+      "epoch": 0.7288479262672811,
+      "grad_norm": 0.3813181221485138,
+      "learning_rate": 2.0759338139827145e-06,
+      "loss": 0.4959,
+      "step": 1318
+    },
+    {
+      "epoch": 0.7294009216589862,
+      "grad_norm": 0.31590506434440613,
+      "learning_rate": 2.0681078029876322e-06,
+      "loss": 0.451,
+      "step": 1319
+    },
+    {
+      "epoch": 0.7299539170506912,
+      "grad_norm": 0.3604491651058197,
+      "learning_rate": 2.0602927233098908e-06,
+      "loss": 0.4817,
+      "step": 1320
+    },
+    {
+      "epoch": 0.7305069124423963,
+      "grad_norm": 0.3593979775905609,
+      "learning_rate": 2.0524886040873676e-06,
+      "loss": 0.4707,
+      "step": 1321
+    },
+    {
+      "epoch": 0.7310599078341014,
+      "grad_norm": 0.38674402236938477,
+      "learning_rate": 2.0446954744170748e-06,
+      "loss": 0.4522,
+      "step": 1322
+    },
+    {
+      "epoch": 0.7316129032258064,
+      "grad_norm": 0.3926672637462616,
+      "learning_rate": 2.036913363355052e-06,
+      "loss": 0.4582,
+      "step": 1323
+    },
+    {
+      "epoch": 0.7321658986175115,
+      "grad_norm": 0.3277740776538849,
+      "learning_rate": 2.029142299916255e-06,
+      "loss": 0.4913,
+      "step": 1324
+    },
+    {
+      "epoch": 0.7327188940092166,
+      "grad_norm": 0.3636924922466278,
+      "learning_rate": 2.0213823130744516e-06,
+      "loss": 0.4624,
+      "step": 1325
+    },
+    {
+      "epoch": 0.7332718894009217,
+      "grad_norm": 0.36329200863838196,
+      "learning_rate": 2.0136334317621093e-06,
+      "loss": 0.4551,
+      "step": 1326
+    },
+    {
+      "epoch": 0.7338248847926268,
+      "grad_norm": 0.3718715012073517,
+      "learning_rate": 2.005895684870291e-06,
+      "loss": 0.4717,
+      "step": 1327
+    },
+    {
+      "epoch": 0.7343778801843318,
+      "grad_norm": 0.3874320685863495,
+      "learning_rate": 1.9981691012485455e-06,
+      "loss": 0.4757,
+      "step": 1328
+    },
+    {
+      "epoch": 0.7349308755760369,
+      "grad_norm": 0.29239824414253235,
+      "learning_rate": 1.9904537097048004e-06,
+      "loss": 0.4554,
+      "step": 1329
+    },
+    {
+      "epoch": 0.7354838709677419,
+      "grad_norm": 0.3096185326576233,
+      "learning_rate": 1.982749539005254e-06,
+      "loss": 0.4559,
+      "step": 1330
+    },
+    {
+      "epoch": 0.736036866359447,
+      "grad_norm": 0.34961435198783875,
+      "learning_rate": 1.9750566178742697e-06,
+      "loss": 0.4376,
+      "step": 1331
+    },
+    {
+      "epoch": 0.736589861751152,
+      "grad_norm": 0.36215296387672424,
+      "learning_rate": 1.9673749749942655e-06,
+      "loss": 0.4544,
+      "step": 1332
+    },
+    {
+      "epoch": 0.7371428571428571,
+      "grad_norm": 0.3274637758731842,
+      "learning_rate": 1.959704639005613e-06,
+      "loss": 0.4468,
+      "step": 1333
+    },
+    {
+      "epoch": 0.7376958525345622,
+      "grad_norm": 0.324142187833786,
+      "learning_rate": 1.952045638506523e-06,
+      "loss": 0.4775,
+      "step": 1334
+    },
+    {
+      "epoch": 0.7382488479262673,
+      "grad_norm": 0.30846402049064636,
+      "learning_rate": 1.9443980020529456e-06,
+      "loss": 0.4507,
+      "step": 1335
+    },
+    {
+      "epoch": 0.7388018433179724,
+      "grad_norm": 0.36270537972450256,
+      "learning_rate": 1.9367617581584606e-06,
+      "loss": 0.4526,
+      "step": 1336
+    },
+    {
+      "epoch": 0.7393548387096774,
+      "grad_norm": 0.35891664028167725,
+      "learning_rate": 1.9291369352941696e-06,
+      "loss": 0.4606,
+      "step": 1337
+    },
+    {
+      "epoch": 0.7399078341013825,
+      "grad_norm": 0.3416483700275421,
+      "learning_rate": 1.9215235618885964e-06,
+      "loss": 0.4976,
+      "step": 1338
+    },
+    {
+      "epoch": 0.7404608294930876,
+      "grad_norm": 0.30863887071609497,
+      "learning_rate": 1.9139216663275727e-06,
+      "loss": 0.44,
+      "step": 1339
+    },
+    {
+      "epoch": 0.7410138248847926,
+      "grad_norm": 0.313449501991272,
+      "learning_rate": 1.9063312769541348e-06,
+      "loss": 0.4643,
+      "step": 1340
+    },
+    {
+      "epoch": 0.7415668202764977,
+      "grad_norm": 0.35145795345306396,
+      "learning_rate": 1.8987524220684222e-06,
+      "loss": 0.4689,
+      "step": 1341
+    },
+    {
+      "epoch": 0.7421198156682027,
+      "grad_norm": 0.34510427713394165,
+      "learning_rate": 1.8911851299275675e-06,
+      "loss": 0.498,
+      "step": 1342
+    },
+    {
+      "epoch": 0.7426728110599078,
+      "grad_norm": 0.3454337418079376,
+      "learning_rate": 1.8836294287455936e-06,
+      "loss": 0.4488,
+      "step": 1343
+    },
+    {
+      "epoch": 0.743225806451613,
+      "grad_norm": 0.3266771733760834,
+      "learning_rate": 1.8760853466933072e-06,
+      "loss": 0.4855,
+      "step": 1344
+    },
+    {
+      "epoch": 0.743778801843318,
+      "grad_norm": 0.3165615200996399,
+      "learning_rate": 1.8685529118981926e-06,
+      "loss": 0.4556,
+      "step": 1345
+    },
+    {
+      "epoch": 0.7443317972350231,
+      "grad_norm": 0.33930885791778564,
+      "learning_rate": 1.8610321524443099e-06,
+      "loss": 0.4401,
+      "step": 1346
+    },
+    {
+      "epoch": 0.7448847926267281,
+      "grad_norm": 0.35839566588401794,
+      "learning_rate": 1.8535230963721884e-06,
+      "loss": 0.4965,
+      "step": 1347
+    },
+    {
+      "epoch": 0.7454377880184332,
+      "grad_norm": 0.33815956115722656,
+      "learning_rate": 1.8460257716787216e-06,
+      "loss": 0.4868,
+      "step": 1348
+    },
+    {
+      "epoch": 0.7459907834101382,
+      "grad_norm": 0.34670326113700867,
+      "learning_rate": 1.8385402063170643e-06,
+      "loss": 0.4696,
+      "step": 1349
+    },
+    {
+      "epoch": 0.7465437788018433,
+      "grad_norm": 0.32649490237236023,
+      "learning_rate": 1.8310664281965268e-06,
+      "loss": 0.4951,
+      "step": 1350
+    },
+    {
+      "epoch": 0.7470967741935484,
+      "grad_norm": 0.32630065083503723,
+      "learning_rate": 1.8236044651824725e-06,
+      "loss": 0.4733,
+      "step": 1351
+    },
+    {
+      "epoch": 0.7476497695852534,
+      "grad_norm": 0.2883811593055725,
+      "learning_rate": 1.8161543450962127e-06,
+      "loss": 0.4637,
+      "step": 1352
+    },
+    {
+      "epoch": 0.7482027649769585,
+      "grad_norm": 0.34161558747291565,
+      "learning_rate": 1.8087160957149036e-06,
+      "loss": 0.4629,
+      "step": 1353
+    },
+    {
+      "epoch": 0.7487557603686636,
+      "grad_norm": 0.3212891221046448,
+      "learning_rate": 1.8012897447714417e-06,
+      "loss": 0.4703,
+      "step": 1354
+    },
+    {
+      "epoch": 0.7493087557603687,
+      "grad_norm": 0.3223973214626312,
+      "learning_rate": 1.7938753199543618e-06,
+      "loss": 0.4536,
+      "step": 1355
+    },
+    {
+      "epoch": 0.7498617511520738,
+      "grad_norm": 0.3456159830093384,
+      "learning_rate": 1.7864728489077332e-06,
+      "loss": 0.4671,
+      "step": 1356
+    },
+    {
+      "epoch": 0.7504147465437788,
+      "grad_norm": 0.35950833559036255,
+      "learning_rate": 1.779082359231057e-06,
+      "loss": 0.4674,
+      "step": 1357
+    },
+    {
+      "epoch": 0.7509677419354839,
+      "grad_norm": 0.33876293897628784,
+      "learning_rate": 1.7717038784791612e-06,
+      "loss": 0.4614,
+      "step": 1358
+    },
+    {
+      "epoch": 0.7515207373271889,
+      "grad_norm": 0.34674006700515747,
+      "learning_rate": 1.7643374341621006e-06,
+      "loss": 0.4434,
+      "step": 1359
+    },
+    {
+      "epoch": 0.752073732718894,
+      "grad_norm": 0.3747800290584564,
+      "learning_rate": 1.7569830537450533e-06,
+      "loss": 0.4607,
+      "step": 1360
+    },
+    {
+      "epoch": 0.752626728110599,
+      "grad_norm": 0.3502337336540222,
+      "learning_rate": 1.7496407646482182e-06,
+      "loss": 0.471,
+      "step": 1361
+    },
+    {
+      "epoch": 0.7531797235023041,
+      "grad_norm": 0.3303203582763672,
+      "learning_rate": 1.7423105942467116e-06,
+      "loss": 0.4801,
+      "step": 1362
+    },
+    {
+      "epoch": 0.7537327188940092,
+      "grad_norm": 0.3284555673599243,
+      "learning_rate": 1.7349925698704673e-06,
+      "loss": 0.4742,
+      "step": 1363
+    },
+    {
+      "epoch": 0.7542857142857143,
+      "grad_norm": 0.3339201509952545,
+      "learning_rate": 1.7276867188041336e-06,
+      "loss": 0.4588,
+      "step": 1364
+    },
+    {
+      "epoch": 0.7548387096774194,
+      "grad_norm": 0.35038480162620544,
+      "learning_rate": 1.7203930682869707e-06,
+      "loss": 0.4921,
+      "step": 1365
+    },
+    {
+      "epoch": 0.7553917050691245,
+      "grad_norm": 0.3145638406276703,
+      "learning_rate": 1.7131116455127518e-06,
+      "loss": 0.428,
+      "step": 1366
+    },
+    {
+      "epoch": 0.7559447004608295,
+      "grad_norm": 0.34270545840263367,
+      "learning_rate": 1.7058424776296583e-06,
+      "loss": 0.4599,
+      "step": 1367
+    },
+    {
+      "epoch": 0.7564976958525346,
+      "grad_norm": 0.33915436267852783,
+      "learning_rate": 1.698585591740181e-06,
+      "loss": 0.4713,
+      "step": 1368
+    },
+    {
+      "epoch": 0.7570506912442396,
+      "grad_norm": 0.3200076222419739,
+      "learning_rate": 1.6913410149010179e-06,
+      "loss": 0.4469,
+      "step": 1369
+    },
+    {
+      "epoch": 0.7576036866359447,
+      "grad_norm": 0.31938573718070984,
+      "learning_rate": 1.6841087741229745e-06,
+      "loss": 0.4563,
+      "step": 1370
+    },
+    {
+      "epoch": 0.7581566820276497,
+      "grad_norm": 0.3671538233757019,
+      "learning_rate": 1.6768888963708612e-06,
+      "loss": 0.471,
+      "step": 1371
+    },
+    {
+      "epoch": 0.7587096774193548,
+      "grad_norm": 0.33220696449279785,
+      "learning_rate": 1.669681408563395e-06,
+      "loss": 0.444,
+      "step": 1372
+    },
+    {
+      "epoch": 0.7592626728110599,
+      "grad_norm": 0.3261774182319641,
+      "learning_rate": 1.6624863375730977e-06,
+      "loss": 0.4775,
+      "step": 1373
+    },
+    {
+      "epoch": 0.759815668202765,
+      "grad_norm": 0.31591543555259705,
+      "learning_rate": 1.6553037102261955e-06,
+      "loss": 0.4763,
+      "step": 1374
+    },
+    {
+      "epoch": 0.7603686635944701,
+      "grad_norm": 0.345234215259552,
+      "learning_rate": 1.6481335533025195e-06,
+      "loss": 0.443,
+      "step": 1375
+    },
+    {
+      "epoch": 0.7609216589861751,
+      "grad_norm": 0.3390832245349884,
+      "learning_rate": 1.6409758935354065e-06,
+      "loss": 0.4622,
+      "step": 1376
+    },
+    {
+      "epoch": 0.7614746543778802,
+      "grad_norm": 0.34640833735466003,
+      "learning_rate": 1.6338307576115987e-06,
+      "loss": 0.4606,
+      "step": 1377
+    },
+    {
+      "epoch": 0.7620276497695853,
+      "grad_norm": 0.3271695375442505,
+      "learning_rate": 1.6266981721711438e-06,
+      "loss": 0.4354,
+      "step": 1378
+    },
+    {
+      "epoch": 0.7625806451612903,
+      "grad_norm": 0.34254854917526245,
+      "learning_rate": 1.619578163807296e-06,
+      "loss": 0.4601,
+      "step": 1379
+    },
+    {
+      "epoch": 0.7631336405529954,
+      "grad_norm": 0.3350234925746918,
+      "learning_rate": 1.6124707590664168e-06,
+      "loss": 0.4652,
+      "step": 1380
+    },
+    {
+      "epoch": 0.7636866359447004,
+      "grad_norm": 0.3148898780345917,
+      "learning_rate": 1.6053759844478768e-06,
+      "loss": 0.4702,
+      "step": 1381
+    },
+    {
+      "epoch": 0.7642396313364055,
+      "grad_norm": 0.3196204602718353,
+      "learning_rate": 1.5982938664039555e-06,
+      "loss": 0.464,
+      "step": 1382
+    },
+    {
+      "epoch": 0.7647926267281105,
+      "grad_norm": 0.3282592296600342,
+      "learning_rate": 1.591224431339744e-06,
+      "loss": 0.4631,
+      "step": 1383
+    },
+    {
+      "epoch": 0.7653456221198157,
+      "grad_norm": 0.34297293424606323,
+      "learning_rate": 1.584167705613046e-06,
+      "loss": 0.4494,
+      "step": 1384
+    },
+    {
+      "epoch": 0.7658986175115208,
+      "grad_norm": 0.34436681866645813,
+      "learning_rate": 1.5771237155342784e-06,
+      "loss": 0.4501,
+      "step": 1385
+    },
+    {
+      "epoch": 0.7664516129032258,
+      "grad_norm": 0.2950928807258606,
+      "learning_rate": 1.5700924873663758e-06,
+      "loss": 0.4493,
+      "step": 1386
+    },
+    {
+      "epoch": 0.7670046082949309,
+      "grad_norm": 0.3339753746986389,
+      "learning_rate": 1.5630740473246896e-06,
+      "loss": 0.4793,
+      "step": 1387
+    },
+    {
+      "epoch": 0.767557603686636,
+      "grad_norm": 0.3539128005504608,
+      "learning_rate": 1.5560684215768935e-06,
+      "loss": 0.4459,
+      "step": 1388
+    },
+    {
+      "epoch": 0.768110599078341,
+      "grad_norm": 0.35491037368774414,
+      "learning_rate": 1.549075636242882e-06,
+      "loss": 0.4814,
+      "step": 1389
+    },
+    {
+      "epoch": 0.7686635944700461,
+      "grad_norm": 0.3396543562412262,
+      "learning_rate": 1.5420957173946772e-06,
+      "loss": 0.4526,
+      "step": 1390
+    },
+    {
+      "epoch": 0.7692165898617511,
+      "grad_norm": 0.35023194551467896,
+      "learning_rate": 1.5351286910563278e-06,
+      "loss": 0.432,
+      "step": 1391
+    },
+    {
+      "epoch": 0.7697695852534562,
+      "grad_norm": 0.3362998366355896,
+      "learning_rate": 1.5281745832038159e-06,
+      "loss": 0.4518,
+      "step": 1392
+    },
+    {
+      "epoch": 0.7703225806451612,
+      "grad_norm": 0.3237576186656952,
+      "learning_rate": 1.5212334197649564e-06,
+      "loss": 0.4732,
+      "step": 1393
+    },
+    {
+      "epoch": 0.7708755760368664,
+      "grad_norm": 0.3349010646343231,
+      "learning_rate": 1.514305226619302e-06,
+      "loss": 0.4793,
+      "step": 1394
+    },
+    {
+      "epoch": 0.7714285714285715,
+      "grad_norm": 0.3398427367210388,
+      "learning_rate": 1.5073900295980481e-06,
+      "loss": 0.471,
+      "step": 1395
+    },
+    {
+      "epoch": 0.7719815668202765,
+      "grad_norm": 0.33397191762924194,
+      "learning_rate": 1.5004878544839335e-06,
+      "loss": 0.4398,
+      "step": 1396
+    },
+    {
+      "epoch": 0.7725345622119816,
+      "grad_norm": 0.3291890025138855,
+      "learning_rate": 1.493598727011148e-06,
+      "loss": 0.4632,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7730875576036866,
+      "grad_norm": 0.33560484647750854,
+      "learning_rate": 1.4867226728652318e-06,
+      "loss": 0.4661,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7736405529953917,
+      "grad_norm": 0.32114970684051514,
+      "learning_rate": 1.4798597176829844e-06,
+      "loss": 0.4558,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.29689663648605347,
+      "learning_rate": 1.4730098870523652e-06,
+      "loss": 0.4567,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7747465437788018,
+      "grad_norm": 0.33287614583969116,
+      "learning_rate": 1.4661732065124012e-06,
+      "loss": 0.4458,
+      "step": 1401
+    },
+    {
+      "epoch": 0.7752995391705069,
+      "grad_norm": 0.32061877846717834,
+      "learning_rate": 1.45934970155309e-06,
+      "loss": 0.4482,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7758525345622119,
+      "grad_norm": 0.32134944200515747,
+      "learning_rate": 1.4525393976153046e-06,
+      "loss": 0.4669,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7764055299539171,
+      "grad_norm": 0.31328532099723816,
+      "learning_rate": 1.4457423200906994e-06,
+      "loss": 0.474,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7769585253456222,
+      "grad_norm": 0.31957975029945374,
+      "learning_rate": 1.4389584943216156e-06,
+      "loss": 0.4478,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7775115207373272,
+      "grad_norm": 0.318330317735672,
+      "learning_rate": 1.4321879456009858e-06,
+      "loss": 0.4738,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7780645161290323,
+      "grad_norm": 0.3517070710659027,
+      "learning_rate": 1.4254306991722406e-06,
+      "loss": 0.49,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7786175115207373,
+      "grad_norm": 0.3362846374511719,
+      "learning_rate": 1.4186867802292132e-06,
+      "loss": 0.4756,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7791705069124424,
+      "grad_norm": 0.32801997661590576,
+      "learning_rate": 1.411956213916048e-06,
+      "loss": 0.4745,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7797235023041474,
+      "grad_norm": 0.32685017585754395,
+      "learning_rate": 1.4052390253271037e-06,
+      "loss": 0.4618,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7802764976958525,
+      "grad_norm": 0.32221168279647827,
+      "learning_rate": 1.3985352395068618e-06,
+      "loss": 0.4373,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7808294930875576,
+      "grad_norm": 0.3476710915565491,
+      "learning_rate": 1.3918448814498336e-06,
+      "loss": 0.4463,
+      "step": 1412
+    },
+    {
+      "epoch": 0.7813824884792627,
+      "grad_norm": 0.3333290219306946,
+      "learning_rate": 1.3851679761004644e-06,
+      "loss": 0.4592,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7819354838709678,
+      "grad_norm": 0.32516762614250183,
+      "learning_rate": 1.3785045483530435e-06,
+      "loss": 0.476,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7824884792626728,
+      "grad_norm": 0.3284921944141388,
+      "learning_rate": 1.3718546230516095e-06,
+      "loss": 0.4669,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7830414746543779,
+      "grad_norm": 0.3187192976474762,
+      "learning_rate": 1.3652182249898583e-06,
+      "loss": 0.478,
+      "step": 1416
+    },
+    {
+      "epoch": 0.783594470046083,
+      "grad_norm": 0.3000369369983673,
+      "learning_rate": 1.3585953789110506e-06,
+      "loss": 0.4763,
+      "step": 1417
+    },
+    {
+      "epoch": 0.784147465437788,
+      "grad_norm": 0.3406629264354706,
+      "learning_rate": 1.3519861095079202e-06,
+      "loss": 0.451,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7847004608294931,
+      "grad_norm": 0.3349382281303406,
+      "learning_rate": 1.3453904414225799e-06,
+      "loss": 0.4573,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7852534562211981,
+      "grad_norm": 0.3069670796394348,
+      "learning_rate": 1.3388083992464335e-06,
+      "loss": 0.4739,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7858064516129032,
+      "grad_norm": 0.31198862195014954,
+      "learning_rate": 1.3322400075200792e-06,
+      "loss": 0.454,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7863594470046082,
+      "grad_norm": 0.30845198035240173,
+      "learning_rate": 1.325685290733223e-06,
+      "loss": 0.4498,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7869124423963134,
+      "grad_norm": 0.31618401408195496,
+      "learning_rate": 1.3191442733245824e-06,
+      "loss": 0.491,
+      "step": 1423
+    },
+    {
+      "epoch": 0.7874654377880185,
+      "grad_norm": 0.356711208820343,
+      "learning_rate": 1.3126169796818001e-06,
+      "loss": 0.4422,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7880184331797235,
+      "grad_norm": 0.35151243209838867,
+      "learning_rate": 1.3061034341413497e-06,
+      "loss": 0.469,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7885714285714286,
+      "grad_norm": 0.30518248677253723,
+      "learning_rate": 1.2996036609884478e-06,
+      "loss": 0.4778,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7891244239631336,
+      "grad_norm": 0.30732637643814087,
+      "learning_rate": 1.2931176844569588e-06,
+      "loss": 0.4627,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7896774193548387,
+      "grad_norm": 0.3301125466823578,
+      "learning_rate": 1.2866455287293094e-06,
+      "loss": 0.4711,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7902304147465438,
+      "grad_norm": 0.3300682008266449,
+      "learning_rate": 1.2801872179363978e-06,
+      "loss": 0.497,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7907834101382488,
+      "grad_norm": 0.3206273019313812,
+      "learning_rate": 1.2737427761575006e-06,
+      "loss": 0.4593,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7913364055299539,
+      "grad_norm": 0.35528701543807983,
+      "learning_rate": 1.2673122274201844e-06,
+      "loss": 0.452,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7918894009216589,
+      "grad_norm": 0.3094896972179413,
+      "learning_rate": 1.2608955957002196e-06,
+      "loss": 0.4426,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7924423963133641,
+      "grad_norm": 0.3296172618865967,
+      "learning_rate": 1.2544929049214843e-06,
+      "loss": 0.4751,
+      "step": 1433
+    },
+    {
+      "epoch": 0.7929953917050692,
+      "grad_norm": 0.3449612855911255,
+      "learning_rate": 1.248104178955883e-06,
+      "loss": 0.473,
+      "step": 1434
+    },
+    {
+      "epoch": 0.7935483870967742,
+      "grad_norm": 0.30491313338279724,
+      "learning_rate": 1.2417294416232505e-06,
+      "loss": 0.447,
+      "step": 1435
+    },
+    {
+      "epoch": 0.7941013824884793,
+      "grad_norm": 0.32028716802597046,
+      "learning_rate": 1.235368716691267e-06,
+      "loss": 0.4358,
+      "step": 1436
+    },
+    {
+      "epoch": 0.7946543778801843,
+      "grad_norm": 0.3335442543029785,
+      "learning_rate": 1.22902202787537e-06,
+      "loss": 0.476,
+      "step": 1437
+    },
+    {
+      "epoch": 0.7952073732718894,
+      "grad_norm": 0.3434390723705292,
+      "learning_rate": 1.2226893988386618e-06,
+      "loss": 0.4447,
+      "step": 1438
+    },
+    {
+      "epoch": 0.7957603686635945,
+      "grad_norm": 0.3305012881755829,
+      "learning_rate": 1.2163708531918267e-06,
+      "loss": 0.4644,
+      "step": 1439
+    },
+    {
+      "epoch": 0.7963133640552995,
+      "grad_norm": 0.32044875621795654,
+      "learning_rate": 1.210066414493039e-06,
+      "loss": 0.4525,
+      "step": 1440
+    },
+    {
+      "epoch": 0.7968663594470046,
+      "grad_norm": 0.3219109773635864,
+      "learning_rate": 1.2037761062478759e-06,
+      "loss": 0.4614,
+      "step": 1441
+    },
+    {
+      "epoch": 0.7974193548387096,
+      "grad_norm": 0.3471842408180237,
+      "learning_rate": 1.1974999519092311e-06,
+      "loss": 0.4541,
+      "step": 1442
+    },
+    {
+      "epoch": 0.7979723502304148,
+      "grad_norm": 0.2974897623062134,
+      "learning_rate": 1.1912379748772267e-06,
+      "loss": 0.4344,
+      "step": 1443
+    },
+    {
+      "epoch": 0.7985253456221199,
+      "grad_norm": 0.31679412722587585,
+      "learning_rate": 1.1849901984991253e-06,
+      "loss": 0.456,
+      "step": 1444
+    },
+    {
+      "epoch": 0.7990783410138249,
+      "grad_norm": 0.30493679642677307,
+      "learning_rate": 1.1787566460692446e-06,
+      "loss": 0.4674,
+      "step": 1445
+    },
+    {
+      "epoch": 0.79963133640553,
+      "grad_norm": 0.3283561170101166,
+      "learning_rate": 1.1725373408288682e-06,
+      "loss": 0.4621,
+      "step": 1446
+    },
+    {
+      "epoch": 0.800184331797235,
+      "grad_norm": 0.32548418641090393,
+      "learning_rate": 1.1663323059661609e-06,
+      "loss": 0.4422,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8007373271889401,
+      "grad_norm": 0.29554906487464905,
+      "learning_rate": 1.1601415646160813e-06,
+      "loss": 0.4535,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8012903225806451,
+      "grad_norm": 0.31748977303504944,
+      "learning_rate": 1.153965139860297e-06,
+      "loss": 0.4579,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8018433179723502,
+      "grad_norm": 0.30266064405441284,
+      "learning_rate": 1.147803054727095e-06,
+      "loss": 0.446,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8023963133640553,
+      "grad_norm": 0.32854562997817993,
+      "learning_rate": 1.1416553321913009e-06,
+      "loss": 0.4632,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8029493087557603,
+      "grad_norm": 0.29594528675079346,
+      "learning_rate": 1.1355219951741881e-06,
+      "loss": 0.4704,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8035023041474655,
+      "grad_norm": 0.2916601300239563,
+      "learning_rate": 1.1294030665433969e-06,
+      "loss": 0.4287,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8040552995391705,
+      "grad_norm": 0.31938451528549194,
+      "learning_rate": 1.1232985691128457e-06,
+      "loss": 0.4242,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8046082949308756,
+      "grad_norm": 0.3538084030151367,
+      "learning_rate": 1.1172085256426473e-06,
+      "loss": 0.4705,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8051612903225807,
+      "grad_norm": 0.3275561034679413,
+      "learning_rate": 1.1111329588390253e-06,
+      "loss": 0.4745,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8057142857142857,
+      "grad_norm": 0.30866336822509766,
+      "learning_rate": 1.1050718913542275e-06,
+      "loss": 0.453,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8062672811059908,
+      "grad_norm": 0.3329750597476959,
+      "learning_rate": 1.0990253457864418e-06,
+      "loss": 0.4598,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8068202764976958,
+      "grad_norm": 0.32834047079086304,
+      "learning_rate": 1.0929933446797136e-06,
+      "loss": 0.4649,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8073732718894009,
+      "grad_norm": 0.3380810022354126,
+      "learning_rate": 1.0869759105238592e-06,
+      "loss": 0.4605,
+      "step": 1460
+    },
+    {
+      "epoch": 0.807926267281106,
+      "grad_norm": 0.3055575489997864,
+      "learning_rate": 1.0809730657543838e-06,
+      "loss": 0.4852,
+      "step": 1461
+    },
+    {
+      "epoch": 0.808479262672811,
+      "grad_norm": 0.30661699175834656,
+      "learning_rate": 1.0749848327523966e-06,
+      "loss": 0.4824,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8090322580645162,
+      "grad_norm": 0.32859787344932556,
+      "learning_rate": 1.0690112338445292e-06,
+      "loss": 0.464,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8095852534562212,
+      "grad_norm": 0.3309606611728668,
+      "learning_rate": 1.0630522913028508e-06,
+      "loss": 0.4585,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8101382488479263,
+      "grad_norm": 0.33870929479599,
+      "learning_rate": 1.0571080273447858e-06,
+      "loss": 0.476,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8106912442396313,
+      "grad_norm": 0.33712446689605713,
+      "learning_rate": 1.0511784641330286e-06,
+      "loss": 0.4476,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8112442396313364,
+      "grad_norm": 0.2965930700302124,
+      "learning_rate": 1.0452636237754642e-06,
+      "loss": 0.4766,
+      "step": 1467
+    },
+    {
+      "epoch": 0.8117972350230415,
+      "grad_norm": 0.3398984372615814,
+      "learning_rate": 1.0393635283250885e-06,
+      "loss": 0.4716,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8123502304147465,
+      "grad_norm": 0.3270639181137085,
+      "learning_rate": 1.0334781997799165e-06,
+      "loss": 0.4556,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8129032258064516,
+      "grad_norm": 0.3468005955219269,
+      "learning_rate": 1.0276076600829094e-06,
+      "loss": 0.4725,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8134562211981566,
+      "grad_norm": 0.31104719638824463,
+      "learning_rate": 1.0217519311218882e-06,
+      "loss": 0.4735,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8140092165898618,
+      "grad_norm": 0.30012935400009155,
+      "learning_rate": 1.0159110347294536e-06,
+      "loss": 0.4624,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8145622119815669,
+      "grad_norm": 0.3302370309829712,
+      "learning_rate": 1.0100849926829043e-06,
+      "loss": 0.4755,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8151152073732719,
+      "grad_norm": 0.3273729681968689,
+      "learning_rate": 1.0042738267041552e-06,
+      "loss": 0.4558,
+      "step": 1474
+    },
+    {
+      "epoch": 0.815668202764977,
+      "grad_norm": 0.313931405544281,
+      "learning_rate": 9.984775584596585e-07,
+      "loss": 0.4778,
+      "step": 1475
+    },
+    {
+      "epoch": 0.816221198156682,
+      "grad_norm": 0.31616419553756714,
+      "learning_rate": 9.926962095603199e-07,
+      "loss": 0.4358,
+      "step": 1476
+    },
+    {
+      "epoch": 0.8167741935483871,
+      "grad_norm": 0.33137091994285583,
+      "learning_rate": 9.869298015614198e-07,
+      "loss": 0.4558,
+      "step": 1477
+    },
+    {
+      "epoch": 0.8173271889400922,
+      "grad_norm": 0.3267715573310852,
+      "learning_rate": 9.811783559625343e-07,
+      "loss": 0.475,
+      "step": 1478
+    },
+    {
+      "epoch": 0.8178801843317972,
+      "grad_norm": 0.37252572178840637,
+      "learning_rate": 9.754418942074512e-07,
+      "loss": 0.4829,
+      "step": 1479
+    },
+    {
+      "epoch": 0.8184331797235023,
+      "grad_norm": 0.3060474395751953,
+      "learning_rate": 9.697204376840936e-07,
+      "loss": 0.4627,
+      "step": 1480
+    },
+    {
+      "epoch": 0.8189861751152073,
+      "grad_norm": 0.32007962465286255,
+      "learning_rate": 9.640140077244382e-07,
+      "loss": 0.4698,
+      "step": 1481
+    },
+    {
+      "epoch": 0.8195391705069125,
+      "grad_norm": 0.2968813478946686,
+      "learning_rate": 9.58322625604437e-07,
+      "loss": 0.4316,
+      "step": 1482
+    },
+    {
+      "epoch": 0.8200921658986176,
+      "grad_norm": 0.31622204184532166,
+      "learning_rate": 9.526463125439372e-07,
+      "loss": 0.4408,
+      "step": 1483
+    },
+    {
+      "epoch": 0.8206451612903226,
+      "grad_norm": 0.2871773838996887,
+      "learning_rate": 9.469850897066014e-07,
+      "loss": 0.4563,
+      "step": 1484
+    },
+    {
+      "epoch": 0.8211981566820277,
+      "grad_norm": 0.30365270376205444,
+      "learning_rate": 9.413389781998311e-07,
+      "loss": 0.4519,
+      "step": 1485
+    },
+    {
+      "epoch": 0.8217511520737327,
+      "grad_norm": 0.3301312029361725,
+      "learning_rate": 9.357079990746853e-07,
+      "loss": 0.4543,
+      "step": 1486
+    },
+    {
+      "epoch": 0.8223041474654378,
+      "grad_norm": 0.3503333628177643,
+      "learning_rate": 9.300921733258039e-07,
+      "loss": 0.4753,
+      "step": 1487
+    },
+    {
+      "epoch": 0.8228571428571428,
+      "grad_norm": 0.30572423338890076,
+      "learning_rate": 9.244915218913275e-07,
+      "loss": 0.4694,
+      "step": 1488
+    },
+    {
+      "epoch": 0.8234101382488479,
+      "grad_norm": 0.30958428978919983,
+      "learning_rate": 9.18906065652822e-07,
+      "loss": 0.4544,
+      "step": 1489
+    },
+    {
+      "epoch": 0.823963133640553,
+      "grad_norm": 0.3019687831401825,
+      "learning_rate": 9.133358254351982e-07,
+      "loss": 0.4751,
+      "step": 1490
+    },
+    {
+      "epoch": 0.824516129032258,
+      "grad_norm": 0.3120213747024536,
+      "learning_rate": 9.077808220066359e-07,
+      "loss": 0.4586,
+      "step": 1491
+    },
+    {
+      "epoch": 0.8250691244239632,
+      "grad_norm": 0.3264663815498352,
+      "learning_rate": 9.022410760785055e-07,
+      "loss": 0.4569,
+      "step": 1492
+    },
+    {
+      "epoch": 0.8256221198156682,
+      "grad_norm": 0.3079594075679779,
+      "learning_rate": 8.967166083052908e-07,
+      "loss": 0.5003,
+      "step": 1493
+    },
+    {
+      "epoch": 0.8261751152073733,
+      "grad_norm": 0.30771124362945557,
+      "learning_rate": 8.912074392845144e-07,
+      "loss": 0.4654,
+      "step": 1494
+    },
+    {
+      "epoch": 0.8267281105990784,
+      "grad_norm": 0.31957051157951355,
+      "learning_rate": 8.857135895566538e-07,
+      "loss": 0.4722,
+      "step": 1495
+    },
+    {
+      "epoch": 0.8272811059907834,
+      "grad_norm": 0.30836620926856995,
+      "learning_rate": 8.80235079605074e-07,
+      "loss": 0.4557,
+      "step": 1496
+    },
+    {
+      "epoch": 0.8278341013824885,
+      "grad_norm": 0.30665266513824463,
+      "learning_rate": 8.747719298559465e-07,
+      "loss": 0.4516,
+      "step": 1497
+    },
+    {
+      "epoch": 0.8283870967741935,
+      "grad_norm": 0.33325862884521484,
+      "learning_rate": 8.693241606781728e-07,
+      "loss": 0.4667,
+      "step": 1498
+    },
+    {
+      "epoch": 0.8289400921658986,
+      "grad_norm": 0.3295386731624603,
+      "learning_rate": 8.638917923833074e-07,
+      "loss": 0.4602,
+      "step": 1499
+    },
+    {
+      "epoch": 0.8294930875576036,
+      "grad_norm": 0.3496408462524414,
+      "learning_rate": 8.584748452254888e-07,
+      "loss": 0.4601,
+      "step": 1500
+    },
+    {
+      "epoch": 0.8300460829493087,
+      "grad_norm": 0.2922409474849701,
+      "learning_rate": 8.530733394013546e-07,
+      "loss": 0.4787,
+      "step": 1501
+    },
+    {
+      "epoch": 0.8305990783410139,
+      "grad_norm": 0.3221864104270935,
+      "learning_rate": 8.476872950499726e-07,
+      "loss": 0.4401,
+      "step": 1502
+    },
+    {
+      "epoch": 0.8311520737327189,
+      "grad_norm": 0.33632180094718933,
+      "learning_rate": 8.423167322527626e-07,
+      "loss": 0.4636,
+      "step": 1503
+    },
+    {
+      "epoch": 0.831705069124424,
+      "grad_norm": 0.3501831293106079,
+      "learning_rate": 8.369616710334233e-07,
+      "loss": 0.4573,
+      "step": 1504
+    },
+    {
+      "epoch": 0.832258064516129,
+      "grad_norm": 0.31366902589797974,
+      "learning_rate": 8.316221313578576e-07,
+      "loss": 0.4583,
+      "step": 1505
+    },
+    {
+      "epoch": 0.8328110599078341,
+      "grad_norm": 0.36373165249824524,
+      "learning_rate": 8.262981331340969e-07,
+      "loss": 0.4633,
+      "step": 1506
+    },
+    {
+      "epoch": 0.8333640552995392,
+      "grad_norm": 0.31518691778182983,
+      "learning_rate": 8.209896962122282e-07,
+      "loss": 0.4599,
+      "step": 1507
+    },
+    {
+      "epoch": 0.8339170506912442,
+      "grad_norm": 0.3142349123954773,
+      "learning_rate": 8.15696840384319e-07,
+      "loss": 0.465,
+      "step": 1508
+    },
+    {
+      "epoch": 0.8344700460829493,
+      "grad_norm": 0.3050268292427063,
+      "learning_rate": 8.104195853843433e-07,
+      "loss": 0.4617,
+      "step": 1509
+    },
+    {
+      "epoch": 0.8350230414746543,
+      "grad_norm": 0.3328079879283905,
+      "learning_rate": 8.051579508881107e-07,
+      "loss": 0.4558,
+      "step": 1510
+    },
+    {
+      "epoch": 0.8355760368663594,
+      "grad_norm": 0.3367290496826172,
+      "learning_rate": 7.999119565131891e-07,
+      "loss": 0.4662,
+      "step": 1511
+    },
+    {
+      "epoch": 0.8361290322580646,
+      "grad_norm": 0.2858041226863861,
+      "learning_rate": 7.946816218188347e-07,
+      "loss": 0.4645,
+      "step": 1512
+    },
+    {
+      "epoch": 0.8366820276497696,
+      "grad_norm": 0.3110063374042511,
+      "learning_rate": 7.894669663059168e-07,
+      "loss": 0.4616,
+      "step": 1513
+    },
+    {
+      "epoch": 0.8372350230414747,
+      "grad_norm": 0.3313262462615967,
+      "learning_rate": 7.84268009416847e-07,
+      "loss": 0.4423,
+      "step": 1514
+    },
+    {
+      "epoch": 0.8377880184331797,
+      "grad_norm": 0.33478546142578125,
+      "learning_rate": 7.790847705355059e-07,
+      "loss": 0.4551,
+      "step": 1515
+    },
+    {
+      "epoch": 0.8383410138248848,
+      "grad_norm": 0.3241022527217865,
+      "learning_rate": 7.739172689871705e-07,
+      "loss": 0.4749,
+      "step": 1516
+    },
+    {
+      "epoch": 0.8388940092165899,
+      "grad_norm": 0.30135831236839294,
+      "learning_rate": 7.687655240384423e-07,
+      "loss": 0.4432,
+      "step": 1517
+    },
+    {
+      "epoch": 0.8394470046082949,
+      "grad_norm": 0.3093847930431366,
+      "learning_rate": 7.636295548971762e-07,
+      "loss": 0.4398,
+      "step": 1518
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.31386277079582214,
+      "learning_rate": 7.585093807124077e-07,
+      "loss": 0.4308,
+      "step": 1519
+    },
+    {
+      "epoch": 0.840552995391705,
+      "grad_norm": 0.32683154940605164,
+      "learning_rate": 7.534050205742827e-07,
+      "loss": 0.4671,
+      "step": 1520
+    },
+    {
+      "epoch": 0.8411059907834101,
+      "grad_norm": 0.32057997584342957,
+      "learning_rate": 7.483164935139847e-07,
+      "loss": 0.4777,
+      "step": 1521
+    },
+    {
+      "epoch": 0.8416589861751153,
+      "grad_norm": 0.3294847011566162,
+      "learning_rate": 7.432438185036667e-07,
+      "loss": 0.4872,
+      "step": 1522
+    },
+    {
+      "epoch": 0.8422119815668203,
+      "grad_norm": 0.3206014931201935,
+      "learning_rate": 7.381870144563763e-07,
+      "loss": 0.4719,
+      "step": 1523
+    },
+    {
+      "epoch": 0.8427649769585254,
+      "grad_norm": 0.291071355342865,
+      "learning_rate": 7.33146100225991e-07,
+      "loss": 0.4496,
+      "step": 1524
+    },
+    {
+      "epoch": 0.8433179723502304,
+      "grad_norm": 0.31741753220558167,
+      "learning_rate": 7.281210946071393e-07,
+      "loss": 0.4305,
+      "step": 1525
+    },
+    {
+      "epoch": 0.8438709677419355,
+      "grad_norm": 0.3081185817718506,
+      "learning_rate": 7.231120163351396e-07,
+      "loss": 0.4796,
+      "step": 1526
+    },
+    {
+      "epoch": 0.8444239631336405,
+      "grad_norm": 0.31715425848960876,
+      "learning_rate": 7.181188840859266e-07,
+      "loss": 0.4827,
+      "step": 1527
+    },
+    {
+      "epoch": 0.8449769585253456,
+      "grad_norm": 0.3110186755657196,
+      "learning_rate": 7.131417164759791e-07,
+      "loss": 0.4652,
+      "step": 1528
+    },
+    {
+      "epoch": 0.8455299539170507,
+      "grad_norm": 0.3567166030406952,
+      "learning_rate": 7.081805320622559e-07,
+      "loss": 0.485,
+      "step": 1529
+    },
+    {
+      "epoch": 0.8460829493087557,
+      "grad_norm": 0.3137590289115906,
+      "learning_rate": 7.032353493421213e-07,
+      "loss": 0.445,
+      "step": 1530
+    },
+    {
+      "epoch": 0.8466359447004608,
+      "grad_norm": 0.3127385675907135,
+      "learning_rate": 6.983061867532798e-07,
+      "loss": 0.4691,
+      "step": 1531
+    },
+    {
+      "epoch": 0.8471889400921659,
+      "grad_norm": 0.3255513608455658,
+      "learning_rate": 6.933930626737057e-07,
+      "loss": 0.4418,
+      "step": 1532
+    },
+    {
+      "epoch": 0.847741935483871,
+      "grad_norm": 0.2911495566368103,
+      "learning_rate": 6.88495995421577e-07,
+      "loss": 0.4578,
+      "step": 1533
+    },
+    {
+      "epoch": 0.8482949308755761,
+      "grad_norm": 0.29997825622558594,
+      "learning_rate": 6.83615003255203e-07,
+      "loss": 0.4543,
+      "step": 1534
+    },
+    {
+      "epoch": 0.8488479262672811,
+      "grad_norm": 0.2882632315158844,
+      "learning_rate": 6.787501043729577e-07,
+      "loss": 0.4531,
+      "step": 1535
+    },
+    {
+      "epoch": 0.8494009216589862,
+      "grad_norm": 0.32424575090408325,
+      "learning_rate": 6.739013169132153e-07,
+      "loss": 0.4516,
+      "step": 1536
+    },
+    {
+      "epoch": 0.8499539170506912,
+      "grad_norm": 0.3189219534397125,
+      "learning_rate": 6.690686589542778e-07,
+      "loss": 0.456,
+      "step": 1537
+    },
+    {
+      "epoch": 0.8505069124423963,
+      "grad_norm": 0.2835007309913635,
+      "learning_rate": 6.642521485143099e-07,
+      "loss": 0.4874,
+      "step": 1538
+    },
+    {
+      "epoch": 0.8510599078341013,
+      "grad_norm": 0.30130714178085327,
+      "learning_rate": 6.594518035512726e-07,
+      "loss": 0.4698,
+      "step": 1539
+    },
+    {
+      "epoch": 0.8516129032258064,
+      "grad_norm": 0.3199770748615265,
+      "learning_rate": 6.546676419628545e-07,
+      "loss": 0.4822,
+      "step": 1540
+    },
+    {
+      "epoch": 0.8521658986175116,
+      "grad_norm": 0.31458160281181335,
+      "learning_rate": 6.498996815864068e-07,
+      "loss": 0.4737,
+      "step": 1541
+    },
+    {
+      "epoch": 0.8527188940092166,
+      "grad_norm": 0.33827102184295654,
+      "learning_rate": 6.451479401988736e-07,
+      "loss": 0.4607,
+      "step": 1542
+    },
+    {
+      "epoch": 0.8532718894009217,
+      "grad_norm": 0.3004834055900574,
+      "learning_rate": 6.404124355167302e-07,
+      "loss": 0.4621,
+      "step": 1543
+    },
+    {
+      "epoch": 0.8538248847926267,
+      "grad_norm": 0.2976799011230469,
+      "learning_rate": 6.356931851959136e-07,
+      "loss": 0.4571,
+      "step": 1544
+    },
+    {
+      "epoch": 0.8543778801843318,
+      "grad_norm": 0.31761622428894043,
+      "learning_rate": 6.30990206831758e-07,
+      "loss": 0.4647,
+      "step": 1545
+    },
+    {
+      "epoch": 0.8549308755760369,
+      "grad_norm": 0.30549031496047974,
+      "learning_rate": 6.263035179589288e-07,
+      "loss": 0.4619,
+      "step": 1546
+    },
+    {
+      "epoch": 0.8554838709677419,
+      "grad_norm": 0.3004506528377533,
+      "learning_rate": 6.216331360513572e-07,
+      "loss": 0.4786,
+      "step": 1547
+    },
+    {
+      "epoch": 0.856036866359447,
+      "grad_norm": 0.3054359257221222,
+      "learning_rate": 6.169790785221763e-07,
+      "loss": 0.4652,
+      "step": 1548
+    },
+    {
+      "epoch": 0.856589861751152,
+      "grad_norm": 0.3147340416908264,
+      "learning_rate": 6.123413627236536e-07,
+      "loss": 0.4573,
+      "step": 1549
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.3203125,
+      "learning_rate": 6.077200059471289e-07,
+      "loss": 0.4498,
+      "step": 1550
+    },
+    {
+      "epoch": 0.8576958525345623,
+      "grad_norm": 0.3389976918697357,
+      "learning_rate": 6.031150254229484e-07,
+      "loss": 0.4756,
+      "step": 1551
+    },
+    {
+      "epoch": 0.8582488479262673,
+      "grad_norm": 0.3188997507095337,
+      "learning_rate": 5.985264383204003e-07,
+      "loss": 0.4579,
+      "step": 1552
+    },
+    {
+      "epoch": 0.8588018433179724,
+      "grad_norm": 0.2967085540294647,
+      "learning_rate": 5.939542617476529e-07,
+      "loss": 0.4648,
+      "step": 1553
+    },
+    {
+      "epoch": 0.8593548387096774,
+      "grad_norm": 0.2970874011516571,
+      "learning_rate": 5.893985127516866e-07,
+      "loss": 0.4654,
+      "step": 1554
+    },
+    {
+      "epoch": 0.8599078341013825,
+      "grad_norm": 0.3088582158088684,
+      "learning_rate": 5.848592083182348e-07,
+      "loss": 0.4444,
+      "step": 1555
+    },
+    {
+      "epoch": 0.8604608294930876,
+      "grad_norm": 0.3454407751560211,
+      "learning_rate": 5.803363653717187e-07,
+      "loss": 0.4796,
+      "step": 1556
+    },
+    {
+      "epoch": 0.8610138248847926,
+      "grad_norm": 0.34124165773391724,
+      "learning_rate": 5.758300007751832e-07,
+      "loss": 0.475,
+      "step": 1557
+    },
+    {
+      "epoch": 0.8615668202764977,
+      "grad_norm": 0.31380608677864075,
+      "learning_rate": 5.713401313302358e-07,
+      "loss": 0.4714,
+      "step": 1558
+    },
+    {
+      "epoch": 0.8621198156682027,
+      "grad_norm": 0.3094062805175781,
+      "learning_rate": 5.668667737769834e-07,
+      "loss": 0.4348,
+      "step": 1559
+    },
+    {
+      "epoch": 0.8626728110599078,
+      "grad_norm": 0.3069341778755188,
+      "learning_rate": 5.624099447939696e-07,
+      "loss": 0.4677,
+      "step": 1560
+    },
+    {
+      "epoch": 0.863225806451613,
+      "grad_norm": 0.34499165415763855,
+      "learning_rate": 5.579696609981117e-07,
+      "loss": 0.4673,
+      "step": 1561
+    },
+    {
+      "epoch": 0.863778801843318,
+      "grad_norm": 0.32358747720718384,
+      "learning_rate": 5.535459389446401e-07,
+      "loss": 0.4627,
+      "step": 1562
+    },
+    {
+      "epoch": 0.8643317972350231,
+      "grad_norm": 0.33145248889923096,
+      "learning_rate": 5.491387951270366e-07,
+      "loss": 0.4535,
+      "step": 1563
+    },
+    {
+      "epoch": 0.8648847926267281,
+      "grad_norm": 0.3262435495853424,
+      "learning_rate": 5.447482459769709e-07,
+      "loss": 0.4613,
+      "step": 1564
+    },
+    {
+      "epoch": 0.8654377880184332,
+      "grad_norm": 0.31697559356689453,
+      "learning_rate": 5.403743078642443e-07,
+      "loss": 0.4503,
+      "step": 1565
+    },
+    {
+      "epoch": 0.8659907834101382,
+      "grad_norm": 0.3151688873767853,
+      "learning_rate": 5.360169970967221e-07,
+      "loss": 0.4816,
+      "step": 1566
+    },
+    {
+      "epoch": 0.8665437788018433,
+      "grad_norm": 0.3067736029624939,
+      "learning_rate": 5.316763299202766e-07,
+      "loss": 0.4631,
+      "step": 1567
+    },
+    {
+      "epoch": 0.8670967741935484,
+      "grad_norm": 0.2979617714881897,
+      "learning_rate": 5.273523225187255e-07,
+      "loss": 0.4403,
+      "step": 1568
+    },
+    {
+      "epoch": 0.8676497695852534,
+      "grad_norm": 0.3092879056930542,
+      "learning_rate": 5.23044991013773e-07,
+      "loss": 0.4669,
+      "step": 1569
+    },
+    {
+      "epoch": 0.8682027649769585,
+      "grad_norm": 0.29821327328681946,
+      "learning_rate": 5.187543514649479e-07,
+      "loss": 0.4678,
+      "step": 1570
+    },
+    {
+      "epoch": 0.8687557603686636,
+      "grad_norm": 0.3069091737270355,
+      "learning_rate": 5.144804198695447e-07,
+      "loss": 0.4834,
+      "step": 1571
+    },
+    {
+      "epoch": 0.8693087557603687,
+      "grad_norm": 0.3062274754047394,
+      "learning_rate": 5.102232121625633e-07,
+      "loss": 0.4583,
+      "step": 1572
+    },
+    {
+      "epoch": 0.8698617511520738,
+      "grad_norm": 0.3618842661380768,
+      "learning_rate": 5.059827442166504e-07,
+      "loss": 0.4348,
+      "step": 1573
+    },
+    {
+      "epoch": 0.8704147465437788,
+      "grad_norm": 0.3038586676120758,
+      "learning_rate": 5.017590318420395e-07,
+      "loss": 0.4601,
+      "step": 1574
+    },
+    {
+      "epoch": 0.8709677419354839,
+      "grad_norm": 0.30661341547966003,
+      "learning_rate": 4.975520907864928e-07,
+      "loss": 0.4432,
+      "step": 1575
+    },
+    {
+      "epoch": 0.8715207373271889,
+      "grad_norm": 0.3259396255016327,
+      "learning_rate": 4.93361936735241e-07,
+      "loss": 0.4602,
+      "step": 1576
+    },
+    {
+      "epoch": 0.872073732718894,
+      "grad_norm": 0.32900482416152954,
+      "learning_rate": 4.891885853109279e-07,
+      "loss": 0.473,
+      "step": 1577
+    },
+    {
+      "epoch": 0.872626728110599,
+      "grad_norm": 0.280129075050354,
+      "learning_rate": 4.85032052073548e-07,
+      "loss": 0.4581,
+      "step": 1578
+    },
+    {
+      "epoch": 0.8731797235023041,
+      "grad_norm": 0.31960612535476685,
+      "learning_rate": 4.808923525203912e-07,
+      "loss": 0.4472,
+      "step": 1579
+    },
+    {
+      "epoch": 0.8737327188940092,
+      "grad_norm": 0.28361162543296814,
+      "learning_rate": 4.767695020859847e-07,
+      "loss": 0.4538,
+      "step": 1580
+    },
+    {
+      "epoch": 0.8742857142857143,
+      "grad_norm": 0.3138439953327179,
+      "learning_rate": 4.726635161420351e-07,
+      "loss": 0.4746,
+      "step": 1581
+    },
+    {
+      "epoch": 0.8748387096774194,
+      "grad_norm": 0.30272024869918823,
+      "learning_rate": 4.685744099973716e-07,
+      "loss": 0.4713,
+      "step": 1582
+    },
+    {
+      "epoch": 0.8753917050691244,
+      "grad_norm": 0.31137874722480774,
+      "learning_rate": 4.6450219889788816e-07,
+      "loss": 0.4569,
+      "step": 1583
+    },
+    {
+      "epoch": 0.8759447004608295,
+      "grad_norm": 0.34966525435447693,
+      "learning_rate": 4.6044689802648534e-07,
+      "loss": 0.4646,
+      "step": 1584
+    },
+    {
+      "epoch": 0.8764976958525346,
+      "grad_norm": 0.29577386379241943,
+      "learning_rate": 4.564085225030174e-07,
+      "loss": 0.4642,
+      "step": 1585
+    },
+    {
+      "epoch": 0.8770506912442396,
+      "grad_norm": 0.2946975529193878,
+      "learning_rate": 4.523870873842329e-07,
+      "loss": 0.4312,
+      "step": 1586
+    },
+    {
+      "epoch": 0.8776036866359447,
+      "grad_norm": 0.30097469687461853,
+      "learning_rate": 4.4838260766372044e-07,
+      "loss": 0.4666,
+      "step": 1587
+    },
+    {
+      "epoch": 0.8781566820276497,
+      "grad_norm": 0.2847994565963745,
+      "learning_rate": 4.4439509827185034e-07,
+      "loss": 0.4165,
+      "step": 1588
+    },
+    {
+      "epoch": 0.8787096774193548,
+      "grad_norm": 0.2944558560848236,
+      "learning_rate": 4.4042457407572235e-07,
+      "loss": 0.4721,
+      "step": 1589
+    },
+    {
+      "epoch": 0.8792626728110599,
+      "grad_norm": 0.3086279034614563,
+      "learning_rate": 4.3647104987910636e-07,
+      "loss": 0.45,
+      "step": 1590
+    },
+    {
+      "epoch": 0.879815668202765,
+      "grad_norm": 0.3032161593437195,
+      "learning_rate": 4.3253454042239016e-07,
+      "loss": 0.458,
+      "step": 1591
+    },
+    {
+      "epoch": 0.8803686635944701,
+      "grad_norm": 0.3120705485343933,
+      "learning_rate": 4.286150603825234e-07,
+      "loss": 0.474,
+      "step": 1592
+    },
+    {
+      "epoch": 0.8809216589861751,
+      "grad_norm": 0.332922101020813,
+      "learning_rate": 4.2471262437296326e-07,
+      "loss": 0.4702,
+      "step": 1593
+    },
+    {
+      "epoch": 0.8814746543778802,
+      "grad_norm": 0.2810650169849396,
+      "learning_rate": 4.208272469436192e-07,
+      "loss": 0.4509,
+      "step": 1594
+    },
+    {
+      "epoch": 0.8820276497695853,
+      "grad_norm": 0.31393900513648987,
+      "learning_rate": 4.1695894258079903e-07,
+      "loss": 0.4734,
+      "step": 1595
+    },
+    {
+      "epoch": 0.8825806451612903,
+      "grad_norm": 0.2897551357746124,
+      "learning_rate": 4.1310772570715586e-07,
+      "loss": 0.4442,
+      "step": 1596
+    },
+    {
+      "epoch": 0.8831336405529954,
+      "grad_norm": 0.3010689616203308,
+      "learning_rate": 4.092736106816314e-07,
+      "loss": 0.4477,
+      "step": 1597
+    },
+    {
+      "epoch": 0.8836866359447004,
+      "grad_norm": 0.29473552107810974,
+      "learning_rate": 4.0545661179940854e-07,
+      "loss": 0.4595,
+      "step": 1598
+    },
+    {
+      "epoch": 0.8842396313364055,
+      "grad_norm": 0.3020123839378357,
+      "learning_rate": 4.0165674329185055e-07,
+      "loss": 0.461,
+      "step": 1599
+    },
+    {
+      "epoch": 0.8847926267281107,
+      "grad_norm": 0.3016417920589447,
+      "learning_rate": 3.978740193264524e-07,
+      "loss": 0.4802,
+      "step": 1600
+    },
+    {
+      "epoch": 0.8853456221198157,
+      "grad_norm": 0.3054351806640625,
+      "learning_rate": 3.941084540067874e-07,
+      "loss": 0.461,
+      "step": 1601
+    },
+    {
+      "epoch": 0.8858986175115208,
+      "grad_norm": 0.31593799591064453,
+      "learning_rate": 3.9036006137245396e-07,
+      "loss": 0.485,
+      "step": 1602
+    },
+    {
+      "epoch": 0.8864516129032258,
+      "grad_norm": 0.2995027005672455,
+      "learning_rate": 3.866288553990233e-07,
+      "loss": 0.4544,
+      "step": 1603
+    },
+    {
+      "epoch": 0.8870046082949309,
+      "grad_norm": 0.3163018524646759,
+      "learning_rate": 3.82914849997989e-07,
+      "loss": 0.4675,
+      "step": 1604
+    },
+    {
+      "epoch": 0.8875576036866359,
+      "grad_norm": 0.3466692268848419,
+      "learning_rate": 3.7921805901671273e-07,
+      "loss": 0.455,
+      "step": 1605
+    },
+    {
+      "epoch": 0.888110599078341,
+      "grad_norm": 0.30836209654808044,
+      "learning_rate": 3.7553849623837237e-07,
+      "loss": 0.4983,
+      "step": 1606
+    },
+    {
+      "epoch": 0.8886635944700461,
+      "grad_norm": 0.3018787205219269,
+      "learning_rate": 3.7187617538191446e-07,
+      "loss": 0.4714,
+      "step": 1607
+    },
+    {
+      "epoch": 0.8892165898617511,
+      "grad_norm": 0.3216530382633209,
+      "learning_rate": 3.682311101019981e-07,
+      "loss": 0.4857,
+      "step": 1608
+    },
+    {
+      "epoch": 0.8897695852534562,
+      "grad_norm": 0.32293516397476196,
+      "learning_rate": 3.646033139889482e-07,
+      "loss": 0.4508,
+      "step": 1609
+    },
+    {
+      "epoch": 0.8903225806451613,
+      "grad_norm": 0.29628053307533264,
+      "learning_rate": 3.6099280056870136e-07,
+      "loss": 0.4271,
+      "step": 1610
+    },
+    {
+      "epoch": 0.8908755760368664,
+      "grad_norm": 0.30662253499031067,
+      "learning_rate": 3.5739958330275936e-07,
+      "loss": 0.4833,
+      "step": 1611
+    },
+    {
+      "epoch": 0.8914285714285715,
+      "grad_norm": 0.30339956283569336,
+      "learning_rate": 3.538236755881341e-07,
+      "loss": 0.4675,
+      "step": 1612
+    },
+    {
+      "epoch": 0.8919815668202765,
+      "grad_norm": 0.29642388224601746,
+      "learning_rate": 3.502650907573024e-07,
+      "loss": 0.4529,
+      "step": 1613
+    },
+    {
+      "epoch": 0.8925345622119816,
+      "grad_norm": 0.3087359666824341,
+      "learning_rate": 3.467238420781527e-07,
+      "loss": 0.4794,
+      "step": 1614
+    },
+    {
+      "epoch": 0.8930875576036866,
+      "grad_norm": 0.34186214208602905,
+      "learning_rate": 3.431999427539368e-07,
+      "loss": 0.4951,
+      "step": 1615
+    },
+    {
+      "epoch": 0.8936405529953917,
+      "grad_norm": 0.31149131059646606,
+      "learning_rate": 3.396934059232226e-07,
+      "loss": 0.4476,
+      "step": 1616
+    },
+    {
+      "epoch": 0.8941935483870967,
+      "grad_norm": 0.3173343241214752,
+      "learning_rate": 3.362042446598418e-07,
+      "loss": 0.4735,
+      "step": 1617
+    },
+    {
+      "epoch": 0.8947465437788018,
+      "grad_norm": 0.29687947034835815,
+      "learning_rate": 3.3273247197284366e-07,
+      "loss": 0.4588,
+      "step": 1618
+    },
+    {
+      "epoch": 0.8952995391705069,
+      "grad_norm": 0.29158180952072144,
+      "learning_rate": 3.292781008064455e-07,
+      "loss": 0.4477,
+      "step": 1619
+    },
+    {
+      "epoch": 0.895852534562212,
+      "grad_norm": 0.31376034021377563,
+      "learning_rate": 3.258411440399839e-07,
+      "loss": 0.4613,
+      "step": 1620
+    },
+    {
+      "epoch": 0.8964055299539171,
+      "grad_norm": 0.3553878366947174,
+      "learning_rate": 3.2242161448786724e-07,
+      "loss": 0.4638,
+      "step": 1621
+    },
+    {
+      "epoch": 0.8969585253456221,
+      "grad_norm": 0.31276994943618774,
+      "learning_rate": 3.1901952489952927e-07,
+      "loss": 0.4297,
+      "step": 1622
+    },
+    {
+      "epoch": 0.8975115207373272,
+      "grad_norm": 0.2925131320953369,
+      "learning_rate": 3.15634887959379e-07,
+      "loss": 0.4642,
+      "step": 1623
+    },
+    {
+      "epoch": 0.8980645161290323,
+      "grad_norm": 0.3111323416233063,
+      "learning_rate": 3.122677162867549e-07,
+      "loss": 0.4628,
+      "step": 1624
+    },
+    {
+      "epoch": 0.8986175115207373,
+      "grad_norm": 0.28449130058288574,
+      "learning_rate": 3.089180224358773e-07,
+      "loss": 0.4619,
+      "step": 1625
+    },
+    {
+      "epoch": 0.8991705069124424,
+      "grad_norm": 0.2895408868789673,
+      "learning_rate": 3.0558581889580383e-07,
+      "loss": 0.4486,
+      "step": 1626
+    },
+    {
+      "epoch": 0.8997235023041474,
+      "grad_norm": 0.28121551871299744,
+      "learning_rate": 3.0227111809037815e-07,
+      "loss": 0.4523,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9002764976958525,
+      "grad_norm": 0.3156479001045227,
+      "learning_rate": 2.9897393237818873e-07,
+      "loss": 0.4504,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9008294930875576,
+      "grad_norm": 0.30593904852867126,
+      "learning_rate": 2.956942740525187e-07,
+      "loss": 0.4832,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9013824884792627,
+      "grad_norm": 0.2923428416252136,
+      "learning_rate": 2.924321553413029e-07,
+      "loss": 0.4473,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9019354838709678,
+      "grad_norm": 0.29185551404953003,
+      "learning_rate": 2.891875884070816e-07,
+      "loss": 0.4514,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9024884792626728,
+      "grad_norm": 0.3090588450431824,
+      "learning_rate": 2.859605853469533e-07,
+      "loss": 0.492,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9030414746543779,
+      "grad_norm": 0.2918623387813568,
+      "learning_rate": 2.8275115819253184e-07,
+      "loss": 0.4378,
+      "step": 1633
+    },
+    {
+      "epoch": 0.903594470046083,
+      "grad_norm": 0.2900620698928833,
+      "learning_rate": 2.795593189099016e-07,
+      "loss": 0.4702,
+      "step": 1634
+    },
+    {
+      "epoch": 0.904147465437788,
+      "grad_norm": 0.30443379282951355,
+      "learning_rate": 2.7638507939956926e-07,
+      "loss": 0.4562,
+      "step": 1635
+    },
+    {
+      "epoch": 0.9047004608294931,
+      "grad_norm": 0.2894691526889801,
+      "learning_rate": 2.7322845149642564e-07,
+      "loss": 0.4405,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9052534562211981,
+      "grad_norm": 0.30414119362831116,
+      "learning_rate": 2.700894469696963e-07,
+      "loss": 0.4618,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9058064516129032,
+      "grad_norm": 0.29795682430267334,
+      "learning_rate": 2.669680775228989e-07,
+      "loss": 0.4717,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9063594470046082,
+      "grad_norm": 0.2925158739089966,
+      "learning_rate": 2.638643547938019e-07,
+      "loss": 0.4788,
+      "step": 1639
+    },
+    {
+      "epoch": 0.9069124423963134,
+      "grad_norm": 0.29635146260261536,
+      "learning_rate": 2.607782903543782e-07,
+      "loss": 0.4521,
+      "step": 1640
+    },
+    {
+      "epoch": 0.9074654377880185,
+      "grad_norm": 0.29401543736457825,
+      "learning_rate": 2.5770989571076375e-07,
+      "loss": 0.4401,
+      "step": 1641
+    },
+    {
+      "epoch": 0.9080184331797235,
+      "grad_norm": 0.2778622508049011,
+      "learning_rate": 2.5465918230321464e-07,
+      "loss": 0.4522,
+      "step": 1642
+    },
+    {
+      "epoch": 0.9085714285714286,
+      "grad_norm": 0.2870117127895355,
+      "learning_rate": 2.5162616150606167e-07,
+      "loss": 0.4686,
+      "step": 1643
+    },
+    {
+      "epoch": 0.9091244239631336,
+      "grad_norm": 0.3375545144081116,
+      "learning_rate": 2.486108446276725e-07,
+      "loss": 0.4741,
+      "step": 1644
+    },
+    {
+      "epoch": 0.9096774193548387,
+      "grad_norm": 0.285396546125412,
+      "learning_rate": 2.4561324291040636e-07,
+      "loss": 0.4749,
+      "step": 1645
+    },
+    {
+      "epoch": 0.9102304147465438,
+      "grad_norm": 0.31609663367271423,
+      "learning_rate": 2.4263336753057364e-07,
+      "loss": 0.4552,
+      "step": 1646
+    },
+    {
+      "epoch": 0.9107834101382488,
+      "grad_norm": 0.32548245787620544,
+      "learning_rate": 2.396712295983922e-07,
+      "loss": 0.4707,
+      "step": 1647
+    },
+    {
+      "epoch": 0.9113364055299539,
+      "grad_norm": 0.28127720952033997,
+      "learning_rate": 2.3672684015794923e-07,
+      "loss": 0.4808,
+      "step": 1648
+    },
+    {
+      "epoch": 0.9118894009216589,
+      "grad_norm": 0.3091859817504883,
+      "learning_rate": 2.338002101871556e-07,
+      "loss": 0.4469,
+      "step": 1649
+    },
+    {
+      "epoch": 0.9124423963133641,
+      "grad_norm": 0.3138236403465271,
+      "learning_rate": 2.3089135059771007e-07,
+      "loss": 0.4848,
+      "step": 1650
+    },
+    {
+      "epoch": 0.9129953917050692,
+      "grad_norm": 0.30295220017433167,
+      "learning_rate": 2.2800027223505438e-07,
+      "loss": 0.4567,
+      "step": 1651
+    },
+    {
+      "epoch": 0.9135483870967742,
+      "grad_norm": 0.3260992467403412,
+      "learning_rate": 2.2512698587833502e-07,
+      "loss": 0.4776,
+      "step": 1652
+    },
+    {
+      "epoch": 0.9141013824884793,
+      "grad_norm": 0.30410924553871155,
+      "learning_rate": 2.222715022403621e-07,
+      "loss": 0.4565,
+      "step": 1653
+    },
+    {
+      "epoch": 0.9146543778801843,
+      "grad_norm": 0.296068400144577,
+      "learning_rate": 2.1943383196756984e-07,
+      "loss": 0.4868,
+      "step": 1654
+    },
+    {
+      "epoch": 0.9152073732718894,
+      "grad_norm": 0.3140951991081238,
+      "learning_rate": 2.1661398563997737e-07,
+      "loss": 0.4645,
+      "step": 1655
+    },
+    {
+      "epoch": 0.9157603686635944,
+      "grad_norm": 0.2961066663265228,
+      "learning_rate": 2.138119737711475e-07,
+      "loss": 0.4376,
+      "step": 1656
+    },
+    {
+      "epoch": 0.9163133640552995,
+      "grad_norm": 0.3166109621524811,
+      "learning_rate": 2.110278068081506e-07,
+      "loss": 0.4713,
+      "step": 1657
+    },
+    {
+      "epoch": 0.9168663594470046,
+      "grad_norm": 0.32516512274742126,
+      "learning_rate": 2.082614951315215e-07,
+      "loss": 0.4516,
+      "step": 1658
+    },
+    {
+      "epoch": 0.9174193548387096,
+      "grad_norm": 0.293089896440506,
+      "learning_rate": 2.055130490552254e-07,
+      "loss": 0.469,
+      "step": 1659
+    },
+    {
+      "epoch": 0.9179723502304148,
+      "grad_norm": 0.31795766949653625,
+      "learning_rate": 2.0278247882661584e-07,
+      "loss": 0.4497,
+      "step": 1660
+    },
+    {
+      "epoch": 0.9185253456221198,
+      "grad_norm": 0.2847367525100708,
+      "learning_rate": 2.0006979462639686e-07,
+      "loss": 0.4447,
+      "step": 1661
+    },
+    {
+      "epoch": 0.9190783410138249,
+      "grad_norm": 0.2693861126899719,
+      "learning_rate": 1.9737500656858754e-07,
+      "loss": 0.4406,
+      "step": 1662
+    },
+    {
+      "epoch": 0.91963133640553,
+      "grad_norm": 0.3001803755760193,
+      "learning_rate": 1.94698124700482e-07,
+      "loss": 0.4964,
+      "step": 1663
+    },
+    {
+      "epoch": 0.920184331797235,
+      "grad_norm": 0.3073177933692932,
+      "learning_rate": 1.9203915900261327e-07,
+      "loss": 0.4737,
+      "step": 1664
+    },
+    {
+      "epoch": 0.9207373271889401,
+      "grad_norm": 0.31089356541633606,
+      "learning_rate": 1.8939811938871456e-07,
+      "loss": 0.4539,
+      "step": 1665
+    },
+    {
+      "epoch": 0.9212903225806451,
+      "grad_norm": 0.27542582154273987,
+      "learning_rate": 1.86775015705683e-07,
+      "loss": 0.434,
+      "step": 1666
+    },
+    {
+      "epoch": 0.9218433179723502,
+      "grad_norm": 0.32208341360092163,
+      "learning_rate": 1.841698577335438e-07,
+      "loss": 0.4476,
+      "step": 1667
+    },
+    {
+      "epoch": 0.9223963133640553,
+      "grad_norm": 0.3119848668575287,
+      "learning_rate": 1.8158265518541274e-07,
+      "loss": 0.4484,
+      "step": 1668
+    },
+    {
+      "epoch": 0.9229493087557604,
+      "grad_norm": 0.2974831759929657,
+      "learning_rate": 1.790134177074604e-07,
+      "loss": 0.4574,
+      "step": 1669
+    },
+    {
+      "epoch": 0.9235023041474655,
+      "grad_norm": 0.29929405450820923,
+      "learning_rate": 1.7646215487887587e-07,
+      "loss": 0.4685,
+      "step": 1670
+    },
+    {
+      "epoch": 0.9240552995391705,
+      "grad_norm": 0.2898726165294647,
+      "learning_rate": 1.739288762118313e-07,
+      "loss": 0.455,
+      "step": 1671
+    },
+    {
+      "epoch": 0.9246082949308756,
+      "grad_norm": 0.3034640848636627,
+      "learning_rate": 1.714135911514475e-07,
+      "loss": 0.4541,
+      "step": 1672
+    },
+    {
+      "epoch": 0.9251612903225807,
+      "grad_norm": 0.3259904682636261,
+      "learning_rate": 1.6891630907575562e-07,
+      "loss": 0.4678,
+      "step": 1673
+    },
+    {
+      "epoch": 0.9257142857142857,
+      "grad_norm": 0.3421642780303955,
+      "learning_rate": 1.664370392956649e-07,
+      "loss": 0.4518,
+      "step": 1674
+    },
+    {
+      "epoch": 0.9262672811059908,
+      "grad_norm": 0.29688096046447754,
+      "learning_rate": 1.6397579105492778e-07,
+      "loss": 0.4222,
+      "step": 1675
+    },
+    {
+      "epoch": 0.9268202764976958,
+      "grad_norm": 0.3092136085033417,
+      "learning_rate": 1.6153257353010433e-07,
+      "loss": 0.4679,
+      "step": 1676
+    },
+    {
+      "epoch": 0.9273732718894009,
+      "grad_norm": 0.3253113627433777,
+      "learning_rate": 1.5910739583053002e-07,
+      "loss": 0.4495,
+      "step": 1677
+    },
+    {
+      "epoch": 0.9279262672811059,
+      "grad_norm": 0.286625474691391,
+      "learning_rate": 1.5670026699827757e-07,
+      "loss": 0.4652,
+      "step": 1678
+    },
+    {
+      "epoch": 0.9284792626728111,
+      "grad_norm": 0.2905667722225189,
+      "learning_rate": 1.5431119600812837e-07,
+      "loss": 0.4529,
+      "step": 1679
+    },
+    {
+      "epoch": 0.9290322580645162,
+      "grad_norm": 0.32414594292640686,
+      "learning_rate": 1.5194019176753615e-07,
+      "loss": 0.4878,
+      "step": 1680
+    },
+    {
+      "epoch": 0.9295852534562212,
+      "grad_norm": 0.29310277104377747,
+      "learning_rate": 1.495872631165929e-07,
+      "loss": 0.476,
+      "step": 1681
+    },
+    {
+      "epoch": 0.9301382488479263,
+      "grad_norm": 0.3012104332447052,
+      "learning_rate": 1.4725241882799946e-07,
+      "loss": 0.4823,
+      "step": 1682
+    },
+    {
+      "epoch": 0.9306912442396313,
+      "grad_norm": 0.2896125018596649,
+      "learning_rate": 1.449356676070285e-07,
+      "loss": 0.4429,
+      "step": 1683
+    },
+    {
+      "epoch": 0.9312442396313364,
+      "grad_norm": 0.2954270839691162,
+      "learning_rate": 1.42637018091496e-07,
+      "loss": 0.473,
+      "step": 1684
+    },
+    {
+      "epoch": 0.9317972350230415,
+      "grad_norm": 0.29838648438453674,
+      "learning_rate": 1.4035647885172533e-07,
+      "loss": 0.4615,
+      "step": 1685
+    },
+    {
+      "epoch": 0.9323502304147465,
+      "grad_norm": 0.3438728451728821,
+      "learning_rate": 1.3809405839051937e-07,
+      "loss": 0.4924,
+      "step": 1686
+    },
+    {
+      "epoch": 0.9329032258064516,
+      "grad_norm": 0.2971719205379486,
+      "learning_rate": 1.358497651431251e-07,
+      "loss": 0.4438,
+      "step": 1687
+    },
+    {
+      "epoch": 0.9334562211981566,
+      "grad_norm": 0.3192031979560852,
+      "learning_rate": 1.3362360747720405e-07,
+      "loss": 0.4559,
+      "step": 1688
+    },
+    {
+      "epoch": 0.9340092165898618,
+      "grad_norm": 0.3306371867656708,
+      "learning_rate": 1.314155936928002e-07,
+      "loss": 0.4633,
+      "step": 1689
+    },
+    {
+      "epoch": 0.9345622119815669,
+      "grad_norm": 0.28587308526039124,
+      "learning_rate": 1.2922573202231114e-07,
+      "loss": 0.4577,
+      "step": 1690
+    },
+    {
+      "epoch": 0.9351152073732719,
+      "grad_norm": 0.31524068117141724,
+      "learning_rate": 1.2705403063045464e-07,
+      "loss": 0.4457,
+      "step": 1691
+    },
+    {
+      "epoch": 0.935668202764977,
+      "grad_norm": 0.2901270389556885,
+      "learning_rate": 1.2490049761423927e-07,
+      "loss": 0.4396,
+      "step": 1692
+    },
+    {
+      "epoch": 0.936221198156682,
+      "grad_norm": 0.2995949685573578,
+      "learning_rate": 1.2276514100293403e-07,
+      "loss": 0.4633,
+      "step": 1693
+    },
+    {
+      "epoch": 0.9367741935483871,
+      "grad_norm": 0.2743770182132721,
+      "learning_rate": 1.206479687580403e-07,
+      "loss": 0.4655,
+      "step": 1694
+    },
+    {
+      "epoch": 0.9373271889400921,
+      "grad_norm": 0.3132480978965759,
+      "learning_rate": 1.1854898877325882e-07,
+      "loss": 0.4641,
+      "step": 1695
+    },
+    {
+      "epoch": 0.9378801843317972,
+      "grad_norm": 0.2994924783706665,
+      "learning_rate": 1.1646820887446342e-07,
+      "loss": 0.4491,
+      "step": 1696
+    },
+    {
+      "epoch": 0.9384331797235023,
+      "grad_norm": 0.2822410762310028,
+      "learning_rate": 1.1440563681966943e-07,
+      "loss": 0.4663,
+      "step": 1697
+    },
+    {
+      "epoch": 0.9389861751152073,
+      "grad_norm": 0.3047800660133362,
+      "learning_rate": 1.1236128029900539e-07,
+      "loss": 0.4638,
+      "step": 1698
+    },
+    {
+      "epoch": 0.9395391705069125,
+      "grad_norm": 0.31529849767684937,
+      "learning_rate": 1.1033514693468584e-07,
+      "loss": 0.4269,
+      "step": 1699
+    },
+    {
+      "epoch": 0.9400921658986175,
+      "grad_norm": 0.28443610668182373,
+      "learning_rate": 1.0832724428098185e-07,
+      "loss": 0.4568,
+      "step": 1700
+    },
+    {
+      "epoch": 0.9406451612903226,
+      "grad_norm": 0.2651902735233307,
+      "learning_rate": 1.0633757982419169e-07,
+      "loss": 0.4508,
+      "step": 1701
+    },
+    {
+      "epoch": 0.9411981566820277,
+      "grad_norm": 0.29510411620140076,
+      "learning_rate": 1.0436616098261409e-07,
+      "loss": 0.4643,
+      "step": 1702
+    },
+    {
+      "epoch": 0.9417511520737327,
+      "grad_norm": 0.2974132299423218,
+      "learning_rate": 1.0241299510652114e-07,
+      "loss": 0.429,
+      "step": 1703
+    },
+    {
+      "epoch": 0.9423041474654378,
+      "grad_norm": 0.2929219901561737,
+      "learning_rate": 1.0047808947813097e-07,
+      "loss": 0.4851,
+      "step": 1704
+    },
+    {
+      "epoch": 0.9428571428571428,
+      "grad_norm": 0.29706960916519165,
+      "learning_rate": 9.85614513115779e-08,
+      "loss": 0.4559,
+      "step": 1705
+    },
+    {
+      "epoch": 0.9434101382488479,
+      "grad_norm": 0.298990935087204,
+      "learning_rate": 9.666308775289013e-08,
+      "loss": 0.479,
+      "step": 1706
+    },
+    {
+      "epoch": 0.943963133640553,
+      "grad_norm": 0.29148900508880615,
+      "learning_rate": 9.47830058799576e-08,
+      "loss": 0.4577,
+      "step": 1707
+    },
+    {
+      "epoch": 0.944516129032258,
+      "grad_norm": 0.3139622211456299,
+      "learning_rate": 9.292121270251031e-08,
+      "loss": 0.4569,
+      "step": 1708
+    },
+    {
+      "epoch": 0.9450691244239632,
+      "grad_norm": 0.29921236634254456,
+      "learning_rate": 9.107771516209008e-08,
+      "loss": 0.4558,
+      "step": 1709
+    },
+    {
+      "epoch": 0.9456221198156682,
+      "grad_norm": 0.28523844480514526,
+      "learning_rate": 8.925252013202545e-08,
+      "loss": 0.4709,
+      "step": 1710
+    },
+    {
+      "epoch": 0.9461751152073733,
+      "grad_norm": 0.3282380998134613,
+      "learning_rate": 8.744563441740461e-08,
+      "loss": 0.4485,
+      "step": 1711
+    },
+    {
+      "epoch": 0.9467281105990784,
+      "grad_norm": 0.30116501450538635,
+      "learning_rate": 8.565706475505086e-08,
+      "loss": 0.4633,
+      "step": 1712
+    },
+    {
+      "epoch": 0.9472811059907834,
+      "grad_norm": 0.2834503948688507,
+      "learning_rate": 8.388681781349828e-08,
+      "loss": 0.4569,
+      "step": 1713
+    },
+    {
+      "epoch": 0.9478341013824885,
+      "grad_norm": 0.31157195568084717,
+      "learning_rate": 8.213490019296666e-08,
+      "loss": 0.4666,
+      "step": 1714
+    },
+    {
+      "epoch": 0.9483870967741935,
+      "grad_norm": 0.30053144693374634,
+      "learning_rate": 8.04013184253355e-08,
+      "loss": 0.4645,
+      "step": 1715
+    },
+    {
+      "epoch": 0.9489400921658986,
+      "grad_norm": 0.3080219328403473,
+      "learning_rate": 7.868607897412062e-08,
+      "loss": 0.4409,
+      "step": 1716
+    },
+    {
+      "epoch": 0.9494930875576036,
+      "grad_norm": 0.2975987493991852,
+      "learning_rate": 7.69891882344509e-08,
+      "loss": 0.4691,
+      "step": 1717
+    },
+    {
+      "epoch": 0.9500460829493087,
+      "grad_norm": 0.3106543719768524,
+      "learning_rate": 7.53106525330427e-08,
+      "loss": 0.4816,
+      "step": 1718
+    },
+    {
+      "epoch": 0.9505990783410139,
+      "grad_norm": 0.29965740442276,
+      "learning_rate": 7.36504781281766e-08,
+      "loss": 0.469,
+      "step": 1719
+    },
+    {
+      "epoch": 0.9511520737327189,
+      "grad_norm": 0.31763914227485657,
+      "learning_rate": 7.20086712096768e-08,
+      "loss": 0.4615,
+      "step": 1720
+    },
+    {
+      "epoch": 0.951705069124424,
+      "grad_norm": 0.2858491837978363,
+      "learning_rate": 7.038523789888397e-08,
+      "loss": 0.457,
+      "step": 1721
+    },
+    {
+      "epoch": 0.952258064516129,
+      "grad_norm": 0.2787969410419464,
+      "learning_rate": 6.878018424863408e-08,
+      "loss": 0.4485,
+      "step": 1722
+    },
+    {
+      "epoch": 0.9528110599078341,
+      "grad_norm": 0.32252055406570435,
+      "learning_rate": 6.7193516243238e-08,
+      "loss": 0.4708,
+      "step": 1723
+    },
+    {
+      "epoch": 0.9533640552995392,
+      "grad_norm": 0.3171110451221466,
+      "learning_rate": 6.562523979845525e-08,
+      "loss": 0.466,
+      "step": 1724
+    },
+    {
+      "epoch": 0.9539170506912442,
+      "grad_norm": 0.31782266497612,
+      "learning_rate": 6.407536076147525e-08,
+      "loss": 0.4624,
+      "step": 1725
+    },
+    {
+      "epoch": 0.9544700460829493,
+      "grad_norm": 0.2952912449836731,
+      "learning_rate": 6.254388491089281e-08,
+      "loss": 0.4537,
+      "step": 1726
+    },
+    {
+      "epoch": 0.9550230414746543,
+      "grad_norm": 0.28047212958335876,
+      "learning_rate": 6.10308179566893e-08,
+      "loss": 0.473,
+      "step": 1727
+    },
+    {
+      "epoch": 0.9555760368663595,
+      "grad_norm": 0.30235886573791504,
+      "learning_rate": 5.9536165540209914e-08,
+      "loss": 0.489,
+      "step": 1728
+    },
+    {
+      "epoch": 0.9561290322580646,
+      "grad_norm": 0.29142141342163086,
+      "learning_rate": 5.80599332341425e-08,
+      "loss": 0.4718,
+      "step": 1729
+    },
+    {
+      "epoch": 0.9566820276497696,
+      "grad_norm": 0.2832047939300537,
+      "learning_rate": 5.6602126542496525e-08,
+      "loss": 0.4562,
+      "step": 1730
+    },
+    {
+      "epoch": 0.9572350230414747,
+      "grad_norm": 0.2992711365222931,
+      "learning_rate": 5.516275090058476e-08,
+      "loss": 0.4816,
+      "step": 1731
+    },
+    {
+      "epoch": 0.9577880184331797,
+      "grad_norm": 0.29784855246543884,
+      "learning_rate": 5.3741811674998254e-08,
+      "loss": 0.4818,
+      "step": 1732
+    },
+    {
+      "epoch": 0.9583410138248848,
+      "grad_norm": 0.3014211058616638,
+      "learning_rate": 5.233931416359195e-08,
+      "loss": 0.4914,
+      "step": 1733
+    },
+    {
+      "epoch": 0.9588940092165898,
+      "grad_norm": 0.30964842438697815,
+      "learning_rate": 5.095526359546243e-08,
+      "loss": 0.4643,
+      "step": 1734
+    },
+    {
+      "epoch": 0.9594470046082949,
+      "grad_norm": 0.3023122251033783,
+      "learning_rate": 4.958966513092689e-08,
+      "loss": 0.4499,
+      "step": 1735
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.295585572719574,
+      "learning_rate": 4.8242523861506405e-08,
+      "loss": 0.4248,
+      "step": 1736
+    },
+    {
+      "epoch": 0.960552995391705,
+      "grad_norm": 0.2833447754383087,
+      "learning_rate": 4.691384480990602e-08,
+      "loss": 0.4491,
+      "step": 1737
+    },
+    {
+      "epoch": 0.9611059907834102,
+      "grad_norm": 0.309023916721344,
+      "learning_rate": 4.56036329299947e-08,
+      "loss": 0.4404,
+      "step": 1738
+    },
+    {
+      "epoch": 0.9616589861751152,
+      "grad_norm": 0.3236503303050995,
+      "learning_rate": 4.4311893106789847e-08,
+      "loss": 0.4591,
+      "step": 1739
+    },
+    {
+      "epoch": 0.9622119815668203,
+      "grad_norm": 0.29482167959213257,
+      "learning_rate": 4.3038630156436166e-08,
+      "loss": 0.4558,
+      "step": 1740
+    },
+    {
+      "epoch": 0.9627649769585254,
+      "grad_norm": 0.2854726016521454,
+      "learning_rate": 4.1783848826189025e-08,
+      "loss": 0.4561,
+      "step": 1741
+    },
+    {
+      "epoch": 0.9633179723502304,
+      "grad_norm": 0.28575026988983154,
+      "learning_rate": 4.0547553794397235e-08,
+      "loss": 0.4541,
+      "step": 1742
+    },
+    {
+      "epoch": 0.9638709677419355,
+      "grad_norm": 0.30080512166023254,
+      "learning_rate": 3.932974967048364e-08,
+      "loss": 0.4475,
+      "step": 1743
+    },
+    {
+      "epoch": 0.9644239631336405,
+      "grad_norm": 0.2857537269592285,
+      "learning_rate": 3.813044099493068e-08,
+      "loss": 0.4521,
+      "step": 1744
+    },
+    {
+      "epoch": 0.9649769585253456,
+      "grad_norm": 0.28055858612060547,
+      "learning_rate": 3.6949632239261514e-08,
+      "loss": 0.4388,
+      "step": 1745
+    },
+    {
+      "epoch": 0.9655299539170507,
+      "grad_norm": 0.27776503562927246,
+      "learning_rate": 3.578732780602334e-08,
+      "loss": 0.4483,
+      "step": 1746
+    },
+    {
+      "epoch": 0.9660829493087557,
+      "grad_norm": 0.322313517332077,
+      "learning_rate": 3.464353202877302e-08,
+      "loss": 0.4589,
+      "step": 1747
+    },
+    {
+      "epoch": 0.9666359447004609,
+      "grad_norm": 0.315902978181839,
+      "learning_rate": 3.351824917205704e-08,
+      "loss": 0.4696,
+      "step": 1748
+    },
+    {
+      "epoch": 0.9671889400921659,
+      "grad_norm": 0.3116399645805359,
+      "learning_rate": 3.2411483431400435e-08,
+      "loss": 0.4532,
+      "step": 1749
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 0.27709704637527466,
+      "learning_rate": 3.1323238933286814e-08,
+      "loss": 0.4808,
+      "step": 1750
+    },
+    {
+      "epoch": 0.968294930875576,
+      "grad_norm": 0.28587040305137634,
+      "learning_rate": 3.0253519735146674e-08,
+      "loss": 0.4498,
+      "step": 1751
+    },
+    {
+      "epoch": 0.9688479262672811,
+      "grad_norm": 0.3236626088619232,
+      "learning_rate": 2.920232982533855e-08,
+      "loss": 0.4618,
+      "step": 1752
+    },
+    {
+      "epoch": 0.9694009216589862,
+      "grad_norm": 0.29707276821136475,
+      "learning_rate": 2.8169673123137896e-08,
+      "loss": 0.4516,
+      "step": 1753
+    },
+    {
+      "epoch": 0.9699539170506912,
+      "grad_norm": 0.3110201954841614,
+      "learning_rate": 2.7155553478719343e-08,
+      "loss": 0.4534,
+      "step": 1754
+    },
+    {
+      "epoch": 0.9705069124423963,
+      "grad_norm": 0.308906227350235,
+      "learning_rate": 2.6159974673143352e-08,
+      "loss": 0.447,
+      "step": 1755
+    },
+    {
+      "epoch": 0.9710599078341013,
+      "grad_norm": 0.30986642837524414,
+      "learning_rate": 2.5182940418344016e-08,
+      "loss": 0.4587,
+      "step": 1756
+    },
+    {
+      "epoch": 0.9716129032258064,
+      "grad_norm": 0.31161195039749146,
+      "learning_rate": 2.4224454357112404e-08,
+      "loss": 0.4616,
+      "step": 1757
+    },
+    {
+      "epoch": 0.9721658986175116,
+      "grad_norm": 0.2983948290348053,
+      "learning_rate": 2.328452006308435e-08,
+      "loss": 0.4696,
+      "step": 1758
+    },
+    {
+      "epoch": 0.9727188940092166,
+      "grad_norm": 0.28888341784477234,
+      "learning_rate": 2.2363141040727123e-08,
+      "loss": 0.4859,
+      "step": 1759
+    },
+    {
+      "epoch": 0.9732718894009217,
+      "grad_norm": 0.2775828242301941,
+      "learning_rate": 2.1460320725326113e-08,
+      "loss": 0.4354,
+      "step": 1760
+    },
+    {
+      "epoch": 0.9738248847926267,
+      "grad_norm": 0.34634268283843994,
+      "learning_rate": 2.057606248297206e-08,
+      "loss": 0.4431,
+      "step": 1761
+    },
+    {
+      "epoch": 0.9743778801843318,
+      "grad_norm": 0.31715697050094604,
+      "learning_rate": 1.971036961054884e-08,
+      "loss": 0.4566,
+      "step": 1762
+    },
+    {
+      "epoch": 0.9749308755760369,
+      "grad_norm": 0.2788535952568054,
+      "learning_rate": 1.886324533572015e-08,
+      "loss": 0.4492,
+      "step": 1763
+    },
+    {
+      "epoch": 0.9754838709677419,
+      "grad_norm": 0.3046756684780121,
+      "learning_rate": 1.8034692816919497e-08,
+      "loss": 0.4461,
+      "step": 1764
+    },
+    {
+      "epoch": 0.976036866359447,
+      "grad_norm": 0.2936743497848511,
+      "learning_rate": 1.7224715143335235e-08,
+      "loss": 0.4635,
+      "step": 1765
+    },
+    {
+      "epoch": 0.976589861751152,
+      "grad_norm": 0.31344956159591675,
+      "learning_rate": 1.643331533490333e-08,
+      "loss": 0.434,
+      "step": 1766
+    },
+    {
+      "epoch": 0.9771428571428571,
+      "grad_norm": 0.2932872176170349,
+      "learning_rate": 1.5660496342291833e-08,
+      "loss": 0.4707,
+      "step": 1767
+    },
+    {
+      "epoch": 0.9776958525345623,
+      "grad_norm": 0.29183509945869446,
+      "learning_rate": 1.4906261046892523e-08,
+      "loss": 0.4711,
+      "step": 1768
+    },
+    {
+      "epoch": 0.9782488479262673,
+      "grad_norm": 0.29039350152015686,
+      "learning_rate": 1.4170612260808736e-08,
+      "loss": 0.4802,
+      "step": 1769
+    },
+    {
+      "epoch": 0.9788018433179724,
+      "grad_norm": 0.30380529165267944,
+      "learning_rate": 1.3453552726847008e-08,
+      "loss": 0.4868,
+      "step": 1770
+    },
+    {
+      "epoch": 0.9793548387096774,
+      "grad_norm": 0.29868564009666443,
+      "learning_rate": 1.2755085118503762e-08,
+      "loss": 0.4625,
+      "step": 1771
+    },
+    {
+      "epoch": 0.9799078341013825,
+      "grad_norm": 0.30664679408073425,
+      "learning_rate": 1.207521203995754e-08,
+      "loss": 0.4369,
+      "step": 1772
+    },
+    {
+      "epoch": 0.9804608294930875,
+      "grad_norm": 0.28604456782341003,
+      "learning_rate": 1.1413936026059558e-08,
+      "loss": 0.4559,
+      "step": 1773
+    },
+    {
+      "epoch": 0.9810138248847926,
+      "grad_norm": 0.33186855912208557,
+      "learning_rate": 1.0771259542322055e-08,
+      "loss": 0.4735,
+      "step": 1774
+    },
+    {
+      "epoch": 0.9815668202764977,
+      "grad_norm": 0.3165326714515686,
+      "learning_rate": 1.0147184984911073e-08,
+      "loss": 0.4411,
+      "step": 1775
+    },
+    {
+      "epoch": 0.9821198156682027,
+      "grad_norm": 0.30226629972457886,
+      "learning_rate": 9.541714680637582e-09,
+      "loss": 0.4274,
+      "step": 1776
+    },
+    {
+      "epoch": 0.9826728110599078,
+      "grad_norm": 0.3025604486465454,
+      "learning_rate": 8.95485088694692e-09,
+      "loss": 0.4461,
+      "step": 1777
+    },
+    {
+      "epoch": 0.983225806451613,
+      "grad_norm": 0.3139297068119049,
+      "learning_rate": 8.386595791912145e-09,
+      "loss": 0.4501,
+      "step": 1778
+    },
+    {
+      "epoch": 0.983778801843318,
+      "grad_norm": 0.28405094146728516,
+      "learning_rate": 7.836951514225144e-09,
+      "loss": 0.4686,
+      "step": 1779
+    },
+    {
+      "epoch": 0.9843317972350231,
+      "grad_norm": 0.2899510860443115,
+      "learning_rate": 7.3059201031899786e-09,
+      "loss": 0.4738,
+      "step": 1780
+    },
+    {
+      "epoch": 0.9848847926267281,
+      "grad_norm": 0.3046450614929199,
+      "learning_rate": 6.7935035387128865e-09,
+      "loss": 0.4531,
+      "step": 1781
+    },
+    {
+      "epoch": 0.9854377880184332,
+      "grad_norm": 0.2701072692871094,
+      "learning_rate": 6.299703731296181e-09,
+      "loss": 0.4532,
+      "step": 1782
+    },
+    {
+      "epoch": 0.9859907834101382,
+      "grad_norm": 0.271697461605072,
+      "learning_rate": 5.8245225220321385e-09,
+      "loss": 0.4349,
+      "step": 1783
+    },
+    {
+      "epoch": 0.9865437788018433,
+      "grad_norm": 0.29448968172073364,
+      "learning_rate": 5.367961682594125e-09,
+      "loss": 0.4664,
+      "step": 1784
+    },
+    {
+      "epoch": 0.9870967741935484,
+      "grad_norm": 0.3071184754371643,
+      "learning_rate": 4.930022915231591e-09,
+      "loss": 0.4397,
+      "step": 1785
+    },
+    {
+      "epoch": 0.9876497695852534,
+      "grad_norm": 0.2995312213897705,
+      "learning_rate": 4.510707852762864e-09,
+      "loss": 0.4592,
+      "step": 1786
+    },
+    {
+      "epoch": 0.9882027649769585,
+      "grad_norm": 0.30661311745643616,
+      "learning_rate": 4.110018058570142e-09,
+      "loss": 0.4835,
+      "step": 1787
+    },
+    {
+      "epoch": 0.9887557603686636,
+      "grad_norm": 0.2768559157848358,
+      "learning_rate": 3.727955026591179e-09,
+      "loss": 0.4506,
+      "step": 1788
+    },
+    {
+      "epoch": 0.9893087557603687,
+      "grad_norm": 0.3139650821685791,
+      "learning_rate": 3.3645201813170546e-09,
+      "loss": 0.4483,
+      "step": 1789
+    },
+    {
+      "epoch": 0.9898617511520738,
+      "grad_norm": 0.318562775850296,
+      "learning_rate": 3.0197148777838524e-09,
+      "loss": 0.4669,
+      "step": 1790
+    },
+    {
+      "epoch": 0.9904147465437788,
+      "grad_norm": 0.31068408489227295,
+      "learning_rate": 2.693540401569883e-09,
+      "loss": 0.4519,
+      "step": 1791
+    },
+    {
+      "epoch": 0.9909677419354839,
+      "grad_norm": 0.30719509720802307,
+      "learning_rate": 2.3859979687901326e-09,
+      "loss": 0.4284,
+      "step": 1792
+    },
+    {
+      "epoch": 0.9915207373271889,
+      "grad_norm": 0.28648641705513,
+      "learning_rate": 2.0970887260907123e-09,
+      "loss": 0.4556,
+      "step": 1793
+    },
+    {
+      "epoch": 0.992073732718894,
+      "grad_norm": 0.2835429608821869,
+      "learning_rate": 1.8268137506455286e-09,
+      "loss": 0.4764,
+      "step": 1794
+    },
+    {
+      "epoch": 0.992626728110599,
+      "grad_norm": 0.3018244206905365,
+      "learning_rate": 1.57517405015295e-09,
+      "loss": 0.4505,
+      "step": 1795
+    },
+    {
+      "epoch": 0.9931797235023041,
+      "grad_norm": 0.30132514238357544,
+      "learning_rate": 1.3421705628302584e-09,
+      "loss": 0.4686,
+      "step": 1796
+    },
+    {
+      "epoch": 0.9937327188940093,
+      "grad_norm": 0.292498379945755,
+      "learning_rate": 1.1278041574125376e-09,
+      "loss": 0.4295,
+      "step": 1797
+    },
+    {
+      "epoch": 0.9942857142857143,
+      "grad_norm": 0.2868688702583313,
+      "learning_rate": 9.320756331465675e-10,
+      "loss": 0.4246,
+      "step": 1798
+    },
+    {
+      "epoch": 0.9948387096774194,
+      "grad_norm": 0.2882429361343384,
+      "learning_rate": 7.549857197897137e-10,
+      "loss": 0.4539,
+      "step": 1799
+    },
+    {
+      "epoch": 0.9953917050691244,
+      "grad_norm": 0.2899753749370575,
+      "learning_rate": 5.965350776071521e-10,
+      "loss": 0.4748,
+      "step": 1800
+    },
+    {
+      "epoch": 0.9959447004608295,
+      "grad_norm": 0.293375700712204,
+      "learning_rate": 4.567242973696484e-10,
+      "loss": 0.4613,
+      "step": 1801
+    },
+    {
+      "epoch": 0.9964976958525346,
+      "grad_norm": 0.2795793116092682,
+      "learning_rate": 3.3555390034967214e-10,
+      "loss": 0.4732,
+      "step": 1802
+    },
+    {
+      "epoch": 0.9970506912442396,
+      "grad_norm": 0.29951760172843933,
+      "learning_rate": 2.330243383208419e-10,
+      "loss": 0.4612,
+      "step": 1803
+    },
+    {
+      "epoch": 0.9976036866359447,
+      "grad_norm": 0.2993388772010803,
+      "learning_rate": 1.4913599355625975e-10,
+      "loss": 0.4653,
+      "step": 1804
+    },
+    {
+      "epoch": 0.9981566820276497,
+      "grad_norm": 0.31436586380004883,
+      "learning_rate": 8.388917882684589e-11,
+      "loss": 0.4712,
+      "step": 1805
+    },
+    {
+      "epoch": 0.9987096774193548,
+      "grad_norm": 0.2935016453266144,
+      "learning_rate": 3.728413739967351e-11,
+      "loss": 0.4381,
+      "step": 1806
+    },
+    {
+      "epoch": 0.99926267281106,
+      "grad_norm": 0.29614418745040894,
+      "learning_rate": 9.321043037968657e-12,
+      "loss": 0.4752,
+      "step": 1807
+    },
+    {
+      "epoch": 0.999815668202765,
+      "grad_norm": 0.294443815946579,
+      "learning_rate": 0.0,
+      "loss": 0.4617,
+      "step": 1808
+    },
+    {
+      "epoch": 0.999815668202765,
+      "step": 1808,
+      "total_flos": 2730506599792640.0,
+      "train_loss": 0.488344039471276,
+      "train_runtime": 63700.3598,
+      "train_samples_per_second": 2.725,
+      "train_steps_per_second": 0.028
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 1808,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2730506599792640.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}